github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-22 21:51:37 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #13433] Model request: AutoGLM-Phone-9B #55380

New Issue
Open
opened 2026-04-29 09:04:47 -05:00 by GiteaMirror · 11 comments
GiteaMirror commented 2026-04-29 09:04:47 -05:00
Owner
Copy Link

Originally created by @pinghe on GitHub (Dec 12, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/13433

https://github.com/zai-org/Open-AutoGLM?tab=readme-ov-file

https://huggingface.co/zai-org/AutoGLM-Phone-9B
https://huggingface.co/zai-org/AutoGLM-Phone-9B-Multilingual

Originally created by @pinghe on GitHub (Dec 12, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/13433 https://github.com/zai-org/Open-AutoGLM?tab=readme-ov-file https://huggingface.co/zai-org/AutoGLM-Phone-9B https://huggingface.co/zai-org/AutoGLM-Phone-9B-Multilingual
GiteaMirror added the model label 2026-04-29 09:04:47 -05:00
GiteaMirror commented 2026-04-29 09:04:48 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Dec 13, 2025):

This model is glm4v architecture so will be supported when https://github.com/ggml-org/llama.cpp/pull/16600 is merged and vendor synced.

<!-- gh-comment-id:3648672982 --> @rick-github commented on GitHub (Dec 13, 2025): This model is `glm4v` architecture so will be supported when https://github.com/ggml-org/llama.cpp/pull/16600 is merged and vendor synced.
GiteaMirror commented 2026-04-29 09:04:48 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Jan 1, 2026):

$ ollama run hf.co/ggml-org/AutoGLM-Phone-9B-GGUF:Q4_K_M describe this image ./image1.jpg 
Added image './image1.jpg'
I can see a small, white, fluffy dog in the image. The dog appears to be a 
young, small breed with long, soft fur. It's wearing a red collar around 
its neck and has a small bell or tag attached to it. The background is a 
concrete or stone surface, suggesting an outdoor setting.

The image shows the dog from a slightly side profile view, looking to the 
right. The lighting is even, with a shallow depth of field that keeps most 
of the focus on the dog itself.
<!-- gh-comment-id:3703309417 --> @rick-github commented on GitHub (Jan 1, 2026): ```console $ ollama run hf.co/ggml-org/AutoGLM-Phone-9B-GGUF:Q4_K_M describe this image ./image1.jpg Added image './image1.jpg' I can see a small, white, fluffy dog in the image. The dog appears to be a young, small breed with long, soft fur. It's wearing a red collar around its neck and has a small bell or tag attached to it. The background is a concrete or stone surface, suggesting an outdoor setting. The image shows the dog from a slightly side profile view, looking to the right. The lighting is even, with a shallow depth of field that keeps most of the focus on the dog itself. ```
GiteaMirror commented 2026-04-29 09:04:49 -05:00
Author
Owner
Copy Link

@pinghe commented on GitHub (Jan 5, 2026):

05 12:12:16 arch ollama[927]: warmup: warmup with image size = 1288 x 1288
05 12:12:16 arch ollama[927]: ggml_backend_cuda_buffer_type_alloc_buffer: allocating 515.05 MiB on device 0: cudaMalloc led: out of memory
05 12:12:16 arch ollama[927]: ggml_gallocr_reserve_n_impl: failed to allocate CUDA0 buffer of size 540070912
05 12:12:16 arch ollama[927]: alloc_compute_meta: CPU compute buffer size = 19.11 MiB
05 12:12:16 arch ollama[927]: alloc_compute_meta: graph splits = 1, nodes = 632
05 12:12:16 arch ollama[927]: warmup: flash attention is enabled
05 12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.324+08:00 level=INFO source=server.go:1376 msg="llama runner rted in 2.61 seconds"
05 12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.324+08:00 level=INFO source=sched.go:517 msg="loaded runners" nt=1
05 12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.324+08:00 level=INFO source=server.go:1338 msg="waiting for ma runner to start responding"
05 12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.325+08:00 level=INFO source=server.go:1376 msg="llama runner rted in 2.61 seconds"
05 12:12:16 arch ollama[927]: add_text: <|begin_of_image|>
05 12:12:16 arch ollama[927]: image_tokens->nx = 85
05 12:12:16 arch ollama[927]: image_tokens->ny = 48
05 12:12:16 arch ollama[927]: batch_f32 size = 1
05 12:12:16 arch ollama[927]: add_text: <|end_of_image|>
05 12:12:16 arch ollama[927]: ggml_backend_cuda_buffer_type_alloc_buffer: allocating 993.11 MiB on device 0: cudaMalloc led: out of memory
05 12:12:16 arch ollama[927]: ggml_gallocr_reserve_n_impl: failed to allocate CUDA0 buffer of size 1041346560
05 12:12:16 arch ollama[927]: SIGSEGV: segmentation violation
05 12:12:16 arch ollama[927]: PC=0x564e038541cb m=15 sigcode=1 addr=0x0
05 12:12:16 arch ollama[927]: signal arrived during cgo execution
05 12:12:16 arch ollama[927]: goroutine 40 gp=0xc000505c00 m=15 mp=0xc000101008 [syscall]:
05 12:12:16 arch ollama[927]: runtime.cgocall(0x564e03841190, 0xc0000491d8)
05 12:12:16 arch ollama[927]: runtime/cgocall.go:167 +0x4b fp=0xc0000491b0 sp=0xc000049178 pc=0x564e02aef6eb
05 12:12:16 arch ollama[927]: github.com/ollama/ollama/llama._Cfunc_mtmd_encode_chunk(0x7f834008dd90, 0x7f82817bc000)
05 12:12:16 arch ollama[927]: _cgo_gotypes.go:1079 +0x4a fp=0xc0000491d8 sp=0xc0000491b0 pc=0x564e02eaa04a
05 12:12:16 arch ollama[927]: github.com/ollama/ollama/llama.(*MtmdContext).MultimodalTokenize.func11(...)
05 12:12:16 arch ollama[927]: github.com/ollama/ollama/llama/llama.go:595
05 12:12:16 arch ollama[927]: github.com/ollama/ollama/llama.(*MtmdContext).MultimodalTokenize(0xc00034c0b8, 0003a6750, {0xc000a82000, 0x117bcb, 0
00049520?})

@rick-github

<!-- gh-comment-id:3708883168 --> @pinghe commented on GitHub (Jan 5, 2026): 05 12:12:16 arch ollama[927]: warmup: warmup with image size = 1288 x 1288 05 12:12:16 arch ollama[927]: ggml_backend_cuda_buffer_type_alloc_buffer: allocating 515.05 MiB on device 0: cudaMalloc led: out of memory 05 12:12:16 arch ollama[927]: ggml_gallocr_reserve_n_impl: failed to allocate CUDA0 buffer of size 540070912 05 12:12:16 arch ollama[927]: alloc_compute_meta: CPU compute buffer size = 19.11 MiB 05 12:12:16 arch ollama[927]: alloc_compute_meta: graph splits = 1, nodes = 632 05 12:12:16 arch ollama[927]: warmup: flash attention is enabled 05 12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.324+08:00 level=INFO source=server.go:1376 msg="llama runner rted in 2.61 seconds" 05 12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.324+08:00 level=INFO source=sched.go:517 msg="loaded runners" nt=1 05 12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.324+08:00 level=INFO source=server.go:1338 msg="waiting for ma runner to start responding" 05 12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.325+08:00 level=INFO source=server.go:1376 msg="llama runner rted in 2.61 seconds" 05 12:12:16 arch ollama[927]: add_text: <|begin_of_image|> 05 12:12:16 arch ollama[927]: image_tokens->nx = 85 05 12:12:16 arch ollama[927]: image_tokens->ny = 48 05 12:12:16 arch ollama[927]: batch_f32 size = 1 05 12:12:16 arch ollama[927]: add_text: <|end_of_image|> 05 12:12:16 arch ollama[927]: ggml_backend_cuda_buffer_type_alloc_buffer: allocating 993.11 MiB on device 0: cudaMalloc led: out of memory 05 12:12:16 arch ollama[927]: ggml_gallocr_reserve_n_impl: failed to allocate CUDA0 buffer of size 1041346560 05 12:12:16 arch ollama[927]: SIGSEGV: segmentation violation 05 12:12:16 arch ollama[927]: PC=0x564e038541cb m=15 sigcode=1 addr=0x0 05 12:12:16 arch ollama[927]: signal arrived during cgo execution 05 12:12:16 arch ollama[927]: goroutine 40 gp=0xc000505c00 m=15 mp=0xc000101008 [syscall]: 05 12:12:16 arch ollama[927]: runtime.cgocall(0x564e03841190, 0xc0000491d8) 05 12:12:16 arch ollama[927]: runtime/cgocall.go:167 +0x4b fp=0xc0000491b0 sp=0xc000049178 pc=0x564e02aef6eb 05 12:12:16 arch ollama[927]: github.com/ollama/ollama/llama._Cfunc_mtmd_encode_chunk(0x7f834008dd90, 0x7f82817bc000) 05 12:12:16 arch ollama[927]: _cgo_gotypes.go:1079 +0x4a fp=0xc0000491d8 sp=0xc0000491b0 pc=0x564e02eaa04a 05 12:12:16 arch ollama[927]: github.com/ollama/ollama/llama.(*MtmdContext).MultimodalTokenize.func11(...) 05 12:12:16 arch ollama[927]: github.com/ollama/ollama/llama/llama.go:595 05 12:12:16 arch ollama[927]: github.com/ollama/ollama/llama.(*MtmdContext).MultimodalTokenize(0xc00034c0b8, 0003a6750, {0xc000a82000, 0x117bcb, 0 00049520?}) @rick-github
GiteaMirror commented 2026-04-29 09:04:49 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Jan 5, 2026):

Full log.

<!-- gh-comment-id:3708910044 --> @rick-github commented on GitHub (Jan 5, 2026): Full log.
GiteaMirror commented 2026-04-29 09:04:50 -05:00
Author
Owner
Copy Link

@pinghe commented on GitHub (Jan 5, 2026):

12:12:12 arch ollama[927]: time=2026-01-05T12:12:12.828+08:00 level=INFO source=runner.go:464 msg="failure during GPU discovery" OLLAMA_LIBRARY_PATH="[/usr/local/lib/ollama /usr/local/lib/ollama/cuda_v13]" extra_envs=map[] error="failed to finish discovery before timeout"
12:12:12 arch ollama[927]: time=2026-01-05T12:12:12.828+08:00 level=WARN source=runner.go:356 msg="unable to refresh free memory, using old values"
12:12:12 arch ollama[927]: time=2026-01-05T12:12:12.829+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 33053"
12:12:13 arch ollama[927]: llama_model_loader: loaded meta data with 33 key-value pairs and 523 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20af820432e1dd55d22a2b0e728e79223e77a (version GGUF V3 (latest))
12:12:13 arch ollama[927]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
12:12:13 arch ollama[927]: llama_model_loader: - kv 0: general.architecture str = glm4
12:12:13 arch ollama[927]: llama_model_loader: - kv 1: general.type str = model
12:12:13 arch ollama[927]: llama_model_loader: - kv 2: general.size_label str = 9.4B
12:12:13 arch ollama[927]: llama_model_loader: - kv 3: general.license str = mit
12:12:13 arch ollama[927]: llama_model_loader: - kv 4: general.base_model.count u32 = 1
12:12:13 arch ollama[927]: llama_model_loader: - kv 5: general.base_model.0.name str = GLM 4.1V 9B Base
12:12:13 arch ollama[927]: llama_model_loader: - kv 6: general.base_model.0.organization str = Zai Org
12:12:13 arch ollama[927]: llama_model_loader: - kv 7: general.base_model.0.repo_url str = https://huggingface.co/zai-org/GLM-4....
12:12:13 arch ollama[927]: llama_model_loader: - kv 8: general.tags arr[str,2] = ["agent", "image-text-to-text"]
12:12:13 arch ollama[927]: llama_model_loader: - kv 9: general.languages arr[str,1] = ["zh"]
12:12:13 arch ollama[927]: llama_model_loader: - kv 10: glm4.block_count u32 = 40
12:12:13 arch ollama[927]: llama_model_loader: - kv 11: glm4.context_length u32 = 65536
12:12:13 arch ollama[927]: llama_model_loader: - kv 12: glm4.embedding_length u32 = 4096
12:12:13 arch ollama[927]: llama_model_loader: - kv 13: glm4.feed_forward_length u32 = 13696
12:12:13 arch ollama[927]: llama_model_loader: - kv 14: glm4.attention.head_count u32 = 32
12:12:13 arch ollama[927]: llama_model_loader: - kv 15: glm4.attention.head_count_kv u32 = 2
12:12:13 arch ollama[927]: llama_model_loader: - kv 16: glm4.rope.dimension_sections arr[i32,4] = [8, 12, 12, 0]
12:12:13 arch ollama[927]: llama_model_loader: - kv 17: glm4.rope.freq_base f32 = 10000.000000
12:12:13 arch ollama[927]: llama_model_loader: - kv 18: glm4.attention.layer_norm_rms_epsilon f32 = 0.000010
12:12:13 arch ollama[927]: llama_model_loader: - kv 19: glm4.rope.dimension_count u32 = 64
12:12:13 arch ollama[927]: llama_model_loader: - kv 20: tokenizer.ggml.model str = gpt2
12:12:13 arch ollama[927]: llama_model_loader: - kv 21: tokenizer.ggml.pre str = glm4
12:12:13 arch ollama[927]: llama_model_loader: - kv 22: tokenizer.ggml.tokens arr[str,151552] = ["!", """, "#", "$", "%", "&", "'", ...
12:12:13 arch ollama[927]: llama_model_loader: - kv 23: tokenizer.ggml.token_type arr[i32,151552] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
12:12:13 arch ollama[927]: llama_model_loader: - kv 24: tokenizer.ggml.merges arr[str,318088] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "...
12:12:13 arch ollama[927]: llama_model_loader: - kv 25: tokenizer.ggml.eos_token_id u32 = 151329
12:12:13 arch ollama[927]: llama_model_loader: - kv 26: tokenizer.ggml.padding_token_id u32 = 151329
12:12:13 arch ollama[927]: llama_model_loader: - kv 27: tokenizer.ggml.eot_token_id u32 = 151336
12:12:13 arch ollama[927]: llama_model_loader: - kv 28: tokenizer.ggml.unknown_token_id u32 = 151329
12:12:13 arch ollama[927]: llama_model_loader: - kv 29: tokenizer.ggml.bos_token_id u32 = 151329
12:12:13 arch ollama[927]: llama_model_loader: - kv 30: tokenizer.chat_template str = [gMASK]\n{%- for msg in messages ...
12:12:13 arch ollama[927]: llama_model_loader: - kv 31: general.quantization_version u32 = 2
12:12:13 arch ollama[927]: llama_model_loader: - kv 32: general.file_type u32 = 15
12:12:13 arch ollama[927]: llama_model_loader: - type f32: 281 tensors
12:12:13 arch ollama[927]: llama_model_loader: - type q5_0: 20 tensors
12:12:13 arch ollama[927]: llama_model_loader: - type q8_0: 20 tensors
12:12:13 arch ollama[927]: llama_model_loader: - type q4_K: 181 tensors
12:12:13 arch ollama[927]: llama_model_loader: - type q6_K: 21 tensors
12:12:13 arch ollama[927]: print_info: file format = GGUF V3 (latest)
12:12:13 arch ollama[927]: print_info: file type = Q4_K - Medium
12:12:13 arch ollama[927]: print_info: file size = 5.73 GiB (5.24 BPW)
12:12:13 arch ollama[927]: load: special_eot_id is not in special_eog_ids - the tokenizer config may be incorrect
12:12:13 arch ollama[927]: load: printing all EOG tokens:
12:12:13 arch ollama[927]: load: - 151329 ('<|endoftext|>')
12:12:13 arch ollama[927]: load: - 151336 ('<|user|>')
12:12:13 arch ollama[927]: load: special tokens cache size = 23
12:12:13 arch ollama[927]: load: token to piece cache size = 0.9711 MB
12:12:13 arch ollama[927]: print_info: arch = glm4
12:12:13 arch ollama[927]: print_info: vocab_only = 1
12:12:13 arch ollama[927]: print_info: no_alloc = 0
12:12:13 arch ollama[927]: print_info: model type = ?B
12:12:13 arch ollama[927]: print_info: model params = 9.40 B
12:12:13 arch ollama[927]: print_info: general.name = n/a
12:12:13 arch ollama[927]: print_info: vocab type = BPE
12:12:13 arch ollama[927]: print_info: n_vocab = 151552
12:12:13 arch ollama[927]: print_info: n_merges = 318088
12:12:13 arch ollama[927]: print_info: BOS token = 151329 '<|endoftext|>'
12:12:13 arch ollama[927]: print_info: EOS token = 151329 '<|endoftext|>'
12:12:13 arch ollama[927]: print_info: EOT token = 151336 '<|user|>'
12:12:13 arch ollama[927]: print_info: UNK token = 151329 '<|endoftext|>'
12:12:13 arch ollama[927]: print_info: PAD token = 151329 '<|endoftext|>'
12:12:13 arch ollama[927]: print_info: LF token = 198 'Ċ'
12:12:13 arch ollama[927]: print_info: EOG token = 151329 '<|endoftext|>'
12:12:13 arch ollama[927]: print_info: EOG token = 151336 '<|user|>'
12:12:13 arch ollama[927]: print_info: max token length = 1024
12:12:13 arch ollama[927]: llama_model_load: vocab only - skipping tensors
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.712+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --model /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20af820432e1dd55d22a2b0e728e79223e77a --port 33773"
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.712+08:00 level=INFO source=sched.go:443 msg="system memory" total="62.6 GiB" free="41.9 GiB" free_swap="3.2 GiB"
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.712+08:00 level=INFO source=sched.go:450 msg="gpu memory" id=GPU-8215c551-6dde-569b-490d-884f3ab7a437 library=CUDA available="6.6 GiB" free="7.1 GiB" minimum="457.0 MiB" overhead="0 B"
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.712+08:00 level=INFO source=server.go:496 msg="loading model" "model layers"=41 requested=-1
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.713+08:00 level=INFO source=device.go:240 msg="model weights" device=CUDA0 size="4.2 GiB"
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.713+08:00 level=INFO source=device.go:245 msg="model weights" device=CPU size="1.3 GiB"
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.713+08:00 level=INFO source=device.go:251 msg="kv cache" device=CUDA0 size="544.0 MiB"
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.713+08:00 level=INFO source=device.go:256 msg="kv cache" device=CPU size="96.0 MiB"
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.713+08:00 level=INFO source=device.go:262 msg="compute graph" device=CUDA0 size="1.7 GiB"
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.713+08:00 level=INFO source=device.go:272 msg="total memory" size="7.7 GiB"
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.728+08:00 level=INFO source=runner.go:965 msg="starting go runner"
12:12:13 arch ollama[927]: load_backend: loaded CPU backend from /usr/local/lib/ollama/libggml-cpu-haswell.so
12:12:13 arch ollama[927]: ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
12:12:13 arch ollama[927]: ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
12:12:13 arch ollama[927]: ggml_cuda_init: found 1 CUDA devices:
12:12:13 arch ollama[927]: Device 0: NVIDIA GeForce RTX 2060 SUPER, compute capability 7.5, VMM: yes, ID: GPU-8215c551-6dde-569b-490d-884f3ab7a437
12:12:13 arch ollama[927]: load_backend: loaded CUDA backend from /usr/local/lib/ollama/cuda_v13/libggml-cuda.so
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.805+08:00 level=INFO source=ggml.go:104 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.BMI2=1 CPU.0.LLAMAFILE=1 CPU.1.LLAMAFILE=1 CUDA.0.ARCHS=750,800,860,870,890,900,1000,1030,1100,1200,1210 CUDA.0.USE_GRAPHS=1 CUDA.0.PEER_MAX_BATCH_SIZE=128 compiler=cgo(gcc)
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.805+08:00 level=INFO source=runner.go:1001 msg="Server listening on 127.0.0.1:33773"
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.811+08:00 level=INFO source=runner.go:895 msg=load request="{Operation:commit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:Auto KvSize:16384 KvCacheType: NumThreads:12 GPULayers:34[ID:GPU-8215c551-6dde-569b-490d-884f3ab7a437 Layers:34(6..39)] MultiUserCache:false ProjectorPath:/usr/share/ollama/.ollama/models/blobs/sha256-454dd441c925c1a81c984204bb9d54feef0ef07b789c6fe1118099014ba2727d MainGPU:0 UseMmap:true}"
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.811+08:00 level=INFO source=server.go:1338 msg="waiting for llama runner to start responding"
12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.811+08:00 level=INFO source=server.go:1372 msg="waiting for server to become available" status="llm server loading model"
12:12:13 arch ollama[927]: ggml_backend_cuda_device_get_memory device GPU-8215c551-6dde-569b-490d-884f3ab7a437 utilizing NVML memory reporting free: 7585398784 total: 8589934592
12:12:13 arch ollama[927]: llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce RTX 2060 SUPER) (0000:03:00.0) - 7234 MiB free
12:12:13 arch ollama[927]: llama_model_loader: loaded meta data with 33 key-value pairs and 523 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20af820432e1dd55d22a2b0e728e79223e77a (version GGUF V3 (latest))
12:12:13 arch ollama[927]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
12:12:13 arch ollama[927]: llama_model_loader: - kv 0: general.architecture str = glm4
12:12:13 arch ollama[927]: llama_model_loader: - kv 1: general.type str = model
12:12:13 arch ollama[927]: llama_model_loader: - kv 2: general.size_label str = 9.4B
12:12:13 arch ollama[927]: llama_model_loader: - kv 3: general.license str = mit
12:12:13 arch ollama[927]: llama_model_loader: - kv 4: general.base_model.count u32 = 1
12:12:13 arch ollama[927]: llama_model_loader: - kv 5: general.base_model.0.name str = GLM 4.1V 9B Base
12:12:13 arch ollama[927]: llama_model_loader: - kv 6: general.base_model.0.organization str = Zai Org
12:12:13 arch ollama[927]: llama_model_loader: - kv 7: general.base_model.0.repo_url str = https://huggingface.co/zai-org/GLM-4....
12:12:13 arch ollama[927]: llama_model_loader: - kv 8: general.tags arr[str,2] = ["agent", "image-text-to-text"]
12:12:13 arch ollama[927]: llama_model_loader: - kv 9: general.languages arr[str,1] = ["zh"]
12:12:13 arch ollama[927]: llama_model_loader: - kv 10: glm4.block_count u32 = 40
12:12:13 arch ollama[927]: llama_model_loader: - kv 11: glm4.context_length u32 = 65536
12:12:13 arch ollama[927]: llama_model_loader: - kv 12: glm4.embedding_length u32 = 4096
12:12:13 arch ollama[927]: llama_model_loader: - kv 13: glm4.feed_forward_length u32 = 13696
12:12:13 arch ollama[927]: llama_model_loader: - kv 14: glm4.attention.head_count u32 = 32
12:12:13 arch ollama[927]: llama_model_loader: - kv 15: glm4.attention.head_count_kv u32 = 2
12:12:13 arch ollama[927]: llama_model_loader: - kv 16: glm4.rope.dimension_sections arr[i32,4] = [8, 12, 12, 0]
12:12:13 arch ollama[927]: llama_model_loader: - kv 17: glm4.rope.freq_base f32 = 10000.000000
12:12:13 arch ollama[927]: llama_model_loader: - kv 18: glm4.attention.layer_norm_rms_epsilon f32 = 0.000010
12:12:13 arch ollama[927]: llama_model_loader: - kv 19: glm4.rope.dimension_count u32 = 64
12:12:13 arch ollama[927]: llama_model_loader: - kv 20: tokenizer.ggml.model str = gpt2
12:12:13 arch ollama[927]: llama_model_loader: - kv 21: tokenizer.ggml.pre str = glm4
12:12:13 arch ollama[927]: llama_model_loader: - kv 22: tokenizer.ggml.tokens arr[str,151552] = ["!", """, "#", "$", "%", "&", "'", ...
12:12:13 arch ollama[927]: llama_model_loader: - kv 23: tokenizer.ggml.token_type arr[i32,151552] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
12:12:14 arch ollama[927]: llama_model_loader: - kv 24: tokenizer.ggml.merges arr[str,318088] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "...
12:12:14 arch ollama[927]: llama_model_loader: - kv 25: tokenizer.ggml.eos_token_id u32 = 151329
12:12:14 arch ollama[927]: llama_model_loader: - kv 26: tokenizer.ggml.padding_token_id u32 = 151329
12:12:14 arch ollama[927]: llama_model_loader: - kv 27: tokenizer.ggml.eot_token_id u32 = 151336
12:12:14 arch ollama[927]: llama_model_loader: - kv 28: tokenizer.ggml.unknown_token_id u32 = 151329
12:12:14 arch ollama[927]: llama_model_loader: - kv 29: tokenizer.ggml.bos_token_id u32 = 151329
12:12:14 arch ollama[927]: llama_model_loader: - kv 30: tokenizer.chat_template str = [gMASK]\n{%- for msg in messages ...
12:12:14 arch ollama[927]: llama_model_loader: - kv 31: general.quantization_version u32 = 2
12:12:14 arch ollama[927]: llama_model_loader: - kv 32: general.file_type u32 = 15
12:12:14 arch ollama[927]: llama_model_loader: - type f32: 281 tensors
12:12:14 arch ollama[927]: llama_model_loader: - type q5_0: 20 tensors
12:12:14 arch ollama[927]: llama_model_loader: - type q8_0: 20 tensors
12:12:14 arch ollama[927]: llama_model_loader: - type q4_K: 181 tensors
12:12:14 arch ollama[927]: llama_model_loader: - type q6_K: 21 tensors
12:12:14 arch ollama[927]: print_info: file format = GGUF V3 (latest)
12:12:14 arch ollama[927]: print_info: file type = Q4_K - Medium
12:12:14 arch ollama[927]: print_info: file size = 5.73 GiB (5.24 BPW)
12:12:14 arch ollama[927]: load: special_eot_id is not in special_eog_ids - the tokenizer config may be incorrect
12:12:14 arch ollama[927]: load: printing all EOG tokens:
12:12:14 arch ollama[927]: load: - 151329 ('<|endoftext|>')
12:12:14 arch ollama[927]: load: - 151336 ('<|user|>')
12:12:14 arch ollama[927]: load: special tokens cache size = 23
12:12:14 arch ollama[927]: load: token to piece cache size = 0.9711 MB
12:12:14 arch ollama[927]: print_info: arch = glm4
12:12:14 arch ollama[927]: print_info: vocab_only = 0
12:12:14 arch ollama[927]: print_info: no_alloc = 0
12:12:14 arch ollama[927]: print_info: n_ctx_train = 65536
12:12:14 arch ollama[927]: print_info: n_embd = 4096
12:12:14 arch ollama[927]: print_info: n_embd_inp = 4096
12:12:14 arch ollama[927]: print_info: n_layer = 40
12:12:14 arch ollama[927]: print_info: n_head = 32
12:12:14 arch ollama[927]: print_info: n_head_kv = 2
12:12:14 arch ollama[927]: print_info: n_rot = 64
12:12:14 arch ollama[927]: print_info: n_swa = 0
12:12:14 arch ollama[927]: print_info: is_swa_any = 0
12:12:14 arch ollama[927]: print_info: n_embd_head_k = 128
12:12:14 arch ollama[927]: print_info: n_embd_head_v = 128
12:12:14 arch ollama[927]: print_info: n_gqa = 16
12:12:14 arch ollama[927]: print_info: n_embd_k_gqa = 256
12:12:14 arch ollama[927]: print_info: n_embd_v_gqa = 256
12:12:14 arch ollama[927]: print_info: f_norm_eps = 0.0e+00
12:12:14 arch ollama[927]: print_info: f_norm_rms_eps = 1.0e-05
12:12:14 arch ollama[927]: print_info: f_clamp_kqv = 0.0e+00
12:12:14 arch ollama[927]: print_info: f_max_alibi_bias = 0.0e+00
12:12:14 arch ollama[927]: print_info: f_logit_scale = 0.0e+00
12:12:14 arch ollama[927]: print_info: f_attn_scale = 0.0e+00
12:12:14 arch ollama[927]: print_info: n_ff = 13696
12:12:14 arch ollama[927]: print_info: n_expert = 0
12:12:14 arch ollama[927]: print_info: n_expert_used = 0
12:12:14 arch ollama[927]: print_info: n_expert_groups = 0
12:12:14 arch ollama[927]: print_info: n_group_used = 0
12:12:14 arch ollama[927]: print_info: causal attn = 1
12:12:14 arch ollama[927]: print_info: pooling type = 0
12:12:14 arch ollama[927]: print_info: rope type = 8
12:12:14 arch ollama[927]: print_info: rope scaling = linear
12:12:14 arch ollama[927]: print_info: freq_base_train = 10000.0
12:12:14 arch ollama[927]: print_info: freq_scale_train = 1
12:12:14 arch ollama[927]: print_info: n_ctx_orig_yarn = 65536
12:12:14 arch ollama[927]: print_info: rope_yarn_log_mul= 0.0000
12:12:14 arch ollama[927]: print_info: rope_finetuned = unknown
12:12:14 arch ollama[927]: print_info: mrope sections = [8, 12, 12, 0]
12:12:14 arch ollama[927]: print_info: model type = 9B
12:12:14 arch ollama[927]: print_info: model params = 9.40 B
12:12:14 arch ollama[927]: print_info: general.name = n/a
12:12:14 arch ollama[927]: print_info: vocab type = BPE
12:12:14 arch ollama[927]: print_info: n_vocab = 151552
12:12:14 arch ollama[927]: print_info: n_merges = 318088
12:12:14 arch ollama[927]: print_info: BOS token = 151329 '<|endoftext|>'
12:12:14 arch ollama[927]: print_info: EOS token = 151329 '<|endoftext|>'
12:12:14 arch ollama[927]: print_info: EOT token = 151336 '<|user|>'
12:12:14 arch ollama[927]: print_info: UNK token = 151329 '<|endoftext|>'
12:12:14 arch ollama[927]: print_info: PAD token = 151329 '<|endoftext|>'
12:12:14 arch ollama[927]: print_info: LF token = 198 'Ċ'
12:12:14 arch ollama[927]: print_info: EOG token = 151329 '<|endoftext|>'
12:12:14 arch ollama[927]: print_info: EOG token = 151336 '<|user|>'
12:12:14 arch ollama[927]: print_info: max token length = 1024
12:12:14 arch ollama[927]: load_tensors: loading model tensors, this can take a while... (mmap = true)
12:12:14 arch ollama[927]: load_tensors: offloading 34 repeating layers to GPU
12:12:14 arch ollama[927]: load_tensors: offloaded 34/41 layers to GPU
12:12:14 arch ollama[927]: load_tensors: CPU_Mapped model buffer size = 1617.29 MiB
12:12:14 arch ollama[927]: load_tensors: CUDA0 model buffer size = 4254.72 MiB
12:12:15 arch ollama[927]: llama_context: constructing llama_context
12:12:15 arch ollama[927]: llama_context: n_seq_max = 1
12:12:15 arch ollama[927]: llama_context: n_ctx = 16384
12:12:15 arch ollama[927]: llama_context: n_ctx_seq = 16384
12:12:15 arch ollama[927]: llama_context: n_batch = 512
12:12:15 arch ollama[927]: llama_context: n_ubatch = 512
12:12:15 arch ollama[927]: llama_context: causal_attn = 1
12:12:15 arch ollama[927]: llama_context: flash_attn = auto
12:12:15 arch ollama[927]: llama_context: kv_unified = false
12:12:15 arch ollama[927]: llama_context: freq_base = 10000.0
12:12:15 arch ollama[927]: llama_context: freq_scale = 1
12:12:15 arch ollama[927]: llama_context: n_ctx_seq (16384) < n_ctx_train (65536) -- the full capacity of the model will not be utilized
12:12:15 arch ollama[927]: llama_context: CPU output buffer size = 0.59 MiB
12:12:15 arch ollama[927]: llama_kv_cache: CPU KV buffer size = 96.00 MiB
12:12:15 arch ollama[927]: llama_kv_cache: CUDA0 KV buffer size = 544.00 MiB
12:12:15 arch ollama[927]: llama_kv_cache: size = 640.00 MiB ( 16384 cells, 40 layers, 1/1 seqs), K (f16): 320.00 MiB, V (f16): 320.00 MiB
12:12:15 arch ollama[927]: llama_context: Flash Attention was auto, set to enabled
12:12:15 arch ollama[927]: llama_context: CUDA0 compute buffer size = 789.62 MiB
12:12:15 arch ollama[927]: llama_context: CUDA_Host compute buffer size = 40.02 MiB
12:12:15 arch ollama[927]: llama_context: graph nodes = 1487
12:12:15 arch ollama[927]: llama_context: graph splits = 94 (with bs=512), 3 (with bs=1)
12:12:15 arch ollama[927]: clip_model_loader: model name:
12:12:15 arch ollama[927]: clip_model_loader: description:
12:12:15 arch ollama[927]: clip_model_loader: GGUF version: 3
12:12:15 arch ollama[927]: clip_model_loader: alignment: 32
12:12:15 arch ollama[927]: clip_model_loader: n_tensors: 182
12:12:15 arch ollama[927]: clip_model_loader: n_kv: 25
12:12:15 arch ollama[927]: clip_model_loader: has vision encoder
12:12:15 arch ollama[927]: clip_model_loader: tensor[0]: n_dims = 2, name = v.blk.0.attn_out.weight, tensor_size=2506752, offset=0, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[1]: n_dims = 2, name = v.blk.0.attn_qkv.weight, tensor_size=7520256, offset=2506752, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[2]: n_dims = 2, name = v.blk.0.ffn_down.weight, tensor_size=6684672, offset=10027008, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[3]: n_dims = 2, name = v.blk.0.ffn_gate.weight, tensor_size=6684672, offset=16711680, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[4]: n_dims = 2, name = v.blk.0.ffn_up.weight, tensor_size=6684672, offset=23396352, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[5]: n_dims = 1, name = v.blk.0.ln1.weight, tensor_size=6144, offset=30081024, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[6]: n_dims = 1, name = v.blk.0.ln2.weight, tensor_size=6144, offset=30087168, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[7]: n_dims = 2, name = v.blk.1.attn_out.weight, tensor_size=2506752, offset=30093312, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[8]: n_dims = 2, name = v.blk.1.attn_qkv.weight, tensor_size=7520256, offset=32600064, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[9]: n_dims = 2, name = v.blk.1.ffn_down.weight, tensor_size=6684672, offset=40120320, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[10]: n_dims = 2, name = v.blk.1.ffn_gate.weight, tensor_size=6684672, offset=46804992, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[11]: n_dims = 2, name = v.blk.1.ffn_up.weight, tensor_size=6684672, offset=53489664, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[12]: n_dims = 1, name = v.blk.1.ln1.weight, tensor_size=6144, offset=60174336, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[13]: n_dims = 1, name = v.blk.1.ln2.weight, tensor_size=6144, offset=60180480, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[14]: n_dims = 2, name = v.blk.10.attn_out.weight, tensor_size=2506752, offset=60186624, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[15]: n_dims = 2, name = v.blk.10.attn_qkv.weight, tensor_size=7520256, offset=62693376, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[16]: n_dims = 2, name = v.blk.10.ffn_down.weight, tensor_size=6684672, offset=70213632, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[17]: n_dims = 2, name = v.blk.10.ffn_gate.weight, tensor_size=6684672, offset=76898304, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[18]: n_dims = 2, name = v.blk.10.ffn_up.weight, tensor_size=6684672, offset=83582976, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[19]: n_dims = 1, name = v.blk.10.ln1.weight, tensor_size=6144, offset=90267648, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[20]: n_dims = 1, name = v.blk.10.ln2.weight, tensor_size=6144, offset=90273792, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[21]: n_dims = 2, name = v.blk.11.attn_out.weight, tensor_size=2506752, offset=90279936, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[22]: n_dims = 2, name = v.blk.11.attn_qkv.weight, tensor_size=7520256, offset=92786688, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[23]: n_dims = 2, name = v.blk.11.ffn_down.weight, tensor_size=6684672, offset=100306944, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[24]: n_dims = 2, name = v.blk.11.ffn_gate.weight, tensor_size=6684672, offset=106991616, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[25]: n_dims = 2, name = v.blk.11.ffn_up.weight, tensor_size=6684672, offset=113676288, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[26]: n_dims = 1, name = v.blk.11.ln1.weight, tensor_size=6144, offset=120360960, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[27]: n_dims = 1, name = v.blk.11.ln2.weight, tensor_size=6144, offset=120367104, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[28]: n_dims = 2, name = v.blk.12.attn_out.weight, tensor_size=2506752, offset=120373248, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[29]: n_dims = 2, name = v.blk.12.attn_qkv.weight, tensor_size=7520256, offset=122880000, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[30]: n_dims = 2, name = v.blk.12.ffn_down.weight, tensor_size=6684672, offset=130400256, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[31]: n_dims = 2, name = v.blk.12.ffn_gate.weight, tensor_size=6684672, offset=137084928, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[32]: n_dims = 2, name = v.blk.12.ffn_up.weight, tensor_size=6684672, offset=143769600, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[33]: n_dims = 1, name = v.blk.12.ln1.weight, tensor_size=6144, offset=150454272, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[34]: n_dims = 1, name = v.blk.12.ln2.weight, tensor_size=6144, offset=150460416, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[35]: n_dims = 2, name = v.blk.13.attn_out.weight, tensor_size=2506752, offset=150466560, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[36]: n_dims = 2, name = v.blk.13.attn_qkv.weight, tensor_size=7520256, offset=152973312, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[37]: n_dims = 2, name = v.blk.13.ffn_down.weight, tensor_size=6684672, offset=160493568, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[38]: n_dims = 2, name = v.blk.13.ffn_gate.weight, tensor_size=6684672, offset=167178240, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[39]: n_dims = 2, name = v.blk.13.ffn_up.weight, tensor_size=6684672, offset=173862912, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[40]: n_dims = 1, name = v.blk.13.ln1.weight, tensor_size=6144, offset=180547584, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[41]: n_dims = 1, name = v.blk.13.ln2.weight, tensor_size=6144, offset=180553728, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[42]: n_dims = 2, name = v.blk.14.attn_out.weight, tensor_size=2506752, offset=180559872, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[43]: n_dims = 2, name = v.blk.14.attn_qkv.weight, tensor_size=7520256, offset=183066624, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[44]: n_dims = 2, name = v.blk.14.ffn_down.weight, tensor_size=6684672, offset=190586880, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[45]: n_dims = 2, name = v.blk.14.ffn_gate.weight, tensor_size=6684672, offset=197271552, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[46]: n_dims = 2, name = v.blk.14.ffn_up.weight, tensor_size=6684672, offset=203956224, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[47]: n_dims = 1, name = v.blk.14.ln1.weight, tensor_size=6144, offset=210640896, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[48]: n_dims = 1, name = v.blk.14.ln2.weight, tensor_size=6144, offset=210647040, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[49]: n_dims = 2, name = v.blk.15.attn_out.weight, tensor_size=2506752, offset=210653184, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[50]: n_dims = 2, name = v.blk.15.attn_qkv.weight, tensor_size=7520256, offset=213159936, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[51]: n_dims = 2, name = v.blk.15.ffn_down.weight, tensor_size=6684672, offset=220680192, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[52]: n_dims = 2, name = v.blk.15.ffn_gate.weight, tensor_size=6684672, offset=227364864, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[53]: n_dims = 2, name = v.blk.15.ffn_up.weight, tensor_size=6684672, offset=234049536, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[54]: n_dims = 1, name = v.blk.15.ln1.weight, tensor_size=6144, offset=240734208, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[55]: n_dims = 1, name = v.blk.15.ln2.weight, tensor_size=6144, offset=240740352, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[56]: n_dims = 2, name = v.blk.16.attn_out.weight, tensor_size=2506752, offset=240746496, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[57]: n_dims = 2, name = v.blk.16.attn_qkv.weight, tensor_size=7520256, offset=243253248, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[58]: n_dims = 2, name = v.blk.16.ffn_down.weight, tensor_size=6684672, offset=250773504, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[59]: n_dims = 2, name = v.blk.16.ffn_gate.weight, tensor_size=6684672, offset=257458176, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[60]: n_dims = 2, name = v.blk.16.ffn_up.weight, tensor_size=6684672, offset=264142848, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[61]: n_dims = 1, name = v.blk.16.ln1.weight, tensor_size=6144, offset=270827520, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[62]: n_dims = 1, name = v.blk.16.ln2.weight, tensor_size=6144, offset=270833664, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[63]: n_dims = 2, name = v.blk.17.attn_out.weight, tensor_size=2506752, offset=270839808, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[64]: n_dims = 2, name = v.blk.17.attn_qkv.weight, tensor_size=7520256, offset=273346560, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[65]: n_dims = 2, name = v.blk.17.ffn_down.weight, tensor_size=6684672, offset=280866816, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[66]: n_dims = 2, name = v.blk.17.ffn_gate.weight, tensor_size=6684672, offset=287551488, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[67]: n_dims = 2, name = v.blk.17.ffn_up.weight, tensor_size=6684672, offset=294236160, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[68]: n_dims = 1, name = v.blk.17.ln1.weight, tensor_size=6144, offset=300920832, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[69]: n_dims = 1, name = v.blk.17.ln2.weight, tensor_size=6144, offset=300926976, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[70]: n_dims = 2, name = v.blk.18.attn_out.weight, tensor_size=2506752, offset=300933120, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[71]: n_dims = 2, name = v.blk.18.attn_qkv.weight, tensor_size=7520256, offset=303439872, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[72]: n_dims = 2, name = v.blk.18.ffn_down.weight, tensor_size=6684672, offset=310960128, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[73]: n_dims = 2, name = v.blk.18.ffn_gate.weight, tensor_size=6684672, offset=317644800, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[74]: n_dims = 2, name = v.blk.18.ffn_up.weight, tensor_size=6684672, offset=324329472, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[75]: n_dims = 1, name = v.blk.18.ln1.weight, tensor_size=6144, offset=331014144, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[76]: n_dims = 1, name = v.blk.18.ln2.weight, tensor_size=6144, offset=331020288, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[77]: n_dims = 2, name = v.blk.19.attn_out.weight, tensor_size=2506752, offset=331026432, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[78]: n_dims = 2, name = v.blk.19.attn_qkv.weight, tensor_size=7520256, offset=333533184, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[79]: n_dims = 2, name = v.blk.19.ffn_down.weight, tensor_size=6684672, offset=341053440, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[80]: n_dims = 2, name = v.blk.19.ffn_gate.weight, tensor_size=6684672, offset=347738112, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[81]: n_dims = 2, name = v.blk.19.ffn_up.weight, tensor_size=6684672, offset=354422784, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[82]: n_dims = 1, name = v.blk.19.ln1.weight, tensor_size=6144, offset=361107456, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[83]: n_dims = 1, name = v.blk.19.ln2.weight, tensor_size=6144, offset=361113600, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[84]: n_dims = 2, name = v.blk.2.attn_out.weight, tensor_size=2506752, offset=361119744, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[85]: n_dims = 2, name = v.blk.2.attn_qkv.weight, tensor_size=7520256, offset=363626496, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[86]: n_dims = 2, name = v.blk.2.ffn_down.weight, tensor_size=6684672, offset=371146752, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[87]: n_dims = 2, name = v.blk.2.ffn_gate.weight, tensor_size=6684672, offset=377831424, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[88]: n_dims = 2, name = v.blk.2.ffn_up.weight, tensor_size=6684672, offset=384516096, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[89]: n_dims = 1, name = v.blk.2.ln1.weight, tensor_size=6144, offset=391200768, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[90]: n_dims = 1, name = v.blk.2.ln2.weight, tensor_size=6144, offset=391206912, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[91]: n_dims = 2, name = v.blk.20.attn_out.weight, tensor_size=2506752, offset=391213056, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[92]: n_dims = 2, name = v.blk.20.attn_qkv.weight, tensor_size=7520256, offset=393719808, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[93]: n_dims = 2, name = v.blk.20.ffn_down.weight, tensor_size=6684672, offset=401240064, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[94]: n_dims = 2, name = v.blk.20.ffn_gate.weight, tensor_size=6684672, offset=407924736, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[95]: n_dims = 2, name = v.blk.20.ffn_up.weight, tensor_size=6684672, offset=414609408, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[96]: n_dims = 1, name = v.blk.20.ln1.weight, tensor_size=6144, offset=421294080, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[97]: n_dims = 1, name = v.blk.20.ln2.weight, tensor_size=6144, offset=421300224, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[98]: n_dims = 2, name = v.blk.21.attn_out.weight, tensor_size=2506752, offset=421306368, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[99]: n_dims = 2, name = v.blk.21.attn_qkv.weight, tensor_size=7520256, offset=423813120, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[100]: n_dims = 2, name = v.blk.21.ffn_down.weight, tensor_size=6684672, offset=431333376, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[101]: n_dims = 2, name = v.blk.21.ffn_gate.weight, tensor_size=6684672, offset=438018048, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[102]: n_dims = 2, name = v.blk.21.ffn_up.weight, tensor_size=6684672, offset=444702720, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[103]: n_dims = 1, name = v.blk.21.ln1.weight, tensor_size=6144, offset=451387392, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[104]: n_dims = 1, name = v.blk.21.ln2.weight, tensor_size=6144, offset=451393536, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[105]: n_dims = 2, name = v.blk.22.attn_out.weight, tensor_size=2506752, offset=451399680, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[106]: n_dims = 2, name = v.blk.22.attn_qkv.weight, tensor_size=7520256, offset=453906432, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[107]: n_dims = 2, name = v.blk.22.ffn_down.weight, tensor_size=6684672, offset=461426688, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[108]: n_dims = 2, name = v.blk.22.ffn_gate.weight, tensor_size=6684672, offset=468111360, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[109]: n_dims = 2, name = v.blk.22.ffn_up.weight, tensor_size=6684672, offset=474796032, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[110]: n_dims = 1, name = v.blk.22.ln1.weight, tensor_size=6144, offset=481480704, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[111]: n_dims = 1, name = v.blk.22.ln2.weight, tensor_size=6144, offset=481486848, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[112]: n_dims = 2, name = v.blk.23.attn_out.weight, tensor_size=2506752, offset=481492992, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[113]: n_dims = 2, name = v.blk.23.attn_qkv.weight, tensor_size=7520256, offset=483999744, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[114]: n_dims = 2, name = v.blk.23.ffn_down.weight, tensor_size=6684672, offset=491520000, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[115]: n_dims = 2, name = v.blk.23.ffn_gate.weight, tensor_size=6684672, offset=498204672, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[116]: n_dims = 2, name = v.blk.23.ffn_up.weight, tensor_size=6684672, offset=504889344, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[117]: n_dims = 1, name = v.blk.23.ln1.weight, tensor_size=6144, offset=511574016, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[118]: n_dims = 1, name = v.blk.23.ln2.weight, tensor_size=6144, offset=511580160, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[119]: n_dims = 2, name = v.blk.3.attn_out.weight, tensor_size=2506752, offset=511586304, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[120]: n_dims = 2, name = v.blk.3.attn_qkv.weight, tensor_size=7520256, offset=514093056, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[121]: n_dims = 2, name = v.blk.3.ffn_down.weight, tensor_size=6684672, offset=521613312, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[122]: n_dims = 2, name = v.blk.3.ffn_gate.weight, tensor_size=6684672, offset=528297984, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[123]: n_dims = 2, name = v.blk.3.ffn_up.weight, tensor_size=6684672, offset=534982656, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[124]: n_dims = 1, name = v.blk.3.ln1.weight, tensor_size=6144, offset=541667328, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[125]: n_dims = 1, name = v.blk.3.ln2.weight, tensor_size=6144, offset=541673472, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[126]: n_dims = 2, name = v.blk.4.attn_out.weight, tensor_size=2506752, offset=541679616, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[127]: n_dims = 2, name = v.blk.4.attn_qkv.weight, tensor_size=7520256, offset=544186368, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[128]: n_dims = 2, name = v.blk.4.ffn_down.weight, tensor_size=6684672, offset=551706624, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[129]: n_dims = 2, name = v.blk.4.ffn_gate.weight, tensor_size=6684672, offset=558391296, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[130]: n_dims = 2, name = v.blk.4.ffn_up.weight, tensor_size=6684672, offset=565075968, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[131]: n_dims = 1, name = v.blk.4.ln1.weight, tensor_size=6144, offset=571760640, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[132]: n_dims = 1, name = v.blk.4.ln2.weight, tensor_size=6144, offset=571766784, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[133]: n_dims = 2, name = v.blk.5.attn_out.weight, tensor_size=2506752, offset=571772928, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[134]: n_dims = 2, name = v.blk.5.attn_qkv.weight, tensor_size=7520256, offset=574279680, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[135]: n_dims = 2, name = v.blk.5.ffn_down.weight, tensor_size=6684672, offset=581799936, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[136]: n_dims = 2, name = v.blk.5.ffn_gate.weight, tensor_size=6684672, offset=588484608, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[137]: n_dims = 2, name = v.blk.5.ffn_up.weight, tensor_size=6684672, offset=595169280, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[138]: n_dims = 1, name = v.blk.5.ln1.weight, tensor_size=6144, offset=601853952, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[139]: n_dims = 1, name = v.blk.5.ln2.weight, tensor_size=6144, offset=601860096, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[140]: n_dims = 2, name = v.blk.6.attn_out.weight, tensor_size=2506752, offset=601866240, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[141]: n_dims = 2, name = v.blk.6.attn_qkv.weight, tensor_size=7520256, offset=604372992, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[142]: n_dims = 2, name = v.blk.6.ffn_down.weight, tensor_size=6684672, offset=611893248, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[143]: n_dims = 2, name = v.blk.6.ffn_gate.weight, tensor_size=6684672, offset=618577920, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[144]: n_dims = 2, name = v.blk.6.ffn_up.weight, tensor_size=6684672, offset=625262592, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[145]: n_dims = 1, name = v.blk.6.ln1.weight, tensor_size=6144, offset=631947264, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[146]: n_dims = 1, name = v.blk.6.ln2.weight, tensor_size=6144, offset=631953408, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[147]: n_dims = 2, name = v.blk.7.attn_out.weight, tensor_size=2506752, offset=631959552, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[148]: n_dims = 2, name = v.blk.7.attn_qkv.weight, tensor_size=7520256, offset=634466304, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[149]: n_dims = 2, name = v.blk.7.ffn_down.weight, tensor_size=6684672, offset=641986560, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[150]: n_dims = 2, name = v.blk.7.ffn_gate.weight, tensor_size=6684672, offset=648671232, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[151]: n_dims = 2, name = v.blk.7.ffn_up.weight, tensor_size=6684672, offset=655355904, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[152]: n_dims = 1, name = v.blk.7.ln1.weight, tensor_size=6144, offset=662040576, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[153]: n_dims = 1, name = v.blk.7.ln2.weight, tensor_size=6144, offset=662046720, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[154]: n_dims = 2, name = v.blk.8.attn_out.weight, tensor_size=2506752, offset=662052864, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[155]: n_dims = 2, name = v.blk.8.attn_qkv.weight, tensor_size=7520256, offset=664559616, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[156]: n_dims = 2, name = v.blk.8.ffn_down.weight, tensor_size=6684672, offset=672079872, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[157]: n_dims = 2, name = v.blk.8.ffn_gate.weight, tensor_size=6684672, offset=678764544, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[158]: n_dims = 2, name = v.blk.8.ffn_up.weight, tensor_size=6684672, offset=685449216, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[159]: n_dims = 1, name = v.blk.8.ln1.weight, tensor_size=6144, offset=692133888, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[160]: n_dims = 1, name = v.blk.8.ln2.weight, tensor_size=6144, offset=692140032, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[161]: n_dims = 2, name = v.blk.9.attn_out.weight, tensor_size=2506752, offset=692146176, shape:[1536, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[162]: n_dims = 2, name = v.blk.9.attn_qkv.weight, tensor_size=7520256, offset=694652928, shape:[1536, 4608, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[163]: n_dims = 2, name = v.blk.9.ffn_down.weight, tensor_size=6684672, offset=702173184, shape:[4096, 1536, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[164]: n_dims = 2, name = v.blk.9.ffn_gate.weight, tensor_size=6684672, offset=708857856, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[165]: n_dims = 2, name = v.blk.9.ffn_up.weight, tensor_size=6684672, offset=715542528, shape:[1536, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[166]: n_dims = 1, name = v.blk.9.ln1.weight, tensor_size=6144, offset=722227200, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[167]: n_dims = 1, name = v.blk.9.ln2.weight, tensor_size=6144, offset=722233344, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[168]: n_dims = 1, name = mm.patch_merger.bias, tensor_size=16384, offset=722239488, shape:[4096, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[169]: n_dims = 4, name = mm.patch_merger.weight, tensor_size=50331648, offset=722255872, shape:[2, 2, 1536, 4096], type = f16
12:12:15 arch ollama[927]: clip_model_loader: tensor[170]: n_dims = 2, name = v.position_embd.weight, tensor_size=3538944, offset=772587520, shape:[1536, 576, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[171]: n_dims = 2, name = mm.down.weight, tensor_size=59604992, offset=776126464, shape:[13696, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[172]: n_dims = 2, name = mm.gate.weight, tensor_size=59604992, offset=835731456, shape:[4096, 13696, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[173]: n_dims = 1, name = mm.post_norm.bias, tensor_size=16384, offset=895336448, shape:[4096, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[174]: n_dims = 1, name = mm.post_norm.weight, tensor_size=16384, offset=895352832, shape:[4096, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[175]: n_dims = 2, name = mm.model.fc.weight, tensor_size=17825792, offset=895369216, shape:[4096, 4096, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[176]: n_dims = 2, name = mm.up.weight, tensor_size=59604992, offset=913195008, shape:[4096, 13696, 1, 1], type = q8_0
12:12:15 arch ollama[927]: clip_model_loader: tensor[177]: n_dims = 1, name = v.patch_embd.bias, tensor_size=6144, offset=972800000, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[178]: n_dims = 4, name = v.patch_embd.weight, tensor_size=3612672, offset=972806144, shape:[14, 14, 3, 1536], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[179]: n_dims = 4, name = v.patch_embd.weight.1, tensor_size=3612672, offset=976418816, shape:[14, 14, 3, 1536], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[180]: n_dims = 1, name = v.norm_embd.weight, tensor_size=6144, offset=980031488, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_model_loader: tensor[181]: n_dims = 1, name = v.post_ln.weight, tensor_size=6144, offset=980037632, shape:[1536, 1, 1, 1], type = f32
12:12:15 arch ollama[927]: clip_ctx: CLIP using CUDA0 backend
12:12:15 arch ollama[927]: load_hparams: projector: glm4v
12:12:15 arch ollama[927]: load_hparams: n_embd: 1536
12:12:15 arch ollama[927]: load_hparams: n_head: 12
12:12:15 arch ollama[927]: load_hparams: n_ff: 13696
12:12:15 arch ollama[927]: load_hparams: n_layer: 24
12:12:15 arch ollama[927]: load_hparams: ffn_op: silu
12:12:15 arch ollama[927]: load_hparams: projection_dim: 4096
12:12:15 arch ollama[927]: --- vision hparams ---
12:12:15 arch ollama[927]: load_hparams: image_size: 336
12:12:15 arch ollama[927]: load_hparams: patch_size: 14
12:12:15 arch ollama[927]: load_hparams: has_llava_proj: 0
12:12:15 arch ollama[927]: load_hparams: minicpmv_version: 0
12:12:15 arch ollama[927]: load_hparams: n_merge: 2
12:12:15 arch ollama[927]: load_hparams: n_wa_pattern: 0
12:12:15 arch ollama[927]: load_hparams: image_min_pixels: 6272
12:12:15 arch ollama[927]: load_hparams: image_max_pixels: 3211264
12:12:15 arch ollama[927]: load_hparams: model size: 934.64 MiB
12:12:15 arch ollama[927]: load_hparams: metadata size: 0.06 MiB
12:12:16 arch ollama[927]: load_tensors: loaded 182 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-454dd441c925c1a81c984204bb9d54feef0ef07b789c6fe1118099014ba2727d
12:12:16 arch ollama[927]: warmup: warmup with image size = 1288 x 1288
12:12:16 arch ollama[927]: ggml_backend_cuda_buffer_type_alloc_buffer: allocating 515.05 MiB on device 0: cudaMalloc failed: out of memory
12:12:16 arch ollama[927]: ggml_gallocr_reserve_n_impl: failed to allocate CUDA0 buffer of size 540070912
12:12:16 arch ollama[927]: alloc_compute_meta: CPU compute buffer size = 19.11 MiB
12:12:16 arch ollama[927]: alloc_compute_meta: graph splits = 1, nodes = 632
12:12:16 arch ollama[927]: warmup: flash attention is enabled
12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.324+08:00 level=INFO source=server.go:1376 msg="llama runner started in 2.61 seconds"
12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.324+08:00 level=INFO source=sched.go:517 msg="loaded runners" count=1
12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.324+08:00 level=INFO source=server.go:1338 msg="waiting for llama runner to start responding"
12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.325+08:00 level=INFO source=server.go:1376 msg="llama runner started in 2.61 seconds"
12:12:16 arch ollama[927]: add_text: <|begin_of_image|>
12:12:16 arch ollama[927]: image_tokens->nx = 85
12:12:16 arch ollama[927]: image_tokens->ny = 48
12:12:16 arch ollama[927]: batch_f32 size = 1
12:12:16 arch ollama[927]: add_text: <|end_of_image|>
12:12:16 arch ollama[927]: ggml_backend_cuda_buffer_type_alloc_buffer: allocating 993.11 MiB on device 0: cudaMalloc failed: out of memory
12:12:16 arch ollama[927]: ggml_gallocr_reserve_n_impl: failed to allocate CUDA0 buffer of size 1041346560
12:12:16 arch ollama[927]: SIGSEGV: segmentation violation
12:12:16 arch ollama[927]: PC=0x564e038541cb m=15 sigcode=1 addr=0x0
12:12:16 arch ollama[927]: signal arrived during cgo execution
12:12:16 arch ollama[927]: goroutine 40 gp=0xc000505c00 m=15 mp=0xc000101008 [syscall]:
12:12:16 arch ollama[927]: runtime.cgocall(0x564e03841190, 0xc0000491d8)
12:12:16 arch ollama[927]: runtime/cgocall.go:167 +0x4b fp=0xc0000491b0 sp=0xc000049178 pc=0x564e02aef6eb
12:12:16 arch ollama[927]: github.com/ollama/ollama/llama._Cfunc_mtmd_encode_chunk(0x7f834008dd90, 0x7f82817bc000)
12:12:16 arch ollama[927]: _cgo_gotypes.go:1079 +0x4a fp=0xc0000491d8 sp=0xc0000491b0 pc=0x564e02eaa04a
12:12:16 arch ollama[927]: github.com/ollama/ollama/llama.(*MtmdContext).MultimodalTokenize.func11(...)
12:12:16 arch ollama[927]: github.com/ollama/ollama/llama/llama.go:595
12:12:16 arch ollama[927]: github.com/ollama/ollama/llama.(*MtmdContext).MultimodalTokenize(0xc00034c0b8, 0xc0003a6750, {0xc000a82000, 0x117bcb, 0xc000049520?})
12:12:16 arch ollama[927]: github.com/ollama/ollama/llama/llama.go:595 +0x6c5 fp=0xc000049490 sp=0xc0000491d8 pc=0x564e02eaf085
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*ImageContext).MultimodalTokenize(0xc00063c240, 0xc0003a6750, {0xc000a82000, 0x117bcb, 0x117bcd})
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/image.go:76 +0x145 fp=0xc000049530 sp=0xc000049490 pc=0x564e02f61365
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).inputs(0xc00013f900, {0xc000042440?, 0x3d?}, {0xc00061c080, 0x1, 0x160?})
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:236 +0x2c6 fp=0xc000049698 sp=0xc000049530 pc=0x564e02f62766
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).NewSequence(0xc00013f900, {0xc000042440, 0x3d}, {0xc00061c080, 0x1, 0x1}, {0x28000, {0xc000044610, 0x1, 0x1}, ...})
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:126 +0x8d fp=0xc000049838 sp=0xc000049698 pc=0x564e02f61d2d
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).completion(0xc00013f900, {0x564e040a5fa0, 0xc0003c21c0}, 0xc0003b2280)
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:659 +0x5f9 fp=0xc000049ac0 sp=0xc000049838 pc=0x564e02f64d99
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).completion-fm({0x564e040a5fa0?, 0xc0003c21c0?}, 0xc000049b40?)
12:12:16 arch ollama[927]: :1 +0x36 fp=0xc000049af0 sp=0xc000049ac0 pc=0x564e02f687f6
12:12:16 arch ollama[927]: net/http.HandlerFunc.ServeHTTP(0xc000544180?, {0x564e040a5fa0?, 0xc0003c21c0?}, 0xc000049b60?)
12:12:16 arch ollama[927]: net/http/server.go:2294 +0x29 fp=0xc000049b18 sp=0xc000049af0 pc=0x564e02df20e9
12:12:16 arch ollama[927]: net/http.(*ServeMux).ServeHTTP(0x564e02a978c5?, {0x564e040a5fa0, 0xc0003c21c0}, 0xc0003b2280)
12:12:16 arch ollama[927]: net/http/server.go:2822 +0x1c4 fp=0xc000049b68 sp=0xc000049b18 pc=0x564e02df3fe4
12:12:16 arch ollama[927]: net/http.serverHandler.ServeHTTP({0x564e040a2590?}, {0x564e040a5fa0?, 0xc0003c21c0?}, 0x1?)
12:12:16 arch ollama[927]: net/http/server.go:3301 +0x8e fp=0xc000049b98 sp=0xc000049b68 pc=0x564e02e11a6e
12:12:16 arch ollama[927]: net/http.(*conn).serve(0xc0001463f0, {0x564e040a83d8, 0xc000144b10})
12:12:16 arch ollama[927]: net/http/server.go:2102 +0x625 fp=0xc000049fb8 sp=0xc000049b98 pc=0x564e02df05e5
12:12:16 arch ollama[927]: net/http.(*Server).Serve.gowrap3()
12:12:16 arch ollama[927]: net/http/server.go:3454 +0x28 fp=0xc000049fe0 sp=0xc000049fb8 pc=0x564e02df5ea8
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000049fe8 sp=0xc000049fe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by net/http.(*Server).Serve in goroutine 1
12:12:16 arch ollama[927]: net/http/server.go:3454 +0x485
12:12:16 arch ollama[927]: goroutine 1 gp=0xc000002380 m=nil [IO wait]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00018f790 sp=0xc00018f770 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.netpollblock(0xc0005197e0?, 0x2a8c2a6?, 0x4e?)
12:12:16 arch ollama[927]: runtime/netpoll.go:575 +0xf7 fp=0xc00018f7c8 sp=0xc00018f790 pc=0x564e02ab7e97
12:12:16 arch ollama[927]: internal/poll.runtime_pollWait(0x7f83f643aeb0, 0x72)
12:12:16 arch ollama[927]: runtime/netpoll.go:351 +0x85 fp=0xc00018f7e8 sp=0xc00018f7c8 pc=0x564e02af1d85
12:12:16 arch ollama[927]: internal/poll.(*pollDesc).wait(0xc000701480?, 0x900000036?, 0x0)
12:12:16 arch ollama[927]: internal/poll/fd_poll_runtime.go:84 +0x27 fp=0xc00018f810 sp=0xc00018f7e8 pc=0x564e02b79f07
12:12:16 arch ollama[927]: internal/poll.(*pollDesc).waitRead(...)
12:12:16 arch ollama[927]: internal/poll/fd_poll_runtime.go:89
12:12:16 arch ollama[927]: internal/poll.(*FD).Accept(0xc000701480)
12:12:16 arch ollama[927]: internal/poll/fd_unix.go:620 +0x295 fp=0xc00018f8b8 sp=0xc00018f810 pc=0x564e02b7f2d5
12:12:16 arch ollama[927]: net.(*netFD).accept(0xc000701480)
12:12:16 arch ollama[927]: net/fd_unix.go:172 +0x29 fp=0xc00018f970 sp=0xc00018f8b8 pc=0x564e02bf21a9
12:12:16 arch ollama[927]: net.(*TCPListener).accept(0xc000531200)
12:12:16 arch ollama[927]: net/tcpsock_posix.go:159 +0x1b fp=0xc00018f9c0 sp=0xc00018f970 pc=0x564e02c07b5b
12:12:16 arch ollama[927]: net.(*TCPListener).Accept(0xc000531200)
12:12:16 arch ollama[927]: net/tcpsock.go:380 +0x30 fp=0xc00018f9f0 sp=0xc00018f9c0 pc=0x564e02c06a10
12:12:16 arch ollama[927]: net/http.(*onceCloseListener).Accept(0xc0001463f0?)
12:12:16 arch ollama[927]: :1 +0x24 fp=0xc00018fa08 sp=0xc00018f9f0 pc=0x564e02e1e1e4
12:12:16 arch ollama[927]: net/http.(*Server).Serve(0xc000213700, {0x564e040a5dc0, 0xc000531200})
12:12:16 arch ollama[927]: net/http/server.go:3424 +0x30c fp=0xc00018fb38 sp=0xc00018fa08 pc=0x564e02df5aac
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.Execute({0xc000034260, 0x4, 0x4})
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:1002 +0x8f5 fp=0xc00018fd08 sp=0xc00018fb38 pc=0x564e02f68175
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner.Execute({0xc000034250?, 0x0?, 0x0?})
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/runner.go:22 +0xd4 fp=0xc00018fd30 sp=0xc00018fd08 pc=0x564e03013cf4
12:12:16 arch ollama[927]: github.com/ollama/ollama/cmd.NewCLI.func2(0xc000213400?, {0x564e03b880ad?, 0x4?, 0x564e03b880b1?})
12:12:16 arch ollama[927]: github.com/ollama/ollama/cmd/cmd.go:1841 +0x45 fp=0xc00018fd58 sp=0xc00018fd30 pc=0x564e037d0f25
12:12:16 arch ollama[927]: github.com/spf13/cobra.(*Command).execute(0xc000149508, {0xc000531000, 0x4, 0x4})
12:12:16 arch ollama[927]: github.com/spf13/cobra@v1.7.0/command.go:940 +0x85c fp=0xc00018fe78 sp=0xc00018fd58 pc=0x564e02c6b7fc
12:12:16 arch ollama[927]: github.com/spf13/cobra.(*Command).ExecuteC(0xc000126908)
12:12:16 arch ollama[927]: github.com/spf13/cobra@v1.7.0/command.go:1068 +0x3a5 fp=0xc00018ff30 sp=0xc00018fe78 pc=0x564e02c6c045
12:12:16 arch ollama[927]: github.com/spf13/cobra.(*Command).Execute(...)
12:12:16 arch ollama[927]: github.com/spf13/cobra@v1.7.0/command.go:992
12:12:16 arch ollama[927]: github.com/spf13/cobra.(*Command).ExecuteContext(...)
12:12:16 arch ollama[927]: github.com/spf13/cobra@v1.7.0/command.go:985
12:12:16 arch ollama[927]: main.main()
12:12:16 arch ollama[927]: github.com/ollama/ollama/main.go:12 +0x4d fp=0xc00018ff50 sp=0xc00018ff30 pc=0x564e037d1a0d
12:12:16 arch ollama[927]: runtime.main()
12:12:16 arch ollama[927]: runtime/proc.go:283 +0x29d fp=0xc00018ffe0 sp=0xc00018ff50 pc=0x564e02abf51d
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00018ffe8 sp=0xc00018ffe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: goroutine 2 gp=0xc000002e00 m=nil [force gc (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008efa8 sp=0xc00008ef88 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.goparkunlock(...)
12:12:16 arch ollama[927]: runtime/proc.go:441
12:12:16 arch ollama[927]: runtime.forcegchelper()
12:12:16 arch ollama[927]: runtime/proc.go:348 +0xb8 fp=0xc00008efe0 sp=0xc00008efa8 pc=0x564e02abf858
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008efe8 sp=0xc00008efe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.init.7 in goroutine 1
12:12:16 arch ollama[927]: runtime/proc.go:336 +0x1a
12:12:16 arch ollama[927]: goroutine 3 gp=0xc000003340 m=nil [GC sweep wait]:
12:12:16 arch ollama[927]: runtime.gopark(0x1?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008f780 sp=0xc00008f760 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.goparkunlock(...)
12:12:16 arch ollama[927]: runtime/proc.go:441
12:12:16 arch ollama[927]: runtime.bgsweep(0xc0000ba000)
12:12:16 arch ollama[927]: runtime/mgcsweep.go:316 +0xdf fp=0xc00008f7c8 sp=0xc00008f780 pc=0x564e02aa9fff
12:12:16 arch ollama[927]: runtime.gcenable.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:204 +0x25 fp=0xc00008f7e0 sp=0xc00008f7c8 pc=0x564e02a9e3e5
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008f7e8 sp=0xc00008f7e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcenable in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:204 +0x66
12:12:16 arch ollama[927]: goroutine 4 gp=0xc000003500 m=nil [GC scavenge wait]:
12:12:16 arch ollama[927]: runtime.gopark(0x10000?, 0x564e03d592a0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008ff78 sp=0xc00008ff58 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.goparkunlock(...)
12:12:16 arch ollama[927]: runtime/proc.go:441
12:12:16 arch ollama[927]: runtime.(*scavengerState).park(0x564e0497c280)
12:12:16 arch ollama[927]: runtime/mgcscavenge.go:425 +0x49 fp=0xc00008ffa8 sp=0xc00008ff78 pc=0x564e02aa7a49
12:12:16 arch ollama[927]: runtime.bgscavenge(0xc0000ba000)
12:12:16 arch ollama[927]: runtime/mgcscavenge.go:658 +0x59 fp=0xc00008ffc8 sp=0xc00008ffa8 pc=0x564e02aa7fd9
12:12:16 arch ollama[927]: runtime.gcenable.gowrap2()
12:12:16 arch ollama[927]: runtime/mgc.go:205 +0x25 fp=0xc00008ffe0 sp=0xc00008ffc8 pc=0x564e02a9e385
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008ffe8 sp=0xc00008ffe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcenable in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:205 +0xa5
12:12:16 arch ollama[927]: goroutine 5 gp=0xc000003dc0 m=nil [finalizer wait]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x564e04092250?, 0x40?, 0x61?, 0x1000000010?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008e630 sp=0xc00008e610 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.runfinq()
12:12:16 arch ollama[927]: runtime/mfinal.go:196 +0x107 fp=0xc00008e7e0 sp=0xc00008e630 pc=0x564e02a9d3a7
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008e7e8 sp=0xc00008e7e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.createfing in goroutine 1
12:12:16 arch ollama[927]: runtime/mfinal.go:166 +0x3d
12:12:16 arch ollama[927]: goroutine 6 gp=0xc0001f08c0 m=nil [chan receive]:
12:12:16 arch ollama[927]: runtime.gopark(0xc000245720?, 0xc000590018?, 0x60?, 0x7?, 0x564e02bd8de8?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000090718 sp=0xc0000906f8 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.chanrecv(0xc0000c6310, 0x0, 0x1)
12:12:16 arch ollama[927]: runtime/chan.go:664 +0x445 fp=0xc000090790 sp=0xc000090718 pc=0x564e02a8ee85
12:12:16 arch ollama[927]: runtime.chanrecv1(0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/chan.go:506 +0x12 fp=0xc0000907b8 sp=0xc000090790 pc=0x564e02a8ea12
12:12:16 arch ollama[927]: runtime.unique_runtime_registerUniqueMapCleanup.func2(...)
12:12:16 arch ollama[927]: runtime/mgc.go:1796
12:12:16 arch ollama[927]: runtime.unique_runtime_registerUniqueMapCleanup.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1799 +0x2f fp=0xc0000907e0 sp=0xc0000907b8 pc=0x564e02aa158f
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0000907e8 sp=0xc0000907e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by unique.runtime_registerUniqueMapCleanup in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1794 +0x85
12:12:16 arch ollama[927]: goroutine 7 gp=0xc0001f0c40 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000090f38 sp=0xc000090f18 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000090fc8 sp=0xc000090f38 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000090fe0 sp=0xc000090fc8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000090fe8 sp=0xc000090fe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 18 gp=0xc000504000 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008a738 sp=0xc00008a718 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008a7c8 sp=0xc00008a738 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008a7e0 sp=0xc00008a7c8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008a7e8 sp=0xc00008a7e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 19 gp=0xc0005041c0 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008af38 sp=0xc00008af18 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008afc8 sp=0xc00008af38 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008afe0 sp=0xc00008afc8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008afe8 sp=0xc00008afe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 34 gp=0xc000102380 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00011a738 sp=0xc00011a718 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00011a7c8 sp=0xc00011a738 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00011a7e0 sp=0xc00011a7c8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00011a7e8 sp=0xc00011a7e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 35 gp=0xc000102540 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00011af38 sp=0xc00011af18 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00011afc8 sp=0xc00011af38 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00011afe0 sp=0xc00011afc8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00011afe8 sp=0xc00011afe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 20 gp=0xc000504380 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008b738 sp=0xc00008b718 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008b7c8 sp=0xc00008b738 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008b7e0 sp=0xc00008b7c8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008b7e8 sp=0xc00008b7e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 21 gp=0xc000504540 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008bf38 sp=0xc00008bf18 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008bfc8 sp=0xc00008bf38 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008bfe0 sp=0xc00008bfc8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008bfe8 sp=0xc00008bfe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 22 gp=0xc000504700 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008c738 sp=0xc00008c718 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008c7c8 sp=0xc00008c738 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008c7e0 sp=0xc00008c7c8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008c7e8 sp=0xc00008c7e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 23 gp=0xc0005048c0 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008cf38 sp=0xc00008cf18 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008cfc8 sp=0xc00008cf38 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008cfe0 sp=0xc00008cfc8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008cfe8 sp=0xc00008cfe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 24 gp=0xc000504a80 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008d738 sp=0xc00008d718 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008d7c8 sp=0xc00008d738 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008d7e0 sp=0xc00008d7c8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008d7e8 sp=0xc00008d7e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 25 gp=0xc000504c40 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008df38 sp=0xc00008df18 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008dfc8 sp=0xc00008df38 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008dfe0 sp=0xc00008dfc8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008dfe8 sp=0xc00008dfe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 8 gp=0xc0001f0e00 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000091738 sp=0xc000091718 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0000917c8 sp=0xc000091738 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0000917e0 sp=0xc0000917c8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0000917e8 sp=0xc0000917e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 36 gp=0xc000102700 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00011b738 sp=0xc00011b718 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00011b7c8 sp=0xc00011b738 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00011b7e0 sp=0xc00011b7c8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00011b7e8 sp=0xc00011b7e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 37 gp=0xc0001028c0 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00011bf38 sp=0xc00011bf18 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00011bfc8 sp=0xc00011bf38 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00011bfe0 sp=0xc00011bfc8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00011bfe8 sp=0xc00011bfe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 9 gp=0xc0001f0fc0 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x16e3afe5ca10f?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000091f38 sp=0xc000091f18 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000091fc8 sp=0xc000091f38 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000091fe0 sp=0xc000091fc8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000091fe8 sp=0xc000091fe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 10 gp=0xc0001f1180 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x564e04a4a680?, 0x3?, 0xda?, 0x8f?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000116738 sp=0xc000116718 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0001167c8 sp=0xc000116738 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0001167e0 sp=0xc0001167c8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0001167e8 sp=0xc0001167e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 11 gp=0xc0001f1340 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x564e04a4a680?, 0x1?, 0x8f?, 0x69?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000116f38 sp=0xc000116f18 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000116fc8 sp=0xc000116f38 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000116fe0 sp=0xc000116fc8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000116fe8 sp=0xc000116fe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 12 gp=0xc0001f1500 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x564e04a4a680?, 0x1?, 0x46?, 0x67?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000117738 sp=0xc000117718 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0001177c8 sp=0xc000117738 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0001177e0 sp=0xc0001177c8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0001177e8 sp=0xc0001177e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 13 gp=0xc0001f16c0 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x16e3afe5baaf9?, 0x0?, 0x0?, 0x0?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000117f38 sp=0xc000117f18 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000117fc8 sp=0xc000117f38 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000117fe0 sp=0xc000117fc8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000117fe8 sp=0xc000117fe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 14 gp=0xc0001f1880 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x16e3afe5bde06?, 0x1?, 0xd5?, 0x34?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000118738 sp=0xc000118718 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0001187c8 sp=0xc000118738 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0001187e0 sp=0xc0001187c8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0001187e8 sp=0xc0001187e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 15 gp=0xc0001f1a40 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x564e04a4a680?, 0x1?, 0x78?, 0x39?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000118f38 sp=0xc000118f18 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000118fc8 sp=0xc000118f38 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000118fe0 sp=0xc000118fc8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000118fe8 sp=0xc000118fe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 38 gp=0xc000102a80 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x564e04a4a680?, 0x1?, 0x6a?, 0xc4?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00011c738 sp=0xc00011c718 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00011c7c8 sp=0xc00011c738 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00011c7e0 sp=0xc00011c7c8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00011c7e8 sp=0xc00011c7e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 16 gp=0xc0001f1c00 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x16e3afe5bc298?, 0x1?, 0xb0?, 0x10?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000119738 sp=0xc000119718 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0001197c8 sp=0xc000119738 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0001197e0 sp=0xc0001197c8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0001197e8 sp=0xc0001197e0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 50 gp=0xc0001f1dc0 m=nil [GC worker (idle)]:
12:12:16 arch ollama[927]: runtime.gopark(0x16e3afe5ba5b0?, 0x1?, 0x1d?, 0x9c?, 0x0?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000119f38 sp=0xc000119f18 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730)
12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000119fc8 sp=0xc000119f38 pc=0x564e02aa08a9
12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000119fe0 sp=0xc000119fc8 pc=0x564e02aa0785
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000119fe8 sp=0xc000119fe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105
12:12:16 arch ollama[927]: goroutine 39 gp=0xc000505a40 m=nil [sync.Cond.Wait]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0xc000165c78?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000165be8 sp=0xc000165bc8 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.goparkunlock(...)
12:12:16 arch ollama[927]: runtime/proc.go:441
12:12:16 arch ollama[927]: sync.runtime_notifyListWait(0xc0005311d0, 0x0)
12:12:16 arch ollama[927]: runtime/sema.go:597 +0x15a fp=0xc000165c38 sp=0xc000165be8 pc=0x564e02af46ba
12:12:16 arch ollama[927]: sync.(*Cond).Wait(0xc000165cb8?)
12:12:16 arch ollama[927]: sync/cond.go:71 +0x85 fp=0xc000165c70 sp=0xc000165c38 pc=0x564e02b047c5
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).processBatch(0xc00013f900, 0xc00033c140, 0xc00033c190)
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:408 +0x93 fp=0xc000165ee8 sp=0xc000165c70 pc=0x564e02f63213
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).run(0xc00013f900, {0x564e040a8410, 0xc000623090})
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:387 +0x1d5 fp=0xc000165fb8 sp=0xc000165ee8 pc=0x564e02f63015
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.Execute.gowrap1()
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:981 +0x28 fp=0xc000165fe0 sp=0xc000165fb8 pc=0x564e02f683e8
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000165fe8 sp=0xc000165fe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by github.com/ollama/ollama/runner/llamarunner.Execute in goroutine 1
12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:981 +0x4c5
12:12:16 arch ollama[927]: goroutine 45 gp=0xc000602fc0 m=nil [IO wait]:
12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0xb?)
12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000163dd8 sp=0xc000163db8 pc=0x564e02af2b6e
12:12:16 arch ollama[927]: runtime.netpollblock(0x564e02b16338?, 0x2a8c2a6?, 0x4e?)
12:12:16 arch ollama[927]: runtime/netpoll.go:575 +0xf7 fp=0xc000163e10 sp=0xc000163dd8 pc=0x564e02ab7e97
12:12:16 arch ollama[927]: internal/poll.runtime_pollWait(0x7f83f643ad98, 0x72)
12:12:16 arch ollama[927]: runtime/netpoll.go:351 +0x85 fp=0xc000163e30 sp=0xc000163e10 pc=0x564e02af1d85
12:12:16 arch ollama[927]: internal/poll.(*pollDesc).wait(0xc000701500?, 0xc000144c11?, 0x0)
12:12:16 arch ollama[927]: internal/poll/fd_poll_runtime.go:84 +0x27 fp=0xc000163e58 sp=0xc000163e30 pc=0x564e02b79f07
12:12:16 arch ollama[927]: internal/poll.(*pollDesc).waitRead(...)
12:12:16 arch ollama[927]: internal/poll/fd_poll_runtime.go:89
12:12:16 arch ollama[927]: internal/poll.(*FD).Read(0xc000701500, {0xc000144c11, 0x1, 0x1})
12:12:16 arch ollama[927]: internal/poll/fd_unix.go:165 +0x27a fp=0xc000163ef0 sp=0xc000163e58 pc=0x564e02b7b1fa
12:12:16 arch ollama[927]: net.(*netFD).Read(0xc000701500, {0xc000144c11?, 0x0?, 0x0?})
12:12:16 arch ollama[927]: net/fd_posix.go:55 +0x25 fp=0xc000163f38 sp=0xc000163ef0 pc=0x564e02bf0205
12:12:16 arch ollama[927]: net.(*conn).Read(0xc000092908, {0xc000144c11?, 0x0?, 0x0?})
12:12:16 arch ollama[927]: net/net.go:194 +0x45 fp=0xc000163f80 sp=0xc000163f38 pc=0x564e02bfe5c5
12:12:16 arch ollama[927]: net/http.(*connReader).backgroundRead(0xc000144c00)
12:12:16 arch ollama[927]: net/http/server.go:690 +0x37 fp=0xc000163fc8 sp=0xc000163f80 pc=0x564e02dea4b7
12:12:16 arch ollama[927]: net/http.(*connReader).startBackgroundRead.gowrap2()
12:12:16 arch ollama[927]: net/http/server.go:686 +0x25 fp=0xc000163fe0 sp=0xc000163fc8 pc=0x564e02dea3e5
12:12:16 arch ollama[927]: runtime.goexit({})
12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000163fe8 sp=0xc000163fe0 pc=0x564e02afaa01
12:12:16 arch ollama[927]: created by net/http.(*connReader).startBackgroundRead in goroutine 40
12:12:16 arch ollama[927]: net/http/server.go:686 +0xb6
12:12:16 arch ollama[927]: rax 0x7f82d2850890
12:12:16 arch ollama[927]: rbx 0x0
12:12:16 arch ollama[927]: rcx 0x0
12:12:16 arch ollama[927]: rdx 0x0
12:12:16 arch ollama[927]: rdi 0x7f82d25dfff0
12:12:16 arch ollama[927]: rsi 0x7f82c622f310
12:12:16 arch ollama[927]: rbp 0x7f82d25e7e68
12:12:16 arch ollama[927]: rsp 0x7f82d5ffcbc0
12:12:16 arch ollama[927]: r8 0x0
12:12:16 arch ollama[927]: r9 0x7f82736c7040
12:12:16 arch ollama[927]: r10 0x24db000
12:12:16 arch ollama[927]: r11 0x246
12:12:16 arch ollama[927]: r12 0x7f81c445a9e0
12:12:16 arch ollama[927]: r13 0x160
12:12:16 arch ollama[927]: r14 0x7f82c622f030
12:12:16 arch ollama[927]: r15 0x1
12:12:16 arch ollama[927]: rip 0x564e038541cb
12:12:16 arch ollama[927]: rflags 0x10246
12:12:16 arch ollama[927]: cs 0x33
12:12:16 arch ollama[927]: fs 0x0
12:12:16 arch ollama[927]: gs 0x0
12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.610+08:00 level=ERROR source=server.go:1583 msg="post predict" error="Post "http://127.0.0.1:33773/completion": EOF"

<!-- gh-comment-id:3709154784 --> @pinghe commented on GitHub (Jan 5, 2026): 12:12:12 arch ollama[927]: time=2026-01-05T12:12:12.828+08:00 level=INFO source=runner.go:464 msg="failure during GPU discovery" OLLAMA_LIBRARY_PATH="[/usr/local/lib/ollama /usr/local/lib/ollama/cuda_v13]" extra_envs=map[] error="failed to finish discovery before timeout" 12:12:12 arch ollama[927]: time=2026-01-05T12:12:12.828+08:00 level=WARN source=runner.go:356 msg="unable to refresh free memory, using old values" 12:12:12 arch ollama[927]: time=2026-01-05T12:12:12.829+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 33053" 12:12:13 arch ollama[927]: llama_model_loader: loaded meta data with 33 key-value pairs and 523 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20af820432e1dd55d22a2b0e728e79223e77a (version GGUF V3 (latest)) 12:12:13 arch ollama[927]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. 12:12:13 arch ollama[927]: llama_model_loader: - kv 0: general.architecture str = glm4 12:12:13 arch ollama[927]: llama_model_loader: - kv 1: general.type str = model 12:12:13 arch ollama[927]: llama_model_loader: - kv 2: general.size_label str = 9.4B 12:12:13 arch ollama[927]: llama_model_loader: - kv 3: general.license str = mit 12:12:13 arch ollama[927]: llama_model_loader: - kv 4: general.base_model.count u32 = 1 12:12:13 arch ollama[927]: llama_model_loader: - kv 5: general.base_model.0.name str = GLM 4.1V 9B Base 12:12:13 arch ollama[927]: llama_model_loader: - kv 6: general.base_model.0.organization str = Zai Org 12:12:13 arch ollama[927]: llama_model_loader: - kv 7: general.base_model.0.repo_url str = https://huggingface.co/zai-org/GLM-4.... 12:12:13 arch ollama[927]: llama_model_loader: - kv 8: general.tags arr[str,2] = ["agent", "image-text-to-text"] 12:12:13 arch ollama[927]: llama_model_loader: - kv 9: general.languages arr[str,1] = ["zh"] 12:12:13 arch ollama[927]: llama_model_loader: - kv 10: glm4.block_count u32 = 40 12:12:13 arch ollama[927]: llama_model_loader: - kv 11: glm4.context_length u32 = 65536 12:12:13 arch ollama[927]: llama_model_loader: - kv 12: glm4.embedding_length u32 = 4096 12:12:13 arch ollama[927]: llama_model_loader: - kv 13: glm4.feed_forward_length u32 = 13696 12:12:13 arch ollama[927]: llama_model_loader: - kv 14: glm4.attention.head_count u32 = 32 12:12:13 arch ollama[927]: llama_model_loader: - kv 15: glm4.attention.head_count_kv u32 = 2 12:12:13 arch ollama[927]: llama_model_loader: - kv 16: glm4.rope.dimension_sections arr[i32,4] = [8, 12, 12, 0] 12:12:13 arch ollama[927]: llama_model_loader: - kv 17: glm4.rope.freq_base f32 = 10000.000000 12:12:13 arch ollama[927]: llama_model_loader: - kv 18: glm4.attention.layer_norm_rms_epsilon f32 = 0.000010 12:12:13 arch ollama[927]: llama_model_loader: - kv 19: glm4.rope.dimension_count u32 = 64 12:12:13 arch ollama[927]: llama_model_loader: - kv 20: tokenizer.ggml.model str = gpt2 12:12:13 arch ollama[927]: llama_model_loader: - kv 21: tokenizer.ggml.pre str = glm4 12:12:13 arch ollama[927]: llama_model_loader: - kv 22: tokenizer.ggml.tokens arr[str,151552] = ["!", "\"", "#", "$", "%", "&", "'", ... 12:12:13 arch ollama[927]: llama_model_loader: - kv 23: tokenizer.ggml.token_type arr[i32,151552] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... 12:12:13 arch ollama[927]: llama_model_loader: - kv 24: tokenizer.ggml.merges arr[str,318088] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "... 12:12:13 arch ollama[927]: llama_model_loader: - kv 25: tokenizer.ggml.eos_token_id u32 = 151329 12:12:13 arch ollama[927]: llama_model_loader: - kv 26: tokenizer.ggml.padding_token_id u32 = 151329 12:12:13 arch ollama[927]: llama_model_loader: - kv 27: tokenizer.ggml.eot_token_id u32 = 151336 12:12:13 arch ollama[927]: llama_model_loader: - kv 28: tokenizer.ggml.unknown_token_id u32 = 151329 12:12:13 arch ollama[927]: llama_model_loader: - kv 29: tokenizer.ggml.bos_token_id u32 = 151329 12:12:13 arch ollama[927]: llama_model_loader: - kv 30: tokenizer.chat_template str = [gMASK]<sop>\n{%- for msg in messages ... 12:12:13 arch ollama[927]: llama_model_loader: - kv 31: general.quantization_version u32 = 2 12:12:13 arch ollama[927]: llama_model_loader: - kv 32: general.file_type u32 = 15 12:12:13 arch ollama[927]: llama_model_loader: - type f32: 281 tensors 12:12:13 arch ollama[927]: llama_model_loader: - type q5_0: 20 tensors 12:12:13 arch ollama[927]: llama_model_loader: - type q8_0: 20 tensors 12:12:13 arch ollama[927]: llama_model_loader: - type q4_K: 181 tensors 12:12:13 arch ollama[927]: llama_model_loader: - type q6_K: 21 tensors 12:12:13 arch ollama[927]: print_info: file format = GGUF V3 (latest) 12:12:13 arch ollama[927]: print_info: file type = Q4_K - Medium 12:12:13 arch ollama[927]: print_info: file size = 5.73 GiB (5.24 BPW) 12:12:13 arch ollama[927]: load: special_eot_id is not in special_eog_ids - the tokenizer config may be incorrect 12:12:13 arch ollama[927]: load: printing all EOG tokens: 12:12:13 arch ollama[927]: load: - 151329 ('<|endoftext|>') 12:12:13 arch ollama[927]: load: - 151336 ('<|user|>') 12:12:13 arch ollama[927]: load: special tokens cache size = 23 12:12:13 arch ollama[927]: load: token to piece cache size = 0.9711 MB 12:12:13 arch ollama[927]: print_info: arch = glm4 12:12:13 arch ollama[927]: print_info: vocab_only = 1 12:12:13 arch ollama[927]: print_info: no_alloc = 0 12:12:13 arch ollama[927]: print_info: model type = ?B 12:12:13 arch ollama[927]: print_info: model params = 9.40 B 12:12:13 arch ollama[927]: print_info: general.name = n/a 12:12:13 arch ollama[927]: print_info: vocab type = BPE 12:12:13 arch ollama[927]: print_info: n_vocab = 151552 12:12:13 arch ollama[927]: print_info: n_merges = 318088 12:12:13 arch ollama[927]: print_info: BOS token = 151329 '<|endoftext|>' 12:12:13 arch ollama[927]: print_info: EOS token = 151329 '<|endoftext|>' 12:12:13 arch ollama[927]: print_info: EOT token = 151336 '<|user|>' 12:12:13 arch ollama[927]: print_info: UNK token = 151329 '<|endoftext|>' 12:12:13 arch ollama[927]: print_info: PAD token = 151329 '<|endoftext|>' 12:12:13 arch ollama[927]: print_info: LF token = 198 'Ċ' 12:12:13 arch ollama[927]: print_info: EOG token = 151329 '<|endoftext|>' 12:12:13 arch ollama[927]: print_info: EOG token = 151336 '<|user|>' 12:12:13 arch ollama[927]: print_info: max token length = 1024 12:12:13 arch ollama[927]: llama_model_load: vocab only - skipping tensors 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.712+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --model /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20af820432e1dd55d22a2b0e728e79223e77a --port 33773" 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.712+08:00 level=INFO source=sched.go:443 msg="system memory" total="62.6 GiB" free="41.9 GiB" free_swap="3.2 GiB" 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.712+08:00 level=INFO source=sched.go:450 msg="gpu memory" id=GPU-8215c551-6dde-569b-490d-884f3ab7a437 library=CUDA available="6.6 GiB" free="7.1 GiB" minimum="457.0 MiB" overhead="0 B" 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.712+08:00 level=INFO source=server.go:496 msg="loading model" "model layers"=41 requested=-1 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.713+08:00 level=INFO source=device.go:240 msg="model weights" device=CUDA0 size="4.2 GiB" 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.713+08:00 level=INFO source=device.go:245 msg="model weights" device=CPU size="1.3 GiB" 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.713+08:00 level=INFO source=device.go:251 msg="kv cache" device=CUDA0 size="544.0 MiB" 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.713+08:00 level=INFO source=device.go:256 msg="kv cache" device=CPU size="96.0 MiB" 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.713+08:00 level=INFO source=device.go:262 msg="compute graph" device=CUDA0 size="1.7 GiB" 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.713+08:00 level=INFO source=device.go:272 msg="total memory" size="7.7 GiB" 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.728+08:00 level=INFO source=runner.go:965 msg="starting go runner" 12:12:13 arch ollama[927]: load_backend: loaded CPU backend from /usr/local/lib/ollama/libggml-cpu-haswell.so 12:12:13 arch ollama[927]: ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no 12:12:13 arch ollama[927]: ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no 12:12:13 arch ollama[927]: ggml_cuda_init: found 1 CUDA devices: 12:12:13 arch ollama[927]: Device 0: NVIDIA GeForce RTX 2060 SUPER, compute capability 7.5, VMM: yes, ID: GPU-8215c551-6dde-569b-490d-884f3ab7a437 12:12:13 arch ollama[927]: load_backend: loaded CUDA backend from /usr/local/lib/ollama/cuda_v13/libggml-cuda.so 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.805+08:00 level=INFO source=ggml.go:104 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.BMI2=1 CPU.0.LLAMAFILE=1 CPU.1.LLAMAFILE=1 CUDA.0.ARCHS=750,800,860,870,890,900,1000,1030,1100,1200,1210 CUDA.0.USE_GRAPHS=1 CUDA.0.PEER_MAX_BATCH_SIZE=128 compiler=cgo(gcc) 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.805+08:00 level=INFO source=runner.go:1001 msg="Server listening on 127.0.0.1:33773" 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.811+08:00 level=INFO source=runner.go:895 msg=load request="{Operation:commit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:Auto KvSize:16384 KvCacheType: NumThreads:12 GPULayers:34[ID:GPU-8215c551-6dde-569b-490d-884f3ab7a437 Layers:34(6..39)] MultiUserCache:false ProjectorPath:/usr/share/ollama/.ollama/models/blobs/sha256-454dd441c925c1a81c984204bb9d54feef0ef07b789c6fe1118099014ba2727d MainGPU:0 UseMmap:true}" 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.811+08:00 level=INFO source=server.go:1338 msg="waiting for llama runner to start responding" 12:12:13 arch ollama[927]: time=2026-01-05T12:12:13.811+08:00 level=INFO source=server.go:1372 msg="waiting for server to become available" status="llm server loading model" 12:12:13 arch ollama[927]: ggml_backend_cuda_device_get_memory device GPU-8215c551-6dde-569b-490d-884f3ab7a437 utilizing NVML memory reporting free: 7585398784 total: 8589934592 12:12:13 arch ollama[927]: llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce RTX 2060 SUPER) (0000:03:00.0) - 7234 MiB free 12:12:13 arch ollama[927]: llama_model_loader: loaded meta data with 33 key-value pairs and 523 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20af820432e1dd55d22a2b0e728e79223e77a (version GGUF V3 (latest)) 12:12:13 arch ollama[927]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. 12:12:13 arch ollama[927]: llama_model_loader: - kv 0: general.architecture str = glm4 12:12:13 arch ollama[927]: llama_model_loader: - kv 1: general.type str = model 12:12:13 arch ollama[927]: llama_model_loader: - kv 2: general.size_label str = 9.4B 12:12:13 arch ollama[927]: llama_model_loader: - kv 3: general.license str = mit 12:12:13 arch ollama[927]: llama_model_loader: - kv 4: general.base_model.count u32 = 1 12:12:13 arch ollama[927]: llama_model_loader: - kv 5: general.base_model.0.name str = GLM 4.1V 9B Base 12:12:13 arch ollama[927]: llama_model_loader: - kv 6: general.base_model.0.organization str = Zai Org 12:12:13 arch ollama[927]: llama_model_loader: - kv 7: general.base_model.0.repo_url str = https://huggingface.co/zai-org/GLM-4.... 12:12:13 arch ollama[927]: llama_model_loader: - kv 8: general.tags arr[str,2] = ["agent", "image-text-to-text"] 12:12:13 arch ollama[927]: llama_model_loader: - kv 9: general.languages arr[str,1] = ["zh"] 12:12:13 arch ollama[927]: llama_model_loader: - kv 10: glm4.block_count u32 = 40 12:12:13 arch ollama[927]: llama_model_loader: - kv 11: glm4.context_length u32 = 65536 12:12:13 arch ollama[927]: llama_model_loader: - kv 12: glm4.embedding_length u32 = 4096 12:12:13 arch ollama[927]: llama_model_loader: - kv 13: glm4.feed_forward_length u32 = 13696 12:12:13 arch ollama[927]: llama_model_loader: - kv 14: glm4.attention.head_count u32 = 32 12:12:13 arch ollama[927]: llama_model_loader: - kv 15: glm4.attention.head_count_kv u32 = 2 12:12:13 arch ollama[927]: llama_model_loader: - kv 16: glm4.rope.dimension_sections arr[i32,4] = [8, 12, 12, 0] 12:12:13 arch ollama[927]: llama_model_loader: - kv 17: glm4.rope.freq_base f32 = 10000.000000 12:12:13 arch ollama[927]: llama_model_loader: - kv 18: glm4.attention.layer_norm_rms_epsilon f32 = 0.000010 12:12:13 arch ollama[927]: llama_model_loader: - kv 19: glm4.rope.dimension_count u32 = 64 12:12:13 arch ollama[927]: llama_model_loader: - kv 20: tokenizer.ggml.model str = gpt2 12:12:13 arch ollama[927]: llama_model_loader: - kv 21: tokenizer.ggml.pre str = glm4 12:12:13 arch ollama[927]: llama_model_loader: - kv 22: tokenizer.ggml.tokens arr[str,151552] = ["!", "\"", "#", "$", "%", "&", "'", ... 12:12:13 arch ollama[927]: llama_model_loader: - kv 23: tokenizer.ggml.token_type arr[i32,151552] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... 12:12:14 arch ollama[927]: llama_model_loader: - kv 24: tokenizer.ggml.merges arr[str,318088] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "... 12:12:14 arch ollama[927]: llama_model_loader: - kv 25: tokenizer.ggml.eos_token_id u32 = 151329 12:12:14 arch ollama[927]: llama_model_loader: - kv 26: tokenizer.ggml.padding_token_id u32 = 151329 12:12:14 arch ollama[927]: llama_model_loader: - kv 27: tokenizer.ggml.eot_token_id u32 = 151336 12:12:14 arch ollama[927]: llama_model_loader: - kv 28: tokenizer.ggml.unknown_token_id u32 = 151329 12:12:14 arch ollama[927]: llama_model_loader: - kv 29: tokenizer.ggml.bos_token_id u32 = 151329 12:12:14 arch ollama[927]: llama_model_loader: - kv 30: tokenizer.chat_template str = [gMASK]<sop>\n{%- for msg in messages ... 12:12:14 arch ollama[927]: llama_model_loader: - kv 31: general.quantization_version u32 = 2 12:12:14 arch ollama[927]: llama_model_loader: - kv 32: general.file_type u32 = 15 12:12:14 arch ollama[927]: llama_model_loader: - type f32: 281 tensors 12:12:14 arch ollama[927]: llama_model_loader: - type q5_0: 20 tensors 12:12:14 arch ollama[927]: llama_model_loader: - type q8_0: 20 tensors 12:12:14 arch ollama[927]: llama_model_loader: - type q4_K: 181 tensors 12:12:14 arch ollama[927]: llama_model_loader: - type q6_K: 21 tensors 12:12:14 arch ollama[927]: print_info: file format = GGUF V3 (latest) 12:12:14 arch ollama[927]: print_info: file type = Q4_K - Medium 12:12:14 arch ollama[927]: print_info: file size = 5.73 GiB (5.24 BPW) 12:12:14 arch ollama[927]: load: special_eot_id is not in special_eog_ids - the tokenizer config may be incorrect 12:12:14 arch ollama[927]: load: printing all EOG tokens: 12:12:14 arch ollama[927]: load: - 151329 ('<|endoftext|>') 12:12:14 arch ollama[927]: load: - 151336 ('<|user|>') 12:12:14 arch ollama[927]: load: special tokens cache size = 23 12:12:14 arch ollama[927]: load: token to piece cache size = 0.9711 MB 12:12:14 arch ollama[927]: print_info: arch = glm4 12:12:14 arch ollama[927]: print_info: vocab_only = 0 12:12:14 arch ollama[927]: print_info: no_alloc = 0 12:12:14 arch ollama[927]: print_info: n_ctx_train = 65536 12:12:14 arch ollama[927]: print_info: n_embd = 4096 12:12:14 arch ollama[927]: print_info: n_embd_inp = 4096 12:12:14 arch ollama[927]: print_info: n_layer = 40 12:12:14 arch ollama[927]: print_info: n_head = 32 12:12:14 arch ollama[927]: print_info: n_head_kv = 2 12:12:14 arch ollama[927]: print_info: n_rot = 64 12:12:14 arch ollama[927]: print_info: n_swa = 0 12:12:14 arch ollama[927]: print_info: is_swa_any = 0 12:12:14 arch ollama[927]: print_info: n_embd_head_k = 128 12:12:14 arch ollama[927]: print_info: n_embd_head_v = 128 12:12:14 arch ollama[927]: print_info: n_gqa = 16 12:12:14 arch ollama[927]: print_info: n_embd_k_gqa = 256 12:12:14 arch ollama[927]: print_info: n_embd_v_gqa = 256 12:12:14 arch ollama[927]: print_info: f_norm_eps = 0.0e+00 12:12:14 arch ollama[927]: print_info: f_norm_rms_eps = 1.0e-05 12:12:14 arch ollama[927]: print_info: f_clamp_kqv = 0.0e+00 12:12:14 arch ollama[927]: print_info: f_max_alibi_bias = 0.0e+00 12:12:14 arch ollama[927]: print_info: f_logit_scale = 0.0e+00 12:12:14 arch ollama[927]: print_info: f_attn_scale = 0.0e+00 12:12:14 arch ollama[927]: print_info: n_ff = 13696 12:12:14 arch ollama[927]: print_info: n_expert = 0 12:12:14 arch ollama[927]: print_info: n_expert_used = 0 12:12:14 arch ollama[927]: print_info: n_expert_groups = 0 12:12:14 arch ollama[927]: print_info: n_group_used = 0 12:12:14 arch ollama[927]: print_info: causal attn = 1 12:12:14 arch ollama[927]: print_info: pooling type = 0 12:12:14 arch ollama[927]: print_info: rope type = 8 12:12:14 arch ollama[927]: print_info: rope scaling = linear 12:12:14 arch ollama[927]: print_info: freq_base_train = 10000.0 12:12:14 arch ollama[927]: print_info: freq_scale_train = 1 12:12:14 arch ollama[927]: print_info: n_ctx_orig_yarn = 65536 12:12:14 arch ollama[927]: print_info: rope_yarn_log_mul= 0.0000 12:12:14 arch ollama[927]: print_info: rope_finetuned = unknown 12:12:14 arch ollama[927]: print_info: mrope sections = [8, 12, 12, 0] 12:12:14 arch ollama[927]: print_info: model type = 9B 12:12:14 arch ollama[927]: print_info: model params = 9.40 B 12:12:14 arch ollama[927]: print_info: general.name = n/a 12:12:14 arch ollama[927]: print_info: vocab type = BPE 12:12:14 arch ollama[927]: print_info: n_vocab = 151552 12:12:14 arch ollama[927]: print_info: n_merges = 318088 12:12:14 arch ollama[927]: print_info: BOS token = 151329 '<|endoftext|>' 12:12:14 arch ollama[927]: print_info: EOS token = 151329 '<|endoftext|>' 12:12:14 arch ollama[927]: print_info: EOT token = 151336 '<|user|>' 12:12:14 arch ollama[927]: print_info: UNK token = 151329 '<|endoftext|>' 12:12:14 arch ollama[927]: print_info: PAD token = 151329 '<|endoftext|>' 12:12:14 arch ollama[927]: print_info: LF token = 198 'Ċ' 12:12:14 arch ollama[927]: print_info: EOG token = 151329 '<|endoftext|>' 12:12:14 arch ollama[927]: print_info: EOG token = 151336 '<|user|>' 12:12:14 arch ollama[927]: print_info: max token length = 1024 12:12:14 arch ollama[927]: load_tensors: loading model tensors, this can take a while... (mmap = true) 12:12:14 arch ollama[927]: load_tensors: offloading 34 repeating layers to GPU 12:12:14 arch ollama[927]: load_tensors: offloaded 34/41 layers to GPU 12:12:14 arch ollama[927]: load_tensors: CPU_Mapped model buffer size = 1617.29 MiB 12:12:14 arch ollama[927]: load_tensors: CUDA0 model buffer size = 4254.72 MiB 12:12:15 arch ollama[927]: llama_context: constructing llama_context 12:12:15 arch ollama[927]: llama_context: n_seq_max = 1 12:12:15 arch ollama[927]: llama_context: n_ctx = 16384 12:12:15 arch ollama[927]: llama_context: n_ctx_seq = 16384 12:12:15 arch ollama[927]: llama_context: n_batch = 512 12:12:15 arch ollama[927]: llama_context: n_ubatch = 512 12:12:15 arch ollama[927]: llama_context: causal_attn = 1 12:12:15 arch ollama[927]: llama_context: flash_attn = auto 12:12:15 arch ollama[927]: llama_context: kv_unified = false 12:12:15 arch ollama[927]: llama_context: freq_base = 10000.0 12:12:15 arch ollama[927]: llama_context: freq_scale = 1 12:12:15 arch ollama[927]: llama_context: n_ctx_seq (16384) < n_ctx_train (65536) -- the full capacity of the model will not be utilized 12:12:15 arch ollama[927]: llama_context: CPU output buffer size = 0.59 MiB 12:12:15 arch ollama[927]: llama_kv_cache: CPU KV buffer size = 96.00 MiB 12:12:15 arch ollama[927]: llama_kv_cache: CUDA0 KV buffer size = 544.00 MiB 12:12:15 arch ollama[927]: llama_kv_cache: size = 640.00 MiB ( 16384 cells, 40 layers, 1/1 seqs), K (f16): 320.00 MiB, V (f16): 320.00 MiB 12:12:15 arch ollama[927]: llama_context: Flash Attention was auto, set to enabled 12:12:15 arch ollama[927]: llama_context: CUDA0 compute buffer size = 789.62 MiB 12:12:15 arch ollama[927]: llama_context: CUDA_Host compute buffer size = 40.02 MiB 12:12:15 arch ollama[927]: llama_context: graph nodes = 1487 12:12:15 arch ollama[927]: llama_context: graph splits = 94 (with bs=512), 3 (with bs=1) 12:12:15 arch ollama[927]: clip_model_loader: model name: 12:12:15 arch ollama[927]: clip_model_loader: description: 12:12:15 arch ollama[927]: clip_model_loader: GGUF version: 3 12:12:15 arch ollama[927]: clip_model_loader: alignment: 32 12:12:15 arch ollama[927]: clip_model_loader: n_tensors: 182 12:12:15 arch ollama[927]: clip_model_loader: n_kv: 25 12:12:15 arch ollama[927]: clip_model_loader: has vision encoder 12:12:15 arch ollama[927]: clip_model_loader: tensor[0]: n_dims = 2, name = v.blk.0.attn_out.weight, tensor_size=2506752, offset=0, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[1]: n_dims = 2, name = v.blk.0.attn_qkv.weight, tensor_size=7520256, offset=2506752, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[2]: n_dims = 2, name = v.blk.0.ffn_down.weight, tensor_size=6684672, offset=10027008, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[3]: n_dims = 2, name = v.blk.0.ffn_gate.weight, tensor_size=6684672, offset=16711680, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[4]: n_dims = 2, name = v.blk.0.ffn_up.weight, tensor_size=6684672, offset=23396352, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[5]: n_dims = 1, name = v.blk.0.ln1.weight, tensor_size=6144, offset=30081024, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[6]: n_dims = 1, name = v.blk.0.ln2.weight, tensor_size=6144, offset=30087168, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[7]: n_dims = 2, name = v.blk.1.attn_out.weight, tensor_size=2506752, offset=30093312, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[8]: n_dims = 2, name = v.blk.1.attn_qkv.weight, tensor_size=7520256, offset=32600064, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[9]: n_dims = 2, name = v.blk.1.ffn_down.weight, tensor_size=6684672, offset=40120320, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[10]: n_dims = 2, name = v.blk.1.ffn_gate.weight, tensor_size=6684672, offset=46804992, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[11]: n_dims = 2, name = v.blk.1.ffn_up.weight, tensor_size=6684672, offset=53489664, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[12]: n_dims = 1, name = v.blk.1.ln1.weight, tensor_size=6144, offset=60174336, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[13]: n_dims = 1, name = v.blk.1.ln2.weight, tensor_size=6144, offset=60180480, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[14]: n_dims = 2, name = v.blk.10.attn_out.weight, tensor_size=2506752, offset=60186624, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[15]: n_dims = 2, name = v.blk.10.attn_qkv.weight, tensor_size=7520256, offset=62693376, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[16]: n_dims = 2, name = v.blk.10.ffn_down.weight, tensor_size=6684672, offset=70213632, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[17]: n_dims = 2, name = v.blk.10.ffn_gate.weight, tensor_size=6684672, offset=76898304, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[18]: n_dims = 2, name = v.blk.10.ffn_up.weight, tensor_size=6684672, offset=83582976, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[19]: n_dims = 1, name = v.blk.10.ln1.weight, tensor_size=6144, offset=90267648, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[20]: n_dims = 1, name = v.blk.10.ln2.weight, tensor_size=6144, offset=90273792, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[21]: n_dims = 2, name = v.blk.11.attn_out.weight, tensor_size=2506752, offset=90279936, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[22]: n_dims = 2, name = v.blk.11.attn_qkv.weight, tensor_size=7520256, offset=92786688, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[23]: n_dims = 2, name = v.blk.11.ffn_down.weight, tensor_size=6684672, offset=100306944, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[24]: n_dims = 2, name = v.blk.11.ffn_gate.weight, tensor_size=6684672, offset=106991616, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[25]: n_dims = 2, name = v.blk.11.ffn_up.weight, tensor_size=6684672, offset=113676288, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[26]: n_dims = 1, name = v.blk.11.ln1.weight, tensor_size=6144, offset=120360960, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[27]: n_dims = 1, name = v.blk.11.ln2.weight, tensor_size=6144, offset=120367104, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[28]: n_dims = 2, name = v.blk.12.attn_out.weight, tensor_size=2506752, offset=120373248, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[29]: n_dims = 2, name = v.blk.12.attn_qkv.weight, tensor_size=7520256, offset=122880000, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[30]: n_dims = 2, name = v.blk.12.ffn_down.weight, tensor_size=6684672, offset=130400256, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[31]: n_dims = 2, name = v.blk.12.ffn_gate.weight, tensor_size=6684672, offset=137084928, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[32]: n_dims = 2, name = v.blk.12.ffn_up.weight, tensor_size=6684672, offset=143769600, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[33]: n_dims = 1, name = v.blk.12.ln1.weight, tensor_size=6144, offset=150454272, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[34]: n_dims = 1, name = v.blk.12.ln2.weight, tensor_size=6144, offset=150460416, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[35]: n_dims = 2, name = v.blk.13.attn_out.weight, tensor_size=2506752, offset=150466560, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[36]: n_dims = 2, name = v.blk.13.attn_qkv.weight, tensor_size=7520256, offset=152973312, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[37]: n_dims = 2, name = v.blk.13.ffn_down.weight, tensor_size=6684672, offset=160493568, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[38]: n_dims = 2, name = v.blk.13.ffn_gate.weight, tensor_size=6684672, offset=167178240, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[39]: n_dims = 2, name = v.blk.13.ffn_up.weight, tensor_size=6684672, offset=173862912, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[40]: n_dims = 1, name = v.blk.13.ln1.weight, tensor_size=6144, offset=180547584, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[41]: n_dims = 1, name = v.blk.13.ln2.weight, tensor_size=6144, offset=180553728, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[42]: n_dims = 2, name = v.blk.14.attn_out.weight, tensor_size=2506752, offset=180559872, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[43]: n_dims = 2, name = v.blk.14.attn_qkv.weight, tensor_size=7520256, offset=183066624, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[44]: n_dims = 2, name = v.blk.14.ffn_down.weight, tensor_size=6684672, offset=190586880, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[45]: n_dims = 2, name = v.blk.14.ffn_gate.weight, tensor_size=6684672, offset=197271552, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[46]: n_dims = 2, name = v.blk.14.ffn_up.weight, tensor_size=6684672, offset=203956224, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[47]: n_dims = 1, name = v.blk.14.ln1.weight, tensor_size=6144, offset=210640896, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[48]: n_dims = 1, name = v.blk.14.ln2.weight, tensor_size=6144, offset=210647040, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[49]: n_dims = 2, name = v.blk.15.attn_out.weight, tensor_size=2506752, offset=210653184, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[50]: n_dims = 2, name = v.blk.15.attn_qkv.weight, tensor_size=7520256, offset=213159936, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[51]: n_dims = 2, name = v.blk.15.ffn_down.weight, tensor_size=6684672, offset=220680192, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[52]: n_dims = 2, name = v.blk.15.ffn_gate.weight, tensor_size=6684672, offset=227364864, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[53]: n_dims = 2, name = v.blk.15.ffn_up.weight, tensor_size=6684672, offset=234049536, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[54]: n_dims = 1, name = v.blk.15.ln1.weight, tensor_size=6144, offset=240734208, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[55]: n_dims = 1, name = v.blk.15.ln2.weight, tensor_size=6144, offset=240740352, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[56]: n_dims = 2, name = v.blk.16.attn_out.weight, tensor_size=2506752, offset=240746496, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[57]: n_dims = 2, name = v.blk.16.attn_qkv.weight, tensor_size=7520256, offset=243253248, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[58]: n_dims = 2, name = v.blk.16.ffn_down.weight, tensor_size=6684672, offset=250773504, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[59]: n_dims = 2, name = v.blk.16.ffn_gate.weight, tensor_size=6684672, offset=257458176, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[60]: n_dims = 2, name = v.blk.16.ffn_up.weight, tensor_size=6684672, offset=264142848, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[61]: n_dims = 1, name = v.blk.16.ln1.weight, tensor_size=6144, offset=270827520, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[62]: n_dims = 1, name = v.blk.16.ln2.weight, tensor_size=6144, offset=270833664, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[63]: n_dims = 2, name = v.blk.17.attn_out.weight, tensor_size=2506752, offset=270839808, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[64]: n_dims = 2, name = v.blk.17.attn_qkv.weight, tensor_size=7520256, offset=273346560, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[65]: n_dims = 2, name = v.blk.17.ffn_down.weight, tensor_size=6684672, offset=280866816, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[66]: n_dims = 2, name = v.blk.17.ffn_gate.weight, tensor_size=6684672, offset=287551488, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[67]: n_dims = 2, name = v.blk.17.ffn_up.weight, tensor_size=6684672, offset=294236160, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[68]: n_dims = 1, name = v.blk.17.ln1.weight, tensor_size=6144, offset=300920832, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[69]: n_dims = 1, name = v.blk.17.ln2.weight, tensor_size=6144, offset=300926976, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[70]: n_dims = 2, name = v.blk.18.attn_out.weight, tensor_size=2506752, offset=300933120, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[71]: n_dims = 2, name = v.blk.18.attn_qkv.weight, tensor_size=7520256, offset=303439872, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[72]: n_dims = 2, name = v.blk.18.ffn_down.weight, tensor_size=6684672, offset=310960128, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[73]: n_dims = 2, name = v.blk.18.ffn_gate.weight, tensor_size=6684672, offset=317644800, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[74]: n_dims = 2, name = v.blk.18.ffn_up.weight, tensor_size=6684672, offset=324329472, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[75]: n_dims = 1, name = v.blk.18.ln1.weight, tensor_size=6144, offset=331014144, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[76]: n_dims = 1, name = v.blk.18.ln2.weight, tensor_size=6144, offset=331020288, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[77]: n_dims = 2, name = v.blk.19.attn_out.weight, tensor_size=2506752, offset=331026432, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[78]: n_dims = 2, name = v.blk.19.attn_qkv.weight, tensor_size=7520256, offset=333533184, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[79]: n_dims = 2, name = v.blk.19.ffn_down.weight, tensor_size=6684672, offset=341053440, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[80]: n_dims = 2, name = v.blk.19.ffn_gate.weight, tensor_size=6684672, offset=347738112, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[81]: n_dims = 2, name = v.blk.19.ffn_up.weight, tensor_size=6684672, offset=354422784, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[82]: n_dims = 1, name = v.blk.19.ln1.weight, tensor_size=6144, offset=361107456, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[83]: n_dims = 1, name = v.blk.19.ln2.weight, tensor_size=6144, offset=361113600, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[84]: n_dims = 2, name = v.blk.2.attn_out.weight, tensor_size=2506752, offset=361119744, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[85]: n_dims = 2, name = v.blk.2.attn_qkv.weight, tensor_size=7520256, offset=363626496, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[86]: n_dims = 2, name = v.blk.2.ffn_down.weight, tensor_size=6684672, offset=371146752, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[87]: n_dims = 2, name = v.blk.2.ffn_gate.weight, tensor_size=6684672, offset=377831424, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[88]: n_dims = 2, name = v.blk.2.ffn_up.weight, tensor_size=6684672, offset=384516096, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[89]: n_dims = 1, name = v.blk.2.ln1.weight, tensor_size=6144, offset=391200768, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[90]: n_dims = 1, name = v.blk.2.ln2.weight, tensor_size=6144, offset=391206912, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[91]: n_dims = 2, name = v.blk.20.attn_out.weight, tensor_size=2506752, offset=391213056, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[92]: n_dims = 2, name = v.blk.20.attn_qkv.weight, tensor_size=7520256, offset=393719808, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[93]: n_dims = 2, name = v.blk.20.ffn_down.weight, tensor_size=6684672, offset=401240064, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[94]: n_dims = 2, name = v.blk.20.ffn_gate.weight, tensor_size=6684672, offset=407924736, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[95]: n_dims = 2, name = v.blk.20.ffn_up.weight, tensor_size=6684672, offset=414609408, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[96]: n_dims = 1, name = v.blk.20.ln1.weight, tensor_size=6144, offset=421294080, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[97]: n_dims = 1, name = v.blk.20.ln2.weight, tensor_size=6144, offset=421300224, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[98]: n_dims = 2, name = v.blk.21.attn_out.weight, tensor_size=2506752, offset=421306368, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[99]: n_dims = 2, name = v.blk.21.attn_qkv.weight, tensor_size=7520256, offset=423813120, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[100]: n_dims = 2, name = v.blk.21.ffn_down.weight, tensor_size=6684672, offset=431333376, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[101]: n_dims = 2, name = v.blk.21.ffn_gate.weight, tensor_size=6684672, offset=438018048, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[102]: n_dims = 2, name = v.blk.21.ffn_up.weight, tensor_size=6684672, offset=444702720, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[103]: n_dims = 1, name = v.blk.21.ln1.weight, tensor_size=6144, offset=451387392, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[104]: n_dims = 1, name = v.blk.21.ln2.weight, tensor_size=6144, offset=451393536, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[105]: n_dims = 2, name = v.blk.22.attn_out.weight, tensor_size=2506752, offset=451399680, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[106]: n_dims = 2, name = v.blk.22.attn_qkv.weight, tensor_size=7520256, offset=453906432, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[107]: n_dims = 2, name = v.blk.22.ffn_down.weight, tensor_size=6684672, offset=461426688, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[108]: n_dims = 2, name = v.blk.22.ffn_gate.weight, tensor_size=6684672, offset=468111360, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[109]: n_dims = 2, name = v.blk.22.ffn_up.weight, tensor_size=6684672, offset=474796032, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[110]: n_dims = 1, name = v.blk.22.ln1.weight, tensor_size=6144, offset=481480704, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[111]: n_dims = 1, name = v.blk.22.ln2.weight, tensor_size=6144, offset=481486848, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[112]: n_dims = 2, name = v.blk.23.attn_out.weight, tensor_size=2506752, offset=481492992, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[113]: n_dims = 2, name = v.blk.23.attn_qkv.weight, tensor_size=7520256, offset=483999744, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[114]: n_dims = 2, name = v.blk.23.ffn_down.weight, tensor_size=6684672, offset=491520000, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[115]: n_dims = 2, name = v.blk.23.ffn_gate.weight, tensor_size=6684672, offset=498204672, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[116]: n_dims = 2, name = v.blk.23.ffn_up.weight, tensor_size=6684672, offset=504889344, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[117]: n_dims = 1, name = v.blk.23.ln1.weight, tensor_size=6144, offset=511574016, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[118]: n_dims = 1, name = v.blk.23.ln2.weight, tensor_size=6144, offset=511580160, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[119]: n_dims = 2, name = v.blk.3.attn_out.weight, tensor_size=2506752, offset=511586304, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[120]: n_dims = 2, name = v.blk.3.attn_qkv.weight, tensor_size=7520256, offset=514093056, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[121]: n_dims = 2, name = v.blk.3.ffn_down.weight, tensor_size=6684672, offset=521613312, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[122]: n_dims = 2, name = v.blk.3.ffn_gate.weight, tensor_size=6684672, offset=528297984, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[123]: n_dims = 2, name = v.blk.3.ffn_up.weight, tensor_size=6684672, offset=534982656, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[124]: n_dims = 1, name = v.blk.3.ln1.weight, tensor_size=6144, offset=541667328, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[125]: n_dims = 1, name = v.blk.3.ln2.weight, tensor_size=6144, offset=541673472, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[126]: n_dims = 2, name = v.blk.4.attn_out.weight, tensor_size=2506752, offset=541679616, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[127]: n_dims = 2, name = v.blk.4.attn_qkv.weight, tensor_size=7520256, offset=544186368, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[128]: n_dims = 2, name = v.blk.4.ffn_down.weight, tensor_size=6684672, offset=551706624, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[129]: n_dims = 2, name = v.blk.4.ffn_gate.weight, tensor_size=6684672, offset=558391296, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[130]: n_dims = 2, name = v.blk.4.ffn_up.weight, tensor_size=6684672, offset=565075968, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[131]: n_dims = 1, name = v.blk.4.ln1.weight, tensor_size=6144, offset=571760640, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[132]: n_dims = 1, name = v.blk.4.ln2.weight, tensor_size=6144, offset=571766784, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[133]: n_dims = 2, name = v.blk.5.attn_out.weight, tensor_size=2506752, offset=571772928, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[134]: n_dims = 2, name = v.blk.5.attn_qkv.weight, tensor_size=7520256, offset=574279680, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[135]: n_dims = 2, name = v.blk.5.ffn_down.weight, tensor_size=6684672, offset=581799936, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[136]: n_dims = 2, name = v.blk.5.ffn_gate.weight, tensor_size=6684672, offset=588484608, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[137]: n_dims = 2, name = v.blk.5.ffn_up.weight, tensor_size=6684672, offset=595169280, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[138]: n_dims = 1, name = v.blk.5.ln1.weight, tensor_size=6144, offset=601853952, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[139]: n_dims = 1, name = v.blk.5.ln2.weight, tensor_size=6144, offset=601860096, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[140]: n_dims = 2, name = v.blk.6.attn_out.weight, tensor_size=2506752, offset=601866240, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[141]: n_dims = 2, name = v.blk.6.attn_qkv.weight, tensor_size=7520256, offset=604372992, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[142]: n_dims = 2, name = v.blk.6.ffn_down.weight, tensor_size=6684672, offset=611893248, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[143]: n_dims = 2, name = v.blk.6.ffn_gate.weight, tensor_size=6684672, offset=618577920, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[144]: n_dims = 2, name = v.blk.6.ffn_up.weight, tensor_size=6684672, offset=625262592, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[145]: n_dims = 1, name = v.blk.6.ln1.weight, tensor_size=6144, offset=631947264, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[146]: n_dims = 1, name = v.blk.6.ln2.weight, tensor_size=6144, offset=631953408, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[147]: n_dims = 2, name = v.blk.7.attn_out.weight, tensor_size=2506752, offset=631959552, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[148]: n_dims = 2, name = v.blk.7.attn_qkv.weight, tensor_size=7520256, offset=634466304, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[149]: n_dims = 2, name = v.blk.7.ffn_down.weight, tensor_size=6684672, offset=641986560, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[150]: n_dims = 2, name = v.blk.7.ffn_gate.weight, tensor_size=6684672, offset=648671232, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[151]: n_dims = 2, name = v.blk.7.ffn_up.weight, tensor_size=6684672, offset=655355904, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[152]: n_dims = 1, name = v.blk.7.ln1.weight, tensor_size=6144, offset=662040576, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[153]: n_dims = 1, name = v.blk.7.ln2.weight, tensor_size=6144, offset=662046720, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[154]: n_dims = 2, name = v.blk.8.attn_out.weight, tensor_size=2506752, offset=662052864, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[155]: n_dims = 2, name = v.blk.8.attn_qkv.weight, tensor_size=7520256, offset=664559616, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[156]: n_dims = 2, name = v.blk.8.ffn_down.weight, tensor_size=6684672, offset=672079872, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[157]: n_dims = 2, name = v.blk.8.ffn_gate.weight, tensor_size=6684672, offset=678764544, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[158]: n_dims = 2, name = v.blk.8.ffn_up.weight, tensor_size=6684672, offset=685449216, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[159]: n_dims = 1, name = v.blk.8.ln1.weight, tensor_size=6144, offset=692133888, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[160]: n_dims = 1, name = v.blk.8.ln2.weight, tensor_size=6144, offset=692140032, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[161]: n_dims = 2, name = v.blk.9.attn_out.weight, tensor_size=2506752, offset=692146176, shape:[1536, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[162]: n_dims = 2, name = v.blk.9.attn_qkv.weight, tensor_size=7520256, offset=694652928, shape:[1536, 4608, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[163]: n_dims = 2, name = v.blk.9.ffn_down.weight, tensor_size=6684672, offset=702173184, shape:[4096, 1536, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[164]: n_dims = 2, name = v.blk.9.ffn_gate.weight, tensor_size=6684672, offset=708857856, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[165]: n_dims = 2, name = v.blk.9.ffn_up.weight, tensor_size=6684672, offset=715542528, shape:[1536, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[166]: n_dims = 1, name = v.blk.9.ln1.weight, tensor_size=6144, offset=722227200, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[167]: n_dims = 1, name = v.blk.9.ln2.weight, tensor_size=6144, offset=722233344, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[168]: n_dims = 1, name = mm.patch_merger.bias, tensor_size=16384, offset=722239488, shape:[4096, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[169]: n_dims = 4, name = mm.patch_merger.weight, tensor_size=50331648, offset=722255872, shape:[2, 2, 1536, 4096], type = f16 12:12:15 arch ollama[927]: clip_model_loader: tensor[170]: n_dims = 2, name = v.position_embd.weight, tensor_size=3538944, offset=772587520, shape:[1536, 576, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[171]: n_dims = 2, name = mm.down.weight, tensor_size=59604992, offset=776126464, shape:[13696, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[172]: n_dims = 2, name = mm.gate.weight, tensor_size=59604992, offset=835731456, shape:[4096, 13696, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[173]: n_dims = 1, name = mm.post_norm.bias, tensor_size=16384, offset=895336448, shape:[4096, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[174]: n_dims = 1, name = mm.post_norm.weight, tensor_size=16384, offset=895352832, shape:[4096, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[175]: n_dims = 2, name = mm.model.fc.weight, tensor_size=17825792, offset=895369216, shape:[4096, 4096, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[176]: n_dims = 2, name = mm.up.weight, tensor_size=59604992, offset=913195008, shape:[4096, 13696, 1, 1], type = q8_0 12:12:15 arch ollama[927]: clip_model_loader: tensor[177]: n_dims = 1, name = v.patch_embd.bias, tensor_size=6144, offset=972800000, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[178]: n_dims = 4, name = v.patch_embd.weight, tensor_size=3612672, offset=972806144, shape:[14, 14, 3, 1536], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[179]: n_dims = 4, name = v.patch_embd.weight.1, tensor_size=3612672, offset=976418816, shape:[14, 14, 3, 1536], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[180]: n_dims = 1, name = v.norm_embd.weight, tensor_size=6144, offset=980031488, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_model_loader: tensor[181]: n_dims = 1, name = v.post_ln.weight, tensor_size=6144, offset=980037632, shape:[1536, 1, 1, 1], type = f32 12:12:15 arch ollama[927]: clip_ctx: CLIP using CUDA0 backend 12:12:15 arch ollama[927]: load_hparams: projector: glm4v 12:12:15 arch ollama[927]: load_hparams: n_embd: 1536 12:12:15 arch ollama[927]: load_hparams: n_head: 12 12:12:15 arch ollama[927]: load_hparams: n_ff: 13696 12:12:15 arch ollama[927]: load_hparams: n_layer: 24 12:12:15 arch ollama[927]: load_hparams: ffn_op: silu 12:12:15 arch ollama[927]: load_hparams: projection_dim: 4096 12:12:15 arch ollama[927]: --- vision hparams --- 12:12:15 arch ollama[927]: load_hparams: image_size: 336 12:12:15 arch ollama[927]: load_hparams: patch_size: 14 12:12:15 arch ollama[927]: load_hparams: has_llava_proj: 0 12:12:15 arch ollama[927]: load_hparams: minicpmv_version: 0 12:12:15 arch ollama[927]: load_hparams: n_merge: 2 12:12:15 arch ollama[927]: load_hparams: n_wa_pattern: 0 12:12:15 arch ollama[927]: load_hparams: image_min_pixels: 6272 12:12:15 arch ollama[927]: load_hparams: image_max_pixels: 3211264 12:12:15 arch ollama[927]: load_hparams: model size: 934.64 MiB 12:12:15 arch ollama[927]: load_hparams: metadata size: 0.06 MiB 12:12:16 arch ollama[927]: load_tensors: loaded 182 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-454dd441c925c1a81c984204bb9d54feef0ef07b789c6fe1118099014ba2727d 12:12:16 arch ollama[927]: warmup: warmup with image size = 1288 x 1288 12:12:16 arch ollama[927]: ggml_backend_cuda_buffer_type_alloc_buffer: allocating 515.05 MiB on device 0: cudaMalloc failed: out of memory 12:12:16 arch ollama[927]: ggml_gallocr_reserve_n_impl: failed to allocate CUDA0 buffer of size 540070912 12:12:16 arch ollama[927]: alloc_compute_meta: CPU compute buffer size = 19.11 MiB 12:12:16 arch ollama[927]: alloc_compute_meta: graph splits = 1, nodes = 632 12:12:16 arch ollama[927]: warmup: flash attention is enabled 12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.324+08:00 level=INFO source=server.go:1376 msg="llama runner started in 2.61 seconds" 12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.324+08:00 level=INFO source=sched.go:517 msg="loaded runners" count=1 12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.324+08:00 level=INFO source=server.go:1338 msg="waiting for llama runner to start responding" 12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.325+08:00 level=INFO source=server.go:1376 msg="llama runner started in 2.61 seconds" 12:12:16 arch ollama[927]: add_text: <|begin_of_image|> 12:12:16 arch ollama[927]: image_tokens->nx = 85 12:12:16 arch ollama[927]: image_tokens->ny = 48 12:12:16 arch ollama[927]: batch_f32 size = 1 12:12:16 arch ollama[927]: add_text: <|end_of_image|> 12:12:16 arch ollama[927]: ggml_backend_cuda_buffer_type_alloc_buffer: allocating 993.11 MiB on device 0: cudaMalloc failed: out of memory 12:12:16 arch ollama[927]: ggml_gallocr_reserve_n_impl: failed to allocate CUDA0 buffer of size 1041346560 12:12:16 arch ollama[927]: SIGSEGV: segmentation violation 12:12:16 arch ollama[927]: PC=0x564e038541cb m=15 sigcode=1 addr=0x0 12:12:16 arch ollama[927]: signal arrived during cgo execution 12:12:16 arch ollama[927]: goroutine 40 gp=0xc000505c00 m=15 mp=0xc000101008 [syscall]: 12:12:16 arch ollama[927]: runtime.cgocall(0x564e03841190, 0xc0000491d8) 12:12:16 arch ollama[927]: runtime/cgocall.go:167 +0x4b fp=0xc0000491b0 sp=0xc000049178 pc=0x564e02aef6eb 12:12:16 arch ollama[927]: github.com/ollama/ollama/llama._Cfunc_mtmd_encode_chunk(0x7f834008dd90, 0x7f82817bc000) 12:12:16 arch ollama[927]: _cgo_gotypes.go:1079 +0x4a fp=0xc0000491d8 sp=0xc0000491b0 pc=0x564e02eaa04a 12:12:16 arch ollama[927]: github.com/ollama/ollama/llama.(*MtmdContext).MultimodalTokenize.func11(...) 12:12:16 arch ollama[927]: github.com/ollama/ollama/llama/llama.go:595 12:12:16 arch ollama[927]: github.com/ollama/ollama/llama.(*MtmdContext).MultimodalTokenize(0xc00034c0b8, 0xc0003a6750, {0xc000a82000, 0x117bcb, 0xc000049520?}) 12:12:16 arch ollama[927]: github.com/ollama/ollama/llama/llama.go:595 +0x6c5 fp=0xc000049490 sp=0xc0000491d8 pc=0x564e02eaf085 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*ImageContext).MultimodalTokenize(0xc00063c240, 0xc0003a6750, {0xc000a82000, 0x117bcb, 0x117bcd}) 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/image.go:76 +0x145 fp=0xc000049530 sp=0xc000049490 pc=0x564e02f61365 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).inputs(0xc00013f900, {0xc000042440?, 0x3d?}, {0xc00061c080, 0x1, 0x160?}) 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:236 +0x2c6 fp=0xc000049698 sp=0xc000049530 pc=0x564e02f62766 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).NewSequence(0xc00013f900, {0xc000042440, 0x3d}, {0xc00061c080, 0x1, 0x1}, {0x28000, {0xc000044610, 0x1, 0x1}, ...}) 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:126 +0x8d fp=0xc000049838 sp=0xc000049698 pc=0x564e02f61d2d 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).completion(0xc00013f900, {0x564e040a5fa0, 0xc0003c21c0}, 0xc0003b2280) 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:659 +0x5f9 fp=0xc000049ac0 sp=0xc000049838 pc=0x564e02f64d99 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).completion-fm({0x564e040a5fa0?, 0xc0003c21c0?}, 0xc000049b40?) 12:12:16 arch ollama[927]: <autogenerated>:1 +0x36 fp=0xc000049af0 sp=0xc000049ac0 pc=0x564e02f687f6 12:12:16 arch ollama[927]: net/http.HandlerFunc.ServeHTTP(0xc000544180?, {0x564e040a5fa0?, 0xc0003c21c0?}, 0xc000049b60?) 12:12:16 arch ollama[927]: net/http/server.go:2294 +0x29 fp=0xc000049b18 sp=0xc000049af0 pc=0x564e02df20e9 12:12:16 arch ollama[927]: net/http.(*ServeMux).ServeHTTP(0x564e02a978c5?, {0x564e040a5fa0, 0xc0003c21c0}, 0xc0003b2280) 12:12:16 arch ollama[927]: net/http/server.go:2822 +0x1c4 fp=0xc000049b68 sp=0xc000049b18 pc=0x564e02df3fe4 12:12:16 arch ollama[927]: net/http.serverHandler.ServeHTTP({0x564e040a2590?}, {0x564e040a5fa0?, 0xc0003c21c0?}, 0x1?) 12:12:16 arch ollama[927]: net/http/server.go:3301 +0x8e fp=0xc000049b98 sp=0xc000049b68 pc=0x564e02e11a6e 12:12:16 arch ollama[927]: net/http.(*conn).serve(0xc0001463f0, {0x564e040a83d8, 0xc000144b10}) 12:12:16 arch ollama[927]: net/http/server.go:2102 +0x625 fp=0xc000049fb8 sp=0xc000049b98 pc=0x564e02df05e5 12:12:16 arch ollama[927]: net/http.(*Server).Serve.gowrap3() 12:12:16 arch ollama[927]: net/http/server.go:3454 +0x28 fp=0xc000049fe0 sp=0xc000049fb8 pc=0x564e02df5ea8 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000049fe8 sp=0xc000049fe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by net/http.(*Server).Serve in goroutine 1 12:12:16 arch ollama[927]: net/http/server.go:3454 +0x485 12:12:16 arch ollama[927]: goroutine 1 gp=0xc000002380 m=nil [IO wait]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00018f790 sp=0xc00018f770 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.netpollblock(0xc0005197e0?, 0x2a8c2a6?, 0x4e?) 12:12:16 arch ollama[927]: runtime/netpoll.go:575 +0xf7 fp=0xc00018f7c8 sp=0xc00018f790 pc=0x564e02ab7e97 12:12:16 arch ollama[927]: internal/poll.runtime_pollWait(0x7f83f643aeb0, 0x72) 12:12:16 arch ollama[927]: runtime/netpoll.go:351 +0x85 fp=0xc00018f7e8 sp=0xc00018f7c8 pc=0x564e02af1d85 12:12:16 arch ollama[927]: internal/poll.(*pollDesc).wait(0xc000701480?, 0x900000036?, 0x0) 12:12:16 arch ollama[927]: internal/poll/fd_poll_runtime.go:84 +0x27 fp=0xc00018f810 sp=0xc00018f7e8 pc=0x564e02b79f07 12:12:16 arch ollama[927]: internal/poll.(*pollDesc).waitRead(...) 12:12:16 arch ollama[927]: internal/poll/fd_poll_runtime.go:89 12:12:16 arch ollama[927]: internal/poll.(*FD).Accept(0xc000701480) 12:12:16 arch ollama[927]: internal/poll/fd_unix.go:620 +0x295 fp=0xc00018f8b8 sp=0xc00018f810 pc=0x564e02b7f2d5 12:12:16 arch ollama[927]: net.(*netFD).accept(0xc000701480) 12:12:16 arch ollama[927]: net/fd_unix.go:172 +0x29 fp=0xc00018f970 sp=0xc00018f8b8 pc=0x564e02bf21a9 12:12:16 arch ollama[927]: net.(*TCPListener).accept(0xc000531200) 12:12:16 arch ollama[927]: net/tcpsock_posix.go:159 +0x1b fp=0xc00018f9c0 sp=0xc00018f970 pc=0x564e02c07b5b 12:12:16 arch ollama[927]: net.(*TCPListener).Accept(0xc000531200) 12:12:16 arch ollama[927]: net/tcpsock.go:380 +0x30 fp=0xc00018f9f0 sp=0xc00018f9c0 pc=0x564e02c06a10 12:12:16 arch ollama[927]: net/http.(*onceCloseListener).Accept(0xc0001463f0?) 12:12:16 arch ollama[927]: <autogenerated>:1 +0x24 fp=0xc00018fa08 sp=0xc00018f9f0 pc=0x564e02e1e1e4 12:12:16 arch ollama[927]: net/http.(*Server).Serve(0xc000213700, {0x564e040a5dc0, 0xc000531200}) 12:12:16 arch ollama[927]: net/http/server.go:3424 +0x30c fp=0xc00018fb38 sp=0xc00018fa08 pc=0x564e02df5aac 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.Execute({0xc000034260, 0x4, 0x4}) 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:1002 +0x8f5 fp=0xc00018fd08 sp=0xc00018fb38 pc=0x564e02f68175 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner.Execute({0xc000034250?, 0x0?, 0x0?}) 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/runner.go:22 +0xd4 fp=0xc00018fd30 sp=0xc00018fd08 pc=0x564e03013cf4 12:12:16 arch ollama[927]: github.com/ollama/ollama/cmd.NewCLI.func2(0xc000213400?, {0x564e03b880ad?, 0x4?, 0x564e03b880b1?}) 12:12:16 arch ollama[927]: github.com/ollama/ollama/cmd/cmd.go:1841 +0x45 fp=0xc00018fd58 sp=0xc00018fd30 pc=0x564e037d0f25 12:12:16 arch ollama[927]: github.com/spf13/cobra.(*Command).execute(0xc000149508, {0xc000531000, 0x4, 0x4}) 12:12:16 arch ollama[927]: github.com/spf13/cobra@v1.7.0/command.go:940 +0x85c fp=0xc00018fe78 sp=0xc00018fd58 pc=0x564e02c6b7fc 12:12:16 arch ollama[927]: github.com/spf13/cobra.(*Command).ExecuteC(0xc000126908) 12:12:16 arch ollama[927]: github.com/spf13/cobra@v1.7.0/command.go:1068 +0x3a5 fp=0xc00018ff30 sp=0xc00018fe78 pc=0x564e02c6c045 12:12:16 arch ollama[927]: github.com/spf13/cobra.(*Command).Execute(...) 12:12:16 arch ollama[927]: github.com/spf13/cobra@v1.7.0/command.go:992 12:12:16 arch ollama[927]: github.com/spf13/cobra.(*Command).ExecuteContext(...) 12:12:16 arch ollama[927]: github.com/spf13/cobra@v1.7.0/command.go:985 12:12:16 arch ollama[927]: main.main() 12:12:16 arch ollama[927]: github.com/ollama/ollama/main.go:12 +0x4d fp=0xc00018ff50 sp=0xc00018ff30 pc=0x564e037d1a0d 12:12:16 arch ollama[927]: runtime.main() 12:12:16 arch ollama[927]: runtime/proc.go:283 +0x29d fp=0xc00018ffe0 sp=0xc00018ff50 pc=0x564e02abf51d 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00018ffe8 sp=0xc00018ffe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: goroutine 2 gp=0xc000002e00 m=nil [force gc (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008efa8 sp=0xc00008ef88 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.goparkunlock(...) 12:12:16 arch ollama[927]: runtime/proc.go:441 12:12:16 arch ollama[927]: runtime.forcegchelper() 12:12:16 arch ollama[927]: runtime/proc.go:348 +0xb8 fp=0xc00008efe0 sp=0xc00008efa8 pc=0x564e02abf858 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008efe8 sp=0xc00008efe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.init.7 in goroutine 1 12:12:16 arch ollama[927]: runtime/proc.go:336 +0x1a 12:12:16 arch ollama[927]: goroutine 3 gp=0xc000003340 m=nil [GC sweep wait]: 12:12:16 arch ollama[927]: runtime.gopark(0x1?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008f780 sp=0xc00008f760 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.goparkunlock(...) 12:12:16 arch ollama[927]: runtime/proc.go:441 12:12:16 arch ollama[927]: runtime.bgsweep(0xc0000ba000) 12:12:16 arch ollama[927]: runtime/mgcsweep.go:316 +0xdf fp=0xc00008f7c8 sp=0xc00008f780 pc=0x564e02aa9fff 12:12:16 arch ollama[927]: runtime.gcenable.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:204 +0x25 fp=0xc00008f7e0 sp=0xc00008f7c8 pc=0x564e02a9e3e5 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008f7e8 sp=0xc00008f7e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcenable in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:204 +0x66 12:12:16 arch ollama[927]: goroutine 4 gp=0xc000003500 m=nil [GC scavenge wait]: 12:12:16 arch ollama[927]: runtime.gopark(0x10000?, 0x564e03d592a0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008ff78 sp=0xc00008ff58 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.goparkunlock(...) 12:12:16 arch ollama[927]: runtime/proc.go:441 12:12:16 arch ollama[927]: runtime.(*scavengerState).park(0x564e0497c280) 12:12:16 arch ollama[927]: runtime/mgcscavenge.go:425 +0x49 fp=0xc00008ffa8 sp=0xc00008ff78 pc=0x564e02aa7a49 12:12:16 arch ollama[927]: runtime.bgscavenge(0xc0000ba000) 12:12:16 arch ollama[927]: runtime/mgcscavenge.go:658 +0x59 fp=0xc00008ffc8 sp=0xc00008ffa8 pc=0x564e02aa7fd9 12:12:16 arch ollama[927]: runtime.gcenable.gowrap2() 12:12:16 arch ollama[927]: runtime/mgc.go:205 +0x25 fp=0xc00008ffe0 sp=0xc00008ffc8 pc=0x564e02a9e385 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008ffe8 sp=0xc00008ffe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcenable in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:205 +0xa5 12:12:16 arch ollama[927]: goroutine 5 gp=0xc000003dc0 m=nil [finalizer wait]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x564e04092250?, 0x40?, 0x61?, 0x1000000010?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008e630 sp=0xc00008e610 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.runfinq() 12:12:16 arch ollama[927]: runtime/mfinal.go:196 +0x107 fp=0xc00008e7e0 sp=0xc00008e630 pc=0x564e02a9d3a7 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008e7e8 sp=0xc00008e7e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.createfing in goroutine 1 12:12:16 arch ollama[927]: runtime/mfinal.go:166 +0x3d 12:12:16 arch ollama[927]: goroutine 6 gp=0xc0001f08c0 m=nil [chan receive]: 12:12:16 arch ollama[927]: runtime.gopark(0xc000245720?, 0xc000590018?, 0x60?, 0x7?, 0x564e02bd8de8?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000090718 sp=0xc0000906f8 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.chanrecv(0xc0000c6310, 0x0, 0x1) 12:12:16 arch ollama[927]: runtime/chan.go:664 +0x445 fp=0xc000090790 sp=0xc000090718 pc=0x564e02a8ee85 12:12:16 arch ollama[927]: runtime.chanrecv1(0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/chan.go:506 +0x12 fp=0xc0000907b8 sp=0xc000090790 pc=0x564e02a8ea12 12:12:16 arch ollama[927]: runtime.unique_runtime_registerUniqueMapCleanup.func2(...) 12:12:16 arch ollama[927]: runtime/mgc.go:1796 12:12:16 arch ollama[927]: runtime.unique_runtime_registerUniqueMapCleanup.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1799 +0x2f fp=0xc0000907e0 sp=0xc0000907b8 pc=0x564e02aa158f 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0000907e8 sp=0xc0000907e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by unique.runtime_registerUniqueMapCleanup in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1794 +0x85 12:12:16 arch ollama[927]: goroutine 7 gp=0xc0001f0c40 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000090f38 sp=0xc000090f18 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000090fc8 sp=0xc000090f38 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000090fe0 sp=0xc000090fc8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000090fe8 sp=0xc000090fe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 18 gp=0xc000504000 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008a738 sp=0xc00008a718 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008a7c8 sp=0xc00008a738 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008a7e0 sp=0xc00008a7c8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008a7e8 sp=0xc00008a7e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 19 gp=0xc0005041c0 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008af38 sp=0xc00008af18 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008afc8 sp=0xc00008af38 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008afe0 sp=0xc00008afc8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008afe8 sp=0xc00008afe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 34 gp=0xc000102380 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00011a738 sp=0xc00011a718 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00011a7c8 sp=0xc00011a738 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00011a7e0 sp=0xc00011a7c8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00011a7e8 sp=0xc00011a7e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 35 gp=0xc000102540 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00011af38 sp=0xc00011af18 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00011afc8 sp=0xc00011af38 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00011afe0 sp=0xc00011afc8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00011afe8 sp=0xc00011afe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 20 gp=0xc000504380 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008b738 sp=0xc00008b718 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008b7c8 sp=0xc00008b738 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008b7e0 sp=0xc00008b7c8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008b7e8 sp=0xc00008b7e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 21 gp=0xc000504540 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008bf38 sp=0xc00008bf18 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008bfc8 sp=0xc00008bf38 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008bfe0 sp=0xc00008bfc8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008bfe8 sp=0xc00008bfe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 22 gp=0xc000504700 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008c738 sp=0xc00008c718 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008c7c8 sp=0xc00008c738 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008c7e0 sp=0xc00008c7c8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008c7e8 sp=0xc00008c7e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 23 gp=0xc0005048c0 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008cf38 sp=0xc00008cf18 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008cfc8 sp=0xc00008cf38 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008cfe0 sp=0xc00008cfc8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008cfe8 sp=0xc00008cfe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 24 gp=0xc000504a80 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008d738 sp=0xc00008d718 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008d7c8 sp=0xc00008d738 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008d7e0 sp=0xc00008d7c8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008d7e8 sp=0xc00008d7e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 25 gp=0xc000504c40 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008df38 sp=0xc00008df18 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008dfc8 sp=0xc00008df38 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008dfe0 sp=0xc00008dfc8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008dfe8 sp=0xc00008dfe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 8 gp=0xc0001f0e00 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000091738 sp=0xc000091718 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0000917c8 sp=0xc000091738 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0000917e0 sp=0xc0000917c8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0000917e8 sp=0xc0000917e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 36 gp=0xc000102700 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00011b738 sp=0xc00011b718 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00011b7c8 sp=0xc00011b738 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00011b7e0 sp=0xc00011b7c8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00011b7e8 sp=0xc00011b7e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 37 gp=0xc0001028c0 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00011bf38 sp=0xc00011bf18 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00011bfc8 sp=0xc00011bf38 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00011bfe0 sp=0xc00011bfc8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00011bfe8 sp=0xc00011bfe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 9 gp=0xc0001f0fc0 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x16e3afe5ca10f?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000091f38 sp=0xc000091f18 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000091fc8 sp=0xc000091f38 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000091fe0 sp=0xc000091fc8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000091fe8 sp=0xc000091fe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 10 gp=0xc0001f1180 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x564e04a4a680?, 0x3?, 0xda?, 0x8f?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000116738 sp=0xc000116718 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0001167c8 sp=0xc000116738 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0001167e0 sp=0xc0001167c8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0001167e8 sp=0xc0001167e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 11 gp=0xc0001f1340 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x564e04a4a680?, 0x1?, 0x8f?, 0x69?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000116f38 sp=0xc000116f18 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000116fc8 sp=0xc000116f38 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000116fe0 sp=0xc000116fc8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000116fe8 sp=0xc000116fe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 12 gp=0xc0001f1500 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x564e04a4a680?, 0x1?, 0x46?, 0x67?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000117738 sp=0xc000117718 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0001177c8 sp=0xc000117738 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0001177e0 sp=0xc0001177c8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0001177e8 sp=0xc0001177e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 13 gp=0xc0001f16c0 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x16e3afe5baaf9?, 0x0?, 0x0?, 0x0?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000117f38 sp=0xc000117f18 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000117fc8 sp=0xc000117f38 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000117fe0 sp=0xc000117fc8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000117fe8 sp=0xc000117fe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 14 gp=0xc0001f1880 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x16e3afe5bde06?, 0x1?, 0xd5?, 0x34?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000118738 sp=0xc000118718 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0001187c8 sp=0xc000118738 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0001187e0 sp=0xc0001187c8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0001187e8 sp=0xc0001187e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 15 gp=0xc0001f1a40 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x564e04a4a680?, 0x1?, 0x78?, 0x39?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000118f38 sp=0xc000118f18 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000118fc8 sp=0xc000118f38 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000118fe0 sp=0xc000118fc8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000118fe8 sp=0xc000118fe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 38 gp=0xc000102a80 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x564e04a4a680?, 0x1?, 0x6a?, 0xc4?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00011c738 sp=0xc00011c718 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00011c7c8 sp=0xc00011c738 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00011c7e0 sp=0xc00011c7c8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00011c7e8 sp=0xc00011c7e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 16 gp=0xc0001f1c00 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x16e3afe5bc298?, 0x1?, 0xb0?, 0x10?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000119738 sp=0xc000119718 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0001197c8 sp=0xc000119738 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0001197e0 sp=0xc0001197c8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0001197e8 sp=0xc0001197e0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 50 gp=0xc0001f1dc0 m=nil [GC worker (idle)]: 12:12:16 arch ollama[927]: runtime.gopark(0x16e3afe5ba5b0?, 0x1?, 0x1d?, 0x9c?, 0x0?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000119f38 sp=0xc000119f18 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.gcBgMarkWorker(0xc0000c7730) 12:12:16 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000119fc8 sp=0xc000119f38 pc=0x564e02aa08a9 12:12:16 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000119fe0 sp=0xc000119fc8 pc=0x564e02aa0785 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000119fe8 sp=0xc000119fe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 12:12:16 arch ollama[927]: runtime/mgc.go:1339 +0x105 12:12:16 arch ollama[927]: goroutine 39 gp=0xc000505a40 m=nil [sync.Cond.Wait]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0xc000165c78?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000165be8 sp=0xc000165bc8 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.goparkunlock(...) 12:12:16 arch ollama[927]: runtime/proc.go:441 12:12:16 arch ollama[927]: sync.runtime_notifyListWait(0xc0005311d0, 0x0) 12:12:16 arch ollama[927]: runtime/sema.go:597 +0x15a fp=0xc000165c38 sp=0xc000165be8 pc=0x564e02af46ba 12:12:16 arch ollama[927]: sync.(*Cond).Wait(0xc000165cb8?) 12:12:16 arch ollama[927]: sync/cond.go:71 +0x85 fp=0xc000165c70 sp=0xc000165c38 pc=0x564e02b047c5 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).processBatch(0xc00013f900, 0xc00033c140, 0xc00033c190) 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:408 +0x93 fp=0xc000165ee8 sp=0xc000165c70 pc=0x564e02f63213 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).run(0xc00013f900, {0x564e040a8410, 0xc000623090}) 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:387 +0x1d5 fp=0xc000165fb8 sp=0xc000165ee8 pc=0x564e02f63015 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.Execute.gowrap1() 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:981 +0x28 fp=0xc000165fe0 sp=0xc000165fb8 pc=0x564e02f683e8 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000165fe8 sp=0xc000165fe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by github.com/ollama/ollama/runner/llamarunner.Execute in goroutine 1 12:12:16 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:981 +0x4c5 12:12:16 arch ollama[927]: goroutine 45 gp=0xc000602fc0 m=nil [IO wait]: 12:12:16 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0xb?) 12:12:16 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000163dd8 sp=0xc000163db8 pc=0x564e02af2b6e 12:12:16 arch ollama[927]: runtime.netpollblock(0x564e02b16338?, 0x2a8c2a6?, 0x4e?) 12:12:16 arch ollama[927]: runtime/netpoll.go:575 +0xf7 fp=0xc000163e10 sp=0xc000163dd8 pc=0x564e02ab7e97 12:12:16 arch ollama[927]: internal/poll.runtime_pollWait(0x7f83f643ad98, 0x72) 12:12:16 arch ollama[927]: runtime/netpoll.go:351 +0x85 fp=0xc000163e30 sp=0xc000163e10 pc=0x564e02af1d85 12:12:16 arch ollama[927]: internal/poll.(*pollDesc).wait(0xc000701500?, 0xc000144c11?, 0x0) 12:12:16 arch ollama[927]: internal/poll/fd_poll_runtime.go:84 +0x27 fp=0xc000163e58 sp=0xc000163e30 pc=0x564e02b79f07 12:12:16 arch ollama[927]: internal/poll.(*pollDesc).waitRead(...) 12:12:16 arch ollama[927]: internal/poll/fd_poll_runtime.go:89 12:12:16 arch ollama[927]: internal/poll.(*FD).Read(0xc000701500, {0xc000144c11, 0x1, 0x1}) 12:12:16 arch ollama[927]: internal/poll/fd_unix.go:165 +0x27a fp=0xc000163ef0 sp=0xc000163e58 pc=0x564e02b7b1fa 12:12:16 arch ollama[927]: net.(*netFD).Read(0xc000701500, {0xc000144c11?, 0x0?, 0x0?}) 12:12:16 arch ollama[927]: net/fd_posix.go:55 +0x25 fp=0xc000163f38 sp=0xc000163ef0 pc=0x564e02bf0205 12:12:16 arch ollama[927]: net.(*conn).Read(0xc000092908, {0xc000144c11?, 0x0?, 0x0?}) 12:12:16 arch ollama[927]: net/net.go:194 +0x45 fp=0xc000163f80 sp=0xc000163f38 pc=0x564e02bfe5c5 12:12:16 arch ollama[927]: net/http.(*connReader).backgroundRead(0xc000144c00) 12:12:16 arch ollama[927]: net/http/server.go:690 +0x37 fp=0xc000163fc8 sp=0xc000163f80 pc=0x564e02dea4b7 12:12:16 arch ollama[927]: net/http.(*connReader).startBackgroundRead.gowrap2() 12:12:16 arch ollama[927]: net/http/server.go:686 +0x25 fp=0xc000163fe0 sp=0xc000163fc8 pc=0x564e02dea3e5 12:12:16 arch ollama[927]: runtime.goexit({}) 12:12:16 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000163fe8 sp=0xc000163fe0 pc=0x564e02afaa01 12:12:16 arch ollama[927]: created by net/http.(*connReader).startBackgroundRead in goroutine 40 12:12:16 arch ollama[927]: net/http/server.go:686 +0xb6 12:12:16 arch ollama[927]: rax 0x7f82d2850890 12:12:16 arch ollama[927]: rbx 0x0 12:12:16 arch ollama[927]: rcx 0x0 12:12:16 arch ollama[927]: rdx 0x0 12:12:16 arch ollama[927]: rdi 0x7f82d25dfff0 12:12:16 arch ollama[927]: rsi 0x7f82c622f310 12:12:16 arch ollama[927]: rbp 0x7f82d25e7e68 12:12:16 arch ollama[927]: rsp 0x7f82d5ffcbc0 12:12:16 arch ollama[927]: r8 0x0 12:12:16 arch ollama[927]: r9 0x7f82736c7040 12:12:16 arch ollama[927]: r10 0x24db000 12:12:16 arch ollama[927]: r11 0x246 12:12:16 arch ollama[927]: r12 0x7f81c445a9e0 12:12:16 arch ollama[927]: r13 0x160 12:12:16 arch ollama[927]: r14 0x7f82c622f030 12:12:16 arch ollama[927]: r15 0x1 12:12:16 arch ollama[927]: rip 0x564e038541cb 12:12:16 arch ollama[927]: rflags 0x10246 12:12:16 arch ollama[927]: cs 0x33 12:12:16 arch ollama[927]: fs 0x0 12:12:16 arch ollama[927]: gs 0x0 12:12:16 arch ollama[927]: time=2026-01-05T12:12:16.610+08:00 level=ERROR source=server.go:1583 msg="post predict" error="Post \"http://127.0.0.1:33773/completion\": EOF"
GiteaMirror commented 2026-04-29 09:04:50 -05:00
Author
Owner
Copy Link

@babysource commented on GitHub (Jan 5, 2026):

open-autoglm check_deployment always response "</answer>"

Image

@rick-github

<!-- gh-comment-id:3709286450 --> @babysource commented on GitHub (Jan 5, 2026): open-autoglm check_deployment always response "<\/answer>" <img width="1143" height="361" alt="Image" src="https://github.com/user-attachments/assets/dea688c0-2f70-4213-8d3d-adc17627149e" /> @rick-github
GiteaMirror commented 2026-04-29 09:04:51 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Jan 5, 2026):

What is "open-autoglm check_deployment"?

<!-- gh-comment-id:3709405855 --> @rick-github commented on GitHub (Jan 5, 2026): What is "open-autoglm check_deployment"?
GiteaMirror commented 2026-04-29 09:04:54 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Jan 5, 2026):

@pinghe Full log, from the start.

<!-- gh-comment-id:3709474468 --> @rick-github commented on GitHub (Jan 5, 2026): @pinghe Full log, from the start.
GiteaMirror commented 2026-04-29 09:04:55 -05:00
Author
Owner
Copy Link

@babysource commented on GitHub (Jan 5, 2026):

Where can I obtain the full log of ollama?

@rick-github

<!-- gh-comment-id:3709537887 --> @babysource commented on GitHub (Jan 5, 2026): Where can I obtain the full log of ollama? @rick-github
GiteaMirror commented 2026-04-29 09:04:56 -05:00
Author
Owner
Copy Link

@babysource commented on GitHub (Jan 5, 2026):

Step:

  1. exec: $ ollama run www.modelscope.cn/ggml-org/AutoGLM-Phone-9B-GGUF:Q4_K_M
  2. Rename the model “autoglm-phone:9b”
  3. git clone and install https://github.com/zai-org/Open-AutoGLM
  4. exec: $ python scripts/check_deployment_cn.py --base-url http://192.168.244.40:11434/v1 --model "autoglm-phone:9b"
  5. ollama api always response content “</answer>”

However, if the images in the script<scripts/check_deployment_cn.py> are removed, a normal response will be returned.

@rick-github

$ journalctl -u ollama --since "1 hour ago"

Full log

1月 05 11:45:17 wythe ollama[1712]: time=2026-01-05T11:45:17.590Z level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 39397"
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: loaded meta data with 33 key-value pairs and 523 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20a>1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv   0:                       general.architecture str              = glm4
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv   1:                               general.type str              = model
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv   2:                         general.size_label str              = 9.4B
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv   3:                            general.license str              = mit
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv   4:                   general.base_model.count u32              = 1
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv   5:                  general.base_model.0.name str              = GLM 4.1V 9B Base
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv   6:          general.base_model.0.organization str              = Zai Org
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv   7:              general.base_model.0.repo_url str              = https://huggingface.co/zai-org/GLM-4....
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv   8:                               general.tags arr[str,2]       = ["agent", "image-text-to-text"]
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv   9:                          general.languages arr[str,1]       = ["zh"]
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  10:                           glm4.block_count u32              = 40
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  11:                        glm4.context_length u32              = 65536
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  12:                      glm4.embedding_length u32              = 4096
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  13:                   glm4.feed_forward_length u32              = 13696
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  14:                  glm4.attention.head_count u32              = 32
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  15:               glm4.attention.head_count_kv u32              = 2
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  16:               glm4.rope.dimension_sections arr[i32,4]       = [8, 12, 12, 0]
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  17:                        glm4.rope.freq_base f32              = 10000.000000
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  18:      glm4.attention.layer_norm_rms_epsilon f32              = 0.000010
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  19:                  glm4.rope.dimension_count u32              = 64
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  20:                       tokenizer.ggml.model str              = gpt2
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  21:                         tokenizer.ggml.pre str              = glm4
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  22:                      tokenizer.ggml.tokens arr[str,151552]  = ["!", "\"", "#", "$", "%", "&", "'", ...
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  23:                  tokenizer.ggml.token_type arr[i32,151552]  = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  24:                      tokenizer.ggml.merges arr[str,318088]  = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "...
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  25:                tokenizer.ggml.eos_token_id u32              = 151329
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  26:            tokenizer.ggml.padding_token_id u32              = 151329
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  27:                tokenizer.ggml.eot_token_id u32              = 151336
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  28:            tokenizer.ggml.unknown_token_id u32              = 151329
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  29:                tokenizer.ggml.bos_token_id u32              = 151329
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  30:                    tokenizer.chat_template str              = [gMASK]<sop>\n{%- for msg in messages ...
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  31:               general.quantization_version u32              = 2
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv  32:                          general.file_type u32              = 15
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - type  f32:  281 tensors
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - type q5_0:   20 tensors
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - type q8_0:   20 tensors
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - type q4_K:  181 tensors
1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - type q6_K:   21 tensors
1月 05 11:45:17 wythe ollama[1712]: print_info: file format = GGUF V3 (latest)
1月 05 11:45:17 wythe ollama[1712]: print_info: file type   = Q4_K - Medium
1月 05 11:45:17 wythe ollama[1712]: print_info: file size   = 5.73 GiB (5.24 BPW)
1月 05 11:45:18 wythe ollama[1712]: load: special_eot_id is not in special_eog_ids - the tokenizer config may be incorrect
1月 05 11:45:18 wythe ollama[1712]: load: printing all EOG tokens:
1月 05 11:45:18 wythe ollama[1712]: load:   - 151329 ('<|endoftext|>')
1月 05 11:45:18 wythe ollama[1712]: load:   - 151336 ('<|user|>')
1月 05 11:45:18 wythe ollama[1712]: load: special tokens cache size = 23
1月 05 11:45:18 wythe ollama[1712]: load: token to piece cache size = 0.9711 MB
1月 05 11:45:18 wythe ollama[1712]: print_info: arch             = glm4
1月 05 11:45:18 wythe ollama[1712]: print_info: vocab_only       = 1
1月 05 11:45:18 wythe ollama[1712]: print_info: no_alloc         = 0
1月 05 11:45:18 wythe ollama[1712]: print_info: model type       = ?B
1月 05 11:45:18 wythe ollama[1712]: print_info: model params     = 9.40 B
1月 05 11:45:18 wythe ollama[1712]: print_info: general.name     = n/a
1月 05 11:45:18 wythe ollama[1712]: print_info: vocab type       = BPE
1月 05 11:45:18 wythe ollama[1712]: print_info: n_vocab          = 151552
1月 05 11:45:18 wythe ollama[1712]: print_info: n_merges         = 318088
1月 05 11:45:18 wythe ollama[1712]: print_info: BOS token        = 151329 '<|endoftext|>'
1月 05 11:45:18 wythe ollama[1712]: print_info: EOS token        = 151329 '<|endoftext|>'
1月 05 11:45:18 wythe ollama[1712]: print_info: EOT token        = 151336 '<|user|>'
1月 05 11:45:18 wythe ollama[1712]: print_info: UNK token        = 151329 '<|endoftext|>'
1月 05 11:45:18 wythe ollama[1712]: print_info: PAD token        = 151329 '<|endoftext|>'
1月 05 11:45:18 wythe ollama[1712]: print_info: LF token         = 198 'Ċ'
1月 05 11:45:18 wythe ollama[1712]: print_info: EOG token        = 151329 '<|endoftext|>'
1月 05 11:45:18 wythe ollama[1712]: print_info: EOG token        = 151336 '<|user|>'
1月 05 11:45:18 wythe ollama[1712]: print_info: max token length = 1024
1月 05 11:45:18 wythe ollama[1712]: llama_model_load: vocab only - skipping tensors
1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.087Z level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --model /usr/share/ollama/.ollama/models>
1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.088Z level=INFO source=sched.go:443 msg="system memory" total="15.5 GiB" free="12.5 GiB" free_swap="4.0 GiB"
1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.088Z level=INFO source=sched.go:450 msg="gpu memory" id=GPU-f74419c7-e9cb-f6fb-75c3-739c9fdc77e8 library=CUDA available="11.5 GiB" fr>
1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.088Z level=INFO source=server.go:496 msg="loading model" "model layers"=41 requested=-1
1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.088Z level=INFO source=device.go:240 msg="model weights" device=CUDA0 size="5.4 GiB"
1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.088Z level=INFO source=device.go:251 msg="kv cache" device=CUDA0 size="160.0 MiB"
1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.088Z level=INFO source=device.go:262 msg="compute graph" device=CUDA0 size="426.7 MiB"
1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.088Z level=INFO source=device.go:272 msg="total memory" size="6.0 GiB"
1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.098Z level=INFO source=runner.go:965 msg="starting go runner"
1月 05 11:45:18 wythe ollama[1712]: load_backend: loaded CPU backend from /usr/local/lib/ollama/libggml-cpu-haswell.so
1月 05 11:45:18 wythe ollama[1712]: ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
1月 05 11:45:18 wythe ollama[1712]: ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
1月 05 11:45:18 wythe ollama[1712]: ggml_cuda_init: found 1 CUDA devices:
1月 05 11:45:18 wythe ollama[1712]:   Device 0: NVIDIA GeForce GTX TITAN X, compute capability 5.2, VMM: yes, ID: GPU-f74419c7-e9cb-f6fb-75c3-739c9fdc77e8
1月 05 11:45:18 wythe ollama[1712]: load_backend: loaded CUDA backend from /usr/local/lib/ollama/cuda_v12/libggml-cuda.so
1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.152Z level=INFO source=ggml.go:104 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.BMI2>
1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.152Z level=INFO source=runner.go:1001 msg="Server listening on 127.0.0.1:44695"
1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.163Z level=INFO source=runner.go:895 msg=load request="{Operation:commit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:Auto KvS>
1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.163Z level=INFO source=server.go:1338 msg="waiting for llama runner to start responding"
1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.164Z level=INFO source=server.go:1372 msg="waiting for server to become available" status="llm server loading model"
1月 05 11:45:18 wythe ollama[1712]: ggml_backend_cuda_device_get_memory device GPU-f74419c7-e9cb-f6fb-75c3-739c9fdc77e8 utilizing NVML memory reporting free: 12778471424 total: 12884901888
1月 05 11:45:18 wythe ollama[1712]: llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce GTX TITAN X) (0000:01:00.0) - 12186 MiB free
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: loaded meta data with 33 key-value pairs and 523 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20a>
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv   0:                       general.architecture str              = glm4
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv   1:                               general.type str              = model
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv   2:                         general.size_label str              = 9.4B
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv   3:                            general.license str              = mit
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv   4:                   general.base_model.count u32              = 1
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv   5:                  general.base_model.0.name str              = GLM 4.1V 9B Base
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv   6:          general.base_model.0.organization str              = Zai Org
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv   7:              general.base_model.0.repo_url str              = https://huggingface.co/zai-org/GLM-4....
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv   8:                               general.tags arr[str,2]       = ["agent", "image-text-to-text"]
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv   9:                          general.languages arr[str,1]       = ["zh"]
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  10:                           glm4.block_count u32              = 40
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  11:                        glm4.context_length u32              = 65536
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  12:                      glm4.embedding_length u32              = 4096
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  13:                   glm4.feed_forward_length u32              = 13696
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  14:                  glm4.attention.head_count u32              = 32
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  15:               glm4.attention.head_count_kv u32              = 2
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  16:               glm4.rope.dimension_sections arr[i32,4]       = [8, 12, 12, 0]
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  17:                        glm4.rope.freq_base f32              = 10000.000000
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  18:      glm4.attention.layer_norm_rms_epsilon f32              = 0.000010
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  19:                  glm4.rope.dimension_count u32              = 64
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  20:                       tokenizer.ggml.model str              = gpt2
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  21:                         tokenizer.ggml.pre str              = glm4
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  22:                      tokenizer.ggml.tokens arr[str,151552]  = ["!", "\"", "#", "$", "%", "&", "'", ...
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  23:                  tokenizer.ggml.token_type arr[i32,151552]  = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  24:                      tokenizer.ggml.merges arr[str,318088]  = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "...
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  25:                tokenizer.ggml.eos_token_id u32              = 151329
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  26:            tokenizer.ggml.padding_token_id u32              = 151329
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  27:                tokenizer.ggml.eot_token_id u32              = 151336
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  28:            tokenizer.ggml.unknown_token_id u32              = 151329
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  29:                tokenizer.ggml.bos_token_id u32              = 151329
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  30:                    tokenizer.chat_template str              = [gMASK]<sop>\n{%- for msg in messages ...
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  31:               general.quantization_version u32              = 2
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv  32:                          general.file_type u32              = 15
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - type  f32:  281 tensors
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - type q5_0:   20 tensors
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - type q8_0:   20 tensors
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - type q4_K:  181 tensors
1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - type q6_K:   21 tensors
1月 05 11:45:18 wythe ollama[1712]: print_info: file format = GGUF V3 (latest)
1月 05 11:45:18 wythe ollama[1712]: print_info: file type   = Q4_K - Medium
1月 05 11:45:18 wythe ollama[1712]: print_info: file size   = 5.73 GiB (5.24 BPW)
1月 05 11:45:18 wythe ollama[1712]: load: special_eot_id is not in special_eog_ids - the tokenizer config may be incorrect
1月 05 11:45:18 wythe ollama[1712]: load: printing all EOG tokens:
1月 05 11:45:18 wythe ollama[1712]: load:   - 151329 ('<|endoftext|>')
1月 05 11:45:18 wythe ollama[1712]: load:   - 151336 ('<|user|>')
1月 05 11:45:18 wythe ollama[1712]: load: special tokens cache size = 23
1月 05 11:45:18 wythe ollama[1712]: load: token to piece cache size = 0.9711 MB
1月 05 11:45:18 wythe ollama[1712]: print_info: arch             = glm4
1月 05 11:45:18 wythe ollama[1712]: print_info: vocab_only       = 0
1月 05 11:45:18 wythe ollama[1712]: print_info: no_alloc         = 0
1月 05 11:45:18 wythe ollama[1712]: print_info: n_ctx_train      = 65536
1月 05 11:45:18 wythe ollama[1712]: print_info: n_embd           = 4096
1月 05 11:45:18 wythe ollama[1712]: print_info: n_embd_inp       = 4096
1月 05 11:45:18 wythe ollama[1712]: print_info: n_layer          = 40
1月 05 11:45:18 wythe ollama[1712]: print_info: n_head           = 32
1月 05 11:45:18 wythe ollama[1712]: print_info: n_head_kv        = 2
1月 05 11:45:18 wythe ollama[1712]: print_info: n_rot            = 64
1月 05 11:45:18 wythe ollama[1712]: print_info: n_swa            = 0
1月 05 11:45:18 wythe ollama[1712]: print_info: is_swa_any       = 0
1月 05 11:45:18 wythe ollama[1712]: print_info: n_embd_head_k    = 128
1月 05 11:45:18 wythe ollama[1712]: print_info: n_embd_head_v    = 128
1月 05 11:45:18 wythe ollama[1712]: print_info: n_gqa            = 16
1月 05 11:45:18 wythe ollama[1712]: print_info: n_embd_k_gqa     = 256
1月 05 11:45:18 wythe ollama[1712]: print_info: n_embd_v_gqa     = 256
1月 05 11:45:18 wythe ollama[1712]: print_info: f_norm_eps       = 0.0e+00
1月 05 11:45:18 wythe ollama[1712]: print_info: f_norm_rms_eps   = 1.0e-05
1月 05 11:45:18 wythe ollama[1712]: print_info: f_clamp_kqv      = 0.0e+00
1月 05 11:45:18 wythe ollama[1712]: print_info: f_max_alibi_bias = 0.0e+00
1月 05 11:45:18 wythe ollama[1712]: print_info: f_logit_scale    = 0.0e+00
1月 05 11:45:18 wythe ollama[1712]: print_info: f_attn_scale     = 0.0e+00
1月 05 11:45:18 wythe ollama[1712]: print_info: n_ff             = 13696
1月 05 11:45:18 wythe ollama[1712]: print_info: n_expert         = 0
1月 05 11:45:18 wythe ollama[1712]: print_info: n_expert_used    = 0
1月 05 11:45:18 wythe ollama[1712]: print_info: n_expert_groups  = 0
1月 05 11:45:18 wythe ollama[1712]: print_info: n_group_used     = 0
1月 05 11:45:18 wythe ollama[1712]: print_info: causal attn      = 1
1月 05 11:45:18 wythe ollama[1712]: print_info: pooling type     = 0
1月 05 11:45:18 wythe ollama[1712]: print_info: rope type        = 8
1月 05 11:45:18 wythe ollama[1712]: print_info: rope scaling     = linear
1月 05 11:45:18 wythe ollama[1712]: print_info: freq_base_train  = 10000.0
1月 05 11:45:18 wythe ollama[1712]: print_info: freq_scale_train = 1
1月 05 11:45:18 wythe ollama[1712]: print_info: n_ctx_orig_yarn  = 65536
1月 05 11:45:18 wythe ollama[1712]: print_info: rope_yarn_log_mul= 0.0000
1月 05 11:45:18 wythe ollama[1712]: print_info: rope_finetuned   = unknown
1月 05 11:45:18 wythe ollama[1712]: print_info: mrope sections   = [8, 12, 12, 0]
1月 05 11:45:18 wythe ollama[1712]: print_info: model type       = 9B
1月 05 11:45:18 wythe ollama[1712]: print_info: model params     = 9.40 B
1月 05 11:45:18 wythe ollama[1712]: print_info: general.name     = n/a
1月 05 11:45:18 wythe ollama[1712]: print_info: vocab type       = BPE
1月 05 11:45:18 wythe ollama[1712]: print_info: n_vocab          = 151552
1月 05 11:45:18 wythe ollama[1712]: print_info: n_merges         = 318088
1月 05 11:45:18 wythe ollama[1712]: print_info: BOS token        = 151329 '<|endoftext|>'
1月 05 11:45:18 wythe ollama[1712]: print_info: EOS token        = 151329 '<|endoftext|>'
1月 05 11:45:18 wythe ollama[1712]: print_info: EOT token        = 151336 '<|user|>'
1月 05 11:45:18 wythe ollama[1712]: print_info: UNK token        = 151329 '<|endoftext|>'
1月 05 11:45:18 wythe ollama[1712]: print_info: PAD token        = 151329 '<|endoftext|>'
1月 05 11:45:18 wythe ollama[1712]: print_info: LF token         = 198 'Ċ'
1月 05 11:45:18 wythe ollama[1712]: print_info: EOG token        = 151329 '<|endoftext|>'
1月 05 11:45:18 wythe ollama[1712]: print_info: EOG token        = 151336 '<|user|>'
1月 05 11:45:18 wythe ollama[1712]: print_info: max token length = 1024
1月 05 11:45:18 wythe ollama[1712]: load_tensors: loading model tensors, this can take a while... (mmap = true)
1月 05 11:45:18 wythe ollama[1712]: load_tensors: offloading 40 repeating layers to GPU
1月 05 11:45:18 wythe ollama[1712]: load_tensors: offloading output layer to GPU
1月 05 11:45:18 wythe ollama[1712]: load_tensors: offloaded 41/41 layers to GPU
1月 05 11:45:18 wythe ollama[1712]: load_tensors:   CPU_Mapped model buffer size =   333.00 MiB
1月 05 11:45:18 wythe ollama[1712]: load_tensors:        CUDA0 model buffer size =  5539.01 MiB
1月 05 11:45:19 wythe ollama[1712]: llama_context: constructing llama_context
1月 05 11:45:19 wythe ollama[1712]: llama_context: n_seq_max     = 1
1月 05 11:45:19 wythe ollama[1712]: llama_context: n_ctx         = 4096
1月 05 11:45:19 wythe ollama[1712]: llama_context: n_ctx_seq     = 4096
1月 05 11:45:19 wythe ollama[1712]: llama_context: n_batch       = 512
1月 05 11:45:19 wythe ollama[1712]: llama_context: n_ubatch      = 512
1月 05 11:45:19 wythe ollama[1712]: llama_context: causal_attn   = 1
1月 05 11:45:19 wythe ollama[1712]: llama_context: flash_attn    = auto
1月 05 11:45:19 wythe ollama[1712]: llama_context: kv_unified    = false
1月 05 11:45:19 wythe ollama[1712]: llama_context: freq_base     = 10000.0
1月 05 11:45:19 wythe ollama[1712]: llama_context: freq_scale    = 1
1月 05 11:45:19 wythe ollama[1712]: llama_context: n_ctx_seq (4096) < n_ctx_train (65536) -- the full capacity of the model will not be utilized
1月 05 11:45:19 wythe ollama[1712]: llama_context:  CUDA_Host  output buffer size =     0.59 MiB
1月 05 11:45:19 wythe ollama[1712]: llama_kv_cache:      CUDA0 KV buffer size =   160.00 MiB
1月 05 11:45:19 wythe ollama[1712]: llama_kv_cache: size =  160.00 MiB (  4096 cells,  40 layers,  1/1 seqs), K (f16):   80.00 MiB, V (f16):   80.00 MiB
1月 05 11:45:19 wythe ollama[1712]: llama_context: Flash Attention was auto, set to enabled
1月 05 11:45:19 wythe ollama[1712]: llama_context:      CUDA0 compute buffer size =   304.00 MiB
1月 05 11:45:19 wythe ollama[1712]: llama_context:  CUDA_Host compute buffer size =    16.02 MiB
1月 05 11:45:19 wythe ollama[1712]: llama_context: graph nodes  = 1487
1月 05 11:45:19 wythe ollama[1712]: llama_context: graph splits = 2
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: model name:
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: description:
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: GGUF version: 3
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: alignment:    32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: n_tensors:    182
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: n_kv:         25
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: has vision encoder
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[0]: n_dims = 2, name = v.blk.0.attn_out.weight, tensor_size=2506752, offset=0, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[1]: n_dims = 2, name = v.blk.0.attn_qkv.weight, tensor_size=7520256, offset=2506752, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[2]: n_dims = 2, name = v.blk.0.ffn_down.weight, tensor_size=6684672, offset=10027008, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[3]: n_dims = 2, name = v.blk.0.ffn_gate.weight, tensor_size=6684672, offset=16711680, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[4]: n_dims = 2, name = v.blk.0.ffn_up.weight, tensor_size=6684672, offset=23396352, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[5]: n_dims = 1, name = v.blk.0.ln1.weight, tensor_size=6144, offset=30081024, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[6]: n_dims = 1, name = v.blk.0.ln2.weight, tensor_size=6144, offset=30087168, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[7]: n_dims = 2, name = v.blk.1.attn_out.weight, tensor_size=2506752, offset=30093312, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[8]: n_dims = 2, name = v.blk.1.attn_qkv.weight, tensor_size=7520256, offset=32600064, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[9]: n_dims = 2, name = v.blk.1.ffn_down.weight, tensor_size=6684672, offset=40120320, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[10]: n_dims = 2, name = v.blk.1.ffn_gate.weight, tensor_size=6684672, offset=46804992, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[11]: n_dims = 2, name = v.blk.1.ffn_up.weight, tensor_size=6684672, offset=53489664, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[12]: n_dims = 1, name = v.blk.1.ln1.weight, tensor_size=6144, offset=60174336, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[13]: n_dims = 1, name = v.blk.1.ln2.weight, tensor_size=6144, offset=60180480, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[14]: n_dims = 2, name = v.blk.10.attn_out.weight, tensor_size=2506752, offset=60186624, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[15]: n_dims = 2, name = v.blk.10.attn_qkv.weight, tensor_size=7520256, offset=62693376, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[16]: n_dims = 2, name = v.blk.10.ffn_down.weight, tensor_size=6684672, offset=70213632, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[17]: n_dims = 2, name = v.blk.10.ffn_gate.weight, tensor_size=6684672, offset=76898304, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[18]: n_dims = 2, name = v.blk.10.ffn_up.weight, tensor_size=6684672, offset=83582976, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[19]: n_dims = 1, name = v.blk.10.ln1.weight, tensor_size=6144, offset=90267648, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[20]: n_dims = 1, name = v.blk.10.ln2.weight, tensor_size=6144, offset=90273792, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[21]: n_dims = 2, name = v.blk.11.attn_out.weight, tensor_size=2506752, offset=90279936, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[22]: n_dims = 2, name = v.blk.11.attn_qkv.weight, tensor_size=7520256, offset=92786688, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[23]: n_dims = 2, name = v.blk.11.ffn_down.weight, tensor_size=6684672, offset=100306944, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[24]: n_dims = 2, name = v.blk.11.ffn_gate.weight, tensor_size=6684672, offset=106991616, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[25]: n_dims = 2, name = v.blk.11.ffn_up.weight, tensor_size=6684672, offset=113676288, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[26]: n_dims = 1, name = v.blk.11.ln1.weight, tensor_size=6144, offset=120360960, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[27]: n_dims = 1, name = v.blk.11.ln2.weight, tensor_size=6144, offset=120367104, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[28]: n_dims = 2, name = v.blk.12.attn_out.weight, tensor_size=2506752, offset=120373248, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[29]: n_dims = 2, name = v.blk.12.attn_qkv.weight, tensor_size=7520256, offset=122880000, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[30]: n_dims = 2, name = v.blk.12.ffn_down.weight, tensor_size=6684672, offset=130400256, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[31]: n_dims = 2, name = v.blk.12.ffn_gate.weight, tensor_size=6684672, offset=137084928, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[32]: n_dims = 2, name = v.blk.12.ffn_up.weight, tensor_size=6684672, offset=143769600, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[33]: n_dims = 1, name = v.blk.12.ln1.weight, tensor_size=6144, offset=150454272, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[34]: n_dims = 1, name = v.blk.12.ln2.weight, tensor_size=6144, offset=150460416, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[35]: n_dims = 2, name = v.blk.13.attn_out.weight, tensor_size=2506752, offset=150466560, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[36]: n_dims = 2, name = v.blk.13.attn_qkv.weight, tensor_size=7520256, offset=152973312, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[37]: n_dims = 2, name = v.blk.13.ffn_down.weight, tensor_size=6684672, offset=160493568, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[38]: n_dims = 2, name = v.blk.13.ffn_gate.weight, tensor_size=6684672, offset=167178240, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[39]: n_dims = 2, name = v.blk.13.ffn_up.weight, tensor_size=6684672, offset=173862912, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[40]: n_dims = 1, name = v.blk.13.ln1.weight, tensor_size=6144, offset=180547584, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[41]: n_dims = 1, name = v.blk.13.ln2.weight, tensor_size=6144, offset=180553728, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[42]: n_dims = 2, name = v.blk.14.attn_out.weight, tensor_size=2506752, offset=180559872, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[43]: n_dims = 2, name = v.blk.14.attn_qkv.weight, tensor_size=7520256, offset=183066624, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[44]: n_dims = 2, name = v.blk.14.ffn_down.weight, tensor_size=6684672, offset=190586880, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[45]: n_dims = 2, name = v.blk.14.ffn_gate.weight, tensor_size=6684672, offset=197271552, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[46]: n_dims = 2, name = v.blk.14.ffn_up.weight, tensor_size=6684672, offset=203956224, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[47]: n_dims = 1, name = v.blk.14.ln1.weight, tensor_size=6144, offset=210640896, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[48]: n_dims = 1, name = v.blk.14.ln2.weight, tensor_size=6144, offset=210647040, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[49]: n_dims = 2, name = v.blk.15.attn_out.weight, tensor_size=2506752, offset=210653184, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[50]: n_dims = 2, name = v.blk.15.attn_qkv.weight, tensor_size=7520256, offset=213159936, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[51]: n_dims = 2, name = v.blk.15.ffn_down.weight, tensor_size=6684672, offset=220680192, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[52]: n_dims = 2, name = v.blk.15.ffn_gate.weight, tensor_size=6684672, offset=227364864, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[53]: n_dims = 2, name = v.blk.15.ffn_up.weight, tensor_size=6684672, offset=234049536, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[54]: n_dims = 1, name = v.blk.15.ln1.weight, tensor_size=6144, offset=240734208, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[55]: n_dims = 1, name = v.blk.15.ln2.weight, tensor_size=6144, offset=240740352, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[56]: n_dims = 2, name = v.blk.16.attn_out.weight, tensor_size=2506752, offset=240746496, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[57]: n_dims = 2, name = v.blk.16.attn_qkv.weight, tensor_size=7520256, offset=243253248, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[58]: n_dims = 2, name = v.blk.16.ffn_down.weight, tensor_size=6684672, offset=250773504, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[59]: n_dims = 2, name = v.blk.16.ffn_gate.weight, tensor_size=6684672, offset=257458176, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[60]: n_dims = 2, name = v.blk.16.ffn_up.weight, tensor_size=6684672, offset=264142848, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[61]: n_dims = 1, name = v.blk.16.ln1.weight, tensor_size=6144, offset=270827520, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[62]: n_dims = 1, name = v.blk.16.ln2.weight, tensor_size=6144, offset=270833664, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[63]: n_dims = 2, name = v.blk.17.attn_out.weight, tensor_size=2506752, offset=270839808, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[64]: n_dims = 2, name = v.blk.17.attn_qkv.weight, tensor_size=7520256, offset=273346560, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[65]: n_dims = 2, name = v.blk.17.ffn_down.weight, tensor_size=6684672, offset=280866816, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[66]: n_dims = 2, name = v.blk.17.ffn_gate.weight, tensor_size=6684672, offset=287551488, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[67]: n_dims = 2, name = v.blk.17.ffn_up.weight, tensor_size=6684672, offset=294236160, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[68]: n_dims = 1, name = v.blk.17.ln1.weight, tensor_size=6144, offset=300920832, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[69]: n_dims = 1, name = v.blk.17.ln2.weight, tensor_size=6144, offset=300926976, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[70]: n_dims = 2, name = v.blk.18.attn_out.weight, tensor_size=2506752, offset=300933120, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[71]: n_dims = 2, name = v.blk.18.attn_qkv.weight, tensor_size=7520256, offset=303439872, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[72]: n_dims = 2, name = v.blk.18.ffn_down.weight, tensor_size=6684672, offset=310960128, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[73]: n_dims = 2, name = v.blk.18.ffn_gate.weight, tensor_size=6684672, offset=317644800, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[74]: n_dims = 2, name = v.blk.18.ffn_up.weight, tensor_size=6684672, offset=324329472, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[75]: n_dims = 1, name = v.blk.18.ln1.weight, tensor_size=6144, offset=331014144, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[76]: n_dims = 1, name = v.blk.18.ln2.weight, tensor_size=6144, offset=331020288, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[77]: n_dims = 2, name = v.blk.19.attn_out.weight, tensor_size=2506752, offset=331026432, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[78]: n_dims = 2, name = v.blk.19.attn_qkv.weight, tensor_size=7520256, offset=333533184, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[79]: n_dims = 2, name = v.blk.19.ffn_down.weight, tensor_size=6684672, offset=341053440, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[80]: n_dims = 2, name = v.blk.19.ffn_gate.weight, tensor_size=6684672, offset=347738112, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[81]: n_dims = 2, name = v.blk.19.ffn_up.weight, tensor_size=6684672, offset=354422784, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[82]: n_dims = 1, name = v.blk.19.ln1.weight, tensor_size=6144, offset=361107456, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[83]: n_dims = 1, name = v.blk.19.ln2.weight, tensor_size=6144, offset=361113600, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[84]: n_dims = 2, name = v.blk.2.attn_out.weight, tensor_size=2506752, offset=361119744, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[85]: n_dims = 2, name = v.blk.2.attn_qkv.weight, tensor_size=7520256, offset=363626496, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[86]: n_dims = 2, name = v.blk.2.ffn_down.weight, tensor_size=6684672, offset=371146752, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[87]: n_dims = 2, name = v.blk.2.ffn_gate.weight, tensor_size=6684672, offset=377831424, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[88]: n_dims = 2, name = v.blk.2.ffn_up.weight, tensor_size=6684672, offset=384516096, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[89]: n_dims = 1, name = v.blk.2.ln1.weight, tensor_size=6144, offset=391200768, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[90]: n_dims = 1, name = v.blk.2.ln2.weight, tensor_size=6144, offset=391206912, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[91]: n_dims = 2, name = v.blk.20.attn_out.weight, tensor_size=2506752, offset=391213056, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[92]: n_dims = 2, name = v.blk.20.attn_qkv.weight, tensor_size=7520256, offset=393719808, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[93]: n_dims = 2, name = v.blk.20.ffn_down.weight, tensor_size=6684672, offset=401240064, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[94]: n_dims = 2, name = v.blk.20.ffn_gate.weight, tensor_size=6684672, offset=407924736, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[95]: n_dims = 2, name = v.blk.20.ffn_up.weight, tensor_size=6684672, offset=414609408, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[96]: n_dims = 1, name = v.blk.20.ln1.weight, tensor_size=6144, offset=421294080, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[97]: n_dims = 1, name = v.blk.20.ln2.weight, tensor_size=6144, offset=421300224, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[98]: n_dims = 2, name = v.blk.21.attn_out.weight, tensor_size=2506752, offset=421306368, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[99]: n_dims = 2, name = v.blk.21.attn_qkv.weight, tensor_size=7520256, offset=423813120, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[100]: n_dims = 2, name = v.blk.21.ffn_down.weight, tensor_size=6684672, offset=431333376, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[101]: n_dims = 2, name = v.blk.21.ffn_gate.weight, tensor_size=6684672, offset=438018048, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[102]: n_dims = 2, name = v.blk.21.ffn_up.weight, tensor_size=6684672, offset=444702720, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[103]: n_dims = 1, name = v.blk.21.ln1.weight, tensor_size=6144, offset=451387392, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[104]: n_dims = 1, name = v.blk.21.ln2.weight, tensor_size=6144, offset=451393536, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[105]: n_dims = 2, name = v.blk.22.attn_out.weight, tensor_size=2506752, offset=451399680, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[106]: n_dims = 2, name = v.blk.22.attn_qkv.weight, tensor_size=7520256, offset=453906432, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[107]: n_dims = 2, name = v.blk.22.ffn_down.weight, tensor_size=6684672, offset=461426688, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[108]: n_dims = 2, name = v.blk.22.ffn_gate.weight, tensor_size=6684672, offset=468111360, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[109]: n_dims = 2, name = v.blk.22.ffn_up.weight, tensor_size=6684672, offset=474796032, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[110]: n_dims = 1, name = v.blk.22.ln1.weight, tensor_size=6144, offset=481480704, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[111]: n_dims = 1, name = v.blk.22.ln2.weight, tensor_size=6144, offset=481486848, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[112]: n_dims = 2, name = v.blk.23.attn_out.weight, tensor_size=2506752, offset=481492992, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[113]: n_dims = 2, name = v.blk.23.attn_qkv.weight, tensor_size=7520256, offset=483999744, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[114]: n_dims = 2, name = v.blk.23.ffn_down.weight, tensor_size=6684672, offset=491520000, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[115]: n_dims = 2, name = v.blk.23.ffn_gate.weight, tensor_size=6684672, offset=498204672, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[116]: n_dims = 2, name = v.blk.23.ffn_up.weight, tensor_size=6684672, offset=504889344, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[117]: n_dims = 1, name = v.blk.23.ln1.weight, tensor_size=6144, offset=511574016, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[118]: n_dims = 1, name = v.blk.23.ln2.weight, tensor_size=6144, offset=511580160, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[119]: n_dims = 2, name = v.blk.3.attn_out.weight, tensor_size=2506752, offset=511586304, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[120]: n_dims = 2, name = v.blk.3.attn_qkv.weight, tensor_size=7520256, offset=514093056, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[121]: n_dims = 2, name = v.blk.3.ffn_down.weight, tensor_size=6684672, offset=521613312, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[122]: n_dims = 2, name = v.blk.3.ffn_gate.weight, tensor_size=6684672, offset=528297984, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[123]: n_dims = 2, name = v.blk.3.ffn_up.weight, tensor_size=6684672, offset=534982656, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[124]: n_dims = 1, name = v.blk.3.ln1.weight, tensor_size=6144, offset=541667328, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[125]: n_dims = 1, name = v.blk.3.ln2.weight, tensor_size=6144, offset=541673472, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[126]: n_dims = 2, name = v.blk.4.attn_out.weight, tensor_size=2506752, offset=541679616, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[127]: n_dims = 2, name = v.blk.4.attn_qkv.weight, tensor_size=7520256, offset=544186368, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[128]: n_dims = 2, name = v.blk.4.ffn_down.weight, tensor_size=6684672, offset=551706624, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[129]: n_dims = 2, name = v.blk.4.ffn_gate.weight, tensor_size=6684672, offset=558391296, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[130]: n_dims = 2, name = v.blk.4.ffn_up.weight, tensor_size=6684672, offset=565075968, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[131]: n_dims = 1, name = v.blk.4.ln1.weight, tensor_size=6144, offset=571760640, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[132]: n_dims = 1, name = v.blk.4.ln2.weight, tensor_size=6144, offset=571766784, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[133]: n_dims = 2, name = v.blk.5.attn_out.weight, tensor_size=2506752, offset=571772928, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[134]: n_dims = 2, name = v.blk.5.attn_qkv.weight, tensor_size=7520256, offset=574279680, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[135]: n_dims = 2, name = v.blk.5.ffn_down.weight, tensor_size=6684672, offset=581799936, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[136]: n_dims = 2, name = v.blk.5.ffn_gate.weight, tensor_size=6684672, offset=588484608, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[137]: n_dims = 2, name = v.blk.5.ffn_up.weight, tensor_size=6684672, offset=595169280, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[138]: n_dims = 1, name = v.blk.5.ln1.weight, tensor_size=6144, offset=601853952, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[139]: n_dims = 1, name = v.blk.5.ln2.weight, tensor_size=6144, offset=601860096, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[140]: n_dims = 2, name = v.blk.6.attn_out.weight, tensor_size=2506752, offset=601866240, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[141]: n_dims = 2, name = v.blk.6.attn_qkv.weight, tensor_size=7520256, offset=604372992, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[142]: n_dims = 2, name = v.blk.6.ffn_down.weight, tensor_size=6684672, offset=611893248, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[143]: n_dims = 2, name = v.blk.6.ffn_gate.weight, tensor_size=6684672, offset=618577920, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[144]: n_dims = 2, name = v.blk.6.ffn_up.weight, tensor_size=6684672, offset=625262592, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[145]: n_dims = 1, name = v.blk.6.ln1.weight, tensor_size=6144, offset=631947264, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[146]: n_dims = 1, name = v.blk.6.ln2.weight, tensor_size=6144, offset=631953408, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[147]: n_dims = 2, name = v.blk.7.attn_out.weight, tensor_size=2506752, offset=631959552, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[148]: n_dims = 2, name = v.blk.7.attn_qkv.weight, tensor_size=7520256, offset=634466304, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[149]: n_dims = 2, name = v.blk.7.ffn_down.weight, tensor_size=6684672, offset=641986560, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[150]: n_dims = 2, name = v.blk.7.ffn_gate.weight, tensor_size=6684672, offset=648671232, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[151]: n_dims = 2, name = v.blk.7.ffn_up.weight, tensor_size=6684672, offset=655355904, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[152]: n_dims = 1, name = v.blk.7.ln1.weight, tensor_size=6144, offset=662040576, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[153]: n_dims = 1, name = v.blk.7.ln2.weight, tensor_size=6144, offset=662046720, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[154]: n_dims = 2, name = v.blk.8.attn_out.weight, tensor_size=2506752, offset=662052864, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[155]: n_dims = 2, name = v.blk.8.attn_qkv.weight, tensor_size=7520256, offset=664559616, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[156]: n_dims = 2, name = v.blk.8.ffn_down.weight, tensor_size=6684672, offset=672079872, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[157]: n_dims = 2, name = v.blk.8.ffn_gate.weight, tensor_size=6684672, offset=678764544, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[158]: n_dims = 2, name = v.blk.8.ffn_up.weight, tensor_size=6684672, offset=685449216, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[159]: n_dims = 1, name = v.blk.8.ln1.weight, tensor_size=6144, offset=692133888, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[160]: n_dims = 1, name = v.blk.8.ln2.weight, tensor_size=6144, offset=692140032, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[161]: n_dims = 2, name = v.blk.9.attn_out.weight, tensor_size=2506752, offset=692146176, shape:[1536, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[162]: n_dims = 2, name = v.blk.9.attn_qkv.weight, tensor_size=7520256, offset=694652928, shape:[1536, 4608, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[163]: n_dims = 2, name = v.blk.9.ffn_down.weight, tensor_size=6684672, offset=702173184, shape:[4096, 1536, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[164]: n_dims = 2, name = v.blk.9.ffn_gate.weight, tensor_size=6684672, offset=708857856, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[165]: n_dims = 2, name = v.blk.9.ffn_up.weight, tensor_size=6684672, offset=715542528, shape:[1536, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[166]: n_dims = 1, name = v.blk.9.ln1.weight, tensor_size=6144, offset=722227200, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[167]: n_dims = 1, name = v.blk.9.ln2.weight, tensor_size=6144, offset=722233344, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[168]: n_dims = 1, name = mm.patch_merger.bias, tensor_size=16384, offset=722239488, shape:[4096, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[169]: n_dims = 4, name = mm.patch_merger.weight, tensor_size=50331648, offset=722255872, shape:[2, 2, 1536, 4096], type = f16
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[170]: n_dims = 2, name = v.position_embd.weight, tensor_size=3538944, offset=772587520, shape:[1536, 576, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[171]: n_dims = 2, name = mm.down.weight, tensor_size=59604992, offset=776126464, shape:[13696, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[172]: n_dims = 2, name = mm.gate.weight, tensor_size=59604992, offset=835731456, shape:[4096, 13696, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[173]: n_dims = 1, name = mm.post_norm.bias, tensor_size=16384, offset=895336448, shape:[4096, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[174]: n_dims = 1, name = mm.post_norm.weight, tensor_size=16384, offset=895352832, shape:[4096, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[175]: n_dims = 2, name = mm.model.fc.weight, tensor_size=17825792, offset=895369216, shape:[4096, 4096, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[176]: n_dims = 2, name = mm.up.weight, tensor_size=59604992, offset=913195008, shape:[4096, 13696, 1, 1], type = q8_0
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[177]: n_dims = 1, name = v.patch_embd.bias, tensor_size=6144, offset=972800000, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[178]: n_dims = 4, name = v.patch_embd.weight, tensor_size=3612672, offset=972806144, shape:[14, 14, 3, 1536], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[179]: n_dims = 4, name = v.patch_embd.weight.1, tensor_size=3612672, offset=976418816, shape:[14, 14, 3, 1536], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[180]: n_dims = 1, name = v.norm_embd.weight, tensor_size=6144, offset=980031488, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[181]: n_dims = 1, name = v.post_ln.weight, tensor_size=6144, offset=980037632, shape:[1536, 1, 1, 1], type = f32
1月 05 11:45:19 wythe ollama[1712]: clip_ctx: CLIP using CUDA0 backend
1月 05 11:45:19 wythe ollama[1712]: load_hparams: projector:          glm4v
1月 05 11:45:19 wythe ollama[1712]: load_hparams: n_embd:             1536
1月 05 11:45:19 wythe ollama[1712]: load_hparams: n_head:             12
1月 05 11:45:19 wythe ollama[1712]: load_hparams: n_ff:               13696
1月 05 11:45:19 wythe ollama[1712]: load_hparams: n_layer:            24
1月 05 11:45:19 wythe ollama[1712]: load_hparams: ffn_op:             silu
1月 05 11:45:19 wythe ollama[1712]: load_hparams: projection_dim:     4096
1月 05 11:45:19 wythe ollama[1712]: --- vision hparams ---
1月 05 11:45:19 wythe ollama[1712]: load_hparams: image_size:         336
1月 05 11:45:19 wythe ollama[1712]: load_hparams: patch_size:         14
1月 05 11:45:19 wythe ollama[1712]: load_hparams: has_llava_proj:     0
1月 05 11:45:19 wythe ollama[1712]: load_hparams: minicpmv_version:   0
1月 05 11:45:19 wythe ollama[1712]: load_hparams: n_merge:            2
1月 05 11:45:19 wythe ollama[1712]: load_hparams: n_wa_pattern:       0
1月 05 11:45:19 wythe ollama[1712]: load_hparams: image_min_pixels:   6272
1月 05 11:45:19 wythe ollama[1712]: load_hparams: image_max_pixels:   3211264
1月 05 11:45:19 wythe ollama[1712]: load_hparams: model size:         934.64 MiB
1月 05 11:45:19 wythe ollama[1712]: load_hparams: metadata size:      0.06 MiB
1月 05 11:45:19 wythe ollama[1712]: load_tensors: loaded 182 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-454dd441c925c1a81c984204bb9d54feef0ef07b789c6fe1118099014ba2727d
1月 05 11:45:19 wythe ollama[1712]: warmup: warmup with image size = 1288 x 1288
1月 05 11:45:19 wythe ollama[1712]: alloc_compute_meta:      CUDA0 compute buffer size =   515.05 MiB
1月 05 11:45:19 wythe ollama[1712]: alloc_compute_meta:        CPU compute buffer size =    19.11 MiB
1月 05 11:45:19 wythe ollama[1712]: alloc_compute_meta: graph splits = 1, nodes = 632
1月 05 11:45:19 wythe ollama[1712]: warmup: flash attention is enabled
1月 05 11:45:20 wythe ollama[1712]: time=2026-01-05T11:45:20.171Z level=INFO source=server.go:1376 msg="llama runner started in 2.08 seconds"
1月 05 11:45:20 wythe ollama[1712]: time=2026-01-05T11:45:20.171Z level=INFO source=sched.go:517 msg="loaded runners" count=1
1月 05 11:45:20 wythe ollama[1712]: time=2026-01-05T11:45:20.171Z level=INFO source=server.go:1338 msg="waiting for llama runner to start responding"
1月 05 11:45:20 wythe ollama[1712]: time=2026-01-05T11:45:20.171Z level=INFO source=server.go:1376 msg="llama runner started in 2.08 seconds"
1月 05 11:45:20 wythe ollama[1712]: add_text: <|begin_of_image|>
1月 05 11:45:20 wythe ollama[1712]: image_tokens->nx = 39
1月 05 11:45:20 wythe ollama[1712]: image_tokens->ny = 86
1月 05 11:45:20 wythe ollama[1712]: batch_f32 size = 1
1月 05 11:45:20 wythe ollama[1712]: add_text: <|end_of_image|>
1月 05 11:45:49 wythe ollama[1712]: [GIN] 2026/01/05 - 11:45:49 | 200 | 31.946132842s |  192.168.244.50 | POST     "/v1/chat/completions"
<!-- gh-comment-id:3709572719 --> @babysource commented on GitHub (Jan 5, 2026): Step: 1. exec: $ ollama run www.modelscope.cn/ggml-org/AutoGLM-Phone-9B-GGUF:Q4_K_M 2. Rename the model “autoglm-phone:9b” 3. git clone and install https://github.com/zai-org/Open-AutoGLM 4. exec: $ python scripts/check_deployment_cn.py --base-url http://192.168.244.40:11434/v1 --model "autoglm-phone:9b" 5. ollama api always response content “<\/answer>” However, if the images in the script<scripts/check_deployment_cn.py> are removed, a normal response will be returned. @rick-github $ journalctl -u ollama --since "1 hour ago" Full log ``` 1月 05 11:45:17 wythe ollama[1712]: time=2026-01-05T11:45:17.590Z level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 39397" 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: loaded meta data with 33 key-value pairs and 523 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20a>1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 0: general.architecture str = glm4 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 1: general.type str = model 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 2: general.size_label str = 9.4B 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 3: general.license str = mit 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 4: general.base_model.count u32 = 1 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 5: general.base_model.0.name str = GLM 4.1V 9B Base 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 6: general.base_model.0.organization str = Zai Org 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 7: general.base_model.0.repo_url str = https://huggingface.co/zai-org/GLM-4.... 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 8: general.tags arr[str,2] = ["agent", "image-text-to-text"] 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 9: general.languages arr[str,1] = ["zh"] 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 10: glm4.block_count u32 = 40 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 11: glm4.context_length u32 = 65536 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 12: glm4.embedding_length u32 = 4096 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 13: glm4.feed_forward_length u32 = 13696 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 14: glm4.attention.head_count u32 = 32 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 15: glm4.attention.head_count_kv u32 = 2 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 16: glm4.rope.dimension_sections arr[i32,4] = [8, 12, 12, 0] 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 17: glm4.rope.freq_base f32 = 10000.000000 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 18: glm4.attention.layer_norm_rms_epsilon f32 = 0.000010 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 19: glm4.rope.dimension_count u32 = 64 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 20: tokenizer.ggml.model str = gpt2 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 21: tokenizer.ggml.pre str = glm4 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 22: tokenizer.ggml.tokens arr[str,151552] = ["!", "\"", "#", "$", "%", "&", "'", ... 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 23: tokenizer.ggml.token_type arr[i32,151552] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 24: tokenizer.ggml.merges arr[str,318088] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "... 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 25: tokenizer.ggml.eos_token_id u32 = 151329 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 26: tokenizer.ggml.padding_token_id u32 = 151329 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 27: tokenizer.ggml.eot_token_id u32 = 151336 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 28: tokenizer.ggml.unknown_token_id u32 = 151329 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 29: tokenizer.ggml.bos_token_id u32 = 151329 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 30: tokenizer.chat_template str = [gMASK]<sop>\n{%- for msg in messages ... 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 31: general.quantization_version u32 = 2 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - kv 32: general.file_type u32 = 15 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - type f32: 281 tensors 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - type q5_0: 20 tensors 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - type q8_0: 20 tensors 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - type q4_K: 181 tensors 1月 05 11:45:17 wythe ollama[1712]: llama_model_loader: - type q6_K: 21 tensors 1月 05 11:45:17 wythe ollama[1712]: print_info: file format = GGUF V3 (latest) 1月 05 11:45:17 wythe ollama[1712]: print_info: file type = Q4_K - Medium 1月 05 11:45:17 wythe ollama[1712]: print_info: file size = 5.73 GiB (5.24 BPW) 1月 05 11:45:18 wythe ollama[1712]: load: special_eot_id is not in special_eog_ids - the tokenizer config may be incorrect 1月 05 11:45:18 wythe ollama[1712]: load: printing all EOG tokens: 1月 05 11:45:18 wythe ollama[1712]: load: - 151329 ('<|endoftext|>') 1月 05 11:45:18 wythe ollama[1712]: load: - 151336 ('<|user|>') 1月 05 11:45:18 wythe ollama[1712]: load: special tokens cache size = 23 1月 05 11:45:18 wythe ollama[1712]: load: token to piece cache size = 0.9711 MB 1月 05 11:45:18 wythe ollama[1712]: print_info: arch = glm4 1月 05 11:45:18 wythe ollama[1712]: print_info: vocab_only = 1 1月 05 11:45:18 wythe ollama[1712]: print_info: no_alloc = 0 1月 05 11:45:18 wythe ollama[1712]: print_info: model type = ?B 1月 05 11:45:18 wythe ollama[1712]: print_info: model params = 9.40 B 1月 05 11:45:18 wythe ollama[1712]: print_info: general.name = n/a 1月 05 11:45:18 wythe ollama[1712]: print_info: vocab type = BPE 1月 05 11:45:18 wythe ollama[1712]: print_info: n_vocab = 151552 1月 05 11:45:18 wythe ollama[1712]: print_info: n_merges = 318088 1月 05 11:45:18 wythe ollama[1712]: print_info: BOS token = 151329 '<|endoftext|>' 1月 05 11:45:18 wythe ollama[1712]: print_info: EOS token = 151329 '<|endoftext|>' 1月 05 11:45:18 wythe ollama[1712]: print_info: EOT token = 151336 '<|user|>' 1月 05 11:45:18 wythe ollama[1712]: print_info: UNK token = 151329 '<|endoftext|>' 1月 05 11:45:18 wythe ollama[1712]: print_info: PAD token = 151329 '<|endoftext|>' 1月 05 11:45:18 wythe ollama[1712]: print_info: LF token = 198 'Ċ' 1月 05 11:45:18 wythe ollama[1712]: print_info: EOG token = 151329 '<|endoftext|>' 1月 05 11:45:18 wythe ollama[1712]: print_info: EOG token = 151336 '<|user|>' 1月 05 11:45:18 wythe ollama[1712]: print_info: max token length = 1024 1月 05 11:45:18 wythe ollama[1712]: llama_model_load: vocab only - skipping tensors 1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.087Z level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --model /usr/share/ollama/.ollama/models> 1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.088Z level=INFO source=sched.go:443 msg="system memory" total="15.5 GiB" free="12.5 GiB" free_swap="4.0 GiB" 1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.088Z level=INFO source=sched.go:450 msg="gpu memory" id=GPU-f74419c7-e9cb-f6fb-75c3-739c9fdc77e8 library=CUDA available="11.5 GiB" fr> 1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.088Z level=INFO source=server.go:496 msg="loading model" "model layers"=41 requested=-1 1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.088Z level=INFO source=device.go:240 msg="model weights" device=CUDA0 size="5.4 GiB" 1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.088Z level=INFO source=device.go:251 msg="kv cache" device=CUDA0 size="160.0 MiB" 1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.088Z level=INFO source=device.go:262 msg="compute graph" device=CUDA0 size="426.7 MiB" 1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.088Z level=INFO source=device.go:272 msg="total memory" size="6.0 GiB" 1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.098Z level=INFO source=runner.go:965 msg="starting go runner" 1月 05 11:45:18 wythe ollama[1712]: load_backend: loaded CPU backend from /usr/local/lib/ollama/libggml-cpu-haswell.so 1月 05 11:45:18 wythe ollama[1712]: ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no 1月 05 11:45:18 wythe ollama[1712]: ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no 1月 05 11:45:18 wythe ollama[1712]: ggml_cuda_init: found 1 CUDA devices: 1月 05 11:45:18 wythe ollama[1712]: Device 0: NVIDIA GeForce GTX TITAN X, compute capability 5.2, VMM: yes, ID: GPU-f74419c7-e9cb-f6fb-75c3-739c9fdc77e8 1月 05 11:45:18 wythe ollama[1712]: load_backend: loaded CUDA backend from /usr/local/lib/ollama/cuda_v12/libggml-cuda.so 1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.152Z level=INFO source=ggml.go:104 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.BMI2> 1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.152Z level=INFO source=runner.go:1001 msg="Server listening on 127.0.0.1:44695" 1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.163Z level=INFO source=runner.go:895 msg=load request="{Operation:commit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:Auto KvS> 1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.163Z level=INFO source=server.go:1338 msg="waiting for llama runner to start responding" 1月 05 11:45:18 wythe ollama[1712]: time=2026-01-05T11:45:18.164Z level=INFO source=server.go:1372 msg="waiting for server to become available" status="llm server loading model" 1月 05 11:45:18 wythe ollama[1712]: ggml_backend_cuda_device_get_memory device GPU-f74419c7-e9cb-f6fb-75c3-739c9fdc77e8 utilizing NVML memory reporting free: 12778471424 total: 12884901888 1月 05 11:45:18 wythe ollama[1712]: llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce GTX TITAN X) (0000:01:00.0) - 12186 MiB free 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: loaded meta data with 33 key-value pairs and 523 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20a> 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 0: general.architecture str = glm4 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 1: general.type str = model 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 2: general.size_label str = 9.4B 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 3: general.license str = mit 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 4: general.base_model.count u32 = 1 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 5: general.base_model.0.name str = GLM 4.1V 9B Base 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 6: general.base_model.0.organization str = Zai Org 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 7: general.base_model.0.repo_url str = https://huggingface.co/zai-org/GLM-4.... 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 8: general.tags arr[str,2] = ["agent", "image-text-to-text"] 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 9: general.languages arr[str,1] = ["zh"] 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 10: glm4.block_count u32 = 40 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 11: glm4.context_length u32 = 65536 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 12: glm4.embedding_length u32 = 4096 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 13: glm4.feed_forward_length u32 = 13696 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 14: glm4.attention.head_count u32 = 32 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 15: glm4.attention.head_count_kv u32 = 2 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 16: glm4.rope.dimension_sections arr[i32,4] = [8, 12, 12, 0] 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 17: glm4.rope.freq_base f32 = 10000.000000 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 18: glm4.attention.layer_norm_rms_epsilon f32 = 0.000010 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 19: glm4.rope.dimension_count u32 = 64 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 20: tokenizer.ggml.model str = gpt2 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 21: tokenizer.ggml.pre str = glm4 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 22: tokenizer.ggml.tokens arr[str,151552] = ["!", "\"", "#", "$", "%", "&", "'", ... 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 23: tokenizer.ggml.token_type arr[i32,151552] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 24: tokenizer.ggml.merges arr[str,318088] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "... 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 25: tokenizer.ggml.eos_token_id u32 = 151329 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 26: tokenizer.ggml.padding_token_id u32 = 151329 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 27: tokenizer.ggml.eot_token_id u32 = 151336 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 28: tokenizer.ggml.unknown_token_id u32 = 151329 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 29: tokenizer.ggml.bos_token_id u32 = 151329 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 30: tokenizer.chat_template str = [gMASK]<sop>\n{%- for msg in messages ... 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 31: general.quantization_version u32 = 2 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - kv 32: general.file_type u32 = 15 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - type f32: 281 tensors 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - type q5_0: 20 tensors 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - type q8_0: 20 tensors 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - type q4_K: 181 tensors 1月 05 11:45:18 wythe ollama[1712]: llama_model_loader: - type q6_K: 21 tensors 1月 05 11:45:18 wythe ollama[1712]: print_info: file format = GGUF V3 (latest) 1月 05 11:45:18 wythe ollama[1712]: print_info: file type = Q4_K - Medium 1月 05 11:45:18 wythe ollama[1712]: print_info: file size = 5.73 GiB (5.24 BPW) 1月 05 11:45:18 wythe ollama[1712]: load: special_eot_id is not in special_eog_ids - the tokenizer config may be incorrect 1月 05 11:45:18 wythe ollama[1712]: load: printing all EOG tokens: 1月 05 11:45:18 wythe ollama[1712]: load: - 151329 ('<|endoftext|>') 1月 05 11:45:18 wythe ollama[1712]: load: - 151336 ('<|user|>') 1月 05 11:45:18 wythe ollama[1712]: load: special tokens cache size = 23 1月 05 11:45:18 wythe ollama[1712]: load: token to piece cache size = 0.9711 MB 1月 05 11:45:18 wythe ollama[1712]: print_info: arch = glm4 1月 05 11:45:18 wythe ollama[1712]: print_info: vocab_only = 0 1月 05 11:45:18 wythe ollama[1712]: print_info: no_alloc = 0 1月 05 11:45:18 wythe ollama[1712]: print_info: n_ctx_train = 65536 1月 05 11:45:18 wythe ollama[1712]: print_info: n_embd = 4096 1月 05 11:45:18 wythe ollama[1712]: print_info: n_embd_inp = 4096 1月 05 11:45:18 wythe ollama[1712]: print_info: n_layer = 40 1月 05 11:45:18 wythe ollama[1712]: print_info: n_head = 32 1月 05 11:45:18 wythe ollama[1712]: print_info: n_head_kv = 2 1月 05 11:45:18 wythe ollama[1712]: print_info: n_rot = 64 1月 05 11:45:18 wythe ollama[1712]: print_info: n_swa = 0 1月 05 11:45:18 wythe ollama[1712]: print_info: is_swa_any = 0 1月 05 11:45:18 wythe ollama[1712]: print_info: n_embd_head_k = 128 1月 05 11:45:18 wythe ollama[1712]: print_info: n_embd_head_v = 128 1月 05 11:45:18 wythe ollama[1712]: print_info: n_gqa = 16 1月 05 11:45:18 wythe ollama[1712]: print_info: n_embd_k_gqa = 256 1月 05 11:45:18 wythe ollama[1712]: print_info: n_embd_v_gqa = 256 1月 05 11:45:18 wythe ollama[1712]: print_info: f_norm_eps = 0.0e+00 1月 05 11:45:18 wythe ollama[1712]: print_info: f_norm_rms_eps = 1.0e-05 1月 05 11:45:18 wythe ollama[1712]: print_info: f_clamp_kqv = 0.0e+00 1月 05 11:45:18 wythe ollama[1712]: print_info: f_max_alibi_bias = 0.0e+00 1月 05 11:45:18 wythe ollama[1712]: print_info: f_logit_scale = 0.0e+00 1月 05 11:45:18 wythe ollama[1712]: print_info: f_attn_scale = 0.0e+00 1月 05 11:45:18 wythe ollama[1712]: print_info: n_ff = 13696 1月 05 11:45:18 wythe ollama[1712]: print_info: n_expert = 0 1月 05 11:45:18 wythe ollama[1712]: print_info: n_expert_used = 0 1月 05 11:45:18 wythe ollama[1712]: print_info: n_expert_groups = 0 1月 05 11:45:18 wythe ollama[1712]: print_info: n_group_used = 0 1月 05 11:45:18 wythe ollama[1712]: print_info: causal attn = 1 1月 05 11:45:18 wythe ollama[1712]: print_info: pooling type = 0 1月 05 11:45:18 wythe ollama[1712]: print_info: rope type = 8 1月 05 11:45:18 wythe ollama[1712]: print_info: rope scaling = linear 1月 05 11:45:18 wythe ollama[1712]: print_info: freq_base_train = 10000.0 1月 05 11:45:18 wythe ollama[1712]: print_info: freq_scale_train = 1 1月 05 11:45:18 wythe ollama[1712]: print_info: n_ctx_orig_yarn = 65536 1月 05 11:45:18 wythe ollama[1712]: print_info: rope_yarn_log_mul= 0.0000 1月 05 11:45:18 wythe ollama[1712]: print_info: rope_finetuned = unknown 1月 05 11:45:18 wythe ollama[1712]: print_info: mrope sections = [8, 12, 12, 0] 1月 05 11:45:18 wythe ollama[1712]: print_info: model type = 9B 1月 05 11:45:18 wythe ollama[1712]: print_info: model params = 9.40 B 1月 05 11:45:18 wythe ollama[1712]: print_info: general.name = n/a 1月 05 11:45:18 wythe ollama[1712]: print_info: vocab type = BPE 1月 05 11:45:18 wythe ollama[1712]: print_info: n_vocab = 151552 1月 05 11:45:18 wythe ollama[1712]: print_info: n_merges = 318088 1月 05 11:45:18 wythe ollama[1712]: print_info: BOS token = 151329 '<|endoftext|>' 1月 05 11:45:18 wythe ollama[1712]: print_info: EOS token = 151329 '<|endoftext|>' 1月 05 11:45:18 wythe ollama[1712]: print_info: EOT token = 151336 '<|user|>' 1月 05 11:45:18 wythe ollama[1712]: print_info: UNK token = 151329 '<|endoftext|>' 1月 05 11:45:18 wythe ollama[1712]: print_info: PAD token = 151329 '<|endoftext|>' 1月 05 11:45:18 wythe ollama[1712]: print_info: LF token = 198 'Ċ' 1月 05 11:45:18 wythe ollama[1712]: print_info: EOG token = 151329 '<|endoftext|>' 1月 05 11:45:18 wythe ollama[1712]: print_info: EOG token = 151336 '<|user|>' 1月 05 11:45:18 wythe ollama[1712]: print_info: max token length = 1024 1月 05 11:45:18 wythe ollama[1712]: load_tensors: loading model tensors, this can take a while... (mmap = true) 1月 05 11:45:18 wythe ollama[1712]: load_tensors: offloading 40 repeating layers to GPU 1月 05 11:45:18 wythe ollama[1712]: load_tensors: offloading output layer to GPU 1月 05 11:45:18 wythe ollama[1712]: load_tensors: offloaded 41/41 layers to GPU 1月 05 11:45:18 wythe ollama[1712]: load_tensors: CPU_Mapped model buffer size = 333.00 MiB 1月 05 11:45:18 wythe ollama[1712]: load_tensors: CUDA0 model buffer size = 5539.01 MiB 1月 05 11:45:19 wythe ollama[1712]: llama_context: constructing llama_context 1月 05 11:45:19 wythe ollama[1712]: llama_context: n_seq_max = 1 1月 05 11:45:19 wythe ollama[1712]: llama_context: n_ctx = 4096 1月 05 11:45:19 wythe ollama[1712]: llama_context: n_ctx_seq = 4096 1月 05 11:45:19 wythe ollama[1712]: llama_context: n_batch = 512 1月 05 11:45:19 wythe ollama[1712]: llama_context: n_ubatch = 512 1月 05 11:45:19 wythe ollama[1712]: llama_context: causal_attn = 1 1月 05 11:45:19 wythe ollama[1712]: llama_context: flash_attn = auto 1月 05 11:45:19 wythe ollama[1712]: llama_context: kv_unified = false 1月 05 11:45:19 wythe ollama[1712]: llama_context: freq_base = 10000.0 1月 05 11:45:19 wythe ollama[1712]: llama_context: freq_scale = 1 1月 05 11:45:19 wythe ollama[1712]: llama_context: n_ctx_seq (4096) < n_ctx_train (65536) -- the full capacity of the model will not be utilized 1月 05 11:45:19 wythe ollama[1712]: llama_context: CUDA_Host output buffer size = 0.59 MiB 1月 05 11:45:19 wythe ollama[1712]: llama_kv_cache: CUDA0 KV buffer size = 160.00 MiB 1月 05 11:45:19 wythe ollama[1712]: llama_kv_cache: size = 160.00 MiB ( 4096 cells, 40 layers, 1/1 seqs), K (f16): 80.00 MiB, V (f16): 80.00 MiB 1月 05 11:45:19 wythe ollama[1712]: llama_context: Flash Attention was auto, set to enabled 1月 05 11:45:19 wythe ollama[1712]: llama_context: CUDA0 compute buffer size = 304.00 MiB 1月 05 11:45:19 wythe ollama[1712]: llama_context: CUDA_Host compute buffer size = 16.02 MiB 1月 05 11:45:19 wythe ollama[1712]: llama_context: graph nodes = 1487 1月 05 11:45:19 wythe ollama[1712]: llama_context: graph splits = 2 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: model name: 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: description: 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: GGUF version: 3 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: alignment: 32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: n_tensors: 182 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: n_kv: 25 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: has vision encoder 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[0]: n_dims = 2, name = v.blk.0.attn_out.weight, tensor_size=2506752, offset=0, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[1]: n_dims = 2, name = v.blk.0.attn_qkv.weight, tensor_size=7520256, offset=2506752, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[2]: n_dims = 2, name = v.blk.0.ffn_down.weight, tensor_size=6684672, offset=10027008, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[3]: n_dims = 2, name = v.blk.0.ffn_gate.weight, tensor_size=6684672, offset=16711680, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[4]: n_dims = 2, name = v.blk.0.ffn_up.weight, tensor_size=6684672, offset=23396352, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[5]: n_dims = 1, name = v.blk.0.ln1.weight, tensor_size=6144, offset=30081024, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[6]: n_dims = 1, name = v.blk.0.ln2.weight, tensor_size=6144, offset=30087168, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[7]: n_dims = 2, name = v.blk.1.attn_out.weight, tensor_size=2506752, offset=30093312, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[8]: n_dims = 2, name = v.blk.1.attn_qkv.weight, tensor_size=7520256, offset=32600064, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[9]: n_dims = 2, name = v.blk.1.ffn_down.weight, tensor_size=6684672, offset=40120320, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[10]: n_dims = 2, name = v.blk.1.ffn_gate.weight, tensor_size=6684672, offset=46804992, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[11]: n_dims = 2, name = v.blk.1.ffn_up.weight, tensor_size=6684672, offset=53489664, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[12]: n_dims = 1, name = v.blk.1.ln1.weight, tensor_size=6144, offset=60174336, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[13]: n_dims = 1, name = v.blk.1.ln2.weight, tensor_size=6144, offset=60180480, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[14]: n_dims = 2, name = v.blk.10.attn_out.weight, tensor_size=2506752, offset=60186624, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[15]: n_dims = 2, name = v.blk.10.attn_qkv.weight, tensor_size=7520256, offset=62693376, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[16]: n_dims = 2, name = v.blk.10.ffn_down.weight, tensor_size=6684672, offset=70213632, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[17]: n_dims = 2, name = v.blk.10.ffn_gate.weight, tensor_size=6684672, offset=76898304, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[18]: n_dims = 2, name = v.blk.10.ffn_up.weight, tensor_size=6684672, offset=83582976, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[19]: n_dims = 1, name = v.blk.10.ln1.weight, tensor_size=6144, offset=90267648, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[20]: n_dims = 1, name = v.blk.10.ln2.weight, tensor_size=6144, offset=90273792, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[21]: n_dims = 2, name = v.blk.11.attn_out.weight, tensor_size=2506752, offset=90279936, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[22]: n_dims = 2, name = v.blk.11.attn_qkv.weight, tensor_size=7520256, offset=92786688, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[23]: n_dims = 2, name = v.blk.11.ffn_down.weight, tensor_size=6684672, offset=100306944, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[24]: n_dims = 2, name = v.blk.11.ffn_gate.weight, tensor_size=6684672, offset=106991616, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[25]: n_dims = 2, name = v.blk.11.ffn_up.weight, tensor_size=6684672, offset=113676288, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[26]: n_dims = 1, name = v.blk.11.ln1.weight, tensor_size=6144, offset=120360960, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[27]: n_dims = 1, name = v.blk.11.ln2.weight, tensor_size=6144, offset=120367104, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[28]: n_dims = 2, name = v.blk.12.attn_out.weight, tensor_size=2506752, offset=120373248, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[29]: n_dims = 2, name = v.blk.12.attn_qkv.weight, tensor_size=7520256, offset=122880000, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[30]: n_dims = 2, name = v.blk.12.ffn_down.weight, tensor_size=6684672, offset=130400256, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[31]: n_dims = 2, name = v.blk.12.ffn_gate.weight, tensor_size=6684672, offset=137084928, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[32]: n_dims = 2, name = v.blk.12.ffn_up.weight, tensor_size=6684672, offset=143769600, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[33]: n_dims = 1, name = v.blk.12.ln1.weight, tensor_size=6144, offset=150454272, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[34]: n_dims = 1, name = v.blk.12.ln2.weight, tensor_size=6144, offset=150460416, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[35]: n_dims = 2, name = v.blk.13.attn_out.weight, tensor_size=2506752, offset=150466560, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[36]: n_dims = 2, name = v.blk.13.attn_qkv.weight, tensor_size=7520256, offset=152973312, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[37]: n_dims = 2, name = v.blk.13.ffn_down.weight, tensor_size=6684672, offset=160493568, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[38]: n_dims = 2, name = v.blk.13.ffn_gate.weight, tensor_size=6684672, offset=167178240, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[39]: n_dims = 2, name = v.blk.13.ffn_up.weight, tensor_size=6684672, offset=173862912, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[40]: n_dims = 1, name = v.blk.13.ln1.weight, tensor_size=6144, offset=180547584, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[41]: n_dims = 1, name = v.blk.13.ln2.weight, tensor_size=6144, offset=180553728, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[42]: n_dims = 2, name = v.blk.14.attn_out.weight, tensor_size=2506752, offset=180559872, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[43]: n_dims = 2, name = v.blk.14.attn_qkv.weight, tensor_size=7520256, offset=183066624, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[44]: n_dims = 2, name = v.blk.14.ffn_down.weight, tensor_size=6684672, offset=190586880, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[45]: n_dims = 2, name = v.blk.14.ffn_gate.weight, tensor_size=6684672, offset=197271552, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[46]: n_dims = 2, name = v.blk.14.ffn_up.weight, tensor_size=6684672, offset=203956224, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[47]: n_dims = 1, name = v.blk.14.ln1.weight, tensor_size=6144, offset=210640896, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[48]: n_dims = 1, name = v.blk.14.ln2.weight, tensor_size=6144, offset=210647040, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[49]: n_dims = 2, name = v.blk.15.attn_out.weight, tensor_size=2506752, offset=210653184, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[50]: n_dims = 2, name = v.blk.15.attn_qkv.weight, tensor_size=7520256, offset=213159936, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[51]: n_dims = 2, name = v.blk.15.ffn_down.weight, tensor_size=6684672, offset=220680192, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[52]: n_dims = 2, name = v.blk.15.ffn_gate.weight, tensor_size=6684672, offset=227364864, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[53]: n_dims = 2, name = v.blk.15.ffn_up.weight, tensor_size=6684672, offset=234049536, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[54]: n_dims = 1, name = v.blk.15.ln1.weight, tensor_size=6144, offset=240734208, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[55]: n_dims = 1, name = v.blk.15.ln2.weight, tensor_size=6144, offset=240740352, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[56]: n_dims = 2, name = v.blk.16.attn_out.weight, tensor_size=2506752, offset=240746496, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[57]: n_dims = 2, name = v.blk.16.attn_qkv.weight, tensor_size=7520256, offset=243253248, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[58]: n_dims = 2, name = v.blk.16.ffn_down.weight, tensor_size=6684672, offset=250773504, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[59]: n_dims = 2, name = v.blk.16.ffn_gate.weight, tensor_size=6684672, offset=257458176, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[60]: n_dims = 2, name = v.blk.16.ffn_up.weight, tensor_size=6684672, offset=264142848, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[61]: n_dims = 1, name = v.blk.16.ln1.weight, tensor_size=6144, offset=270827520, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[62]: n_dims = 1, name = v.blk.16.ln2.weight, tensor_size=6144, offset=270833664, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[63]: n_dims = 2, name = v.blk.17.attn_out.weight, tensor_size=2506752, offset=270839808, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[64]: n_dims = 2, name = v.blk.17.attn_qkv.weight, tensor_size=7520256, offset=273346560, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[65]: n_dims = 2, name = v.blk.17.ffn_down.weight, tensor_size=6684672, offset=280866816, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[66]: n_dims = 2, name = v.blk.17.ffn_gate.weight, tensor_size=6684672, offset=287551488, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[67]: n_dims = 2, name = v.blk.17.ffn_up.weight, tensor_size=6684672, offset=294236160, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[68]: n_dims = 1, name = v.blk.17.ln1.weight, tensor_size=6144, offset=300920832, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[69]: n_dims = 1, name = v.blk.17.ln2.weight, tensor_size=6144, offset=300926976, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[70]: n_dims = 2, name = v.blk.18.attn_out.weight, tensor_size=2506752, offset=300933120, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[71]: n_dims = 2, name = v.blk.18.attn_qkv.weight, tensor_size=7520256, offset=303439872, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[72]: n_dims = 2, name = v.blk.18.ffn_down.weight, tensor_size=6684672, offset=310960128, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[73]: n_dims = 2, name = v.blk.18.ffn_gate.weight, tensor_size=6684672, offset=317644800, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[74]: n_dims = 2, name = v.blk.18.ffn_up.weight, tensor_size=6684672, offset=324329472, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[75]: n_dims = 1, name = v.blk.18.ln1.weight, tensor_size=6144, offset=331014144, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[76]: n_dims = 1, name = v.blk.18.ln2.weight, tensor_size=6144, offset=331020288, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[77]: n_dims = 2, name = v.blk.19.attn_out.weight, tensor_size=2506752, offset=331026432, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[78]: n_dims = 2, name = v.blk.19.attn_qkv.weight, tensor_size=7520256, offset=333533184, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[79]: n_dims = 2, name = v.blk.19.ffn_down.weight, tensor_size=6684672, offset=341053440, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[80]: n_dims = 2, name = v.blk.19.ffn_gate.weight, tensor_size=6684672, offset=347738112, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[81]: n_dims = 2, name = v.blk.19.ffn_up.weight, tensor_size=6684672, offset=354422784, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[82]: n_dims = 1, name = v.blk.19.ln1.weight, tensor_size=6144, offset=361107456, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[83]: n_dims = 1, name = v.blk.19.ln2.weight, tensor_size=6144, offset=361113600, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[84]: n_dims = 2, name = v.blk.2.attn_out.weight, tensor_size=2506752, offset=361119744, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[85]: n_dims = 2, name = v.blk.2.attn_qkv.weight, tensor_size=7520256, offset=363626496, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[86]: n_dims = 2, name = v.blk.2.ffn_down.weight, tensor_size=6684672, offset=371146752, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[87]: n_dims = 2, name = v.blk.2.ffn_gate.weight, tensor_size=6684672, offset=377831424, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[88]: n_dims = 2, name = v.blk.2.ffn_up.weight, tensor_size=6684672, offset=384516096, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[89]: n_dims = 1, name = v.blk.2.ln1.weight, tensor_size=6144, offset=391200768, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[90]: n_dims = 1, name = v.blk.2.ln2.weight, tensor_size=6144, offset=391206912, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[91]: n_dims = 2, name = v.blk.20.attn_out.weight, tensor_size=2506752, offset=391213056, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[92]: n_dims = 2, name = v.blk.20.attn_qkv.weight, tensor_size=7520256, offset=393719808, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[93]: n_dims = 2, name = v.blk.20.ffn_down.weight, tensor_size=6684672, offset=401240064, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[94]: n_dims = 2, name = v.blk.20.ffn_gate.weight, tensor_size=6684672, offset=407924736, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[95]: n_dims = 2, name = v.blk.20.ffn_up.weight, tensor_size=6684672, offset=414609408, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[96]: n_dims = 1, name = v.blk.20.ln1.weight, tensor_size=6144, offset=421294080, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[97]: n_dims = 1, name = v.blk.20.ln2.weight, tensor_size=6144, offset=421300224, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[98]: n_dims = 2, name = v.blk.21.attn_out.weight, tensor_size=2506752, offset=421306368, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[99]: n_dims = 2, name = v.blk.21.attn_qkv.weight, tensor_size=7520256, offset=423813120, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[100]: n_dims = 2, name = v.blk.21.ffn_down.weight, tensor_size=6684672, offset=431333376, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[101]: n_dims = 2, name = v.blk.21.ffn_gate.weight, tensor_size=6684672, offset=438018048, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[102]: n_dims = 2, name = v.blk.21.ffn_up.weight, tensor_size=6684672, offset=444702720, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[103]: n_dims = 1, name = v.blk.21.ln1.weight, tensor_size=6144, offset=451387392, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[104]: n_dims = 1, name = v.blk.21.ln2.weight, tensor_size=6144, offset=451393536, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[105]: n_dims = 2, name = v.blk.22.attn_out.weight, tensor_size=2506752, offset=451399680, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[106]: n_dims = 2, name = v.blk.22.attn_qkv.weight, tensor_size=7520256, offset=453906432, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[107]: n_dims = 2, name = v.blk.22.ffn_down.weight, tensor_size=6684672, offset=461426688, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[108]: n_dims = 2, name = v.blk.22.ffn_gate.weight, tensor_size=6684672, offset=468111360, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[109]: n_dims = 2, name = v.blk.22.ffn_up.weight, tensor_size=6684672, offset=474796032, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[110]: n_dims = 1, name = v.blk.22.ln1.weight, tensor_size=6144, offset=481480704, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[111]: n_dims = 1, name = v.blk.22.ln2.weight, tensor_size=6144, offset=481486848, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[112]: n_dims = 2, name = v.blk.23.attn_out.weight, tensor_size=2506752, offset=481492992, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[113]: n_dims = 2, name = v.blk.23.attn_qkv.weight, tensor_size=7520256, offset=483999744, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[114]: n_dims = 2, name = v.blk.23.ffn_down.weight, tensor_size=6684672, offset=491520000, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[115]: n_dims = 2, name = v.blk.23.ffn_gate.weight, tensor_size=6684672, offset=498204672, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[116]: n_dims = 2, name = v.blk.23.ffn_up.weight, tensor_size=6684672, offset=504889344, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[117]: n_dims = 1, name = v.blk.23.ln1.weight, tensor_size=6144, offset=511574016, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[118]: n_dims = 1, name = v.blk.23.ln2.weight, tensor_size=6144, offset=511580160, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[119]: n_dims = 2, name = v.blk.3.attn_out.weight, tensor_size=2506752, offset=511586304, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[120]: n_dims = 2, name = v.blk.3.attn_qkv.weight, tensor_size=7520256, offset=514093056, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[121]: n_dims = 2, name = v.blk.3.ffn_down.weight, tensor_size=6684672, offset=521613312, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[122]: n_dims = 2, name = v.blk.3.ffn_gate.weight, tensor_size=6684672, offset=528297984, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[123]: n_dims = 2, name = v.blk.3.ffn_up.weight, tensor_size=6684672, offset=534982656, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[124]: n_dims = 1, name = v.blk.3.ln1.weight, tensor_size=6144, offset=541667328, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[125]: n_dims = 1, name = v.blk.3.ln2.weight, tensor_size=6144, offset=541673472, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[126]: n_dims = 2, name = v.blk.4.attn_out.weight, tensor_size=2506752, offset=541679616, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[127]: n_dims = 2, name = v.blk.4.attn_qkv.weight, tensor_size=7520256, offset=544186368, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[128]: n_dims = 2, name = v.blk.4.ffn_down.weight, tensor_size=6684672, offset=551706624, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[129]: n_dims = 2, name = v.blk.4.ffn_gate.weight, tensor_size=6684672, offset=558391296, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[130]: n_dims = 2, name = v.blk.4.ffn_up.weight, tensor_size=6684672, offset=565075968, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[131]: n_dims = 1, name = v.blk.4.ln1.weight, tensor_size=6144, offset=571760640, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[132]: n_dims = 1, name = v.blk.4.ln2.weight, tensor_size=6144, offset=571766784, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[133]: n_dims = 2, name = v.blk.5.attn_out.weight, tensor_size=2506752, offset=571772928, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[134]: n_dims = 2, name = v.blk.5.attn_qkv.weight, tensor_size=7520256, offset=574279680, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[135]: n_dims = 2, name = v.blk.5.ffn_down.weight, tensor_size=6684672, offset=581799936, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[136]: n_dims = 2, name = v.blk.5.ffn_gate.weight, tensor_size=6684672, offset=588484608, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[137]: n_dims = 2, name = v.blk.5.ffn_up.weight, tensor_size=6684672, offset=595169280, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[138]: n_dims = 1, name = v.blk.5.ln1.weight, tensor_size=6144, offset=601853952, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[139]: n_dims = 1, name = v.blk.5.ln2.weight, tensor_size=6144, offset=601860096, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[140]: n_dims = 2, name = v.blk.6.attn_out.weight, tensor_size=2506752, offset=601866240, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[141]: n_dims = 2, name = v.blk.6.attn_qkv.weight, tensor_size=7520256, offset=604372992, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[142]: n_dims = 2, name = v.blk.6.ffn_down.weight, tensor_size=6684672, offset=611893248, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[143]: n_dims = 2, name = v.blk.6.ffn_gate.weight, tensor_size=6684672, offset=618577920, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[144]: n_dims = 2, name = v.blk.6.ffn_up.weight, tensor_size=6684672, offset=625262592, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[145]: n_dims = 1, name = v.blk.6.ln1.weight, tensor_size=6144, offset=631947264, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[146]: n_dims = 1, name = v.blk.6.ln2.weight, tensor_size=6144, offset=631953408, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[147]: n_dims = 2, name = v.blk.7.attn_out.weight, tensor_size=2506752, offset=631959552, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[148]: n_dims = 2, name = v.blk.7.attn_qkv.weight, tensor_size=7520256, offset=634466304, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[149]: n_dims = 2, name = v.blk.7.ffn_down.weight, tensor_size=6684672, offset=641986560, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[150]: n_dims = 2, name = v.blk.7.ffn_gate.weight, tensor_size=6684672, offset=648671232, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[151]: n_dims = 2, name = v.blk.7.ffn_up.weight, tensor_size=6684672, offset=655355904, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[152]: n_dims = 1, name = v.blk.7.ln1.weight, tensor_size=6144, offset=662040576, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[153]: n_dims = 1, name = v.blk.7.ln2.weight, tensor_size=6144, offset=662046720, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[154]: n_dims = 2, name = v.blk.8.attn_out.weight, tensor_size=2506752, offset=662052864, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[155]: n_dims = 2, name = v.blk.8.attn_qkv.weight, tensor_size=7520256, offset=664559616, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[156]: n_dims = 2, name = v.blk.8.ffn_down.weight, tensor_size=6684672, offset=672079872, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[157]: n_dims = 2, name = v.blk.8.ffn_gate.weight, tensor_size=6684672, offset=678764544, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[158]: n_dims = 2, name = v.blk.8.ffn_up.weight, tensor_size=6684672, offset=685449216, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[159]: n_dims = 1, name = v.blk.8.ln1.weight, tensor_size=6144, offset=692133888, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[160]: n_dims = 1, name = v.blk.8.ln2.weight, tensor_size=6144, offset=692140032, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[161]: n_dims = 2, name = v.blk.9.attn_out.weight, tensor_size=2506752, offset=692146176, shape:[1536, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[162]: n_dims = 2, name = v.blk.9.attn_qkv.weight, tensor_size=7520256, offset=694652928, shape:[1536, 4608, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[163]: n_dims = 2, name = v.blk.9.ffn_down.weight, tensor_size=6684672, offset=702173184, shape:[4096, 1536, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[164]: n_dims = 2, name = v.blk.9.ffn_gate.weight, tensor_size=6684672, offset=708857856, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[165]: n_dims = 2, name = v.blk.9.ffn_up.weight, tensor_size=6684672, offset=715542528, shape:[1536, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[166]: n_dims = 1, name = v.blk.9.ln1.weight, tensor_size=6144, offset=722227200, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[167]: n_dims = 1, name = v.blk.9.ln2.weight, tensor_size=6144, offset=722233344, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[168]: n_dims = 1, name = mm.patch_merger.bias, tensor_size=16384, offset=722239488, shape:[4096, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[169]: n_dims = 4, name = mm.patch_merger.weight, tensor_size=50331648, offset=722255872, shape:[2, 2, 1536, 4096], type = f16 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[170]: n_dims = 2, name = v.position_embd.weight, tensor_size=3538944, offset=772587520, shape:[1536, 576, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[171]: n_dims = 2, name = mm.down.weight, tensor_size=59604992, offset=776126464, shape:[13696, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[172]: n_dims = 2, name = mm.gate.weight, tensor_size=59604992, offset=835731456, shape:[4096, 13696, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[173]: n_dims = 1, name = mm.post_norm.bias, tensor_size=16384, offset=895336448, shape:[4096, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[174]: n_dims = 1, name = mm.post_norm.weight, tensor_size=16384, offset=895352832, shape:[4096, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[175]: n_dims = 2, name = mm.model.fc.weight, tensor_size=17825792, offset=895369216, shape:[4096, 4096, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[176]: n_dims = 2, name = mm.up.weight, tensor_size=59604992, offset=913195008, shape:[4096, 13696, 1, 1], type = q8_0 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[177]: n_dims = 1, name = v.patch_embd.bias, tensor_size=6144, offset=972800000, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[178]: n_dims = 4, name = v.patch_embd.weight, tensor_size=3612672, offset=972806144, shape:[14, 14, 3, 1536], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[179]: n_dims = 4, name = v.patch_embd.weight.1, tensor_size=3612672, offset=976418816, shape:[14, 14, 3, 1536], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[180]: n_dims = 1, name = v.norm_embd.weight, tensor_size=6144, offset=980031488, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_model_loader: tensor[181]: n_dims = 1, name = v.post_ln.weight, tensor_size=6144, offset=980037632, shape:[1536, 1, 1, 1], type = f32 1月 05 11:45:19 wythe ollama[1712]: clip_ctx: CLIP using CUDA0 backend 1月 05 11:45:19 wythe ollama[1712]: load_hparams: projector: glm4v 1月 05 11:45:19 wythe ollama[1712]: load_hparams: n_embd: 1536 1月 05 11:45:19 wythe ollama[1712]: load_hparams: n_head: 12 1月 05 11:45:19 wythe ollama[1712]: load_hparams: n_ff: 13696 1月 05 11:45:19 wythe ollama[1712]: load_hparams: n_layer: 24 1月 05 11:45:19 wythe ollama[1712]: load_hparams: ffn_op: silu 1月 05 11:45:19 wythe ollama[1712]: load_hparams: projection_dim: 4096 1月 05 11:45:19 wythe ollama[1712]: --- vision hparams --- 1月 05 11:45:19 wythe ollama[1712]: load_hparams: image_size: 336 1月 05 11:45:19 wythe ollama[1712]: load_hparams: patch_size: 14 1月 05 11:45:19 wythe ollama[1712]: load_hparams: has_llava_proj: 0 1月 05 11:45:19 wythe ollama[1712]: load_hparams: minicpmv_version: 0 1月 05 11:45:19 wythe ollama[1712]: load_hparams: n_merge: 2 1月 05 11:45:19 wythe ollama[1712]: load_hparams: n_wa_pattern: 0 1月 05 11:45:19 wythe ollama[1712]: load_hparams: image_min_pixels: 6272 1月 05 11:45:19 wythe ollama[1712]: load_hparams: image_max_pixels: 3211264 1月 05 11:45:19 wythe ollama[1712]: load_hparams: model size: 934.64 MiB 1月 05 11:45:19 wythe ollama[1712]: load_hparams: metadata size: 0.06 MiB 1月 05 11:45:19 wythe ollama[1712]: load_tensors: loaded 182 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-454dd441c925c1a81c984204bb9d54feef0ef07b789c6fe1118099014ba2727d 1月 05 11:45:19 wythe ollama[1712]: warmup: warmup with image size = 1288 x 1288 1月 05 11:45:19 wythe ollama[1712]: alloc_compute_meta: CUDA0 compute buffer size = 515.05 MiB 1月 05 11:45:19 wythe ollama[1712]: alloc_compute_meta: CPU compute buffer size = 19.11 MiB 1月 05 11:45:19 wythe ollama[1712]: alloc_compute_meta: graph splits = 1, nodes = 632 1月 05 11:45:19 wythe ollama[1712]: warmup: flash attention is enabled 1月 05 11:45:20 wythe ollama[1712]: time=2026-01-05T11:45:20.171Z level=INFO source=server.go:1376 msg="llama runner started in 2.08 seconds" 1月 05 11:45:20 wythe ollama[1712]: time=2026-01-05T11:45:20.171Z level=INFO source=sched.go:517 msg="loaded runners" count=1 1月 05 11:45:20 wythe ollama[1712]: time=2026-01-05T11:45:20.171Z level=INFO source=server.go:1338 msg="waiting for llama runner to start responding" 1月 05 11:45:20 wythe ollama[1712]: time=2026-01-05T11:45:20.171Z level=INFO source=server.go:1376 msg="llama runner started in 2.08 seconds" 1月 05 11:45:20 wythe ollama[1712]: add_text: <|begin_of_image|> 1月 05 11:45:20 wythe ollama[1712]: image_tokens->nx = 39 1月 05 11:45:20 wythe ollama[1712]: image_tokens->ny = 86 1月 05 11:45:20 wythe ollama[1712]: batch_f32 size = 1 1月 05 11:45:20 wythe ollama[1712]: add_text: <|end_of_image|> 1月 05 11:45:49 wythe ollama[1712]: [GIN] 2026/01/05 - 11:45:49 | 200 | 31.946132842s | 192.168.244.50 | POST "/v1/chat/completions" ```
GiteaMirror commented 2026-04-29 09:04:56 -05:00
Author
Owner
Copy Link

@pinghe commented on GitHub (Jan 5, 2026):

journalctl -u ollama --since "1 hour ago"

Full log

19:10:44 arch ollama[927]: [GIN] 2026/01/05 - 19:10:44 | 200 |      44.465µs |       127.0.0.1 | HEAD     "/"
19:10:44 arch ollama[927]: [GIN] 2026/01/05 - 19:10:44 | 200 |   150.97054ms |       127.0.0.1 | POST     "/api/show"
19:10:45 arch ollama[927]: time=2026-01-05T19:10:45.073+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 36103"
19:10:45 arch ollama[927]: time=2026-01-05T19:10:45.440+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 45011"
19:10:45 arch ollama[927]: time=2026-01-05T19:10:45.690+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 38851"
19:10:45 arch ollama[927]: time=2026-01-05T19:10:45.940+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 45395"
19:10:46 arch ollama[927]: time=2026-01-05T19:10:46.190+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 45955"
19:10:46 arch ollama[927]: time=2026-01-05T19:10:46.439+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 37417"
19:10:46 arch ollama[927]: time=2026-01-05T19:10:46.690+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 39245"
19:10:46 arch ollama[927]: time=2026-01-05T19:10:46.940+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 45121"
19:10:47 arch ollama[927]: time=2026-01-05T19:10:47.190+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 37815"
19:10:47 arch ollama[927]: time=2026-01-05T19:10:47.440+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 40155"
19:10:47 arch ollama[927]: time=2026-01-05T19:10:47.690+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 43019"
19:10:47 arch ollama[927]: time=2026-01-05T19:10:47.940+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 37159"
19:10:48 arch ollama[927]: time=2026-01-05T19:10:48.190+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 36817"
19:10:48 arch ollama[927]: time=2026-01-05T19:10:48.440+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 34387"
19:10:48 arch ollama[927]: time=2026-01-05T19:10:48.690+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 45519"
19:10:48 arch ollama[927]: time=2026-01-05T19:10:48.940+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 44193"
19:10:49 arch ollama[927]: time=2026-01-05T19:10:49.190+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 42021"
19:10:49 arch ollama[927]: time=2026-01-05T19:10:49.440+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 37207"
19:10:49 arch ollama[927]: time=2026-01-05T19:10:49.690+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 46387"
19:10:49 arch ollama[927]: time=2026-01-05T19:10:49.940+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 45197"
19:10:50 arch ollama[927]: time=2026-01-05T19:10:50.190+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 44683"
19:10:50 arch ollama[927]: llama_model_loader: loaded meta data with 33 key-value pairs and 523 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20af820432e1dd55d22a2b0e728e79223e77a (version GGUF V3 (latest))
19:10:50 arch ollama[927]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
19:10:50 arch ollama[927]: llama_model_loader: - kv   0:                       general.architecture str              = glm4
19:10:50 arch ollama[927]: llama_model_loader: - kv   1:                               general.type str              = model
19:10:50 arch ollama[927]: llama_model_loader: - kv   2:                         general.size_label str              = 9.4B
19:10:50 arch ollama[927]: llama_model_loader: - kv   3:                            general.license str              = mit
19:10:50 arch ollama[927]: llama_model_loader: - kv   4:                   general.base_model.count u32              = 1
19:10:50 arch ollama[927]: llama_model_loader: - kv   5:                  general.base_model.0.name str              = GLM 4.1V 9B Base
19:10:50 arch ollama[927]: llama_model_loader: - kv   6:          general.base_model.0.organization str              = Zai Org
19:10:50 arch ollama[927]: llama_model_loader: - kv   7:              general.base_model.0.repo_url str              = https://huggingface.co/zai-org/GLM-4....
19:10:50 arch ollama[927]: llama_model_loader: - kv   8:                               general.tags arr[str,2]       = ["agent", "image-text-to-text"]
19:10:50 arch ollama[927]: llama_model_loader: - kv   9:                          general.languages arr[str,1]       = ["zh"]
19:10:50 arch ollama[927]: llama_model_loader: - kv  10:                           glm4.block_count u32              = 40
19:10:50 arch ollama[927]: llama_model_loader: - kv  11:                        glm4.context_length u32              = 65536
19:10:50 arch ollama[927]: llama_model_loader: - kv  12:                      glm4.embedding_length u32              = 4096
19:10:50 arch ollama[927]: llama_model_loader: - kv  13:                   glm4.feed_forward_length u32              = 13696
19:10:50 arch ollama[927]: llama_model_loader: - kv  14:                  glm4.attention.head_count u32              = 32
19:10:50 arch ollama[927]: llama_model_loader: - kv  15:               glm4.attention.head_count_kv u32              = 2
19:10:50 arch ollama[927]: llama_model_loader: - kv  16:               glm4.rope.dimension_sections arr[i32,4]       = [8, 12, 12, 0]
19:10:50 arch ollama[927]: llama_model_loader: - kv  17:                        glm4.rope.freq_base f32              = 10000.000000
19:10:50 arch ollama[927]: llama_model_loader: - kv  18:      glm4.attention.layer_norm_rms_epsilon f32              = 0.000010
19:10:50 arch ollama[927]: llama_model_loader: - kv  19:                  glm4.rope.dimension_count u32              = 64
19:10:50 arch ollama[927]: llama_model_loader: - kv  20:                       tokenizer.ggml.model str              = gpt2
19:10:50 arch ollama[927]: llama_model_loader: - kv  21:                         tokenizer.ggml.pre str              = glm4
19:10:50 arch ollama[927]: llama_model_loader: - kv  22:                      tokenizer.ggml.tokens arr[str,151552]  = ["!", "\"", "#", "$", "%", "&", "'", ...
19:10:50 arch ollama[927]: llama_model_loader: - kv  23:                  tokenizer.ggml.token_type arr[i32,151552]  = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
19:10:50 arch ollama[927]: llama_model_loader: - kv  24:                      tokenizer.ggml.merges arr[str,318088]  = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "...
19:10:50 arch ollama[927]: llama_model_loader: - kv  25:                tokenizer.ggml.eos_token_id u32              = 151329
19:10:50 arch ollama[927]: llama_model_loader: - kv  26:            tokenizer.ggml.padding_token_id u32              = 151329
19:10:50 arch ollama[927]: llama_model_loader: - kv  27:                tokenizer.ggml.eot_token_id u32              = 151336
19:10:50 arch ollama[927]: llama_model_loader: - kv  28:            tokenizer.ggml.unknown_token_id u32              = 151329
19:10:50 arch ollama[927]: llama_model_loader: - kv  29:                tokenizer.ggml.bos_token_id u32              = 151329
19:10:50 arch ollama[927]: llama_model_loader: - kv  30:                    tokenizer.chat_template str              = [gMASK]<sop>\n{%- for msg in messages ...
19:10:50 arch ollama[927]: llama_model_loader: - kv  31:               general.quantization_version u32              = 2
19:10:50 arch ollama[927]: llama_model_loader: - kv  32:                          general.file_type u32              = 15
19:10:50 arch ollama[927]: llama_model_loader: - type  f32:  281 tensors
19:10:50 arch ollama[927]: llama_model_loader: - type q5_0:   20 tensors
19:10:50 arch ollama[927]: llama_model_loader: - type q8_0:   20 tensors
19:10:50 arch ollama[927]: llama_model_loader: - type q4_K:  181 tensors
19:10:50 arch ollama[927]: llama_model_loader: - type q6_K:   21 tensors
19:10:50 arch ollama[927]: print_info: file format = GGUF V3 (latest)
19:10:50 arch ollama[927]: print_info: file type   = Q4_K - Medium
19:10:50 arch ollama[927]: print_info: file size   = 5.73 GiB (5.24 BPW)
19:10:51 arch ollama[927]: load: special_eot_id is not in special_eog_ids - the tokenizer config may be incorrect
19:10:51 arch ollama[927]: load: printing all EOG tokens:
19:10:51 arch ollama[927]: load:   - 151329 ('<|endoftext|>')
19:10:51 arch ollama[927]: load:   - 151336 ('<|user|>')
19:10:51 arch ollama[927]: load: special tokens cache size = 23
19:10:51 arch ollama[927]: load: token to piece cache size = 0.9711 MB
19:10:51 arch ollama[927]: print_info: arch             = glm4
19:10:51 arch ollama[927]: print_info: vocab_only       = 1
19:10:51 arch ollama[927]: print_info: no_alloc         = 0
19:10:51 arch ollama[927]: print_info: model type       = ?B
19:10:51 arch ollama[927]: print_info: model params     = 9.40 B
19:10:51 arch ollama[927]: print_info: general.name     = n/a
19:10:51 arch ollama[927]: print_info: vocab type       = BPE
19:10:51 arch ollama[927]: print_info: n_vocab          = 151552
19:10:51 arch ollama[927]: print_info: n_merges         = 318088
19:10:51 arch ollama[927]: print_info: BOS token        = 151329 '<|endoftext|>'
19:10:51 arch ollama[927]: print_info: EOS token        = 151329 '<|endoftext|>'
19:10:51 arch ollama[927]: print_info: EOT token        = 151336 '<|user|>'
19:10:51 arch ollama[927]: print_info: UNK token        = 151329 '<|endoftext|>'
19:10:51 arch ollama[927]: print_info: PAD token        = 151329 '<|endoftext|>'
19:10:51 arch ollama[927]: print_info: LF token         = 198 'Ċ'
19:10:51 arch ollama[927]: print_info: EOG token        = 151329 '<|endoftext|>'
19:10:51 arch ollama[927]: print_info: EOG token        = 151336 '<|user|>'
19:10:51 arch ollama[927]: print_info: max token length = 1024
19:10:51 arch ollama[927]: llama_model_load: vocab only - skipping tensors
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.249+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --model /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20af820432e1dd55d22a2b0e728e79223e77a --port 41297"
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=sched.go:443 msg="system memory" total="62.6 GiB" free="43.2 GiB" free_swap="71.6 MiB"
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=sched.go:450 msg="gpu memory" id=GPU-8215c551-6dde-569b-490d-884f3ab7a437 library=CUDA available="6.6 GiB" free="7.1 GiB" minimum="457.0 MiB" overhead="0 B"
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=server.go:496 msg="loading model" "model layers"=41 requested=-1
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=device.go:240 msg="model weights" device=CUDA0 size="4.2 GiB"
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=device.go:245 msg="model weights" device=CPU size="1.3 GiB"
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=device.go:251 msg="kv cache" device=CUDA0 size="544.0 MiB"
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=device.go:256 msg="kv cache" device=CPU size="96.0 MiB"
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=device.go:262 msg="compute graph" device=CUDA0 size="1.7 GiB"
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=device.go:272 msg="total memory" size="7.7 GiB"
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.266+08:00 level=INFO source=runner.go:965 msg="starting go runner"
19:10:51 arch ollama[927]: load_backend: loaded CPU backend from /usr/local/lib/ollama/libggml-cpu-haswell.so
19:10:51 arch ollama[927]: ggml_cuda_init: GGML_CUDA_FORCE_MMQ:    no
19:10:51 arch ollama[927]: ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
19:10:51 arch ollama[927]: ggml_cuda_init: found 1 CUDA devices:
19:10:51 arch ollama[927]:   Device 0: NVIDIA GeForce RTX 2060 SUPER, compute capability 7.5, VMM: yes, ID: GPU-8215c551-6dde-569b-490d-884f3ab7a437
19:10:51 arch ollama[927]: load_backend: loaded CUDA backend from /usr/local/lib/ollama/cuda_v13/libggml-cuda.so
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.341+08:00 level=INFO source=ggml.go:104 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.BMI2=1 CPU.0.LLAMAFILE=1 CPU.1.LLAMAFILE=1 CUDA.0.ARCHS=750,800,860,870,890,900,1000,1030,1100,1200,1210 CUDA.0.USE_GRAPHS=1 CUDA.0.PEER_MAX_BATCH_SIZE=128 compiler=cgo(gcc)
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.342+08:00 level=INFO source=runner.go:1001 msg="Server listening on 127.0.0.1:41297"
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.348+08:00 level=INFO source=runner.go:895 msg=load request="{Operation:commit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:Auto KvSize:16384 KvCacheType: NumThreads:12 GPULayers:34[ID:GPU-8215c551-6dde-569b-490d-884f3ab7a437 Layers:34(6..39)] MultiUserCache:false ProjectorPath:/usr/share/ollama/.ollama/models/blobs/sha256-454dd441c925c1a81c984204bb9d54feef0ef07b789c6fe1118099014ba2727d MainGPU:0 UseMmap:true}"
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.348+08:00 level=INFO source=server.go:1338 msg="waiting for llama runner to start responding"
19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.349+08:00 level=INFO source=server.go:1372 msg="waiting for server to become available" status="llm server loading model"
19:10:51 arch ollama[927]: ggml_backend_cuda_device_get_memory device GPU-8215c551-6dde-569b-490d-884f3ab7a437 utilizing NVML memory reporting free: 7585398784 total: 8589934592
19:10:51 arch ollama[927]: llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce RTX 2060 SUPER) (0000:03:00.0) - 7234 MiB free
19:10:51 arch ollama[927]: llama_model_loader: loaded meta data with 33 key-value pairs and 523 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20af820432e1dd55d22a2b0e728e79223e77a (version GGUF V3 (latest))
19:10:51 arch ollama[927]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
19:10:51 arch ollama[927]: llama_model_loader: - kv   0:                       general.architecture str              = glm4
19:10:51 arch ollama[927]: llama_model_loader: - kv   1:                               general.type str              = model
19:10:51 arch ollama[927]: llama_model_loader: - kv   2:                         general.size_label str              = 9.4B
19:10:51 arch ollama[927]: llama_model_loader: - kv   3:                            general.license str              = mit
19:10:51 arch ollama[927]: llama_model_loader: - kv   4:                   general.base_model.count u32              = 1
19:10:51 arch ollama[927]: llama_model_loader: - kv   5:                  general.base_model.0.name str              = GLM 4.1V 9B Base
19:10:51 arch ollama[927]: llama_model_loader: - kv   6:          general.base_model.0.organization str              = Zai Org
19:10:51 arch ollama[927]: llama_model_loader: - kv   7:              general.base_model.0.repo_url str              = https://huggingface.co/zai-org/GLM-4....
19:10:51 arch ollama[927]: llama_model_loader: - kv   8:                               general.tags arr[str,2]       = ["agent", "image-text-to-text"]
19:10:51 arch ollama[927]: llama_model_loader: - kv   9:                          general.languages arr[str,1]       = ["zh"]
19:10:51 arch ollama[927]: llama_model_loader: - kv  10:                           glm4.block_count u32              = 40
19:10:51 arch ollama[927]: llama_model_loader: - kv  11:                        glm4.context_length u32              = 65536
19:10:51 arch ollama[927]: llama_model_loader: - kv  12:                      glm4.embedding_length u32              = 4096
19:10:51 arch ollama[927]: llama_model_loader: - kv  13:                   glm4.feed_forward_length u32              = 13696
19:10:51 arch ollama[927]: llama_model_loader: - kv  14:                  glm4.attention.head_count u32              = 32
19:10:51 arch ollama[927]: llama_model_loader: - kv  15:               glm4.attention.head_count_kv u32              = 2
19:10:51 arch ollama[927]: llama_model_loader: - kv  16:               glm4.rope.dimension_sections arr[i32,4]       = [8, 12, 12, 0]
19:10:51 arch ollama[927]: llama_model_loader: - kv  17:                        glm4.rope.freq_base f32              = 10000.000000
19:10:51 arch ollama[927]: llama_model_loader: - kv  18:      glm4.attention.layer_norm_rms_epsilon f32              = 0.000010
19:10:51 arch ollama[927]: llama_model_loader: - kv  19:                  glm4.rope.dimension_count u32              = 64
19:10:51 arch ollama[927]: llama_model_loader: - kv  20:                       tokenizer.ggml.model str              = gpt2
19:10:51 arch ollama[927]: llama_model_loader: - kv  21:                         tokenizer.ggml.pre str              = glm4
19:10:51 arch ollama[927]: llama_model_loader: - kv  22:                      tokenizer.ggml.tokens arr[str,151552]  = ["!", "\"", "#", "$", "%", "&", "'", ...
19:10:51 arch ollama[927]: llama_model_loader: - kv  23:                  tokenizer.ggml.token_type arr[i32,151552]  = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
19:10:51 arch ollama[927]: llama_model_loader: - kv  24:                      tokenizer.ggml.merges arr[str,318088]  = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "...
19:10:51 arch ollama[927]: llama_model_loader: - kv  25:                tokenizer.ggml.eos_token_id u32              = 151329
19:10:51 arch ollama[927]: llama_model_loader: - kv  26:            tokenizer.ggml.padding_token_id u32              = 151329
19:10:51 arch ollama[927]: llama_model_loader: - kv  27:                tokenizer.ggml.eot_token_id u32              = 151336
19:10:51 arch ollama[927]: llama_model_loader: - kv  28:            tokenizer.ggml.unknown_token_id u32              = 151329
19:10:51 arch ollama[927]: llama_model_loader: - kv  29:                tokenizer.ggml.bos_token_id u32              = 151329
19:10:51 arch ollama[927]: llama_model_loader: - kv  30:                    tokenizer.chat_template str              = [gMASK]<sop>\n{%- for msg in messages ...
19:10:51 arch ollama[927]: llama_model_loader: - kv  31:               general.quantization_version u32              = 2
19:10:51 arch ollama[927]: llama_model_loader: - kv  32:                          general.file_type u32              = 15
19:10:51 arch ollama[927]: llama_model_loader: - type  f32:  281 tensors
19:10:51 arch ollama[927]: llama_model_loader: - type q5_0:   20 tensors
19:10:51 arch ollama[927]: llama_model_loader: - type q8_0:   20 tensors
19:10:51 arch ollama[927]: llama_model_loader: - type q4_K:  181 tensors
19:10:51 arch ollama[927]: llama_model_loader: - type q6_K:   21 tensors
19:10:51 arch ollama[927]: print_info: file format = GGUF V3 (latest)
19:10:51 arch ollama[927]: print_info: file type   = Q4_K - Medium
19:10:51 arch ollama[927]: print_info: file size   = 5.73 GiB (5.24 BPW)
19:10:51 arch ollama[927]: load: special_eot_id is not in special_eog_ids - the tokenizer config may be incorrect
19:10:51 arch ollama[927]: load: printing all EOG tokens:
19:10:51 arch ollama[927]: load:   - 151329 ('<|endoftext|>')
19:10:51 arch ollama[927]: load:   - 151336 ('<|user|>')
19:10:51 arch ollama[927]: load: special tokens cache size = 23
19:10:51 arch ollama[927]: load: token to piece cache size = 0.9711 MB
19:10:51 arch ollama[927]: print_info: arch             = glm4
19:10:51 arch ollama[927]: print_info: vocab_only       = 0
19:10:51 arch ollama[927]: print_info: no_alloc         = 0
19:10:51 arch ollama[927]: print_info: n_ctx_train      = 65536
19:10:51 arch ollama[927]: print_info: n_embd           = 4096
19:10:51 arch ollama[927]: print_info: n_embd_inp       = 4096
19:10:51 arch ollama[927]: print_info: n_layer          = 40
19:10:51 arch ollama[927]: print_info: n_head           = 32
19:10:51 arch ollama[927]: print_info: n_head_kv        = 2
19:10:51 arch ollama[927]: print_info: n_rot            = 64
19:10:51 arch ollama[927]: print_info: n_swa            = 0
19:10:51 arch ollama[927]: print_info: is_swa_any       = 0
19:10:51 arch ollama[927]: print_info: n_embd_head_k    = 128
19:10:51 arch ollama[927]: print_info: n_embd_head_v    = 128
19:10:51 arch ollama[927]: print_info: n_gqa            = 16
19:10:51 arch ollama[927]: print_info: n_embd_k_gqa     = 256
19:10:51 arch ollama[927]: print_info: n_embd_v_gqa     = 256
19:10:51 arch ollama[927]: print_info: f_norm_eps       = 0.0e+00
19:10:51 arch ollama[927]: print_info: f_norm_rms_eps   = 1.0e-05
19:10:51 arch ollama[927]: print_info: f_clamp_kqv      = 0.0e+00
19:10:51 arch ollama[927]: print_info: f_max_alibi_bias = 0.0e+00
19:10:51 arch ollama[927]: print_info: f_logit_scale    = 0.0e+00
19:10:51 arch ollama[927]: print_info: f_attn_scale     = 0.0e+00
19:10:51 arch ollama[927]: print_info: n_ff             = 13696
19:10:51 arch ollama[927]: print_info: n_expert         = 0
19:10:51 arch ollama[927]: print_info: n_expert_used    = 0
19:10:51 arch ollama[927]: print_info: n_expert_groups  = 0
19:10:51 arch ollama[927]: print_info: n_group_used     = 0
19:10:51 arch ollama[927]: print_info: causal attn      = 1
19:10:51 arch ollama[927]: print_info: pooling type     = 0
19:10:51 arch ollama[927]: print_info: rope type        = 8
19:10:51 arch ollama[927]: print_info: rope scaling     = linear
19:10:51 arch ollama[927]: print_info: freq_base_train  = 10000.0
19:10:51 arch ollama[927]: print_info: freq_scale_train = 1
19:10:51 arch ollama[927]: print_info: n_ctx_orig_yarn  = 65536
19:10:51 arch ollama[927]: print_info: rope_yarn_log_mul= 0.0000
19:10:51 arch ollama[927]: print_info: rope_finetuned   = unknown
19:10:51 arch ollama[927]: print_info: mrope sections   = [8, 12, 12, 0]
19:10:51 arch ollama[927]: print_info: model type       = 9B
19:10:51 arch ollama[927]: print_info: model params     = 9.40 B
19:10:51 arch ollama[927]: print_info: general.name     = n/a
19:10:51 arch ollama[927]: print_info: vocab type       = BPE
19:10:51 arch ollama[927]: print_info: n_vocab          = 151552
19:10:51 arch ollama[927]: print_info: n_merges         = 318088
19:10:51 arch ollama[927]: print_info: BOS token        = 151329 '<|endoftext|>'
19:10:51 arch ollama[927]: print_info: EOS token        = 151329 '<|endoftext|>'
19:10:51 arch ollama[927]: print_info: EOT token        = 151336 '<|user|>'
19:10:51 arch ollama[927]: print_info: UNK token        = 151329 '<|endoftext|>'
19:10:51 arch ollama[927]: print_info: PAD token        = 151329 '<|endoftext|>'
19:10:51 arch ollama[927]: print_info: LF token         = 198 'Ċ'
19:10:51 arch ollama[927]: print_info: EOG token        = 151329 '<|endoftext|>'
19:10:51 arch ollama[927]: print_info: EOG token        = 151336 '<|user|>'
19:10:51 arch ollama[927]: print_info: max token length = 1024
19:10:51 arch ollama[927]: load_tensors: loading model tensors, this can take a while... (mmap = true)
19:10:52 arch ollama[927]: load_tensors: offloading 34 repeating layers to GPU
19:10:52 arch ollama[927]: load_tensors: offloaded 34/41 layers to GPU
19:10:52 arch ollama[927]: load_tensors:   CPU_Mapped model buffer size =  1617.29 MiB
19:10:52 arch ollama[927]: load_tensors:        CUDA0 model buffer size =  4254.72 MiB
19:10:53 arch ollama[927]: llama_context: constructing llama_context
19:10:53 arch ollama[927]: llama_context: n_seq_max     = 1
19:10:53 arch ollama[927]: llama_context: n_ctx         = 16384
19:10:53 arch ollama[927]: llama_context: n_ctx_seq     = 16384
19:10:53 arch ollama[927]: llama_context: n_batch       = 512
19:10:53 arch ollama[927]: llama_context: n_ubatch      = 512
19:10:53 arch ollama[927]: llama_context: causal_attn   = 1
19:10:53 arch ollama[927]: llama_context: flash_attn    = auto
19:10:53 arch ollama[927]: llama_context: kv_unified    = false
19:10:53 arch ollama[927]: llama_context: freq_base     = 10000.0
19:10:53 arch ollama[927]: llama_context: freq_scale    = 1
19:10:53 arch ollama[927]: llama_context: n_ctx_seq (16384) < n_ctx_train (65536) -- the full capacity of the model will not be utilized
19:10:53 arch ollama[927]: llama_context:        CPU  output buffer size =     0.59 MiB
19:10:53 arch ollama[927]: llama_kv_cache:        CPU KV buffer size =    96.00 MiB
19:10:53 arch ollama[927]: llama_kv_cache:      CUDA0 KV buffer size =   544.00 MiB
19:10:53 arch ollama[927]: llama_kv_cache: size =  640.00 MiB ( 16384 cells,  40 layers,  1/1 seqs), K (f16):  320.00 MiB, V (f16):  320.00 MiB
19:10:53 arch ollama[927]: llama_context: Flash Attention was auto, set to enabled
19:10:53 arch ollama[927]: llama_context:      CUDA0 compute buffer size =   789.62 MiB
19:10:53 arch ollama[927]: llama_context:  CUDA_Host compute buffer size =    40.02 MiB
19:10:53 arch ollama[927]: llama_context: graph nodes  = 1487
19:10:53 arch ollama[927]: llama_context: graph splits = 94 (with bs=512), 3 (with bs=1)
19:10:53 arch ollama[927]: clip_model_loader: model name:
19:10:53 arch ollama[927]: clip_model_loader: description:
19:10:53 arch ollama[927]: clip_model_loader: GGUF version: 3
19:10:53 arch ollama[927]: clip_model_loader: alignment:    32
19:10:53 arch ollama[927]: clip_model_loader: n_tensors:    182
19:10:53 arch ollama[927]: clip_model_loader: n_kv:         25
19:10:53 arch ollama[927]: clip_model_loader: has vision encoder
19:10:53 arch ollama[927]: clip_model_loader: tensor[0]: n_dims = 2, name = v.blk.0.attn_out.weight, tensor_size=2506752, offset=0, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[1]: n_dims = 2, name = v.blk.0.attn_qkv.weight, tensor_size=7520256, offset=2506752, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[2]: n_dims = 2, name = v.blk.0.ffn_down.weight, tensor_size=6684672, offset=10027008, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[3]: n_dims = 2, name = v.blk.0.ffn_gate.weight, tensor_size=6684672, offset=16711680, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[4]: n_dims = 2, name = v.blk.0.ffn_up.weight, tensor_size=6684672, offset=23396352, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[5]: n_dims = 1, name = v.blk.0.ln1.weight, tensor_size=6144, offset=30081024, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[6]: n_dims = 1, name = v.blk.0.ln2.weight, tensor_size=6144, offset=30087168, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[7]: n_dims = 2, name = v.blk.1.attn_out.weight, tensor_size=2506752, offset=30093312, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[8]: n_dims = 2, name = v.blk.1.attn_qkv.weight, tensor_size=7520256, offset=32600064, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[9]: n_dims = 2, name = v.blk.1.ffn_down.weight, tensor_size=6684672, offset=40120320, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[10]: n_dims = 2, name = v.blk.1.ffn_gate.weight, tensor_size=6684672, offset=46804992, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[11]: n_dims = 2, name = v.blk.1.ffn_up.weight, tensor_size=6684672, offset=53489664, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[12]: n_dims = 1, name = v.blk.1.ln1.weight, tensor_size=6144, offset=60174336, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[13]: n_dims = 1, name = v.blk.1.ln2.weight, tensor_size=6144, offset=60180480, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[14]: n_dims = 2, name = v.blk.10.attn_out.weight, tensor_size=2506752, offset=60186624, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[15]: n_dims = 2, name = v.blk.10.attn_qkv.weight, tensor_size=7520256, offset=62693376, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[16]: n_dims = 2, name = v.blk.10.ffn_down.weight, tensor_size=6684672, offset=70213632, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[17]: n_dims = 2, name = v.blk.10.ffn_gate.weight, tensor_size=6684672, offset=76898304, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[18]: n_dims = 2, name = v.blk.10.ffn_up.weight, tensor_size=6684672, offset=83582976, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[19]: n_dims = 1, name = v.blk.10.ln1.weight, tensor_size=6144, offset=90267648, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[20]: n_dims = 1, name = v.blk.10.ln2.weight, tensor_size=6144, offset=90273792, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[21]: n_dims = 2, name = v.blk.11.attn_out.weight, tensor_size=2506752, offset=90279936, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[22]: n_dims = 2, name = v.blk.11.attn_qkv.weight, tensor_size=7520256, offset=92786688, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[23]: n_dims = 2, name = v.blk.11.ffn_down.weight, tensor_size=6684672, offset=100306944, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[24]: n_dims = 2, name = v.blk.11.ffn_gate.weight, tensor_size=6684672, offset=106991616, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[25]: n_dims = 2, name = v.blk.11.ffn_up.weight, tensor_size=6684672, offset=113676288, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[26]: n_dims = 1, name = v.blk.11.ln1.weight, tensor_size=6144, offset=120360960, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[27]: n_dims = 1, name = v.blk.11.ln2.weight, tensor_size=6144, offset=120367104, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[28]: n_dims = 2, name = v.blk.12.attn_out.weight, tensor_size=2506752, offset=120373248, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[29]: n_dims = 2, name = v.blk.12.attn_qkv.weight, tensor_size=7520256, offset=122880000, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[30]: n_dims = 2, name = v.blk.12.ffn_down.weight, tensor_size=6684672, offset=130400256, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[31]: n_dims = 2, name = v.blk.12.ffn_gate.weight, tensor_size=6684672, offset=137084928, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[32]: n_dims = 2, name = v.blk.12.ffn_up.weight, tensor_size=6684672, offset=143769600, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[33]: n_dims = 1, name = v.blk.12.ln1.weight, tensor_size=6144, offset=150454272, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[34]: n_dims = 1, name = v.blk.12.ln2.weight, tensor_size=6144, offset=150460416, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[35]: n_dims = 2, name = v.blk.13.attn_out.weight, tensor_size=2506752, offset=150466560, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[36]: n_dims = 2, name = v.blk.13.attn_qkv.weight, tensor_size=7520256, offset=152973312, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[37]: n_dims = 2, name = v.blk.13.ffn_down.weight, tensor_size=6684672, offset=160493568, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[38]: n_dims = 2, name = v.blk.13.ffn_gate.weight, tensor_size=6684672, offset=167178240, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[39]: n_dims = 2, name = v.blk.13.ffn_up.weight, tensor_size=6684672, offset=173862912, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[40]: n_dims = 1, name = v.blk.13.ln1.weight, tensor_size=6144, offset=180547584, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[41]: n_dims = 1, name = v.blk.13.ln2.weight, tensor_size=6144, offset=180553728, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[42]: n_dims = 2, name = v.blk.14.attn_out.weight, tensor_size=2506752, offset=180559872, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[43]: n_dims = 2, name = v.blk.14.attn_qkv.weight, tensor_size=7520256, offset=183066624, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[44]: n_dims = 2, name = v.blk.14.ffn_down.weight, tensor_size=6684672, offset=190586880, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[45]: n_dims = 2, name = v.blk.14.ffn_gate.weight, tensor_size=6684672, offset=197271552, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[46]: n_dims = 2, name = v.blk.14.ffn_up.weight, tensor_size=6684672, offset=203956224, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[47]: n_dims = 1, name = v.blk.14.ln1.weight, tensor_size=6144, offset=210640896, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[48]: n_dims = 1, name = v.blk.14.ln2.weight, tensor_size=6144, offset=210647040, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[49]: n_dims = 2, name = v.blk.15.attn_out.weight, tensor_size=2506752, offset=210653184, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[50]: n_dims = 2, name = v.blk.15.attn_qkv.weight, tensor_size=7520256, offset=213159936, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[51]: n_dims = 2, name = v.blk.15.ffn_down.weight, tensor_size=6684672, offset=220680192, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[52]: n_dims = 2, name = v.blk.15.ffn_gate.weight, tensor_size=6684672, offset=227364864, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[53]: n_dims = 2, name = v.blk.15.ffn_up.weight, tensor_size=6684672, offset=234049536, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[54]: n_dims = 1, name = v.blk.15.ln1.weight, tensor_size=6144, offset=240734208, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[55]: n_dims = 1, name = v.blk.15.ln2.weight, tensor_size=6144, offset=240740352, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[56]: n_dims = 2, name = v.blk.16.attn_out.weight, tensor_size=2506752, offset=240746496, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[57]: n_dims = 2, name = v.blk.16.attn_qkv.weight, tensor_size=7520256, offset=243253248, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[58]: n_dims = 2, name = v.blk.16.ffn_down.weight, tensor_size=6684672, offset=250773504, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[59]: n_dims = 2, name = v.blk.16.ffn_gate.weight, tensor_size=6684672, offset=257458176, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[60]: n_dims = 2, name = v.blk.16.ffn_up.weight, tensor_size=6684672, offset=264142848, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[61]: n_dims = 1, name = v.blk.16.ln1.weight, tensor_size=6144, offset=270827520, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[62]: n_dims = 1, name = v.blk.16.ln2.weight, tensor_size=6144, offset=270833664, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[63]: n_dims = 2, name = v.blk.17.attn_out.weight, tensor_size=2506752, offset=270839808, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[64]: n_dims = 2, name = v.blk.17.attn_qkv.weight, tensor_size=7520256, offset=273346560, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[65]: n_dims = 2, name = v.blk.17.ffn_down.weight, tensor_size=6684672, offset=280866816, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[66]: n_dims = 2, name = v.blk.17.ffn_gate.weight, tensor_size=6684672, offset=287551488, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[67]: n_dims = 2, name = v.blk.17.ffn_up.weight, tensor_size=6684672, offset=294236160, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[68]: n_dims = 1, name = v.blk.17.ln1.weight, tensor_size=6144, offset=300920832, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[69]: n_dims = 1, name = v.blk.17.ln2.weight, tensor_size=6144, offset=300926976, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[70]: n_dims = 2, name = v.blk.18.attn_out.weight, tensor_size=2506752, offset=300933120, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[71]: n_dims = 2, name = v.blk.18.attn_qkv.weight, tensor_size=7520256, offset=303439872, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[72]: n_dims = 2, name = v.blk.18.ffn_down.weight, tensor_size=6684672, offset=310960128, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[73]: n_dims = 2, name = v.blk.18.ffn_gate.weight, tensor_size=6684672, offset=317644800, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[74]: n_dims = 2, name = v.blk.18.ffn_up.weight, tensor_size=6684672, offset=324329472, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[75]: n_dims = 1, name = v.blk.18.ln1.weight, tensor_size=6144, offset=331014144, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[76]: n_dims = 1, name = v.blk.18.ln2.weight, tensor_size=6144, offset=331020288, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[77]: n_dims = 2, name = v.blk.19.attn_out.weight, tensor_size=2506752, offset=331026432, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[78]: n_dims = 2, name = v.blk.19.attn_qkv.weight, tensor_size=7520256, offset=333533184, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[79]: n_dims = 2, name = v.blk.19.ffn_down.weight, tensor_size=6684672, offset=341053440, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[80]: n_dims = 2, name = v.blk.19.ffn_gate.weight, tensor_size=6684672, offset=347738112, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[81]: n_dims = 2, name = v.blk.19.ffn_up.weight, tensor_size=6684672, offset=354422784, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[82]: n_dims = 1, name = v.blk.19.ln1.weight, tensor_size=6144, offset=361107456, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[83]: n_dims = 1, name = v.blk.19.ln2.weight, tensor_size=6144, offset=361113600, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[84]: n_dims = 2, name = v.blk.2.attn_out.weight, tensor_size=2506752, offset=361119744, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[85]: n_dims = 2, name = v.blk.2.attn_qkv.weight, tensor_size=7520256, offset=363626496, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[86]: n_dims = 2, name = v.blk.2.ffn_down.weight, tensor_size=6684672, offset=371146752, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[87]: n_dims = 2, name = v.blk.2.ffn_gate.weight, tensor_size=6684672, offset=377831424, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[88]: n_dims = 2, name = v.blk.2.ffn_up.weight, tensor_size=6684672, offset=384516096, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[89]: n_dims = 1, name = v.blk.2.ln1.weight, tensor_size=6144, offset=391200768, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[90]: n_dims = 1, name = v.blk.2.ln2.weight, tensor_size=6144, offset=391206912, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[91]: n_dims = 2, name = v.blk.20.attn_out.weight, tensor_size=2506752, offset=391213056, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[92]: n_dims = 2, name = v.blk.20.attn_qkv.weight, tensor_size=7520256, offset=393719808, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[93]: n_dims = 2, name = v.blk.20.ffn_down.weight, tensor_size=6684672, offset=401240064, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[94]: n_dims = 2, name = v.blk.20.ffn_gate.weight, tensor_size=6684672, offset=407924736, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[95]: n_dims = 2, name = v.blk.20.ffn_up.weight, tensor_size=6684672, offset=414609408, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[96]: n_dims = 1, name = v.blk.20.ln1.weight, tensor_size=6144, offset=421294080, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[97]: n_dims = 1, name = v.blk.20.ln2.weight, tensor_size=6144, offset=421300224, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[98]: n_dims = 2, name = v.blk.21.attn_out.weight, tensor_size=2506752, offset=421306368, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[99]: n_dims = 2, name = v.blk.21.attn_qkv.weight, tensor_size=7520256, offset=423813120, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[100]: n_dims = 2, name = v.blk.21.ffn_down.weight, tensor_size=6684672, offset=431333376, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[101]: n_dims = 2, name = v.blk.21.ffn_gate.weight, tensor_size=6684672, offset=438018048, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[102]: n_dims = 2, name = v.blk.21.ffn_up.weight, tensor_size=6684672, offset=444702720, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[103]: n_dims = 1, name = v.blk.21.ln1.weight, tensor_size=6144, offset=451387392, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[104]: n_dims = 1, name = v.blk.21.ln2.weight, tensor_size=6144, offset=451393536, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[105]: n_dims = 2, name = v.blk.22.attn_out.weight, tensor_size=2506752, offset=451399680, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[106]: n_dims = 2, name = v.blk.22.attn_qkv.weight, tensor_size=7520256, offset=453906432, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[107]: n_dims = 2, name = v.blk.22.ffn_down.weight, tensor_size=6684672, offset=461426688, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[108]: n_dims = 2, name = v.blk.22.ffn_gate.weight, tensor_size=6684672, offset=468111360, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[109]: n_dims = 2, name = v.blk.22.ffn_up.weight, tensor_size=6684672, offset=474796032, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[110]: n_dims = 1, name = v.blk.22.ln1.weight, tensor_size=6144, offset=481480704, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[111]: n_dims = 1, name = v.blk.22.ln2.weight, tensor_size=6144, offset=481486848, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[112]: n_dims = 2, name = v.blk.23.attn_out.weight, tensor_size=2506752, offset=481492992, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[113]: n_dims = 2, name = v.blk.23.attn_qkv.weight, tensor_size=7520256, offset=483999744, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[114]: n_dims = 2, name = v.blk.23.ffn_down.weight, tensor_size=6684672, offset=491520000, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[115]: n_dims = 2, name = v.blk.23.ffn_gate.weight, tensor_size=6684672, offset=498204672, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[116]: n_dims = 2, name = v.blk.23.ffn_up.weight, tensor_size=6684672, offset=504889344, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[117]: n_dims = 1, name = v.blk.23.ln1.weight, tensor_size=6144, offset=511574016, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[118]: n_dims = 1, name = v.blk.23.ln2.weight, tensor_size=6144, offset=511580160, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[119]: n_dims = 2, name = v.blk.3.attn_out.weight, tensor_size=2506752, offset=511586304, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[120]: n_dims = 2, name = v.blk.3.attn_qkv.weight, tensor_size=7520256, offset=514093056, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[121]: n_dims = 2, name = v.blk.3.ffn_down.weight, tensor_size=6684672, offset=521613312, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[122]: n_dims = 2, name = v.blk.3.ffn_gate.weight, tensor_size=6684672, offset=528297984, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[123]: n_dims = 2, name = v.blk.3.ffn_up.weight, tensor_size=6684672, offset=534982656, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[124]: n_dims = 1, name = v.blk.3.ln1.weight, tensor_size=6144, offset=541667328, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[125]: n_dims = 1, name = v.blk.3.ln2.weight, tensor_size=6144, offset=541673472, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[126]: n_dims = 2, name = v.blk.4.attn_out.weight, tensor_size=2506752, offset=541679616, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[127]: n_dims = 2, name = v.blk.4.attn_qkv.weight, tensor_size=7520256, offset=544186368, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[128]: n_dims = 2, name = v.blk.4.ffn_down.weight, tensor_size=6684672, offset=551706624, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[129]: n_dims = 2, name = v.blk.4.ffn_gate.weight, tensor_size=6684672, offset=558391296, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[130]: n_dims = 2, name = v.blk.4.ffn_up.weight, tensor_size=6684672, offset=565075968, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[131]: n_dims = 1, name = v.blk.4.ln1.weight, tensor_size=6144, offset=571760640, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[132]: n_dims = 1, name = v.blk.4.ln2.weight, tensor_size=6144, offset=571766784, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[133]: n_dims = 2, name = v.blk.5.attn_out.weight, tensor_size=2506752, offset=571772928, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[134]: n_dims = 2, name = v.blk.5.attn_qkv.weight, tensor_size=7520256, offset=574279680, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[135]: n_dims = 2, name = v.blk.5.ffn_down.weight, tensor_size=6684672, offset=581799936, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[136]: n_dims = 2, name = v.blk.5.ffn_gate.weight, tensor_size=6684672, offset=588484608, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[137]: n_dims = 2, name = v.blk.5.ffn_up.weight, tensor_size=6684672, offset=595169280, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[138]: n_dims = 1, name = v.blk.5.ln1.weight, tensor_size=6144, offset=601853952, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[139]: n_dims = 1, name = v.blk.5.ln2.weight, tensor_size=6144, offset=601860096, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[140]: n_dims = 2, name = v.blk.6.attn_out.weight, tensor_size=2506752, offset=601866240, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[141]: n_dims = 2, name = v.blk.6.attn_qkv.weight, tensor_size=7520256, offset=604372992, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[142]: n_dims = 2, name = v.blk.6.ffn_down.weight, tensor_size=6684672, offset=611893248, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[143]: n_dims = 2, name = v.blk.6.ffn_gate.weight, tensor_size=6684672, offset=618577920, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[144]: n_dims = 2, name = v.blk.6.ffn_up.weight, tensor_size=6684672, offset=625262592, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[145]: n_dims = 1, name = v.blk.6.ln1.weight, tensor_size=6144, offset=631947264, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[146]: n_dims = 1, name = v.blk.6.ln2.weight, tensor_size=6144, offset=631953408, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[147]: n_dims = 2, name = v.blk.7.attn_out.weight, tensor_size=2506752, offset=631959552, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[148]: n_dims = 2, name = v.blk.7.attn_qkv.weight, tensor_size=7520256, offset=634466304, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[149]: n_dims = 2, name = v.blk.7.ffn_down.weight, tensor_size=6684672, offset=641986560, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[150]: n_dims = 2, name = v.blk.7.ffn_gate.weight, tensor_size=6684672, offset=648671232, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[151]: n_dims = 2, name = v.blk.7.ffn_up.weight, tensor_size=6684672, offset=655355904, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[152]: n_dims = 1, name = v.blk.7.ln1.weight, tensor_size=6144, offset=662040576, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[153]: n_dims = 1, name = v.blk.7.ln2.weight, tensor_size=6144, offset=662046720, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[154]: n_dims = 2, name = v.blk.8.attn_out.weight, tensor_size=2506752, offset=662052864, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[155]: n_dims = 2, name = v.blk.8.attn_qkv.weight, tensor_size=7520256, offset=664559616, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[156]: n_dims = 2, name = v.blk.8.ffn_down.weight, tensor_size=6684672, offset=672079872, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[157]: n_dims = 2, name = v.blk.8.ffn_gate.weight, tensor_size=6684672, offset=678764544, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[158]: n_dims = 2, name = v.blk.8.ffn_up.weight, tensor_size=6684672, offset=685449216, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[159]: n_dims = 1, name = v.blk.8.ln1.weight, tensor_size=6144, offset=692133888, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[160]: n_dims = 1, name = v.blk.8.ln2.weight, tensor_size=6144, offset=692140032, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[161]: n_dims = 2, name = v.blk.9.attn_out.weight, tensor_size=2506752, offset=692146176, shape:[1536, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[162]: n_dims = 2, name = v.blk.9.attn_qkv.weight, tensor_size=7520256, offset=694652928, shape:[1536, 4608, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[163]: n_dims = 2, name = v.blk.9.ffn_down.weight, tensor_size=6684672, offset=702173184, shape:[4096, 1536, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[164]: n_dims = 2, name = v.blk.9.ffn_gate.weight, tensor_size=6684672, offset=708857856, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[165]: n_dims = 2, name = v.blk.9.ffn_up.weight, tensor_size=6684672, offset=715542528, shape:[1536, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[166]: n_dims = 1, name = v.blk.9.ln1.weight, tensor_size=6144, offset=722227200, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[167]: n_dims = 1, name = v.blk.9.ln2.weight, tensor_size=6144, offset=722233344, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[168]: n_dims = 1, name = mm.patch_merger.bias, tensor_size=16384, offset=722239488, shape:[4096, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[169]: n_dims = 4, name = mm.patch_merger.weight, tensor_size=50331648, offset=722255872, shape:[2, 2, 1536, 4096], type = f16
19:10:53 arch ollama[927]: clip_model_loader: tensor[170]: n_dims = 2, name = v.position_embd.weight, tensor_size=3538944, offset=772587520, shape:[1536, 576, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[171]: n_dims = 2, name = mm.down.weight, tensor_size=59604992, offset=776126464, shape:[13696, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[172]: n_dims = 2, name = mm.gate.weight, tensor_size=59604992, offset=835731456, shape:[4096, 13696, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[173]: n_dims = 1, name = mm.post_norm.bias, tensor_size=16384, offset=895336448, shape:[4096, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[174]: n_dims = 1, name = mm.post_norm.weight, tensor_size=16384, offset=895352832, shape:[4096, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[175]: n_dims = 2, name = mm.model.fc.weight, tensor_size=17825792, offset=895369216, shape:[4096, 4096, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[176]: n_dims = 2, name = mm.up.weight, tensor_size=59604992, offset=913195008, shape:[4096, 13696, 1, 1], type = q8_0
19:10:53 arch ollama[927]: clip_model_loader: tensor[177]: n_dims = 1, name = v.patch_embd.bias, tensor_size=6144, offset=972800000, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[178]: n_dims = 4, name = v.patch_embd.weight, tensor_size=3612672, offset=972806144, shape:[14, 14, 3, 1536], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[179]: n_dims = 4, name = v.patch_embd.weight.1, tensor_size=3612672, offset=976418816, shape:[14, 14, 3, 1536], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[180]: n_dims = 1, name = v.norm_embd.weight, tensor_size=6144, offset=980031488, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_model_loader: tensor[181]: n_dims = 1, name = v.post_ln.weight, tensor_size=6144, offset=980037632, shape:[1536, 1, 1, 1], type = f32
19:10:53 arch ollama[927]: clip_ctx: CLIP using CUDA0 backend
19:10:53 arch ollama[927]: load_hparams: projector:          glm4v
19:10:53 arch ollama[927]: load_hparams: n_embd:             1536
19:10:53 arch ollama[927]: load_hparams: n_head:             12
19:10:53 arch ollama[927]: load_hparams: n_ff:               13696
19:10:53 arch ollama[927]: load_hparams: n_layer:            24
19:10:53 arch ollama[927]: load_hparams: ffn_op:             silu
19:10:53 arch ollama[927]: load_hparams: projection_dim:     4096
19:10:53 arch ollama[927]: --- vision hparams ---
19:10:53 arch ollama[927]: load_hparams: image_size:         336
19:10:53 arch ollama[927]: load_hparams: patch_size:         14
19:10:53 arch ollama[927]: load_hparams: has_llava_proj:     0
19:10:53 arch ollama[927]: load_hparams: minicpmv_version:   0
19:10:53 arch ollama[927]: load_hparams: n_merge:            2
19:10:53 arch ollama[927]: load_hparams: n_wa_pattern:       0
19:10:53 arch ollama[927]: load_hparams: image_min_pixels:   6272
19:10:53 arch ollama[927]: load_hparams: image_max_pixels:   3211264
19:10:53 arch ollama[927]: load_hparams: model size:         934.64 MiB
19:10:53 arch ollama[927]: load_hparams: metadata size:      0.06 MiB
19:10:53 arch ollama[927]: load_tensors: loaded 182 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-454dd441c925c1a81c984204bb9d54feef0ef07b789c6fe1118099014ba2727d
19:10:53 arch ollama[927]: warmup: warmup with image size = 1288 x 1288
19:10:53 arch ollama[927]: ggml_backend_cuda_buffer_type_alloc_buffer: allocating 515.05 MiB on device 0: cudaMalloc failed: out of memory
19:10:53 arch ollama[927]: ggml_gallocr_reserve_n_impl: failed to allocate CUDA0 buffer of size 540070912
19:10:53 arch ollama[927]: alloc_compute_meta:        CPU compute buffer size =    19.11 MiB
19:10:53 arch ollama[927]: alloc_compute_meta: graph splits = 1, nodes = 632
19:10:53 arch ollama[927]: warmup: flash attention is enabled
19:10:53 arch ollama[927]: time=2026-01-05T19:10:53.861+08:00 level=INFO source=server.go:1376 msg="llama runner started in 2.61 seconds"
19:10:53 arch ollama[927]: time=2026-01-05T19:10:53.861+08:00 level=INFO source=sched.go:517 msg="loaded runners" count=1
19:10:53 arch ollama[927]: time=2026-01-05T19:10:53.861+08:00 level=INFO source=server.go:1338 msg="waiting for llama runner to start responding"
19:10:53 arch ollama[927]: time=2026-01-05T19:10:53.861+08:00 level=INFO source=server.go:1376 msg="llama runner started in 2.61 seconds"
19:10:54 arch ollama[927]: add_text: <|begin_of_image|>
19:10:54 arch ollama[927]: image_tokens->nx = 85
19:10:54 arch ollama[927]: image_tokens->ny = 48
19:10:54 arch ollama[927]: batch_f32 size = 1
19:10:54 arch ollama[927]: add_text: <|end_of_image|>
19:10:54 arch ollama[927]: ggml_backend_cuda_buffer_type_alloc_buffer: allocating 993.11 MiB on device 0: cudaMalloc failed: out of memory
19:10:54 arch ollama[927]: ggml_gallocr_reserve_n_impl: failed to allocate CUDA0 buffer of size 1041346560
19:10:54 arch ollama[927]: SIGSEGV: segmentation violation
19:10:54 arch ollama[927]: PC=0x5625c497b1cb m=5 sigcode=1 addr=0x0
19:10:54 arch ollama[927]: signal arrived during cgo execution
19:10:54 arch ollama[927]: goroutine 10 gp=0xc00058b180 m=5 mp=0xc000100008 [syscall]:
19:10:54 arch ollama[927]: runtime.cgocall(0x5625c4968190, 0xc0000491d8)
19:10:54 arch ollama[927]:         runtime/cgocall.go:167 +0x4b fp=0xc0000491b0 sp=0xc000049178 pc=0x5625c3c166eb
19:10:54 arch ollama[927]: github.com/ollama/ollama/llama._Cfunc_mtmd_encode_chunk(0x7fd89008df10, 0x7fd7cd7bc180)
19:10:54 arch ollama[927]:         _cgo_gotypes.go:1079 +0x4a fp=0xc0000491d8 sp=0xc0000491b0 pc=0x5625c3fd104a
19:10:54 arch ollama[927]: github.com/ollama/ollama/llama.(*MtmdContext).MultimodalTokenize.func11(...)
19:10:54 arch ollama[927]:         github.com/ollama/ollama/llama/llama.go:595
19:10:54 arch ollama[927]: github.com/ollama/ollama/llama.(*MtmdContext).MultimodalTokenize(0xc0004b2028, 0xc000346010, {0xc000a86000, 0x117bcb, 0xc000049520?})
19:10:54 arch ollama[927]:         github.com/ollama/ollama/llama/llama.go:595 +0x6c5 fp=0xc000049490 sp=0xc0000491d8 pc=0x5625c3fd6085
19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*ImageContext).MultimodalTokenize(0xc000051d40, 0xc000346010, {0xc000a86000, 0x117bcb, 0x117bcd})
19:10:54 arch ollama[927]:         github.com/ollama/ollama/runner/llamarunner/image.go:76 +0x145 fp=0xc000049530 sp=0xc000049490 pc=0x5625c4088365
19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).inputs(0xc000254780, {0xc000160400?, 0x3d?}, {0xc0000fa080, 0x1, 0x160?})
19:10:54 arch ollama[927]:         github.com/ollama/ollama/runner/llamarunner/runner.go:236 +0x2c6 fp=0xc000049698 sp=0xc000049530 pc=0x5625c4089766
19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).NewSequence(0xc000254780, {0xc000160400, 0x3d}, {0xc0000fa080, 0x1, 0x1}, {0x28000, {0xc0003d6060, 0x1, 0x1}, ...})
19:10:54 arch ollama[927]:         github.com/ollama/ollama/runner/llamarunner/runner.go:126 +0x8d fp=0xc000049838 sp=0xc000049698 pc=0x5625c4088d2d
19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).completion(0xc000254780, {0x5625c51ccfa0, 0xc0000e3b20}, 0xc000177040)
19:10:54 arch ollama[927]:         github.com/ollama/ollama/runner/llamarunner/runner.go:659 +0x5f9 fp=0xc000049ac0 sp=0xc000049838 pc=0x5625c408bd99
19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).completion-fm({0x5625c51ccfa0?, 0xc0000e3b20?}, 0xc000049b40?)
19:10:54 arch ollama[927]:         <autogenerated>:1 +0x36 fp=0xc000049af0 sp=0xc000049ac0 pc=0x5625c408f7f6
19:10:54 arch ollama[927]: net/http.HandlerFunc.ServeHTTP(0xc000051c80?, {0x5625c51ccfa0?, 0xc0000e3b20?}, 0xc000049b60?)
19:10:54 arch ollama[927]:         net/http/server.go:2294 +0x29 fp=0xc000049b18 sp=0xc000049af0 pc=0x5625c3f190e9
19:10:54 arch ollama[927]: net/http.(*ServeMux).ServeHTTP(0x5625c3bbe8c5?, {0x5625c51ccfa0, 0xc0000e3b20}, 0xc000177040)
19:10:54 arch ollama[927]:         net/http/server.go:2822 +0x1c4 fp=0xc000049b68 sp=0xc000049b18 pc=0x5625c3f1afe4
19:10:54 arch ollama[927]: net/http.serverHandler.ServeHTTP({0x5625c51c9590?}, {0x5625c51ccfa0?, 0xc0000e3b20?}, 0x1?)
19:10:54 arch ollama[927]:         net/http/server.go:3301 +0x8e fp=0xc000049b98 sp=0xc000049b68 pc=0x5625c3f38a6e
19:10:54 arch ollama[927]: net/http.(*conn).serve(0xc00023e3f0, {0x5625c51cf3d8, 0xc00023cfc0})
19:10:54 arch ollama[927]:         net/http/server.go:2102 +0x625 fp=0xc000049fb8 sp=0xc000049b98 pc=0x5625c3f175e5
19:10:54 arch ollama[927]: net/http.(*Server).Serve.gowrap3()
19:10:54 arch ollama[927]:         net/http/server.go:3454 +0x28 fp=0xc000049fe0 sp=0xc000049fb8 pc=0x5625c3f1cea8
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc000049fe8 sp=0xc000049fe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by net/http.(*Server).Serve in goroutine 1
19:10:54 arch ollama[927]:         net/http/server.go:3454 +0x485
19:10:54 arch ollama[927]: goroutine 1 gp=0xc000002380 m=nil [IO wait]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc0003fd790 sp=0xc0003fd770 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.netpollblock(0xc0005877e0?, 0xc3bb32a6?, 0x25?)
19:10:54 arch ollama[927]:         runtime/netpoll.go:575 +0xf7 fp=0xc0003fd7c8 sp=0xc0003fd790 pc=0x5625c3bdee97
19:10:54 arch ollama[927]: internal/poll.runtime_pollWait(0x7fd89dec5de0, 0x72)
19:10:54 arch ollama[927]:         runtime/netpoll.go:351 +0x85 fp=0xc0003fd7e8 sp=0xc0003fd7c8 pc=0x5625c3c18d85
19:10:54 arch ollama[927]: internal/poll.(*pollDesc).wait(0xc000170980?, 0x900000036?, 0x0)
19:10:54 arch ollama[927]:         internal/poll/fd_poll_runtime.go:84 +0x27 fp=0xc0003fd810 sp=0xc0003fd7e8 pc=0x5625c3ca0f07
19:10:54 arch ollama[927]: internal/poll.(*pollDesc).waitRead(...)
19:10:54 arch ollama[927]:         internal/poll/fd_poll_runtime.go:89
19:10:54 arch ollama[927]: internal/poll.(*FD).Accept(0xc000170980)
19:10:54 arch ollama[927]:         internal/poll/fd_unix.go:620 +0x295 fp=0xc0003fd8b8 sp=0xc0003fd810 pc=0x5625c3ca62d5
19:10:54 arch ollama[927]: net.(*netFD).accept(0xc000170980)
19:10:54 arch ollama[927]:         net/fd_unix.go:172 +0x29 fp=0xc0003fd970 sp=0xc0003fd8b8 pc=0x5625c3d191a9
19:10:54 arch ollama[927]: net.(*TCPListener).accept(0xc000305340)
19:10:54 arch ollama[927]:         net/tcpsock_posix.go:159 +0x1b fp=0xc0003fd9c0 sp=0xc0003fd970 pc=0x5625c3d2eb5b
19:10:54 arch ollama[927]: net.(*TCPListener).Accept(0xc000305340)
19:10:54 arch ollama[927]:         net/tcpsock.go:380 +0x30 fp=0xc0003fd9f0 sp=0xc0003fd9c0 pc=0x5625c3d2da10
19:10:54 arch ollama[927]: net/http.(*onceCloseListener).Accept(0xc00023e3f0?)
19:10:54 arch ollama[927]:         <autogenerated>:1 +0x24 fp=0xc0003fda08 sp=0xc0003fd9f0 pc=0x5625c3f451e4
19:10:54 arch ollama[927]: net/http.(*Server).Serve(0xc0001fb700, {0x5625c51ccdc0, 0xc000305340})
19:10:54 arch ollama[927]:         net/http/server.go:3424 +0x30c fp=0xc0003fdb38 sp=0xc0003fda08 pc=0x5625c3f1caac
19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.Execute({0xc000130140, 0x4, 0x4})
19:10:54 arch ollama[927]:         github.com/ollama/ollama/runner/llamarunner/runner.go:1002 +0x8f5 fp=0xc0003fdd08 sp=0xc0003fdb38 pc=0x5625c408f175
19:10:54 arch ollama[927]: github.com/ollama/ollama/runner.Execute({0xc000130130?, 0x0?, 0x0?})
19:10:54 arch ollama[927]:         github.com/ollama/ollama/runner/runner.go:22 +0xd4 fp=0xc0003fdd30 sp=0xc0003fdd08 pc=0x5625c413acf4
19:10:54 arch ollama[927]: github.com/ollama/ollama/cmd.NewCLI.func2(0xc0001fb400?, {0x5625c4caf0ad?, 0x4?, 0x5625c4caf0b1?})
19:10:54 arch ollama[927]:         github.com/ollama/ollama/cmd/cmd.go:1841 +0x45 fp=0xc0003fdd58 sp=0xc0003fdd30 pc=0x5625c48f7f25
19:10:54 arch ollama[927]: github.com/spf13/cobra.(*Command).execute(0xc00027d508, {0xc000305140, 0x4, 0x4})
19:10:54 arch ollama[927]:         github.com/spf13/cobra@v1.7.0/command.go:940 +0x85c fp=0xc0003fde78 sp=0xc0003fdd58 pc=0x5625c3d927fc
19:10:54 arch ollama[927]: github.com/spf13/cobra.(*Command).ExecuteC(0xc0000f4908)
19:10:54 arch ollama[927]:         github.com/spf13/cobra@v1.7.0/command.go:1068 +0x3a5 fp=0xc0003fdf30 sp=0xc0003fde78 pc=0x5625c3d93045
19:10:54 arch ollama[927]: github.com/spf13/cobra.(*Command).Execute(...)
19:10:54 arch ollama[927]:         github.com/spf13/cobra@v1.7.0/command.go:992
19:10:54 arch ollama[927]: github.com/spf13/cobra.(*Command).ExecuteContext(...)
19:10:54 arch ollama[927]:         github.com/spf13/cobra@v1.7.0/command.go:985
19:10:54 arch ollama[927]: main.main()
19:10:54 arch ollama[927]:         github.com/ollama/ollama/main.go:12 +0x4d fp=0xc0003fdf50 sp=0xc0003fdf30 pc=0x5625c48f8a0d
19:10:54 arch ollama[927]: runtime.main()
19:10:54 arch ollama[927]:         runtime/proc.go:283 +0x29d fp=0xc0003fdfe0 sp=0xc0003fdf50 pc=0x5625c3be651d
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc0003fdfe8 sp=0xc0003fdfe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: goroutine 2 gp=0xc000002e00 m=nil [force gc (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc00008efa8 sp=0xc00008ef88 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.goparkunlock(...)
19:10:54 arch ollama[927]:         runtime/proc.go:441
19:10:54 arch ollama[927]: runtime.forcegchelper()
19:10:54 arch ollama[927]:         runtime/proc.go:348 +0xb8 fp=0xc00008efe0 sp=0xc00008efa8 pc=0x5625c3be6858
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc00008efe8 sp=0xc00008efe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.init.7 in goroutine 1
19:10:54 arch ollama[927]:         runtime/proc.go:336 +0x1a
19:10:54 arch ollama[927]: goroutine 3 gp=0xc000003340 m=nil [GC sweep wait]:
19:10:54 arch ollama[927]: runtime.gopark(0x1?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc00008f780 sp=0xc00008f760 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.goparkunlock(...)
19:10:54 arch ollama[927]:         runtime/proc.go:441
19:10:54 arch ollama[927]: runtime.bgsweep(0xc0000ba000)
19:10:54 arch ollama[927]:         runtime/mgcsweep.go:316 +0xdf fp=0xc00008f7c8 sp=0xc00008f780 pc=0x5625c3bd0fff
19:10:54 arch ollama[927]: runtime.gcenable.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:204 +0x25 fp=0xc00008f7e0 sp=0xc00008f7c8 pc=0x5625c3bc53e5
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc00008f7e8 sp=0xc00008f7e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcenable in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:204 +0x66
19:10:54 arch ollama[927]: goroutine 4 gp=0xc000003500 m=nil [GC scavenge wait]:
19:10:54 arch ollama[927]: runtime.gopark(0x10000?, 0x5625c4e802a0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc00008ff78 sp=0xc00008ff58 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.goparkunlock(...)
19:10:54 arch ollama[927]:         runtime/proc.go:441
19:10:54 arch ollama[927]: runtime.(*scavengerState).park(0x5625c5aa3280)
19:10:54 arch ollama[927]:         runtime/mgcscavenge.go:425 +0x49 fp=0xc00008ffa8 sp=0xc00008ff78 pc=0x5625c3bcea49
19:10:54 arch ollama[927]: runtime.bgscavenge(0xc0000ba000)
19:10:54 arch ollama[927]:         runtime/mgcscavenge.go:658 +0x59 fp=0xc00008ffc8 sp=0xc00008ffa8 pc=0x5625c3bcefd9
19:10:54 arch ollama[927]: runtime.gcenable.gowrap2()
19:10:54 arch ollama[927]:         runtime/mgc.go:205 +0x25 fp=0xc00008ffe0 sp=0xc00008ffc8 pc=0x5625c3bc5385
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc00008ffe8 sp=0xc00008ffe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcenable in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:205 +0xa5
19:10:54 arch ollama[927]: goroutine 18 gp=0xc000102700 m=nil [finalizer wait]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x5625c51b9250?, 0x0?, 0x80?, 0x1000000010?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc00008e630 sp=0xc00008e610 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.runfinq()
19:10:54 arch ollama[927]:         runtime/mfinal.go:196 +0x107 fp=0xc00008e7e0 sp=0xc00008e630 pc=0x5625c3bc43a7
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc00008e7e8 sp=0xc00008e7e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.createfing in goroutine 1
19:10:54 arch ollama[927]:         runtime/mfinal.go:166 +0x3d
19:10:54 arch ollama[927]: goroutine 19 gp=0xc000103180 m=nil [chan receive]:
19:10:54 arch ollama[927]: runtime.gopark(0xc00022d720?, 0xc0004a0030?, 0x60?, 0xa7?, 0x5625c3cffde8?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc00008a718 sp=0xc00008a6f8 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.chanrecv(0xc000110310, 0x0, 0x1)
19:10:54 arch ollama[927]:         runtime/chan.go:664 +0x445 fp=0xc00008a790 sp=0xc00008a718 pc=0x5625c3bb5e85
19:10:54 arch ollama[927]: runtime.chanrecv1(0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/chan.go:506 +0x12 fp=0xc00008a7b8 sp=0xc00008a790 pc=0x5625c3bb5a12
19:10:54 arch ollama[927]: runtime.unique_runtime_registerUniqueMapCleanup.func2(...)
19:10:54 arch ollama[927]:         runtime/mgc.go:1796
19:10:54 arch ollama[927]: runtime.unique_runtime_registerUniqueMapCleanup.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1799 +0x2f fp=0xc00008a7e0 sp=0xc00008a7b8 pc=0x5625c3bc858f
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc00008a7e8 sp=0xc00008a7e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by unique.runtime_registerUniqueMapCleanup in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1794 +0x85
19:10:54 arch ollama[927]: goroutine 20 gp=0xc000103500 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc00008af38 sp=0xc00008af18 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc00008afc8 sp=0xc00008af38 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc00008afe0 sp=0xc00008afc8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc00008afe8 sp=0xc00008afe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 34 gp=0xc0003b6000 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000484738 sp=0xc000484718 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc0004847c8 sp=0xc000484738 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc0004847e0 sp=0xc0004847c8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc0004847e8 sp=0xc0004847e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 35 gp=0xc0003b61c0 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000484f38 sp=0xc000484f18 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc000484fc8 sp=0xc000484f38 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc000484fe0 sp=0xc000484fc8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc000484fe8 sp=0xc000484fe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 36 gp=0xc0003b6380 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000485738 sp=0xc000485718 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc0004857c8 sp=0xc000485738 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc0004857e0 sp=0xc0004857c8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc0004857e8 sp=0xc0004857e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 37 gp=0xc0003b6540 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000485f38 sp=0xc000485f18 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc000485fc8 sp=0xc000485f38 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc000485fe0 sp=0xc000485fc8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc000485fe8 sp=0xc000485fe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 38 gp=0xc0003b6700 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000486738 sp=0xc000486718 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc0004867c8 sp=0xc000486738 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc0004867e0 sp=0xc0004867c8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc0004867e8 sp=0xc0004867e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 39 gp=0xc0003b68c0 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000486f38 sp=0xc000486f18 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc000486fc8 sp=0xc000486f38 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc000486fe0 sp=0xc000486fc8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc000486fe8 sp=0xc000486fe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 21 gp=0xc0001036c0 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc00008b738 sp=0xc00008b718 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc00008b7c8 sp=0xc00008b738 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc00008b7e0 sp=0xc00008b7c8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc00008b7e8 sp=0xc00008b7e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 5 gp=0xc000003a40 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000090738 sp=0xc000090718 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc0000907c8 sp=0xc000090738 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc0000907e0 sp=0xc0000907c8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc0000907e8 sp=0xc0000907e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 40 gp=0xc0003b6a80 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000487738 sp=0xc000487718 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc0004877c8 sp=0xc000487738 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc0004877e0 sp=0xc0004877c8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc0004877e8 sp=0xc0004877e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 6 gp=0xc000003c00 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000090f38 sp=0xc000090f18 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc000090fc8 sp=0xc000090f38 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc000090fe0 sp=0xc000090fc8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc000090fe8 sp=0xc000090fe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 22 gp=0xc000103880 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc00008bf38 sp=0xc00008bf18 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc00008bfc8 sp=0xc00008bf38 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc00008bfe0 sp=0xc00008bfc8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc00008bfe8 sp=0xc00008bfe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 41 gp=0xc0003b6c40 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000487f38 sp=0xc000487f18 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc000487fc8 sp=0xc000487f38 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc000487fe0 sp=0xc000487fc8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc000487fe8 sp=0xc000487fe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 42 gp=0xc0003b6e00 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000480738 sp=0xc000480718 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc0004807c8 sp=0xc000480738 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc0004807e0 sp=0xc0004807c8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc0004807e8 sp=0xc0004807e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 43 gp=0xc0003b6fc0 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000480f38 sp=0xc000480f18 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc000480fc8 sp=0xc000480f38 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc000480fe0 sp=0xc000480fc8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc000480fe8 sp=0xc000480fe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 44 gp=0xc0003b7180 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000481738 sp=0xc000481718 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc0004817c8 sp=0xc000481738 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc0004817e0 sp=0xc0004817c8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc0004817e8 sp=0xc0004817e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 45 gp=0xc0003b7340 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x185132099240a?, 0x0?, 0x0?, 0x0?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000481f38 sp=0xc000481f18 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc000481fc8 sp=0xc000481f38 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc000481fe0 sp=0xc000481fc8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc000481fe8 sp=0xc000481fe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 7 gp=0xc000003dc0 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x5625c5b71680?, 0x1?, 0xfb?, 0x4c?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000091738 sp=0xc000091718 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc0000917c8 sp=0xc000091738 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc0000917e0 sp=0xc0000917c8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc0000917e8 sp=0xc0000917e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 23 gp=0xc000103a40 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x1851320999d4c?, 0x1?, 0xd9?, 0xbd?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc00008c738 sp=0xc00008c718 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc00008c7c8 sp=0xc00008c738 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc00008c7e0 sp=0xc00008c7c8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc00008c7e8 sp=0xc00008c7e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 24 gp=0xc000103c00 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x5625c5b71680?, 0x1?, 0xcb?, 0xaa?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc00008cf38 sp=0xc00008cf18 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc00008cfc8 sp=0xc00008cf38 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc00008cfe0 sp=0xc00008cfc8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc00008cfe8 sp=0xc00008cfe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 46 gp=0xc0003b7500 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x5625c5b71680?, 0x1?, 0x7a?, 0x93?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000482738 sp=0xc000482718 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc0004827c8 sp=0xc000482738 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc0004827e0 sp=0xc0004827c8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc0004827e8 sp=0xc0004827e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 25 gp=0xc000103dc0 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x5625c5b71680?, 0x1?, 0xbd?, 0x7f?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc00008d738 sp=0xc00008d718 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc00008d7c8 sp=0xc00008d738 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc00008d7e0 sp=0xc00008d7c8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc00008d7e8 sp=0xc00008d7e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 8 gp=0xc0000c6000 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x185132099abe5?, 0x3?, 0x9?, 0x81?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000091f38 sp=0xc000091f18 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc000091fc8 sp=0xc000091f38 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc000091fe0 sp=0xc000091fc8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc000091fe8 sp=0xc000091fe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 47 gp=0xc0003b76c0 m=nil [GC worker (idle)]:
19:10:54 arch ollama[927]: runtime.gopark(0x18513209989f8?, 0x1?, 0xc?, 0x18?, 0x0?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc000482f38 sp=0xc000482f18 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730)
19:10:54 arch ollama[927]:         runtime/mgc.go:1423 +0xe9 fp=0xc000482fc8 sp=0xc000482f38 pc=0x5625c3bc78a9
19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1()
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x25 fp=0xc000482fe0 sp=0xc000482fc8 pc=0x5625c3bc7785
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc000482fe8 sp=0xc000482fe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1
19:10:54 arch ollama[927]:         runtime/mgc.go:1339 +0x105
19:10:54 arch ollama[927]: goroutine 9 gp=0xc00058afc0 m=nil [sync.Cond.Wait]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0xc0003adc78?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc0003adbe8 sp=0xc0003adbc8 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.goparkunlock(...)
19:10:54 arch ollama[927]:         runtime/proc.go:441
19:10:54 arch ollama[927]: sync.runtime_notifyListWait(0xc000305310, 0x0)
19:10:54 arch ollama[927]:         runtime/sema.go:597 +0x15a fp=0xc0003adc38 sp=0xc0003adbe8 pc=0x5625c3c1b6ba
19:10:54 arch ollama[927]: sync.(*Cond).Wait(0xc0003adcb8?)
19:10:54 arch ollama[927]:         sync/cond.go:71 +0x85 fp=0xc0003adc70 sp=0xc0003adc38 pc=0x5625c3c2b7c5
19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).processBatch(0xc000254780, 0xc00048a190, 0xc00048a1e0)
19:10:54 arch ollama[927]:         github.com/ollama/ollama/runner/llamarunner/runner.go:408 +0x93 fp=0xc0003adee8 sp=0xc0003adc70 pc=0x5625c408a213
19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).run(0xc000254780, {0x5625c51cf410, 0xc000018730})
19:10:54 arch ollama[927]:         github.com/ollama/ollama/runner/llamarunner/runner.go:387 +0x1d5 fp=0xc0003adfb8 sp=0xc0003adee8 pc=0x5625c408a015
19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.Execute.gowrap1()
19:10:54 arch ollama[927]:         github.com/ollama/ollama/runner/llamarunner/runner.go:981 +0x28 fp=0xc0003adfe0 sp=0xc0003adfb8 pc=0x5625c408f3e8
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc0003adfe8 sp=0xc0003adfe0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by github.com/ollama/ollama/runner/llamarunner.Execute in goroutine 1
19:10:54 arch ollama[927]:         github.com/ollama/ollama/runner/llamarunner/runner.go:981 +0x4c5
19:10:54 arch ollama[927]: goroutine 98 gp=0xc0003b7dc0 m=nil [IO wait]:
19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0xb?)
19:10:54 arch ollama[927]:         runtime/proc.go:435 +0xce fp=0xc0003ac5d8 sp=0xc0003ac5b8 pc=0x5625c3c19b6e
19:10:54 arch ollama[927]: runtime.netpollblock(0x5625c3c3d338?, 0xc3bb32a6?, 0x25?)
19:10:54 arch ollama[927]:         runtime/netpoll.go:575 +0xf7 fp=0xc0003ac610 sp=0xc0003ac5d8 pc=0x5625c3bdee97
19:10:54 arch ollama[927]: internal/poll.runtime_pollWait(0x7fd89dec5cc8, 0x72)
19:10:54 arch ollama[927]:         runtime/netpoll.go:351 +0x85 fp=0xc0003ac630 sp=0xc0003ac610 pc=0x5625c3c18d85
19:10:54 arch ollama[927]: internal/poll.(*pollDesc).wait(0xc000170a00?, 0xc00023d0c1?, 0x0)
19:10:54 arch ollama[927]:         internal/poll/fd_poll_runtime.go:84 +0x27 fp=0xc0003ac658 sp=0xc0003ac630 pc=0x5625c3ca0f07
19:10:54 arch ollama[927]: internal/poll.(*pollDesc).waitRead(...)
19:10:54 arch ollama[927]:         internal/poll/fd_poll_runtime.go:89
19:10:54 arch ollama[927]: internal/poll.(*FD).Read(0xc000170a00, {0xc00023d0c1, 0x1, 0x1})
19:10:54 arch ollama[927]:         internal/poll/fd_unix.go:165 +0x27a fp=0xc0003ac6f0 sp=0xc0003ac658 pc=0x5625c3ca21fa
19:10:54 arch ollama[927]: net.(*netFD).Read(0xc000170a00, {0xc00023d0c1?, 0x0?, 0x0?})
19:10:54 arch ollama[927]:         net/fd_posix.go:55 +0x25 fp=0xc0003ac738 sp=0xc0003ac6f0 pc=0x5625c3d17205
19:10:54 arch ollama[927]: net.(*conn).Read(0xc00011ca30, {0xc00023d0c1?, 0xc0003ac700?, 0x0?})
19:10:54 arch ollama[927]:         net/net.go:194 +0x45 fp=0xc0003ac780 sp=0xc0003ac738 pc=0x5625c3d255c5
19:10:54 arch ollama[927]: net/http.(*connReader).backgroundRead(0xc00023d0b0)
19:10:54 arch ollama[927]:         net/http/server.go:690 +0x37 fp=0xc0003ac7c8 sp=0xc0003ac780 pc=0x5625c3f114b7
19:10:54 arch ollama[927]: net/http.(*connReader).startBackgroundRead.gowrap2()
19:10:54 arch ollama[927]:         net/http/server.go:686 +0x25 fp=0xc0003ac7e0 sp=0xc0003ac7c8 pc=0x5625c3f113e5
19:10:54 arch ollama[927]: runtime.goexit({})
19:10:54 arch ollama[927]:         runtime/asm_amd64.s:1700 +0x1 fp=0xc0003ac7e8 sp=0xc0003ac7e0 pc=0x5625c3c21a01
19:10:54 arch ollama[927]: created by net/http.(*connReader).startBackgroundRead in goroutine 10
19:10:54 arch ollama[927]:         net/http/server.go:686 +0xb6
19:10:54 arch ollama[927]: rax    0x7fd7ca850b90
19:10:54 arch ollama[927]: rbx    0x0
19:10:54 arch ollama[927]: rcx    0x0
19:10:54 arch ollama[927]: rdx    0x0
19:10:54 arch ollama[927]: rdi    0x7fd7cd7bc0e0
19:10:54 arch ollama[927]: rsi    0x7fd7be22f310
19:10:54 arch ollama[927]: rbp    0x7fd7ca5e8148
19:10:54 arch ollama[927]: rsp    0x7fd89deb9bc0
19:10:54 arch ollama[927]: r8     0x0
19:10:54 arch ollama[927]: r9     0x7fd771f0b040
19:10:54 arch ollama[927]: r10    0x24db000
19:10:54 arch ollama[927]: r11    0x246
19:10:54 arch ollama[927]: r12    0x7fd6bc45ac80
19:10:54 arch ollama[927]: r13    0x160
19:10:54 arch ollama[927]: r14    0x7fd7be22f030
19:10:54 arch ollama[927]: r15    0x1
19:10:54 arch ollama[927]: rip    0x5625c497b1cb
19:10:54 arch ollama[927]: rflags 0x10246
19:10:54 arch ollama[927]: cs     0x33
19:10:54 arch ollama[927]: fs     0x0
19:10:54 arch ollama[927]: gs     0x0
19:10:54 arch ollama[927]: time=2026-01-05T19:10:54.181+08:00 level=ERROR source=server.go:1583 msg="post predict" error="Post \"http://127.0.0.1:41297/completion\": EOF"
19:10:54 arch ollama[927]: [GIN] 2026/01/05 - 19:10:54 | 500 |  9.456008317s |       127.0.0.1 | POST     "/api/generate"
19:10:54 arch ollama[927]: time=2026-01-05T19:10:54.331+08:00 level=ERROR source=server.go:302 msg="llama runner terminated" error="exit status 2"
<!-- gh-comment-id:3710011375 --> @pinghe commented on GitHub (Jan 5, 2026): journalctl -u ollama --since "1 hour ago" Full log ``` 19:10:44 arch ollama[927]: [GIN] 2026/01/05 - 19:10:44 | 200 | 44.465µs | 127.0.0.1 | HEAD "/" 19:10:44 arch ollama[927]: [GIN] 2026/01/05 - 19:10:44 | 200 | 150.97054ms | 127.0.0.1 | POST "/api/show" 19:10:45 arch ollama[927]: time=2026-01-05T19:10:45.073+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 36103" 19:10:45 arch ollama[927]: time=2026-01-05T19:10:45.440+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 45011" 19:10:45 arch ollama[927]: time=2026-01-05T19:10:45.690+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 38851" 19:10:45 arch ollama[927]: time=2026-01-05T19:10:45.940+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 45395" 19:10:46 arch ollama[927]: time=2026-01-05T19:10:46.190+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 45955" 19:10:46 arch ollama[927]: time=2026-01-05T19:10:46.439+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 37417" 19:10:46 arch ollama[927]: time=2026-01-05T19:10:46.690+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 39245" 19:10:46 arch ollama[927]: time=2026-01-05T19:10:46.940+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 45121" 19:10:47 arch ollama[927]: time=2026-01-05T19:10:47.190+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 37815" 19:10:47 arch ollama[927]: time=2026-01-05T19:10:47.440+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 40155" 19:10:47 arch ollama[927]: time=2026-01-05T19:10:47.690+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 43019" 19:10:47 arch ollama[927]: time=2026-01-05T19:10:47.940+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 37159" 19:10:48 arch ollama[927]: time=2026-01-05T19:10:48.190+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 36817" 19:10:48 arch ollama[927]: time=2026-01-05T19:10:48.440+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 34387" 19:10:48 arch ollama[927]: time=2026-01-05T19:10:48.690+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 45519" 19:10:48 arch ollama[927]: time=2026-01-05T19:10:48.940+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 44193" 19:10:49 arch ollama[927]: time=2026-01-05T19:10:49.190+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 42021" 19:10:49 arch ollama[927]: time=2026-01-05T19:10:49.440+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 37207" 19:10:49 arch ollama[927]: time=2026-01-05T19:10:49.690+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 46387" 19:10:49 arch ollama[927]: time=2026-01-05T19:10:49.940+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 45197" 19:10:50 arch ollama[927]: time=2026-01-05T19:10:50.190+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --ollama-engine --port 44683" 19:10:50 arch ollama[927]: llama_model_loader: loaded meta data with 33 key-value pairs and 523 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20af820432e1dd55d22a2b0e728e79223e77a (version GGUF V3 (latest)) 19:10:50 arch ollama[927]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. 19:10:50 arch ollama[927]: llama_model_loader: - kv 0: general.architecture str = glm4 19:10:50 arch ollama[927]: llama_model_loader: - kv 1: general.type str = model 19:10:50 arch ollama[927]: llama_model_loader: - kv 2: general.size_label str = 9.4B 19:10:50 arch ollama[927]: llama_model_loader: - kv 3: general.license str = mit 19:10:50 arch ollama[927]: llama_model_loader: - kv 4: general.base_model.count u32 = 1 19:10:50 arch ollama[927]: llama_model_loader: - kv 5: general.base_model.0.name str = GLM 4.1V 9B Base 19:10:50 arch ollama[927]: llama_model_loader: - kv 6: general.base_model.0.organization str = Zai Org 19:10:50 arch ollama[927]: llama_model_loader: - kv 7: general.base_model.0.repo_url str = https://huggingface.co/zai-org/GLM-4.... 19:10:50 arch ollama[927]: llama_model_loader: - kv 8: general.tags arr[str,2] = ["agent", "image-text-to-text"] 19:10:50 arch ollama[927]: llama_model_loader: - kv 9: general.languages arr[str,1] = ["zh"] 19:10:50 arch ollama[927]: llama_model_loader: - kv 10: glm4.block_count u32 = 40 19:10:50 arch ollama[927]: llama_model_loader: - kv 11: glm4.context_length u32 = 65536 19:10:50 arch ollama[927]: llama_model_loader: - kv 12: glm4.embedding_length u32 = 4096 19:10:50 arch ollama[927]: llama_model_loader: - kv 13: glm4.feed_forward_length u32 = 13696 19:10:50 arch ollama[927]: llama_model_loader: - kv 14: glm4.attention.head_count u32 = 32 19:10:50 arch ollama[927]: llama_model_loader: - kv 15: glm4.attention.head_count_kv u32 = 2 19:10:50 arch ollama[927]: llama_model_loader: - kv 16: glm4.rope.dimension_sections arr[i32,4] = [8, 12, 12, 0] 19:10:50 arch ollama[927]: llama_model_loader: - kv 17: glm4.rope.freq_base f32 = 10000.000000 19:10:50 arch ollama[927]: llama_model_loader: - kv 18: glm4.attention.layer_norm_rms_epsilon f32 = 0.000010 19:10:50 arch ollama[927]: llama_model_loader: - kv 19: glm4.rope.dimension_count u32 = 64 19:10:50 arch ollama[927]: llama_model_loader: - kv 20: tokenizer.ggml.model str = gpt2 19:10:50 arch ollama[927]: llama_model_loader: - kv 21: tokenizer.ggml.pre str = glm4 19:10:50 arch ollama[927]: llama_model_loader: - kv 22: tokenizer.ggml.tokens arr[str,151552] = ["!", "\"", "#", "$", "%", "&", "'", ... 19:10:50 arch ollama[927]: llama_model_loader: - kv 23: tokenizer.ggml.token_type arr[i32,151552] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... 19:10:50 arch ollama[927]: llama_model_loader: - kv 24: tokenizer.ggml.merges arr[str,318088] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "... 19:10:50 arch ollama[927]: llama_model_loader: - kv 25: tokenizer.ggml.eos_token_id u32 = 151329 19:10:50 arch ollama[927]: llama_model_loader: - kv 26: tokenizer.ggml.padding_token_id u32 = 151329 19:10:50 arch ollama[927]: llama_model_loader: - kv 27: tokenizer.ggml.eot_token_id u32 = 151336 19:10:50 arch ollama[927]: llama_model_loader: - kv 28: tokenizer.ggml.unknown_token_id u32 = 151329 19:10:50 arch ollama[927]: llama_model_loader: - kv 29: tokenizer.ggml.bos_token_id u32 = 151329 19:10:50 arch ollama[927]: llama_model_loader: - kv 30: tokenizer.chat_template str = [gMASK]<sop>\n{%- for msg in messages ... 19:10:50 arch ollama[927]: llama_model_loader: - kv 31: general.quantization_version u32 = 2 19:10:50 arch ollama[927]: llama_model_loader: - kv 32: general.file_type u32 = 15 19:10:50 arch ollama[927]: llama_model_loader: - type f32: 281 tensors 19:10:50 arch ollama[927]: llama_model_loader: - type q5_0: 20 tensors 19:10:50 arch ollama[927]: llama_model_loader: - type q8_0: 20 tensors 19:10:50 arch ollama[927]: llama_model_loader: - type q4_K: 181 tensors 19:10:50 arch ollama[927]: llama_model_loader: - type q6_K: 21 tensors 19:10:50 arch ollama[927]: print_info: file format = GGUF V3 (latest) 19:10:50 arch ollama[927]: print_info: file type = Q4_K - Medium 19:10:50 arch ollama[927]: print_info: file size = 5.73 GiB (5.24 BPW) 19:10:51 arch ollama[927]: load: special_eot_id is not in special_eog_ids - the tokenizer config may be incorrect 19:10:51 arch ollama[927]: load: printing all EOG tokens: 19:10:51 arch ollama[927]: load: - 151329 ('<|endoftext|>') 19:10:51 arch ollama[927]: load: - 151336 ('<|user|>') 19:10:51 arch ollama[927]: load: special tokens cache size = 23 19:10:51 arch ollama[927]: load: token to piece cache size = 0.9711 MB 19:10:51 arch ollama[927]: print_info: arch = glm4 19:10:51 arch ollama[927]: print_info: vocab_only = 1 19:10:51 arch ollama[927]: print_info: no_alloc = 0 19:10:51 arch ollama[927]: print_info: model type = ?B 19:10:51 arch ollama[927]: print_info: model params = 9.40 B 19:10:51 arch ollama[927]: print_info: general.name = n/a 19:10:51 arch ollama[927]: print_info: vocab type = BPE 19:10:51 arch ollama[927]: print_info: n_vocab = 151552 19:10:51 arch ollama[927]: print_info: n_merges = 318088 19:10:51 arch ollama[927]: print_info: BOS token = 151329 '<|endoftext|>' 19:10:51 arch ollama[927]: print_info: EOS token = 151329 '<|endoftext|>' 19:10:51 arch ollama[927]: print_info: EOT token = 151336 '<|user|>' 19:10:51 arch ollama[927]: print_info: UNK token = 151329 '<|endoftext|>' 19:10:51 arch ollama[927]: print_info: PAD token = 151329 '<|endoftext|>' 19:10:51 arch ollama[927]: print_info: LF token = 198 'Ċ' 19:10:51 arch ollama[927]: print_info: EOG token = 151329 '<|endoftext|>' 19:10:51 arch ollama[927]: print_info: EOG token = 151336 '<|user|>' 19:10:51 arch ollama[927]: print_info: max token length = 1024 19:10:51 arch ollama[927]: llama_model_load: vocab only - skipping tensors 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.249+08:00 level=INFO source=server.go:429 msg="starting runner" cmd="/usr/local/bin/ollama runner --model /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20af820432e1dd55d22a2b0e728e79223e77a --port 41297" 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=sched.go:443 msg="system memory" total="62.6 GiB" free="43.2 GiB" free_swap="71.6 MiB" 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=sched.go:450 msg="gpu memory" id=GPU-8215c551-6dde-569b-490d-884f3ab7a437 library=CUDA available="6.6 GiB" free="7.1 GiB" minimum="457.0 MiB" overhead="0 B" 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=server.go:496 msg="loading model" "model layers"=41 requested=-1 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=device.go:240 msg="model weights" device=CUDA0 size="4.2 GiB" 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=device.go:245 msg="model weights" device=CPU size="1.3 GiB" 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=device.go:251 msg="kv cache" device=CUDA0 size="544.0 MiB" 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=device.go:256 msg="kv cache" device=CPU size="96.0 MiB" 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=device.go:262 msg="compute graph" device=CUDA0 size="1.7 GiB" 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.250+08:00 level=INFO source=device.go:272 msg="total memory" size="7.7 GiB" 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.266+08:00 level=INFO source=runner.go:965 msg="starting go runner" 19:10:51 arch ollama[927]: load_backend: loaded CPU backend from /usr/local/lib/ollama/libggml-cpu-haswell.so 19:10:51 arch ollama[927]: ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no 19:10:51 arch ollama[927]: ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no 19:10:51 arch ollama[927]: ggml_cuda_init: found 1 CUDA devices: 19:10:51 arch ollama[927]: Device 0: NVIDIA GeForce RTX 2060 SUPER, compute capability 7.5, VMM: yes, ID: GPU-8215c551-6dde-569b-490d-884f3ab7a437 19:10:51 arch ollama[927]: load_backend: loaded CUDA backend from /usr/local/lib/ollama/cuda_v13/libggml-cuda.so 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.341+08:00 level=INFO source=ggml.go:104 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.BMI2=1 CPU.0.LLAMAFILE=1 CPU.1.LLAMAFILE=1 CUDA.0.ARCHS=750,800,860,870,890,900,1000,1030,1100,1200,1210 CUDA.0.USE_GRAPHS=1 CUDA.0.PEER_MAX_BATCH_SIZE=128 compiler=cgo(gcc) 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.342+08:00 level=INFO source=runner.go:1001 msg="Server listening on 127.0.0.1:41297" 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.348+08:00 level=INFO source=runner.go:895 msg=load request="{Operation:commit LoraPath:[] Parallel:1 BatchSize:512 FlashAttention:Auto KvSize:16384 KvCacheType: NumThreads:12 GPULayers:34[ID:GPU-8215c551-6dde-569b-490d-884f3ab7a437 Layers:34(6..39)] MultiUserCache:false ProjectorPath:/usr/share/ollama/.ollama/models/blobs/sha256-454dd441c925c1a81c984204bb9d54feef0ef07b789c6fe1118099014ba2727d MainGPU:0 UseMmap:true}" 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.348+08:00 level=INFO source=server.go:1338 msg="waiting for llama runner to start responding" 19:10:51 arch ollama[927]: time=2026-01-05T19:10:51.349+08:00 level=INFO source=server.go:1372 msg="waiting for server to become available" status="llm server loading model" 19:10:51 arch ollama[927]: ggml_backend_cuda_device_get_memory device GPU-8215c551-6dde-569b-490d-884f3ab7a437 utilizing NVML memory reporting free: 7585398784 total: 8589934592 19:10:51 arch ollama[927]: llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce RTX 2060 SUPER) (0000:03:00.0) - 7234 MiB free 19:10:51 arch ollama[927]: llama_model_loader: loaded meta data with 33 key-value pairs and 523 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-004fd25079bfce8caa7363df20c20af820432e1dd55d22a2b0e728e79223e77a (version GGUF V3 (latest)) 19:10:51 arch ollama[927]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. 19:10:51 arch ollama[927]: llama_model_loader: - kv 0: general.architecture str = glm4 19:10:51 arch ollama[927]: llama_model_loader: - kv 1: general.type str = model 19:10:51 arch ollama[927]: llama_model_loader: - kv 2: general.size_label str = 9.4B 19:10:51 arch ollama[927]: llama_model_loader: - kv 3: general.license str = mit 19:10:51 arch ollama[927]: llama_model_loader: - kv 4: general.base_model.count u32 = 1 19:10:51 arch ollama[927]: llama_model_loader: - kv 5: general.base_model.0.name str = GLM 4.1V 9B Base 19:10:51 arch ollama[927]: llama_model_loader: - kv 6: general.base_model.0.organization str = Zai Org 19:10:51 arch ollama[927]: llama_model_loader: - kv 7: general.base_model.0.repo_url str = https://huggingface.co/zai-org/GLM-4.... 19:10:51 arch ollama[927]: llama_model_loader: - kv 8: general.tags arr[str,2] = ["agent", "image-text-to-text"] 19:10:51 arch ollama[927]: llama_model_loader: - kv 9: general.languages arr[str,1] = ["zh"] 19:10:51 arch ollama[927]: llama_model_loader: - kv 10: glm4.block_count u32 = 40 19:10:51 arch ollama[927]: llama_model_loader: - kv 11: glm4.context_length u32 = 65536 19:10:51 arch ollama[927]: llama_model_loader: - kv 12: glm4.embedding_length u32 = 4096 19:10:51 arch ollama[927]: llama_model_loader: - kv 13: glm4.feed_forward_length u32 = 13696 19:10:51 arch ollama[927]: llama_model_loader: - kv 14: glm4.attention.head_count u32 = 32 19:10:51 arch ollama[927]: llama_model_loader: - kv 15: glm4.attention.head_count_kv u32 = 2 19:10:51 arch ollama[927]: llama_model_loader: - kv 16: glm4.rope.dimension_sections arr[i32,4] = [8, 12, 12, 0] 19:10:51 arch ollama[927]: llama_model_loader: - kv 17: glm4.rope.freq_base f32 = 10000.000000 19:10:51 arch ollama[927]: llama_model_loader: - kv 18: glm4.attention.layer_norm_rms_epsilon f32 = 0.000010 19:10:51 arch ollama[927]: llama_model_loader: - kv 19: glm4.rope.dimension_count u32 = 64 19:10:51 arch ollama[927]: llama_model_loader: - kv 20: tokenizer.ggml.model str = gpt2 19:10:51 arch ollama[927]: llama_model_loader: - kv 21: tokenizer.ggml.pre str = glm4 19:10:51 arch ollama[927]: llama_model_loader: - kv 22: tokenizer.ggml.tokens arr[str,151552] = ["!", "\"", "#", "$", "%", "&", "'", ... 19:10:51 arch ollama[927]: llama_model_loader: - kv 23: tokenizer.ggml.token_type arr[i32,151552] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... 19:10:51 arch ollama[927]: llama_model_loader: - kv 24: tokenizer.ggml.merges arr[str,318088] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "... 19:10:51 arch ollama[927]: llama_model_loader: - kv 25: tokenizer.ggml.eos_token_id u32 = 151329 19:10:51 arch ollama[927]: llama_model_loader: - kv 26: tokenizer.ggml.padding_token_id u32 = 151329 19:10:51 arch ollama[927]: llama_model_loader: - kv 27: tokenizer.ggml.eot_token_id u32 = 151336 19:10:51 arch ollama[927]: llama_model_loader: - kv 28: tokenizer.ggml.unknown_token_id u32 = 151329 19:10:51 arch ollama[927]: llama_model_loader: - kv 29: tokenizer.ggml.bos_token_id u32 = 151329 19:10:51 arch ollama[927]: llama_model_loader: - kv 30: tokenizer.chat_template str = [gMASK]<sop>\n{%- for msg in messages ... 19:10:51 arch ollama[927]: llama_model_loader: - kv 31: general.quantization_version u32 = 2 19:10:51 arch ollama[927]: llama_model_loader: - kv 32: general.file_type u32 = 15 19:10:51 arch ollama[927]: llama_model_loader: - type f32: 281 tensors 19:10:51 arch ollama[927]: llama_model_loader: - type q5_0: 20 tensors 19:10:51 arch ollama[927]: llama_model_loader: - type q8_0: 20 tensors 19:10:51 arch ollama[927]: llama_model_loader: - type q4_K: 181 tensors 19:10:51 arch ollama[927]: llama_model_loader: - type q6_K: 21 tensors 19:10:51 arch ollama[927]: print_info: file format = GGUF V3 (latest) 19:10:51 arch ollama[927]: print_info: file type = Q4_K - Medium 19:10:51 arch ollama[927]: print_info: file size = 5.73 GiB (5.24 BPW) 19:10:51 arch ollama[927]: load: special_eot_id is not in special_eog_ids - the tokenizer config may be incorrect 19:10:51 arch ollama[927]: load: printing all EOG tokens: 19:10:51 arch ollama[927]: load: - 151329 ('<|endoftext|>') 19:10:51 arch ollama[927]: load: - 151336 ('<|user|>') 19:10:51 arch ollama[927]: load: special tokens cache size = 23 19:10:51 arch ollama[927]: load: token to piece cache size = 0.9711 MB 19:10:51 arch ollama[927]: print_info: arch = glm4 19:10:51 arch ollama[927]: print_info: vocab_only = 0 19:10:51 arch ollama[927]: print_info: no_alloc = 0 19:10:51 arch ollama[927]: print_info: n_ctx_train = 65536 19:10:51 arch ollama[927]: print_info: n_embd = 4096 19:10:51 arch ollama[927]: print_info: n_embd_inp = 4096 19:10:51 arch ollama[927]: print_info: n_layer = 40 19:10:51 arch ollama[927]: print_info: n_head = 32 19:10:51 arch ollama[927]: print_info: n_head_kv = 2 19:10:51 arch ollama[927]: print_info: n_rot = 64 19:10:51 arch ollama[927]: print_info: n_swa = 0 19:10:51 arch ollama[927]: print_info: is_swa_any = 0 19:10:51 arch ollama[927]: print_info: n_embd_head_k = 128 19:10:51 arch ollama[927]: print_info: n_embd_head_v = 128 19:10:51 arch ollama[927]: print_info: n_gqa = 16 19:10:51 arch ollama[927]: print_info: n_embd_k_gqa = 256 19:10:51 arch ollama[927]: print_info: n_embd_v_gqa = 256 19:10:51 arch ollama[927]: print_info: f_norm_eps = 0.0e+00 19:10:51 arch ollama[927]: print_info: f_norm_rms_eps = 1.0e-05 19:10:51 arch ollama[927]: print_info: f_clamp_kqv = 0.0e+00 19:10:51 arch ollama[927]: print_info: f_max_alibi_bias = 0.0e+00 19:10:51 arch ollama[927]: print_info: f_logit_scale = 0.0e+00 19:10:51 arch ollama[927]: print_info: f_attn_scale = 0.0e+00 19:10:51 arch ollama[927]: print_info: n_ff = 13696 19:10:51 arch ollama[927]: print_info: n_expert = 0 19:10:51 arch ollama[927]: print_info: n_expert_used = 0 19:10:51 arch ollama[927]: print_info: n_expert_groups = 0 19:10:51 arch ollama[927]: print_info: n_group_used = 0 19:10:51 arch ollama[927]: print_info: causal attn = 1 19:10:51 arch ollama[927]: print_info: pooling type = 0 19:10:51 arch ollama[927]: print_info: rope type = 8 19:10:51 arch ollama[927]: print_info: rope scaling = linear 19:10:51 arch ollama[927]: print_info: freq_base_train = 10000.0 19:10:51 arch ollama[927]: print_info: freq_scale_train = 1 19:10:51 arch ollama[927]: print_info: n_ctx_orig_yarn = 65536 19:10:51 arch ollama[927]: print_info: rope_yarn_log_mul= 0.0000 19:10:51 arch ollama[927]: print_info: rope_finetuned = unknown 19:10:51 arch ollama[927]: print_info: mrope sections = [8, 12, 12, 0] 19:10:51 arch ollama[927]: print_info: model type = 9B 19:10:51 arch ollama[927]: print_info: model params = 9.40 B 19:10:51 arch ollama[927]: print_info: general.name = n/a 19:10:51 arch ollama[927]: print_info: vocab type = BPE 19:10:51 arch ollama[927]: print_info: n_vocab = 151552 19:10:51 arch ollama[927]: print_info: n_merges = 318088 19:10:51 arch ollama[927]: print_info: BOS token = 151329 '<|endoftext|>' 19:10:51 arch ollama[927]: print_info: EOS token = 151329 '<|endoftext|>' 19:10:51 arch ollama[927]: print_info: EOT token = 151336 '<|user|>' 19:10:51 arch ollama[927]: print_info: UNK token = 151329 '<|endoftext|>' 19:10:51 arch ollama[927]: print_info: PAD token = 151329 '<|endoftext|>' 19:10:51 arch ollama[927]: print_info: LF token = 198 'Ċ' 19:10:51 arch ollama[927]: print_info: EOG token = 151329 '<|endoftext|>' 19:10:51 arch ollama[927]: print_info: EOG token = 151336 '<|user|>' 19:10:51 arch ollama[927]: print_info: max token length = 1024 19:10:51 arch ollama[927]: load_tensors: loading model tensors, this can take a while... (mmap = true) 19:10:52 arch ollama[927]: load_tensors: offloading 34 repeating layers to GPU 19:10:52 arch ollama[927]: load_tensors: offloaded 34/41 layers to GPU 19:10:52 arch ollama[927]: load_tensors: CPU_Mapped model buffer size = 1617.29 MiB 19:10:52 arch ollama[927]: load_tensors: CUDA0 model buffer size = 4254.72 MiB 19:10:53 arch ollama[927]: llama_context: constructing llama_context 19:10:53 arch ollama[927]: llama_context: n_seq_max = 1 19:10:53 arch ollama[927]: llama_context: n_ctx = 16384 19:10:53 arch ollama[927]: llama_context: n_ctx_seq = 16384 19:10:53 arch ollama[927]: llama_context: n_batch = 512 19:10:53 arch ollama[927]: llama_context: n_ubatch = 512 19:10:53 arch ollama[927]: llama_context: causal_attn = 1 19:10:53 arch ollama[927]: llama_context: flash_attn = auto 19:10:53 arch ollama[927]: llama_context: kv_unified = false 19:10:53 arch ollama[927]: llama_context: freq_base = 10000.0 19:10:53 arch ollama[927]: llama_context: freq_scale = 1 19:10:53 arch ollama[927]: llama_context: n_ctx_seq (16384) < n_ctx_train (65536) -- the full capacity of the model will not be utilized 19:10:53 arch ollama[927]: llama_context: CPU output buffer size = 0.59 MiB 19:10:53 arch ollama[927]: llama_kv_cache: CPU KV buffer size = 96.00 MiB 19:10:53 arch ollama[927]: llama_kv_cache: CUDA0 KV buffer size = 544.00 MiB 19:10:53 arch ollama[927]: llama_kv_cache: size = 640.00 MiB ( 16384 cells, 40 layers, 1/1 seqs), K (f16): 320.00 MiB, V (f16): 320.00 MiB 19:10:53 arch ollama[927]: llama_context: Flash Attention was auto, set to enabled 19:10:53 arch ollama[927]: llama_context: CUDA0 compute buffer size = 789.62 MiB 19:10:53 arch ollama[927]: llama_context: CUDA_Host compute buffer size = 40.02 MiB 19:10:53 arch ollama[927]: llama_context: graph nodes = 1487 19:10:53 arch ollama[927]: llama_context: graph splits = 94 (with bs=512), 3 (with bs=1) 19:10:53 arch ollama[927]: clip_model_loader: model name: 19:10:53 arch ollama[927]: clip_model_loader: description: 19:10:53 arch ollama[927]: clip_model_loader: GGUF version: 3 19:10:53 arch ollama[927]: clip_model_loader: alignment: 32 19:10:53 arch ollama[927]: clip_model_loader: n_tensors: 182 19:10:53 arch ollama[927]: clip_model_loader: n_kv: 25 19:10:53 arch ollama[927]: clip_model_loader: has vision encoder 19:10:53 arch ollama[927]: clip_model_loader: tensor[0]: n_dims = 2, name = v.blk.0.attn_out.weight, tensor_size=2506752, offset=0, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[1]: n_dims = 2, name = v.blk.0.attn_qkv.weight, tensor_size=7520256, offset=2506752, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[2]: n_dims = 2, name = v.blk.0.ffn_down.weight, tensor_size=6684672, offset=10027008, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[3]: n_dims = 2, name = v.blk.0.ffn_gate.weight, tensor_size=6684672, offset=16711680, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[4]: n_dims = 2, name = v.blk.0.ffn_up.weight, tensor_size=6684672, offset=23396352, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[5]: n_dims = 1, name = v.blk.0.ln1.weight, tensor_size=6144, offset=30081024, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[6]: n_dims = 1, name = v.blk.0.ln2.weight, tensor_size=6144, offset=30087168, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[7]: n_dims = 2, name = v.blk.1.attn_out.weight, tensor_size=2506752, offset=30093312, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[8]: n_dims = 2, name = v.blk.1.attn_qkv.weight, tensor_size=7520256, offset=32600064, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[9]: n_dims = 2, name = v.blk.1.ffn_down.weight, tensor_size=6684672, offset=40120320, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[10]: n_dims = 2, name = v.blk.1.ffn_gate.weight, tensor_size=6684672, offset=46804992, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[11]: n_dims = 2, name = v.blk.1.ffn_up.weight, tensor_size=6684672, offset=53489664, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[12]: n_dims = 1, name = v.blk.1.ln1.weight, tensor_size=6144, offset=60174336, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[13]: n_dims = 1, name = v.blk.1.ln2.weight, tensor_size=6144, offset=60180480, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[14]: n_dims = 2, name = v.blk.10.attn_out.weight, tensor_size=2506752, offset=60186624, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[15]: n_dims = 2, name = v.blk.10.attn_qkv.weight, tensor_size=7520256, offset=62693376, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[16]: n_dims = 2, name = v.blk.10.ffn_down.weight, tensor_size=6684672, offset=70213632, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[17]: n_dims = 2, name = v.blk.10.ffn_gate.weight, tensor_size=6684672, offset=76898304, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[18]: n_dims = 2, name = v.blk.10.ffn_up.weight, tensor_size=6684672, offset=83582976, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[19]: n_dims = 1, name = v.blk.10.ln1.weight, tensor_size=6144, offset=90267648, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[20]: n_dims = 1, name = v.blk.10.ln2.weight, tensor_size=6144, offset=90273792, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[21]: n_dims = 2, name = v.blk.11.attn_out.weight, tensor_size=2506752, offset=90279936, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[22]: n_dims = 2, name = v.blk.11.attn_qkv.weight, tensor_size=7520256, offset=92786688, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[23]: n_dims = 2, name = v.blk.11.ffn_down.weight, tensor_size=6684672, offset=100306944, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[24]: n_dims = 2, name = v.blk.11.ffn_gate.weight, tensor_size=6684672, offset=106991616, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[25]: n_dims = 2, name = v.blk.11.ffn_up.weight, tensor_size=6684672, offset=113676288, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[26]: n_dims = 1, name = v.blk.11.ln1.weight, tensor_size=6144, offset=120360960, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[27]: n_dims = 1, name = v.blk.11.ln2.weight, tensor_size=6144, offset=120367104, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[28]: n_dims = 2, name = v.blk.12.attn_out.weight, tensor_size=2506752, offset=120373248, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[29]: n_dims = 2, name = v.blk.12.attn_qkv.weight, tensor_size=7520256, offset=122880000, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[30]: n_dims = 2, name = v.blk.12.ffn_down.weight, tensor_size=6684672, offset=130400256, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[31]: n_dims = 2, name = v.blk.12.ffn_gate.weight, tensor_size=6684672, offset=137084928, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[32]: n_dims = 2, name = v.blk.12.ffn_up.weight, tensor_size=6684672, offset=143769600, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[33]: n_dims = 1, name = v.blk.12.ln1.weight, tensor_size=6144, offset=150454272, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[34]: n_dims = 1, name = v.blk.12.ln2.weight, tensor_size=6144, offset=150460416, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[35]: n_dims = 2, name = v.blk.13.attn_out.weight, tensor_size=2506752, offset=150466560, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[36]: n_dims = 2, name = v.blk.13.attn_qkv.weight, tensor_size=7520256, offset=152973312, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[37]: n_dims = 2, name = v.blk.13.ffn_down.weight, tensor_size=6684672, offset=160493568, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[38]: n_dims = 2, name = v.blk.13.ffn_gate.weight, tensor_size=6684672, offset=167178240, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[39]: n_dims = 2, name = v.blk.13.ffn_up.weight, tensor_size=6684672, offset=173862912, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[40]: n_dims = 1, name = v.blk.13.ln1.weight, tensor_size=6144, offset=180547584, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[41]: n_dims = 1, name = v.blk.13.ln2.weight, tensor_size=6144, offset=180553728, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[42]: n_dims = 2, name = v.blk.14.attn_out.weight, tensor_size=2506752, offset=180559872, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[43]: n_dims = 2, name = v.blk.14.attn_qkv.weight, tensor_size=7520256, offset=183066624, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[44]: n_dims = 2, name = v.blk.14.ffn_down.weight, tensor_size=6684672, offset=190586880, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[45]: n_dims = 2, name = v.blk.14.ffn_gate.weight, tensor_size=6684672, offset=197271552, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[46]: n_dims = 2, name = v.blk.14.ffn_up.weight, tensor_size=6684672, offset=203956224, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[47]: n_dims = 1, name = v.blk.14.ln1.weight, tensor_size=6144, offset=210640896, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[48]: n_dims = 1, name = v.blk.14.ln2.weight, tensor_size=6144, offset=210647040, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[49]: n_dims = 2, name = v.blk.15.attn_out.weight, tensor_size=2506752, offset=210653184, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[50]: n_dims = 2, name = v.blk.15.attn_qkv.weight, tensor_size=7520256, offset=213159936, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[51]: n_dims = 2, name = v.blk.15.ffn_down.weight, tensor_size=6684672, offset=220680192, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[52]: n_dims = 2, name = v.blk.15.ffn_gate.weight, tensor_size=6684672, offset=227364864, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[53]: n_dims = 2, name = v.blk.15.ffn_up.weight, tensor_size=6684672, offset=234049536, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[54]: n_dims = 1, name = v.blk.15.ln1.weight, tensor_size=6144, offset=240734208, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[55]: n_dims = 1, name = v.blk.15.ln2.weight, tensor_size=6144, offset=240740352, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[56]: n_dims = 2, name = v.blk.16.attn_out.weight, tensor_size=2506752, offset=240746496, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[57]: n_dims = 2, name = v.blk.16.attn_qkv.weight, tensor_size=7520256, offset=243253248, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[58]: n_dims = 2, name = v.blk.16.ffn_down.weight, tensor_size=6684672, offset=250773504, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[59]: n_dims = 2, name = v.blk.16.ffn_gate.weight, tensor_size=6684672, offset=257458176, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[60]: n_dims = 2, name = v.blk.16.ffn_up.weight, tensor_size=6684672, offset=264142848, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[61]: n_dims = 1, name = v.blk.16.ln1.weight, tensor_size=6144, offset=270827520, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[62]: n_dims = 1, name = v.blk.16.ln2.weight, tensor_size=6144, offset=270833664, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[63]: n_dims = 2, name = v.blk.17.attn_out.weight, tensor_size=2506752, offset=270839808, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[64]: n_dims = 2, name = v.blk.17.attn_qkv.weight, tensor_size=7520256, offset=273346560, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[65]: n_dims = 2, name = v.blk.17.ffn_down.weight, tensor_size=6684672, offset=280866816, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[66]: n_dims = 2, name = v.blk.17.ffn_gate.weight, tensor_size=6684672, offset=287551488, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[67]: n_dims = 2, name = v.blk.17.ffn_up.weight, tensor_size=6684672, offset=294236160, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[68]: n_dims = 1, name = v.blk.17.ln1.weight, tensor_size=6144, offset=300920832, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[69]: n_dims = 1, name = v.blk.17.ln2.weight, tensor_size=6144, offset=300926976, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[70]: n_dims = 2, name = v.blk.18.attn_out.weight, tensor_size=2506752, offset=300933120, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[71]: n_dims = 2, name = v.blk.18.attn_qkv.weight, tensor_size=7520256, offset=303439872, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[72]: n_dims = 2, name = v.blk.18.ffn_down.weight, tensor_size=6684672, offset=310960128, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[73]: n_dims = 2, name = v.blk.18.ffn_gate.weight, tensor_size=6684672, offset=317644800, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[74]: n_dims = 2, name = v.blk.18.ffn_up.weight, tensor_size=6684672, offset=324329472, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[75]: n_dims = 1, name = v.blk.18.ln1.weight, tensor_size=6144, offset=331014144, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[76]: n_dims = 1, name = v.blk.18.ln2.weight, tensor_size=6144, offset=331020288, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[77]: n_dims = 2, name = v.blk.19.attn_out.weight, tensor_size=2506752, offset=331026432, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[78]: n_dims = 2, name = v.blk.19.attn_qkv.weight, tensor_size=7520256, offset=333533184, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[79]: n_dims = 2, name = v.blk.19.ffn_down.weight, tensor_size=6684672, offset=341053440, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[80]: n_dims = 2, name = v.blk.19.ffn_gate.weight, tensor_size=6684672, offset=347738112, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[81]: n_dims = 2, name = v.blk.19.ffn_up.weight, tensor_size=6684672, offset=354422784, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[82]: n_dims = 1, name = v.blk.19.ln1.weight, tensor_size=6144, offset=361107456, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[83]: n_dims = 1, name = v.blk.19.ln2.weight, tensor_size=6144, offset=361113600, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[84]: n_dims = 2, name = v.blk.2.attn_out.weight, tensor_size=2506752, offset=361119744, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[85]: n_dims = 2, name = v.blk.2.attn_qkv.weight, tensor_size=7520256, offset=363626496, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[86]: n_dims = 2, name = v.blk.2.ffn_down.weight, tensor_size=6684672, offset=371146752, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[87]: n_dims = 2, name = v.blk.2.ffn_gate.weight, tensor_size=6684672, offset=377831424, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[88]: n_dims = 2, name = v.blk.2.ffn_up.weight, tensor_size=6684672, offset=384516096, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[89]: n_dims = 1, name = v.blk.2.ln1.weight, tensor_size=6144, offset=391200768, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[90]: n_dims = 1, name = v.blk.2.ln2.weight, tensor_size=6144, offset=391206912, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[91]: n_dims = 2, name = v.blk.20.attn_out.weight, tensor_size=2506752, offset=391213056, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[92]: n_dims = 2, name = v.blk.20.attn_qkv.weight, tensor_size=7520256, offset=393719808, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[93]: n_dims = 2, name = v.blk.20.ffn_down.weight, tensor_size=6684672, offset=401240064, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[94]: n_dims = 2, name = v.blk.20.ffn_gate.weight, tensor_size=6684672, offset=407924736, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[95]: n_dims = 2, name = v.blk.20.ffn_up.weight, tensor_size=6684672, offset=414609408, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[96]: n_dims = 1, name = v.blk.20.ln1.weight, tensor_size=6144, offset=421294080, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[97]: n_dims = 1, name = v.blk.20.ln2.weight, tensor_size=6144, offset=421300224, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[98]: n_dims = 2, name = v.blk.21.attn_out.weight, tensor_size=2506752, offset=421306368, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[99]: n_dims = 2, name = v.blk.21.attn_qkv.weight, tensor_size=7520256, offset=423813120, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[100]: n_dims = 2, name = v.blk.21.ffn_down.weight, tensor_size=6684672, offset=431333376, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[101]: n_dims = 2, name = v.blk.21.ffn_gate.weight, tensor_size=6684672, offset=438018048, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[102]: n_dims = 2, name = v.blk.21.ffn_up.weight, tensor_size=6684672, offset=444702720, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[103]: n_dims = 1, name = v.blk.21.ln1.weight, tensor_size=6144, offset=451387392, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[104]: n_dims = 1, name = v.blk.21.ln2.weight, tensor_size=6144, offset=451393536, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[105]: n_dims = 2, name = v.blk.22.attn_out.weight, tensor_size=2506752, offset=451399680, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[106]: n_dims = 2, name = v.blk.22.attn_qkv.weight, tensor_size=7520256, offset=453906432, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[107]: n_dims = 2, name = v.blk.22.ffn_down.weight, tensor_size=6684672, offset=461426688, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[108]: n_dims = 2, name = v.blk.22.ffn_gate.weight, tensor_size=6684672, offset=468111360, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[109]: n_dims = 2, name = v.blk.22.ffn_up.weight, tensor_size=6684672, offset=474796032, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[110]: n_dims = 1, name = v.blk.22.ln1.weight, tensor_size=6144, offset=481480704, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[111]: n_dims = 1, name = v.blk.22.ln2.weight, tensor_size=6144, offset=481486848, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[112]: n_dims = 2, name = v.blk.23.attn_out.weight, tensor_size=2506752, offset=481492992, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[113]: n_dims = 2, name = v.blk.23.attn_qkv.weight, tensor_size=7520256, offset=483999744, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[114]: n_dims = 2, name = v.blk.23.ffn_down.weight, tensor_size=6684672, offset=491520000, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[115]: n_dims = 2, name = v.blk.23.ffn_gate.weight, tensor_size=6684672, offset=498204672, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[116]: n_dims = 2, name = v.blk.23.ffn_up.weight, tensor_size=6684672, offset=504889344, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[117]: n_dims = 1, name = v.blk.23.ln1.weight, tensor_size=6144, offset=511574016, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[118]: n_dims = 1, name = v.blk.23.ln2.weight, tensor_size=6144, offset=511580160, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[119]: n_dims = 2, name = v.blk.3.attn_out.weight, tensor_size=2506752, offset=511586304, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[120]: n_dims = 2, name = v.blk.3.attn_qkv.weight, tensor_size=7520256, offset=514093056, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[121]: n_dims = 2, name = v.blk.3.ffn_down.weight, tensor_size=6684672, offset=521613312, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[122]: n_dims = 2, name = v.blk.3.ffn_gate.weight, tensor_size=6684672, offset=528297984, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[123]: n_dims = 2, name = v.blk.3.ffn_up.weight, tensor_size=6684672, offset=534982656, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[124]: n_dims = 1, name = v.blk.3.ln1.weight, tensor_size=6144, offset=541667328, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[125]: n_dims = 1, name = v.blk.3.ln2.weight, tensor_size=6144, offset=541673472, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[126]: n_dims = 2, name = v.blk.4.attn_out.weight, tensor_size=2506752, offset=541679616, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[127]: n_dims = 2, name = v.blk.4.attn_qkv.weight, tensor_size=7520256, offset=544186368, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[128]: n_dims = 2, name = v.blk.4.ffn_down.weight, tensor_size=6684672, offset=551706624, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[129]: n_dims = 2, name = v.blk.4.ffn_gate.weight, tensor_size=6684672, offset=558391296, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[130]: n_dims = 2, name = v.blk.4.ffn_up.weight, tensor_size=6684672, offset=565075968, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[131]: n_dims = 1, name = v.blk.4.ln1.weight, tensor_size=6144, offset=571760640, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[132]: n_dims = 1, name = v.blk.4.ln2.weight, tensor_size=6144, offset=571766784, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[133]: n_dims = 2, name = v.blk.5.attn_out.weight, tensor_size=2506752, offset=571772928, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[134]: n_dims = 2, name = v.blk.5.attn_qkv.weight, tensor_size=7520256, offset=574279680, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[135]: n_dims = 2, name = v.blk.5.ffn_down.weight, tensor_size=6684672, offset=581799936, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[136]: n_dims = 2, name = v.blk.5.ffn_gate.weight, tensor_size=6684672, offset=588484608, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[137]: n_dims = 2, name = v.blk.5.ffn_up.weight, tensor_size=6684672, offset=595169280, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[138]: n_dims = 1, name = v.blk.5.ln1.weight, tensor_size=6144, offset=601853952, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[139]: n_dims = 1, name = v.blk.5.ln2.weight, tensor_size=6144, offset=601860096, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[140]: n_dims = 2, name = v.blk.6.attn_out.weight, tensor_size=2506752, offset=601866240, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[141]: n_dims = 2, name = v.blk.6.attn_qkv.weight, tensor_size=7520256, offset=604372992, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[142]: n_dims = 2, name = v.blk.6.ffn_down.weight, tensor_size=6684672, offset=611893248, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[143]: n_dims = 2, name = v.blk.6.ffn_gate.weight, tensor_size=6684672, offset=618577920, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[144]: n_dims = 2, name = v.blk.6.ffn_up.weight, tensor_size=6684672, offset=625262592, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[145]: n_dims = 1, name = v.blk.6.ln1.weight, tensor_size=6144, offset=631947264, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[146]: n_dims = 1, name = v.blk.6.ln2.weight, tensor_size=6144, offset=631953408, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[147]: n_dims = 2, name = v.blk.7.attn_out.weight, tensor_size=2506752, offset=631959552, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[148]: n_dims = 2, name = v.blk.7.attn_qkv.weight, tensor_size=7520256, offset=634466304, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[149]: n_dims = 2, name = v.blk.7.ffn_down.weight, tensor_size=6684672, offset=641986560, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[150]: n_dims = 2, name = v.blk.7.ffn_gate.weight, tensor_size=6684672, offset=648671232, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[151]: n_dims = 2, name = v.blk.7.ffn_up.weight, tensor_size=6684672, offset=655355904, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[152]: n_dims = 1, name = v.blk.7.ln1.weight, tensor_size=6144, offset=662040576, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[153]: n_dims = 1, name = v.blk.7.ln2.weight, tensor_size=6144, offset=662046720, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[154]: n_dims = 2, name = v.blk.8.attn_out.weight, tensor_size=2506752, offset=662052864, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[155]: n_dims = 2, name = v.blk.8.attn_qkv.weight, tensor_size=7520256, offset=664559616, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[156]: n_dims = 2, name = v.blk.8.ffn_down.weight, tensor_size=6684672, offset=672079872, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[157]: n_dims = 2, name = v.blk.8.ffn_gate.weight, tensor_size=6684672, offset=678764544, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[158]: n_dims = 2, name = v.blk.8.ffn_up.weight, tensor_size=6684672, offset=685449216, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[159]: n_dims = 1, name = v.blk.8.ln1.weight, tensor_size=6144, offset=692133888, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[160]: n_dims = 1, name = v.blk.8.ln2.weight, tensor_size=6144, offset=692140032, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[161]: n_dims = 2, name = v.blk.9.attn_out.weight, tensor_size=2506752, offset=692146176, shape:[1536, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[162]: n_dims = 2, name = v.blk.9.attn_qkv.weight, tensor_size=7520256, offset=694652928, shape:[1536, 4608, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[163]: n_dims = 2, name = v.blk.9.ffn_down.weight, tensor_size=6684672, offset=702173184, shape:[4096, 1536, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[164]: n_dims = 2, name = v.blk.9.ffn_gate.weight, tensor_size=6684672, offset=708857856, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[165]: n_dims = 2, name = v.blk.9.ffn_up.weight, tensor_size=6684672, offset=715542528, shape:[1536, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[166]: n_dims = 1, name = v.blk.9.ln1.weight, tensor_size=6144, offset=722227200, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[167]: n_dims = 1, name = v.blk.9.ln2.weight, tensor_size=6144, offset=722233344, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[168]: n_dims = 1, name = mm.patch_merger.bias, tensor_size=16384, offset=722239488, shape:[4096, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[169]: n_dims = 4, name = mm.patch_merger.weight, tensor_size=50331648, offset=722255872, shape:[2, 2, 1536, 4096], type = f16 19:10:53 arch ollama[927]: clip_model_loader: tensor[170]: n_dims = 2, name = v.position_embd.weight, tensor_size=3538944, offset=772587520, shape:[1536, 576, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[171]: n_dims = 2, name = mm.down.weight, tensor_size=59604992, offset=776126464, shape:[13696, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[172]: n_dims = 2, name = mm.gate.weight, tensor_size=59604992, offset=835731456, shape:[4096, 13696, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[173]: n_dims = 1, name = mm.post_norm.bias, tensor_size=16384, offset=895336448, shape:[4096, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[174]: n_dims = 1, name = mm.post_norm.weight, tensor_size=16384, offset=895352832, shape:[4096, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[175]: n_dims = 2, name = mm.model.fc.weight, tensor_size=17825792, offset=895369216, shape:[4096, 4096, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[176]: n_dims = 2, name = mm.up.weight, tensor_size=59604992, offset=913195008, shape:[4096, 13696, 1, 1], type = q8_0 19:10:53 arch ollama[927]: clip_model_loader: tensor[177]: n_dims = 1, name = v.patch_embd.bias, tensor_size=6144, offset=972800000, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[178]: n_dims = 4, name = v.patch_embd.weight, tensor_size=3612672, offset=972806144, shape:[14, 14, 3, 1536], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[179]: n_dims = 4, name = v.patch_embd.weight.1, tensor_size=3612672, offset=976418816, shape:[14, 14, 3, 1536], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[180]: n_dims = 1, name = v.norm_embd.weight, tensor_size=6144, offset=980031488, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_model_loader: tensor[181]: n_dims = 1, name = v.post_ln.weight, tensor_size=6144, offset=980037632, shape:[1536, 1, 1, 1], type = f32 19:10:53 arch ollama[927]: clip_ctx: CLIP using CUDA0 backend 19:10:53 arch ollama[927]: load_hparams: projector: glm4v 19:10:53 arch ollama[927]: load_hparams: n_embd: 1536 19:10:53 arch ollama[927]: load_hparams: n_head: 12 19:10:53 arch ollama[927]: load_hparams: n_ff: 13696 19:10:53 arch ollama[927]: load_hparams: n_layer: 24 19:10:53 arch ollama[927]: load_hparams: ffn_op: silu 19:10:53 arch ollama[927]: load_hparams: projection_dim: 4096 19:10:53 arch ollama[927]: --- vision hparams --- 19:10:53 arch ollama[927]: load_hparams: image_size: 336 19:10:53 arch ollama[927]: load_hparams: patch_size: 14 19:10:53 arch ollama[927]: load_hparams: has_llava_proj: 0 19:10:53 arch ollama[927]: load_hparams: minicpmv_version: 0 19:10:53 arch ollama[927]: load_hparams: n_merge: 2 19:10:53 arch ollama[927]: load_hparams: n_wa_pattern: 0 19:10:53 arch ollama[927]: load_hparams: image_min_pixels: 6272 19:10:53 arch ollama[927]: load_hparams: image_max_pixels: 3211264 19:10:53 arch ollama[927]: load_hparams: model size: 934.64 MiB 19:10:53 arch ollama[927]: load_hparams: metadata size: 0.06 MiB 19:10:53 arch ollama[927]: load_tensors: loaded 182 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-454dd441c925c1a81c984204bb9d54feef0ef07b789c6fe1118099014ba2727d 19:10:53 arch ollama[927]: warmup: warmup with image size = 1288 x 1288 19:10:53 arch ollama[927]: ggml_backend_cuda_buffer_type_alloc_buffer: allocating 515.05 MiB on device 0: cudaMalloc failed: out of memory 19:10:53 arch ollama[927]: ggml_gallocr_reserve_n_impl: failed to allocate CUDA0 buffer of size 540070912 19:10:53 arch ollama[927]: alloc_compute_meta: CPU compute buffer size = 19.11 MiB 19:10:53 arch ollama[927]: alloc_compute_meta: graph splits = 1, nodes = 632 19:10:53 arch ollama[927]: warmup: flash attention is enabled 19:10:53 arch ollama[927]: time=2026-01-05T19:10:53.861+08:00 level=INFO source=server.go:1376 msg="llama runner started in 2.61 seconds" 19:10:53 arch ollama[927]: time=2026-01-05T19:10:53.861+08:00 level=INFO source=sched.go:517 msg="loaded runners" count=1 19:10:53 arch ollama[927]: time=2026-01-05T19:10:53.861+08:00 level=INFO source=server.go:1338 msg="waiting for llama runner to start responding" 19:10:53 arch ollama[927]: time=2026-01-05T19:10:53.861+08:00 level=INFO source=server.go:1376 msg="llama runner started in 2.61 seconds" 19:10:54 arch ollama[927]: add_text: <|begin_of_image|> 19:10:54 arch ollama[927]: image_tokens->nx = 85 19:10:54 arch ollama[927]: image_tokens->ny = 48 19:10:54 arch ollama[927]: batch_f32 size = 1 19:10:54 arch ollama[927]: add_text: <|end_of_image|> 19:10:54 arch ollama[927]: ggml_backend_cuda_buffer_type_alloc_buffer: allocating 993.11 MiB on device 0: cudaMalloc failed: out of memory 19:10:54 arch ollama[927]: ggml_gallocr_reserve_n_impl: failed to allocate CUDA0 buffer of size 1041346560 19:10:54 arch ollama[927]: SIGSEGV: segmentation violation 19:10:54 arch ollama[927]: PC=0x5625c497b1cb m=5 sigcode=1 addr=0x0 19:10:54 arch ollama[927]: signal arrived during cgo execution 19:10:54 arch ollama[927]: goroutine 10 gp=0xc00058b180 m=5 mp=0xc000100008 [syscall]: 19:10:54 arch ollama[927]: runtime.cgocall(0x5625c4968190, 0xc0000491d8) 19:10:54 arch ollama[927]: runtime/cgocall.go:167 +0x4b fp=0xc0000491b0 sp=0xc000049178 pc=0x5625c3c166eb 19:10:54 arch ollama[927]: github.com/ollama/ollama/llama._Cfunc_mtmd_encode_chunk(0x7fd89008df10, 0x7fd7cd7bc180) 19:10:54 arch ollama[927]: _cgo_gotypes.go:1079 +0x4a fp=0xc0000491d8 sp=0xc0000491b0 pc=0x5625c3fd104a 19:10:54 arch ollama[927]: github.com/ollama/ollama/llama.(*MtmdContext).MultimodalTokenize.func11(...) 19:10:54 arch ollama[927]: github.com/ollama/ollama/llama/llama.go:595 19:10:54 arch ollama[927]: github.com/ollama/ollama/llama.(*MtmdContext).MultimodalTokenize(0xc0004b2028, 0xc000346010, {0xc000a86000, 0x117bcb, 0xc000049520?}) 19:10:54 arch ollama[927]: github.com/ollama/ollama/llama/llama.go:595 +0x6c5 fp=0xc000049490 sp=0xc0000491d8 pc=0x5625c3fd6085 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*ImageContext).MultimodalTokenize(0xc000051d40, 0xc000346010, {0xc000a86000, 0x117bcb, 0x117bcd}) 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/image.go:76 +0x145 fp=0xc000049530 sp=0xc000049490 pc=0x5625c4088365 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).inputs(0xc000254780, {0xc000160400?, 0x3d?}, {0xc0000fa080, 0x1, 0x160?}) 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:236 +0x2c6 fp=0xc000049698 sp=0xc000049530 pc=0x5625c4089766 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).NewSequence(0xc000254780, {0xc000160400, 0x3d}, {0xc0000fa080, 0x1, 0x1}, {0x28000, {0xc0003d6060, 0x1, 0x1}, ...}) 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:126 +0x8d fp=0xc000049838 sp=0xc000049698 pc=0x5625c4088d2d 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).completion(0xc000254780, {0x5625c51ccfa0, 0xc0000e3b20}, 0xc000177040) 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:659 +0x5f9 fp=0xc000049ac0 sp=0xc000049838 pc=0x5625c408bd99 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).completion-fm({0x5625c51ccfa0?, 0xc0000e3b20?}, 0xc000049b40?) 19:10:54 arch ollama[927]: <autogenerated>:1 +0x36 fp=0xc000049af0 sp=0xc000049ac0 pc=0x5625c408f7f6 19:10:54 arch ollama[927]: net/http.HandlerFunc.ServeHTTP(0xc000051c80?, {0x5625c51ccfa0?, 0xc0000e3b20?}, 0xc000049b60?) 19:10:54 arch ollama[927]: net/http/server.go:2294 +0x29 fp=0xc000049b18 sp=0xc000049af0 pc=0x5625c3f190e9 19:10:54 arch ollama[927]: net/http.(*ServeMux).ServeHTTP(0x5625c3bbe8c5?, {0x5625c51ccfa0, 0xc0000e3b20}, 0xc000177040) 19:10:54 arch ollama[927]: net/http/server.go:2822 +0x1c4 fp=0xc000049b68 sp=0xc000049b18 pc=0x5625c3f1afe4 19:10:54 arch ollama[927]: net/http.serverHandler.ServeHTTP({0x5625c51c9590?}, {0x5625c51ccfa0?, 0xc0000e3b20?}, 0x1?) 19:10:54 arch ollama[927]: net/http/server.go:3301 +0x8e fp=0xc000049b98 sp=0xc000049b68 pc=0x5625c3f38a6e 19:10:54 arch ollama[927]: net/http.(*conn).serve(0xc00023e3f0, {0x5625c51cf3d8, 0xc00023cfc0}) 19:10:54 arch ollama[927]: net/http/server.go:2102 +0x625 fp=0xc000049fb8 sp=0xc000049b98 pc=0x5625c3f175e5 19:10:54 arch ollama[927]: net/http.(*Server).Serve.gowrap3() 19:10:54 arch ollama[927]: net/http/server.go:3454 +0x28 fp=0xc000049fe0 sp=0xc000049fb8 pc=0x5625c3f1cea8 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000049fe8 sp=0xc000049fe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by net/http.(*Server).Serve in goroutine 1 19:10:54 arch ollama[927]: net/http/server.go:3454 +0x485 19:10:54 arch ollama[927]: goroutine 1 gp=0xc000002380 m=nil [IO wait]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc0003fd790 sp=0xc0003fd770 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.netpollblock(0xc0005877e0?, 0xc3bb32a6?, 0x25?) 19:10:54 arch ollama[927]: runtime/netpoll.go:575 +0xf7 fp=0xc0003fd7c8 sp=0xc0003fd790 pc=0x5625c3bdee97 19:10:54 arch ollama[927]: internal/poll.runtime_pollWait(0x7fd89dec5de0, 0x72) 19:10:54 arch ollama[927]: runtime/netpoll.go:351 +0x85 fp=0xc0003fd7e8 sp=0xc0003fd7c8 pc=0x5625c3c18d85 19:10:54 arch ollama[927]: internal/poll.(*pollDesc).wait(0xc000170980?, 0x900000036?, 0x0) 19:10:54 arch ollama[927]: internal/poll/fd_poll_runtime.go:84 +0x27 fp=0xc0003fd810 sp=0xc0003fd7e8 pc=0x5625c3ca0f07 19:10:54 arch ollama[927]: internal/poll.(*pollDesc).waitRead(...) 19:10:54 arch ollama[927]: internal/poll/fd_poll_runtime.go:89 19:10:54 arch ollama[927]: internal/poll.(*FD).Accept(0xc000170980) 19:10:54 arch ollama[927]: internal/poll/fd_unix.go:620 +0x295 fp=0xc0003fd8b8 sp=0xc0003fd810 pc=0x5625c3ca62d5 19:10:54 arch ollama[927]: net.(*netFD).accept(0xc000170980) 19:10:54 arch ollama[927]: net/fd_unix.go:172 +0x29 fp=0xc0003fd970 sp=0xc0003fd8b8 pc=0x5625c3d191a9 19:10:54 arch ollama[927]: net.(*TCPListener).accept(0xc000305340) 19:10:54 arch ollama[927]: net/tcpsock_posix.go:159 +0x1b fp=0xc0003fd9c0 sp=0xc0003fd970 pc=0x5625c3d2eb5b 19:10:54 arch ollama[927]: net.(*TCPListener).Accept(0xc000305340) 19:10:54 arch ollama[927]: net/tcpsock.go:380 +0x30 fp=0xc0003fd9f0 sp=0xc0003fd9c0 pc=0x5625c3d2da10 19:10:54 arch ollama[927]: net/http.(*onceCloseListener).Accept(0xc00023e3f0?) 19:10:54 arch ollama[927]: <autogenerated>:1 +0x24 fp=0xc0003fda08 sp=0xc0003fd9f0 pc=0x5625c3f451e4 19:10:54 arch ollama[927]: net/http.(*Server).Serve(0xc0001fb700, {0x5625c51ccdc0, 0xc000305340}) 19:10:54 arch ollama[927]: net/http/server.go:3424 +0x30c fp=0xc0003fdb38 sp=0xc0003fda08 pc=0x5625c3f1caac 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.Execute({0xc000130140, 0x4, 0x4}) 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:1002 +0x8f5 fp=0xc0003fdd08 sp=0xc0003fdb38 pc=0x5625c408f175 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner.Execute({0xc000130130?, 0x0?, 0x0?}) 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/runner.go:22 +0xd4 fp=0xc0003fdd30 sp=0xc0003fdd08 pc=0x5625c413acf4 19:10:54 arch ollama[927]: github.com/ollama/ollama/cmd.NewCLI.func2(0xc0001fb400?, {0x5625c4caf0ad?, 0x4?, 0x5625c4caf0b1?}) 19:10:54 arch ollama[927]: github.com/ollama/ollama/cmd/cmd.go:1841 +0x45 fp=0xc0003fdd58 sp=0xc0003fdd30 pc=0x5625c48f7f25 19:10:54 arch ollama[927]: github.com/spf13/cobra.(*Command).execute(0xc00027d508, {0xc000305140, 0x4, 0x4}) 19:10:54 arch ollama[927]: github.com/spf13/cobra@v1.7.0/command.go:940 +0x85c fp=0xc0003fde78 sp=0xc0003fdd58 pc=0x5625c3d927fc 19:10:54 arch ollama[927]: github.com/spf13/cobra.(*Command).ExecuteC(0xc0000f4908) 19:10:54 arch ollama[927]: github.com/spf13/cobra@v1.7.0/command.go:1068 +0x3a5 fp=0xc0003fdf30 sp=0xc0003fde78 pc=0x5625c3d93045 19:10:54 arch ollama[927]: github.com/spf13/cobra.(*Command).Execute(...) 19:10:54 arch ollama[927]: github.com/spf13/cobra@v1.7.0/command.go:992 19:10:54 arch ollama[927]: github.com/spf13/cobra.(*Command).ExecuteContext(...) 19:10:54 arch ollama[927]: github.com/spf13/cobra@v1.7.0/command.go:985 19:10:54 arch ollama[927]: main.main() 19:10:54 arch ollama[927]: github.com/ollama/ollama/main.go:12 +0x4d fp=0xc0003fdf50 sp=0xc0003fdf30 pc=0x5625c48f8a0d 19:10:54 arch ollama[927]: runtime.main() 19:10:54 arch ollama[927]: runtime/proc.go:283 +0x29d fp=0xc0003fdfe0 sp=0xc0003fdf50 pc=0x5625c3be651d 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0003fdfe8 sp=0xc0003fdfe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: goroutine 2 gp=0xc000002e00 m=nil [force gc (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008efa8 sp=0xc00008ef88 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.goparkunlock(...) 19:10:54 arch ollama[927]: runtime/proc.go:441 19:10:54 arch ollama[927]: runtime.forcegchelper() 19:10:54 arch ollama[927]: runtime/proc.go:348 +0xb8 fp=0xc00008efe0 sp=0xc00008efa8 pc=0x5625c3be6858 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008efe8 sp=0xc00008efe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.init.7 in goroutine 1 19:10:54 arch ollama[927]: runtime/proc.go:336 +0x1a 19:10:54 arch ollama[927]: goroutine 3 gp=0xc000003340 m=nil [GC sweep wait]: 19:10:54 arch ollama[927]: runtime.gopark(0x1?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008f780 sp=0xc00008f760 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.goparkunlock(...) 19:10:54 arch ollama[927]: runtime/proc.go:441 19:10:54 arch ollama[927]: runtime.bgsweep(0xc0000ba000) 19:10:54 arch ollama[927]: runtime/mgcsweep.go:316 +0xdf fp=0xc00008f7c8 sp=0xc00008f780 pc=0x5625c3bd0fff 19:10:54 arch ollama[927]: runtime.gcenable.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:204 +0x25 fp=0xc00008f7e0 sp=0xc00008f7c8 pc=0x5625c3bc53e5 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008f7e8 sp=0xc00008f7e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcenable in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:204 +0x66 19:10:54 arch ollama[927]: goroutine 4 gp=0xc000003500 m=nil [GC scavenge wait]: 19:10:54 arch ollama[927]: runtime.gopark(0x10000?, 0x5625c4e802a0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008ff78 sp=0xc00008ff58 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.goparkunlock(...) 19:10:54 arch ollama[927]: runtime/proc.go:441 19:10:54 arch ollama[927]: runtime.(*scavengerState).park(0x5625c5aa3280) 19:10:54 arch ollama[927]: runtime/mgcscavenge.go:425 +0x49 fp=0xc00008ffa8 sp=0xc00008ff78 pc=0x5625c3bcea49 19:10:54 arch ollama[927]: runtime.bgscavenge(0xc0000ba000) 19:10:54 arch ollama[927]: runtime/mgcscavenge.go:658 +0x59 fp=0xc00008ffc8 sp=0xc00008ffa8 pc=0x5625c3bcefd9 19:10:54 arch ollama[927]: runtime.gcenable.gowrap2() 19:10:54 arch ollama[927]: runtime/mgc.go:205 +0x25 fp=0xc00008ffe0 sp=0xc00008ffc8 pc=0x5625c3bc5385 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008ffe8 sp=0xc00008ffe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcenable in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:205 +0xa5 19:10:54 arch ollama[927]: goroutine 18 gp=0xc000102700 m=nil [finalizer wait]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x5625c51b9250?, 0x0?, 0x80?, 0x1000000010?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008e630 sp=0xc00008e610 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.runfinq() 19:10:54 arch ollama[927]: runtime/mfinal.go:196 +0x107 fp=0xc00008e7e0 sp=0xc00008e630 pc=0x5625c3bc43a7 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008e7e8 sp=0xc00008e7e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.createfing in goroutine 1 19:10:54 arch ollama[927]: runtime/mfinal.go:166 +0x3d 19:10:54 arch ollama[927]: goroutine 19 gp=0xc000103180 m=nil [chan receive]: 19:10:54 arch ollama[927]: runtime.gopark(0xc00022d720?, 0xc0004a0030?, 0x60?, 0xa7?, 0x5625c3cffde8?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008a718 sp=0xc00008a6f8 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.chanrecv(0xc000110310, 0x0, 0x1) 19:10:54 arch ollama[927]: runtime/chan.go:664 +0x445 fp=0xc00008a790 sp=0xc00008a718 pc=0x5625c3bb5e85 19:10:54 arch ollama[927]: runtime.chanrecv1(0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/chan.go:506 +0x12 fp=0xc00008a7b8 sp=0xc00008a790 pc=0x5625c3bb5a12 19:10:54 arch ollama[927]: runtime.unique_runtime_registerUniqueMapCleanup.func2(...) 19:10:54 arch ollama[927]: runtime/mgc.go:1796 19:10:54 arch ollama[927]: runtime.unique_runtime_registerUniqueMapCleanup.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1799 +0x2f fp=0xc00008a7e0 sp=0xc00008a7b8 pc=0x5625c3bc858f 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008a7e8 sp=0xc00008a7e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by unique.runtime_registerUniqueMapCleanup in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1794 +0x85 19:10:54 arch ollama[927]: goroutine 20 gp=0xc000103500 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008af38 sp=0xc00008af18 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008afc8 sp=0xc00008af38 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008afe0 sp=0xc00008afc8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008afe8 sp=0xc00008afe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 34 gp=0xc0003b6000 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000484738 sp=0xc000484718 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0004847c8 sp=0xc000484738 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0004847e0 sp=0xc0004847c8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0004847e8 sp=0xc0004847e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 35 gp=0xc0003b61c0 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000484f38 sp=0xc000484f18 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000484fc8 sp=0xc000484f38 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000484fe0 sp=0xc000484fc8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000484fe8 sp=0xc000484fe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 36 gp=0xc0003b6380 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000485738 sp=0xc000485718 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0004857c8 sp=0xc000485738 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0004857e0 sp=0xc0004857c8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0004857e8 sp=0xc0004857e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 37 gp=0xc0003b6540 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000485f38 sp=0xc000485f18 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000485fc8 sp=0xc000485f38 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000485fe0 sp=0xc000485fc8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000485fe8 sp=0xc000485fe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 38 gp=0xc0003b6700 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000486738 sp=0xc000486718 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0004867c8 sp=0xc000486738 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0004867e0 sp=0xc0004867c8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0004867e8 sp=0xc0004867e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 39 gp=0xc0003b68c0 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000486f38 sp=0xc000486f18 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000486fc8 sp=0xc000486f38 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000486fe0 sp=0xc000486fc8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000486fe8 sp=0xc000486fe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 21 gp=0xc0001036c0 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008b738 sp=0xc00008b718 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008b7c8 sp=0xc00008b738 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008b7e0 sp=0xc00008b7c8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008b7e8 sp=0xc00008b7e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 5 gp=0xc000003a40 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000090738 sp=0xc000090718 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0000907c8 sp=0xc000090738 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0000907e0 sp=0xc0000907c8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0000907e8 sp=0xc0000907e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 40 gp=0xc0003b6a80 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000487738 sp=0xc000487718 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0004877c8 sp=0xc000487738 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0004877e0 sp=0xc0004877c8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0004877e8 sp=0xc0004877e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 6 gp=0xc000003c00 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000090f38 sp=0xc000090f18 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000090fc8 sp=0xc000090f38 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000090fe0 sp=0xc000090fc8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000090fe8 sp=0xc000090fe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 22 gp=0xc000103880 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008bf38 sp=0xc00008bf18 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008bfc8 sp=0xc00008bf38 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008bfe0 sp=0xc00008bfc8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008bfe8 sp=0xc00008bfe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 41 gp=0xc0003b6c40 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000487f38 sp=0xc000487f18 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000487fc8 sp=0xc000487f38 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000487fe0 sp=0xc000487fc8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000487fe8 sp=0xc000487fe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 42 gp=0xc0003b6e00 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000480738 sp=0xc000480718 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0004807c8 sp=0xc000480738 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0004807e0 sp=0xc0004807c8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0004807e8 sp=0xc0004807e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 43 gp=0xc0003b6fc0 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000480f38 sp=0xc000480f18 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000480fc8 sp=0xc000480f38 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000480fe0 sp=0xc000480fc8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000480fe8 sp=0xc000480fe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 44 gp=0xc0003b7180 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000481738 sp=0xc000481718 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0004817c8 sp=0xc000481738 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0004817e0 sp=0xc0004817c8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0004817e8 sp=0xc0004817e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 45 gp=0xc0003b7340 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x185132099240a?, 0x0?, 0x0?, 0x0?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000481f38 sp=0xc000481f18 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000481fc8 sp=0xc000481f38 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000481fe0 sp=0xc000481fc8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000481fe8 sp=0xc000481fe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 7 gp=0xc000003dc0 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x5625c5b71680?, 0x1?, 0xfb?, 0x4c?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000091738 sp=0xc000091718 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0000917c8 sp=0xc000091738 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0000917e0 sp=0xc0000917c8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0000917e8 sp=0xc0000917e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 23 gp=0xc000103a40 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x1851320999d4c?, 0x1?, 0xd9?, 0xbd?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008c738 sp=0xc00008c718 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008c7c8 sp=0xc00008c738 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008c7e0 sp=0xc00008c7c8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008c7e8 sp=0xc00008c7e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 24 gp=0xc000103c00 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x5625c5b71680?, 0x1?, 0xcb?, 0xaa?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008cf38 sp=0xc00008cf18 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008cfc8 sp=0xc00008cf38 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008cfe0 sp=0xc00008cfc8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008cfe8 sp=0xc00008cfe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 46 gp=0xc0003b7500 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x5625c5b71680?, 0x1?, 0x7a?, 0x93?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000482738 sp=0xc000482718 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc0004827c8 sp=0xc000482738 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc0004827e0 sp=0xc0004827c8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0004827e8 sp=0xc0004827e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 25 gp=0xc000103dc0 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x5625c5b71680?, 0x1?, 0xbd?, 0x7f?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc00008d738 sp=0xc00008d718 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc00008d7c8 sp=0xc00008d738 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc00008d7e0 sp=0xc00008d7c8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc00008d7e8 sp=0xc00008d7e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 8 gp=0xc0000c6000 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x185132099abe5?, 0x3?, 0x9?, 0x81?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000091f38 sp=0xc000091f18 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000091fc8 sp=0xc000091f38 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000091fe0 sp=0xc000091fc8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000091fe8 sp=0xc000091fe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 47 gp=0xc0003b76c0 m=nil [GC worker (idle)]: 19:10:54 arch ollama[927]: runtime.gopark(0x18513209989f8?, 0x1?, 0xc?, 0x18?, 0x0?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc000482f38 sp=0xc000482f18 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.gcBgMarkWorker(0xc000111730) 19:10:54 arch ollama[927]: runtime/mgc.go:1423 +0xe9 fp=0xc000482fc8 sp=0xc000482f38 pc=0x5625c3bc78a9 19:10:54 arch ollama[927]: runtime.gcBgMarkStartWorkers.gowrap1() 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x25 fp=0xc000482fe0 sp=0xc000482fc8 pc=0x5625c3bc7785 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc000482fe8 sp=0xc000482fe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by runtime.gcBgMarkStartWorkers in goroutine 1 19:10:54 arch ollama[927]: runtime/mgc.go:1339 +0x105 19:10:54 arch ollama[927]: goroutine 9 gp=0xc00058afc0 m=nil [sync.Cond.Wait]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0xc0003adc78?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc0003adbe8 sp=0xc0003adbc8 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.goparkunlock(...) 19:10:54 arch ollama[927]: runtime/proc.go:441 19:10:54 arch ollama[927]: sync.runtime_notifyListWait(0xc000305310, 0x0) 19:10:54 arch ollama[927]: runtime/sema.go:597 +0x15a fp=0xc0003adc38 sp=0xc0003adbe8 pc=0x5625c3c1b6ba 19:10:54 arch ollama[927]: sync.(*Cond).Wait(0xc0003adcb8?) 19:10:54 arch ollama[927]: sync/cond.go:71 +0x85 fp=0xc0003adc70 sp=0xc0003adc38 pc=0x5625c3c2b7c5 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).processBatch(0xc000254780, 0xc00048a190, 0xc00048a1e0) 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:408 +0x93 fp=0xc0003adee8 sp=0xc0003adc70 pc=0x5625c408a213 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.(*Server).run(0xc000254780, {0x5625c51cf410, 0xc000018730}) 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:387 +0x1d5 fp=0xc0003adfb8 sp=0xc0003adee8 pc=0x5625c408a015 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner.Execute.gowrap1() 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:981 +0x28 fp=0xc0003adfe0 sp=0xc0003adfb8 pc=0x5625c408f3e8 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0003adfe8 sp=0xc0003adfe0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by github.com/ollama/ollama/runner/llamarunner.Execute in goroutine 1 19:10:54 arch ollama[927]: github.com/ollama/ollama/runner/llamarunner/runner.go:981 +0x4c5 19:10:54 arch ollama[927]: goroutine 98 gp=0xc0003b7dc0 m=nil [IO wait]: 19:10:54 arch ollama[927]: runtime.gopark(0x0?, 0x0?, 0x0?, 0x0?, 0xb?) 19:10:54 arch ollama[927]: runtime/proc.go:435 +0xce fp=0xc0003ac5d8 sp=0xc0003ac5b8 pc=0x5625c3c19b6e 19:10:54 arch ollama[927]: runtime.netpollblock(0x5625c3c3d338?, 0xc3bb32a6?, 0x25?) 19:10:54 arch ollama[927]: runtime/netpoll.go:575 +0xf7 fp=0xc0003ac610 sp=0xc0003ac5d8 pc=0x5625c3bdee97 19:10:54 arch ollama[927]: internal/poll.runtime_pollWait(0x7fd89dec5cc8, 0x72) 19:10:54 arch ollama[927]: runtime/netpoll.go:351 +0x85 fp=0xc0003ac630 sp=0xc0003ac610 pc=0x5625c3c18d85 19:10:54 arch ollama[927]: internal/poll.(*pollDesc).wait(0xc000170a00?, 0xc00023d0c1?, 0x0) 19:10:54 arch ollama[927]: internal/poll/fd_poll_runtime.go:84 +0x27 fp=0xc0003ac658 sp=0xc0003ac630 pc=0x5625c3ca0f07 19:10:54 arch ollama[927]: internal/poll.(*pollDesc).waitRead(...) 19:10:54 arch ollama[927]: internal/poll/fd_poll_runtime.go:89 19:10:54 arch ollama[927]: internal/poll.(*FD).Read(0xc000170a00, {0xc00023d0c1, 0x1, 0x1}) 19:10:54 arch ollama[927]: internal/poll/fd_unix.go:165 +0x27a fp=0xc0003ac6f0 sp=0xc0003ac658 pc=0x5625c3ca21fa 19:10:54 arch ollama[927]: net.(*netFD).Read(0xc000170a00, {0xc00023d0c1?, 0x0?, 0x0?}) 19:10:54 arch ollama[927]: net/fd_posix.go:55 +0x25 fp=0xc0003ac738 sp=0xc0003ac6f0 pc=0x5625c3d17205 19:10:54 arch ollama[927]: net.(*conn).Read(0xc00011ca30, {0xc00023d0c1?, 0xc0003ac700?, 0x0?}) 19:10:54 arch ollama[927]: net/net.go:194 +0x45 fp=0xc0003ac780 sp=0xc0003ac738 pc=0x5625c3d255c5 19:10:54 arch ollama[927]: net/http.(*connReader).backgroundRead(0xc00023d0b0) 19:10:54 arch ollama[927]: net/http/server.go:690 +0x37 fp=0xc0003ac7c8 sp=0xc0003ac780 pc=0x5625c3f114b7 19:10:54 arch ollama[927]: net/http.(*connReader).startBackgroundRead.gowrap2() 19:10:54 arch ollama[927]: net/http/server.go:686 +0x25 fp=0xc0003ac7e0 sp=0xc0003ac7c8 pc=0x5625c3f113e5 19:10:54 arch ollama[927]: runtime.goexit({}) 19:10:54 arch ollama[927]: runtime/asm_amd64.s:1700 +0x1 fp=0xc0003ac7e8 sp=0xc0003ac7e0 pc=0x5625c3c21a01 19:10:54 arch ollama[927]: created by net/http.(*connReader).startBackgroundRead in goroutine 10 19:10:54 arch ollama[927]: net/http/server.go:686 +0xb6 19:10:54 arch ollama[927]: rax 0x7fd7ca850b90 19:10:54 arch ollama[927]: rbx 0x0 19:10:54 arch ollama[927]: rcx 0x0 19:10:54 arch ollama[927]: rdx 0x0 19:10:54 arch ollama[927]: rdi 0x7fd7cd7bc0e0 19:10:54 arch ollama[927]: rsi 0x7fd7be22f310 19:10:54 arch ollama[927]: rbp 0x7fd7ca5e8148 19:10:54 arch ollama[927]: rsp 0x7fd89deb9bc0 19:10:54 arch ollama[927]: r8 0x0 19:10:54 arch ollama[927]: r9 0x7fd771f0b040 19:10:54 arch ollama[927]: r10 0x24db000 19:10:54 arch ollama[927]: r11 0x246 19:10:54 arch ollama[927]: r12 0x7fd6bc45ac80 19:10:54 arch ollama[927]: r13 0x160 19:10:54 arch ollama[927]: r14 0x7fd7be22f030 19:10:54 arch ollama[927]: r15 0x1 19:10:54 arch ollama[927]: rip 0x5625c497b1cb 19:10:54 arch ollama[927]: rflags 0x10246 19:10:54 arch ollama[927]: cs 0x33 19:10:54 arch ollama[927]: fs 0x0 19:10:54 arch ollama[927]: gs 0x0 19:10:54 arch ollama[927]: time=2026-01-05T19:10:54.181+08:00 level=ERROR source=server.go:1583 msg="post predict" error="Post \"http://127.0.0.1:41297/completion\": EOF" 19:10:54 arch ollama[927]: [GIN] 2026/01/05 - 19:10:54 | 500 | 9.456008317s | 127.0.0.1 | POST "/api/generate" 19:10:54 arch ollama[927]: time=2026-01-05T19:10:54.331+08:00 level=ERROR source=server.go:302 msg="llama runner terminated" error="exit status 2" ```
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label model
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#55380
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?