github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-17 21:55:43 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #6118] panic: runtime error: integer divide by zero in memory.go on bad model create #3823

New Issue
Open
opened 2026-04-12 14:39:19 -05:00 by GiteaMirror · 9 comments
GiteaMirror commented 2026-04-12 14:39:19 -05:00
Owner
Copy Link

Originally created by @SongXiaoMao on GitHub (Aug 1, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/6118

What is the issue?

I installed ollama today, the system is Ubuntu2204,I downloaded llama3.1-405b-Q2.gguf,There are 9 split files in total. Ollama create llama -f Modelfile.txt is completed successfully.
The ollama list is displayed normally, but an error occurs when running.Error: Post "http://127.0.0.1:11434/api/chat": EOF
Open this link and get an error 404 page not found,Open http://127.0.0.1:11434 and it will show that Ollama is running.
The model file size is 151GB, my computer memory is 128G, and the video memory is 96G. Please help me, thank you.

OS

Linux

GPU

Nvidia

CPU

Intel

Ollama version

ollama version is 0.3.2

Originally created by @SongXiaoMao on GitHub (Aug 1, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/6118 ### What is the issue? I installed ollama today, the system is Ubuntu2204,I downloaded llama3.1-405b-Q2.gguf,There are 9 split files in total. Ollama create llama -f Modelfile.txt is completed successfully. The ollama list is displayed normally, but an error occurs when running.Error: Post "http://127.0.0.1:11434/api/chat": EOF Open this link and get an error 404 page not found,Open http://127.0.0.1:11434 and it will show that Ollama is running. The model file size is 151GB, my computer memory is 128G, and the video memory is 96G. Please help me, thank you. ### OS Linux ### GPU Nvidia ### CPU Intel ### Ollama version ollama version is 0.3.2
GiteaMirror added the bug label 2026-04-12 14:39:19 -05:00
GiteaMirror commented 2026-04-12 14:39:20 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Aug 1, 2024):

Server logs will aid in diagnosis.

<!-- gh-comment-id:2263027621 --> @rick-github commented on GitHub (Aug 1, 2024): Server logs will aid in diagnosis.
GiteaMirror commented 2026-04-12 14:39:20 -05:00
Author
Owner
Copy Link

@igorschlum commented on GitHub (Aug 1, 2024):

Hi @SongXiaoMao Can you share this model on Ollama? I would like to see if I can run it on my 192GB Mac Station.

<!-- gh-comment-id:2263159920 --> @igorschlum commented on GitHub (Aug 1, 2024): Hi @SongXiaoMao Can you share this model on Ollama? I would like to see if I can run it on my 192GB Mac Station.
GiteaMirror commented 2026-04-12 14:39:20 -05:00
Author
Owner
Copy Link

@SongXiaoMao commented on GitHub (Aug 2, 2024):

Server logs will aid in diagnosis.服务器日志将有助于诊断。

How to view?

<!-- gh-comment-id:2264243256 --> @SongXiaoMao commented on GitHub (Aug 2, 2024): > Server logs will aid in diagnosis.服务器日志将有助于诊断。 How to view?
GiteaMirror commented 2026-04-12 14:39:21 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Aug 2, 2024):

https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues

<!-- gh-comment-id:2264245856 --> @rick-github commented on GitHub (Aug 2, 2024): https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues
GiteaMirror commented 2026-04-12 14:39:22 -05:00
Author
Owner
Copy Link

@SongXiaoMao commented on GitHub (Aug 2, 2024):

(base) cheng@cheng:$ ollama run qwen
Error: Post "http://127.0.0.1:11434/api/chat": EOF
(base) cheng@cheng:$ journalctl -u ollama
8月 01 15:09:29 cheng systemd[1]: Started Ollama Service.
8月 01 15:09:29 cheng ollama[31790]: Couldn't find '/usr/share/ollama/.ollama/id_ed25519'. Generating new private key.
8月 01 15:09:29 cheng ollama[31790]: Your new public key is:
8月 01 15:09:29 cheng ollama[31790]: ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAILEdfF57fBFLPZGwkUEVkasIBOjp34OgZqC6P5gOckLi
8月 01 15:09:29 cheng ollama[31790]: 2024/08/01 15:09:29 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES>
8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.826+08:00 level=INFO source=images.go:781 msg="total blobs:>
8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.826+08:00 level=INFO source=images.go:788 msg="total unused>
8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.826+08:00 level=INFO source=routes.go:1156 msg="Listening o>
8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.827+08:00 level=INFO source=payload.go:30 msg="extracting e>
8月 01 15:09:36 cheng ollama[31790]: time=2024-08-01T15:09:36.397+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM >
8月 01 15:09:36 cheng ollama[31790]: time=2024-08-01T15:09:36.397+08:00 level=INFO source=gpu.go:205 msg="looking for com>
8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference com>
8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference com>
8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference com>
8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference com>
8月 01 15:20:48 cheng ollama[31790]: [GIN] 2024/08/01 - 15:20:48 | 200 | 94.041µs | 127.0.0.1 | HEAD "/"
8月 01 15:21:00 cheng ollama[31790]: [GIN] 2024/08/01 - 15:21:00 | 200 | 32.908µs | 127.0.0.1 | HEAD "/"
8月 01 15:21:33 cheng ollama[31790]: [GIN] 2024/08/01 - 15:21:33 | 200 | 61.512µs | 127.0.0.1 | HEAD "/"
8月 01 15:21:50 cheng ollama[31790]: [GIN] 2024/08/01 - 15:21:50 | 200 | 48.418µs | 127.0.0.1 | HEAD "/"
8月 01 15:23:25 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:25 | 200 | 32.858µs | 127.0.0.1 | HEAD "/"
8月 01 15:23:25 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:25 | 200 | 301.909µs | 127.0.0.1 | GET "/api>
8月 01 15:23:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:40 | 200 | 30.281µs | 127.0.0.1 | HEAD "/"
8月 01 15:23:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:40 | 200 | 87.991µs | 127.0.0.1 | GET "/api>
8月 01 15:23:47 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:47 | 200 | 29.357µs | 127.0.0.1 | HEAD "/"
8月 01 15:23:47 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:47 | 404 | 223.898µs | 127.0.0.1 | POST "/api>
8月 01 15:23:51 cheng ollama[31790]: time=2024-08-01T15:23:51.495+08:00 level=INFO source=download.go:173 msg="downloadin>
8月 01 15:23:55 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:55 | 200 | 7.982914121s | 127.0.0.1 | POST "/api>
8月 01 15:25:17 cheng ollama[31790]: [GIN] 2024/08/01 - 15:25:17 | 200 | 61.437µs | 127.0.0.1 | HEAD "/"
8月 01 15:26:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:26:56 | 200 | 29.932µs | 127.0.0.1 | HEAD "/"
8月 01 15:27:14 cheng ollama[31790]: [GIN] 2024/08/01 - 15:27:14 | 200 | 33.123µs | 127.0.0.1 | HEAD "/"
8月 01 15:28:51 cheng ollama[31790]: [GIN] 2024/08/01 - 15:28:51 | 200 | 34.332µs | 127.0.0.1 | HEAD "/"
8月 01 15:29:39 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:39 | 200 | 34.374µs | 127.0.0.1 | HEAD "/"
8月 01 15:29:52 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:52 | 200 | 30.588µs | 127.0.0.1 | HEAD "/"
8月 01 15:29:54 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:54 | 201 | 1.639126644s | 127.0.0.1 | POST "/api>
8月 01 15:29:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:56 | 200 | 1.677066999s | 127.0.0.1 | POST "/api>
8月 01 15:30:01 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:01 | 200 | 35.043µs | 127.0.0.1 | HEAD "/"
8月 01 15:30:01 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:01 | 200 | 381.711µs | 127.0.0.1 | GET "/api>
8月 01 15:30:07 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:07 | 200 | 31.294µs | 127.0.0.1 | HEAD "/"
8月 01 15:30:07 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:07 | 200 | 45.74282ms | 127.0.0.1 | POST "/api>
8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.622+08:00 level=INFO source=sched.go:710 msg="new model wil>
8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.623+08:00 level=INFO source=memory.go:309 msg="offload to c>
8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.624+08:00 level=INFO source=server.go:384 msg="starting lla>
8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.625+08:00 level=INFO source=sched.go:445 msg="loaded runner>
8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.625+08:00 level=INFO source=server.go:584 msg="waiting for >
8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.625+08:00 level=INFO source=server.go:618 msg="waiting for >
8月 01 15:30:07 cheng ollama[38287]: INFO [main] build info | build=1 commit="6eeaeba" tid="133149442916352" timestamp=17>
8月 01 15:30:07 cheng ollama[38287]: INFO [main] system info | n_threads=12 n_threads_batch=-1 system_info="AVX = 1 | AVX>
8月 01 15:30:07 cheng ollama[38287]: INFO [main] HTTP server listening | hostname="127.0.0.1" n_threads_http="23" port="3>
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: loaded meta data with 26 key-value pairs and 338 tensors from /u>
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in>
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 0: general.architecture str >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 1: general.name str >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 2: qwen2.block_count u32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 3: qwen2.context_length u32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 4: qwen2.embedding_length u32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 5: qwen2.feed_forward_length u32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 6: qwen2.attention.head_count u32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 8: qwen2.rope.freq_base f32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 10: general.file_type u32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 11: tokenizer.ggml.model str >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 12: tokenizer.ggml.pre str >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151>
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151>
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151>
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 19: tokenizer.chat_template str >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 20: tokenizer.ggml.add_bos_token bool >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 21: general.quantization_version u32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 22: quantize.imatrix.file str >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 23: quantize.imatrix.dataset str >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 24: quantize.imatrix.entries_count i32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 25: quantize.imatrix.chunks_count i32 >
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type f32: 141 tensors
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q2_K: 112 tensors
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q3_K: 56 tensors
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q4_K: 28 tensors
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q6_K: 1 tensors
8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.878+08:00 level=INFO source=server.go:618 msg="waiting for >
8月 01 15:30:08 cheng ollama[31790]: llm_load_vocab: special tokens cache size = 293
8月 01 15:30:08 cheng ollama[31790]: llm_load_vocab: token to piece cache size = 0.9338 MB
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: format = GGUF V3 (latest)
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: arch = qwen2
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: vocab type = BPE
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_vocab = 151936
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_merges = 151387
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: vocab_only = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_ctx_train = 32768
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd = 1536
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_layer = 28
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_head = 12
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_head_kv = 2
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_rot = 128
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_swa = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_head_k = 128
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_head_v = 128
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_gqa = 6
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_k_gqa = 256
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_v_gqa = 256
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_norm_eps = 0.0e+00
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_norm_rms_eps = 1.0e-06
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_clamp_kqv = 0.0e+00
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_max_alibi_bias = 0.0e+00
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_logit_scale = 0.0e+00
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_ff = 8960
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_expert = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_expert_used = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: causal attn = 1
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: pooling type = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: rope type = 2
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: rope scaling = linear
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: freq_base_train = 1000000.0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: freq_scale_train = 1
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_ctx_orig_yarn = 32768
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: rope_finetuned = unknown
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_d_conv = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_d_inner = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_d_state = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_dt_rank = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model type = ?B
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model ftype = Q2_K - Medium
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model params = 1.54 B
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model size = 639.30 MiB (3.47 BPW)
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: general.name = qwen2-1_5b-instruct
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: BOS token = 151643 '<|endoftext|>'
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: EOS token = 151645 '<|im_end|>'
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: PAD token = 151643 '<|endoftext|>'
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: LF token = 148848 'ÄĬ'
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: EOT token = 151645 '<|im_end|>'
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: max token length = 256
8月 01 15:30:08 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
8月 01 15:30:08 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
8月 01 15:30:08 cheng ollama[31790]: ggml_cuda_init: found 1 CUDA devices:
8月 01 15:30:08 cheng ollama[31790]: Device 0: NVIDIA GeForce RTX 3090, compute capability 8.6, VMM: yes
8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: ggml ctx size = 0.30 MiB
8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: offloading 28 repeating layers to GPU
8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: offloading non-repeating layers to GPU
8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: offloaded 29/29 layers to GPU
8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: CPU buffer size = 182.57 MiB
8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: CUDA0 buffer size = 639.30 MiB
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: n_ctx = 8192
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: n_batch = 512
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: n_ubatch = 512
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: flash_attn = 0
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: freq_base = 1000000.0
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: freq_scale = 1
8月 01 15:30:08 cheng ollama[31790]: llama_kv_cache_init: CUDA0 KV buffer size = 224.00 MiB
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: KV self size = 224.00 MiB, K (f16): 112.00 MiB, V (>
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host output buffer size = 2.34 MiB
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: CUDA0 compute buffer size = 299.75 MiB
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host compute buffer size = 19.01 MiB
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: graph nodes = 986
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: graph splits = 2
8月 01 15:30:09 cheng ollama[38287]: INFO [main] model loaded | tid="133149442916352" timestamp=1722497409
8月 01 15:30:10 cheng ollama[31790]: time=2024-08-01T15:30:10.139+08:00 level=INFO source=server.go:623 msg="llama runner>
8月 01 15:30:10 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:10 | 200 | 3.085087137s | 127.0.0.1 | POST "/api>
8月 01 15:30:14 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:14 | 200 | 180.746309ms | 127.0.0.1 | POST "/api>
8月 01 15:36:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:36:56 | 200 | 31.499µs | 127.0.0.1 | HEAD "/"
8月 01 15:36:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:36:56 | 200 | 194.946µs | 127.0.0.1 | GET "/api>
8月 01 15:37:02 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:02 | 200 | 65.7µs | 127.0.0.1 | HEAD "/"
8月 01 15:37:02 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:02 | 200 | 87.954µs | 127.0.0.1 | GET "/api>
8月 01 15:37:31 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:31 | 200 | 33.094µs | 127.0.0.1 | HEAD "/"
8月 01 15:37:33 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:33 | 201 | 1.651416565s | 127.0.0.1 | POST "/api>
8月 01 15:37:35 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:35 | 200 | 1.725709949s | 127.0.0.1 | POST "/api>
8月 01 15:37:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:40 | 200 | 28.411µs | 127.0.0.1 | HEAD "/"
8月 01 15:37:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:40 | 200 | 355.484µs | 127.0.0.1 | GET "/api>
8月 01 15:38:10 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:10 | 200 | 60.68µs | 127.0.0.1 | HEAD "/"
8月 01 15:38:10 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:10 | 200 | 22.792904ms | 127.0.0.1 | POST "/api>
8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.163+08:00 level=INFO source=sched.go:710 msg="new model wil>
8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.164+08:00 level=INFO source=memory.go:309 msg="offload to c>
8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.164+08:00 level=INFO source=server.go:384 msg="starting lla>
8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.165+08:00 level=INFO source=sched.go:445 msg="loaded runner>
8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.165+08:00 level=INFO source=server.go:584 msg="waiting for >
8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.165+08:00 level=INFO source=server.go:618 msg="waiting for >
8月 01 15:38:11 cheng ollama[41534]: INFO [main] build info | build=1 commit="6eeaeba" tid="125225824997376" timestamp=17>
8月 01 15:38:11 cheng ollama[41534]: INFO [main] system info | n_threads=12 n_threads_batch=-1 system_info="AVX = 1 | AVX>
8月 01 15:38:11 cheng ollama[41534]: INFO [main] HTTP server listening | hostname="127.0.0.1" n_threads_http="23" port="3>
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: loaded meta data with 26 key-value pairs and 338 tensors from /u>
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in>
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 0: general.architecture str >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 1: general.name str >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 2: qwen2.block_count u32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 3: qwen2.context_length u32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 4: qwen2.embedding_length u32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 5: qwen2.feed_forward_length u32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 6: qwen2.attention.head_count u32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 8: qwen2.rope.freq_base f32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 10: general.file_type u32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 11: tokenizer.ggml.model str >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 12: tokenizer.ggml.pre str >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151>
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151>
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151>
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 19: tokenizer.chat_template str >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 20: tokenizer.ggml.add_bos_token bool >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 21: general.quantization_version u32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 22: quantize.imatrix.file str >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 23: quantize.imatrix.dataset str >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 24: quantize.imatrix.entries_count i32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 25: quantize.imatrix.chunks_count i32 >
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type f32: 141 tensors
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q2_K: 112 tensors
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q3_K: 56 tensors
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q4_K: 28 tensors
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q6_K: 1 tensors
8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.417+08:00 level=INFO source=server.go:618 msg="waiting for >
8月 01 15:38:11 cheng ollama[31790]: llm_load_vocab: special tokens cache size = 293
8月 01 15:38:11 cheng ollama[31790]: llm_load_vocab: token to piece cache size = 0.9338 MB
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: format = GGUF V3 (latest)
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: arch = qwen2
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: vocab type = BPE
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_vocab = 151936
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_merges = 151387
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: vocab_only = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_ctx_train = 32768
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd = 1536
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_layer = 28
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_head = 12
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_head_kv = 2
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_rot = 128
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_swa = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_head_k = 128
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_head_v = 128
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_gqa = 6
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_k_gqa = 256
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_v_gqa = 256
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_norm_eps = 0.0e+00
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_norm_rms_eps = 1.0e-06
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_clamp_kqv = 0.0e+00
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_max_alibi_bias = 0.0e+00
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_logit_scale = 0.0e+00
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_ff = 8960
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_expert = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_expert_used = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: causal attn = 1
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: pooling type = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: rope type = 2
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: rope scaling = linear
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: freq_base_train = 1000000.0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: freq_scale_train = 1
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_ctx_orig_yarn = 32768
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: rope_finetuned = unknown
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_d_conv = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_d_inner = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_d_state = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_dt_rank = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model type = ?B
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model ftype = Q2_K - Medium
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model params = 1.54 B
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model size = 639.30 MiB (3.47 BPW)
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: general.name = qwen2-1_5b-instruct
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: BOS token = 151643 '<|endoftext|>'
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: EOS token = 151645 '<|im_end|>'
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: PAD token = 151643 '<|endoftext|>'
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: LF token = 148848 'ÄĬ'
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: EOT token = 151645 '<|im_end|>'
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: max token length = 256
8月 01 15:38:11 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
8月 01 15:38:11 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
8月 01 15:38:11 cheng ollama[31790]: ggml_cuda_init: found 1 CUDA devices:
8月 01 15:38:11 cheng ollama[31790]: Device 0: NVIDIA GeForce RTX 3090, compute capability 8.6, VMM: yes
8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: ggml ctx size = 0.30 MiB
8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: offloading 28 repeating layers to GPU
8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: offloading non-repeating layers to GPU
8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: offloaded 29/29 layers to GPU
8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: CPU buffer size = 182.57 MiB
8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: CUDA0 buffer size = 639.30 MiB
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: n_ctx = 8192
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: n_batch = 512
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: n_ubatch = 512
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: flash_attn = 0
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: freq_base = 1000000.0
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: freq_scale = 1
8月 01 15:38:12 cheng ollama[31790]: llama_kv_cache_init: CUDA0 KV buffer size = 224.00 MiB
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: KV self size = 224.00 MiB, K (f16): 112.00 MiB, V (>
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host output buffer size = 2.34 MiB
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: CUDA0 compute buffer size = 299.75 MiB
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host compute buffer size = 19.01 MiB
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: graph nodes = 986
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: graph splits = 2
8月 01 15:38:13 cheng ollama[41534]: INFO [main] model loaded | tid="125225824997376" timestamp=1722497893
8月 01 15:38:13 cheng ollama[31790]: time=2024-08-01T15:38:13.428+08:00 level=INFO source=server.go:623 msg="llama runner>
8月 01 15:38:13 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:13 | 200 | 2.81848391s | 127.0.0.1 | POST "/api>
8月 01 15:38:16 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:16 | 200 | 131.417236ms | 127.0.0.1 | POST "/api>
8月 01 15:38:18 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:18 | 200 | 337.686697ms | 127.0.0.1 | POST "/api>
8月 01 15:38:19 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:19 | 200 | 528.190585ms | 127.0.0.1 | POST "/api>
8月 01 15:38:21 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:21 | 200 | 744.627142ms | 127.0.0.1 | POST "/api>
8月 01 15:41:00 cheng ollama[31790]: [GIN] 2024/08/01 - 15:41:00 | 200 | 31.118µs | 127.0.0.1 | HEAD "/"
8月 01 15:41:00 cheng ollama[31790]: [GIN] 2024/08/01 - 15:41:00 | 200 | 218.55µs | 127.0.0.1 | GET "/api>
8月 01 15:45:20 cheng systemd[1]: Stopping Ollama Service...
8月 01 15:45:20 cheng systemd[1]: ollama.service: Deactivated successfully.
8月 01 15:45:20 cheng systemd[1]: Stopped Ollama Service.
8月 01 15:45:20 cheng systemd[1]: ollama.service: Consumed 40.519s CPU time.
-- Boot 33083ba5d27a4c399c75fd0c8948f208 --
8月 01 19:27:21 cheng systemd[1]: Started Ollama Service.
8月 01 19:27:21 cheng ollama[1889]: Couldn't find '/usr/share/ollama/.ollama/id_ed25519'. Generating new private key.
8月 01 19:27:21 cheng ollama[1889]: Your new public key is:
8月 01 19:27:21 cheng ollama[1889]: ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIMnZh7qZa0+q9j5qj9n/YsCevtxy0d/bNe+D6kJ9tT6p
8月 01 19:27:21 cheng ollama[1889]: 2024/08/01 19:27:21 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:>
8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.102+08:00 level=INFO source=images.go:781 msg="total blobs: >
8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.102+08:00 level=INFO source=images.go:788 msg="total unused >
8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.102+08:00 level=INFO source=routes.go:1156 msg="Listening on>
8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.103+08:00 level=INFO source=payload.go:30 msg="extracting em>
8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.372+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l>
8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.373+08:00 level=INFO source=gpu.go:205 msg="looking for comp>
8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 20:04:56 cheng ollama[1889]: [GIN] 2024/08/01 - 20:04:56 | 200 | 363.118µs | 127.0.0.1 | HEAD "/"
8月 01 20:05:18 cheng ollama[1889]: [GIN] 2024/08/01 - 20:05:18 | 200 | 30.807µs | 127.0.0.1 | HEAD "/"
8月 01 20:06:07 cheng ollama[1889]: [GIN] 2024/08/01 - 20:06:07 | 201 | 30.463162163s | 127.0.0.1 | POST "/api/>
8月 01 20:07:01 cheng ollama[1889]: [GIN] 2024/08/01 - 20:07:01 | 201 | 31.905709158s | 127.0.0.1 | POST "/api/>
8月 01 20:07:56 cheng ollama[1889]: [GIN] 2024/08/01 - 20:07:56 | 201 | 31.758401406s | 127.0.0.1 | POST "/api/>
8月 01 20:08:52 cheng ollama[1889]: [GIN] 2024/08/01 - 20:08:52 | 201 | 33.256550159s | 127.0.0.1 | POST "/api/>
8月 01 20:09:48 cheng ollama[1889]: [GIN] 2024/08/01 - 20:09:48 | 201 | 31.697320718s | 127.0.0.1 | POST "/api/>
8月 01 20:10:45 cheng ollama[1889]: [GIN] 2024/08/01 - 20:10:45 | 201 | 30.805938252s | 127.0.0.1 | POST "/api/>
8月 01 20:11:44 cheng ollama[1889]: [GIN] 2024/08/01 - 20:11:44 | 201 | 33.848623115s | 127.0.0.1 | POST "/api/>
8月 01 20:12:50 cheng ollama[1889]: [GIN] 2024/08/01 - 20:12:50 | 201 | 40.17151261s | 127.0.0.1 | POST "/api/>
8月 01 20:13:40 cheng ollama[1889]: [GIN] 2024/08/01 - 20:13:40 | 201 | 29.724228454s | 127.0.0.1 | POST "/api/>
8月 01 20:18:33 cheng ollama[1889]: [GIN] 2024/08/01 - 20:18:33 | 200 | 31.811µs | 127.0.0.1 | HEAD "/"
8月 01 20:18:43 cheng ollama[1889]: [GIN] 2024/08/01 - 20:18:43 | 200 | 5m3s | 127.0.0.1 | POST "/api/>
8月 01 20:19:22 cheng ollama[1889]: [GIN] 2024/08/01 - 20:19:22 | 201 | 30.191676368s | 127.0.0.1 | POST "/api/>
8月 01 20:19:51 cheng ollama[1889]: [GIN] 2024/08/01 - 20:19:51 | 200 | 28.187662838s | 127.0.0.1 | POST "/api/>
8月 01 20:22:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:22:36 | 200 | 31.05µs | 127.0.0.1 | HEAD "/"
8月 01 20:23:33 cheng ollama[1889]: [GIN] 2024/08/01 - 20:23:33 | 201 | 40.226746855s | 127.0.0.1 | POST "/api/>
8月 01 20:24:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:24:36 | 201 | 41.01948605s | 127.0.0.1 | POST "/api/>
8月 01 20:25:32 cheng ollama[1889]: [GIN] 2024/08/01 - 20:25:32 | 201 | 34.131806305s | 127.0.0.1 | POST "/api/>
8月 01 20:26:27 cheng ollama[1889]: [GIN] 2024/08/01 - 20:26:27 | 201 | 35.313518063s | 127.0.0.1 | POST "/api/>
8月 01 20:27:26 cheng ollama[1889]: [GIN] 2024/08/01 - 20:27:26 | 201 | 39.101495054s | 127.0.0.1 | POST "/api/>
8月 01 20:28:19 cheng ollama[1889]: [GIN] 2024/08/01 - 20:28:19 | 201 | 32.1910056s | 127.0.0.1 | POST "/api/>
8月 01 20:29:12 cheng ollama[1889]: [GIN] 2024/08/01 - 20:29:12 | 201 | 32.608388797s | 127.0.0.1 | POST "/api/>
8月 01 20:30:11 cheng ollama[1889]: [GIN] 2024/08/01 - 20:30:11 | 201 | 32.208379126s | 127.0.0.1 | POST "/api/>
8月 01 20:31:13 cheng ollama[1889]: [GIN] 2024/08/01 - 20:31:13 | 201 | 33.420692973s | 127.0.0.1 | POST "/api/>
8月 01 20:36:43 cheng ollama[1889]: [GIN] 2024/08/01 - 20:36:43 | 200 | 5m29s | 127.0.0.1 | POST "/api/>
8月 01 20:49:17 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:17 | 200 | 34.271µs | 127.0.0.1 | HEAD "/"
8月 01 20:49:17 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:17 | 200 | 488.318µs | 127.0.0.1 | GET "/api/>
8月 01 20:49:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:36 | 200 | 29.398µs | 127.0.0.1 | HEAD "/"
8月 01 20:49:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:36 | 200 | 404.433µs | 127.0.0.1 | GET "/api/>
8月 01 20:49:41 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:41 | 200 | 59.252µs | 127.0.0.1 | HEAD "/"
8月 01 20:49:41 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:41 | 200 | 9.883069ms | 127.0.0.1 | POST "/api/>
8月 01 20:49:42 cheng ollama[1889]: time=2024-08-01T20:49:42.325+08:00 level=WARN source=memory.go:115 msg="model missing>
8月 01 20:49:42 cheng ollama[1889]: panic: runtime error: integer divide by zero
8月 01 20:49:42 cheng ollama[1889]: goroutine 13 [running]:
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, , {, _, _}, {{0x0, 0x2000>
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm.PredictServerFit({0xc00074bbd8?, 0x4?, 0x4?}, , {, , >
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc00016a700, 0xc000709780, >
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc000116ba0, {0x22aeab50>
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server.(*Scheduler).Run.func1()
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 01 20:49:42 cheng ollama[1889]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 01 20:49:42 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 01 20:49:42 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 01 20:49:42 cheng systemd[1]: ollama.service: Consumed 21min 49.078s CPU time.
8月 01 20:49:45 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1.
8月 01 20:49:45 cheng systemd[1]: Stopped Ollama Service.
8月 01 20:49:45 cheng systemd[1]: ollama.service: Consumed 21min 49.078s CPU time.
8月 01 20:49:45 cheng systemd[1]: Started Ollama Service.
8月 01 20:49:45 cheng ollama[28228]: Couldn't find '/usr/share/ollama/.ollama/id_ed25519'. Generating new private key.
8月 01 20:49:45 cheng ollama[28228]: Your new public key is:
8月 01 20:49:45 cheng ollama[28228]: ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAILxBQqiTdXAnV3aWrQ5ZtN1EpdtgIZKBEST7iqBy8u2p
8月 01 20:49:45 cheng ollama[28228]: 2024/08/01 20:49:45 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES>
8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.626+08:00 level=INFO source=images.go:781 msg="total blobs:>
8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.626+08:00 level=INFO source=images.go:788 msg="total unused>
8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.626+08:00 level=INFO source=routes.go:1156 msg="Listening o>
8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.635+08:00 level=WARN source=assets.go:100 msg="unable to cl>
8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.635+08:00 level=INFO source=payload.go:30 msg="extracting e>
8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.347+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM >
8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.347+08:00 level=INFO source=gpu.go:205 msg="looking for com>
8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference com>
8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference com>
8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference com>
8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference com>
8月 01 20:51:45 cheng ollama[28228]: [GIN] 2024/08/01 - 20:51:45 | 200 | 53.369µs | 127.0.0.1 | HEAD "/"
8月 01 20:51:45 cheng ollama[28228]: [GIN] 2024/08/01 - 20:51:45 | 200 | 1.366054ms | 127.0.0.1 | POST "/api>
8月 01 20:51:45 cheng ollama[28228]: time=2024-08-01T20:51:45.643+08:00 level=WARN source=memory.go:115 msg="model missin>
8月 01 20:51:45 cheng ollama[28228]: panic: runtime error: integer divide by zero
8月 01 20:51:45 cheng ollama[28228]: goroutine 66 [running]:
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm.EstimateGPULayers({, _, _}, , {, _, _}, {{0x0, 0x200>
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm.PredictServerFit({0xc000893bd8?, 0x4?, 0x4?}, , {, ,>
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc000558200, 0xc0003d8020,>
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc0003b3020, {0x22aeab5>
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server.(*Scheduler).Run.func1()
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 01 20:51:45 cheng ollama[28228]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 01 20:51:45 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 01 20:51:45 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 01 20:51:45 cheng systemd[1]: ollama.service: Consumed 19.730s CPU time.
8月 01 20:51:48 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 2.
8月 01 20:51:48 cheng systemd[1]: Stopped Ollama Service.
8月 01 20:51:48 cheng systemd[1]: ollama.service: Consumed 19.730s CPU time.
8月 01 20:51:48 cheng systemd[1]: Started Ollama Service.
8月 01 20:51:48 cheng ollama[28388]: 2024/08/01 20:51:48 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES>
8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=INFO source=images.go:781 msg="total blobs:>
8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=INFO source=images.go:788 msg="total unused>
8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=INFO source=routes.go:1156 msg="Listening o>
8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=WARN source=assets.go:100 msg="unable to cl>
8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.857+08:00 level=WARN source=assets.go:100 msg="unable to cl>
8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.857+08:00 level=INFO source=payload.go:30 msg="extracting e>
8月 01 20:51:50 cheng systemd[1]: Stopping Ollama Service...
8月 01 20:51:55 cheng ollama[28388]: time=2024-08-01T20:51:55.741+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM >
8月 01 20:51:55 cheng ollama[28388]: time=2024-08-01T20:51:55.741+08:00 level=INFO source=gpu.go:205 msg="looking for com>
8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference com>
8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference com>
8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference com>
8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference com>
8月 01 20:51:56 cheng systemd[1]: ollama.service: Deactivated successfully.
8月 01 20:51:56 cheng systemd[1]: Stopped Ollama Service.
8月 01 20:51:56 cheng systemd[1]: ollama.service: Consumed 19.925s CPU time.
-- Boot 2c6de2a332144d9a8ee2924b08fb1feb --
8月 01 20:53:23 cheng systemd[1]: Started Ollama Service.
8月 01 20:53:23 cheng ollama[1877]: 2024/08/01 20:53:23 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:>
8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.988+08:00 level=INFO source=images.go:781 msg="total blobs: >
8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.989+08:00 level=INFO source=images.go:788 msg="total unused >
8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.989+08:00 level=INFO source=routes.go:1156 msg="Listening on>
8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.990+08:00 level=INFO source=payload.go:30 msg="extracting em>
8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.110+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l>
8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.110+08:00 level=INFO source=gpu.go:205 msg="looking for comp>
8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.664+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.665+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.665+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.665+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 20:55:12 cheng ollama[1877]: [GIN] 2024/08/01 - 20:55:12 | 200 | 347.797µs | 127.0.0.1 | HEAD "/"
8月 01 20:55:12 cheng ollama[1877]: [GIN] 2024/08/01 - 20:55:12 | 200 | 1.771204ms | 127.0.0.1 | POST "/api/>
8月 01 20:55:12 cheng ollama[1877]: time=2024-08-01T20:55:12.930+08:00 level=WARN source=memory.go:115 msg="model missing>
8月 01 20:55:12 cheng ollama[1877]: panic: runtime error: integer divide by zero
8月 01 20:55:12 cheng ollama[1877]: goroutine 54 [running]:
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm.EstimateGPULayers({, _, _}, , {, _, _}, {{0x0, 0x2000>
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm.PredictServerFit({0xc00039bbd8?, 0x4?, 0x4?}, , {, , >
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc0004b0b00, 0xc000037640, >
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc00043af00, {0x22aeab50>
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server.(*Scheduler).Run.func1()
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 01 20:55:12 cheng ollama[1877]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 01 20:55:12 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 01 20:55:12 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 01 20:55:12 cheng systemd[1]: ollama.service: Consumed 18.734s CPU time.
8月 01 20:55:16 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1.
8月 01 20:55:16 cheng systemd[1]: Stopped Ollama Service.
8月 01 20:55:16 cheng systemd[1]: ollama.service: Consumed 18.734s CPU time.
8月 01 20:55:16 cheng systemd[1]: Started Ollama Service.
8月 01 20:55:16 cheng ollama[4792]: 2024/08/01 20:55:16 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:>
8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.067+08:00 level=INFO source=images.go:781 msg="total blobs: >
8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.067+08:00 level=INFO source=images.go:788 msg="total unused >
8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.067+08:00 level=INFO source=routes.go:1156 msg="Listening on>
8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.068+08:00 level=WARN source=assets.go:100 msg="unable to cle>
8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.068+08:00 level=INFO source=payload.go:30 msg="extracting em>
8月 01 20:55:22 cheng ollama[4792]: time=2024-08-01T20:55:22.577+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l>
8月 01 20:55:22 cheng ollama[4792]: time=2024-08-01T20:55:22.577+08:00 level=INFO source=gpu.go:205 msg="looking for comp>
8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 20:56:40 cheng ollama[4792]: [GIN] 2024/08/01 - 20:56:40 | 200 | 58.151µs | 127.0.0.1 | HEAD "/"
8月 01 20:56:40 cheng ollama[4792]: [GIN] 2024/08/01 - 20:56:40 | 200 | 465.523µs | 127.0.0.1 | GET "/api/>
8月 01 20:59:09 cheng ollama[4792]: [GIN] 2024/08/01 - 20:59:09 | 200 | 50.178µs | 127.0.0.1 | HEAD "/"
8月 01 20:59:09 cheng ollama[4792]: [GIN] 2024/08/01 - 20:59:09 | 200 | 2.818931ms | 127.0.0.1 | POST "/api/>
8月 01 20:59:09 cheng ollama[4792]: time=2024-08-01T20:59:09.773+08:00 level=WARN source=memory.go:115 msg="model missing>
8月 01 20:59:09 cheng ollama[4792]: panic: runtime error: integer divide by zero
8月 01 20:59:09 cheng ollama[4792]: goroutine 66 [running]:
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm.EstimateGPULayers({, _, _}, , {, _, _}, {{0x0, 0x2000>
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm.PredictServerFit({0xc0004e1bd8?, 0x4?, 0x4?}, , {, , >
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc000255100, 0xc0004943c0, >
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc0003b0f60, {0x22aeab50>
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server.(*Scheduler).Run.func1()
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 01 20:59:09 cheng ollama[4792]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 01 20:59:09 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 01 20:59:09 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 01 20:59:09 cheng systemd[1]: ollama.service: Consumed 20.184s CPU time.
8月 01 20:59:13 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 2.
8月 01 20:59:13 cheng systemd[1]: Stopped Ollama Service.
8月 01 20:59:13 cheng systemd[1]: ollama.service: Consumed 20.184s CPU time.
8月 01 20:59:13 cheng systemd[1]: Started Ollama Service.
8月 01 20:59:13 cheng ollama[5138]: 2024/08/01 20:59:13 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:>
8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.067+08:00 level=INFO source=images.go:781 msg="total blobs: >
8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=INFO source=images.go:788 msg="total unused >
8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=INFO source=routes.go:1156 msg="Listening on>
8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=WARN source=assets.go:100 msg="unable to cle>
8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=WARN source=assets.go:100 msg="unable to cle>
8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.069+08:00 level=INFO source=payload.go:30 msg="extracting em>
8月 01 20:59:19 cheng ollama[5138]: time=2024-08-01T20:59:19.727+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l>
8月 01 20:59:19 cheng ollama[5138]: time=2024-08-01T20:59:19.727+08:00 level=INFO source=gpu.go:205 msg="looking for comp>
8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 21:04:01 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:01 | 200 | 51.242µs | 127.0.0.1 | HEAD "/"
8月 01 21:04:01 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:01 | 200 | 442.689µs | 127.0.0.1 | GET "/api/>
8月 01 21:04:12 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:12 | 200 | 35.143µs | 127.0.0.1 | HEAD "/"
8月 01 21:04:12 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:12 | 200 | 1.312058ms | 127.0.0.1 | POST "/api/>
8月 01 21:04:13 cheng ollama[5138]: time=2024-08-01T21:04:13.208+08:00 level=WARN source=memory.go:115 msg="model missing>
8月 01 21:04:13 cheng ollama[5138]: panic: runtime error: integer divide by zero
8月 01 21:04:13 cheng ollama[5138]: goroutine 82 [running]:
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm.EstimateGPULayers({, _, _}, , {, _, _}, {{0x0, 0x2000>
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm.PredictServerFit({0xc000107bd8?, 0x4?, 0x4?}, , {, , >
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc0003d2400, 0xc000037ea0, >
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc0003b0ea0, {0x22aeab50>
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server.(*Scheduler).Run.func1()
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 01 21:04:13 cheng ollama[5138]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 01 21:04:13 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 01 21:04:13 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 01 21:04:13 cheng systemd[1]: ollama.service: Consumed 19.781s CPU time.
8月 01 21:04:16 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 3.
8月 01 21:04:16 cheng systemd[1]: Stopped Ollama Service.
8月 01 21:04:16 cheng systemd[1]: ollama.service: Consumed 19.781s CPU time.
8月 01 21:04:16 cheng systemd[1]: Started Ollama Service.
8月 01 21:04:16 cheng ollama[5398]: 2024/08/01 21:04:16 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:>
8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.318+08:00 level=INFO source=images.go:781 msg="total blobs: >
8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.318+08:00 level=INFO source=images.go:788 msg="total unused >
8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=INFO source=routes.go:1156 msg="Listening on>
8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=WARN source=assets.go:100 msg="unable to cle>
8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=WARN source=assets.go:100 msg="unable to cle>
8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=WARN source=assets.go:100 msg="unable to cle>
8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.320+08:00 level=INFO source=payload.go:30 msg="extracting em>
8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.069+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l>
8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.069+08:00 level=INFO source=gpu.go:205 msg="looking for comp>
8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 01 21:04:23 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:23 | 404 | 86.36µs | 127.0.0.1 | GET "/api/>
8月 01 21:04:24 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:24 | 404 | 9.775µs | 127.0.0.1 | GET "/favi>
8月 01 21:04:38 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:38 | 404 | 13.876µs | 127.0.0.1 | GET "/api"
8月 01 21:04:41 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:41 | 200 | 27.389µs | 127.0.0.1 | GET "/"
8月 01 21:05:17 cheng ollama[5398]: [GIN] 2024/08/01 - 21:05:17 | 404 | 11.268µs | 127.0.0.1 | GET "/api/>
8月 01 21:09:25 cheng ollama[5398]: [GIN] 2024/08/01 - 21:09:25 | 200 | 141.497µs | 127.0.0.1 | GET "/"
8月 01 21:13:26 cheng ollama[5398]: [GIN] 2024/08/01 - 21:13:26 | 200 | 70.515µs | 127.0.0.1 | GET "/"
8月 01 21:15:11 cheng ollama[5398]: [GIN] 2024/08/01 - 21:15:11 | 200 | 56.712µs | 127.0.0.1 | GET "/api/>
8月 01 21:17:26 cheng systemd[1]: Stopping Ollama Service...
8月 01 21:17:26 cheng systemd[1]: ollama.service: Deactivated successfully.
8月 01 21:17:26 cheng systemd[1]: Stopped Ollama Service.
8月 01 21:17:26 cheng systemd[1]: ollama.service: Consumed 20.123s CPU time.
-- Boot 9bceaafe62e14190ba2fd13cfe4b660c --
8月 02 07:54:47 cheng systemd[1]: Started Ollama Service.
8月 02 07:54:47 cheng ollama[1896]: 2024/08/02 07:54:47 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:>
8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.944+08:00 level=INFO source=images.go:781 msg="total blobs: >
8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.944+08:00 level=INFO source=images.go:788 msg="total unused >
8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.945+08:00 level=INFO source=routes.go:1156 msg="Listening on>
8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.946+08:00 level=INFO source=payload.go:30 msg="extracting em>
8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.037+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l>
8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.037+08:00 level=INFO source=gpu.go:205 msg="looking for comp>
8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 08:02:01 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:01 | 200 | 399.071µs | 127.0.0.1 | HEAD "/"
8月 02 08:02:01 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:01 | 200 | 1.195064ms | 127.0.0.1 | GET "/api/>
8月 02 08:02:12 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:12 | 200 | 44.049µs | 127.0.0.1 | HEAD "/"
8月 02 08:02:12 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:12 | 200 | 1.793469ms | 127.0.0.1 | POST "/api/>
8月 02 08:02:13 cheng ollama[1896]: time=2024-08-02T08:02:13.426+08:00 level=WARN source=memory.go:115 msg="model missing>
8月 02 08:02:13 cheng ollama[1896]: panic: runtime error: integer divide by zero
8月 02 08:02:13 cheng ollama[1896]: goroutine 70 [running]:
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm.EstimateGPULayers({, _, _}, , {, _, _}, {{0x0, 0x2000>
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm.PredictServerFit({0xc00040dbd8?, 0x4?, 0x4?}, , {, , >
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc0001d9900, 0xc000037660, >
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc0000cf080, {0x22aeab50>
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server.(*Scheduler).Run.func1()
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 02 08:02:13 cheng ollama[1896]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 02 08:02:13 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 02 08:02:13 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 02 08:02:13 cheng systemd[1]: ollama.service: Consumed 17.958s CPU time.
8月 02 08:02:16 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1.
8月 02 08:02:16 cheng systemd[1]: Stopped Ollama Service.
8月 02 08:02:16 cheng systemd[1]: ollama.service: Consumed 17.958s CPU time.
8月 02 08:02:16 cheng systemd[1]: Started Ollama Service.
8月 02 08:02:16 cheng ollama[4728]: 2024/08/02 08:02:16 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:>
8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=INFO source=images.go:781 msg="total blobs: >
8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=INFO source=images.go:788 msg="total unused >
8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=INFO source=routes.go:1156 msg="Listening on>
8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=WARN source=assets.go:100 msg="unable to cle>
8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.580+08:00 level=INFO source=payload.go:30 msg="extracting em>
8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.278+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l>
8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.278+08:00 level=INFO source=gpu.go:205 msg="looking for comp>
8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 08:08:06 cheng systemd[1]: Stopping Ollama Service...
8月 02 08:08:06 cheng systemd[1]: ollama.service: Deactivated successfully.
8月 02 08:08:06 cheng systemd[1]: Stopped Ollama Service.
8月 02 08:08:06 cheng systemd[1]: ollama.service: Consumed 20.017s CPU time.
-- Boot 88a56500e37f4b6eaaaeaca00780c960 --
8月 02 08:10:53 cheng systemd[1]: Started Ollama Service.
8月 02 08:10:53 cheng ollama[1889]: 2024/08/02 08:10:53 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:>
8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.056+08:00 level=INFO source=images.go:781 msg="total blobs: >
8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.057+08:00 level=INFO source=images.go:788 msg="total unused >
8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.057+08:00 level=INFO source=routes.go:1156 msg="Listening on>
8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.057+08:00 level=INFO source=payload.go:30 msg="extracting em>
8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.295+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l>
8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.295+08:00 level=INFO source=gpu.go:205 msg="looking for comp>
8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 08:15:44 cheng ollama[1889]: [GIN] 2024/08/02 - 08:15:44 | 200 | 386.787µs | 127.0.0.1 | HEAD "/"
8月 02 08:15:44 cheng ollama[1889]: [GIN] 2024/08/02 - 08:15:44 | 200 | 3.462854ms | 127.0.0.1 | POST "/api/>
8月 02 08:15:44 cheng ollama[1889]: time=2024-08-02T08:15:44.622+08:00 level=WARN source=memory.go:115 msg="model missing>
8月 02 08:15:44 cheng ollama[1889]: panic: runtime error: integer divide by zero
8月 02 08:15:44 cheng ollama[1889]: goroutine 68 [running]:
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm.EstimateGPULayers({, _, _}, , {, _, _}, {{0x0, 0x2000>
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm.PredictServerFit({0xc0008a9bd8?, 0x4?, 0x4?}, , {, _, >
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc00046c200, 0xc00037c6a0, >
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc00043af60, {0x22aeab50>
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server.(*Scheduler).Run.func1()
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 02 08:15:44 cheng ollama[1889]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 02 08:15:44 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 02 08:15:44 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 02 08:15:44 cheng systemd[1]: ollama.service: Consumed 18.710s CPU time.
8月 02 08:15:47 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1.
8月 02 08:15:47 cheng systemd[1]: Stopped Ollama Service.
8月 02 08:15:47 cheng systemd[1]: ollama.service: Consumed 18.710s CPU time.
8月 02 08:15:47 cheng systemd[1]: Started Ollama Service.
8月 02 08:15:47 cheng ollama[4056]: 2024/08/02 08:15:47 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:>
8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=INFO source=images.go:781 msg="total blobs: >
8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=INFO source=images.go:788 msg="total unused >
8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=INFO source=routes.go:1156 msg="Listening on>
8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=WARN source=assets.go:100 msg="unable to cle>
8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.845+08:00 level=INFO source=payload.go:30 msg="extracting em>
8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.330+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l>
8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.330+08:00 level=INFO source=gpu.go:205 msg="looking for comp>
8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference comp>
8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference comp>
lines 637-671/671 (END)

<!-- gh-comment-id:2264257312 --> @SongXiaoMao commented on GitHub (Aug 2, 2024): (base) cheng@cheng:~$ ollama run qwen Error: Post "http://127.0.0.1:11434/api/chat": EOF (base) cheng@cheng:~$ journalctl -u ollama 8月 01 15:09:29 cheng systemd[1]: Started Ollama Service. 8月 01 15:09:29 cheng ollama[31790]: Couldn't find '/usr/share/ollama/.ollama/id_ed25519'. Generating new private key. 8月 01 15:09:29 cheng ollama[31790]: Your new public key is: 8月 01 15:09:29 cheng ollama[31790]: ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAILEdfF57fBFLPZGwkUEVkasIBOjp34OgZqC6P5gOckLi 8月 01 15:09:29 cheng ollama[31790]: 2024/08/01 15:09:29 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES> 8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.826+08:00 level=INFO source=images.go:781 msg="total blobs:> 8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.826+08:00 level=INFO source=images.go:788 msg="total unused> 8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.826+08:00 level=INFO source=routes.go:1156 msg="Listening o> 8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.827+08:00 level=INFO source=payload.go:30 msg="extracting e> 8月 01 15:09:36 cheng ollama[31790]: time=2024-08-01T15:09:36.397+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM > 8月 01 15:09:36 cheng ollama[31790]: time=2024-08-01T15:09:36.397+08:00 level=INFO source=gpu.go:205 msg="looking for com> 8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference com> 8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference com> 8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference com> 8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference com> 8月 01 15:20:48 cheng ollama[31790]: [GIN] 2024/08/01 - 15:20:48 | 200 | 94.041µs | 127.0.0.1 | HEAD "/" 8月 01 15:21:00 cheng ollama[31790]: [GIN] 2024/08/01 - 15:21:00 | 200 | 32.908µs | 127.0.0.1 | HEAD "/" 8月 01 15:21:33 cheng ollama[31790]: [GIN] 2024/08/01 - 15:21:33 | 200 | 61.512µs | 127.0.0.1 | HEAD "/" 8月 01 15:21:50 cheng ollama[31790]: [GIN] 2024/08/01 - 15:21:50 | 200 | 48.418µs | 127.0.0.1 | HEAD "/" 8月 01 15:23:25 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:25 | 200 | 32.858µs | 127.0.0.1 | HEAD "/" 8月 01 15:23:25 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:25 | 200 | 301.909µs | 127.0.0.1 | GET "/api> 8月 01 15:23:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:40 | 200 | 30.281µs | 127.0.0.1 | HEAD "/" 8月 01 15:23:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:40 | 200 | 87.991µs | 127.0.0.1 | GET "/api> 8月 01 15:23:47 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:47 | 200 | 29.357µs | 127.0.0.1 | HEAD "/" 8月 01 15:23:47 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:47 | 404 | 223.898µs | 127.0.0.1 | POST "/api> 8月 01 15:23:51 cheng ollama[31790]: time=2024-08-01T15:23:51.495+08:00 level=INFO source=download.go:173 msg="downloadin> 8月 01 15:23:55 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:55 | 200 | 7.982914121s | 127.0.0.1 | POST "/api> 8月 01 15:25:17 cheng ollama[31790]: [GIN] 2024/08/01 - 15:25:17 | 200 | 61.437µs | 127.0.0.1 | HEAD "/" 8月 01 15:26:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:26:56 | 200 | 29.932µs | 127.0.0.1 | HEAD "/" 8月 01 15:27:14 cheng ollama[31790]: [GIN] 2024/08/01 - 15:27:14 | 200 | 33.123µs | 127.0.0.1 | HEAD "/" 8月 01 15:28:51 cheng ollama[31790]: [GIN] 2024/08/01 - 15:28:51 | 200 | 34.332µs | 127.0.0.1 | HEAD "/" 8月 01 15:29:39 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:39 | 200 | 34.374µs | 127.0.0.1 | HEAD "/" 8月 01 15:29:52 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:52 | 200 | 30.588µs | 127.0.0.1 | HEAD "/" 8月 01 15:29:54 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:54 | 201 | 1.639126644s | 127.0.0.1 | POST "/api> 8月 01 15:29:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:56 | 200 | 1.677066999s | 127.0.0.1 | POST "/api> 8月 01 15:30:01 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:01 | 200 | 35.043µs | 127.0.0.1 | HEAD "/" 8月 01 15:30:01 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:01 | 200 | 381.711µs | 127.0.0.1 | GET "/api> 8月 01 15:30:07 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:07 | 200 | 31.294µs | 127.0.0.1 | HEAD "/" 8月 01 15:30:07 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:07 | 200 | 45.74282ms | 127.0.0.1 | POST "/api> 8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.622+08:00 level=INFO source=sched.go:710 msg="new model wil> 8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.623+08:00 level=INFO source=memory.go:309 msg="offload to c> 8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.624+08:00 level=INFO source=server.go:384 msg="starting lla> 8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.625+08:00 level=INFO source=sched.go:445 msg="loaded runner> 8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.625+08:00 level=INFO source=server.go:584 msg="waiting for > 8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.625+08:00 level=INFO source=server.go:618 msg="waiting for > 8月 01 15:30:07 cheng ollama[38287]: INFO [main] build info | build=1 commit="6eeaeba" tid="133149442916352" timestamp=17> 8月 01 15:30:07 cheng ollama[38287]: INFO [main] system info | n_threads=12 n_threads_batch=-1 system_info="AVX = 1 | AVX> 8月 01 15:30:07 cheng ollama[38287]: INFO [main] HTTP server listening | hostname="127.0.0.1" n_threads_http="23" port="3> 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: loaded meta data with 26 key-value pairs and 338 tensors from /u> 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in> 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 0: general.architecture str > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 1: general.name str > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 2: qwen2.block_count u32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 3: qwen2.context_length u32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 4: qwen2.embedding_length u32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 5: qwen2.feed_forward_length u32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 6: qwen2.attention.head_count u32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 8: qwen2.rope.freq_base f32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 10: general.file_type u32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 11: tokenizer.ggml.model str > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 12: tokenizer.ggml.pre str > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151> 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151> 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151> 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 19: tokenizer.chat_template str > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 20: tokenizer.ggml.add_bos_token bool > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 21: general.quantization_version u32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 22: quantize.imatrix.file str > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 23: quantize.imatrix.dataset str > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 24: quantize.imatrix.entries_count i32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 25: quantize.imatrix.chunks_count i32 > 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type f32: 141 tensors 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q2_K: 112 tensors 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q3_K: 56 tensors 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q4_K: 28 tensors 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q6_K: 1 tensors 8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.878+08:00 level=INFO source=server.go:618 msg="waiting for > 8月 01 15:30:08 cheng ollama[31790]: llm_load_vocab: special tokens cache size = 293 8月 01 15:30:08 cheng ollama[31790]: llm_load_vocab: token to piece cache size = 0.9338 MB 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: format = GGUF V3 (latest) 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: arch = qwen2 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: vocab type = BPE 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_vocab = 151936 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_merges = 151387 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: vocab_only = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_ctx_train = 32768 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd = 1536 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_layer = 28 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_head = 12 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_head_kv = 2 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_rot = 128 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_swa = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_head_k = 128 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_head_v = 128 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_gqa = 6 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_k_gqa = 256 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_v_gqa = 256 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_norm_eps = 0.0e+00 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_norm_rms_eps = 1.0e-06 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_clamp_kqv = 0.0e+00 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_max_alibi_bias = 0.0e+00 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_logit_scale = 0.0e+00 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_ff = 8960 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_expert = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_expert_used = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: causal attn = 1 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: pooling type = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: rope type = 2 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: rope scaling = linear 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: freq_base_train = 1000000.0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: freq_scale_train = 1 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_ctx_orig_yarn = 32768 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: rope_finetuned = unknown 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_d_conv = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_d_inner = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_d_state = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_dt_rank = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model type = ?B 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model ftype = Q2_K - Medium 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model params = 1.54 B 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model size = 639.30 MiB (3.47 BPW) 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: general.name = qwen2-1_5b-instruct 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: BOS token = 151643 '<|endoftext|>' 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: EOS token = 151645 '<|im_end|>' 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: PAD token = 151643 '<|endoftext|>' 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: LF token = 148848 'ÄĬ' 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: EOT token = 151645 '<|im_end|>' 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: max token length = 256 8月 01 15:30:08 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no 8月 01 15:30:08 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no 8月 01 15:30:08 cheng ollama[31790]: ggml_cuda_init: found 1 CUDA devices: 8月 01 15:30:08 cheng ollama[31790]: Device 0: NVIDIA GeForce RTX 3090, compute capability 8.6, VMM: yes 8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: ggml ctx size = 0.30 MiB 8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: offloading 28 repeating layers to GPU 8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: offloading non-repeating layers to GPU 8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: offloaded 29/29 layers to GPU 8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: CPU buffer size = 182.57 MiB 8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: CUDA0 buffer size = 639.30 MiB 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: n_ctx = 8192 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: n_batch = 512 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: n_ubatch = 512 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: flash_attn = 0 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: freq_base = 1000000.0 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: freq_scale = 1 8月 01 15:30:08 cheng ollama[31790]: llama_kv_cache_init: CUDA0 KV buffer size = 224.00 MiB 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: KV self size = 224.00 MiB, K (f16): 112.00 MiB, V (> 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host output buffer size = 2.34 MiB 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: CUDA0 compute buffer size = 299.75 MiB 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host compute buffer size = 19.01 MiB 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: graph nodes = 986 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: graph splits = 2 8月 01 15:30:09 cheng ollama[38287]: INFO [main] model loaded | tid="133149442916352" timestamp=1722497409 8月 01 15:30:10 cheng ollama[31790]: time=2024-08-01T15:30:10.139+08:00 level=INFO source=server.go:623 msg="llama runner> 8月 01 15:30:10 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:10 | 200 | 3.085087137s | 127.0.0.1 | POST "/api> 8月 01 15:30:14 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:14 | 200 | 180.746309ms | 127.0.0.1 | POST "/api> 8月 01 15:36:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:36:56 | 200 | 31.499µs | 127.0.0.1 | HEAD "/" 8月 01 15:36:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:36:56 | 200 | 194.946µs | 127.0.0.1 | GET "/api> 8月 01 15:37:02 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:02 | 200 | 65.7µs | 127.0.0.1 | HEAD "/" 8月 01 15:37:02 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:02 | 200 | 87.954µs | 127.0.0.1 | GET "/api> 8月 01 15:37:31 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:31 | 200 | 33.094µs | 127.0.0.1 | HEAD "/" 8月 01 15:37:33 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:33 | 201 | 1.651416565s | 127.0.0.1 | POST "/api> 8月 01 15:37:35 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:35 | 200 | 1.725709949s | 127.0.0.1 | POST "/api> 8月 01 15:37:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:40 | 200 | 28.411µs | 127.0.0.1 | HEAD "/" 8月 01 15:37:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:40 | 200 | 355.484µs | 127.0.0.1 | GET "/api> 8月 01 15:38:10 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:10 | 200 | 60.68µs | 127.0.0.1 | HEAD "/" 8月 01 15:38:10 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:10 | 200 | 22.792904ms | 127.0.0.1 | POST "/api> 8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.163+08:00 level=INFO source=sched.go:710 msg="new model wil> 8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.164+08:00 level=INFO source=memory.go:309 msg="offload to c> 8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.164+08:00 level=INFO source=server.go:384 msg="starting lla> 8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.165+08:00 level=INFO source=sched.go:445 msg="loaded runner> 8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.165+08:00 level=INFO source=server.go:584 msg="waiting for > 8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.165+08:00 level=INFO source=server.go:618 msg="waiting for > 8月 01 15:38:11 cheng ollama[41534]: INFO [main] build info | build=1 commit="6eeaeba" tid="125225824997376" timestamp=17> 8月 01 15:38:11 cheng ollama[41534]: INFO [main] system info | n_threads=12 n_threads_batch=-1 system_info="AVX = 1 | AVX> 8月 01 15:38:11 cheng ollama[41534]: INFO [main] HTTP server listening | hostname="127.0.0.1" n_threads_http="23" port="3> 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: loaded meta data with 26 key-value pairs and 338 tensors from /u> 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in> 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 0: general.architecture str > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 1: general.name str > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 2: qwen2.block_count u32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 3: qwen2.context_length u32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 4: qwen2.embedding_length u32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 5: qwen2.feed_forward_length u32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 6: qwen2.attention.head_count u32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 8: qwen2.rope.freq_base f32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 10: general.file_type u32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 11: tokenizer.ggml.model str > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 12: tokenizer.ggml.pre str > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151> 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151> 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151> 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 19: tokenizer.chat_template str > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 20: tokenizer.ggml.add_bos_token bool > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 21: general.quantization_version u32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 22: quantize.imatrix.file str > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 23: quantize.imatrix.dataset str > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 24: quantize.imatrix.entries_count i32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 25: quantize.imatrix.chunks_count i32 > 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type f32: 141 tensors 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q2_K: 112 tensors 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q3_K: 56 tensors 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q4_K: 28 tensors 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q6_K: 1 tensors 8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.417+08:00 level=INFO source=server.go:618 msg="waiting for > 8月 01 15:38:11 cheng ollama[31790]: llm_load_vocab: special tokens cache size = 293 8月 01 15:38:11 cheng ollama[31790]: llm_load_vocab: token to piece cache size = 0.9338 MB 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: format = GGUF V3 (latest) 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: arch = qwen2 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: vocab type = BPE 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_vocab = 151936 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_merges = 151387 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: vocab_only = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_ctx_train = 32768 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd = 1536 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_layer = 28 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_head = 12 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_head_kv = 2 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_rot = 128 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_swa = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_head_k = 128 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_head_v = 128 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_gqa = 6 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_k_gqa = 256 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_v_gqa = 256 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_norm_eps = 0.0e+00 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_norm_rms_eps = 1.0e-06 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_clamp_kqv = 0.0e+00 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_max_alibi_bias = 0.0e+00 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_logit_scale = 0.0e+00 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_ff = 8960 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_expert = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_expert_used = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: causal attn = 1 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: pooling type = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: rope type = 2 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: rope scaling = linear 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: freq_base_train = 1000000.0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: freq_scale_train = 1 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_ctx_orig_yarn = 32768 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: rope_finetuned = unknown 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_d_conv = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_d_inner = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_d_state = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_dt_rank = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model type = ?B 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model ftype = Q2_K - Medium 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model params = 1.54 B 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model size = 639.30 MiB (3.47 BPW) 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: general.name = qwen2-1_5b-instruct 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: BOS token = 151643 '<|endoftext|>' 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: EOS token = 151645 '<|im_end|>' 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: PAD token = 151643 '<|endoftext|>' 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: LF token = 148848 'ÄĬ' 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: EOT token = 151645 '<|im_end|>' 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: max token length = 256 8月 01 15:38:11 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no 8月 01 15:38:11 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no 8月 01 15:38:11 cheng ollama[31790]: ggml_cuda_init: found 1 CUDA devices: 8月 01 15:38:11 cheng ollama[31790]: Device 0: NVIDIA GeForce RTX 3090, compute capability 8.6, VMM: yes 8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: ggml ctx size = 0.30 MiB 8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: offloading 28 repeating layers to GPU 8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: offloading non-repeating layers to GPU 8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: offloaded 29/29 layers to GPU 8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: CPU buffer size = 182.57 MiB 8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: CUDA0 buffer size = 639.30 MiB 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: n_ctx = 8192 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: n_batch = 512 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: n_ubatch = 512 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: flash_attn = 0 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: freq_base = 1000000.0 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: freq_scale = 1 8月 01 15:38:12 cheng ollama[31790]: llama_kv_cache_init: CUDA0 KV buffer size = 224.00 MiB 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: KV self size = 224.00 MiB, K (f16): 112.00 MiB, V (> 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host output buffer size = 2.34 MiB 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: CUDA0 compute buffer size = 299.75 MiB 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host compute buffer size = 19.01 MiB 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: graph nodes = 986 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: graph splits = 2 8月 01 15:38:13 cheng ollama[41534]: INFO [main] model loaded | tid="125225824997376" timestamp=1722497893 8月 01 15:38:13 cheng ollama[31790]: time=2024-08-01T15:38:13.428+08:00 level=INFO source=server.go:623 msg="llama runner> 8月 01 15:38:13 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:13 | 200 | 2.81848391s | 127.0.0.1 | POST "/api> 8月 01 15:38:16 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:16 | 200 | 131.417236ms | 127.0.0.1 | POST "/api> 8月 01 15:38:18 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:18 | 200 | 337.686697ms | 127.0.0.1 | POST "/api> 8月 01 15:38:19 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:19 | 200 | 528.190585ms | 127.0.0.1 | POST "/api> 8月 01 15:38:21 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:21 | 200 | 744.627142ms | 127.0.0.1 | POST "/api> 8月 01 15:41:00 cheng ollama[31790]: [GIN] 2024/08/01 - 15:41:00 | 200 | 31.118µs | 127.0.0.1 | HEAD "/" 8月 01 15:41:00 cheng ollama[31790]: [GIN] 2024/08/01 - 15:41:00 | 200 | 218.55µs | 127.0.0.1 | GET "/api> 8月 01 15:45:20 cheng systemd[1]: Stopping Ollama Service... 8月 01 15:45:20 cheng systemd[1]: ollama.service: Deactivated successfully. 8月 01 15:45:20 cheng systemd[1]: Stopped Ollama Service. 8月 01 15:45:20 cheng systemd[1]: ollama.service: Consumed 40.519s CPU time. -- Boot 33083ba5d27a4c399c75fd0c8948f208 -- 8月 01 19:27:21 cheng systemd[1]: Started Ollama Service. 8月 01 19:27:21 cheng ollama[1889]: Couldn't find '/usr/share/ollama/.ollama/id_ed25519'. Generating new private key. 8月 01 19:27:21 cheng ollama[1889]: Your new public key is: 8月 01 19:27:21 cheng ollama[1889]: ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIMnZh7qZa0+q9j5qj9n/YsCevtxy0d/bNe+D6kJ9tT6p 8月 01 19:27:21 cheng ollama[1889]: 2024/08/01 19:27:21 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:> 8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.102+08:00 level=INFO source=images.go:781 msg="total blobs: > 8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.102+08:00 level=INFO source=images.go:788 msg="total unused > 8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.102+08:00 level=INFO source=routes.go:1156 msg="Listening on> 8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.103+08:00 level=INFO source=payload.go:30 msg="extracting em> 8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.372+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l> 8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.373+08:00 level=INFO source=gpu.go:205 msg="looking for comp> 8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 20:04:56 cheng ollama[1889]: [GIN] 2024/08/01 - 20:04:56 | 200 | 363.118µs | 127.0.0.1 | HEAD "/" 8月 01 20:05:18 cheng ollama[1889]: [GIN] 2024/08/01 - 20:05:18 | 200 | 30.807µs | 127.0.0.1 | HEAD "/" 8月 01 20:06:07 cheng ollama[1889]: [GIN] 2024/08/01 - 20:06:07 | 201 | 30.463162163s | 127.0.0.1 | POST "/api/> 8月 01 20:07:01 cheng ollama[1889]: [GIN] 2024/08/01 - 20:07:01 | 201 | 31.905709158s | 127.0.0.1 | POST "/api/> 8月 01 20:07:56 cheng ollama[1889]: [GIN] 2024/08/01 - 20:07:56 | 201 | 31.758401406s | 127.0.0.1 | POST "/api/> 8月 01 20:08:52 cheng ollama[1889]: [GIN] 2024/08/01 - 20:08:52 | 201 | 33.256550159s | 127.0.0.1 | POST "/api/> 8月 01 20:09:48 cheng ollama[1889]: [GIN] 2024/08/01 - 20:09:48 | 201 | 31.697320718s | 127.0.0.1 | POST "/api/> 8月 01 20:10:45 cheng ollama[1889]: [GIN] 2024/08/01 - 20:10:45 | 201 | 30.805938252s | 127.0.0.1 | POST "/api/> 8月 01 20:11:44 cheng ollama[1889]: [GIN] 2024/08/01 - 20:11:44 | 201 | 33.848623115s | 127.0.0.1 | POST "/api/> 8月 01 20:12:50 cheng ollama[1889]: [GIN] 2024/08/01 - 20:12:50 | 201 | 40.17151261s | 127.0.0.1 | POST "/api/> 8月 01 20:13:40 cheng ollama[1889]: [GIN] 2024/08/01 - 20:13:40 | 201 | 29.724228454s | 127.0.0.1 | POST "/api/> 8月 01 20:18:33 cheng ollama[1889]: [GIN] 2024/08/01 - 20:18:33 | 200 | 31.811µs | 127.0.0.1 | HEAD "/" 8月 01 20:18:43 cheng ollama[1889]: [GIN] 2024/08/01 - 20:18:43 | 200 | 5m3s | 127.0.0.1 | POST "/api/> 8月 01 20:19:22 cheng ollama[1889]: [GIN] 2024/08/01 - 20:19:22 | 201 | 30.191676368s | 127.0.0.1 | POST "/api/> 8月 01 20:19:51 cheng ollama[1889]: [GIN] 2024/08/01 - 20:19:51 | 200 | 28.187662838s | 127.0.0.1 | POST "/api/> 8月 01 20:22:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:22:36 | 200 | 31.05µs | 127.0.0.1 | HEAD "/" 8月 01 20:23:33 cheng ollama[1889]: [GIN] 2024/08/01 - 20:23:33 | 201 | 40.226746855s | 127.0.0.1 | POST "/api/> 8月 01 20:24:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:24:36 | 201 | 41.01948605s | 127.0.0.1 | POST "/api/> 8月 01 20:25:32 cheng ollama[1889]: [GIN] 2024/08/01 - 20:25:32 | 201 | 34.131806305s | 127.0.0.1 | POST "/api/> 8月 01 20:26:27 cheng ollama[1889]: [GIN] 2024/08/01 - 20:26:27 | 201 | 35.313518063s | 127.0.0.1 | POST "/api/> 8月 01 20:27:26 cheng ollama[1889]: [GIN] 2024/08/01 - 20:27:26 | 201 | 39.101495054s | 127.0.0.1 | POST "/api/> 8月 01 20:28:19 cheng ollama[1889]: [GIN] 2024/08/01 - 20:28:19 | 201 | 32.1910056s | 127.0.0.1 | POST "/api/> 8月 01 20:29:12 cheng ollama[1889]: [GIN] 2024/08/01 - 20:29:12 | 201 | 32.608388797s | 127.0.0.1 | POST "/api/> 8月 01 20:30:11 cheng ollama[1889]: [GIN] 2024/08/01 - 20:30:11 | 201 | 32.208379126s | 127.0.0.1 | POST "/api/> 8月 01 20:31:13 cheng ollama[1889]: [GIN] 2024/08/01 - 20:31:13 | 201 | 33.420692973s | 127.0.0.1 | POST "/api/> 8月 01 20:36:43 cheng ollama[1889]: [GIN] 2024/08/01 - 20:36:43 | 200 | 5m29s | 127.0.0.1 | POST "/api/> 8月 01 20:49:17 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:17 | 200 | 34.271µs | 127.0.0.1 | HEAD "/" 8月 01 20:49:17 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:17 | 200 | 488.318µs | 127.0.0.1 | GET "/api/> 8月 01 20:49:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:36 | 200 | 29.398µs | 127.0.0.1 | HEAD "/" 8月 01 20:49:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:36 | 200 | 404.433µs | 127.0.0.1 | GET "/api/> 8月 01 20:49:41 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:41 | 200 | 59.252µs | 127.0.0.1 | HEAD "/" 8月 01 20:49:41 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:41 | 200 | 9.883069ms | 127.0.0.1 | POST "/api/> 8月 01 20:49:42 cheng ollama[1889]: time=2024-08-01T20:49:42.325+08:00 level=WARN source=memory.go:115 msg="model missing> 8月 01 20:49:42 cheng ollama[1889]: panic: runtime error: integer divide by zero 8月 01 20:49:42 cheng ollama[1889]: goroutine 13 [running]: 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000> 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm.PredictServerFit({0xc00074bbd8?, 0x4?, 0x4?}, _, {_, _, > 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc00016a700, 0xc000709780, > 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc000116ba0, {0x22aeab50> 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 01 20:49:42 cheng ollama[1889]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 01 20:49:42 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 01 20:49:42 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 01 20:49:42 cheng systemd[1]: ollama.service: Consumed 21min 49.078s CPU time. 8月 01 20:49:45 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1. 8月 01 20:49:45 cheng systemd[1]: Stopped Ollama Service. 8月 01 20:49:45 cheng systemd[1]: ollama.service: Consumed 21min 49.078s CPU time. 8月 01 20:49:45 cheng systemd[1]: Started Ollama Service. 8月 01 20:49:45 cheng ollama[28228]: Couldn't find '/usr/share/ollama/.ollama/id_ed25519'. Generating new private key. 8月 01 20:49:45 cheng ollama[28228]: Your new public key is: 8月 01 20:49:45 cheng ollama[28228]: ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAILxBQqiTdXAnV3aWrQ5ZtN1EpdtgIZKBEST7iqBy8u2p 8月 01 20:49:45 cheng ollama[28228]: 2024/08/01 20:49:45 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES> 8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.626+08:00 level=INFO source=images.go:781 msg="total blobs:> 8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.626+08:00 level=INFO source=images.go:788 msg="total unused> 8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.626+08:00 level=INFO source=routes.go:1156 msg="Listening o> 8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.635+08:00 level=WARN source=assets.go:100 msg="unable to cl> 8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.635+08:00 level=INFO source=payload.go:30 msg="extracting e> 8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.347+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM > 8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.347+08:00 level=INFO source=gpu.go:205 msg="looking for com> 8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference com> 8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference com> 8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference com> 8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference com> 8月 01 20:51:45 cheng ollama[28228]: [GIN] 2024/08/01 - 20:51:45 | 200 | 53.369µs | 127.0.0.1 | HEAD "/" 8月 01 20:51:45 cheng ollama[28228]: [GIN] 2024/08/01 - 20:51:45 | 200 | 1.366054ms | 127.0.0.1 | POST "/api> 8月 01 20:51:45 cheng ollama[28228]: time=2024-08-01T20:51:45.643+08:00 level=WARN source=memory.go:115 msg="model missin> 8月 01 20:51:45 cheng ollama[28228]: panic: runtime error: integer divide by zero 8月 01 20:51:45 cheng ollama[28228]: goroutine 66 [running]: 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x200> 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm.PredictServerFit({0xc000893bd8?, 0x4?, 0x4?}, _, {_, _,> 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc000558200, 0xc0003d8020,> 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc0003b3020, {0x22aeab5> 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 01 20:51:45 cheng ollama[28228]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 01 20:51:45 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 01 20:51:45 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 01 20:51:45 cheng systemd[1]: ollama.service: Consumed 19.730s CPU time. 8月 01 20:51:48 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 2. 8月 01 20:51:48 cheng systemd[1]: Stopped Ollama Service. 8月 01 20:51:48 cheng systemd[1]: ollama.service: Consumed 19.730s CPU time. 8月 01 20:51:48 cheng systemd[1]: Started Ollama Service. 8月 01 20:51:48 cheng ollama[28388]: 2024/08/01 20:51:48 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES> 8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=INFO source=images.go:781 msg="total blobs:> 8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=INFO source=images.go:788 msg="total unused> 8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=INFO source=routes.go:1156 msg="Listening o> 8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=WARN source=assets.go:100 msg="unable to cl> 8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.857+08:00 level=WARN source=assets.go:100 msg="unable to cl> 8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.857+08:00 level=INFO source=payload.go:30 msg="extracting e> 8月 01 20:51:50 cheng systemd[1]: Stopping Ollama Service... 8月 01 20:51:55 cheng ollama[28388]: time=2024-08-01T20:51:55.741+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM > 8月 01 20:51:55 cheng ollama[28388]: time=2024-08-01T20:51:55.741+08:00 level=INFO source=gpu.go:205 msg="looking for com> 8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference com> 8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference com> 8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference com> 8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference com> 8月 01 20:51:56 cheng systemd[1]: ollama.service: Deactivated successfully. 8月 01 20:51:56 cheng systemd[1]: Stopped Ollama Service. 8月 01 20:51:56 cheng systemd[1]: ollama.service: Consumed 19.925s CPU time. -- Boot 2c6de2a332144d9a8ee2924b08fb1feb -- 8月 01 20:53:23 cheng systemd[1]: Started Ollama Service. 8月 01 20:53:23 cheng ollama[1877]: 2024/08/01 20:53:23 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:> 8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.988+08:00 level=INFO source=images.go:781 msg="total blobs: > 8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.989+08:00 level=INFO source=images.go:788 msg="total unused > 8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.989+08:00 level=INFO source=routes.go:1156 msg="Listening on> 8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.990+08:00 level=INFO source=payload.go:30 msg="extracting em> 8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.110+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l> 8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.110+08:00 level=INFO source=gpu.go:205 msg="looking for comp> 8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.664+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.665+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.665+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.665+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 20:55:12 cheng ollama[1877]: [GIN] 2024/08/01 - 20:55:12 | 200 | 347.797µs | 127.0.0.1 | HEAD "/" 8月 01 20:55:12 cheng ollama[1877]: [GIN] 2024/08/01 - 20:55:12 | 200 | 1.771204ms | 127.0.0.1 | POST "/api/> 8月 01 20:55:12 cheng ollama[1877]: time=2024-08-01T20:55:12.930+08:00 level=WARN source=memory.go:115 msg="model missing> 8月 01 20:55:12 cheng ollama[1877]: panic: runtime error: integer divide by zero 8月 01 20:55:12 cheng ollama[1877]: goroutine 54 [running]: 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000> 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm.PredictServerFit({0xc00039bbd8?, 0x4?, 0x4?}, _, {_, _, > 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc0004b0b00, 0xc000037640, > 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc00043af00, {0x22aeab50> 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 01 20:55:12 cheng ollama[1877]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 01 20:55:12 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 01 20:55:12 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 01 20:55:12 cheng systemd[1]: ollama.service: Consumed 18.734s CPU time. 8月 01 20:55:16 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1. 8月 01 20:55:16 cheng systemd[1]: Stopped Ollama Service. 8月 01 20:55:16 cheng systemd[1]: ollama.service: Consumed 18.734s CPU time. 8月 01 20:55:16 cheng systemd[1]: Started Ollama Service. 8月 01 20:55:16 cheng ollama[4792]: 2024/08/01 20:55:16 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:> 8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.067+08:00 level=INFO source=images.go:781 msg="total blobs: > 8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.067+08:00 level=INFO source=images.go:788 msg="total unused > 8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.067+08:00 level=INFO source=routes.go:1156 msg="Listening on> 8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.068+08:00 level=WARN source=assets.go:100 msg="unable to cle> 8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.068+08:00 level=INFO source=payload.go:30 msg="extracting em> 8月 01 20:55:22 cheng ollama[4792]: time=2024-08-01T20:55:22.577+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l> 8月 01 20:55:22 cheng ollama[4792]: time=2024-08-01T20:55:22.577+08:00 level=INFO source=gpu.go:205 msg="looking for comp> 8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 20:56:40 cheng ollama[4792]: [GIN] 2024/08/01 - 20:56:40 | 200 | 58.151µs | 127.0.0.1 | HEAD "/" 8月 01 20:56:40 cheng ollama[4792]: [GIN] 2024/08/01 - 20:56:40 | 200 | 465.523µs | 127.0.0.1 | GET "/api/> 8月 01 20:59:09 cheng ollama[4792]: [GIN] 2024/08/01 - 20:59:09 | 200 | 50.178µs | 127.0.0.1 | HEAD "/" 8月 01 20:59:09 cheng ollama[4792]: [GIN] 2024/08/01 - 20:59:09 | 200 | 2.818931ms | 127.0.0.1 | POST "/api/> 8月 01 20:59:09 cheng ollama[4792]: time=2024-08-01T20:59:09.773+08:00 level=WARN source=memory.go:115 msg="model missing> 8月 01 20:59:09 cheng ollama[4792]: panic: runtime error: integer divide by zero 8月 01 20:59:09 cheng ollama[4792]: goroutine 66 [running]: 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000> 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm.PredictServerFit({0xc0004e1bd8?, 0x4?, 0x4?}, _, {_, _, > 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc000255100, 0xc0004943c0, > 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc0003b0f60, {0x22aeab50> 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 01 20:59:09 cheng ollama[4792]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 01 20:59:09 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 01 20:59:09 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 01 20:59:09 cheng systemd[1]: ollama.service: Consumed 20.184s CPU time. 8月 01 20:59:13 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 2. 8月 01 20:59:13 cheng systemd[1]: Stopped Ollama Service. 8月 01 20:59:13 cheng systemd[1]: ollama.service: Consumed 20.184s CPU time. 8月 01 20:59:13 cheng systemd[1]: Started Ollama Service. 8月 01 20:59:13 cheng ollama[5138]: 2024/08/01 20:59:13 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:> 8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.067+08:00 level=INFO source=images.go:781 msg="total blobs: > 8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=INFO source=images.go:788 msg="total unused > 8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=INFO source=routes.go:1156 msg="Listening on> 8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=WARN source=assets.go:100 msg="unable to cle> 8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=WARN source=assets.go:100 msg="unable to cle> 8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.069+08:00 level=INFO source=payload.go:30 msg="extracting em> 8月 01 20:59:19 cheng ollama[5138]: time=2024-08-01T20:59:19.727+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l> 8月 01 20:59:19 cheng ollama[5138]: time=2024-08-01T20:59:19.727+08:00 level=INFO source=gpu.go:205 msg="looking for comp> 8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 21:04:01 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:01 | 200 | 51.242µs | 127.0.0.1 | HEAD "/" 8月 01 21:04:01 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:01 | 200 | 442.689µs | 127.0.0.1 | GET "/api/> 8月 01 21:04:12 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:12 | 200 | 35.143µs | 127.0.0.1 | HEAD "/" 8月 01 21:04:12 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:12 | 200 | 1.312058ms | 127.0.0.1 | POST "/api/> 8月 01 21:04:13 cheng ollama[5138]: time=2024-08-01T21:04:13.208+08:00 level=WARN source=memory.go:115 msg="model missing> 8月 01 21:04:13 cheng ollama[5138]: panic: runtime error: integer divide by zero 8月 01 21:04:13 cheng ollama[5138]: goroutine 82 [running]: 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000> 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm.PredictServerFit({0xc000107bd8?, 0x4?, 0x4?}, _, {_, _, > 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc0003d2400, 0xc000037ea0, > 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc0003b0ea0, {0x22aeab50> 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 01 21:04:13 cheng ollama[5138]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 01 21:04:13 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 01 21:04:13 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 01 21:04:13 cheng systemd[1]: ollama.service: Consumed 19.781s CPU time. 8月 01 21:04:16 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 3. 8月 01 21:04:16 cheng systemd[1]: Stopped Ollama Service. 8月 01 21:04:16 cheng systemd[1]: ollama.service: Consumed 19.781s CPU time. 8月 01 21:04:16 cheng systemd[1]: Started Ollama Service. 8月 01 21:04:16 cheng ollama[5398]: 2024/08/01 21:04:16 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:> 8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.318+08:00 level=INFO source=images.go:781 msg="total blobs: > 8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.318+08:00 level=INFO source=images.go:788 msg="total unused > 8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=INFO source=routes.go:1156 msg="Listening on> 8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=WARN source=assets.go:100 msg="unable to cle> 8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=WARN source=assets.go:100 msg="unable to cle> 8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=WARN source=assets.go:100 msg="unable to cle> 8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.320+08:00 level=INFO source=payload.go:30 msg="extracting em> 8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.069+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l> 8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.069+08:00 level=INFO source=gpu.go:205 msg="looking for comp> 8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 01 21:04:23 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:23 | 404 | 86.36µs | 127.0.0.1 | GET "/api/> 8月 01 21:04:24 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:24 | 404 | 9.775µs | 127.0.0.1 | GET "/favi> 8月 01 21:04:38 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:38 | 404 | 13.876µs | 127.0.0.1 | GET "/api" 8月 01 21:04:41 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:41 | 200 | 27.389µs | 127.0.0.1 | GET "/" 8月 01 21:05:17 cheng ollama[5398]: [GIN] 2024/08/01 - 21:05:17 | 404 | 11.268µs | 127.0.0.1 | GET "/api/> 8月 01 21:09:25 cheng ollama[5398]: [GIN] 2024/08/01 - 21:09:25 | 200 | 141.497µs | 127.0.0.1 | GET "/" 8月 01 21:13:26 cheng ollama[5398]: [GIN] 2024/08/01 - 21:13:26 | 200 | 70.515µs | 127.0.0.1 | GET "/" 8月 01 21:15:11 cheng ollama[5398]: [GIN] 2024/08/01 - 21:15:11 | 200 | 56.712µs | 127.0.0.1 | GET "/api/> 8月 01 21:17:26 cheng systemd[1]: Stopping Ollama Service... 8月 01 21:17:26 cheng systemd[1]: ollama.service: Deactivated successfully. 8月 01 21:17:26 cheng systemd[1]: Stopped Ollama Service. 8月 01 21:17:26 cheng systemd[1]: ollama.service: Consumed 20.123s CPU time. -- Boot 9bceaafe62e14190ba2fd13cfe4b660c -- 8月 02 07:54:47 cheng systemd[1]: Started Ollama Service. 8月 02 07:54:47 cheng ollama[1896]: 2024/08/02 07:54:47 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:> 8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.944+08:00 level=INFO source=images.go:781 msg="total blobs: > 8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.944+08:00 level=INFO source=images.go:788 msg="total unused > 8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.945+08:00 level=INFO source=routes.go:1156 msg="Listening on> 8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.946+08:00 level=INFO source=payload.go:30 msg="extracting em> 8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.037+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l> 8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.037+08:00 level=INFO source=gpu.go:205 msg="looking for comp> 8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 08:02:01 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:01 | 200 | 399.071µs | 127.0.0.1 | HEAD "/" 8月 02 08:02:01 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:01 | 200 | 1.195064ms | 127.0.0.1 | GET "/api/> 8月 02 08:02:12 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:12 | 200 | 44.049µs | 127.0.0.1 | HEAD "/" 8月 02 08:02:12 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:12 | 200 | 1.793469ms | 127.0.0.1 | POST "/api/> 8月 02 08:02:13 cheng ollama[1896]: time=2024-08-02T08:02:13.426+08:00 level=WARN source=memory.go:115 msg="model missing> 8月 02 08:02:13 cheng ollama[1896]: panic: runtime error: integer divide by zero 8月 02 08:02:13 cheng ollama[1896]: goroutine 70 [running]: 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000> 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm.PredictServerFit({0xc00040dbd8?, 0x4?, 0x4?}, _, {_, _, > 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc0001d9900, 0xc000037660, > 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc0000cf080, {0x22aeab50> 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 02 08:02:13 cheng ollama[1896]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 02 08:02:13 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 02 08:02:13 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 02 08:02:13 cheng systemd[1]: ollama.service: Consumed 17.958s CPU time. 8月 02 08:02:16 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1. 8月 02 08:02:16 cheng systemd[1]: Stopped Ollama Service. 8月 02 08:02:16 cheng systemd[1]: ollama.service: Consumed 17.958s CPU time. 8月 02 08:02:16 cheng systemd[1]: Started Ollama Service. 8月 02 08:02:16 cheng ollama[4728]: 2024/08/02 08:02:16 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:> 8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=INFO source=images.go:781 msg="total blobs: > 8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=INFO source=images.go:788 msg="total unused > 8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=INFO source=routes.go:1156 msg="Listening on> 8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=WARN source=assets.go:100 msg="unable to cle> 8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.580+08:00 level=INFO source=payload.go:30 msg="extracting em> 8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.278+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l> 8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.278+08:00 level=INFO source=gpu.go:205 msg="looking for comp> 8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 08:08:06 cheng systemd[1]: Stopping Ollama Service... 8月 02 08:08:06 cheng systemd[1]: ollama.service: Deactivated successfully. 8月 02 08:08:06 cheng systemd[1]: Stopped Ollama Service. 8月 02 08:08:06 cheng systemd[1]: ollama.service: Consumed 20.017s CPU time. -- Boot 88a56500e37f4b6eaaaeaca00780c960 -- 8月 02 08:10:53 cheng systemd[1]: Started Ollama Service. 8月 02 08:10:53 cheng ollama[1889]: 2024/08/02 08:10:53 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:> 8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.056+08:00 level=INFO source=images.go:781 msg="total blobs: > 8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.057+08:00 level=INFO source=images.go:788 msg="total unused > 8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.057+08:00 level=INFO source=routes.go:1156 msg="Listening on> 8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.057+08:00 level=INFO source=payload.go:30 msg="extracting em> 8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.295+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l> 8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.295+08:00 level=INFO source=gpu.go:205 msg="looking for comp> 8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 08:15:44 cheng ollama[1889]: [GIN] 2024/08/02 - 08:15:44 | 200 | 386.787µs | 127.0.0.1 | HEAD "/" 8月 02 08:15:44 cheng ollama[1889]: [GIN] 2024/08/02 - 08:15:44 | 200 | 3.462854ms | 127.0.0.1 | POST "/api/> 8月 02 08:15:44 cheng ollama[1889]: time=2024-08-02T08:15:44.622+08:00 level=WARN source=memory.go:115 msg="model missing> 8月 02 08:15:44 cheng ollama[1889]: panic: runtime error: integer divide by zero 8月 02 08:15:44 cheng ollama[1889]: goroutine 68 [running]: 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000> 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm.PredictServerFit({0xc0008a9bd8?, 0x4?, 0x4?}, _, {_, _, > 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc00046c200, 0xc00037c6a0, > 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc00043af60, {0x22aeab50> 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 02 08:15:44 cheng ollama[1889]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 02 08:15:44 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 02 08:15:44 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 02 08:15:44 cheng systemd[1]: ollama.service: Consumed 18.710s CPU time. 8月 02 08:15:47 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1. 8月 02 08:15:47 cheng systemd[1]: Stopped Ollama Service. 8月 02 08:15:47 cheng systemd[1]: ollama.service: Consumed 18.710s CPU time. 8月 02 08:15:47 cheng systemd[1]: Started Ollama Service. 8月 02 08:15:47 cheng ollama[4056]: 2024/08/02 08:15:47 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES:> 8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=INFO source=images.go:781 msg="total blobs: > 8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=INFO source=images.go:788 msg="total unused > 8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=INFO source=routes.go:1156 msg="Listening on> 8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=WARN source=assets.go:100 msg="unable to cle> 8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.845+08:00 level=INFO source=payload.go:30 msg="extracting em> 8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.330+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM l> 8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.330+08:00 level=INFO source=gpu.go:205 msg="looking for comp> 8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference comp> 8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference comp> lines 637-671/671 (END)
GiteaMirror commented 2026-04-12 14:39:22 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Aug 2, 2024):

There is missing information because of the pager, run journalctl -u ollama --no-pager. When you paste it in, use triple-backticks (```) before and after the paste. If the output is very large, perhaps add as an attachment.

<!-- gh-comment-id:2264271012 --> @rick-github commented on GitHub (Aug 2, 2024): There is missing information because of the pager, run `journalctl -u ollama --no-pager`. When you paste it in, use triple-backticks (```) before and after the paste. If the output is very large, perhaps add as an attachment.
GiteaMirror commented 2026-04-12 14:39:23 -05:00
Author
Owner
Copy Link

@SongXiaoMao commented on GitHub (Aug 2, 2024):

'''(base) cheng@cheng:$ ollama run qwen
Error: Post "http://127.0.0.1:11434/api/chat": EOF
(base) cheng@cheng:$ journalctl -u ollama --no-pager
8月 01 15:09:29 cheng systemd[1]: Started Ollama Service.
8月 01 15:09:29 cheng ollama[31790]: Couldn't find '/usr/share/ollama/.ollama/id_ed25519'. Generating new private key.
8月 01 15:09:29 cheng ollama[31790]: Your new public key is:
8月 01 15:09:29 cheng ollama[31790]: ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAILEdfF57fBFLPZGwkUEVkasIBOjp34OgZqC6P5gOckLi
8月 01 15:09:29 cheng ollama[31790]: 2024/08/01 15:09:29 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.826+08:00 level=INFO source=images.go:781 msg="total blobs: 0"
8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.826+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0"
8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.826+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)"
8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.827+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama2755471342/runners
8月 01 15:09:36 cheng ollama[31790]: time=2024-08-01T15:09:36.397+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11 rocm_v60102]"
8月 01 15:09:36 cheng ollama[31790]: time=2024-08-01T15:09:36.397+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs"
8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 15:20:48 cheng ollama[31790]: [GIN] 2024/08/01 - 15:20:48 | 200 | 94.041µs | 127.0.0.1 | HEAD "/"
8月 01 15:21:00 cheng ollama[31790]: [GIN] 2024/08/01 - 15:21:00 | 200 | 32.908µs | 127.0.0.1 | HEAD "/"
8月 01 15:21:33 cheng ollama[31790]: [GIN] 2024/08/01 - 15:21:33 | 200 | 61.512µs | 127.0.0.1 | HEAD "/"
8月 01 15:21:50 cheng ollama[31790]: [GIN] 2024/08/01 - 15:21:50 | 200 | 48.418µs | 127.0.0.1 | HEAD "/"
8月 01 15:23:25 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:25 | 200 | 32.858µs | 127.0.0.1 | HEAD "/"
8月 01 15:23:25 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:25 | 200 | 301.909µs | 127.0.0.1 | GET "/api/tags"
8月 01 15:23:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:40 | 200 | 30.281µs | 127.0.0.1 | HEAD "/"
8月 01 15:23:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:40 | 200 | 87.991µs | 127.0.0.1 | GET "/api/tags"
8月 01 15:23:47 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:47 | 200 | 29.357µs | 127.0.0.1 | HEAD "/"
8月 01 15:23:47 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:47 | 404 | 223.898µs | 127.0.0.1 | POST "/api/show"
8月 01 15:23:51 cheng ollama[31790]: time=2024-08-01T15:23:51.495+08:00 level=INFO source=download.go:173 msg="downloading 46bb65206e0e in 24 100 MB part(s)"
8月 01 15:23:55 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:55 | 200 | 7.982914121s | 127.0.0.1 | POST "/api/pull"
8月 01 15:25:17 cheng ollama[31790]: [GIN] 2024/08/01 - 15:25:17 | 200 | 61.437µs | 127.0.0.1 | HEAD "/"
8月 01 15:26:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:26:56 | 200 | 29.932µs | 127.0.0.1 | HEAD "/"
8月 01 15:27:14 cheng ollama[31790]: [GIN] 2024/08/01 - 15:27:14 | 200 | 33.123µs | 127.0.0.1 | HEAD "/"
8月 01 15:28:51 cheng ollama[31790]: [GIN] 2024/08/01 - 15:28:51 | 200 | 34.332µs | 127.0.0.1 | HEAD "/"
8月 01 15:29:39 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:39 | 200 | 34.374µs | 127.0.0.1 | HEAD "/"
8月 01 15:29:52 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:52 | 200 | 30.588µs | 127.0.0.1 | HEAD "/"
8月 01 15:29:54 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:54 | 201 | 1.639126644s | 127.0.0.1 | POST "/api/blobs/sha256:7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306"
8月 01 15:29:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:56 | 200 | 1.677066999s | 127.0.0.1 | POST "/api/create"
8月 01 15:30:01 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:01 | 200 | 35.043µs | 127.0.0.1 | HEAD "/"
8月 01 15:30:01 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:01 | 200 | 381.711µs | 127.0.0.1 | GET "/api/tags"
8月 01 15:30:07 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:07 | 200 | 31.294µs | 127.0.0.1 | HEAD "/"
8月 01 15:30:07 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:07 | 200 | 45.74282ms | 127.0.0.1 | POST "/api/show"
8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.622+08:00 level=INFO source=sched.go:710 msg="new model will fit in available VRAM in single GPU, loading" model=/usr/share/ollama/.ollama/models/blobs/sha256-7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306 gpu=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 parallel=4 available=25155272704 required="1.6 GiB"
8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.623+08:00 level=INFO source=memory.go:309 msg="offload to cuda" layers.requested=-1 layers.model=29 layers.offload=29 layers.split="" memory.available="[23.4 GiB]" memory.required.full="1.6 GiB" memory.required.partial="1.6 GiB" memory.required.kv="224.0 MiB" memory.required.allocations="[1.6 GiB]" memory.weights.total="680.7 MiB" memory.weights.repeating="498.1 MiB" memory.weights.nonrepeating="182.6 MiB" memory.graph.full="299.8 MiB" memory.graph.partial="482.3 MiB"
8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.624+08:00 level=INFO source=server.go:384 msg="starting llama server" cmd="/tmp/ollama2755471342/runners/cuda_v11/ollama_llama_server --model /usr/share/ollama/.ollama/models/blobs/sha256-7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306 --ctx-size 8192 --batch-size 512 --embedding --log-disable --n-gpu-layers 29 --parallel 4 --port 39233"
8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.625+08:00 level=INFO source=sched.go:445 msg="loaded runners" count=1
8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.625+08:00 level=INFO source=server.go:584 msg="waiting for llama runner to start responding"
8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.625+08:00 level=INFO source=server.go:618 msg="waiting for server to become available" status="llm server error"
8月 01 15:30:07 cheng ollama[38287]: INFO [main] build info | build=1 commit="6eeaeba" tid="133149442916352" timestamp=1722497407
8月 01 15:30:07 cheng ollama[38287]: INFO [main] system info | n_threads=12 n_threads_batch=-1 system_info="AVX = 1 | AVX_VNNI = 0 | AVX2 = 0 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | AVX512_BF16 = 0 | FMA = 0 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 0 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | " tid="133149442916352" timestamp=1722497407 total_threads=24
8月 01 15:30:07 cheng ollama[38287]: INFO [main] HTTP server listening | hostname="127.0.0.1" n_threads_http="23" port="39233" tid="133149442916352" timestamp=1722497407
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: loaded meta data with 26 key-value pairs and 338 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306 (version GGUF V3 (latest))
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 0: general.architecture str = qwen2
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 1: general.name str = qwen2-1_5b-instruct
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 2: qwen2.block_count u32 = 28
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 3: qwen2.context_length u32 = 32768
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 4: qwen2.embedding_length u32 = 1536
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 5: qwen2.feed_forward_length u32 = 8960
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 6: qwen2.attention.head_count u32 = 12
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 = 2
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 8: qwen2.rope.freq_base f32 = 1000000.000000
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000001
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 10: general.file_type u32 = 10
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 11: tokenizer.ggml.model str = gpt2
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 12: tokenizer.ggml.pre str = qwen2
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151936] = ["!", """, "#", "$", "%", "&", "'", ...
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151936] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 = 151645
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 = 151643
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 151643
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 19: tokenizer.chat_template str = {% for message in messages %}{% if lo...
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 20: tokenizer.ggml.add_bos_token bool = false
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 21: general.quantization_version u32 = 2
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 22: quantize.imatrix.file str = ../Qwen2/gguf/qwen2-1_5b-imatrix/imat...
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 23: quantize.imatrix.dataset str = ../sft_2406.txt
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 24: quantize.imatrix.entries_count i32 = 196
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 25: quantize.imatrix.chunks_count i32 = 1937
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type f32: 141 tensors
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q2_K: 112 tensors
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q3_K: 56 tensors
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q4_K: 28 tensors
8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q6_K: 1 tensors
8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.878+08:00 level=INFO source=server.go:618 msg="waiting for server to become available" status="llm server loading model"
8月 01 15:30:08 cheng ollama[31790]: llm_load_vocab: special tokens cache size = 293
8月 01 15:30:08 cheng ollama[31790]: llm_load_vocab: token to piece cache size = 0.9338 MB
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: format = GGUF V3 (latest)
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: arch = qwen2
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: vocab type = BPE
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_vocab = 151936
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_merges = 151387
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: vocab_only = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_ctx_train = 32768
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd = 1536
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_layer = 28
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_head = 12
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_head_kv = 2
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_rot = 128
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_swa = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_head_k = 128
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_head_v = 128
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_gqa = 6
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_k_gqa = 256
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_v_gqa = 256
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_norm_eps = 0.0e+00
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_norm_rms_eps = 1.0e-06
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_clamp_kqv = 0.0e+00
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_max_alibi_bias = 0.0e+00
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_logit_scale = 0.0e+00
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_ff = 8960
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_expert = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_expert_used = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: causal attn = 1
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: pooling type = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: rope type = 2
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: rope scaling = linear
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: freq_base_train = 1000000.0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: freq_scale_train = 1
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_ctx_orig_yarn = 32768
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: rope_finetuned = unknown
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_d_conv = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_d_inner = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_d_state = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_dt_rank = 0
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model type = ?B
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model ftype = Q2_K - Medium
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model params = 1.54 B
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model size = 639.30 MiB (3.47 BPW)
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: general.name = qwen2-1_5b-instruct
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: BOS token = 151643 '<|endoftext|>'
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: EOS token = 151645 '<|im_end|>'
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: PAD token = 151643 '<|endoftext|>'
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: LF token = 148848 'ÄĬ'
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: EOT token = 151645 '<|im_end|>'
8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: max token length = 256
8月 01 15:30:08 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
8月 01 15:30:08 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
8月 01 15:30:08 cheng ollama[31790]: ggml_cuda_init: found 1 CUDA devices:
8月 01 15:30:08 cheng ollama[31790]: Device 0: NVIDIA GeForce RTX 3090, compute capability 8.6, VMM: yes
8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: ggml ctx size = 0.30 MiB
8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: offloading 28 repeating layers to GPU
8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: offloading non-repeating layers to GPU
8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: offloaded 29/29 layers to GPU
8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: CPU buffer size = 182.57 MiB
8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: CUDA0 buffer size = 639.30 MiB
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: n_ctx = 8192
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: n_batch = 512
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: n_ubatch = 512
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: flash_attn = 0
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: freq_base = 1000000.0
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: freq_scale = 1
8月 01 15:30:08 cheng ollama[31790]: llama_kv_cache_init: CUDA0 KV buffer size = 224.00 MiB
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: KV self size = 224.00 MiB, K (f16): 112.00 MiB, V (f16): 112.00 MiB
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host output buffer size = 2.34 MiB
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: CUDA0 compute buffer size = 299.75 MiB
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host compute buffer size = 19.01 MiB
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: graph nodes = 986
8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: graph splits = 2
8月 01 15:30:09 cheng ollama[38287]: INFO [main] model loaded | tid="133149442916352" timestamp=1722497409
8月 01 15:30:10 cheng ollama[31790]: time=2024-08-01T15:30:10.139+08:00 level=INFO source=server.go:623 msg="llama runner started in 2.51 seconds"
8月 01 15:30:10 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:10 | 200 | 3.085087137s | 127.0.0.1 | POST "/api/chat"
8月 01 15:30:14 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:14 | 200 | 180.746309ms | 127.0.0.1 | POST "/api/chat"
8月 01 15:36:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:36:56 | 200 | 31.499µs | 127.0.0.1 | HEAD "/"
8月 01 15:36:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:36:56 | 200 | 194.946µs | 127.0.0.1 | GET "/api/tags"
8月 01 15:37:02 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:02 | 200 | 65.7µs | 127.0.0.1 | HEAD "/"
8月 01 15:37:02 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:02 | 200 | 87.954µs | 127.0.0.1 | GET "/api/tags"
8月 01 15:37:31 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:31 | 200 | 33.094µs | 127.0.0.1 | HEAD "/"
8月 01 15:37:33 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:33 | 201 | 1.651416565s | 127.0.0.1 | POST "/api/blobs/sha256:7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306"
8月 01 15:37:35 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:35 | 200 | 1.725709949s | 127.0.0.1 | POST "/api/create"
8月 01 15:37:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:40 | 200 | 28.411µs | 127.0.0.1 | HEAD "/"
8月 01 15:37:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:40 | 200 | 355.484µs | 127.0.0.1 | GET "/api/tags"
8月 01 15:38:10 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:10 | 200 | 60.68µs | 127.0.0.1 | HEAD "/"
8月 01 15:38:10 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:10 | 200 | 22.792904ms | 127.0.0.1 | POST "/api/show"
8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.163+08:00 level=INFO source=sched.go:710 msg="new model will fit in available VRAM in single GPU, loading" model=/usr/share/ollama/.ollama/models/blobs/sha256-7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306 gpu=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 parallel=4 available=25155272704 required="1.6 GiB"
8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.164+08:00 level=INFO source=memory.go:309 msg="offload to cuda" layers.requested=-1 layers.model=29 layers.offload=29 layers.split="" memory.available="[23.4 GiB]" memory.required.full="1.6 GiB" memory.required.partial="1.6 GiB" memory.required.kv="224.0 MiB" memory.required.allocations="[1.6 GiB]" memory.weights.total="680.7 MiB" memory.weights.repeating="498.1 MiB" memory.weights.nonrepeating="182.6 MiB" memory.graph.full="299.8 MiB" memory.graph.partial="482.3 MiB"
8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.164+08:00 level=INFO source=server.go:384 msg="starting llama server" cmd="/tmp/ollama2755471342/runners/cuda_v11/ollama_llama_server --model /usr/share/ollama/.ollama/models/blobs/sha256-7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306 --ctx-size 8192 --batch-size 512 --embedding --log-disable --n-gpu-layers 29 --parallel 4 --port 39549"
8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.165+08:00 level=INFO source=sched.go:445 msg="loaded runners" count=1
8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.165+08:00 level=INFO source=server.go:584 msg="waiting for llama runner to start responding"
8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.165+08:00 level=INFO source=server.go:618 msg="waiting for server to become available" status="llm server error"
8月 01 15:38:11 cheng ollama[41534]: INFO [main] build info | build=1 commit="6eeaeba" tid="125225824997376" timestamp=1722497891
8月 01 15:38:11 cheng ollama[41534]: INFO [main] system info | n_threads=12 n_threads_batch=-1 system_info="AVX = 1 | AVX_VNNI = 0 | AVX2 = 0 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | AVX512_BF16 = 0 | FMA = 0 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 0 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | " tid="125225824997376" timestamp=1722497891 total_threads=24
8月 01 15:38:11 cheng ollama[41534]: INFO [main] HTTP server listening | hostname="127.0.0.1" n_threads_http="23" port="39549" tid="125225824997376" timestamp=1722497891
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: loaded meta data with 26 key-value pairs and 338 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306 (version GGUF V3 (latest))
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 0: general.architecture str = qwen2
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 1: general.name str = qwen2-1_5b-instruct
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 2: qwen2.block_count u32 = 28
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 3: qwen2.context_length u32 = 32768
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 4: qwen2.embedding_length u32 = 1536
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 5: qwen2.feed_forward_length u32 = 8960
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 6: qwen2.attention.head_count u32 = 12
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 = 2
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 8: qwen2.rope.freq_base f32 = 1000000.000000
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000001
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 10: general.file_type u32 = 10
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 11: tokenizer.ggml.model str = gpt2
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 12: tokenizer.ggml.pre str = qwen2
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151936] = ["!", """, "#", "$", "%", "&", "'", ...
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151936] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 = 151645
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 = 151643
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 151643
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 19: tokenizer.chat_template str = {% for message in messages %}{% if lo...
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 20: tokenizer.ggml.add_bos_token bool = false
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 21: general.quantization_version u32 = 2
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 22: quantize.imatrix.file str = ../Qwen2/gguf/qwen2-1_5b-imatrix/imat...
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 23: quantize.imatrix.dataset str = ../sft_2406.txt
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 24: quantize.imatrix.entries_count i32 = 196
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 25: quantize.imatrix.chunks_count i32 = 1937
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type f32: 141 tensors
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q2_K: 112 tensors
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q3_K: 56 tensors
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q4_K: 28 tensors
8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q6_K: 1 tensors
8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.417+08:00 level=INFO source=server.go:618 msg="waiting for server to become available" status="llm server loading model"
8月 01 15:38:11 cheng ollama[31790]: llm_load_vocab: special tokens cache size = 293
8月 01 15:38:11 cheng ollama[31790]: llm_load_vocab: token to piece cache size = 0.9338 MB
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: format = GGUF V3 (latest)
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: arch = qwen2
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: vocab type = BPE
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_vocab = 151936
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_merges = 151387
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: vocab_only = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_ctx_train = 32768
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd = 1536
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_layer = 28
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_head = 12
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_head_kv = 2
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_rot = 128
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_swa = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_head_k = 128
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_head_v = 128
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_gqa = 6
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_k_gqa = 256
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_v_gqa = 256
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_norm_eps = 0.0e+00
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_norm_rms_eps = 1.0e-06
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_clamp_kqv = 0.0e+00
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_max_alibi_bias = 0.0e+00
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_logit_scale = 0.0e+00
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_ff = 8960
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_expert = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_expert_used = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: causal attn = 1
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: pooling type = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: rope type = 2
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: rope scaling = linear
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: freq_base_train = 1000000.0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: freq_scale_train = 1
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_ctx_orig_yarn = 32768
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: rope_finetuned = unknown
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_d_conv = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_d_inner = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_d_state = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_dt_rank = 0
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model type = ?B
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model ftype = Q2_K - Medium
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model params = 1.54 B
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model size = 639.30 MiB (3.47 BPW)
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: general.name = qwen2-1_5b-instruct
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: BOS token = 151643 '<|endoftext|>'
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: EOS token = 151645 '<|im_end|>'
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: PAD token = 151643 '<|endoftext|>'
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: LF token = 148848 'ÄĬ'
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: EOT token = 151645 '<|im_end|>'
8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: max token length = 256
8月 01 15:38:11 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
8月 01 15:38:11 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
8月 01 15:38:11 cheng ollama[31790]: ggml_cuda_init: found 1 CUDA devices:
8月 01 15:38:11 cheng ollama[31790]: Device 0: NVIDIA GeForce RTX 3090, compute capability 8.6, VMM: yes
8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: ggml ctx size = 0.30 MiB
8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: offloading 28 repeating layers to GPU
8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: offloading non-repeating layers to GPU
8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: offloaded 29/29 layers to GPU
8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: CPU buffer size = 182.57 MiB
8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: CUDA0 buffer size = 639.30 MiB
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: n_ctx = 8192
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: n_batch = 512
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: n_ubatch = 512
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: flash_attn = 0
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: freq_base = 1000000.0
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: freq_scale = 1
8月 01 15:38:12 cheng ollama[31790]: llama_kv_cache_init: CUDA0 KV buffer size = 224.00 MiB
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: KV self size = 224.00 MiB, K (f16): 112.00 MiB, V (f16): 112.00 MiB
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host output buffer size = 2.34 MiB
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: CUDA0 compute buffer size = 299.75 MiB
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host compute buffer size = 19.01 MiB
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: graph nodes = 986
8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: graph splits = 2
8月 01 15:38:13 cheng ollama[41534]: INFO [main] model loaded | tid="125225824997376" timestamp=1722497893
8月 01 15:38:13 cheng ollama[31790]: time=2024-08-01T15:38:13.428+08:00 level=INFO source=server.go:623 msg="llama runner started in 2.26 seconds"
8月 01 15:38:13 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:13 | 200 | 2.81848391s | 127.0.0.1 | POST "/api/chat"
8月 01 15:38:16 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:16 | 200 | 131.417236ms | 127.0.0.1 | POST "/api/chat"
8月 01 15:38:18 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:18 | 200 | 337.686697ms | 127.0.0.1 | POST "/api/chat"
8月 01 15:38:19 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:19 | 200 | 528.190585ms | 127.0.0.1 | POST "/api/chat"
8月 01 15:38:21 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:21 | 200 | 744.627142ms | 127.0.0.1 | POST "/api/chat"
8月 01 15:41:00 cheng ollama[31790]: [GIN] 2024/08/01 - 15:41:00 | 200 | 31.118µs | 127.0.0.1 | HEAD "/"
8月 01 15:41:00 cheng ollama[31790]: [GIN] 2024/08/01 - 15:41:00 | 200 | 218.55µs | 127.0.0.1 | GET "/api/tags"
8月 01 15:45:20 cheng systemd[1]: Stopping Ollama Service...
8月 01 15:45:20 cheng systemd[1]: ollama.service: Deactivated successfully.
8月 01 15:45:20 cheng systemd[1]: Stopped Ollama Service.
8月 01 15:45:20 cheng systemd[1]: ollama.service: Consumed 40.519s CPU time.
-- Boot 33083ba5d27a4c399c75fd0c8948f208 --
8月 01 19:27:21 cheng systemd[1]: Started Ollama Service.
8月 01 19:27:21 cheng ollama[1889]: Couldn't find '/usr/share/ollama/.ollama/id_ed25519'. Generating new private key.
8月 01 19:27:21 cheng ollama[1889]: Your new public key is:
8月 01 19:27:21 cheng ollama[1889]: ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIMnZh7qZa0+q9j5qj9n/YsCevtxy0d/bNe+D6kJ9tT6p
8月 01 19:27:21 cheng ollama[1889]: 2024/08/01 19:27:21 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.102+08:00 level=INFO source=images.go:781 msg="total blobs: 0"
8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.102+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0"
8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.102+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)"
8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.103+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama2457549556/runners
8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.372+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu_avx cpu_avx2 cuda_v11 rocm_v60102 cpu]"
8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.373+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs"
8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:04:56 cheng ollama[1889]: [GIN] 2024/08/01 - 20:04:56 | 200 | 363.118µs | 127.0.0.1 | HEAD "/"
8月 01 20:05:18 cheng ollama[1889]: [GIN] 2024/08/01 - 20:05:18 | 200 | 30.807µs | 127.0.0.1 | HEAD "/"
8月 01 20:06:07 cheng ollama[1889]: [GIN] 2024/08/01 - 20:06:07 | 201 | 30.463162163s | 127.0.0.1 | POST "/api/blobs/sha256:84e82bd14926b67751d38f2a4fb0bf9e48c73bbadaaad5a1e7742b98cea5a0ca"
8月 01 20:07:01 cheng ollama[1889]: [GIN] 2024/08/01 - 20:07:01 | 201 | 31.905709158s | 127.0.0.1 | POST "/api/blobs/sha256:3d25eef820731f8bc2d4fda6dfb241573c096ac672d3fb98b78ceb393dc2c84f"
8月 01 20:07:56 cheng ollama[1889]: [GIN] 2024/08/01 - 20:07:56 | 201 | 31.758401406s | 127.0.0.1 | POST "/api/blobs/sha256:6e8f1d80f216dfb32d57e617de7d8c4e0f58513edf026ec614626848556394eb"
8月 01 20:08:52 cheng ollama[1889]: [GIN] 2024/08/01 - 20:08:52 | 201 | 33.256550159s | 127.0.0.1 | POST "/api/blobs/sha256:cd2156d736336dcd878bd91ab823fd8b5e6c8ecae951fa6a34894b00ad0ca2f6"
8月 01 20:09:48 cheng ollama[1889]: [GIN] 2024/08/01 - 20:09:48 | 201 | 31.697320718s | 127.0.0.1 | POST "/api/blobs/sha256:ffa8fc60b1df9a0195de7c3171b89e0df69586d38466b42b8528aa0c06392e93"
8月 01 20:10:45 cheng ollama[1889]: [GIN] 2024/08/01 - 20:10:45 | 201 | 30.805938252s | 127.0.0.1 | POST "/api/blobs/sha256:bf6f42aaf95fb86a14a1b74ada6ac2780b38ed358d707f611b7f69c2ab37a494"
8月 01 20:11:44 cheng ollama[1889]: [GIN] 2024/08/01 - 20:11:44 | 201 | 33.848623115s | 127.0.0.1 | POST "/api/blobs/sha256:ede18cfa7e112bb2a9d9e5e148b9ff7ebb95279c1c5e41d4a20506a4b61b6f03"
8月 01 20:12:50 cheng ollama[1889]: [GIN] 2024/08/01 - 20:12:50 | 201 | 40.17151261s | 127.0.0.1 | POST "/api/blobs/sha256:296fdee32a474a1752805609620005d68907034d8b48f89381951cf85bdcc9bd"
8月 01 20:13:40 cheng ollama[1889]: [GIN] 2024/08/01 - 20:13:40 | 201 | 29.724228454s | 127.0.0.1 | POST "/api/blobs/sha256:e1266b80a6ca3779079d8c9bdf1d1cf777f5c5c75201f4f567b03f6990ee8e7d"
8月 01 20:18:33 cheng ollama[1889]: [GIN] 2024/08/01 - 20:18:33 | 200 | 31.811µs | 127.0.0.1 | HEAD "/"
8月 01 20:18:43 cheng ollama[1889]: [GIN] 2024/08/01 - 20:18:43 | 200 | 5m3s | 127.0.0.1 | POST "/api/create"
8月 01 20:19:22 cheng ollama[1889]: [GIN] 2024/08/01 - 20:19:22 | 201 | 30.191676368s | 127.0.0.1 | POST "/api/blobs/sha256:84e82bd14926b67751d38f2a4fb0bf9e48c73bbadaaad5a1e7742b98cea5a0ca"
8月 01 20:19:51 cheng ollama[1889]: [GIN] 2024/08/01 - 20:19:51 | 200 | 28.187662838s | 127.0.0.1 | POST "/api/create"
8月 01 20:22:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:22:36 | 200 | 31.05µs | 127.0.0.1 | HEAD "/"
8月 01 20:23:33 cheng ollama[1889]: [GIN] 2024/08/01 - 20:23:33 | 201 | 40.226746855s | 127.0.0.1 | POST "/api/blobs/sha256:84e82bd14926b67751d38f2a4fb0bf9e48c73bbadaaad5a1e7742b98cea5a0ca"
8月 01 20:24:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:24:36 | 201 | 41.01948605s | 127.0.0.1 | POST "/api/blobs/sha256:3d25eef820731f8bc2d4fda6dfb241573c096ac672d3fb98b78ceb393dc2c84f"
8月 01 20:25:32 cheng ollama[1889]: [GIN] 2024/08/01 - 20:25:32 | 201 | 34.131806305s | 127.0.0.1 | POST "/api/blobs/sha256:6e8f1d80f216dfb32d57e617de7d8c4e0f58513edf026ec614626848556394eb"
8月 01 20:26:27 cheng ollama[1889]: [GIN] 2024/08/01 - 20:26:27 | 201 | 35.313518063s | 127.0.0.1 | POST "/api/blobs/sha256:cd2156d736336dcd878bd91ab823fd8b5e6c8ecae951fa6a34894b00ad0ca2f6"
8月 01 20:27:26 cheng ollama[1889]: [GIN] 2024/08/01 - 20:27:26 | 201 | 39.101495054s | 127.0.0.1 | POST "/api/blobs/sha256:ffa8fc60b1df9a0195de7c3171b89e0df69586d38466b42b8528aa0c06392e93"
8月 01 20:28:19 cheng ollama[1889]: [GIN] 2024/08/01 - 20:28:19 | 201 | 32.1910056s | 127.0.0.1 | POST "/api/blobs/sha256:bf6f42aaf95fb86a14a1b74ada6ac2780b38ed358d707f611b7f69c2ab37a494"
8月 01 20:29:12 cheng ollama[1889]: [GIN] 2024/08/01 - 20:29:12 | 201 | 32.608388797s | 127.0.0.1 | POST "/api/blobs/sha256:ede18cfa7e112bb2a9d9e5e148b9ff7ebb95279c1c5e41d4a20506a4b61b6f03"
8月 01 20:30:11 cheng ollama[1889]: [GIN] 2024/08/01 - 20:30:11 | 201 | 32.208379126s | 127.0.0.1 | POST "/api/blobs/sha256:296fdee32a474a1752805609620005d68907034d8b48f89381951cf85bdcc9bd"
8月 01 20:31:13 cheng ollama[1889]: [GIN] 2024/08/01 - 20:31:13 | 201 | 33.420692973s | 127.0.0.1 | POST "/api/blobs/sha256:e1266b80a6ca3779079d8c9bdf1d1cf777f5c5c75201f4f567b03f6990ee8e7d"
8月 01 20:36:43 cheng ollama[1889]: [GIN] 2024/08/01 - 20:36:43 | 200 | 5m29s | 127.0.0.1 | POST "/api/create"
8月 01 20:49:17 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:17 | 200 | 34.271µs | 127.0.0.1 | HEAD "/"
8月 01 20:49:17 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:17 | 200 | 488.318µs | 127.0.0.1 | GET "/api/tags"
8月 01 20:49:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:36 | 200 | 29.398µs | 127.0.0.1 | HEAD "/"
8月 01 20:49:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:36 | 200 | 404.433µs | 127.0.0.1 | GET "/api/tags"
8月 01 20:49:41 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:41 | 200 | 59.252µs | 127.0.0.1 | HEAD "/"
8月 01 20:49:41 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:41 | 200 | 9.883069ms | 127.0.0.1 | POST "/api/show"
8月 01 20:49:42 cheng ollama[1889]: time=2024-08-01T20:49:42.325+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size"
8月 01 20:49:42 cheng ollama[1889]: panic: runtime error: integer divide by zero
8月 01 20:49:42 cheng ollama[1889]: goroutine 13 [running]:
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, , {, _, _}, {{0x0, 0x2000, 0x200, ...}, ...})
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm.PredictServerFit({0xc00074bbd8?, 0x4?, 0x4?}, , {, _, _}, {0x0, 0x0, 0x0}, ...)
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc00016a700, 0xc000709780, {0xc00016cdc8?, 0x4?, 0x4?}, 0xc0002edd18)
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server.(Scheduler).processPending(0xc000116ba0, {0x22aeab50, 0xc0005480f0})
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server.(Scheduler).Run.func1()
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 01 20:49:42 cheng ollama[1889]: created by github.com/ollama/ollama/server.(Scheduler).Run in goroutine 1
8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 01 20:49:42 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 01 20:49:42 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 01 20:49:42 cheng systemd[1]: ollama.service: Consumed 21min 49.078s CPU time.
8月 01 20:49:45 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1.
8月 01 20:49:45 cheng systemd[1]: Stopped Ollama Service.
8月 01 20:49:45 cheng systemd[1]: ollama.service: Consumed 21min 49.078s CPU time.
8月 01 20:49:45 cheng systemd[1]: Started Ollama Service.
8月 01 20:49:45 cheng ollama[28228]: Couldn't find '/usr/share/ollama/.ollama/id_ed25519'. Generating new private key.
8月 01 20:49:45 cheng ollama[28228]: Your new public key is:
8月 01 20:49:45 cheng ollama[28228]: ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAILxBQqiTdXAnV3aWrQ5ZtN1EpdtgIZKBEST7iqBy8u2p
8月 01 20:49:45 cheng ollama[28228]: 2024/08/01 20:49:45 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.626+08:00 level=INFO source=images.go:781 msg="total blobs: 12"
8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.626+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0"
8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.626+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)"
8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.635+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama2457549556 error="remove /tmp/ollama2457549556: directory not empty"
8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.635+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama3981261121/runners
8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.347+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu_avx2 cuda_v11 rocm_v60102 cpu cpu_avx]"
8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.347+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs"
8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:51:45 cheng ollama[28228]: [GIN] 2024/08/01 - 20:51:45 | 200 | 53.369µs | 127.0.0.1 | HEAD "/"
8月 01 20:51:45 cheng ollama[28228]: [GIN] 2024/08/01 - 20:51:45 | 200 | 1.366054ms | 127.0.0.1 | POST "/api/show"
8月 01 20:51:45 cheng ollama[28228]: time=2024-08-01T20:51:45.643+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size"
8月 01 20:51:45 cheng ollama[28228]: panic: runtime error: integer divide by zero
8月 01 20:51:45 cheng ollama[28228]: goroutine 66 [running]:
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, , {, _, _}, {{0x0, 0x2000, 0x200, ...}, ...})
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm.PredictServerFit({0xc000893bd8?, 0x4?, 0x4?}, , {, _, _}, {0x0, 0x0, 0x0}, ...)
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc000558200, 0xc0003d8020, {0xc00054cb08?, 0x4?, 0x4?}, 0xc000443d18)
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server.(Scheduler).processPending(0xc0003b3020, {0x22aeab50, 0xc0000bfa90})
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server.(Scheduler).Run.func1()
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 01 20:51:45 cheng ollama[28228]: created by github.com/ollama/ollama/server.(Scheduler).Run in goroutine 1
8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 01 20:51:45 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 01 20:51:45 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 01 20:51:45 cheng systemd[1]: ollama.service: Consumed 19.730s CPU time.
8月 01 20:51:48 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 2.
8月 01 20:51:48 cheng systemd[1]: Stopped Ollama Service.
8月 01 20:51:48 cheng systemd[1]: ollama.service: Consumed 19.730s CPU time.
8月 01 20:51:48 cheng systemd[1]: Started Ollama Service.
8月 01 20:51:48 cheng ollama[28388]: 2024/08/01 20:51:48 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=INFO source=images.go:781 msg="total blobs: 12"
8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0"
8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)"
8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama2457549556 error="remove /tmp/ollama2457549556: directory not empty"
8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.857+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3981261121 error="remove /tmp/ollama3981261121: directory not empty"
8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.857+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama2557344928/runners
8月 01 20:51:50 cheng systemd[1]: Stopping Ollama Service...
8月 01 20:51:55 cheng ollama[28388]: time=2024-08-01T20:51:55.741+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries []"
8月 01 20:51:55 cheng ollama[28388]: time=2024-08-01T20:51:55.741+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs"
8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:51:56 cheng systemd[1]: ollama.service: Deactivated successfully.
8月 01 20:51:56 cheng systemd[1]: Stopped Ollama Service.
8月 01 20:51:56 cheng systemd[1]: ollama.service: Consumed 19.925s CPU time.
-- Boot 2c6de2a332144d9a8ee2924b08fb1feb --
8月 01 20:53:23 cheng systemd[1]: Started Ollama Service.
8月 01 20:53:23 cheng ollama[1877]: 2024/08/01 20:53:23 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.988+08:00 level=INFO source=images.go:781 msg="total blobs: 12"
8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.989+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0"
8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.989+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)"
8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.990+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama3377558664/runners
8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.110+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11 rocm_v60102]"
8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.110+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs"
8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.664+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.665+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.665+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.665+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:55:12 cheng ollama[1877]: [GIN] 2024/08/01 - 20:55:12 | 200 | 347.797µs | 127.0.0.1 | HEAD "/"
8月 01 20:55:12 cheng ollama[1877]: [GIN] 2024/08/01 - 20:55:12 | 200 | 1.771204ms | 127.0.0.1 | POST "/api/show"
8月 01 20:55:12 cheng ollama[1877]: time=2024-08-01T20:55:12.930+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size"
8月 01 20:55:12 cheng ollama[1877]: panic: runtime error: integer divide by zero
8月 01 20:55:12 cheng ollama[1877]: goroutine 54 [running]:
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, , {, _, _}, {{0x0, 0x2000, 0x200, ...}, ...})
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm.PredictServerFit({0xc00039bbd8?, 0x4?, 0x4?}, , {, _, _}, {0x0, 0x0, 0x0}, ...)
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc0004b0b00, 0xc000037640, {0xc0004c4b08?, 0x4?, 0x4?}, 0xc0003a5d18)
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server.(Scheduler).processPending(0xc00043af00, {0x22aeab50, 0xc0003188c0})
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server.(Scheduler).Run.func1()
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 01 20:55:12 cheng ollama[1877]: created by github.com/ollama/ollama/server.(Scheduler).Run in goroutine 1
8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 01 20:55:12 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 01 20:55:12 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 01 20:55:12 cheng systemd[1]: ollama.service: Consumed 18.734s CPU time.
8月 01 20:55:16 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1.
8月 01 20:55:16 cheng systemd[1]: Stopped Ollama Service.
8月 01 20:55:16 cheng systemd[1]: ollama.service: Consumed 18.734s CPU time.
8月 01 20:55:16 cheng systemd[1]: Started Ollama Service.
8月 01 20:55:16 cheng ollama[4792]: 2024/08/01 20:55:16 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.067+08:00 level=INFO source=images.go:781 msg="total blobs: 12"
8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.067+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0"
8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.067+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)"
8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.068+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3377558664 error="remove /tmp/ollama3377558664: directory not empty"
8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.068+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama1815270941/runners
8月 01 20:55:22 cheng ollama[4792]: time=2024-08-01T20:55:22.577+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11 rocm_v60102]"
8月 01 20:55:22 cheng ollama[4792]: time=2024-08-01T20:55:22.577+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs"
8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:56:40 cheng ollama[4792]: [GIN] 2024/08/01 - 20:56:40 | 200 | 58.151µs | 127.0.0.1 | HEAD "/"
8月 01 20:56:40 cheng ollama[4792]: [GIN] 2024/08/01 - 20:56:40 | 200 | 465.523µs | 127.0.0.1 | GET "/api/tags"
8月 01 20:59:09 cheng ollama[4792]: [GIN] 2024/08/01 - 20:59:09 | 200 | 50.178µs | 127.0.0.1 | HEAD "/"
8月 01 20:59:09 cheng ollama[4792]: [GIN] 2024/08/01 - 20:59:09 | 200 | 2.818931ms | 127.0.0.1 | POST "/api/show"
8月 01 20:59:09 cheng ollama[4792]: time=2024-08-01T20:59:09.773+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size"
8月 01 20:59:09 cheng ollama[4792]: panic: runtime error: integer divide by zero
8月 01 20:59:09 cheng ollama[4792]: goroutine 66 [running]:
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, , {, _, _}, {{0x0, 0x2000, 0x200, ...}, ...})
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm.PredictServerFit({0xc0004e1bd8?, 0x4?, 0x4?}, , {, _, _}, {0x0, 0x0, 0x0}, ...)
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc000255100, 0xc0004943c0, {0xc0004222c8?, 0x4?, 0x4?}, 0xc0004f9d18)
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server.(Scheduler).processPending(0xc0003b0f60, {0x22aeab50, 0xc0003a6f50})
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server.(Scheduler).Run.func1()
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 01 20:59:09 cheng ollama[4792]: created by github.com/ollama/ollama/server.(Scheduler).Run in goroutine 1
8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 01 20:59:09 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 01 20:59:09 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 01 20:59:09 cheng systemd[1]: ollama.service: Consumed 20.184s CPU time.
8月 01 20:59:13 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 2.
8月 01 20:59:13 cheng systemd[1]: Stopped Ollama Service.
8月 01 20:59:13 cheng systemd[1]: ollama.service: Consumed 20.184s CPU time.
8月 01 20:59:13 cheng systemd[1]: Started Ollama Service.
8月 01 20:59:13 cheng ollama[5138]: 2024/08/01 20:59:13 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.067+08:00 level=INFO source=images.go:781 msg="total blobs: 12"
8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0"
8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)"
8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama1815270941 error="remove /tmp/ollama1815270941: directory not empty"
8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3377558664 error="remove /tmp/ollama3377558664: directory not empty"
8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.069+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama3269178229/runners
8月 01 20:59:19 cheng ollama[5138]: time=2024-08-01T20:59:19.727+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu_avx2 cuda_v11 rocm_v60102 cpu cpu_avx]"
8月 01 20:59:19 cheng ollama[5138]: time=2024-08-01T20:59:19.727+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs"
8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 21:04:01 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:01 | 200 | 51.242µs | 127.0.0.1 | HEAD "/"
8月 01 21:04:01 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:01 | 200 | 442.689µs | 127.0.0.1 | GET "/api/tags"
8月 01 21:04:12 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:12 | 200 | 35.143µs | 127.0.0.1 | HEAD "/"
8月 01 21:04:12 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:12 | 200 | 1.312058ms | 127.0.0.1 | POST "/api/show"
8月 01 21:04:13 cheng ollama[5138]: time=2024-08-01T21:04:13.208+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size"
8月 01 21:04:13 cheng ollama[5138]: panic: runtime error: integer divide by zero
8月 01 21:04:13 cheng ollama[5138]: goroutine 82 [running]:
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, , {, _, _}, {{0x0, 0x2000, 0x200, ...}, ...})
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm.PredictServerFit({0xc000107bd8?, 0x4?, 0x4?}, , {, _, _}, {0x0, 0x0, 0x0}, ...)
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc0003d2400, 0xc000037ea0, {0xc0003fa2c8?, 0x4?, 0x4?}, 0xc0004ddd18)
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server.(Scheduler).processPending(0xc0003b0ea0, {0x22aeab50, 0xc0004a2f50})
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server.(Scheduler).Run.func1()
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 01 21:04:13 cheng ollama[5138]: created by github.com/ollama/ollama/server.(Scheduler).Run in goroutine 1
8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 01 21:04:13 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 01 21:04:13 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 01 21:04:13 cheng systemd[1]: ollama.service: Consumed 19.781s CPU time.
8月 01 21:04:16 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 3.
8月 01 21:04:16 cheng systemd[1]: Stopped Ollama Service.
8月 01 21:04:16 cheng systemd[1]: ollama.service: Consumed 19.781s CPU time.
8月 01 21:04:16 cheng systemd[1]: Started Ollama Service.
8月 01 21:04:16 cheng ollama[5398]: 2024/08/01 21:04:16 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.318+08:00 level=INFO source=images.go:781 msg="total blobs: 12"
8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.318+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0"
8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)"
8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama1815270941 error="remove /tmp/ollama1815270941: directory not empty"
8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3269178229 error="remove /tmp/ollama3269178229: directory not empty"
8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3377558664 error="remove /tmp/ollama3377558664: directory not empty"
8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.320+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama347671123/runners
8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.069+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu_avx cpu_avx2 cuda_v11 rocm_v60102 cpu]"
8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.069+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs"
8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 01 21:04:23 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:23 | 404 | 86.36µs | 127.0.0.1 | GET "/api/chat"
8月 01 21:04:24 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:24 | 404 | 9.775µs | 127.0.0.1 | GET "/favicon.ico"
8月 01 21:04:38 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:38 | 404 | 13.876µs | 127.0.0.1 | GET "/api"
8月 01 21:04:41 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:41 | 200 | 27.389µs | 127.0.0.1 | GET "/"
8月 01 21:05:17 cheng ollama[5398]: [GIN] 2024/08/01 - 21:05:17 | 404 | 11.268µs | 127.0.0.1 | GET "/api/chat"
8月 01 21:09:25 cheng ollama[5398]: [GIN] 2024/08/01 - 21:09:25 | 200 | 141.497µs | 127.0.0.1 | GET "/"
8月 01 21:13:26 cheng ollama[5398]: [GIN] 2024/08/01 - 21:13:26 | 200 | 70.515µs | 127.0.0.1 | GET "/"
8月 01 21:15:11 cheng ollama[5398]: [GIN] 2024/08/01 - 21:15:11 | 200 | 56.712µs | 127.0.0.1 | GET "/api/version"
8月 01 21:17:26 cheng systemd[1]: Stopping Ollama Service...
8月 01 21:17:26 cheng systemd[1]: ollama.service: Deactivated successfully.
8月 01 21:17:26 cheng systemd[1]: Stopped Ollama Service.
8月 01 21:17:26 cheng systemd[1]: ollama.service: Consumed 20.123s CPU time.
-- Boot 9bceaafe62e14190ba2fd13cfe4b660c --
8月 02 07:54:47 cheng systemd[1]: Started Ollama Service.
8月 02 07:54:47 cheng ollama[1896]: 2024/08/02 07:54:47 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.944+08:00 level=INFO source=images.go:781 msg="total blobs: 12"
8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.944+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0"
8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.945+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)"
8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.946+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama2135090776/runners
8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.037+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11 rocm_v60102]"
8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.037+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs"
8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:02:01 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:01 | 200 | 399.071µs | 127.0.0.1 | HEAD "/"
8月 02 08:02:01 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:01 | 200 | 1.195064ms | 127.0.0.1 | GET "/api/tags"
8月 02 08:02:12 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:12 | 200 | 44.049µs | 127.0.0.1 | HEAD "/"
8月 02 08:02:12 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:12 | 200 | 1.793469ms | 127.0.0.1 | POST "/api/show"
8月 02 08:02:13 cheng ollama[1896]: time=2024-08-02T08:02:13.426+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size"
8月 02 08:02:13 cheng ollama[1896]: panic: runtime error: integer divide by zero
8月 02 08:02:13 cheng ollama[1896]: goroutine 70 [running]:
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, , {, _, _}, {{0x0, 0x2000, 0x200, ...}, ...})
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm.PredictServerFit({0xc00040dbd8?, 0x4?, 0x4?}, , {, _, _}, {0x0, 0x0, 0x0}, ...)
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc0001d9900, 0xc000037660, {0xc000246848?, 0x4?, 0x4?}, 0xc000419d18)
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server.(Scheduler).processPending(0xc0000cf080, {0x22aeab50, 0xc0004f88c0})
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server.(Scheduler).Run.func1()
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 02 08:02:13 cheng ollama[1896]: created by github.com/ollama/ollama/server.(Scheduler).Run in goroutine 1
8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 02 08:02:13 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 02 08:02:13 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 02 08:02:13 cheng systemd[1]: ollama.service: Consumed 17.958s CPU time.
8月 02 08:02:16 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1.
8月 02 08:02:16 cheng systemd[1]: Stopped Ollama Service.
8月 02 08:02:16 cheng systemd[1]: ollama.service: Consumed 17.958s CPU time.
8月 02 08:02:16 cheng systemd[1]: Started Ollama Service.
8月 02 08:02:16 cheng ollama[4728]: 2024/08/02 08:02:16 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=INFO source=images.go:781 msg="total blobs: 12"
8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0"
8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)"
8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama2135090776 error="remove /tmp/ollama2135090776: directory not empty"
8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.580+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama4077334318/runners
8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.278+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cuda_v11 rocm_v60102 cpu cpu_avx cpu_avx2]"
8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.278+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs"
8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:08:06 cheng systemd[1]: Stopping Ollama Service...
8月 02 08:08:06 cheng systemd[1]: ollama.service: Deactivated successfully.
8月 02 08:08:06 cheng systemd[1]: Stopped Ollama Service.
8月 02 08:08:06 cheng systemd[1]: ollama.service: Consumed 20.017s CPU time.
-- Boot 88a56500e37f4b6eaaaeaca00780c960 --
8月 02 08:10:53 cheng systemd[1]: Started Ollama Service.
8月 02 08:10:53 cheng ollama[1889]: 2024/08/02 08:10:53 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.056+08:00 level=INFO source=images.go:781 msg="total blobs: 12"
8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.057+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0"
8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.057+08:00 level=INFO source=routes.go:1156 msg="Listening on [::]:11434 (version 0.3.2)"
8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.057+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama2561891386/runners
8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.295+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11 rocm_v60102]"
8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.295+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs"
8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:15:44 cheng ollama[1889]: [GIN] 2024/08/02 - 08:15:44 | 200 | 386.787µs | 127.0.0.1 | HEAD "/"
8月 02 08:15:44 cheng ollama[1889]: [GIN] 2024/08/02 - 08:15:44 | 200 | 3.462854ms | 127.0.0.1 | POST "/api/show"
8月 02 08:15:44 cheng ollama[1889]: time=2024-08-02T08:15:44.622+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size"
8月 02 08:15:44 cheng ollama[1889]: panic: runtime error: integer divide by zero
8月 02 08:15:44 cheng ollama[1889]: goroutine 68 [running]:
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, , {, _, _}, {{0x0, 0x2000, 0x200, ...}, ...})
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm.PredictServerFit({0xc0008a9bd8?, 0x4?, 0x4?}, , {, _, _}, {0x0, 0x0, 0x0}, ...)
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc00046c200, 0xc00037c6a0, {0xc0004b0008?, 0x4?, 0x4?}, 0xc000047d18)
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server.(Scheduler).processPending(0xc00043af60, {0x22aeab50, 0xc0003308c0})
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server.(Scheduler).Run.func1()
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 02 08:15:44 cheng ollama[1889]: created by github.com/ollama/ollama/server.(Scheduler).Run in goroutine 1
8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 02 08:15:44 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 02 08:15:44 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 02 08:15:44 cheng systemd[1]: ollama.service: Consumed 18.710s CPU time.
8月 02 08:15:47 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1.
8月 02 08:15:47 cheng systemd[1]: Stopped Ollama Service.
8月 02 08:15:47 cheng systemd[1]: ollama.service: Consumed 18.710s CPU time.
8月 02 08:15:47 cheng systemd[1]: Started Ollama Service.
8月 02 08:15:47 cheng ollama[4056]: 2024/08/02 08:15:47 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=INFO source=images.go:781 msg="total blobs: 12"
8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0"
8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=INFO source=routes.go:1156 msg="Listening on [::]:11434 (version 0.3.2)"
8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama2561891386 error="remove /tmp/ollama2561891386: directory not empty"
8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.845+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama3369741500/runners
8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.330+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [rocm_v60102 cpu cpu_avx cpu_avx2 cuda_v11]"
8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.330+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs"
8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:18:51 cheng ollama[4056]: [GIN] 2024/08/02 - 08:18:51 | 200 | 42.468µs | 127.0.0.1 | HEAD "/"
8月 02 08:18:51 cheng ollama[4056]: [GIN] 2024/08/02 - 08:18:51 | 200 | 452.507µs | 127.0.0.1 | GET "/api/tags"
8月 02 08:19:00 cheng ollama[4056]: [GIN] 2024/08/02 - 08:19:00 | 200 | 33.594µs | 127.0.0.1 | HEAD "/"
8月 02 08:19:00 cheng ollama[4056]: [GIN] 2024/08/02 - 08:19:00 | 200 | 1.193366ms | 127.0.0.1 | POST "/api/show"
8月 02 08:19:01 cheng ollama[4056]: time=2024-08-02T08:19:01.303+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size"
8月 02 08:19:01 cheng ollama[4056]: panic: runtime error: integer divide by zero
8月 02 08:19:01 cheng ollama[4056]: goroutine 38 [running]:
8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, , {, _, _}, {{0x0, 0x2000, 0x200, ...}, ...})
8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/llm.PredictServerFit({0xc000779bd8?, 0x4?, 0x4?}, , {, _, _}, {0x0, 0x0, 0x0}, ...)
8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc0002d5000, 0xc000442ec0, {0xc0004622c8?, 0x4?, 0x4?}, 0xc0004ebd18)
8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/server.(Scheduler).processPending(0xc0000b4f00, {0x22aeab50, 0xc0000bf950})
8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/server.(Scheduler).Run.func1()
8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 02 08:19:01 cheng ollama[4056]: created by github.com/ollama/ollama/server.(Scheduler).Run in goroutine 1
8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 02 08:19:01 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 02 08:19:01 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 02 08:19:01 cheng systemd[1]: ollama.service: Consumed 20.346s CPU time.
8月 02 08:19:04 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 2.
8月 02 08:19:04 cheng systemd[1]: Stopped Ollama Service.
8月 02 08:19:04 cheng systemd[1]: ollama.service: Consumed 20.346s CPU time.
8月 02 08:19:04 cheng systemd[1]: Started Ollama Service.
8月 02 08:19:04 cheng ollama[5419]: 2024/08/02 08:19:04 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
8月 02 08:19:04 cheng ollama[5419]: time=2024-08-02T08:19:04.590+08:00 level=INFO source=images.go:781 msg="total blobs: 12"
8月 02 08:19:04 cheng ollama[5419]: time=2024-08-02T08:19:04.590+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0"
8月 02 08:19:04 cheng ollama[5419]: time=2024-08-02T08:19:04.591+08:00 level=INFO source=routes.go:1156 msg="Listening on [::]:11434 (version 0.3.2)"
8月 02 08:19:04 cheng ollama[5419]: time=2024-08-02T08:19:04.591+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama2561891386 error="remove /tmp/ollama2561891386: directory not empty"
8月 02 08:19:04 cheng ollama[5419]: time=2024-08-02T08:19:04.591+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3369741500 error="remove /tmp/ollama3369741500: directory not empty"
8月 02 08:19:04 cheng ollama[5419]: time=2024-08-02T08:19:04.591+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama3427016686/runners
8月 02 08:19:11 cheng ollama[5419]: time=2024-08-02T08:19:11.468+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11 rocm_v60102]"
8月 02 08:19:11 cheng ollama[5419]: time=2024-08-02T08:19:11.468+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs"
8月 02 08:19:12 cheng ollama[5419]: time=2024-08-02T08:19:12.013+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:19:12 cheng ollama[5419]: time=2024-08-02T08:19:12.013+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:19:12 cheng ollama[5419]: time=2024-08-02T08:19:12.013+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 08:19:12 cheng ollama[5419]: time=2024-08-02T08:19:12.013+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB"
8月 02 09:31:44 cheng ollama[5419]: [GIN] 2024/08/02 - 09:31:44 | 200 | 53.704µs | 127.0.0.1 | HEAD "/"
8月 02 09:31:44 cheng ollama[5419]: [GIN] 2024/08/02 - 09:31:44 | 200 | 1.251938ms | 127.0.0.1 | POST "/api/show"
8月 02 09:31:44 cheng ollama[5419]: time=2024-08-02T09:31:44.984+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size"
8月 02 09:31:44 cheng ollama[5419]: panic: runtime error: integer divide by zero
8月 02 09:31:44 cheng ollama[5419]: goroutine 66 [running]:
8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, , {, _, _}, {{0x0, 0x2000, 0x200, ...}, ...})
8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b
8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/llm.PredictServerFit({0xc0004abbd8?, 0x4?, 0x4?}, , {, _, _}, {0x0, 0x0, 0x0}, ...)
8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/llm/memory.go:20 +0xbd
8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc00031f000, 0xc0003d6f40, {0xc00047c008?, 0x4?, 0x4?}, 0xc0004d1d18)
8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/server/sched.go:709 +0x6f3
8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/server.(Scheduler).processPending(0xc0003b0ea0, {0x22aeab50, 0xc00061ceb0})
8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/server/sched.go:220 +0xdfd
8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/server.(Scheduler).Run.func1()
8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/server/sched.go:107 +0x1f
8月 02 09:31:44 cheng ollama[5419]: created by github.com/ollama/ollama/server.(Scheduler).Run in goroutine 1
8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/server/sched.go:106 +0xb4
8月 02 09:31:45 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
8月 02 09:31:45 cheng systemd[1]: ollama.service: Failed with result 'exit-code'.
8月 02 09:31:45 cheng systemd[1]: ollama.service: Consumed 21.014s CPU time.
8月 02 09:31:48 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 3.
8月 02 09:31:48 cheng systemd[1]: Stopped Ollama Service.
8月 02 09:31:48 cheng systemd[1]: ollama.service: Consumed 21.014s CPU time.
8月 02 09:31:48 cheng systemd[1]: Started Ollama Service.
8月 02 09:31:48 cheng ollama[9466]: 2024/08/02 09:31:48 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
8月 02 09:31:48 cheng ollama[9466]: time=2024-08-02T09:31:48.107+08:00 level=INFO source=images.go:781 msg="total blobs: 12"
8月 02 09:31:48 cheng ollama[9466]: time=2024-08-02T09:31:48.107+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0"
8月 02 09:31:48 cheng ollama[9466]: time=2024-08-02T09:31:48.107+08:00 level=INFO source=routes.go:1156 msg="Listening on [::]:11434 (version 0.3.2)"
8月 02 09:31:48 cheng ollama[9466]: time=2024-08-02T09:31:48.108+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama2561891386 error="remove /tmp/ollama2561891386: directory not empty"
8月 02 09:31:48 cheng ollama[9466]: time=2024-08-02T09:31:48.108+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3369741500 error="remove /tmp/ollama3369741500: directory not empty"
8月 02 09:31:48 cheng ollama[9466]: time=2024-08-02T09:31:48.108+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3427016686 error="remove /tmp/ollama3427016686: directory not empty"
8月 02 09:31:48 cheng ollama[9466]: time=2024-08-02T09:31:48.108+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama865072436/runners'''

<!-- gh-comment-id:2264322896 --> @SongXiaoMao commented on GitHub (Aug 2, 2024): '''(base) cheng@cheng:~$ ollama run qwen Error: Post "http://127.0.0.1:11434/api/chat": EOF (base) cheng@cheng:~$ journalctl -u ollama --no-pager 8月 01 15:09:29 cheng systemd[1]: Started Ollama Service. 8月 01 15:09:29 cheng ollama[31790]: Couldn't find '/usr/share/ollama/.ollama/id_ed25519'. Generating new private key. 8月 01 15:09:29 cheng ollama[31790]: Your new public key is: 8月 01 15:09:29 cheng ollama[31790]: ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAILEdfF57fBFLPZGwkUEVkasIBOjp34OgZqC6P5gOckLi 8月 01 15:09:29 cheng ollama[31790]: 2024/08/01 15:09:29 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" 8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.826+08:00 level=INFO source=images.go:781 msg="total blobs: 0" 8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.826+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0" 8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.826+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)" 8月 01 15:09:29 cheng ollama[31790]: time=2024-08-01T15:09:29.827+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama2755471342/runners 8月 01 15:09:36 cheng ollama[31790]: time=2024-08-01T15:09:36.397+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11 rocm_v60102]" 8月 01 15:09:36 cheng ollama[31790]: time=2024-08-01T15:09:36.397+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs" 8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 15:09:37 cheng ollama[31790]: time=2024-08-01T15:09:37.019+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 15:20:48 cheng ollama[31790]: [GIN] 2024/08/01 - 15:20:48 | 200 | 94.041µs | 127.0.0.1 | HEAD "/" 8月 01 15:21:00 cheng ollama[31790]: [GIN] 2024/08/01 - 15:21:00 | 200 | 32.908µs | 127.0.0.1 | HEAD "/" 8月 01 15:21:33 cheng ollama[31790]: [GIN] 2024/08/01 - 15:21:33 | 200 | 61.512µs | 127.0.0.1 | HEAD "/" 8月 01 15:21:50 cheng ollama[31790]: [GIN] 2024/08/01 - 15:21:50 | 200 | 48.418µs | 127.0.0.1 | HEAD "/" 8月 01 15:23:25 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:25 | 200 | 32.858µs | 127.0.0.1 | HEAD "/" 8月 01 15:23:25 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:25 | 200 | 301.909µs | 127.0.0.1 | GET "/api/tags" 8月 01 15:23:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:40 | 200 | 30.281µs | 127.0.0.1 | HEAD "/" 8月 01 15:23:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:40 | 200 | 87.991µs | 127.0.0.1 | GET "/api/tags" 8月 01 15:23:47 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:47 | 200 | 29.357µs | 127.0.0.1 | HEAD "/" 8月 01 15:23:47 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:47 | 404 | 223.898µs | 127.0.0.1 | POST "/api/show" 8月 01 15:23:51 cheng ollama[31790]: time=2024-08-01T15:23:51.495+08:00 level=INFO source=download.go:173 msg="downloading 46bb65206e0e in 24 100 MB part(s)" 8月 01 15:23:55 cheng ollama[31790]: [GIN] 2024/08/01 - 15:23:55 | 200 | 7.982914121s | 127.0.0.1 | POST "/api/pull" 8月 01 15:25:17 cheng ollama[31790]: [GIN] 2024/08/01 - 15:25:17 | 200 | 61.437µs | 127.0.0.1 | HEAD "/" 8月 01 15:26:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:26:56 | 200 | 29.932µs | 127.0.0.1 | HEAD "/" 8月 01 15:27:14 cheng ollama[31790]: [GIN] 2024/08/01 - 15:27:14 | 200 | 33.123µs | 127.0.0.1 | HEAD "/" 8月 01 15:28:51 cheng ollama[31790]: [GIN] 2024/08/01 - 15:28:51 | 200 | 34.332µs | 127.0.0.1 | HEAD "/" 8月 01 15:29:39 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:39 | 200 | 34.374µs | 127.0.0.1 | HEAD "/" 8月 01 15:29:52 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:52 | 200 | 30.588µs | 127.0.0.1 | HEAD "/" 8月 01 15:29:54 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:54 | 201 | 1.639126644s | 127.0.0.1 | POST "/api/blobs/sha256:7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306" 8月 01 15:29:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:29:56 | 200 | 1.677066999s | 127.0.0.1 | POST "/api/create" 8月 01 15:30:01 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:01 | 200 | 35.043µs | 127.0.0.1 | HEAD "/" 8月 01 15:30:01 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:01 | 200 | 381.711µs | 127.0.0.1 | GET "/api/tags" 8月 01 15:30:07 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:07 | 200 | 31.294µs | 127.0.0.1 | HEAD "/" 8月 01 15:30:07 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:07 | 200 | 45.74282ms | 127.0.0.1 | POST "/api/show" 8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.622+08:00 level=INFO source=sched.go:710 msg="new model will fit in available VRAM in single GPU, loading" model=/usr/share/ollama/.ollama/models/blobs/sha256-7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306 gpu=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 parallel=4 available=25155272704 required="1.6 GiB" 8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.623+08:00 level=INFO source=memory.go:309 msg="offload to cuda" layers.requested=-1 layers.model=29 layers.offload=29 layers.split="" memory.available="[23.4 GiB]" memory.required.full="1.6 GiB" memory.required.partial="1.6 GiB" memory.required.kv="224.0 MiB" memory.required.allocations="[1.6 GiB]" memory.weights.total="680.7 MiB" memory.weights.repeating="498.1 MiB" memory.weights.nonrepeating="182.6 MiB" memory.graph.full="299.8 MiB" memory.graph.partial="482.3 MiB" 8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.624+08:00 level=INFO source=server.go:384 msg="starting llama server" cmd="/tmp/ollama2755471342/runners/cuda_v11/ollama_llama_server --model /usr/share/ollama/.ollama/models/blobs/sha256-7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306 --ctx-size 8192 --batch-size 512 --embedding --log-disable --n-gpu-layers 29 --parallel 4 --port 39233" 8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.625+08:00 level=INFO source=sched.go:445 msg="loaded runners" count=1 8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.625+08:00 level=INFO source=server.go:584 msg="waiting for llama runner to start responding" 8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.625+08:00 level=INFO source=server.go:618 msg="waiting for server to become available" status="llm server error" 8月 01 15:30:07 cheng ollama[38287]: INFO [main] build info | build=1 commit="6eeaeba" tid="133149442916352" timestamp=1722497407 8月 01 15:30:07 cheng ollama[38287]: INFO [main] system info | n_threads=12 n_threads_batch=-1 system_info="AVX = 1 | AVX_VNNI = 0 | AVX2 = 0 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | AVX512_BF16 = 0 | FMA = 0 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 0 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | " tid="133149442916352" timestamp=1722497407 total_threads=24 8月 01 15:30:07 cheng ollama[38287]: INFO [main] HTTP server listening | hostname="127.0.0.1" n_threads_http="23" port="39233" tid="133149442916352" timestamp=1722497407 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: loaded meta data with 26 key-value pairs and 338 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306 (version GGUF V3 (latest)) 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 0: general.architecture str = qwen2 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 1: general.name str = qwen2-1_5b-instruct 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 2: qwen2.block_count u32 = 28 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 3: qwen2.context_length u32 = 32768 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 4: qwen2.embedding_length u32 = 1536 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 5: qwen2.feed_forward_length u32 = 8960 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 6: qwen2.attention.head_count u32 = 12 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 = 2 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 8: qwen2.rope.freq_base f32 = 1000000.000000 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000001 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 10: general.file_type u32 = 10 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 11: tokenizer.ggml.model str = gpt2 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 12: tokenizer.ggml.pre str = qwen2 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151936] = ["!", "\"", "#", "$", "%", "&", "'", ... 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151936] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 = 151645 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 = 151643 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 151643 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 19: tokenizer.chat_template str = {% for message in messages %}{% if lo... 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 20: tokenizer.ggml.add_bos_token bool = false 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 21: general.quantization_version u32 = 2 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 22: quantize.imatrix.file str = ../Qwen2/gguf/qwen2-1_5b-imatrix/imat... 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 23: quantize.imatrix.dataset str = ../sft_2406.txt 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 24: quantize.imatrix.entries_count i32 = 196 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - kv 25: quantize.imatrix.chunks_count i32 = 1937 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type f32: 141 tensors 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q2_K: 112 tensors 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q3_K: 56 tensors 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q4_K: 28 tensors 8月 01 15:30:07 cheng ollama[31790]: llama_model_loader: - type q6_K: 1 tensors 8月 01 15:30:07 cheng ollama[31790]: time=2024-08-01T15:30:07.878+08:00 level=INFO source=server.go:618 msg="waiting for server to become available" status="llm server loading model" 8月 01 15:30:08 cheng ollama[31790]: llm_load_vocab: special tokens cache size = 293 8月 01 15:30:08 cheng ollama[31790]: llm_load_vocab: token to piece cache size = 0.9338 MB 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: format = GGUF V3 (latest) 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: arch = qwen2 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: vocab type = BPE 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_vocab = 151936 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_merges = 151387 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: vocab_only = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_ctx_train = 32768 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd = 1536 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_layer = 28 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_head = 12 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_head_kv = 2 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_rot = 128 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_swa = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_head_k = 128 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_head_v = 128 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_gqa = 6 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_k_gqa = 256 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_embd_v_gqa = 256 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_norm_eps = 0.0e+00 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_norm_rms_eps = 1.0e-06 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_clamp_kqv = 0.0e+00 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_max_alibi_bias = 0.0e+00 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: f_logit_scale = 0.0e+00 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_ff = 8960 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_expert = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_expert_used = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: causal attn = 1 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: pooling type = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: rope type = 2 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: rope scaling = linear 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: freq_base_train = 1000000.0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: freq_scale_train = 1 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: n_ctx_orig_yarn = 32768 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: rope_finetuned = unknown 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_d_conv = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_d_inner = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_d_state = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: ssm_dt_rank = 0 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model type = ?B 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model ftype = Q2_K - Medium 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model params = 1.54 B 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: model size = 639.30 MiB (3.47 BPW) 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: general.name = qwen2-1_5b-instruct 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: BOS token = 151643 '<|endoftext|>' 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: EOS token = 151645 '<|im_end|>' 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: PAD token = 151643 '<|endoftext|>' 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: LF token = 148848 'ÄĬ' 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: EOT token = 151645 '<|im_end|>' 8月 01 15:30:08 cheng ollama[31790]: llm_load_print_meta: max token length = 256 8月 01 15:30:08 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no 8月 01 15:30:08 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no 8月 01 15:30:08 cheng ollama[31790]: ggml_cuda_init: found 1 CUDA devices: 8月 01 15:30:08 cheng ollama[31790]: Device 0: NVIDIA GeForce RTX 3090, compute capability 8.6, VMM: yes 8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: ggml ctx size = 0.30 MiB 8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: offloading 28 repeating layers to GPU 8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: offloading non-repeating layers to GPU 8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: offloaded 29/29 layers to GPU 8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: CPU buffer size = 182.57 MiB 8月 01 15:30:08 cheng ollama[31790]: llm_load_tensors: CUDA0 buffer size = 639.30 MiB 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: n_ctx = 8192 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: n_batch = 512 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: n_ubatch = 512 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: flash_attn = 0 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: freq_base = 1000000.0 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: freq_scale = 1 8月 01 15:30:08 cheng ollama[31790]: llama_kv_cache_init: CUDA0 KV buffer size = 224.00 MiB 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: KV self size = 224.00 MiB, K (f16): 112.00 MiB, V (f16): 112.00 MiB 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host output buffer size = 2.34 MiB 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: CUDA0 compute buffer size = 299.75 MiB 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host compute buffer size = 19.01 MiB 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: graph nodes = 986 8月 01 15:30:08 cheng ollama[31790]: llama_new_context_with_model: graph splits = 2 8月 01 15:30:09 cheng ollama[38287]: INFO [main] model loaded | tid="133149442916352" timestamp=1722497409 8月 01 15:30:10 cheng ollama[31790]: time=2024-08-01T15:30:10.139+08:00 level=INFO source=server.go:623 msg="llama runner started in 2.51 seconds" 8月 01 15:30:10 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:10 | 200 | 3.085087137s | 127.0.0.1 | POST "/api/chat" 8月 01 15:30:14 cheng ollama[31790]: [GIN] 2024/08/01 - 15:30:14 | 200 | 180.746309ms | 127.0.0.1 | POST "/api/chat" 8月 01 15:36:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:36:56 | 200 | 31.499µs | 127.0.0.1 | HEAD "/" 8月 01 15:36:56 cheng ollama[31790]: [GIN] 2024/08/01 - 15:36:56 | 200 | 194.946µs | 127.0.0.1 | GET "/api/tags" 8月 01 15:37:02 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:02 | 200 | 65.7µs | 127.0.0.1 | HEAD "/" 8月 01 15:37:02 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:02 | 200 | 87.954µs | 127.0.0.1 | GET "/api/tags" 8月 01 15:37:31 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:31 | 200 | 33.094µs | 127.0.0.1 | HEAD "/" 8月 01 15:37:33 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:33 | 201 | 1.651416565s | 127.0.0.1 | POST "/api/blobs/sha256:7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306" 8月 01 15:37:35 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:35 | 200 | 1.725709949s | 127.0.0.1 | POST "/api/create" 8月 01 15:37:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:40 | 200 | 28.411µs | 127.0.0.1 | HEAD "/" 8月 01 15:37:40 cheng ollama[31790]: [GIN] 2024/08/01 - 15:37:40 | 200 | 355.484µs | 127.0.0.1 | GET "/api/tags" 8月 01 15:38:10 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:10 | 200 | 60.68µs | 127.0.0.1 | HEAD "/" 8月 01 15:38:10 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:10 | 200 | 22.792904ms | 127.0.0.1 | POST "/api/show" 8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.163+08:00 level=INFO source=sched.go:710 msg="new model will fit in available VRAM in single GPU, loading" model=/usr/share/ollama/.ollama/models/blobs/sha256-7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306 gpu=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 parallel=4 available=25155272704 required="1.6 GiB" 8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.164+08:00 level=INFO source=memory.go:309 msg="offload to cuda" layers.requested=-1 layers.model=29 layers.offload=29 layers.split="" memory.available="[23.4 GiB]" memory.required.full="1.6 GiB" memory.required.partial="1.6 GiB" memory.required.kv="224.0 MiB" memory.required.allocations="[1.6 GiB]" memory.weights.total="680.7 MiB" memory.weights.repeating="498.1 MiB" memory.weights.nonrepeating="182.6 MiB" memory.graph.full="299.8 MiB" memory.graph.partial="482.3 MiB" 8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.164+08:00 level=INFO source=server.go:384 msg="starting llama server" cmd="/tmp/ollama2755471342/runners/cuda_v11/ollama_llama_server --model /usr/share/ollama/.ollama/models/blobs/sha256-7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306 --ctx-size 8192 --batch-size 512 --embedding --log-disable --n-gpu-layers 29 --parallel 4 --port 39549" 8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.165+08:00 level=INFO source=sched.go:445 msg="loaded runners" count=1 8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.165+08:00 level=INFO source=server.go:584 msg="waiting for llama runner to start responding" 8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.165+08:00 level=INFO source=server.go:618 msg="waiting for server to become available" status="llm server error" 8月 01 15:38:11 cheng ollama[41534]: INFO [main] build info | build=1 commit="6eeaeba" tid="125225824997376" timestamp=1722497891 8月 01 15:38:11 cheng ollama[41534]: INFO [main] system info | n_threads=12 n_threads_batch=-1 system_info="AVX = 1 | AVX_VNNI = 0 | AVX2 = 0 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | AVX512_BF16 = 0 | FMA = 0 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 0 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | " tid="125225824997376" timestamp=1722497891 total_threads=24 8月 01 15:38:11 cheng ollama[41534]: INFO [main] HTTP server listening | hostname="127.0.0.1" n_threads_http="23" port="39549" tid="125225824997376" timestamp=1722497891 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: loaded meta data with 26 key-value pairs and 338 tensors from /usr/share/ollama/.ollama/models/blobs/sha256-7072d042ff33c22b6462c736cc4167ed6f4986543234734eac38bc24dfb9f306 (version GGUF V3 (latest)) 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 0: general.architecture str = qwen2 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 1: general.name str = qwen2-1_5b-instruct 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 2: qwen2.block_count u32 = 28 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 3: qwen2.context_length u32 = 32768 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 4: qwen2.embedding_length u32 = 1536 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 5: qwen2.feed_forward_length u32 = 8960 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 6: qwen2.attention.head_count u32 = 12 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 7: qwen2.attention.head_count_kv u32 = 2 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 8: qwen2.rope.freq_base f32 = 1000000.000000 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 9: qwen2.attention.layer_norm_rms_epsilon f32 = 0.000001 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 10: general.file_type u32 = 10 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 11: tokenizer.ggml.model str = gpt2 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 12: tokenizer.ggml.pre str = qwen2 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 13: tokenizer.ggml.tokens arr[str,151936] = ["!", "\"", "#", "$", "%", "&", "'", ... 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 14: tokenizer.ggml.token_type arr[i32,151936] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 15: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 16: tokenizer.ggml.eos_token_id u32 = 151645 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 17: tokenizer.ggml.padding_token_id u32 = 151643 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 151643 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 19: tokenizer.chat_template str = {% for message in messages %}{% if lo... 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 20: tokenizer.ggml.add_bos_token bool = false 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 21: general.quantization_version u32 = 2 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 22: quantize.imatrix.file str = ../Qwen2/gguf/qwen2-1_5b-imatrix/imat... 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 23: quantize.imatrix.dataset str = ../sft_2406.txt 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 24: quantize.imatrix.entries_count i32 = 196 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - kv 25: quantize.imatrix.chunks_count i32 = 1937 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type f32: 141 tensors 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q2_K: 112 tensors 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q3_K: 56 tensors 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q4_K: 28 tensors 8月 01 15:38:11 cheng ollama[31790]: llama_model_loader: - type q6_K: 1 tensors 8月 01 15:38:11 cheng ollama[31790]: time=2024-08-01T15:38:11.417+08:00 level=INFO source=server.go:618 msg="waiting for server to become available" status="llm server loading model" 8月 01 15:38:11 cheng ollama[31790]: llm_load_vocab: special tokens cache size = 293 8月 01 15:38:11 cheng ollama[31790]: llm_load_vocab: token to piece cache size = 0.9338 MB 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: format = GGUF V3 (latest) 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: arch = qwen2 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: vocab type = BPE 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_vocab = 151936 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_merges = 151387 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: vocab_only = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_ctx_train = 32768 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd = 1536 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_layer = 28 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_head = 12 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_head_kv = 2 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_rot = 128 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_swa = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_head_k = 128 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_head_v = 128 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_gqa = 6 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_k_gqa = 256 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_embd_v_gqa = 256 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_norm_eps = 0.0e+00 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_norm_rms_eps = 1.0e-06 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_clamp_kqv = 0.0e+00 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_max_alibi_bias = 0.0e+00 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: f_logit_scale = 0.0e+00 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_ff = 8960 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_expert = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_expert_used = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: causal attn = 1 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: pooling type = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: rope type = 2 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: rope scaling = linear 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: freq_base_train = 1000000.0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: freq_scale_train = 1 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: n_ctx_orig_yarn = 32768 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: rope_finetuned = unknown 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_d_conv = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_d_inner = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_d_state = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: ssm_dt_rank = 0 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model type = ?B 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model ftype = Q2_K - Medium 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model params = 1.54 B 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: model size = 639.30 MiB (3.47 BPW) 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: general.name = qwen2-1_5b-instruct 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: BOS token = 151643 '<|endoftext|>' 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: EOS token = 151645 '<|im_end|>' 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: PAD token = 151643 '<|endoftext|>' 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: LF token = 148848 'ÄĬ' 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: EOT token = 151645 '<|im_end|>' 8月 01 15:38:11 cheng ollama[31790]: llm_load_print_meta: max token length = 256 8月 01 15:38:11 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no 8月 01 15:38:11 cheng ollama[31790]: ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no 8月 01 15:38:11 cheng ollama[31790]: ggml_cuda_init: found 1 CUDA devices: 8月 01 15:38:11 cheng ollama[31790]: Device 0: NVIDIA GeForce RTX 3090, compute capability 8.6, VMM: yes 8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: ggml ctx size = 0.30 MiB 8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: offloading 28 repeating layers to GPU 8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: offloading non-repeating layers to GPU 8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: offloaded 29/29 layers to GPU 8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: CPU buffer size = 182.57 MiB 8月 01 15:38:12 cheng ollama[31790]: llm_load_tensors: CUDA0 buffer size = 639.30 MiB 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: n_ctx = 8192 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: n_batch = 512 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: n_ubatch = 512 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: flash_attn = 0 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: freq_base = 1000000.0 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: freq_scale = 1 8月 01 15:38:12 cheng ollama[31790]: llama_kv_cache_init: CUDA0 KV buffer size = 224.00 MiB 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: KV self size = 224.00 MiB, K (f16): 112.00 MiB, V (f16): 112.00 MiB 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host output buffer size = 2.34 MiB 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: CUDA0 compute buffer size = 299.75 MiB 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: CUDA_Host compute buffer size = 19.01 MiB 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: graph nodes = 986 8月 01 15:38:12 cheng ollama[31790]: llama_new_context_with_model: graph splits = 2 8月 01 15:38:13 cheng ollama[41534]: INFO [main] model loaded | tid="125225824997376" timestamp=1722497893 8月 01 15:38:13 cheng ollama[31790]: time=2024-08-01T15:38:13.428+08:00 level=INFO source=server.go:623 msg="llama runner started in 2.26 seconds" 8月 01 15:38:13 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:13 | 200 | 2.81848391s | 127.0.0.1 | POST "/api/chat" 8月 01 15:38:16 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:16 | 200 | 131.417236ms | 127.0.0.1 | POST "/api/chat" 8月 01 15:38:18 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:18 | 200 | 337.686697ms | 127.0.0.1 | POST "/api/chat" 8月 01 15:38:19 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:19 | 200 | 528.190585ms | 127.0.0.1 | POST "/api/chat" 8月 01 15:38:21 cheng ollama[31790]: [GIN] 2024/08/01 - 15:38:21 | 200 | 744.627142ms | 127.0.0.1 | POST "/api/chat" 8月 01 15:41:00 cheng ollama[31790]: [GIN] 2024/08/01 - 15:41:00 | 200 | 31.118µs | 127.0.0.1 | HEAD "/" 8月 01 15:41:00 cheng ollama[31790]: [GIN] 2024/08/01 - 15:41:00 | 200 | 218.55µs | 127.0.0.1 | GET "/api/tags" 8月 01 15:45:20 cheng systemd[1]: Stopping Ollama Service... 8月 01 15:45:20 cheng systemd[1]: ollama.service: Deactivated successfully. 8月 01 15:45:20 cheng systemd[1]: Stopped Ollama Service. 8月 01 15:45:20 cheng systemd[1]: ollama.service: Consumed 40.519s CPU time. -- Boot 33083ba5d27a4c399c75fd0c8948f208 -- 8月 01 19:27:21 cheng systemd[1]: Started Ollama Service. 8月 01 19:27:21 cheng ollama[1889]: Couldn't find '/usr/share/ollama/.ollama/id_ed25519'. Generating new private key. 8月 01 19:27:21 cheng ollama[1889]: Your new public key is: 8月 01 19:27:21 cheng ollama[1889]: ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAIMnZh7qZa0+q9j5qj9n/YsCevtxy0d/bNe+D6kJ9tT6p 8月 01 19:27:21 cheng ollama[1889]: 2024/08/01 19:27:21 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" 8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.102+08:00 level=INFO source=images.go:781 msg="total blobs: 0" 8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.102+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0" 8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.102+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)" 8月 01 19:27:21 cheng ollama[1889]: time=2024-08-01T19:27:21.103+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama2457549556/runners 8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.372+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu_avx cpu_avx2 cuda_v11 rocm_v60102 cpu]" 8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.373+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs" 8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 19:27:27 cheng ollama[1889]: time=2024-08-01T19:27:27.959+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:04:56 cheng ollama[1889]: [GIN] 2024/08/01 - 20:04:56 | 200 | 363.118µs | 127.0.0.1 | HEAD "/" 8月 01 20:05:18 cheng ollama[1889]: [GIN] 2024/08/01 - 20:05:18 | 200 | 30.807µs | 127.0.0.1 | HEAD "/" 8月 01 20:06:07 cheng ollama[1889]: [GIN] 2024/08/01 - 20:06:07 | 201 | 30.463162163s | 127.0.0.1 | POST "/api/blobs/sha256:84e82bd14926b67751d38f2a4fb0bf9e48c73bbadaaad5a1e7742b98cea5a0ca" 8月 01 20:07:01 cheng ollama[1889]: [GIN] 2024/08/01 - 20:07:01 | 201 | 31.905709158s | 127.0.0.1 | POST "/api/blobs/sha256:3d25eef820731f8bc2d4fda6dfb241573c096ac672d3fb98b78ceb393dc2c84f" 8月 01 20:07:56 cheng ollama[1889]: [GIN] 2024/08/01 - 20:07:56 | 201 | 31.758401406s | 127.0.0.1 | POST "/api/blobs/sha256:6e8f1d80f216dfb32d57e617de7d8c4e0f58513edf026ec614626848556394eb" 8月 01 20:08:52 cheng ollama[1889]: [GIN] 2024/08/01 - 20:08:52 | 201 | 33.256550159s | 127.0.0.1 | POST "/api/blobs/sha256:cd2156d736336dcd878bd91ab823fd8b5e6c8ecae951fa6a34894b00ad0ca2f6" 8月 01 20:09:48 cheng ollama[1889]: [GIN] 2024/08/01 - 20:09:48 | 201 | 31.697320718s | 127.0.0.1 | POST "/api/blobs/sha256:ffa8fc60b1df9a0195de7c3171b89e0df69586d38466b42b8528aa0c06392e93" 8月 01 20:10:45 cheng ollama[1889]: [GIN] 2024/08/01 - 20:10:45 | 201 | 30.805938252s | 127.0.0.1 | POST "/api/blobs/sha256:bf6f42aaf95fb86a14a1b74ada6ac2780b38ed358d707f611b7f69c2ab37a494" 8月 01 20:11:44 cheng ollama[1889]: [GIN] 2024/08/01 - 20:11:44 | 201 | 33.848623115s | 127.0.0.1 | POST "/api/blobs/sha256:ede18cfa7e112bb2a9d9e5e148b9ff7ebb95279c1c5e41d4a20506a4b61b6f03" 8月 01 20:12:50 cheng ollama[1889]: [GIN] 2024/08/01 - 20:12:50 | 201 | 40.17151261s | 127.0.0.1 | POST "/api/blobs/sha256:296fdee32a474a1752805609620005d68907034d8b48f89381951cf85bdcc9bd" 8月 01 20:13:40 cheng ollama[1889]: [GIN] 2024/08/01 - 20:13:40 | 201 | 29.724228454s | 127.0.0.1 | POST "/api/blobs/sha256:e1266b80a6ca3779079d8c9bdf1d1cf777f5c5c75201f4f567b03f6990ee8e7d" 8月 01 20:18:33 cheng ollama[1889]: [GIN] 2024/08/01 - 20:18:33 | 200 | 31.811µs | 127.0.0.1 | HEAD "/" 8月 01 20:18:43 cheng ollama[1889]: [GIN] 2024/08/01 - 20:18:43 | 200 | 5m3s | 127.0.0.1 | POST "/api/create" 8月 01 20:19:22 cheng ollama[1889]: [GIN] 2024/08/01 - 20:19:22 | 201 | 30.191676368s | 127.0.0.1 | POST "/api/blobs/sha256:84e82bd14926b67751d38f2a4fb0bf9e48c73bbadaaad5a1e7742b98cea5a0ca" 8月 01 20:19:51 cheng ollama[1889]: [GIN] 2024/08/01 - 20:19:51 | 200 | 28.187662838s | 127.0.0.1 | POST "/api/create" 8月 01 20:22:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:22:36 | 200 | 31.05µs | 127.0.0.1 | HEAD "/" 8月 01 20:23:33 cheng ollama[1889]: [GIN] 2024/08/01 - 20:23:33 | 201 | 40.226746855s | 127.0.0.1 | POST "/api/blobs/sha256:84e82bd14926b67751d38f2a4fb0bf9e48c73bbadaaad5a1e7742b98cea5a0ca" 8月 01 20:24:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:24:36 | 201 | 41.01948605s | 127.0.0.1 | POST "/api/blobs/sha256:3d25eef820731f8bc2d4fda6dfb241573c096ac672d3fb98b78ceb393dc2c84f" 8月 01 20:25:32 cheng ollama[1889]: [GIN] 2024/08/01 - 20:25:32 | 201 | 34.131806305s | 127.0.0.1 | POST "/api/blobs/sha256:6e8f1d80f216dfb32d57e617de7d8c4e0f58513edf026ec614626848556394eb" 8月 01 20:26:27 cheng ollama[1889]: [GIN] 2024/08/01 - 20:26:27 | 201 | 35.313518063s | 127.0.0.1 | POST "/api/blobs/sha256:cd2156d736336dcd878bd91ab823fd8b5e6c8ecae951fa6a34894b00ad0ca2f6" 8月 01 20:27:26 cheng ollama[1889]: [GIN] 2024/08/01 - 20:27:26 | 201 | 39.101495054s | 127.0.0.1 | POST "/api/blobs/sha256:ffa8fc60b1df9a0195de7c3171b89e0df69586d38466b42b8528aa0c06392e93" 8月 01 20:28:19 cheng ollama[1889]: [GIN] 2024/08/01 - 20:28:19 | 201 | 32.1910056s | 127.0.0.1 | POST "/api/blobs/sha256:bf6f42aaf95fb86a14a1b74ada6ac2780b38ed358d707f611b7f69c2ab37a494" 8月 01 20:29:12 cheng ollama[1889]: [GIN] 2024/08/01 - 20:29:12 | 201 | 32.608388797s | 127.0.0.1 | POST "/api/blobs/sha256:ede18cfa7e112bb2a9d9e5e148b9ff7ebb95279c1c5e41d4a20506a4b61b6f03" 8月 01 20:30:11 cheng ollama[1889]: [GIN] 2024/08/01 - 20:30:11 | 201 | 32.208379126s | 127.0.0.1 | POST "/api/blobs/sha256:296fdee32a474a1752805609620005d68907034d8b48f89381951cf85bdcc9bd" 8月 01 20:31:13 cheng ollama[1889]: [GIN] 2024/08/01 - 20:31:13 | 201 | 33.420692973s | 127.0.0.1 | POST "/api/blobs/sha256:e1266b80a6ca3779079d8c9bdf1d1cf777f5c5c75201f4f567b03f6990ee8e7d" 8月 01 20:36:43 cheng ollama[1889]: [GIN] 2024/08/01 - 20:36:43 | 200 | 5m29s | 127.0.0.1 | POST "/api/create" 8月 01 20:49:17 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:17 | 200 | 34.271µs | 127.0.0.1 | HEAD "/" 8月 01 20:49:17 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:17 | 200 | 488.318µs | 127.0.0.1 | GET "/api/tags" 8月 01 20:49:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:36 | 200 | 29.398µs | 127.0.0.1 | HEAD "/" 8月 01 20:49:36 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:36 | 200 | 404.433µs | 127.0.0.1 | GET "/api/tags" 8月 01 20:49:41 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:41 | 200 | 59.252µs | 127.0.0.1 | HEAD "/" 8月 01 20:49:41 cheng ollama[1889]: [GIN] 2024/08/01 - 20:49:41 | 200 | 9.883069ms | 127.0.0.1 | POST "/api/show" 8月 01 20:49:42 cheng ollama[1889]: time=2024-08-01T20:49:42.325+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size" 8月 01 20:49:42 cheng ollama[1889]: panic: runtime error: integer divide by zero 8月 01 20:49:42 cheng ollama[1889]: goroutine 13 [running]: 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000, 0x200, ...}, ...}) 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm.PredictServerFit({0xc00074bbd8?, 0x4?, 0x4?}, _, {_, _, _}, {0x0, 0x0, 0x0}, ...) 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc00016a700, 0xc000709780, {0xc00016cdc8?, 0x4?, 0x4?}, 0xc0002edd18) 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc000116ba0, {0x22aeab50, 0xc0005480f0}) 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 01 20:49:42 cheng ollama[1889]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 01 20:49:42 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 01 20:49:42 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 01 20:49:42 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 01 20:49:42 cheng systemd[1]: ollama.service: Consumed 21min 49.078s CPU time. 8月 01 20:49:45 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1. 8月 01 20:49:45 cheng systemd[1]: Stopped Ollama Service. 8月 01 20:49:45 cheng systemd[1]: ollama.service: Consumed 21min 49.078s CPU time. 8月 01 20:49:45 cheng systemd[1]: Started Ollama Service. 8月 01 20:49:45 cheng ollama[28228]: Couldn't find '/usr/share/ollama/.ollama/id_ed25519'. Generating new private key. 8月 01 20:49:45 cheng ollama[28228]: Your new public key is: 8月 01 20:49:45 cheng ollama[28228]: ssh-ed25519 AAAAC3NzaC1lZDI1NTE5AAAAILxBQqiTdXAnV3aWrQ5ZtN1EpdtgIZKBEST7iqBy8u2p 8月 01 20:49:45 cheng ollama[28228]: 2024/08/01 20:49:45 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" 8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.626+08:00 level=INFO source=images.go:781 msg="total blobs: 12" 8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.626+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0" 8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.626+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)" 8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.635+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama2457549556 error="remove /tmp/ollama2457549556: directory not empty" 8月 01 20:49:45 cheng ollama[28228]: time=2024-08-01T20:49:45.635+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama3981261121/runners 8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.347+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu_avx2 cuda_v11 rocm_v60102 cpu cpu_avx]" 8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.347+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs" 8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:49:52 cheng ollama[28228]: time=2024-08-01T20:49:52.920+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:51:45 cheng ollama[28228]: [GIN] 2024/08/01 - 20:51:45 | 200 | 53.369µs | 127.0.0.1 | HEAD "/" 8月 01 20:51:45 cheng ollama[28228]: [GIN] 2024/08/01 - 20:51:45 | 200 | 1.366054ms | 127.0.0.1 | POST "/api/show" 8月 01 20:51:45 cheng ollama[28228]: time=2024-08-01T20:51:45.643+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size" 8月 01 20:51:45 cheng ollama[28228]: panic: runtime error: integer divide by zero 8月 01 20:51:45 cheng ollama[28228]: goroutine 66 [running]: 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000, 0x200, ...}, ...}) 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm.PredictServerFit({0xc000893bd8?, 0x4?, 0x4?}, _, {_, _, _}, {0x0, 0x0, 0x0}, ...) 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc000558200, 0xc0003d8020, {0xc00054cb08?, 0x4?, 0x4?}, 0xc000443d18) 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc0003b3020, {0x22aeab50, 0xc0000bfa90}) 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 01 20:51:45 cheng ollama[28228]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 01 20:51:45 cheng ollama[28228]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 01 20:51:45 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 01 20:51:45 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 01 20:51:45 cheng systemd[1]: ollama.service: Consumed 19.730s CPU time. 8月 01 20:51:48 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 2. 8月 01 20:51:48 cheng systemd[1]: Stopped Ollama Service. 8月 01 20:51:48 cheng systemd[1]: ollama.service: Consumed 19.730s CPU time. 8月 01 20:51:48 cheng systemd[1]: Started Ollama Service. 8月 01 20:51:48 cheng ollama[28388]: 2024/08/01 20:51:48 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" 8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=INFO source=images.go:781 msg="total blobs: 12" 8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0" 8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)" 8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.856+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama2457549556 error="remove /tmp/ollama2457549556: directory not empty" 8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.857+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3981261121 error="remove /tmp/ollama3981261121: directory not empty" 8月 01 20:51:48 cheng ollama[28388]: time=2024-08-01T20:51:48.857+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama2557344928/runners 8月 01 20:51:50 cheng systemd[1]: Stopping Ollama Service... 8月 01 20:51:55 cheng ollama[28388]: time=2024-08-01T20:51:55.741+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries []" 8月 01 20:51:55 cheng ollama[28388]: time=2024-08-01T20:51:55.741+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs" 8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:51:56 cheng ollama[28388]: time=2024-08-01T20:51:56.281+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:51:56 cheng systemd[1]: ollama.service: Deactivated successfully. 8月 01 20:51:56 cheng systemd[1]: Stopped Ollama Service. 8月 01 20:51:56 cheng systemd[1]: ollama.service: Consumed 19.925s CPU time. -- Boot 2c6de2a332144d9a8ee2924b08fb1feb -- 8月 01 20:53:23 cheng systemd[1]: Started Ollama Service. 8月 01 20:53:23 cheng ollama[1877]: 2024/08/01 20:53:23 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" 8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.988+08:00 level=INFO source=images.go:781 msg="total blobs: 12" 8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.989+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0" 8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.989+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)" 8月 01 20:53:23 cheng ollama[1877]: time=2024-08-01T20:53:23.990+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama3377558664/runners 8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.110+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11 rocm_v60102]" 8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.110+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs" 8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.664+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.665+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.665+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:53:30 cheng ollama[1877]: time=2024-08-01T20:53:30.665+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:55:12 cheng ollama[1877]: [GIN] 2024/08/01 - 20:55:12 | 200 | 347.797µs | 127.0.0.1 | HEAD "/" 8月 01 20:55:12 cheng ollama[1877]: [GIN] 2024/08/01 - 20:55:12 | 200 | 1.771204ms | 127.0.0.1 | POST "/api/show" 8月 01 20:55:12 cheng ollama[1877]: time=2024-08-01T20:55:12.930+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size" 8月 01 20:55:12 cheng ollama[1877]: panic: runtime error: integer divide by zero 8月 01 20:55:12 cheng ollama[1877]: goroutine 54 [running]: 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000, 0x200, ...}, ...}) 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm.PredictServerFit({0xc00039bbd8?, 0x4?, 0x4?}, _, {_, _, _}, {0x0, 0x0, 0x0}, ...) 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc0004b0b00, 0xc000037640, {0xc0004c4b08?, 0x4?, 0x4?}, 0xc0003a5d18) 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc00043af00, {0x22aeab50, 0xc0003188c0}) 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 01 20:55:12 cheng ollama[1877]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 01 20:55:12 cheng ollama[1877]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 01 20:55:12 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 01 20:55:12 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 01 20:55:12 cheng systemd[1]: ollama.service: Consumed 18.734s CPU time. 8月 01 20:55:16 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1. 8月 01 20:55:16 cheng systemd[1]: Stopped Ollama Service. 8月 01 20:55:16 cheng systemd[1]: ollama.service: Consumed 18.734s CPU time. 8月 01 20:55:16 cheng systemd[1]: Started Ollama Service. 8月 01 20:55:16 cheng ollama[4792]: 2024/08/01 20:55:16 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" 8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.067+08:00 level=INFO source=images.go:781 msg="total blobs: 12" 8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.067+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0" 8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.067+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)" 8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.068+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3377558664 error="remove /tmp/ollama3377558664: directory not empty" 8月 01 20:55:16 cheng ollama[4792]: time=2024-08-01T20:55:16.068+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama1815270941/runners 8月 01 20:55:22 cheng ollama[4792]: time=2024-08-01T20:55:22.577+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11 rocm_v60102]" 8月 01 20:55:22 cheng ollama[4792]: time=2024-08-01T20:55:22.577+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs" 8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:55:23 cheng ollama[4792]: time=2024-08-01T20:55:23.134+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:56:40 cheng ollama[4792]: [GIN] 2024/08/01 - 20:56:40 | 200 | 58.151µs | 127.0.0.1 | HEAD "/" 8月 01 20:56:40 cheng ollama[4792]: [GIN] 2024/08/01 - 20:56:40 | 200 | 465.523µs | 127.0.0.1 | GET "/api/tags" 8月 01 20:59:09 cheng ollama[4792]: [GIN] 2024/08/01 - 20:59:09 | 200 | 50.178µs | 127.0.0.1 | HEAD "/" 8月 01 20:59:09 cheng ollama[4792]: [GIN] 2024/08/01 - 20:59:09 | 200 | 2.818931ms | 127.0.0.1 | POST "/api/show" 8月 01 20:59:09 cheng ollama[4792]: time=2024-08-01T20:59:09.773+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size" 8月 01 20:59:09 cheng ollama[4792]: panic: runtime error: integer divide by zero 8月 01 20:59:09 cheng ollama[4792]: goroutine 66 [running]: 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000, 0x200, ...}, ...}) 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm.PredictServerFit({0xc0004e1bd8?, 0x4?, 0x4?}, _, {_, _, _}, {0x0, 0x0, 0x0}, ...) 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc000255100, 0xc0004943c0, {0xc0004222c8?, 0x4?, 0x4?}, 0xc0004f9d18) 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc0003b0f60, {0x22aeab50, 0xc0003a6f50}) 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 01 20:59:09 cheng ollama[4792]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 01 20:59:09 cheng ollama[4792]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 01 20:59:09 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 01 20:59:09 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 01 20:59:09 cheng systemd[1]: ollama.service: Consumed 20.184s CPU time. 8月 01 20:59:13 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 2. 8月 01 20:59:13 cheng systemd[1]: Stopped Ollama Service. 8月 01 20:59:13 cheng systemd[1]: ollama.service: Consumed 20.184s CPU time. 8月 01 20:59:13 cheng systemd[1]: Started Ollama Service. 8月 01 20:59:13 cheng ollama[5138]: 2024/08/01 20:59:13 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" 8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.067+08:00 level=INFO source=images.go:781 msg="total blobs: 12" 8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0" 8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)" 8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama1815270941 error="remove /tmp/ollama1815270941: directory not empty" 8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.068+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3377558664 error="remove /tmp/ollama3377558664: directory not empty" 8月 01 20:59:13 cheng ollama[5138]: time=2024-08-01T20:59:13.069+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama3269178229/runners 8月 01 20:59:19 cheng ollama[5138]: time=2024-08-01T20:59:19.727+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu_avx2 cuda_v11 rocm_v60102 cpu cpu_avx]" 8月 01 20:59:19 cheng ollama[5138]: time=2024-08-01T20:59:19.727+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs" 8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 20:59:20 cheng ollama[5138]: time=2024-08-01T20:59:20.251+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 21:04:01 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:01 | 200 | 51.242µs | 127.0.0.1 | HEAD "/" 8月 01 21:04:01 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:01 | 200 | 442.689µs | 127.0.0.1 | GET "/api/tags" 8月 01 21:04:12 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:12 | 200 | 35.143µs | 127.0.0.1 | HEAD "/" 8月 01 21:04:12 cheng ollama[5138]: [GIN] 2024/08/01 - 21:04:12 | 200 | 1.312058ms | 127.0.0.1 | POST "/api/show" 8月 01 21:04:13 cheng ollama[5138]: time=2024-08-01T21:04:13.208+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size" 8月 01 21:04:13 cheng ollama[5138]: panic: runtime error: integer divide by zero 8月 01 21:04:13 cheng ollama[5138]: goroutine 82 [running]: 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000, 0x200, ...}, ...}) 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm.PredictServerFit({0xc000107bd8?, 0x4?, 0x4?}, _, {_, _, _}, {0x0, 0x0, 0x0}, ...) 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc0003d2400, 0xc000037ea0, {0xc0003fa2c8?, 0x4?, 0x4?}, 0xc0004ddd18) 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc0003b0ea0, {0x22aeab50, 0xc0004a2f50}) 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 01 21:04:13 cheng ollama[5138]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 01 21:04:13 cheng ollama[5138]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 01 21:04:13 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 01 21:04:13 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 01 21:04:13 cheng systemd[1]: ollama.service: Consumed 19.781s CPU time. 8月 01 21:04:16 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 3. 8月 01 21:04:16 cheng systemd[1]: Stopped Ollama Service. 8月 01 21:04:16 cheng systemd[1]: ollama.service: Consumed 19.781s CPU time. 8月 01 21:04:16 cheng systemd[1]: Started Ollama Service. 8月 01 21:04:16 cheng ollama[5398]: 2024/08/01 21:04:16 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" 8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.318+08:00 level=INFO source=images.go:781 msg="total blobs: 12" 8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.318+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0" 8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)" 8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama1815270941 error="remove /tmp/ollama1815270941: directory not empty" 8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3269178229 error="remove /tmp/ollama3269178229: directory not empty" 8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.319+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3377558664 error="remove /tmp/ollama3377558664: directory not empty" 8月 01 21:04:16 cheng ollama[5398]: time=2024-08-01T21:04:16.320+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama347671123/runners 8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.069+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu_avx cpu_avx2 cuda_v11 rocm_v60102 cpu]" 8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.069+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs" 8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 21:04:23 cheng ollama[5398]: time=2024-08-01T21:04:23.683+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 01 21:04:23 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:23 | 404 | 86.36µs | 127.0.0.1 | GET "/api/chat" 8月 01 21:04:24 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:24 | 404 | 9.775µs | 127.0.0.1 | GET "/favicon.ico" 8月 01 21:04:38 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:38 | 404 | 13.876µs | 127.0.0.1 | GET "/api" 8月 01 21:04:41 cheng ollama[5398]: [GIN] 2024/08/01 - 21:04:41 | 200 | 27.389µs | 127.0.0.1 | GET "/" 8月 01 21:05:17 cheng ollama[5398]: [GIN] 2024/08/01 - 21:05:17 | 404 | 11.268µs | 127.0.0.1 | GET "/api/chat" 8月 01 21:09:25 cheng ollama[5398]: [GIN] 2024/08/01 - 21:09:25 | 200 | 141.497µs | 127.0.0.1 | GET "/" 8月 01 21:13:26 cheng ollama[5398]: [GIN] 2024/08/01 - 21:13:26 | 200 | 70.515µs | 127.0.0.1 | GET "/" 8月 01 21:15:11 cheng ollama[5398]: [GIN] 2024/08/01 - 21:15:11 | 200 | 56.712µs | 127.0.0.1 | GET "/api/version" 8月 01 21:17:26 cheng systemd[1]: Stopping Ollama Service... 8月 01 21:17:26 cheng systemd[1]: ollama.service: Deactivated successfully. 8月 01 21:17:26 cheng systemd[1]: Stopped Ollama Service. 8月 01 21:17:26 cheng systemd[1]: ollama.service: Consumed 20.123s CPU time. -- Boot 9bceaafe62e14190ba2fd13cfe4b660c -- 8月 02 07:54:47 cheng systemd[1]: Started Ollama Service. 8月 02 07:54:47 cheng ollama[1896]: 2024/08/02 07:54:47 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" 8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.944+08:00 level=INFO source=images.go:781 msg="total blobs: 12" 8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.944+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0" 8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.945+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)" 8月 02 07:54:47 cheng ollama[1896]: time=2024-08-02T07:54:47.946+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama2135090776/runners 8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.037+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11 rocm_v60102]" 8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.037+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs" 8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 07:54:54 cheng ollama[1896]: time=2024-08-02T07:54:54.596+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:02:01 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:01 | 200 | 399.071µs | 127.0.0.1 | HEAD "/" 8月 02 08:02:01 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:01 | 200 | 1.195064ms | 127.0.0.1 | GET "/api/tags" 8月 02 08:02:12 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:12 | 200 | 44.049µs | 127.0.0.1 | HEAD "/" 8月 02 08:02:12 cheng ollama[1896]: [GIN] 2024/08/02 - 08:02:12 | 200 | 1.793469ms | 127.0.0.1 | POST "/api/show" 8月 02 08:02:13 cheng ollama[1896]: time=2024-08-02T08:02:13.426+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size" 8月 02 08:02:13 cheng ollama[1896]: panic: runtime error: integer divide by zero 8月 02 08:02:13 cheng ollama[1896]: goroutine 70 [running]: 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000, 0x200, ...}, ...}) 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm.PredictServerFit({0xc00040dbd8?, 0x4?, 0x4?}, _, {_, _, _}, {0x0, 0x0, 0x0}, ...) 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc0001d9900, 0xc000037660, {0xc000246848?, 0x4?, 0x4?}, 0xc000419d18) 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc0000cf080, {0x22aeab50, 0xc0004f88c0}) 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 02 08:02:13 cheng ollama[1896]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 02 08:02:13 cheng ollama[1896]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 02 08:02:13 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 02 08:02:13 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 02 08:02:13 cheng systemd[1]: ollama.service: Consumed 17.958s CPU time. 8月 02 08:02:16 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1. 8月 02 08:02:16 cheng systemd[1]: Stopped Ollama Service. 8月 02 08:02:16 cheng systemd[1]: ollama.service: Consumed 17.958s CPU time. 8月 02 08:02:16 cheng systemd[1]: Started Ollama Service. 8月 02 08:02:16 cheng ollama[4728]: 2024/08/02 08:02:16 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" 8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=INFO source=images.go:781 msg="total blobs: 12" 8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0" 8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=INFO source=routes.go:1156 msg="Listening on 127.0.0.1:11434 (version 0.3.2)" 8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.579+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama2135090776 error="remove /tmp/ollama2135090776: directory not empty" 8月 02 08:02:16 cheng ollama[4728]: time=2024-08-02T08:02:16.580+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama4077334318/runners 8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.278+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cuda_v11 rocm_v60102 cpu cpu_avx cpu_avx2]" 8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.278+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs" 8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:02:23 cheng ollama[4728]: time=2024-08-02T08:02:23.822+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:08:06 cheng systemd[1]: Stopping Ollama Service... 8月 02 08:08:06 cheng systemd[1]: ollama.service: Deactivated successfully. 8月 02 08:08:06 cheng systemd[1]: Stopped Ollama Service. 8月 02 08:08:06 cheng systemd[1]: ollama.service: Consumed 20.017s CPU time. -- Boot 88a56500e37f4b6eaaaeaca00780c960 -- 8月 02 08:10:53 cheng systemd[1]: Started Ollama Service. 8月 02 08:10:53 cheng ollama[1889]: 2024/08/02 08:10:53 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" 8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.056+08:00 level=INFO source=images.go:781 msg="total blobs: 12" 8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.057+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0" 8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.057+08:00 level=INFO source=routes.go:1156 msg="Listening on [::]:11434 (version 0.3.2)" 8月 02 08:10:53 cheng ollama[1889]: time=2024-08-02T08:10:53.057+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama2561891386/runners 8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.295+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11 rocm_v60102]" 8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.295+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs" 8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:10:59 cheng ollama[1889]: time=2024-08-02T08:10:59.821+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:15:44 cheng ollama[1889]: [GIN] 2024/08/02 - 08:15:44 | 200 | 386.787µs | 127.0.0.1 | HEAD "/" 8月 02 08:15:44 cheng ollama[1889]: [GIN] 2024/08/02 - 08:15:44 | 200 | 3.462854ms | 127.0.0.1 | POST "/api/show" 8月 02 08:15:44 cheng ollama[1889]: time=2024-08-02T08:15:44.622+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size" 8月 02 08:15:44 cheng ollama[1889]: panic: runtime error: integer divide by zero 8月 02 08:15:44 cheng ollama[1889]: goroutine 68 [running]: 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000, 0x200, ...}, ...}) 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm.PredictServerFit({0xc0008a9bd8?, 0x4?, 0x4?}, _, {_, _, _}, {0x0, 0x0, 0x0}, ...) 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc00046c200, 0xc00037c6a0, {0xc0004b0008?, 0x4?, 0x4?}, 0xc000047d18) 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc00043af60, {0x22aeab50, 0xc0003308c0}) 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 02 08:15:44 cheng ollama[1889]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 02 08:15:44 cheng ollama[1889]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 02 08:15:44 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 02 08:15:44 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 02 08:15:44 cheng systemd[1]: ollama.service: Consumed 18.710s CPU time. 8月 02 08:15:47 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 1. 8月 02 08:15:47 cheng systemd[1]: Stopped Ollama Service. 8月 02 08:15:47 cheng systemd[1]: ollama.service: Consumed 18.710s CPU time. 8月 02 08:15:47 cheng systemd[1]: Started Ollama Service. 8月 02 08:15:47 cheng ollama[4056]: 2024/08/02 08:15:47 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" 8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=INFO source=images.go:781 msg="total blobs: 12" 8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0" 8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=INFO source=routes.go:1156 msg="Listening on [::]:11434 (version 0.3.2)" 8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.844+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama2561891386 error="remove /tmp/ollama2561891386: directory not empty" 8月 02 08:15:47 cheng ollama[4056]: time=2024-08-02T08:15:47.845+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama3369741500/runners 8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.330+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [rocm_v60102 cpu cpu_avx cpu_avx2 cuda_v11]" 8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.330+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs" 8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:15:54 cheng ollama[4056]: time=2024-08-02T08:15:54.884+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:18:51 cheng ollama[4056]: [GIN] 2024/08/02 - 08:18:51 | 200 | 42.468µs | 127.0.0.1 | HEAD "/" 8月 02 08:18:51 cheng ollama[4056]: [GIN] 2024/08/02 - 08:18:51 | 200 | 452.507µs | 127.0.0.1 | GET "/api/tags" 8月 02 08:19:00 cheng ollama[4056]: [GIN] 2024/08/02 - 08:19:00 | 200 | 33.594µs | 127.0.0.1 | HEAD "/" 8月 02 08:19:00 cheng ollama[4056]: [GIN] 2024/08/02 - 08:19:00 | 200 | 1.193366ms | 127.0.0.1 | POST "/api/show" 8月 02 08:19:01 cheng ollama[4056]: time=2024-08-02T08:19:01.303+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size" 8月 02 08:19:01 cheng ollama[4056]: panic: runtime error: integer divide by zero 8月 02 08:19:01 cheng ollama[4056]: goroutine 38 [running]: 8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000, 0x200, ...}, ...}) 8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/llm.PredictServerFit({0xc000779bd8?, 0x4?, 0x4?}, _, {_, _, _}, {0x0, 0x0, 0x0}, ...) 8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc0002d5000, 0xc000442ec0, {0xc0004622c8?, 0x4?, 0x4?}, 0xc0004ebd18) 8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc0000b4f00, {0x22aeab50, 0xc0000bf950}) 8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 02 08:19:01 cheng ollama[4056]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 02 08:19:01 cheng ollama[4056]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 02 08:19:01 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 02 08:19:01 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 02 08:19:01 cheng systemd[1]: ollama.service: Consumed 20.346s CPU time. 8月 02 08:19:04 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 2. 8月 02 08:19:04 cheng systemd[1]: Stopped Ollama Service. 8月 02 08:19:04 cheng systemd[1]: ollama.service: Consumed 20.346s CPU time. 8月 02 08:19:04 cheng systemd[1]: Started Ollama Service. 8月 02 08:19:04 cheng ollama[5419]: 2024/08/02 08:19:04 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" 8月 02 08:19:04 cheng ollama[5419]: time=2024-08-02T08:19:04.590+08:00 level=INFO source=images.go:781 msg="total blobs: 12" 8月 02 08:19:04 cheng ollama[5419]: time=2024-08-02T08:19:04.590+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0" 8月 02 08:19:04 cheng ollama[5419]: time=2024-08-02T08:19:04.591+08:00 level=INFO source=routes.go:1156 msg="Listening on [::]:11434 (version 0.3.2)" 8月 02 08:19:04 cheng ollama[5419]: time=2024-08-02T08:19:04.591+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama2561891386 error="remove /tmp/ollama2561891386: directory not empty" 8月 02 08:19:04 cheng ollama[5419]: time=2024-08-02T08:19:04.591+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3369741500 error="remove /tmp/ollama3369741500: directory not empty" 8月 02 08:19:04 cheng ollama[5419]: time=2024-08-02T08:19:04.591+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama3427016686/runners 8月 02 08:19:11 cheng ollama[5419]: time=2024-08-02T08:19:11.468+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11 rocm_v60102]" 8月 02 08:19:11 cheng ollama[5419]: time=2024-08-02T08:19:11.468+08:00 level=INFO source=gpu.go:205 msg="looking for compatible GPUs" 8月 02 08:19:12 cheng ollama[5419]: time=2024-08-02T08:19:12.013+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-5dcf737a-6e6e-f833-26db-a11d4d8a01e5 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:19:12 cheng ollama[5419]: time=2024-08-02T08:19:12.013+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ddeda9e-1883-a42a-533f-2ba63921dde2 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:19:12 cheng ollama[5419]: time=2024-08-02T08:19:12.013+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-2ccb721c-af3a-e060-accc-b64e78758119 library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 08:19:12 cheng ollama[5419]: time=2024-08-02T08:19:12.013+08:00 level=INFO source=types.go:105 msg="inference compute" id=GPU-eef18c96-f6fc-b36f-02a7-545b0e3e856a library=cuda compute=8.6 driver=12.2 name="NVIDIA GeForce RTX 3090" total="23.7 GiB" available="23.4 GiB" 8月 02 09:31:44 cheng ollama[5419]: [GIN] 2024/08/02 - 09:31:44 | 200 | 53.704µs | 127.0.0.1 | HEAD "/" 8月 02 09:31:44 cheng ollama[5419]: [GIN] 2024/08/02 - 09:31:44 | 200 | 1.251938ms | 127.0.0.1 | POST "/api/show" 8月 02 09:31:44 cheng ollama[5419]: time=2024-08-02T09:31:44.984+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size" 8月 02 09:31:44 cheng ollama[5419]: panic: runtime error: integer divide by zero 8月 02 09:31:44 cheng ollama[5419]: goroutine 66 [running]: 8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/llm.EstimateGPULayers({_, _, _}, _, {_, _, _}, {{0x0, 0x2000, 0x200, ...}, ...}) 8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/llm/memory.go:122 +0x1a1b 8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/llm.PredictServerFit({0xc0004abbd8?, 0x4?, 0x4?}, _, {_, _, _}, {0x0, 0x0, 0x0}, ...) 8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/llm/memory.go:20 +0xbd 8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/server.pickBestFullFitByLibrary(0xc00031f000, 0xc0003d6f40, {0xc00047c008?, 0x4?, 0x4?}, 0xc0004d1d18) 8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/server/sched.go:709 +0x6f3 8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/server.(*Scheduler).processPending(0xc0003b0ea0, {0x22aeab50, 0xc00061ceb0}) 8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/server/sched.go:220 +0xdfd 8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/server.(*Scheduler).Run.func1() 8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/server/sched.go:107 +0x1f 8月 02 09:31:44 cheng ollama[5419]: created by github.com/ollama/ollama/server.(*Scheduler).Run in goroutine 1 8月 02 09:31:44 cheng ollama[5419]: github.com/ollama/ollama/server/sched.go:106 +0xb4 8月 02 09:31:45 cheng systemd[1]: ollama.service: Main process exited, code=exited, status=2/INVALIDARGUMENT 8月 02 09:31:45 cheng systemd[1]: ollama.service: Failed with result 'exit-code'. 8月 02 09:31:45 cheng systemd[1]: ollama.service: Consumed 21.014s CPU time. 8月 02 09:31:48 cheng systemd[1]: ollama.service: Scheduled restart job, restart counter is at 3. 8月 02 09:31:48 cheng systemd[1]: Stopped Ollama Service. 8月 02 09:31:48 cheng systemd[1]: ollama.service: Consumed 21.014s CPU time. 8月 02 09:31:48 cheng systemd[1]: Started Ollama Service. 8月 02 09:31:48 cheng ollama[9466]: 2024/08/02 09:31:48 routes.go:1109: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/usr/share/ollama/.ollama/models 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://*] OLLAMA_RUNNERS_DIR: OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" 8月 02 09:31:48 cheng ollama[9466]: time=2024-08-02T09:31:48.107+08:00 level=INFO source=images.go:781 msg="total blobs: 12" 8月 02 09:31:48 cheng ollama[9466]: time=2024-08-02T09:31:48.107+08:00 level=INFO source=images.go:788 msg="total unused blobs removed: 0" 8月 02 09:31:48 cheng ollama[9466]: time=2024-08-02T09:31:48.107+08:00 level=INFO source=routes.go:1156 msg="Listening on [::]:11434 (version 0.3.2)" 8月 02 09:31:48 cheng ollama[9466]: time=2024-08-02T09:31:48.108+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama2561891386 error="remove /tmp/ollama2561891386: directory not empty" 8月 02 09:31:48 cheng ollama[9466]: time=2024-08-02T09:31:48.108+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3369741500 error="remove /tmp/ollama3369741500: directory not empty" 8月 02 09:31:48 cheng ollama[9466]: time=2024-08-02T09:31:48.108+08:00 level=WARN source=assets.go:100 msg="unable to cleanup stale tmpdir" path=/tmp/ollama3427016686 error="remove /tmp/ollama3427016686: directory not empty" 8月 02 09:31:48 cheng ollama[9466]: time=2024-08-02T09:31:48.108+08:00 level=INFO source=payload.go:30 msg="extracting embedded files" dir=/tmp/ollama865072436/runners'''
GiteaMirror commented 2026-04-12 14:39:23 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Aug 2, 2024):

8月 01 20:49:42 cheng ollama[1889]: time=2024-08-01T20:49:42.325+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size"
8月 01 20:49:42 cheng ollama[1889]: panic: runtime error: integer divide by zero

The model is broken, where did you download the files from?

<!-- gh-comment-id:2264949498 --> @rick-github commented on GitHub (Aug 2, 2024): ``` 8月 01 20:49:42 cheng ollama[1889]: time=2024-08-01T20:49:42.325+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size" 8月 01 20:49:42 cheng ollama[1889]: panic: runtime error: integer divide by zero ``` The model is broken, where did you download the files from?
GiteaMirror commented 2026-04-12 14:39:23 -05:00
Author
Owner
Copy Link

@SongXiaoMao commented on GitHub (Aug 4, 2024):

8月 01 20:49:42 cheng ollama[1889]: time=2024-08-01T20:49:42.325+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size"
8月 01 20:49:42 cheng ollama[1889]: panic: runtime error: integer divide by zero

The model is broken, where did you download the files from?

微信截图_20240804103214

<!-- gh-comment-id:2267260437 --> @SongXiaoMao commented on GitHub (Aug 4, 2024): > ``` > 8月 01 20:49:42 cheng ollama[1889]: time=2024-08-01T20:49:42.325+08:00 level=WARN source=memory.go:115 msg="model missing blk.0 layer size" > 8月 01 20:49:42 cheng ollama[1889]: panic: runtime error: integer divide by zero > ``` > > The model is broken, where did you download the files from? ![微信截图_20240804103214](https://github.com/user-attachments/assets/c0c5a749-759d-47e6-bf8b-153bc142853a)
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 bug
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#3823
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?