github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-07 00:22:43 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[GH-ISSUE #9221] 想问下怎么在ollama启动模型的时候设置其上下文长度? #31771

New Issue
Closed
opened 2026-04-22 12:32:06 -05:00 by GiteaMirror · 6 comments
GiteaMirror commented 2026-04-22 12:32:06 -05:00
Owner
Copy Link

Originally created by @948024326 on GitHub (Feb 19, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/9221

我当前的命令是这样，因为我想让模型永久挂在GPU中，但是我设置了num_ctx 为8192 好像没有效果

curl -X POST http://localhost:11434/api/generate ^
-H "Content-Type: application/json" ^
-d "{"model": "deepseek-r1:32b", "num_ctx": 8192, "keep_alive": -1}"

Originally created by @948024326 on GitHub (Feb 19, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/9221 我当前的命令是这样，因为我想让模型永久挂在GPU中，但是我设置了num_ctx 为8192 好像没有效果 curl -X POST http://localhost:11434/api/generate ^ -H "Content-Type: application/json" ^ -d "{\"model\": \"deepseek-r1:32b\", \"num_ctx\": 8192, \"keep_alive\": -1}"
GiteaMirror commented 2026-04-22 12:32:07 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Feb 19, 2025):

curl -X POST http://localhost:11434/api/generate ^
-H "Content-Type: application/json" ^
-d "{\"model\": \"deepseek-r1:32b\", \"options\":{\"num_ctx\": 8192}, \"keep_alive\": -1}"
<!-- gh-comment-id:2668466070 --> @rick-github commented on GitHub (Feb 19, 2025): ``` curl -X POST http://localhost:11434/api/generate ^ -H "Content-Type: application/json" ^ -d "{\"model\": \"deepseek-r1:32b\", \"options\":{\"num_ctx\": 8192}, \"keep_alive\": -1}" ```
GiteaMirror commented 2026-04-22 12:32:08 -05:00
Author
Owner
Copy Link

@948024326 commented on GitHub (Feb 20, 2025):

curl -X POST http://localhost:11434/api/generate ^
-H "Content-Type: application/json" ^
-d "{\"model\": \"deepseek-r1:32b\", \"options\":{\"num_ctx\": 8192}, \"keep_alive\": -1}"

it occurs a bug?
when i use your command to start the model, the result is like this about 43G GPU

Image

but when i start to chat, it suddenly down to 22G, and the model time from forever change to 5 Min

Image

can you help me?

<!-- gh-comment-id:2670804355 --> @948024326 commented on GitHub (Feb 20, 2025): > ``` > curl -X POST http://localhost:11434/api/generate ^ > -H "Content-Type: application/json" ^ > -d "{\"model\": \"deepseek-r1:32b\", \"options\":{\"num_ctx\": 8192}, \"keep_alive\": -1}" > ``` it occurs a bug? when i use your command to start the model, the result is like this about 43G GPU ![Image](https://github.com/user-attachments/assets/d8488fd1-096f-4b0e-8cd1-4aa8c4e68de8) but when i start to chat, it suddenly down to 22G, and the model time from forever change to 5 Min ![Image](https://github.com/user-attachments/assets/8b9995bc-17c7-42d2-a4da-e373b07ce108) can you help me?
GiteaMirror commented 2026-04-22 12:32:10 -05:00
Author
Owner
Copy Link

@948024326 commented on GitHub (Feb 20, 2025):

curl -X POST http://localhost:11434/api/generate ^
-H "Content-Type: application/json" ^
-d "{\"model\": \"deepseek-r1:32b\", \"options\":{\"num_ctx\": 8192}, \"keep_alive\": -1}"

and here is my log

2025/02/20 16:32:58 routes.go:1187: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://192.168.242.247:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:C:\Users\Administrator\.ollama\models OLLAMA_MULTIUSER_CACHE:false OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:0 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://* vscode-webview://*] OLLAMA_SCHED_SPREAD:false ROCR_VISIBLE_DEVICES:]"
time=2025-02-20T16:32:58.754+08:00 level=INFO source=images.go:432 msg="total blobs: 10"
time=2025-02-20T16:32:58.756+08:00 level=INFO source=images.go:439 msg="total unused blobs removed: 0"
time=2025-02-20T16:32:58.757+08:00 level=INFO source=routes.go:1238 msg="Listening on 192.168.242.247:11434 (version 0.5.7)"
time=2025-02-20T16:32:58.757+08:00 level=INFO source=routes.go:1267 msg="Dynamic LLM libraries" runners="[cpu_avx cpu_avx2 cuda_v11_avx cuda_v12_avx rocm_avx cpu]"
time=2025-02-20T16:32:58.758+08:00 level=INFO source=gpu.go:226 msg="looking for compatible GPUs"
time=2025-02-20T16:32:58.758+08:00 level=INFO source=gpu_windows.go:167 msg=packages count=1
time=2025-02-20T16:32:58.758+08:00 level=INFO source=gpu_windows.go:214 msg="" package=0 cores=64 efficiency=0 threads=128
time=2025-02-20T16:32:59.016+08:00 level=INFO source=gpu.go:334 msg="detected OS VRAM overhead" id=GPU-27ef0894-cb58-204e-0830-e9f6cf9a3c86 library=cuda compute=8.6 driver=12.4 name="NVIDIA RTX A6000" overhead="459.5 MiB"
time=2025-02-20T16:32:59.017+08:00 level=INFO source=types.go:131 msg="inference compute" id=GPU-27ef0894-cb58-204e-0830-e9f6cf9a3c86 library=cuda variant=v12 compute=8.6 driver=12.4 name="NVIDIA RTX A6000" total="48.0 GiB" available="46.2 GiB"
[GIN] 2025/02/20 - 16:32:59 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:32:59 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:33:19 | 200 | 285.5µs | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:33:19 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
time=2025-02-20T16:33:20.954+08:00 level=INFO source=sched.go:714 msg="new model will fit in available VRAM in single GPU, loading" model=C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 gpu=GPU-27ef0894-cb58-204e-0830-e9f6cf9a3c86 parallel=4 available=49596465152 required="40.2 GiB"
time=2025-02-20T16:33:20.964+08:00 level=INFO source=server.go:104 msg="system memory" total="127.8 GiB" free="106.2 GiB" free_swap="123.8 GiB"
time=2025-02-20T16:33:20.966+08:00 level=INFO source=memory.go:356 msg="offload to cuda" layers.requested=-1 layers.model=65 layers.offload=65 layers.split="" memory.available="[46.2 GiB]" memory.gpu_overhead="0 B" memory.required.full="40.2 GiB" memory.required.partial="40.2 GiB" memory.required.kv="16.0 GiB" memory.required.allocations="[40.2 GiB]" memory.weights.total="33.5 GiB" memory.weights.repeating="32.9 GiB" memory.weights.nonrepeating="609.1 MiB" memory.graph.full="5.1 GiB" memory.graph.partial="6.4 GiB"
time=2025-02-20T16:33:20.977+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="C:\Users\Administrator\AppData\Local\Programs\Ollama\lib\ollama\runners\cuda_v12_avx\ollama_llama_server.exe runner --model C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 --ctx-size 65536 --batch-size 512 --n-gpu-layers 65 --threads 64 --no-mmap --parallel 4 --port 55922"
time=2025-02-20T16:33:21.022+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1
time=2025-02-20T16:33:21.022+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding"
time=2025-02-20T16:33:21.022+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error"
time=2025-02-20T16:33:21.107+08:00 level=INFO source=runner.go:936 msg="starting go runner"
ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 1 CUDA devices:
Device 0: NVIDIA RTX A6000, compute capability 8.6, VMM: yes
time=2025-02-20T16:33:21.151+08:00 level=INFO source=runner.go:937 msg=system info="CUDA : ARCHS = 600,610,620,700,720,750,800,860,870,890,900 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | LLAMAFILE = 1 | AARCH64_REPACK = 1 | cgo(clang)" threads=64
time=2025-02-20T16:33:21.152+08:00 level=INFO source=.:0 msg="Server listening on 127.0.0.1:55922"
time=2025-02-20T16:33:21.274+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server loading model"
llama_load_model_from_file: using device CUDA0 (NVIDIA RTX A6000) - 47299 MiB free
llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = qwen2
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B
llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen
llama_model_loader: - kv 4: general.size_label str = 32B
llama_model_loader: - kv 5: qwen2.block_count u32 = 64
llama_model_loader: - kv 6: qwen2.context_length u32 = 131072
llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120
llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648
llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40
llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8
llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 13: general.file_type u32 = 15
llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen
llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", """, "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646
llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643
llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643
llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true
llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false
llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de...
llama_model_loader: - kv 25: general.quantization_version u32 = 2
llama_model_loader: - type f32: 321 tensors
llama_model_loader: - type q4_K: 385 tensors
llama_model_loader: - type q6_K: 65 tensors
llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default'
llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect
llm_load_vocab: special tokens cache size = 22
llm_load_vocab: token to piece cache size = 0.9310 MB
llm_load_print_meta: format = GGUF V3 (latest)
llm_load_print_meta: arch = qwen2
llm_load_print_meta: vocab type = BPE
llm_load_print_meta: n_vocab = 152064
llm_load_print_meta: n_merges = 151387
llm_load_print_meta: vocab_only = 0
llm_load_print_meta: n_ctx_train = 131072
llm_load_print_meta: n_embd = 5120
llm_load_print_meta: n_layer = 64
llm_load_print_meta: n_head = 40
llm_load_print_meta: n_head_kv = 8
llm_load_print_meta: n_rot = 128
llm_load_print_meta: n_swa = 0
llm_load_print_meta: n_embd_head_k = 128
llm_load_print_meta: n_embd_head_v = 128
llm_load_print_meta: n_gqa = 5
llm_load_print_meta: n_embd_k_gqa = 1024
llm_load_print_meta: n_embd_v_gqa = 1024
llm_load_print_meta: f_norm_eps = 0.0e+00
llm_load_print_meta: f_norm_rms_eps = 1.0e-05
llm_load_print_meta: f_clamp_kqv = 0.0e+00
llm_load_print_meta: f_max_alibi_bias = 0.0e+00
llm_load_print_meta: f_logit_scale = 0.0e+00
llm_load_print_meta: n_ff = 27648
llm_load_print_meta: n_expert = 0
llm_load_print_meta: n_expert_used = 0
llm_load_print_meta: causal attn = 1
llm_load_print_meta: pooling type = 0
llm_load_print_meta: rope type = 2
llm_load_print_meta: rope scaling = linear
llm_load_print_meta: freq_base_train = 1000000.0
llm_load_print_meta: freq_scale_train = 1
llm_load_print_meta: n_ctx_orig_yarn = 131072
llm_load_print_meta: rope_finetuned = unknown
llm_load_print_meta: ssm_d_conv = 0
llm_load_print_meta: ssm_d_inner = 0
llm_load_print_meta: ssm_d_state = 0
llm_load_print_meta: ssm_dt_rank = 0
llm_load_print_meta: ssm_dt_b_c_rms = 0
llm_load_print_meta: model type = 32B
llm_load_print_meta: model ftype = Q4_K - Medium
llm_load_print_meta: model params = 32.76 B
llm_load_print_meta: model size = 18.48 GiB (4.85 BPW)
llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B
llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>'
llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: LF token = 148848 'ÄĬ'
llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>'
llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>'
llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>'
llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>'
llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>'
llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>'
llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOG token = 151662 '<|fim_pad|>'
llm_load_print_meta: EOG token = 151663 '<|repo_name|>'
llm_load_print_meta: EOG token = 151664 '<|file_sep|>'
llm_load_print_meta: max token length = 256
llm_load_tensors: offloading 64 repeating layers to GPU
llm_load_tensors: offloading output layer to GPU
llm_load_tensors: offloaded 65/65 layers to GPU
llm_load_tensors: CPU model buffer size = 417.66 MiB
llm_load_tensors: CUDA0 model buffer size = 18508.35 MiB
llama_new_context_with_model: n_seq_max = 4
llama_new_context_with_model: n_ctx = 65536
llama_new_context_with_model: n_ctx_per_seq = 16384
llama_new_context_with_model: n_batch = 2048
llama_new_context_with_model: n_ubatch = 512
llama_new_context_with_model: flash_attn = 0
llama_new_context_with_model: freq_base = 1000000.0
llama_new_context_with_model: freq_scale = 1
llama_new_context_with_model: n_ctx_per_seq (16384) < n_ctx_train (131072) -- the full capacity of the model will not be utilized
llama_kv_cache_init: kv_size = 65536, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 64, can_shift = 1
llama_kv_cache_init: CUDA0 KV buffer size = 16384.00 MiB
llama_new_context_with_model: KV self size = 16384.00 MiB, K (f16): 8192.00 MiB, V (f16): 8192.00 MiB
llama_new_context_with_model: CUDA_Host output buffer size = 2.40 MiB
llama_new_context_with_model: CUDA0 compute buffer size = 5288.00 MiB
llama_new_context_with_model: CUDA_Host compute buffer size = 138.01 MiB
llama_new_context_with_model: graph nodes = 2246
llama_new_context_with_model: graph splits = 2
time=2025-02-20T16:33:25.286+08:00 level=INFO source=server.go:594 msg="llama runner started in 4.26 seconds"
[GIN] 2025/02/20 - 16:33:25 | 200 | 4.3991734s | 192.168.242.247 | POST "/api/generate"
[GIN] 2025/02/20 - 16:33:51 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:33:51 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:33:53 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:33:53 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
time=2025-02-20T16:34:06.948+08:00 level=INFO source=sched.go:714 msg="new model will fit in available VRAM in single GPU, loading" model=C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 gpu=GPU-27ef0894-cb58-204e-0830-e9f6cf9a3c86 parallel=4 available=49588076544 required="21.5 GiB"
time=2025-02-20T16:34:06.958+08:00 level=INFO source=server.go:104 msg="system memory" total="127.8 GiB" free="106.3 GiB" free_swap="123.8 GiB"
time=2025-02-20T16:34:06.959+08:00 level=INFO source=memory.go:356 msg="offload to cuda" layers.requested=-1 layers.model=65 layers.offload=65 layers.split="" memory.available="[46.2 GiB]" memory.gpu_overhead="0 B" memory.required.full="21.5 GiB" memory.required.partial="21.5 GiB" memory.required.kv="2.0 GiB" memory.required.allocations="[21.5 GiB]" memory.weights.total="19.5 GiB" memory.weights.repeating="18.9 GiB" memory.weights.nonrepeating="609.1 MiB" memory.graph.full="676.0 MiB" memory.graph.partial="916.1 MiB"
time=2025-02-20T16:34:06.964+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="C:\Users\Administrator\AppData\Local\Programs\Ollama\lib\ollama\runners\cuda_v12_avx\ollama_llama_server.exe runner --model C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 --ctx-size 8192 --batch-size 512 --n-gpu-layers 65 --threads 64 --no-mmap --parallel 4 --port 55954"
time=2025-02-20T16:34:07.010+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1
time=2025-02-20T16:34:07.010+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding"
time=2025-02-20T16:34:07.010+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error"
time=2025-02-20T16:34:07.094+08:00 level=INFO source=runner.go:936 msg="starting go runner"
ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 1 CUDA devices:
Device 0: NVIDIA RTX A6000, compute capability 8.6, VMM: yes
time=2025-02-20T16:34:07.149+08:00 level=INFO source=runner.go:937 msg=system info="CUDA : ARCHS = 600,610,620,700,720,750,800,860,870,890,900 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | LLAMAFILE = 1 | AARCH64_REPACK = 1 | cgo(clang)" threads=64
time=2025-02-20T16:34:07.150+08:00 level=INFO source=.:0 msg="Server listening on 127.0.0.1:55954"
time=2025-02-20T16:34:07.263+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server loading model"
llama_load_model_from_file: using device CUDA0 (NVIDIA RTX A6000) - 47299 MiB free
llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = qwen2
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B
llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen
llama_model_loader: - kv 4: general.size_label str = 32B
llama_model_loader: - kv 5: qwen2.block_count u32 = 64
llama_model_loader: - kv 6: qwen2.context_length u32 = 131072
llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120
llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648
llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40
llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8
llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 13: general.file_type u32 = 15
llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen
llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", """, "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646
llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643
llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643
llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true
llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false
llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de...
llama_model_loader: - kv 25: general.quantization_version u32 = 2
llama_model_loader: - type f32: 321 tensors
llama_model_loader: - type q4_K: 385 tensors
llama_model_loader: - type q6_K: 65 tensors
llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default'
llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect
llm_load_vocab: special tokens cache size = 22
llm_load_vocab: token to piece cache size = 0.9310 MB
llm_load_print_meta: format = GGUF V3 (latest)
llm_load_print_meta: arch = qwen2
llm_load_print_meta: vocab type = BPE
llm_load_print_meta: n_vocab = 152064
llm_load_print_meta: n_merges = 151387
llm_load_print_meta: vocab_only = 0
llm_load_print_meta: n_ctx_train = 131072
llm_load_print_meta: n_embd = 5120
llm_load_print_meta: n_layer = 64
llm_load_print_meta: n_head = 40
llm_load_print_meta: n_head_kv = 8
llm_load_print_meta: n_rot = 128
llm_load_print_meta: n_swa = 0
llm_load_print_meta: n_embd_head_k = 128
llm_load_print_meta: n_embd_head_v = 128
llm_load_print_meta: n_gqa = 5
llm_load_print_meta: n_embd_k_gqa = 1024
llm_load_print_meta: n_embd_v_gqa = 1024
llm_load_print_meta: f_norm_eps = 0.0e+00
llm_load_print_meta: f_norm_rms_eps = 1.0e-05
llm_load_print_meta: f_clamp_kqv = 0.0e+00
llm_load_print_meta: f_max_alibi_bias = 0.0e+00
llm_load_print_meta: f_logit_scale = 0.0e+00
llm_load_print_meta: n_ff = 27648
llm_load_print_meta: n_expert = 0
llm_load_print_meta: n_expert_used = 0
llm_load_print_meta: causal attn = 1
llm_load_print_meta: pooling type = 0
llm_load_print_meta: rope type = 2
llm_load_print_meta: rope scaling = linear
llm_load_print_meta: freq_base_train = 1000000.0
llm_load_print_meta: freq_scale_train = 1
llm_load_print_meta: n_ctx_orig_yarn = 131072
llm_load_print_meta: rope_finetuned = unknown
llm_load_print_meta: ssm_d_conv = 0
llm_load_print_meta: ssm_d_inner = 0
llm_load_print_meta: ssm_d_state = 0
llm_load_print_meta: ssm_dt_rank = 0
llm_load_print_meta: ssm_dt_b_c_rms = 0
llm_load_print_meta: model type = 32B
llm_load_print_meta: model ftype = Q4_K - Medium
llm_load_print_meta: model params = 32.76 B
llm_load_print_meta: model size = 18.48 GiB (4.85 BPW)
llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B
llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>'
llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: LF token = 148848 'ÄĬ'
llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>'
llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>'
llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>'
llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>'
llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>'
llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>'
llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOG token = 151662 '<|fim_pad|>'
llm_load_print_meta: EOG token = 151663 '<|repo_name|>'
llm_load_print_meta: EOG token = 151664 '<|file_sep|>'
llm_load_print_meta: max token length = 256
llm_load_tensors: offloading 64 repeating layers to GPU
llm_load_tensors: offloading output layer to GPU
llm_load_tensors: offloaded 65/65 layers to GPU
llm_load_tensors: CPU model buffer size = 417.66 MiB
llm_load_tensors: CUDA0 model buffer size = 18508.35 MiB
llama_new_context_with_model: n_seq_max = 4
llama_new_context_with_model: n_ctx = 8192
llama_new_context_with_model: n_ctx_per_seq = 2048
llama_new_context_with_model: n_batch = 2048
llama_new_context_with_model: n_ubatch = 512
llama_new_context_with_model: flash_attn = 0
llama_new_context_with_model: freq_base = 1000000.0
llama_new_context_with_model: freq_scale = 1
llama_new_context_with_model: n_ctx_per_seq (2048) < n_ctx_train (131072) -- the full capacity of the model will not be utilized
llama_kv_cache_init: kv_size = 8192, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 64, can_shift = 1
llama_kv_cache_init: CUDA0 KV buffer size = 2048.00 MiB
llama_new_context_with_model: KV self size = 2048.00 MiB, K (f16): 1024.00 MiB, V (f16): 1024.00 MiB
llama_new_context_with_model: CUDA_Host output buffer size = 2.40 MiB
llama_new_context_with_model: CUDA0 compute buffer size = 696.00 MiB
llama_new_context_with_model: CUDA_Host compute buffer size = 26.01 MiB
llama_new_context_with_model: graph nodes = 2246
llama_new_context_with_model: graph splits = 2
time=2025-02-20T16:34:11.025+08:00 level=INFO source=server.go:594 msg="llama runner started in 4.01 seconds"
llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = qwen2
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B
llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen
llama_model_loader: - kv 4: general.size_label str = 32B
llama_model_loader: - kv 5: qwen2.block_count u32 = 64
llama_model_loader: - kv 6: qwen2.context_length u32 = 131072
llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120
llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648
llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40
llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8
llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 13: general.file_type u32 = 15
llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen
llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", """, "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646
llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643
llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643
llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true
llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false
llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de...
llama_model_loader: - kv 25: general.quantization_version u32 = 2
llama_model_loader: - type f32: 321 tensors
llama_model_loader: - type q4_K: 385 tensors
llama_model_loader: - type q6_K: 65 tensors
llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default'
llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect
llm_load_vocab: special tokens cache size = 22
llm_load_vocab: token to piece cache size = 0.9310 MB
llm_load_print_meta: format = GGUF V3 (latest)
llm_load_print_meta: arch = qwen2
llm_load_print_meta: vocab type = BPE
llm_load_print_meta: n_vocab = 152064
llm_load_print_meta: n_merges = 151387
llm_load_print_meta: vocab_only = 1
llm_load_print_meta: model type = ?B
llm_load_print_meta: model ftype = all F32
llm_load_print_meta: model params = 32.76 B
llm_load_print_meta: model size = 18.48 GiB (4.85 BPW)
llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B
llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>'
llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: LF token = 148848 'ÄĬ'
llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>'
llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>'
llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>'
llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>'
llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>'
llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>'
llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOG token = 151662 '<|fim_pad|>'
llm_load_print_meta: EOG token = 151663 '<|repo_name|>'
llm_load_print_meta: EOG token = 151664 '<|file_sep|>'
llm_load_print_meta: max token length = 256
llama_model_load: vocab only - skipping tensors
time=2025-02-20T16:34:11.359+08:00 level=WARN source=runner.go:129 msg="truncating input prompt" limit=2048 prompt=3963 keep=5 new=2048
[GIN] 2025/02/20 - 16:34:14 | 200 | 7.6964767s | 192.168.242.247 | POST "/api/chat"
[GIN] 2025/02/20 - 16:34:28 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:34:28 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:34:29 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:34:29 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:34:30 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:34:30 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:34:30 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:34:30 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:36:08 | 200 | 156.8µs | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:36:08 | 200 | 1.0341ms | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:36:32 | 200 | 7.4467209s | 192.168.242.247 | POST "/api/chat"
[GIN] 2025/02/20 - 16:38:08 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:38:08 | 200 | 611.4µs | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:38:09 | 200 | 208.6µs | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:38:09 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:38:10 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:38:10 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:38:29 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:38:29 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:38:30 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:38:30 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:41:04 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:41:04 | 200 | 20.5651ms | 192.168.242.247 | POST "/api/show"
time=2025-02-20T16:42:07.019+08:00 level=INFO source=sched.go:714 msg="new model will fit in available VRAM in single GPU, loading" model=C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 gpu=GPU-27ef0894-cb58-204e-0830-e9f6cf9a3c86 parallel=4 available=49617698816 required="21.5 GiB"
time=2025-02-20T16:42:07.030+08:00 level=INFO source=server.go:104 msg="system memory" total="127.8 GiB" free="106.3 GiB" free_swap="123.9 GiB"
time=2025-02-20T16:42:07.031+08:00 level=INFO source=memory.go:356 msg="offload to cuda" layers.requested=-1 layers.model=65 layers.offload=65 layers.split="" memory.available="[46.2 GiB]" memory.gpu_overhead="0 B" memory.required.full="21.5 GiB" memory.required.partial="21.5 GiB" memory.required.kv="2.0 GiB" memory.required.allocations="[21.5 GiB]" memory.weights.total="19.5 GiB" memory.weights.repeating="18.9 GiB" memory.weights.nonrepeating="609.1 MiB" memory.graph.full="676.0 MiB" memory.graph.partial="916.1 MiB"
time=2025-02-20T16:42:07.036+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="C:\Users\Administrator\AppData\Local\Programs\Ollama\lib\ollama\runners\cuda_v12_avx\ollama_llama_server.exe runner --model C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 --ctx-size 8192 --batch-size 512 --n-gpu-layers 65 --threads 64 --no-mmap --parallel 4 --port 56318"
time=2025-02-20T16:42:07.080+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1
time=2025-02-20T16:42:07.080+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding"
time=2025-02-20T16:42:07.080+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error"
time=2025-02-20T16:42:07.164+08:00 level=INFO source=runner.go:936 msg="starting go runner"
ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 1 CUDA devices:
Device 0: NVIDIA RTX A6000, compute capability 8.6, VMM: yes
time=2025-02-20T16:42:07.210+08:00 level=INFO source=runner.go:937 msg=system info="CUDA : ARCHS = 600,610,620,700,720,750,800,860,870,890,900 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | LLAMAFILE = 1 | AARCH64_REPACK = 1 | cgo(clang)" threads=64
time=2025-02-20T16:42:07.212+08:00 level=INFO source=.:0 msg="Server listening on 127.0.0.1:56318"
time=2025-02-20T16:42:07.331+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server loading model"
llama_load_model_from_file: using device CUDA0 (NVIDIA RTX A6000) - 47299 MiB free
llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = qwen2
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B
llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen
llama_model_loader: - kv 4: general.size_label str = 32B
llama_model_loader: - kv 5: qwen2.block_count u32 = 64
llama_model_loader: - kv 6: qwen2.context_length u32 = 131072
llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120
llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648
llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40
llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8
llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 13: general.file_type u32 = 15
llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen
llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", """, "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646
llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643
llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643
llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true
llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false
llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de...
llama_model_loader: - kv 25: general.quantization_version u32 = 2
llama_model_loader: - type f32: 321 tensors
llama_model_loader: - type q4_K: 385 tensors
llama_model_loader: - type q6_K: 65 tensors
llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default'
llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect
llm_load_vocab: special tokens cache size = 22
llm_load_vocab: token to piece cache size = 0.9310 MB
llm_load_print_meta: format = GGUF V3 (latest)
llm_load_print_meta: arch = qwen2
llm_load_print_meta: vocab type = BPE
llm_load_print_meta: n_vocab = 152064
llm_load_print_meta: n_merges = 151387
llm_load_print_meta: vocab_only = 0
llm_load_print_meta: n_ctx_train = 131072
llm_load_print_meta: n_embd = 5120
llm_load_print_meta: n_layer = 64
llm_load_print_meta: n_head = 40
llm_load_print_meta: n_head_kv = 8
llm_load_print_meta: n_rot = 128
llm_load_print_meta: n_swa = 0
llm_load_print_meta: n_embd_head_k = 128
llm_load_print_meta: n_embd_head_v = 128
llm_load_print_meta: n_gqa = 5
llm_load_print_meta: n_embd_k_gqa = 1024
llm_load_print_meta: n_embd_v_gqa = 1024
llm_load_print_meta: f_norm_eps = 0.0e+00
llm_load_print_meta: f_norm_rms_eps = 1.0e-05
llm_load_print_meta: f_clamp_kqv = 0.0e+00
llm_load_print_meta: f_max_alibi_bias = 0.0e+00
llm_load_print_meta: f_logit_scale = 0.0e+00
llm_load_print_meta: n_ff = 27648
llm_load_print_meta: n_expert = 0
llm_load_print_meta: n_expert_used = 0
llm_load_print_meta: causal attn = 1
llm_load_print_meta: pooling type = 0
llm_load_print_meta: rope type = 2
llm_load_print_meta: rope scaling = linear
llm_load_print_meta: freq_base_train = 1000000.0
llm_load_print_meta: freq_scale_train = 1
llm_load_print_meta: n_ctx_orig_yarn = 131072
llm_load_print_meta: rope_finetuned = unknown
llm_load_print_meta: ssm_d_conv = 0
llm_load_print_meta: ssm_d_inner = 0
llm_load_print_meta: ssm_d_state = 0
llm_load_print_meta: ssm_dt_rank = 0
llm_load_print_meta: ssm_dt_b_c_rms = 0
llm_load_print_meta: model type = 32B
llm_load_print_meta: model ftype = Q4_K - Medium
llm_load_print_meta: model params = 32.76 B
llm_load_print_meta: model size = 18.48 GiB (4.85 BPW)
llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B
llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>'
llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: LF token = 148848 'ÄĬ'
llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>'
llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>'
llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>'
llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>'
llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>'
llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>'
llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOG token = 151662 '<|fim_pad|>'
llm_load_print_meta: EOG token = 151663 '<|repo_name|>'
llm_load_print_meta: EOG token = 151664 '<|file_sep|>'
llm_load_print_meta: max token length = 256
llm_load_tensors: offloading 64 repeating layers to GPU
llm_load_tensors: offloading output layer to GPU
llm_load_tensors: offloaded 65/65 layers to GPU
llm_load_tensors: CPU model buffer size = 417.66 MiB
llm_load_tensors: CUDA0 model buffer size = 18508.35 MiB
llama_new_context_with_model: n_seq_max = 4
llama_new_context_with_model: n_ctx = 8192
llama_new_context_with_model: n_ctx_per_seq = 2048
llama_new_context_with_model: n_batch = 2048
llama_new_context_with_model: n_ubatch = 512
llama_new_context_with_model: flash_attn = 0
llama_new_context_with_model: freq_base = 1000000.0
llama_new_context_with_model: freq_scale = 1
llama_new_context_with_model: n_ctx_per_seq (2048) < n_ctx_train (131072) -- the full capacity of the model will not be utilized
llama_kv_cache_init: kv_size = 8192, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 64, can_shift = 1
llama_kv_cache_init: CUDA0 KV buffer size = 2048.00 MiB
llama_new_context_with_model: KV self size = 2048.00 MiB, K (f16): 1024.00 MiB, V (f16): 1024.00 MiB
llama_new_context_with_model: CUDA_Host output buffer size = 2.40 MiB
llama_new_context_with_model: CUDA0 compute buffer size = 696.00 MiB
llama_new_context_with_model: CUDA_Host compute buffer size = 26.01 MiB
llama_new_context_with_model: graph nodes = 2246
llama_new_context_with_model: graph splits = 2
time=2025-02-20T16:42:11.094+08:00 level=INFO source=server.go:594 msg="llama runner started in 4.01 seconds"
llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = qwen2
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B
llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen
llama_model_loader: - kv 4: general.size_label str = 32B
llama_model_loader: - kv 5: qwen2.block_count u32 = 64
llama_model_loader: - kv 6: qwen2.context_length u32 = 131072
llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120
llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648
llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40
llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8
llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 13: general.file_type u32 = 15
llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen
llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", """, "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646
llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643
llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643
llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true
llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false
llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de...
llama_model_loader: - kv 25: general.quantization_version u32 = 2
llama_model_loader: - type f32: 321 tensors
llama_model_loader: - type q4_K: 385 tensors
llama_model_loader: - type q6_K: 65 tensors
llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default'
llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect
llm_load_vocab: special tokens cache size = 22
llm_load_vocab: token to piece cache size = 0.9310 MB
llm_load_print_meta: format = GGUF V3 (latest)
llm_load_print_meta: arch = qwen2
llm_load_print_meta: vocab type = BPE
llm_load_print_meta: n_vocab = 152064
llm_load_print_meta: n_merges = 151387
llm_load_print_meta: vocab_only = 1
llm_load_print_meta: model type = ?B
llm_load_print_meta: model ftype = all F32
llm_load_print_meta: model params = 32.76 B
llm_load_print_meta: model size = 18.48 GiB (4.85 BPW)
llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B
llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>'
llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: LF token = 148848 'ÄĬ'
llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>'
llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>'
llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>'
llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>'
llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>'
llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>'
llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOG token = 151662 '<|fim_pad|>'
llm_load_print_meta: EOG token = 151663 '<|repo_name|>'
llm_load_print_meta: EOG token = 151664 '<|file_sep|>'
llm_load_print_meta: max token length = 256
llama_model_load: vocab only - skipping tensors
[GIN] 2025/02/20 - 16:42:36 | 200 | 29.3566393s | 192.168.242.247 | POST "/api/chat"
[GIN] 2025/02/20 - 16:43:10 | 200 | 31.1044914s | 192.168.242.247 | POST "/api/chat"
[GIN] 2025/02/20 - 16:44:45 | 200 | 35.5680529s | 192.168.242.247 | POST "/api/chat"
time=2025-02-20T16:49:29.342+08:00 level=INFO source=sched.go:714 msg="new model will fit in available VRAM in single GPU, loading" model=C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 gpu=GPU-27ef0894-cb58-204e-0830-e9f6cf9a3c86 parallel=4 available=49609834496 required="40.2 GiB"
time=2025-02-20T16:49:29.354+08:00 level=INFO source=server.go:104 msg="system memory" total="127.8 GiB" free="106.3 GiB" free_swap="123.8 GiB"
time=2025-02-20T16:49:29.356+08:00 level=INFO source=memory.go:356 msg="offload to cuda" layers.requested=-1 layers.model=65 layers.offload=65 layers.split="" memory.available="[46.2 GiB]" memory.gpu_overhead="0 B" memory.required.full="40.2 GiB" memory.required.partial="40.2 GiB" memory.required.kv="16.0 GiB" memory.required.allocations="[40.2 GiB]" memory.weights.total="33.5 GiB" memory.weights.repeating="32.9 GiB" memory.weights.nonrepeating="609.1 MiB" memory.graph.full="5.1 GiB" memory.graph.partial="6.4 GiB"
time=2025-02-20T16:49:29.360+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="C:\Users\Administrator\AppData\Local\Programs\Ollama\lib\ollama\runners\cuda_v12_avx\ollama_llama_server.exe runner --model C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 --ctx-size 65536 --batch-size 512 --n-gpu-layers 65 --threads 64 --no-mmap --parallel 4 --port 56698"
time=2025-02-20T16:49:29.403+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1
time=2025-02-20T16:49:29.403+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding"
time=2025-02-20T16:49:29.403+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error"
time=2025-02-20T16:49:29.493+08:00 level=INFO source=runner.go:936 msg="starting go runner"
ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 1 CUDA devices:
Device 0: NVIDIA RTX A6000, compute capability 8.6, VMM: yes
time=2025-02-20T16:49:29.547+08:00 level=INFO source=runner.go:937 msg=system info="CUDA : ARCHS = 600,610,620,700,720,750,800,860,870,890,900 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | LLAMAFILE = 1 | AARCH64_REPACK = 1 | cgo(clang)" threads=64
time=2025-02-20T16:49:29.548+08:00 level=INFO source=.:0 msg="Server listening on 127.0.0.1:56698"
time=2025-02-20T16:49:29.656+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server loading model"
llama_load_model_from_file: using device CUDA0 (NVIDIA RTX A6000) - 47299 MiB free
llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = qwen2
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B
llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen
llama_model_loader: - kv 4: general.size_label str = 32B
llama_model_loader: - kv 5: qwen2.block_count u32 = 64
llama_model_loader: - kv 6: qwen2.context_length u32 = 131072
llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120
llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648
llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40
llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8
llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 13: general.file_type u32 = 15
llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen
llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", """, "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646
llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643
llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643
llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true
llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false
llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de...
llama_model_loader: - kv 25: general.quantization_version u32 = 2
llama_model_loader: - type f32: 321 tensors
llama_model_loader: - type q4_K: 385 tensors
llama_model_loader: - type q6_K: 65 tensors
llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default'
llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect
llm_load_vocab: special tokens cache size = 22
llm_load_vocab: token to piece cache size = 0.9310 MB
llm_load_print_meta: format = GGUF V3 (latest)
llm_load_print_meta: arch = qwen2
llm_load_print_meta: vocab type = BPE
llm_load_print_meta: n_vocab = 152064
llm_load_print_meta: n_merges = 151387
llm_load_print_meta: vocab_only = 0
llm_load_print_meta: n_ctx_train = 131072
llm_load_print_meta: n_embd = 5120
llm_load_print_meta: n_layer = 64
llm_load_print_meta: n_head = 40
llm_load_print_meta: n_head_kv = 8
llm_load_print_meta: n_rot = 128
llm_load_print_meta: n_swa = 0
llm_load_print_meta: n_embd_head_k = 128
llm_load_print_meta: n_embd_head_v = 128
llm_load_print_meta: n_gqa = 5
llm_load_print_meta: n_embd_k_gqa = 1024
llm_load_print_meta: n_embd_v_gqa = 1024
llm_load_print_meta: f_norm_eps = 0.0e+00
llm_load_print_meta: f_norm_rms_eps = 1.0e-05
llm_load_print_meta: f_clamp_kqv = 0.0e+00
llm_load_print_meta: f_max_alibi_bias = 0.0e+00
llm_load_print_meta: f_logit_scale = 0.0e+00
llm_load_print_meta: n_ff = 27648
llm_load_print_meta: n_expert = 0
llm_load_print_meta: n_expert_used = 0
llm_load_print_meta: causal attn = 1
llm_load_print_meta: pooling type = 0
llm_load_print_meta: rope type = 2
llm_load_print_meta: rope scaling = linear
llm_load_print_meta: freq_base_train = 1000000.0
llm_load_print_meta: freq_scale_train = 1
llm_load_print_meta: n_ctx_orig_yarn = 131072
llm_load_print_meta: rope_finetuned = unknown
llm_load_print_meta: ssm_d_conv = 0
llm_load_print_meta: ssm_d_inner = 0
llm_load_print_meta: ssm_d_state = 0
llm_load_print_meta: ssm_dt_rank = 0
llm_load_print_meta: ssm_dt_b_c_rms = 0
llm_load_print_meta: model type = 32B
llm_load_print_meta: model ftype = Q4_K - Medium
llm_load_print_meta: model params = 32.76 B
llm_load_print_meta: model size = 18.48 GiB (4.85 BPW)
llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B
llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>'
llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: LF token = 148848 'ÄĬ'
llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>'
llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>'
llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>'
llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>'
llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>'
llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>'
llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOG token = 151662 '<|fim_pad|>'
llm_load_print_meta: EOG token = 151663 '<|repo_name|>'
llm_load_print_meta: EOG token = 151664 '<|file_sep|>'
llm_load_print_meta: max token length = 256
llm_load_tensors: offloading 64 repeating layers to GPU
llm_load_tensors: offloading output layer to GPU
llm_load_tensors: offloaded 65/65 layers to GPU
llm_load_tensors: CPU model buffer size = 417.66 MiB
llm_load_tensors: CUDA0 model buffer size = 18508.35 MiB
time=2025-02-20T16:49:30.659+08:00 level=WARN source=server.go:562 msg="client connection closed before server finished loading, aborting load"
time=2025-02-20T16:49:30.659+08:00 level=ERROR source=sched.go:455 msg="error loading llama server" error="timed out waiting for llama runner to start: context canceled"
[GIN] 2025/02/20 - 16:49:30 | 499 | 1.6502837s | 192.168.242.247 | POST "/api/generate"
time=2025-02-20T16:49:35.670+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.0106594 model=C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93
time=2025-02-20T16:49:35.919+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.2599917 model=C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93
time=2025-02-20T16:49:36.169+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.5101071 model=C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93
time=2025-02-20T16:50:49.357+08:00 level=INFO source=sched.go:714 msg="new model will fit in available VRAM in single GPU, loading" model=C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 gpu=GPU-27ef0894-cb58-204e-0830-e9f6cf9a3c86 parallel=4 available=49607868416 required="21.5 GiB"
time=2025-02-20T16:50:49.367+08:00 level=INFO source=server.go:104 msg="system memory" total="127.8 GiB" free="106.2 GiB" free_swap="123.8 GiB"
time=2025-02-20T16:50:49.369+08:00 level=INFO source=memory.go:356 msg="offload to cuda" layers.requested=-1 layers.model=65 layers.offload=65 layers.split="" memory.available="[46.2 GiB]" memory.gpu_overhead="0 B" memory.required.full="21.5 GiB" memory.required.partial="21.5 GiB" memory.required.kv="2.0 GiB" memory.required.allocations="[21.5 GiB]" memory.weights.total="19.5 GiB" memory.weights.repeating="18.9 GiB" memory.weights.nonrepeating="609.1 MiB" memory.graph.full="676.0 MiB" memory.graph.partial="916.1 MiB"
time=2025-02-20T16:50:49.373+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="C:\Users\Administrator\AppData\Local\Programs\Ollama\lib\ollama\runners\cuda_v12_avx\ollama_llama_server.exe runner --model C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 --ctx-size 8192 --batch-size 512 --n-gpu-layers 65 --threads 64 --no-mmap --parallel 4 --port 56884"
time=2025-02-20T16:50:49.417+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1
time=2025-02-20T16:50:49.417+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding"
time=2025-02-20T16:50:49.417+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error"
time=2025-02-20T16:50:49.509+08:00 level=INFO source=runner.go:936 msg="starting go runner"
ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 1 CUDA devices:
Device 0: NVIDIA RTX A6000, compute capability 8.6, VMM: yes
time=2025-02-20T16:50:49.556+08:00 level=INFO source=runner.go:937 msg=system info="CUDA : ARCHS = 600,610,620,700,720,750,800,860,870,890,900 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | LLAMAFILE = 1 | AARCH64_REPACK = 1 | cgo(clang)" threads=64
time=2025-02-20T16:50:49.558+08:00 level=INFO source=.:0 msg="Server listening on 127.0.0.1:56884"
time=2025-02-20T16:50:49.670+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server loading model"
llama_load_model_from_file: using device CUDA0 (NVIDIA RTX A6000) - 47299 MiB free
llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = qwen2
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B
llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen
llama_model_loader: - kv 4: general.size_label str = 32B
llama_model_loader: - kv 5: qwen2.block_count u32 = 64
llama_model_loader: - kv 6: qwen2.context_length u32 = 131072
llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120
llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648
llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40
llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8
llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 13: general.file_type u32 = 15
llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen
llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", """, "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646
llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643
llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643
llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true
llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false
llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de...
llama_model_loader: - kv 25: general.quantization_version u32 = 2
llama_model_loader: - type f32: 321 tensors
llama_model_loader: - type q4_K: 385 tensors
llama_model_loader: - type q6_K: 65 tensors
llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default'
llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect
llm_load_vocab: special tokens cache size = 22
llm_load_vocab: token to piece cache size = 0.9310 MB
llm_load_print_meta: format = GGUF V3 (latest)
llm_load_print_meta: arch = qwen2
llm_load_print_meta: vocab type = BPE
llm_load_print_meta: n_vocab = 152064
llm_load_print_meta: n_merges = 151387
llm_load_print_meta: vocab_only = 0
llm_load_print_meta: n_ctx_train = 131072
llm_load_print_meta: n_embd = 5120
llm_load_print_meta: n_layer = 64
llm_load_print_meta: n_head = 40
llm_load_print_meta: n_head_kv = 8
llm_load_print_meta: n_rot = 128
llm_load_print_meta: n_swa = 0
llm_load_print_meta: n_embd_head_k = 128
llm_load_print_meta: n_embd_head_v = 128
llm_load_print_meta: n_gqa = 5
llm_load_print_meta: n_embd_k_gqa = 1024
llm_load_print_meta: n_embd_v_gqa = 1024
llm_load_print_meta: f_norm_eps = 0.0e+00
llm_load_print_meta: f_norm_rms_eps = 1.0e-05
llm_load_print_meta: f_clamp_kqv = 0.0e+00
llm_load_print_meta: f_max_alibi_bias = 0.0e+00
llm_load_print_meta: f_logit_scale = 0.0e+00
llm_load_print_meta: n_ff = 27648
llm_load_print_meta: n_expert = 0
llm_load_print_meta: n_expert_used = 0
llm_load_print_meta: causal attn = 1
llm_load_print_meta: pooling type = 0
llm_load_print_meta: rope type = 2
llm_load_print_meta: rope scaling = linear
llm_load_print_meta: freq_base_train = 1000000.0
llm_load_print_meta: freq_scale_train = 1
llm_load_print_meta: n_ctx_orig_yarn = 131072
llm_load_print_meta: rope_finetuned = unknown
llm_load_print_meta: ssm_d_conv = 0
llm_load_print_meta: ssm_d_inner = 0
llm_load_print_meta: ssm_d_state = 0
llm_load_print_meta: ssm_dt_rank = 0
llm_load_print_meta: ssm_dt_b_c_rms = 0
llm_load_print_meta: model type = 32B
llm_load_print_meta: model ftype = Q4_K - Medium
llm_load_print_meta: model params = 32.76 B
llm_load_print_meta: model size = 18.48 GiB (4.85 BPW)
llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B
llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>'
llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: LF token = 148848 'ÄĬ'
llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>'
llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>'
llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>'
llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>'
llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>'
llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>'
llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOG token = 151662 '<|fim_pad|>'
llm_load_print_meta: EOG token = 151663 '<|repo_name|>'
llm_load_print_meta: EOG token = 151664 '<|file_sep|>'
llm_load_print_meta: max token length = 256
llm_load_tensors: offloading 64 repeating layers to GPU
llm_load_tensors: offloading output layer to GPU
llm_load_tensors: offloaded 65/65 layers to GPU
llm_load_tensors: CPU model buffer size = 417.66 MiB
llm_load_tensors: CUDA0 model buffer size = 18508.35 MiB
llama_new_context_with_model: n_seq_max = 4
llama_new_context_with_model: n_ctx = 8192
llama_new_context_with_model: n_ctx_per_seq = 2048
llama_new_context_with_model: n_batch = 2048
llama_new_context_with_model: n_ubatch = 512
llama_new_context_with_model: flash_attn = 0
llama_new_context_with_model: freq_base = 1000000.0
llama_new_context_with_model: freq_scale = 1
llama_new_context_with_model: n_ctx_per_seq (2048) < n_ctx_train (131072) -- the full capacity of the model will not be utilized
llama_kv_cache_init: kv_size = 8192, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 64, can_shift = 1
llama_kv_cache_init: CUDA0 KV buffer size = 2048.00 MiB
llama_new_context_with_model: KV self size = 2048.00 MiB, K (f16): 1024.00 MiB, V (f16): 1024.00 MiB
llama_new_context_with_model: CUDA_Host output buffer size = 2.40 MiB
llama_new_context_with_model: CUDA0 compute buffer size = 696.00 MiB
llama_new_context_with_model: CUDA_Host compute buffer size = 26.01 MiB
llama_new_context_with_model: graph nodes = 2246
llama_new_context_with_model: graph splits = 2
time=2025-02-20T16:50:53.429+08:00 level=INFO source=server.go:594 msg="llama runner started in 4.01 seconds"
llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = qwen2
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B
llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen
llama_model_loader: - kv 4: general.size_label str = 32B
llama_model_loader: - kv 5: qwen2.block_count u32 = 64
llama_model_loader: - kv 6: qwen2.context_length u32 = 131072
llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120
llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648
llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40
llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8
llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 13: general.file_type u32 = 15
llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen
llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", """, "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646
llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643
llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643
llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true
llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false
llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de...
llama_model_loader: - kv 25: general.quantization_version u32 = 2
llama_model_loader: - type f32: 321 tensors
llama_model_loader: - type q4_K: 385 tensors
llama_model_loader: - type q6_K: 65 tensors
llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default'
llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect
llm_load_vocab: special tokens cache size = 22
llm_load_vocab: token to piece cache size = 0.9310 MB
llm_load_print_meta: format = GGUF V3 (latest)
llm_load_print_meta: arch = qwen2
llm_load_print_meta: vocab type = BPE
llm_load_print_meta: n_vocab = 152064
llm_load_print_meta: n_merges = 151387
llm_load_print_meta: vocab_only = 1
llm_load_print_meta: model type = ?B
llm_load_print_meta: model ftype = all F32
llm_load_print_meta: model params = 32.76 B
llm_load_print_meta: model size = 18.48 GiB (4.85 BPW)
llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B
llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>'
llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: LF token = 148848 'ÄĬ'
llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>'
llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>'
llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>'
llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>'
llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>'
llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>'
llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>'
llm_load_print_meta: EOG token = 151662 '<|fim_pad|>'
llm_load_print_meta: EOG token = 151663 '<|repo_name|>'
llm_load_print_meta: EOG token = 151664 '<|file_sep|>'
llm_load_print_meta: max token length = 256
llama_model_load: vocab only - skipping tensors
time=2025-02-20T16:50:53.705+08:00 level=WARN source=runner.go:129 msg="truncating input prompt" limit=2048 prompt=3085 keep=5 new=2048
[GIN] 2025/02/20 - 16:51:35 | 200 | 45.8058413s | 192.168.242.247 | POST "/api/chat"
[GIN] 2025/02/20 - 16:51:48 | 200 | 26.03855s | 192.168.242.247 | POST "/api/chat"
[GIN] 2025/02/20 - 16:52:15 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:52:15 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:52:16 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:52:16 | 200 | 507.4µs | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:52:17 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:52:17 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:52:17 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:52:17 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:52:18 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:52:18 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
[GIN] 2025/02/20 - 16:52:18 | 200 | 0s | 192.168.242.247 | HEAD "/"
[GIN] 2025/02/20 - 16:52:18 | 200 | 0s | 192.168.242.247 | GET "/api/ps"

<!-- gh-comment-id:2670842225 --> @948024326 commented on GitHub (Feb 20, 2025): > ``` > curl -X POST http://localhost:11434/api/generate ^ > -H "Content-Type: application/json" ^ > -d "{\"model\": \"deepseek-r1:32b\", \"options\":{\"num_ctx\": 8192}, \"keep_alive\": -1}" > ``` and here is my log 2025/02/20 16:32:58 routes.go:1187: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://192.168.242.247:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:C:\\Users\\Administrator\\.ollama\\models OLLAMA_MULTIUSER_CACHE:false OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:0 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://* vscode-webview://*] OLLAMA_SCHED_SPREAD:false ROCR_VISIBLE_DEVICES:]" time=2025-02-20T16:32:58.754+08:00 level=INFO source=images.go:432 msg="total blobs: 10" time=2025-02-20T16:32:58.756+08:00 level=INFO source=images.go:439 msg="total unused blobs removed: 0" time=2025-02-20T16:32:58.757+08:00 level=INFO source=routes.go:1238 msg="Listening on 192.168.242.247:11434 (version 0.5.7)" time=2025-02-20T16:32:58.757+08:00 level=INFO source=routes.go:1267 msg="Dynamic LLM libraries" runners="[cpu_avx cpu_avx2 cuda_v11_avx cuda_v12_avx rocm_avx cpu]" time=2025-02-20T16:32:58.758+08:00 level=INFO source=gpu.go:226 msg="looking for compatible GPUs" time=2025-02-20T16:32:58.758+08:00 level=INFO source=gpu_windows.go:167 msg=packages count=1 time=2025-02-20T16:32:58.758+08:00 level=INFO source=gpu_windows.go:214 msg="" package=0 cores=64 efficiency=0 threads=128 time=2025-02-20T16:32:59.016+08:00 level=INFO source=gpu.go:334 msg="detected OS VRAM overhead" id=GPU-27ef0894-cb58-204e-0830-e9f6cf9a3c86 library=cuda compute=8.6 driver=12.4 name="NVIDIA RTX A6000" overhead="459.5 MiB" time=2025-02-20T16:32:59.017+08:00 level=INFO source=types.go:131 msg="inference compute" id=GPU-27ef0894-cb58-204e-0830-e9f6cf9a3c86 library=cuda variant=v12 compute=8.6 driver=12.4 name="NVIDIA RTX A6000" total="48.0 GiB" available="46.2 GiB" [GIN] 2025/02/20 - 16:32:59 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:32:59 | 200 | 0s | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:33:19 | 200 | 285.5µs | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:33:19 | 200 | 0s | 192.168.242.247 | GET "/api/ps" time=2025-02-20T16:33:20.954+08:00 level=INFO source=sched.go:714 msg="new model will fit in available VRAM in single GPU, loading" model=C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 gpu=GPU-27ef0894-cb58-204e-0830-e9f6cf9a3c86 parallel=4 available=49596465152 required="40.2 GiB" time=2025-02-20T16:33:20.964+08:00 level=INFO source=server.go:104 msg="system memory" total="127.8 GiB" free="106.2 GiB" free_swap="123.8 GiB" time=2025-02-20T16:33:20.966+08:00 level=INFO source=memory.go:356 msg="offload to cuda" layers.requested=-1 layers.model=65 layers.offload=65 layers.split="" memory.available="[46.2 GiB]" memory.gpu_overhead="0 B" memory.required.full="40.2 GiB" memory.required.partial="40.2 GiB" memory.required.kv="16.0 GiB" memory.required.allocations="[40.2 GiB]" memory.weights.total="33.5 GiB" memory.weights.repeating="32.9 GiB" memory.weights.nonrepeating="609.1 MiB" memory.graph.full="5.1 GiB" memory.graph.partial="6.4 GiB" time=2025-02-20T16:33:20.977+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="C:\\Users\\Administrator\\AppData\\Local\\Programs\\Ollama\\lib\\ollama\\runners\\cuda_v12_avx\\ollama_llama_server.exe runner --model C:\\Users\\Administrator\\.ollama\\models\\blobs\\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 --ctx-size 65536 --batch-size 512 --n-gpu-layers 65 --threads 64 --no-mmap --parallel 4 --port 55922" time=2025-02-20T16:33:21.022+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1 time=2025-02-20T16:33:21.022+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding" time=2025-02-20T16:33:21.022+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error" time=2025-02-20T16:33:21.107+08:00 level=INFO source=runner.go:936 msg="starting go runner" ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 CUDA devices: Device 0: NVIDIA RTX A6000, compute capability 8.6, VMM: yes time=2025-02-20T16:33:21.151+08:00 level=INFO source=runner.go:937 msg=system info="CUDA : ARCHS = 600,610,620,700,720,750,800,860,870,890,900 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | LLAMAFILE = 1 | AARCH64_REPACK = 1 | cgo(clang)" threads=64 time=2025-02-20T16:33:21.152+08:00 level=INFO source=.:0 msg="Server listening on 127.0.0.1:55922" time=2025-02-20T16:33:21.274+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server loading model" llama_load_model_from_file: using device CUDA0 (NVIDIA RTX A6000) - 47299 MiB free llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest)) llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. llama_model_loader: - kv 0: general.architecture str = qwen2 llama_model_loader: - kv 1: general.type str = model llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen llama_model_loader: - kv 4: general.size_label str = 32B llama_model_loader: - kv 5: qwen2.block_count u32 = 64 llama_model_loader: - kv 6: qwen2.context_length u32 = 131072 llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120 llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648 llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40 llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8 llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 13: general.file_type u32 = 15 llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646 llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de... llama_model_loader: - kv 25: general.quantization_version u32 = 2 llama_model_loader: - type f32: 321 tensors llama_model_loader: - type q4_K: 385 tensors llama_model_loader: - type q6_K: 65 tensors llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default' llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect llm_load_vocab: special tokens cache size = 22 llm_load_vocab: token to piece cache size = 0.9310 MB llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 152064 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: vocab_only = 0 llm_load_print_meta: n_ctx_train = 131072 llm_load_print_meta: n_embd = 5120 llm_load_print_meta: n_layer = 64 llm_load_print_meta: n_head = 40 llm_load_print_meta: n_head_kv = 8 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_swa = 0 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 5 llm_load_print_meta: n_embd_k_gqa = 1024 llm_load_print_meta: n_embd_v_gqa = 1024 llm_load_print_meta: f_norm_eps = 0.0e+00 llm_load_print_meta: f_norm_rms_eps = 1.0e-05 llm_load_print_meta: f_clamp_kqv = 0.0e+00 llm_load_print_meta: f_max_alibi_bias = 0.0e+00 llm_load_print_meta: f_logit_scale = 0.0e+00 llm_load_print_meta: n_ff = 27648 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 llm_load_print_meta: causal attn = 1 llm_load_print_meta: pooling type = 0 llm_load_print_meta: rope type = 2 llm_load_print_meta: rope scaling = linear llm_load_print_meta: freq_base_train = 1000000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_ctx_orig_yarn = 131072 llm_load_print_meta: rope_finetuned = unknown llm_load_print_meta: ssm_d_conv = 0 llm_load_print_meta: ssm_d_inner = 0 llm_load_print_meta: ssm_d_state = 0 llm_load_print_meta: ssm_dt_rank = 0 llm_load_print_meta: ssm_dt_b_c_rms = 0 llm_load_print_meta: model type = 32B llm_load_print_meta: model ftype = Q4_K - Medium llm_load_print_meta: model params = 32.76 B llm_load_print_meta: model size = 18.48 GiB (4.85 BPW) llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>' llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>' llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>' llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>' llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>' llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>' llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>' llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOG token = 151662 '<|fim_pad|>' llm_load_print_meta: EOG token = 151663 '<|repo_name|>' llm_load_print_meta: EOG token = 151664 '<|file_sep|>' llm_load_print_meta: max token length = 256 llm_load_tensors: offloading 64 repeating layers to GPU llm_load_tensors: offloading output layer to GPU llm_load_tensors: offloaded 65/65 layers to GPU llm_load_tensors: CPU model buffer size = 417.66 MiB llm_load_tensors: CUDA0 model buffer size = 18508.35 MiB llama_new_context_with_model: n_seq_max = 4 llama_new_context_with_model: n_ctx = 65536 llama_new_context_with_model: n_ctx_per_seq = 16384 llama_new_context_with_model: n_batch = 2048 llama_new_context_with_model: n_ubatch = 512 llama_new_context_with_model: flash_attn = 0 llama_new_context_with_model: freq_base = 1000000.0 llama_new_context_with_model: freq_scale = 1 llama_new_context_with_model: n_ctx_per_seq (16384) < n_ctx_train (131072) -- the full capacity of the model will not be utilized llama_kv_cache_init: kv_size = 65536, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 64, can_shift = 1 llama_kv_cache_init: CUDA0 KV buffer size = 16384.00 MiB llama_new_context_with_model: KV self size = 16384.00 MiB, K (f16): 8192.00 MiB, V (f16): 8192.00 MiB llama_new_context_with_model: CUDA_Host output buffer size = 2.40 MiB llama_new_context_with_model: CUDA0 compute buffer size = 5288.00 MiB llama_new_context_with_model: CUDA_Host compute buffer size = 138.01 MiB llama_new_context_with_model: graph nodes = 2246 llama_new_context_with_model: graph splits = 2 time=2025-02-20T16:33:25.286+08:00 level=INFO source=server.go:594 msg="llama runner started in 4.26 seconds" [GIN] 2025/02/20 - 16:33:25 | 200 | 4.3991734s | 192.168.242.247 | POST "/api/generate" [GIN] 2025/02/20 - 16:33:51 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:33:51 | 200 | 0s | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:33:53 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:33:53 | 200 | 0s | 192.168.242.247 | GET "/api/ps" time=2025-02-20T16:34:06.948+08:00 level=INFO source=sched.go:714 msg="new model will fit in available VRAM in single GPU, loading" model=C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 gpu=GPU-27ef0894-cb58-204e-0830-e9f6cf9a3c86 parallel=4 available=49588076544 required="21.5 GiB" time=2025-02-20T16:34:06.958+08:00 level=INFO source=server.go:104 msg="system memory" total="127.8 GiB" free="106.3 GiB" free_swap="123.8 GiB" time=2025-02-20T16:34:06.959+08:00 level=INFO source=memory.go:356 msg="offload to cuda" layers.requested=-1 layers.model=65 layers.offload=65 layers.split="" memory.available="[46.2 GiB]" memory.gpu_overhead="0 B" memory.required.full="21.5 GiB" memory.required.partial="21.5 GiB" memory.required.kv="2.0 GiB" memory.required.allocations="[21.5 GiB]" memory.weights.total="19.5 GiB" memory.weights.repeating="18.9 GiB" memory.weights.nonrepeating="609.1 MiB" memory.graph.full="676.0 MiB" memory.graph.partial="916.1 MiB" time=2025-02-20T16:34:06.964+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="C:\\Users\\Administrator\\AppData\\Local\\Programs\\Ollama\\lib\\ollama\\runners\\cuda_v12_avx\\ollama_llama_server.exe runner --model C:\\Users\\Administrator\\.ollama\\models\\blobs\\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 --ctx-size 8192 --batch-size 512 --n-gpu-layers 65 --threads 64 --no-mmap --parallel 4 --port 55954" time=2025-02-20T16:34:07.010+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1 time=2025-02-20T16:34:07.010+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding" time=2025-02-20T16:34:07.010+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error" time=2025-02-20T16:34:07.094+08:00 level=INFO source=runner.go:936 msg="starting go runner" ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 CUDA devices: Device 0: NVIDIA RTX A6000, compute capability 8.6, VMM: yes time=2025-02-20T16:34:07.149+08:00 level=INFO source=runner.go:937 msg=system info="CUDA : ARCHS = 600,610,620,700,720,750,800,860,870,890,900 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | LLAMAFILE = 1 | AARCH64_REPACK = 1 | cgo(clang)" threads=64 time=2025-02-20T16:34:07.150+08:00 level=INFO source=.:0 msg="Server listening on 127.0.0.1:55954" time=2025-02-20T16:34:07.263+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server loading model" llama_load_model_from_file: using device CUDA0 (NVIDIA RTX A6000) - 47299 MiB free llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest)) llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. llama_model_loader: - kv 0: general.architecture str = qwen2 llama_model_loader: - kv 1: general.type str = model llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen llama_model_loader: - kv 4: general.size_label str = 32B llama_model_loader: - kv 5: qwen2.block_count u32 = 64 llama_model_loader: - kv 6: qwen2.context_length u32 = 131072 llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120 llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648 llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40 llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8 llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 13: general.file_type u32 = 15 llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646 llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de... llama_model_loader: - kv 25: general.quantization_version u32 = 2 llama_model_loader: - type f32: 321 tensors llama_model_loader: - type q4_K: 385 tensors llama_model_loader: - type q6_K: 65 tensors llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default' llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect llm_load_vocab: special tokens cache size = 22 llm_load_vocab: token to piece cache size = 0.9310 MB llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 152064 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: vocab_only = 0 llm_load_print_meta: n_ctx_train = 131072 llm_load_print_meta: n_embd = 5120 llm_load_print_meta: n_layer = 64 llm_load_print_meta: n_head = 40 llm_load_print_meta: n_head_kv = 8 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_swa = 0 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 5 llm_load_print_meta: n_embd_k_gqa = 1024 llm_load_print_meta: n_embd_v_gqa = 1024 llm_load_print_meta: f_norm_eps = 0.0e+00 llm_load_print_meta: f_norm_rms_eps = 1.0e-05 llm_load_print_meta: f_clamp_kqv = 0.0e+00 llm_load_print_meta: f_max_alibi_bias = 0.0e+00 llm_load_print_meta: f_logit_scale = 0.0e+00 llm_load_print_meta: n_ff = 27648 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 llm_load_print_meta: causal attn = 1 llm_load_print_meta: pooling type = 0 llm_load_print_meta: rope type = 2 llm_load_print_meta: rope scaling = linear llm_load_print_meta: freq_base_train = 1000000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_ctx_orig_yarn = 131072 llm_load_print_meta: rope_finetuned = unknown llm_load_print_meta: ssm_d_conv = 0 llm_load_print_meta: ssm_d_inner = 0 llm_load_print_meta: ssm_d_state = 0 llm_load_print_meta: ssm_dt_rank = 0 llm_load_print_meta: ssm_dt_b_c_rms = 0 llm_load_print_meta: model type = 32B llm_load_print_meta: model ftype = Q4_K - Medium llm_load_print_meta: model params = 32.76 B llm_load_print_meta: model size = 18.48 GiB (4.85 BPW) llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>' llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>' llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>' llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>' llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>' llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>' llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>' llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOG token = 151662 '<|fim_pad|>' llm_load_print_meta: EOG token = 151663 '<|repo_name|>' llm_load_print_meta: EOG token = 151664 '<|file_sep|>' llm_load_print_meta: max token length = 256 llm_load_tensors: offloading 64 repeating layers to GPU llm_load_tensors: offloading output layer to GPU llm_load_tensors: offloaded 65/65 layers to GPU llm_load_tensors: CPU model buffer size = 417.66 MiB llm_load_tensors: CUDA0 model buffer size = 18508.35 MiB llama_new_context_with_model: n_seq_max = 4 llama_new_context_with_model: n_ctx = 8192 llama_new_context_with_model: n_ctx_per_seq = 2048 llama_new_context_with_model: n_batch = 2048 llama_new_context_with_model: n_ubatch = 512 llama_new_context_with_model: flash_attn = 0 llama_new_context_with_model: freq_base = 1000000.0 llama_new_context_with_model: freq_scale = 1 llama_new_context_with_model: n_ctx_per_seq (2048) < n_ctx_train (131072) -- the full capacity of the model will not be utilized llama_kv_cache_init: kv_size = 8192, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 64, can_shift = 1 llama_kv_cache_init: CUDA0 KV buffer size = 2048.00 MiB llama_new_context_with_model: KV self size = 2048.00 MiB, K (f16): 1024.00 MiB, V (f16): 1024.00 MiB llama_new_context_with_model: CUDA_Host output buffer size = 2.40 MiB llama_new_context_with_model: CUDA0 compute buffer size = 696.00 MiB llama_new_context_with_model: CUDA_Host compute buffer size = 26.01 MiB llama_new_context_with_model: graph nodes = 2246 llama_new_context_with_model: graph splits = 2 time=2025-02-20T16:34:11.025+08:00 level=INFO source=server.go:594 msg="llama runner started in 4.01 seconds" llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest)) llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. llama_model_loader: - kv 0: general.architecture str = qwen2 llama_model_loader: - kv 1: general.type str = model llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen llama_model_loader: - kv 4: general.size_label str = 32B llama_model_loader: - kv 5: qwen2.block_count u32 = 64 llama_model_loader: - kv 6: qwen2.context_length u32 = 131072 llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120 llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648 llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40 llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8 llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 13: general.file_type u32 = 15 llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646 llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de... llama_model_loader: - kv 25: general.quantization_version u32 = 2 llama_model_loader: - type f32: 321 tensors llama_model_loader: - type q4_K: 385 tensors llama_model_loader: - type q6_K: 65 tensors llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default' llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect llm_load_vocab: special tokens cache size = 22 llm_load_vocab: token to piece cache size = 0.9310 MB llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 152064 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: vocab_only = 1 llm_load_print_meta: model type = ?B llm_load_print_meta: model ftype = all F32 llm_load_print_meta: model params = 32.76 B llm_load_print_meta: model size = 18.48 GiB (4.85 BPW) llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>' llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>' llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>' llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>' llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>' llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>' llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>' llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOG token = 151662 '<|fim_pad|>' llm_load_print_meta: EOG token = 151663 '<|repo_name|>' llm_load_print_meta: EOG token = 151664 '<|file_sep|>' llm_load_print_meta: max token length = 256 llama_model_load: vocab only - skipping tensors time=2025-02-20T16:34:11.359+08:00 level=WARN source=runner.go:129 msg="truncating input prompt" limit=2048 prompt=3963 keep=5 new=2048 [GIN] 2025/02/20 - 16:34:14 | 200 | 7.6964767s | 192.168.242.247 | POST "/api/chat" [GIN] 2025/02/20 - 16:34:28 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:34:28 | 200 | 0s | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:34:29 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:34:29 | 200 | 0s | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:34:30 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:34:30 | 200 | 0s | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:34:30 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:34:30 | 200 | 0s | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:36:08 | 200 | 156.8µs | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:36:08 | 200 | 1.0341ms | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:36:32 | 200 | 7.4467209s | 192.168.242.247 | POST "/api/chat" [GIN] 2025/02/20 - 16:38:08 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:38:08 | 200 | 611.4µs | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:38:09 | 200 | 208.6µs | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:38:09 | 200 | 0s | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:38:10 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:38:10 | 200 | 0s | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:38:29 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:38:29 | 200 | 0s | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:38:30 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:38:30 | 200 | 0s | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:41:04 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:41:04 | 200 | 20.5651ms | 192.168.242.247 | POST "/api/show" time=2025-02-20T16:42:07.019+08:00 level=INFO source=sched.go:714 msg="new model will fit in available VRAM in single GPU, loading" model=C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 gpu=GPU-27ef0894-cb58-204e-0830-e9f6cf9a3c86 parallel=4 available=49617698816 required="21.5 GiB" time=2025-02-20T16:42:07.030+08:00 level=INFO source=server.go:104 msg="system memory" total="127.8 GiB" free="106.3 GiB" free_swap="123.9 GiB" time=2025-02-20T16:42:07.031+08:00 level=INFO source=memory.go:356 msg="offload to cuda" layers.requested=-1 layers.model=65 layers.offload=65 layers.split="" memory.available="[46.2 GiB]" memory.gpu_overhead="0 B" memory.required.full="21.5 GiB" memory.required.partial="21.5 GiB" memory.required.kv="2.0 GiB" memory.required.allocations="[21.5 GiB]" memory.weights.total="19.5 GiB" memory.weights.repeating="18.9 GiB" memory.weights.nonrepeating="609.1 MiB" memory.graph.full="676.0 MiB" memory.graph.partial="916.1 MiB" time=2025-02-20T16:42:07.036+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="C:\\Users\\Administrator\\AppData\\Local\\Programs\\Ollama\\lib\\ollama\\runners\\cuda_v12_avx\\ollama_llama_server.exe runner --model C:\\Users\\Administrator\\.ollama\\models\\blobs\\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 --ctx-size 8192 --batch-size 512 --n-gpu-layers 65 --threads 64 --no-mmap --parallel 4 --port 56318" time=2025-02-20T16:42:07.080+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1 time=2025-02-20T16:42:07.080+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding" time=2025-02-20T16:42:07.080+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error" time=2025-02-20T16:42:07.164+08:00 level=INFO source=runner.go:936 msg="starting go runner" ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 CUDA devices: Device 0: NVIDIA RTX A6000, compute capability 8.6, VMM: yes time=2025-02-20T16:42:07.210+08:00 level=INFO source=runner.go:937 msg=system info="CUDA : ARCHS = 600,610,620,700,720,750,800,860,870,890,900 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | LLAMAFILE = 1 | AARCH64_REPACK = 1 | cgo(clang)" threads=64 time=2025-02-20T16:42:07.212+08:00 level=INFO source=.:0 msg="Server listening on 127.0.0.1:56318" time=2025-02-20T16:42:07.331+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server loading model" llama_load_model_from_file: using device CUDA0 (NVIDIA RTX A6000) - 47299 MiB free llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest)) llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. llama_model_loader: - kv 0: general.architecture str = qwen2 llama_model_loader: - kv 1: general.type str = model llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen llama_model_loader: - kv 4: general.size_label str = 32B llama_model_loader: - kv 5: qwen2.block_count u32 = 64 llama_model_loader: - kv 6: qwen2.context_length u32 = 131072 llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120 llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648 llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40 llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8 llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 13: general.file_type u32 = 15 llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646 llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de... llama_model_loader: - kv 25: general.quantization_version u32 = 2 llama_model_loader: - type f32: 321 tensors llama_model_loader: - type q4_K: 385 tensors llama_model_loader: - type q6_K: 65 tensors llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default' llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect llm_load_vocab: special tokens cache size = 22 llm_load_vocab: token to piece cache size = 0.9310 MB llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 152064 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: vocab_only = 0 llm_load_print_meta: n_ctx_train = 131072 llm_load_print_meta: n_embd = 5120 llm_load_print_meta: n_layer = 64 llm_load_print_meta: n_head = 40 llm_load_print_meta: n_head_kv = 8 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_swa = 0 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 5 llm_load_print_meta: n_embd_k_gqa = 1024 llm_load_print_meta: n_embd_v_gqa = 1024 llm_load_print_meta: f_norm_eps = 0.0e+00 llm_load_print_meta: f_norm_rms_eps = 1.0e-05 llm_load_print_meta: f_clamp_kqv = 0.0e+00 llm_load_print_meta: f_max_alibi_bias = 0.0e+00 llm_load_print_meta: f_logit_scale = 0.0e+00 llm_load_print_meta: n_ff = 27648 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 llm_load_print_meta: causal attn = 1 llm_load_print_meta: pooling type = 0 llm_load_print_meta: rope type = 2 llm_load_print_meta: rope scaling = linear llm_load_print_meta: freq_base_train = 1000000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_ctx_orig_yarn = 131072 llm_load_print_meta: rope_finetuned = unknown llm_load_print_meta: ssm_d_conv = 0 llm_load_print_meta: ssm_d_inner = 0 llm_load_print_meta: ssm_d_state = 0 llm_load_print_meta: ssm_dt_rank = 0 llm_load_print_meta: ssm_dt_b_c_rms = 0 llm_load_print_meta: model type = 32B llm_load_print_meta: model ftype = Q4_K - Medium llm_load_print_meta: model params = 32.76 B llm_load_print_meta: model size = 18.48 GiB (4.85 BPW) llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>' llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>' llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>' llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>' llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>' llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>' llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>' llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOG token = 151662 '<|fim_pad|>' llm_load_print_meta: EOG token = 151663 '<|repo_name|>' llm_load_print_meta: EOG token = 151664 '<|file_sep|>' llm_load_print_meta: max token length = 256 llm_load_tensors: offloading 64 repeating layers to GPU llm_load_tensors: offloading output layer to GPU llm_load_tensors: offloaded 65/65 layers to GPU llm_load_tensors: CPU model buffer size = 417.66 MiB llm_load_tensors: CUDA0 model buffer size = 18508.35 MiB llama_new_context_with_model: n_seq_max = 4 llama_new_context_with_model: n_ctx = 8192 llama_new_context_with_model: n_ctx_per_seq = 2048 llama_new_context_with_model: n_batch = 2048 llama_new_context_with_model: n_ubatch = 512 llama_new_context_with_model: flash_attn = 0 llama_new_context_with_model: freq_base = 1000000.0 llama_new_context_with_model: freq_scale = 1 llama_new_context_with_model: n_ctx_per_seq (2048) < n_ctx_train (131072) -- the full capacity of the model will not be utilized llama_kv_cache_init: kv_size = 8192, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 64, can_shift = 1 llama_kv_cache_init: CUDA0 KV buffer size = 2048.00 MiB llama_new_context_with_model: KV self size = 2048.00 MiB, K (f16): 1024.00 MiB, V (f16): 1024.00 MiB llama_new_context_with_model: CUDA_Host output buffer size = 2.40 MiB llama_new_context_with_model: CUDA0 compute buffer size = 696.00 MiB llama_new_context_with_model: CUDA_Host compute buffer size = 26.01 MiB llama_new_context_with_model: graph nodes = 2246 llama_new_context_with_model: graph splits = 2 time=2025-02-20T16:42:11.094+08:00 level=INFO source=server.go:594 msg="llama runner started in 4.01 seconds" llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest)) llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. llama_model_loader: - kv 0: general.architecture str = qwen2 llama_model_loader: - kv 1: general.type str = model llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen llama_model_loader: - kv 4: general.size_label str = 32B llama_model_loader: - kv 5: qwen2.block_count u32 = 64 llama_model_loader: - kv 6: qwen2.context_length u32 = 131072 llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120 llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648 llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40 llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8 llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 13: general.file_type u32 = 15 llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646 llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de... llama_model_loader: - kv 25: general.quantization_version u32 = 2 llama_model_loader: - type f32: 321 tensors llama_model_loader: - type q4_K: 385 tensors llama_model_loader: - type q6_K: 65 tensors llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default' llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect llm_load_vocab: special tokens cache size = 22 llm_load_vocab: token to piece cache size = 0.9310 MB llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 152064 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: vocab_only = 1 llm_load_print_meta: model type = ?B llm_load_print_meta: model ftype = all F32 llm_load_print_meta: model params = 32.76 B llm_load_print_meta: model size = 18.48 GiB (4.85 BPW) llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>' llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>' llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>' llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>' llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>' llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>' llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>' llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOG token = 151662 '<|fim_pad|>' llm_load_print_meta: EOG token = 151663 '<|repo_name|>' llm_load_print_meta: EOG token = 151664 '<|file_sep|>' llm_load_print_meta: max token length = 256 llama_model_load: vocab only - skipping tensors [GIN] 2025/02/20 - 16:42:36 | 200 | 29.3566393s | 192.168.242.247 | POST "/api/chat" [GIN] 2025/02/20 - 16:43:10 | 200 | 31.1044914s | 192.168.242.247 | POST "/api/chat" [GIN] 2025/02/20 - 16:44:45 | 200 | 35.5680529s | 192.168.242.247 | POST "/api/chat" time=2025-02-20T16:49:29.342+08:00 level=INFO source=sched.go:714 msg="new model will fit in available VRAM in single GPU, loading" model=C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 gpu=GPU-27ef0894-cb58-204e-0830-e9f6cf9a3c86 parallel=4 available=49609834496 required="40.2 GiB" time=2025-02-20T16:49:29.354+08:00 level=INFO source=server.go:104 msg="system memory" total="127.8 GiB" free="106.3 GiB" free_swap="123.8 GiB" time=2025-02-20T16:49:29.356+08:00 level=INFO source=memory.go:356 msg="offload to cuda" layers.requested=-1 layers.model=65 layers.offload=65 layers.split="" memory.available="[46.2 GiB]" memory.gpu_overhead="0 B" memory.required.full="40.2 GiB" memory.required.partial="40.2 GiB" memory.required.kv="16.0 GiB" memory.required.allocations="[40.2 GiB]" memory.weights.total="33.5 GiB" memory.weights.repeating="32.9 GiB" memory.weights.nonrepeating="609.1 MiB" memory.graph.full="5.1 GiB" memory.graph.partial="6.4 GiB" time=2025-02-20T16:49:29.360+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="C:\\Users\\Administrator\\AppData\\Local\\Programs\\Ollama\\lib\\ollama\\runners\\cuda_v12_avx\\ollama_llama_server.exe runner --model C:\\Users\\Administrator\\.ollama\\models\\blobs\\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 --ctx-size 65536 --batch-size 512 --n-gpu-layers 65 --threads 64 --no-mmap --parallel 4 --port 56698" time=2025-02-20T16:49:29.403+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1 time=2025-02-20T16:49:29.403+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding" time=2025-02-20T16:49:29.403+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error" time=2025-02-20T16:49:29.493+08:00 level=INFO source=runner.go:936 msg="starting go runner" ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 CUDA devices: Device 0: NVIDIA RTX A6000, compute capability 8.6, VMM: yes time=2025-02-20T16:49:29.547+08:00 level=INFO source=runner.go:937 msg=system info="CUDA : ARCHS = 600,610,620,700,720,750,800,860,870,890,900 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | LLAMAFILE = 1 | AARCH64_REPACK = 1 | cgo(clang)" threads=64 time=2025-02-20T16:49:29.548+08:00 level=INFO source=.:0 msg="Server listening on 127.0.0.1:56698" time=2025-02-20T16:49:29.656+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server loading model" llama_load_model_from_file: using device CUDA0 (NVIDIA RTX A6000) - 47299 MiB free llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest)) llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. llama_model_loader: - kv 0: general.architecture str = qwen2 llama_model_loader: - kv 1: general.type str = model llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen llama_model_loader: - kv 4: general.size_label str = 32B llama_model_loader: - kv 5: qwen2.block_count u32 = 64 llama_model_loader: - kv 6: qwen2.context_length u32 = 131072 llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120 llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648 llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40 llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8 llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 13: general.file_type u32 = 15 llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646 llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de... llama_model_loader: - kv 25: general.quantization_version u32 = 2 llama_model_loader: - type f32: 321 tensors llama_model_loader: - type q4_K: 385 tensors llama_model_loader: - type q6_K: 65 tensors llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default' llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect llm_load_vocab: special tokens cache size = 22 llm_load_vocab: token to piece cache size = 0.9310 MB llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 152064 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: vocab_only = 0 llm_load_print_meta: n_ctx_train = 131072 llm_load_print_meta: n_embd = 5120 llm_load_print_meta: n_layer = 64 llm_load_print_meta: n_head = 40 llm_load_print_meta: n_head_kv = 8 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_swa = 0 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 5 llm_load_print_meta: n_embd_k_gqa = 1024 llm_load_print_meta: n_embd_v_gqa = 1024 llm_load_print_meta: f_norm_eps = 0.0e+00 llm_load_print_meta: f_norm_rms_eps = 1.0e-05 llm_load_print_meta: f_clamp_kqv = 0.0e+00 llm_load_print_meta: f_max_alibi_bias = 0.0e+00 llm_load_print_meta: f_logit_scale = 0.0e+00 llm_load_print_meta: n_ff = 27648 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 llm_load_print_meta: causal attn = 1 llm_load_print_meta: pooling type = 0 llm_load_print_meta: rope type = 2 llm_load_print_meta: rope scaling = linear llm_load_print_meta: freq_base_train = 1000000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_ctx_orig_yarn = 131072 llm_load_print_meta: rope_finetuned = unknown llm_load_print_meta: ssm_d_conv = 0 llm_load_print_meta: ssm_d_inner = 0 llm_load_print_meta: ssm_d_state = 0 llm_load_print_meta: ssm_dt_rank = 0 llm_load_print_meta: ssm_dt_b_c_rms = 0 llm_load_print_meta: model type = 32B llm_load_print_meta: model ftype = Q4_K - Medium llm_load_print_meta: model params = 32.76 B llm_load_print_meta: model size = 18.48 GiB (4.85 BPW) llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>' llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>' llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>' llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>' llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>' llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>' llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>' llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOG token = 151662 '<|fim_pad|>' llm_load_print_meta: EOG token = 151663 '<|repo_name|>' llm_load_print_meta: EOG token = 151664 '<|file_sep|>' llm_load_print_meta: max token length = 256 llm_load_tensors: offloading 64 repeating layers to GPU llm_load_tensors: offloading output layer to GPU llm_load_tensors: offloaded 65/65 layers to GPU llm_load_tensors: CPU model buffer size = 417.66 MiB llm_load_tensors: CUDA0 model buffer size = 18508.35 MiB time=2025-02-20T16:49:30.659+08:00 level=WARN source=server.go:562 msg="client connection closed before server finished loading, aborting load" time=2025-02-20T16:49:30.659+08:00 level=ERROR source=sched.go:455 msg="error loading llama server" error="timed out waiting for llama runner to start: context canceled" [GIN] 2025/02/20 - 16:49:30 | 499 | 1.6502837s | 192.168.242.247 | POST "/api/generate" time=2025-02-20T16:49:35.670+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.0106594 model=C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 time=2025-02-20T16:49:35.919+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.2599917 model=C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 time=2025-02-20T16:49:36.169+08:00 level=WARN source=sched.go:646 msg="gpu VRAM usage didn't recover within timeout" seconds=5.5101071 model=C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 time=2025-02-20T16:50:49.357+08:00 level=INFO source=sched.go:714 msg="new model will fit in available VRAM in single GPU, loading" model=C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 gpu=GPU-27ef0894-cb58-204e-0830-e9f6cf9a3c86 parallel=4 available=49607868416 required="21.5 GiB" time=2025-02-20T16:50:49.367+08:00 level=INFO source=server.go:104 msg="system memory" total="127.8 GiB" free="106.2 GiB" free_swap="123.8 GiB" time=2025-02-20T16:50:49.369+08:00 level=INFO source=memory.go:356 msg="offload to cuda" layers.requested=-1 layers.model=65 layers.offload=65 layers.split="" memory.available="[46.2 GiB]" memory.gpu_overhead="0 B" memory.required.full="21.5 GiB" memory.required.partial="21.5 GiB" memory.required.kv="2.0 GiB" memory.required.allocations="[21.5 GiB]" memory.weights.total="19.5 GiB" memory.weights.repeating="18.9 GiB" memory.weights.nonrepeating="609.1 MiB" memory.graph.full="676.0 MiB" memory.graph.partial="916.1 MiB" time=2025-02-20T16:50:49.373+08:00 level=INFO source=server.go:376 msg="starting llama server" cmd="C:\\Users\\Administrator\\AppData\\Local\\Programs\\Ollama\\lib\\ollama\\runners\\cuda_v12_avx\\ollama_llama_server.exe runner --model C:\\Users\\Administrator\\.ollama\\models\\blobs\\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 --ctx-size 8192 --batch-size 512 --n-gpu-layers 65 --threads 64 --no-mmap --parallel 4 --port 56884" time=2025-02-20T16:50:49.417+08:00 level=INFO source=sched.go:449 msg="loaded runners" count=1 time=2025-02-20T16:50:49.417+08:00 level=INFO source=server.go:555 msg="waiting for llama runner to start responding" time=2025-02-20T16:50:49.417+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server error" time=2025-02-20T16:50:49.509+08:00 level=INFO source=runner.go:936 msg="starting go runner" ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 CUDA devices: Device 0: NVIDIA RTX A6000, compute capability 8.6, VMM: yes time=2025-02-20T16:50:49.556+08:00 level=INFO source=runner.go:937 msg=system info="CUDA : ARCHS = 600,610,620,700,720,750,800,860,870,890,900 | USE_GRAPHS = 1 | PEER_MAX_BATCH_SIZE = 128 | CPU : SSE3 = 1 | SSSE3 = 1 | AVX = 1 | LLAMAFILE = 1 | AARCH64_REPACK = 1 | cgo(clang)" threads=64 time=2025-02-20T16:50:49.558+08:00 level=INFO source=.:0 msg="Server listening on 127.0.0.1:56884" time=2025-02-20T16:50:49.670+08:00 level=INFO source=server.go:589 msg="waiting for server to become available" status="llm server loading model" llama_load_model_from_file: using device CUDA0 (NVIDIA RTX A6000) - 47299 MiB free llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest)) llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. llama_model_loader: - kv 0: general.architecture str = qwen2 llama_model_loader: - kv 1: general.type str = model llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen llama_model_loader: - kv 4: general.size_label str = 32B llama_model_loader: - kv 5: qwen2.block_count u32 = 64 llama_model_loader: - kv 6: qwen2.context_length u32 = 131072 llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120 llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648 llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40 llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8 llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 13: general.file_type u32 = 15 llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646 llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de... llama_model_loader: - kv 25: general.quantization_version u32 = 2 llama_model_loader: - type f32: 321 tensors llama_model_loader: - type q4_K: 385 tensors llama_model_loader: - type q6_K: 65 tensors llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default' llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect llm_load_vocab: special tokens cache size = 22 llm_load_vocab: token to piece cache size = 0.9310 MB llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 152064 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: vocab_only = 0 llm_load_print_meta: n_ctx_train = 131072 llm_load_print_meta: n_embd = 5120 llm_load_print_meta: n_layer = 64 llm_load_print_meta: n_head = 40 llm_load_print_meta: n_head_kv = 8 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_swa = 0 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 5 llm_load_print_meta: n_embd_k_gqa = 1024 llm_load_print_meta: n_embd_v_gqa = 1024 llm_load_print_meta: f_norm_eps = 0.0e+00 llm_load_print_meta: f_norm_rms_eps = 1.0e-05 llm_load_print_meta: f_clamp_kqv = 0.0e+00 llm_load_print_meta: f_max_alibi_bias = 0.0e+00 llm_load_print_meta: f_logit_scale = 0.0e+00 llm_load_print_meta: n_ff = 27648 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 llm_load_print_meta: causal attn = 1 llm_load_print_meta: pooling type = 0 llm_load_print_meta: rope type = 2 llm_load_print_meta: rope scaling = linear llm_load_print_meta: freq_base_train = 1000000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_ctx_orig_yarn = 131072 llm_load_print_meta: rope_finetuned = unknown llm_load_print_meta: ssm_d_conv = 0 llm_load_print_meta: ssm_d_inner = 0 llm_load_print_meta: ssm_d_state = 0 llm_load_print_meta: ssm_dt_rank = 0 llm_load_print_meta: ssm_dt_b_c_rms = 0 llm_load_print_meta: model type = 32B llm_load_print_meta: model ftype = Q4_K - Medium llm_load_print_meta: model params = 32.76 B llm_load_print_meta: model size = 18.48 GiB (4.85 BPW) llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>' llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>' llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>' llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>' llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>' llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>' llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>' llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOG token = 151662 '<|fim_pad|>' llm_load_print_meta: EOG token = 151663 '<|repo_name|>' llm_load_print_meta: EOG token = 151664 '<|file_sep|>' llm_load_print_meta: max token length = 256 llm_load_tensors: offloading 64 repeating layers to GPU llm_load_tensors: offloading output layer to GPU llm_load_tensors: offloaded 65/65 layers to GPU llm_load_tensors: CPU model buffer size = 417.66 MiB llm_load_tensors: CUDA0 model buffer size = 18508.35 MiB llama_new_context_with_model: n_seq_max = 4 llama_new_context_with_model: n_ctx = 8192 llama_new_context_with_model: n_ctx_per_seq = 2048 llama_new_context_with_model: n_batch = 2048 llama_new_context_with_model: n_ubatch = 512 llama_new_context_with_model: flash_attn = 0 llama_new_context_with_model: freq_base = 1000000.0 llama_new_context_with_model: freq_scale = 1 llama_new_context_with_model: n_ctx_per_seq (2048) < n_ctx_train (131072) -- the full capacity of the model will not be utilized llama_kv_cache_init: kv_size = 8192, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 64, can_shift = 1 llama_kv_cache_init: CUDA0 KV buffer size = 2048.00 MiB llama_new_context_with_model: KV self size = 2048.00 MiB, K (f16): 1024.00 MiB, V (f16): 1024.00 MiB llama_new_context_with_model: CUDA_Host output buffer size = 2.40 MiB llama_new_context_with_model: CUDA0 compute buffer size = 696.00 MiB llama_new_context_with_model: CUDA_Host compute buffer size = 26.01 MiB llama_new_context_with_model: graph nodes = 2246 llama_new_context_with_model: graph splits = 2 time=2025-02-20T16:50:53.429+08:00 level=INFO source=server.go:594 msg="llama runner started in 4.01 seconds" llama_model_loader: loaded meta data with 26 key-value pairs and 771 tensors from C:\Users\Administrator\.ollama\models\blobs\sha256-6150cb382311b69f09cc0f9a1b69fc029cbd742b66bb8ec531aa5ecf5c613e93 (version GGUF V3 (latest)) llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. llama_model_loader: - kv 0: general.architecture str = qwen2 llama_model_loader: - kv 1: general.type str = model llama_model_loader: - kv 2: general.name str = DeepSeek R1 Distill Qwen 32B llama_model_loader: - kv 3: general.basename str = DeepSeek-R1-Distill-Qwen llama_model_loader: - kv 4: general.size_label str = 32B llama_model_loader: - kv 5: qwen2.block_count u32 = 64 llama_model_loader: - kv 6: qwen2.context_length u32 = 131072 llama_model_loader: - kv 7: qwen2.embedding_length u32 = 5120 llama_model_loader: - kv 8: qwen2.feed_forward_length u32 = 27648 llama_model_loader: - kv 9: qwen2.attention.head_count u32 = 40 llama_model_loader: - kv 10: qwen2.attention.head_count_kv u32 = 8 llama_model_loader: - kv 11: qwen2.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 12: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 13: general.file_type u32 = 15 llama_model_loader: - kv 14: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 15: tokenizer.ggml.pre str = deepseek-r1-qwen llama_model_loader: - kv 16: tokenizer.ggml.tokens arr[str,152064] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,152064] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 151646 llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 151643 llama_model_loader: - kv 21: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false llama_model_loader: - kv 24: tokenizer.chat_template str = {% if not add_generation_prompt is de... llama_model_loader: - kv 25: general.quantization_version u32 = 2 llama_model_loader: - type f32: 321 tensors llama_model_loader: - type q4_K: 385 tensors llama_model_loader: - type q6_K: 65 tensors llm_load_vocab: missing or unrecognized pre-tokenizer type, using: 'default' llm_load_vocab: special_eos_id is not in special_eog_ids - the tokenizer config may be incorrect llm_load_vocab: special tokens cache size = 22 llm_load_vocab: token to piece cache size = 0.9310 MB llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = qwen2 llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 152064 llm_load_print_meta: n_merges = 151387 llm_load_print_meta: vocab_only = 1 llm_load_print_meta: model type = ?B llm_load_print_meta: model ftype = all F32 llm_load_print_meta: model params = 32.76 B llm_load_print_meta: model size = 18.48 GiB (4.85 BPW) llm_load_print_meta: general.name = DeepSeek R1 Distill Qwen 32B llm_load_print_meta: BOS token = 151646 '<｜begin▁of▁sentence｜>' llm_load_print_meta: EOS token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOT token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: PAD token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: LF token = 148848 'ÄĬ' llm_load_print_meta: FIM PRE token = 151659 '<|fim_prefix|>' llm_load_print_meta: FIM SUF token = 151661 '<|fim_suffix|>' llm_load_print_meta: FIM MID token = 151660 '<|fim_middle|>' llm_load_print_meta: FIM PAD token = 151662 '<|fim_pad|>' llm_load_print_meta: FIM REP token = 151663 '<|repo_name|>' llm_load_print_meta: FIM SEP token = 151664 '<|file_sep|>' llm_load_print_meta: EOG token = 151643 '<｜end▁of▁sentence｜>' llm_load_print_meta: EOG token = 151662 '<|fim_pad|>' llm_load_print_meta: EOG token = 151663 '<|repo_name|>' llm_load_print_meta: EOG token = 151664 '<|file_sep|>' llm_load_print_meta: max token length = 256 llama_model_load: vocab only - skipping tensors time=2025-02-20T16:50:53.705+08:00 level=WARN source=runner.go:129 msg="truncating input prompt" limit=2048 prompt=3085 keep=5 new=2048 [GIN] 2025/02/20 - 16:51:35 | 200 | 45.8058413s | 192.168.242.247 | POST "/api/chat" [GIN] 2025/02/20 - 16:51:48 | 200 | 26.03855s | 192.168.242.247 | POST "/api/chat" [GIN] 2025/02/20 - 16:52:15 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:52:15 | 200 | 0s | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:52:16 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:52:16 | 200 | 507.4µs | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:52:17 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:52:17 | 200 | 0s | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:52:17 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:52:17 | 200 | 0s | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:52:18 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:52:18 | 200 | 0s | 192.168.242.247 | GET "/api/ps" [GIN] 2025/02/20 - 16:52:18 | 200 | 0s | 192.168.242.247 | HEAD "/" [GIN] 2025/02/20 - 16:52:18 | 200 | 0s | 192.168.242.247 | GET "/api/ps"
GiteaMirror commented 2026-04-22 12:32:13 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Feb 20, 2025):

If you are using the ollama CLI to chat, do this:

C:\> ollama run deepseek-r1:32b --keepalive -1m
>>> /set parameter num_ctx 8192
Set parameter 'num_ctx' to '8192'
>>> hello
<think>

</think>

Hello! How can I assist you today? 😊

>>> /bye

If you are using a different client, you need to change its configuration to set num_ctx and keep_alive.

<!-- gh-comment-id:2671001073 --> @rick-github commented on GitHub (Feb 20, 2025): If you are using the ollama CLI to chat, do this: ```console C:\> ollama run deepseek-r1:32b --keepalive -1m >>> /set parameter num_ctx 8192 Set parameter 'num_ctx' to '8192' >>> hello <think> </think> Hello! How can I assist you today? 😊 >>> /bye ``` If you are using a different client, you need to change its configuration to set `num_ctx` and `keep_alive`.
GiteaMirror commented 2026-04-22 12:32:15 -05:00
Author
Owner
Copy Link

@948024326 commented on GitHub (Feb 20, 2025):

If you are using the ollama CLI to chat, do this:

C:> ollama run deepseek-r1:32b --keepalive -1m

/set parameter num_ctx 8192
Set parameter 'num_ctx' to '8192'
hello

Hello! How can I assist you today? 😊

/bye
If you are using a different client, you need to change its configuration to set num_ctx and keep_alive.

thanks for your answer, i use ollama and dify, via API method,

"you need to change its configuration to set num_ctx and keep_alive" how to do this? thanks a lot!

<!-- gh-comment-id:2671013823 --> @948024326 commented on GitHub (Feb 20, 2025): > If you are using the ollama CLI to chat, do this: > > C:\> ollama run deepseek-r1:32b --keepalive -1m > >>> /set parameter num_ctx 8192 > Set parameter 'num_ctx' to '8192' > >>> hello > <think> > > </think> > > Hello! How can I assist you today? 😊 > > >>> /bye > If you are using a different client, you need to change its configuration to set `num_ctx` and `keep_alive`. thanks for your answer, i use ollama and dify, via API method, "you need to change its configuration to set num_ctx and keep_alive" how to do this? thanks a lot!
GiteaMirror commented 2026-04-22 12:32:15 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Feb 20, 2025):

You will have to read the dify documentation.

<!-- gh-comment-id:2671024877 --> @rick-github commented on GitHub (Feb 20, 2025): You will have to read the dify documentation.
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
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#31771
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?