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

[GH-ISSUE #9704] ps showing 100% GPU but CPU is used #6337

New Issue
Closed
opened 2026-04-12 17:51:08 -05:00 by GiteaMirror · 5 comments
GiteaMirror commented 2026-04-12 17:51:08 -05:00
Owner
Copy Link

Originally created by @hyperu on GitHub (Mar 13, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/9704

What is the issue?

I checked #7323 & #9068 but no luck

seems an "llm server error"

ollama ps
NAME ID SIZE PROCESSOR UNTIL
opencoder:latest cd882db52297 6.8 GB 100% GPU 4 minutes from now

ollama -v
ollama version is 0.6.0

kernel version 6.13.6-arch1-1

logs concerned:
3 Mar 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.567+08:00 level=INFO source=server.go:619 msg="waiting for server to become available" status="llm server error"

nvidia-smi
Thu Mar 13 09:04:05 2025
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 570.124.04 Driver Version: 570.124.04 CUDA Version: 12.8 |
|-----------------------------------------+------------------------+----------------------+
| GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|=========================================+========================+======================|
| 0 NVIDIA GeForce RTX 4060 Ti Off | 00000000:02:00.0 On | N/A |
| 0% 38C P8 5W / 165W | 312MiB / 16380MiB | 0% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+
| 1 NVIDIA GeForce RTX 4060 Ti Off | 00000000:82:00.0 Off | N/A |
| 30% 43C P8 4W / 165W | 18MiB / 16380MiB | 0% Default |
| | | N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=========================================================================================|
| 0 N/A N/A 1087 G /usr/lib/Xorg 54MiB |
| 0 N/A N/A 1109 G /usr/bin/sddm-greeter-qt6 202MiB |
| 1 N/A N/A 1087 G /usr/lib/Xorg 4MiB |
+-----------------------------------------------------------------------------------------+

Relevant log output

3月 13 09:01:18 oldpc systemd[1]: Stopping Ollama Service...
3月 13 09:01:19 oldpc systemd[1]: ollama.service: Deactivated successfully.
3月 13 09:01:19 oldpc systemd[1]: Stopped Ollama Service.
3月 13 09:01:19 oldpc systemd[1]: ollama.service: Consumed 31min 1.585s CPU time, 1.1G memory peak.
-- Boot d8a7bf5c3c0f4d90bbfbe8e60a87e2c1 --
3月 13 09:02:20 oldpc systemd[1]: Started Ollama Service.
3月 13 09:02:20 oldpc ollama[1141]: 2025/03/13 09:02:20 routes.go:1225: INFO server config env="map[CUDA_VISIBLE_DEVICES:0,1 GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_CONTEXT_LENGTH:2048 OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/var/lib/ollama OLLAMA_MULTIUSER_CACHE:false OLLAMA_NEW_ENGINE:false OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:0 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://* vscode-webview://* vscode-file://*] OLLAMA_SCHED_SPREAD:false ROCR_VISIBLE_DEVICES: http_proxy: https_proxy: no_proxy:]"
3月 13 09:02:20 oldpc ollama[1141]: time=2025-03-13T09:02:20.456+08:00 level=INFO source=images.go:432 msg="total blobs: 24"
3月 13 09:02:20 oldpc ollama[1141]: time=2025-03-13T09:02:20.456+08:00 level=INFO source=images.go:439 msg="total unused blobs removed: 0"
3月 13 09:02:20 oldpc ollama[1141]: time=2025-03-13T09:02:20.457+08:00 level=INFO source=routes.go:1292 msg="Listening on 127.0.0.1:11434 (version 0.6.0)"
3月 13 09:02:20 oldpc ollama[1141]: time=2025-03-13T09:02:20.458+08:00 level=INFO source=gpu.go:217 msg="looking for compatible GPUs"
3月 13 09:02:21 oldpc ollama[1141]: time=2025-03-13T09:02:21.019+08:00 level=INFO source=types.go:130 msg="inference compute" id=GPU-b5db1abb-5b09-c949-8288-7bea7c68d35c library=cuda variant=v12 compute=8.9 driver=12.8 name="NVIDIA GeForce RTX 4060 Ti" total="15.6 GiB" available="15.2 GiB"
3月 13 09:02:21 oldpc ollama[1141]: time=2025-03-13T09:02:21.019+08:00 level=INFO source=types.go:130 msg="inference compute" id=GPU-a7e500ed-9c23-3cc4-c794-ca39e481c42a library=cuda variant=v12 compute=8.9 driver=12.8 name="NVIDIA GeForce RTX 4060 Ti" total="15.6 GiB" available="15.4 GiB"
3月 13 09:03:53 oldpc ollama[1141]: [GIN] 2025/03/13 - 09:03:53 | 200 |      884.22µs |       127.0.0.1 | HEAD     "/"
3月 13 09:03:53 oldpc ollama[1141]: [GIN] 2025/03/13 - 09:03:53 | 200 |   17.752334ms |       127.0.0.1 | POST     "/api/show"
3月 13 09:03:53 oldpc ollama[1141]: time=2025-03-13T09:03:53.998+08:00 level=WARN source=ggml.go:149 msg="key not found" key=llama.attention.key_length default=128
3月 13 09:03:53 oldpc ollama[1141]: time=2025-03-13T09:03:53.999+08:00 level=WARN source=ggml.go:149 msg="key not found" key=llama.attention.value_length default=128
3月 13 09:03:53 oldpc ollama[1141]: time=2025-03-13T09:03:53.999+08:00 level=INFO source=sched.go:715 msg="new model will fit in available VRAM in single GPU, loading" model=/var/lib/ollama/blobs/sha256-27d438e34ebae3b1c7aa2f43bb1ce1b053a0c039fa1dee927abd147ff4e55c55 gpu=GPU-a7e500ed-9c23-3cc4-c794-ca39e481c42a parallel=4 available=16586506240 required="6.4 GiB"
3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.408+08:00 level=INFO source=server.go:105 msg="system memory" total="62.8 GiB" free="61.1 GiB" free_swap="0 B"
3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.409+08:00 level=WARN source=ggml.go:149 msg="key not found" key=llama.attention.key_length default=128
3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.409+08:00 level=WARN source=ggml.go:149 msg="key not found" key=llama.attention.value_length default=128
3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.409+08:00 level=INFO source=server.go:138 msg=offload library=cuda layers.requested=-1 layers.model=33 layers.offload=33 layers.split="" memory.available="[15.4 GiB]" memory.gpu_overhead="0 B" memory.required.full="6.4 GiB" memory.required.partial="6.4 GiB" memory.required.kv="1.0 GiB" memory.required.allocations="[6.4 GiB]" memory.weights.total="4.9 GiB" memory.weights.repeating="4.6 GiB" memory.weights.nonrepeating="309.7 MiB" memory.graph.full="560.0 MiB" memory.graph.partial="585.0 MiB"
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: loaded meta data with 39 key-value pairs and 291 tensors from /var/lib/ollama/blobs/sha256-27d438e34ebae3b1c7aa2f43bb1ce1b053a0c039fa1dee927abd147ff4e55c55 (version GGUF V3 (latest))
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   0:                       general.architecture str              = llama
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   1:                               general.type str              = model
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   2:                               general.name str              = OpenCoder 8B Instruct
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   3:                           general.finetune str              = Instruct
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   4:                           general.basename str              = OpenCoder
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   5:                         general.size_label str              = 8B
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   6:                            general.license str              = other
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   7:                       general.license.name str              = inf
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   8:                       general.license.link str              = https://huggingface.co/infly/OpenCode...
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   9:                   general.base_model.count u32              = 1
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  10:                  general.base_model.0.name str              = OpenCoder 8B Base
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  11:          general.base_model.0.organization str              = Infly
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  12:              general.base_model.0.repo_url str              = https://huggingface.co/infly/OpenCode...
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  13:                               general.tags arr[str,1]       = ["text-generation"]
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  14:                          general.languages arr[str,2]       = ["en", "zh"]
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  15:                           general.datasets arr[str,2]       = ["OpenCoder-LLM/opencoder-sft-stage1"...
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  16:                          llama.block_count u32              = 32
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  17:                       llama.context_length u32              = 8192
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  18:                     llama.embedding_length u32              = 4096
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  19:                  llama.feed_forward_length u32              = 14336
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  20:                 llama.attention.head_count u32              = 32
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  21:              llama.attention.head_count_kv u32              = 8
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  22:                       llama.rope.freq_base f32              = 500000.000000
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  23:     llama.attention.layer_norm_rms_epsilon f32              = 0.000010
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  24:                          general.file_type u32              = 15
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  25:                           llama.vocab_size u32              = 96640
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  26:                 llama.rope.dimension_count u32              = 128
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  27:            tokenizer.ggml.add_space_prefix bool             = false
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  28:                       tokenizer.ggml.model str              = llama
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  29:                         tokenizer.ggml.pre str              = default
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  30:                      tokenizer.ggml.tokens arr[str,96640]   = ["<unk>", "<s>", "</s>", "<pad>", "<0...
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  31:                      tokenizer.ggml.scores arr[f32,96640]   = [-1000.000000, -1000.000000, -1000.00...
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  32:                  tokenizer.ggml.token_type arr[i32,96640]   = [3, 3, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, ...
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  33:                tokenizer.ggml.bos_token_id u32              = 96540
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  34:                tokenizer.ggml.eos_token_id u32              = 96539
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  35:               tokenizer.ggml.add_bos_token bool             = false
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  36:               tokenizer.ggml.add_eos_token bool             = false
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  37:                    tokenizer.chat_template str              = {% for message in messages %}{% if lo...
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  38:               general.quantization_version u32              = 2
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - type  f32:   65 tensors
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - type q4_K:  193 tensors
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - type q6_K:   33 tensors
3月 13 09:03:54 oldpc ollama[1141]: print_info: file format = GGUF V3 (latest)
3月 13 09:03:54 oldpc ollama[1141]: print_info: file type   = Q4_K - Medium
3月 13 09:03:54 oldpc ollama[1141]: print_info: file size   = 4.41 GiB (4.87 BPW)
3月 13 09:03:54 oldpc ollama[1141]: load: special tokens cache size = 46
3月 13 09:03:54 oldpc ollama[1141]: load: token to piece cache size = 0.5697 MB
3月 13 09:03:54 oldpc ollama[1141]: print_info: arch             = llama
3月 13 09:03:54 oldpc ollama[1141]: print_info: vocab_only       = 1
3月 13 09:03:54 oldpc ollama[1141]: print_info: model type       = ?B
3月 13 09:03:54 oldpc ollama[1141]: print_info: model params     = 7.77 B
3月 13 09:03:54 oldpc ollama[1141]: print_info: general.name     = OpenCoder 8B Instruct
3月 13 09:03:54 oldpc ollama[1141]: print_info: vocab type       = SPM
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_vocab          = 96640
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_merges         = 0
3月 13 09:03:54 oldpc ollama[1141]: print_info: BOS token        = 96540 '<|im_start|>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: EOS token        = 96539 '<|im_end|>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: EOT token        = 96506 '<|endoftext|>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: UNK token        = 0 '<unk>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: LF token         = 14 '<0x0A>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: EOG token        = 96500 '<|end|>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: EOG token        = 96506 '<|endoftext|>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: EOG token        = 96539 '<|im_end|>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: max token length = 384
3月 13 09:03:54 oldpc ollama[1141]: llama_model_load: vocab only - skipping tensors
3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.563+08:00 level=INFO source=server.go:405 msg="starting llama server" cmd="/usr/bin/ollama runner --model /var/lib/ollama/blobs/sha256-27d438e34ebae3b1c7aa2f43bb1ce1b053a0c039fa1dee927abd147ff4e55c55 --ctx-size 8192 --batch-size 512 --n-gpu-layers 33 --threads 20 --parallel 4 --port 33243"
3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.564+08:00 level=INFO source=sched.go:450 msg="loaded runners" count=1
3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.566+08:00 level=INFO source=server.go:585 msg="waiting for llama runner to start responding"
3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.567+08:00 level=INFO source=server.go:619 msg="waiting for server to become available" status="llm server error"
3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.580+08:00 level=INFO source=runner.go:931 msg="starting go runner"
3月 13 09:03:54 oldpc ollama[1141]: load_backend: loaded CPU backend from /usr/lib/ollama/libggml-cpu-haswell.so
3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.592+08:00 level=INFO source=ggml.go:109 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.LLAMAFILE=1 CPU.1.LLAMAFILE=1 compiler=cgo(gcc)
3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.592+08:00 level=INFO source=runner.go:991 msg="Server listening on 127.0.0.1:33243"
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: loaded meta data with 39 key-value pairs and 291 tensors from /var/lib/ollama/blobs/sha256-27d438e34ebae3b1c7aa2f43bb1ce1b053a0c039fa1dee927abd147ff4e55c55 (version GGUF V3 (latest))
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   0:                       general.architecture str              = llama
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   1:                               general.type str              = model
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   2:                               general.name str              = OpenCoder 8B Instruct
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   3:                           general.finetune str              = Instruct
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   4:                           general.basename str              = OpenCoder
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   5:                         general.size_label str              = 8B
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   6:                            general.license str              = other
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   7:                       general.license.name str              = inf
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   8:                       general.license.link str              = https://huggingface.co/infly/OpenCode...
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv   9:                   general.base_model.count u32              = 1
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  10:                  general.base_model.0.name str              = OpenCoder 8B Base
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  11:          general.base_model.0.organization str              = Infly
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  12:              general.base_model.0.repo_url str              = https://huggingface.co/infly/OpenCode...
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  13:                               general.tags arr[str,1]       = ["text-generation"]
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  14:                          general.languages arr[str,2]       = ["en", "zh"]
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  15:                           general.datasets arr[str,2]       = ["OpenCoder-LLM/opencoder-sft-stage1"...
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  16:                          llama.block_count u32              = 32
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  17:                       llama.context_length u32              = 8192
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  18:                     llama.embedding_length u32              = 4096
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  19:                  llama.feed_forward_length u32              = 14336
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  20:                 llama.attention.head_count u32              = 32
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  21:              llama.attention.head_count_kv u32              = 8
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  22:                       llama.rope.freq_base f32              = 500000.000000
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  23:     llama.attention.layer_norm_rms_epsilon f32              = 0.000010
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  24:                          general.file_type u32              = 15
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  25:                           llama.vocab_size u32              = 96640
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  26:                 llama.rope.dimension_count u32              = 128
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  27:            tokenizer.ggml.add_space_prefix bool             = false
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  28:                       tokenizer.ggml.model str              = llama
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  29:                         tokenizer.ggml.pre str              = default
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  30:                      tokenizer.ggml.tokens arr[str,96640]   = ["<unk>", "<s>", "</s>", "<pad>", "<0...
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  31:                      tokenizer.ggml.scores arr[f32,96640]   = [-1000.000000, -1000.000000, -1000.00...
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  32:                  tokenizer.ggml.token_type arr[i32,96640]   = [3, 3, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, ...
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  33:                tokenizer.ggml.bos_token_id u32              = 96540
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  34:                tokenizer.ggml.eos_token_id u32              = 96539
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  35:               tokenizer.ggml.add_bos_token bool             = false
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  36:               tokenizer.ggml.add_eos_token bool             = false
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  37:                    tokenizer.chat_template str              = {% for message in messages %}{% if lo...
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv  38:               general.quantization_version u32              = 2
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - type  f32:   65 tensors
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - type q4_K:  193 tensors
3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - type q6_K:   33 tensors
3月 13 09:03:54 oldpc ollama[1141]: print_info: file format = GGUF V3 (latest)
3月 13 09:03:54 oldpc ollama[1141]: print_info: file type   = Q4_K - Medium
3月 13 09:03:54 oldpc ollama[1141]: print_info: file size   = 4.41 GiB (4.87 BPW)
3月 13 09:03:54 oldpc ollama[1141]: load: special tokens cache size = 46
3月 13 09:03:54 oldpc ollama[1141]: load: token to piece cache size = 0.5697 MB
3月 13 09:03:54 oldpc ollama[1141]: print_info: arch             = llama
3月 13 09:03:54 oldpc ollama[1141]: print_info: vocab_only       = 0
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_ctx_train      = 8192
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_embd           = 4096
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_layer          = 32
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_head           = 32
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_head_kv        = 8
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_rot            = 128
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_swa            = 0
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_embd_head_k    = 128
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_embd_head_v    = 128
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_gqa            = 4
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_embd_k_gqa     = 1024
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_embd_v_gqa     = 1024
3月 13 09:03:54 oldpc ollama[1141]: print_info: f_norm_eps       = 0.0e+00
3月 13 09:03:54 oldpc ollama[1141]: print_info: f_norm_rms_eps   = 1.0e-05
3月 13 09:03:54 oldpc ollama[1141]: print_info: f_clamp_kqv      = 0.0e+00
3月 13 09:03:54 oldpc ollama[1141]: print_info: f_max_alibi_bias = 0.0e+00
3月 13 09:03:54 oldpc ollama[1141]: print_info: f_logit_scale    = 0.0e+00
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_ff             = 14336
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_expert         = 0
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_expert_used    = 0
3月 13 09:03:54 oldpc ollama[1141]: print_info: causal attn      = 1
3月 13 09:03:54 oldpc ollama[1141]: print_info: pooling type     = 0
3月 13 09:03:54 oldpc ollama[1141]: print_info: rope type        = 0
3月 13 09:03:54 oldpc ollama[1141]: print_info: rope scaling     = linear
3月 13 09:03:54 oldpc ollama[1141]: print_info: freq_base_train  = 500000.0
3月 13 09:03:54 oldpc ollama[1141]: print_info: freq_scale_train = 1
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_ctx_orig_yarn  = 8192
3月 13 09:03:54 oldpc ollama[1141]: print_info: rope_finetuned   = unknown
3月 13 09:03:54 oldpc ollama[1141]: print_info: ssm_d_conv       = 0
3月 13 09:03:54 oldpc ollama[1141]: print_info: ssm_d_inner      = 0
3月 13 09:03:54 oldpc ollama[1141]: print_info: ssm_d_state      = 0
3月 13 09:03:54 oldpc ollama[1141]: print_info: ssm_dt_rank      = 0
3月 13 09:03:54 oldpc ollama[1141]: print_info: ssm_dt_b_c_rms   = 0
3月 13 09:03:54 oldpc ollama[1141]: print_info: model type       = 8B
3月 13 09:03:54 oldpc ollama[1141]: print_info: model params     = 7.77 B
3月 13 09:03:54 oldpc ollama[1141]: print_info: general.name     = OpenCoder 8B Instruct
3月 13 09:03:54 oldpc ollama[1141]: print_info: vocab type       = SPM
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_vocab          = 96640
3月 13 09:03:54 oldpc ollama[1141]: print_info: n_merges         = 0
3月 13 09:03:54 oldpc ollama[1141]: print_info: BOS token        = 96540 '<|im_start|>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: EOS token        = 96539 '<|im_end|>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: EOT token        = 96506 '<|endoftext|>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: UNK token        = 0 '<unk>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: LF token         = 14 '<0x0A>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: EOG token        = 96500 '<|end|>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: EOG token        = 96506 '<|endoftext|>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: EOG token        = 96539 '<|im_end|>'
3月 13 09:03:54 oldpc ollama[1141]: print_info: max token length = 384
3月 13 09:03:54 oldpc ollama[1141]: load_tensors: loading model tensors, this can take a while... (mmap = true)
3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.819+08:00 level=INFO source=server.go:619 msg="waiting for server to become available" status="llm server loading model"
3月 13 09:03:59 oldpc ollama[1141]: load_tensors:   CPU_Mapped model buffer size =  4514.53 MiB
3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: n_seq_max     = 4
3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: n_ctx         = 8192
3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: n_ctx_per_seq = 2048
3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: n_batch       = 2048
3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: n_ubatch      = 512
3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: flash_attn    = 0
3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: freq_base     = 500000.0
3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: freq_scale    = 1
3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: n_ctx_per_seq (2048) < n_ctx_train (8192) -- the full capacity of the model will not be utilized
3月 13 09:03:59 oldpc ollama[1141]: llama_kv_cache_init: kv_size = 8192, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 32, can_shift = 1
3月 13 09:04:00 oldpc ollama[1141]: llama_kv_cache_init:        CPU KV buffer size =  1024.00 MiB
3月 13 09:04:00 oldpc ollama[1141]: llama_init_from_model: KV self size  = 1024.00 MiB, K (f16):  512.00 MiB, V (f16):  512.00 MiB
3月 13 09:04:00 oldpc ollama[1141]: llama_init_from_model:        CPU  output buffer size =     1.54 MiB
3月 13 09:04:00 oldpc ollama[1141]: llama_init_from_model:        CPU compute buffer size =   560.01 MiB
3月 13 09:04:00 oldpc ollama[1141]: llama_init_from_model: graph nodes  = 1030
3月 13 09:04:00 oldpc ollama[1141]: llama_init_from_model: graph splits = 1
3月 13 09:04:00 oldpc ollama[1141]: time=2025-03-13T09:04:00.341+08:00 level=INFO source=server.go:624 msg="llama runner started in 5.78 seconds"
3月 13 09:04:00 oldpc ollama[1141]: [GIN] 2025/03/13 - 09:04:00 | 200 |  6.816201089s |       127.0.0.1 | POST     "/api/generate"

OS

Linux

GPU

Nvidia

CPU

Intel

Ollama version

0.60

Originally created by @hyperu on GitHub (Mar 13, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/9704 ### What is the issue? I checked #7323 & #9068 but no luck seems an "llm server error" ollama ps NAME ID SIZE PROCESSOR UNTIL opencoder:latest cd882db52297 6.8 GB 100% GPU 4 minutes from now ollama -v ollama version is 0.6.0 kernel version 6.13.6-arch1-1 logs concerned: 3 Mar 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.567+08:00 level=INFO source=server.go:619 msg="waiting for server to become available" status="llm server error" nvidia-smi Thu Mar 13 09:04:05 2025 +-----------------------------------------------------------------------------------------+ | NVIDIA-SMI 570.124.04 Driver Version: 570.124.04 CUDA Version: 12.8 | |-----------------------------------------+------------------------+----------------------+ | GPU Name Persistence-M | Bus-Id Disp.A | Volatile Uncorr. ECC | | Fan Temp Perf Pwr:Usage/Cap | Memory-Usage | GPU-Util Compute M. | | | | MIG M. | |=========================================+========================+======================| | 0 NVIDIA GeForce RTX 4060 Ti Off | 00000000:02:00.0 On | N/A | | 0% 38C P8 5W / 165W | 312MiB / 16380MiB | 0% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+ | 1 NVIDIA GeForce RTX 4060 Ti Off | 00000000:82:00.0 Off | N/A | | 30% 43C P8 4W / 165W | 18MiB / 16380MiB | 0% Default | | | | N/A | +-----------------------------------------+------------------------+----------------------+ +-----------------------------------------------------------------------------------------+ | Processes: | | GPU GI CI PID Type Process name GPU Memory | | ID ID Usage | |=========================================================================================| | 0 N/A N/A 1087 G /usr/lib/Xorg 54MiB | | 0 N/A N/A 1109 G /usr/bin/sddm-greeter-qt6 202MiB | | 1 N/A N/A 1087 G /usr/lib/Xorg 4MiB | +-----------------------------------------------------------------------------------------+ ### Relevant log output ```shell 3月 13 09:01:18 oldpc systemd[1]: Stopping Ollama Service... 3月 13 09:01:19 oldpc systemd[1]: ollama.service: Deactivated successfully. 3月 13 09:01:19 oldpc systemd[1]: Stopped Ollama Service. 3月 13 09:01:19 oldpc systemd[1]: ollama.service: Consumed 31min 1.585s CPU time, 1.1G memory peak. -- Boot d8a7bf5c3c0f4d90bbfbe8e60a87e2c1 -- 3月 13 09:02:20 oldpc systemd[1]: Started Ollama Service. 3月 13 09:02:20 oldpc ollama[1141]: 2025/03/13 09:02:20 routes.go:1225: INFO server config env="map[CUDA_VISIBLE_DEVICES:0,1 GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_CONTEXT_LENGTH:2048 OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://127.0.0.1:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:/var/lib/ollama OLLAMA_MULTIUSER_CACHE:false OLLAMA_NEW_ENGINE:false OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:0 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://* vscode-webview://* vscode-file://*] OLLAMA_SCHED_SPREAD:false ROCR_VISIBLE_DEVICES: http_proxy: https_proxy: no_proxy:]" 3月 13 09:02:20 oldpc ollama[1141]: time=2025-03-13T09:02:20.456+08:00 level=INFO source=images.go:432 msg="total blobs: 24" 3月 13 09:02:20 oldpc ollama[1141]: time=2025-03-13T09:02:20.456+08:00 level=INFO source=images.go:439 msg="total unused blobs removed: 0" 3月 13 09:02:20 oldpc ollama[1141]: time=2025-03-13T09:02:20.457+08:00 level=INFO source=routes.go:1292 msg="Listening on 127.0.0.1:11434 (version 0.6.0)" 3月 13 09:02:20 oldpc ollama[1141]: time=2025-03-13T09:02:20.458+08:00 level=INFO source=gpu.go:217 msg="looking for compatible GPUs" 3月 13 09:02:21 oldpc ollama[1141]: time=2025-03-13T09:02:21.019+08:00 level=INFO source=types.go:130 msg="inference compute" id=GPU-b5db1abb-5b09-c949-8288-7bea7c68d35c library=cuda variant=v12 compute=8.9 driver=12.8 name="NVIDIA GeForce RTX 4060 Ti" total="15.6 GiB" available="15.2 GiB" 3月 13 09:02:21 oldpc ollama[1141]: time=2025-03-13T09:02:21.019+08:00 level=INFO source=types.go:130 msg="inference compute" id=GPU-a7e500ed-9c23-3cc4-c794-ca39e481c42a library=cuda variant=v12 compute=8.9 driver=12.8 name="NVIDIA GeForce RTX 4060 Ti" total="15.6 GiB" available="15.4 GiB" 3月 13 09:03:53 oldpc ollama[1141]: [GIN] 2025/03/13 - 09:03:53 | 200 | 884.22µs | 127.0.0.1 | HEAD "/" 3月 13 09:03:53 oldpc ollama[1141]: [GIN] 2025/03/13 - 09:03:53 | 200 | 17.752334ms | 127.0.0.1 | POST "/api/show" 3月 13 09:03:53 oldpc ollama[1141]: time=2025-03-13T09:03:53.998+08:00 level=WARN source=ggml.go:149 msg="key not found" key=llama.attention.key_length default=128 3月 13 09:03:53 oldpc ollama[1141]: time=2025-03-13T09:03:53.999+08:00 level=WARN source=ggml.go:149 msg="key not found" key=llama.attention.value_length default=128 3月 13 09:03:53 oldpc ollama[1141]: time=2025-03-13T09:03:53.999+08:00 level=INFO source=sched.go:715 msg="new model will fit in available VRAM in single GPU, loading" model=/var/lib/ollama/blobs/sha256-27d438e34ebae3b1c7aa2f43bb1ce1b053a0c039fa1dee927abd147ff4e55c55 gpu=GPU-a7e500ed-9c23-3cc4-c794-ca39e481c42a parallel=4 available=16586506240 required="6.4 GiB" 3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.408+08:00 level=INFO source=server.go:105 msg="system memory" total="62.8 GiB" free="61.1 GiB" free_swap="0 B" 3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.409+08:00 level=WARN source=ggml.go:149 msg="key not found" key=llama.attention.key_length default=128 3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.409+08:00 level=WARN source=ggml.go:149 msg="key not found" key=llama.attention.value_length default=128 3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.409+08:00 level=INFO source=server.go:138 msg=offload library=cuda layers.requested=-1 layers.model=33 layers.offload=33 layers.split="" memory.available="[15.4 GiB]" memory.gpu_overhead="0 B" memory.required.full="6.4 GiB" memory.required.partial="6.4 GiB" memory.required.kv="1.0 GiB" memory.required.allocations="[6.4 GiB]" memory.weights.total="4.9 GiB" memory.weights.repeating="4.6 GiB" memory.weights.nonrepeating="309.7 MiB" memory.graph.full="560.0 MiB" memory.graph.partial="585.0 MiB" 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: loaded meta data with 39 key-value pairs and 291 tensors from /var/lib/ollama/blobs/sha256-27d438e34ebae3b1c7aa2f43bb1ce1b053a0c039fa1dee927abd147ff4e55c55 (version GGUF V3 (latest)) 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 0: general.architecture str = llama 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 1: general.type str = model 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 2: general.name str = OpenCoder 8B Instruct 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 3: general.finetune str = Instruct 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 4: general.basename str = OpenCoder 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 5: general.size_label str = 8B 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 6: general.license str = other 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 7: general.license.name str = inf 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 8: general.license.link str = https://huggingface.co/infly/OpenCode... 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 9: general.base_model.count u32 = 1 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 10: general.base_model.0.name str = OpenCoder 8B Base 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 11: general.base_model.0.organization str = Infly 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 12: general.base_model.0.repo_url str = https://huggingface.co/infly/OpenCode... 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 13: general.tags arr[str,1] = ["text-generation"] 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 14: general.languages arr[str,2] = ["en", "zh"] 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 15: general.datasets arr[str,2] = ["OpenCoder-LLM/opencoder-sft-stage1"... 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 16: llama.block_count u32 = 32 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 17: llama.context_length u32 = 8192 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 18: llama.embedding_length u32 = 4096 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 19: llama.feed_forward_length u32 = 14336 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 20: llama.attention.head_count u32 = 32 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 21: llama.attention.head_count_kv u32 = 8 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 22: llama.rope.freq_base f32 = 500000.000000 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 23: llama.attention.layer_norm_rms_epsilon f32 = 0.000010 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 24: general.file_type u32 = 15 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 25: llama.vocab_size u32 = 96640 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 26: llama.rope.dimension_count u32 = 128 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 27: tokenizer.ggml.add_space_prefix bool = false 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 28: tokenizer.ggml.model str = llama 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 29: tokenizer.ggml.pre str = default 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 30: tokenizer.ggml.tokens arr[str,96640] = ["<unk>", "<s>", "</s>", "<pad>", "<0... 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 31: tokenizer.ggml.scores arr[f32,96640] = [-1000.000000, -1000.000000, -1000.00... 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 32: tokenizer.ggml.token_type arr[i32,96640] = [3, 3, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, ... 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 33: tokenizer.ggml.bos_token_id u32 = 96540 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 34: tokenizer.ggml.eos_token_id u32 = 96539 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 35: tokenizer.ggml.add_bos_token bool = false 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 36: tokenizer.ggml.add_eos_token bool = false 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 37: tokenizer.chat_template str = {% for message in messages %}{% if lo... 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 38: general.quantization_version u32 = 2 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - type f32: 65 tensors 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - type q4_K: 193 tensors 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - type q6_K: 33 tensors 3月 13 09:03:54 oldpc ollama[1141]: print_info: file format = GGUF V3 (latest) 3月 13 09:03:54 oldpc ollama[1141]: print_info: file type = Q4_K - Medium 3月 13 09:03:54 oldpc ollama[1141]: print_info: file size = 4.41 GiB (4.87 BPW) 3月 13 09:03:54 oldpc ollama[1141]: load: special tokens cache size = 46 3月 13 09:03:54 oldpc ollama[1141]: load: token to piece cache size = 0.5697 MB 3月 13 09:03:54 oldpc ollama[1141]: print_info: arch = llama 3月 13 09:03:54 oldpc ollama[1141]: print_info: vocab_only = 1 3月 13 09:03:54 oldpc ollama[1141]: print_info: model type = ?B 3月 13 09:03:54 oldpc ollama[1141]: print_info: model params = 7.77 B 3月 13 09:03:54 oldpc ollama[1141]: print_info: general.name = OpenCoder 8B Instruct 3月 13 09:03:54 oldpc ollama[1141]: print_info: vocab type = SPM 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_vocab = 96640 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_merges = 0 3月 13 09:03:54 oldpc ollama[1141]: print_info: BOS token = 96540 '<|im_start|>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: EOS token = 96539 '<|im_end|>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: EOT token = 96506 '<|endoftext|>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: UNK token = 0 '<unk>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: LF token = 14 '<0x0A>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: EOG token = 96500 '<|end|>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: EOG token = 96506 '<|endoftext|>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: EOG token = 96539 '<|im_end|>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: max token length = 384 3月 13 09:03:54 oldpc ollama[1141]: llama_model_load: vocab only - skipping tensors 3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.563+08:00 level=INFO source=server.go:405 msg="starting llama server" cmd="/usr/bin/ollama runner --model /var/lib/ollama/blobs/sha256-27d438e34ebae3b1c7aa2f43bb1ce1b053a0c039fa1dee927abd147ff4e55c55 --ctx-size 8192 --batch-size 512 --n-gpu-layers 33 --threads 20 --parallel 4 --port 33243" 3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.564+08:00 level=INFO source=sched.go:450 msg="loaded runners" count=1 3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.566+08:00 level=INFO source=server.go:585 msg="waiting for llama runner to start responding" 3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.567+08:00 level=INFO source=server.go:619 msg="waiting for server to become available" status="llm server error" 3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.580+08:00 level=INFO source=runner.go:931 msg="starting go runner" 3月 13 09:03:54 oldpc ollama[1141]: load_backend: loaded CPU backend from /usr/lib/ollama/libggml-cpu-haswell.so 3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.592+08:00 level=INFO source=ggml.go:109 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.LLAMAFILE=1 CPU.1.LLAMAFILE=1 compiler=cgo(gcc) 3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.592+08:00 level=INFO source=runner.go:991 msg="Server listening on 127.0.0.1:33243" 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: loaded meta data with 39 key-value pairs and 291 tensors from /var/lib/ollama/blobs/sha256-27d438e34ebae3b1c7aa2f43bb1ce1b053a0c039fa1dee927abd147ff4e55c55 (version GGUF V3 (latest)) 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 0: general.architecture str = llama 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 1: general.type str = model 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 2: general.name str = OpenCoder 8B Instruct 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 3: general.finetune str = Instruct 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 4: general.basename str = OpenCoder 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 5: general.size_label str = 8B 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 6: general.license str = other 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 7: general.license.name str = inf 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 8: general.license.link str = https://huggingface.co/infly/OpenCode... 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 9: general.base_model.count u32 = 1 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 10: general.base_model.0.name str = OpenCoder 8B Base 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 11: general.base_model.0.organization str = Infly 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 12: general.base_model.0.repo_url str = https://huggingface.co/infly/OpenCode... 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 13: general.tags arr[str,1] = ["text-generation"] 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 14: general.languages arr[str,2] = ["en", "zh"] 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 15: general.datasets arr[str,2] = ["OpenCoder-LLM/opencoder-sft-stage1"... 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 16: llama.block_count u32 = 32 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 17: llama.context_length u32 = 8192 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 18: llama.embedding_length u32 = 4096 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 19: llama.feed_forward_length u32 = 14336 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 20: llama.attention.head_count u32 = 32 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 21: llama.attention.head_count_kv u32 = 8 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 22: llama.rope.freq_base f32 = 500000.000000 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 23: llama.attention.layer_norm_rms_epsilon f32 = 0.000010 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 24: general.file_type u32 = 15 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 25: llama.vocab_size u32 = 96640 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 26: llama.rope.dimension_count u32 = 128 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 27: tokenizer.ggml.add_space_prefix bool = false 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 28: tokenizer.ggml.model str = llama 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 29: tokenizer.ggml.pre str = default 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 30: tokenizer.ggml.tokens arr[str,96640] = ["<unk>", "<s>", "</s>", "<pad>", "<0... 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 31: tokenizer.ggml.scores arr[f32,96640] = [-1000.000000, -1000.000000, -1000.00... 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 32: tokenizer.ggml.token_type arr[i32,96640] = [3, 3, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, ... 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 33: tokenizer.ggml.bos_token_id u32 = 96540 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 34: tokenizer.ggml.eos_token_id u32 = 96539 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 35: tokenizer.ggml.add_bos_token bool = false 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 36: tokenizer.ggml.add_eos_token bool = false 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 37: tokenizer.chat_template str = {% for message in messages %}{% if lo... 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - kv 38: general.quantization_version u32 = 2 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - type f32: 65 tensors 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - type q4_K: 193 tensors 3月 13 09:03:54 oldpc ollama[1141]: llama_model_loader: - type q6_K: 33 tensors 3月 13 09:03:54 oldpc ollama[1141]: print_info: file format = GGUF V3 (latest) 3月 13 09:03:54 oldpc ollama[1141]: print_info: file type = Q4_K - Medium 3月 13 09:03:54 oldpc ollama[1141]: print_info: file size = 4.41 GiB (4.87 BPW) 3月 13 09:03:54 oldpc ollama[1141]: load: special tokens cache size = 46 3月 13 09:03:54 oldpc ollama[1141]: load: token to piece cache size = 0.5697 MB 3月 13 09:03:54 oldpc ollama[1141]: print_info: arch = llama 3月 13 09:03:54 oldpc ollama[1141]: print_info: vocab_only = 0 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_ctx_train = 8192 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_embd = 4096 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_layer = 32 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_head = 32 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_head_kv = 8 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_rot = 128 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_swa = 0 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_embd_head_k = 128 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_embd_head_v = 128 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_gqa = 4 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_embd_k_gqa = 1024 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_embd_v_gqa = 1024 3月 13 09:03:54 oldpc ollama[1141]: print_info: f_norm_eps = 0.0e+00 3月 13 09:03:54 oldpc ollama[1141]: print_info: f_norm_rms_eps = 1.0e-05 3月 13 09:03:54 oldpc ollama[1141]: print_info: f_clamp_kqv = 0.0e+00 3月 13 09:03:54 oldpc ollama[1141]: print_info: f_max_alibi_bias = 0.0e+00 3月 13 09:03:54 oldpc ollama[1141]: print_info: f_logit_scale = 0.0e+00 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_ff = 14336 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_expert = 0 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_expert_used = 0 3月 13 09:03:54 oldpc ollama[1141]: print_info: causal attn = 1 3月 13 09:03:54 oldpc ollama[1141]: print_info: pooling type = 0 3月 13 09:03:54 oldpc ollama[1141]: print_info: rope type = 0 3月 13 09:03:54 oldpc ollama[1141]: print_info: rope scaling = linear 3月 13 09:03:54 oldpc ollama[1141]: print_info: freq_base_train = 500000.0 3月 13 09:03:54 oldpc ollama[1141]: print_info: freq_scale_train = 1 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_ctx_orig_yarn = 8192 3月 13 09:03:54 oldpc ollama[1141]: print_info: rope_finetuned = unknown 3月 13 09:03:54 oldpc ollama[1141]: print_info: ssm_d_conv = 0 3月 13 09:03:54 oldpc ollama[1141]: print_info: ssm_d_inner = 0 3月 13 09:03:54 oldpc ollama[1141]: print_info: ssm_d_state = 0 3月 13 09:03:54 oldpc ollama[1141]: print_info: ssm_dt_rank = 0 3月 13 09:03:54 oldpc ollama[1141]: print_info: ssm_dt_b_c_rms = 0 3月 13 09:03:54 oldpc ollama[1141]: print_info: model type = 8B 3月 13 09:03:54 oldpc ollama[1141]: print_info: model params = 7.77 B 3月 13 09:03:54 oldpc ollama[1141]: print_info: general.name = OpenCoder 8B Instruct 3月 13 09:03:54 oldpc ollama[1141]: print_info: vocab type = SPM 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_vocab = 96640 3月 13 09:03:54 oldpc ollama[1141]: print_info: n_merges = 0 3月 13 09:03:54 oldpc ollama[1141]: print_info: BOS token = 96540 '<|im_start|>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: EOS token = 96539 '<|im_end|>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: EOT token = 96506 '<|endoftext|>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: UNK token = 0 '<unk>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: LF token = 14 '<0x0A>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: EOG token = 96500 '<|end|>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: EOG token = 96506 '<|endoftext|>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: EOG token = 96539 '<|im_end|>' 3月 13 09:03:54 oldpc ollama[1141]: print_info: max token length = 384 3月 13 09:03:54 oldpc ollama[1141]: load_tensors: loading model tensors, this can take a while... (mmap = true) 3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.819+08:00 level=INFO source=server.go:619 msg="waiting for server to become available" status="llm server loading model" 3月 13 09:03:59 oldpc ollama[1141]: load_tensors: CPU_Mapped model buffer size = 4514.53 MiB 3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: n_seq_max = 4 3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: n_ctx = 8192 3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: n_ctx_per_seq = 2048 3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: n_batch = 2048 3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: n_ubatch = 512 3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: flash_attn = 0 3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: freq_base = 500000.0 3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: freq_scale = 1 3月 13 09:03:59 oldpc ollama[1141]: llama_init_from_model: n_ctx_per_seq (2048) < n_ctx_train (8192) -- the full capacity of the model will not be utilized 3月 13 09:03:59 oldpc ollama[1141]: llama_kv_cache_init: kv_size = 8192, offload = 1, type_k = 'f16', type_v = 'f16', n_layer = 32, can_shift = 1 3月 13 09:04:00 oldpc ollama[1141]: llama_kv_cache_init: CPU KV buffer size = 1024.00 MiB 3月 13 09:04:00 oldpc ollama[1141]: llama_init_from_model: KV self size = 1024.00 MiB, K (f16): 512.00 MiB, V (f16): 512.00 MiB 3月 13 09:04:00 oldpc ollama[1141]: llama_init_from_model: CPU output buffer size = 1.54 MiB 3月 13 09:04:00 oldpc ollama[1141]: llama_init_from_model: CPU compute buffer size = 560.01 MiB 3月 13 09:04:00 oldpc ollama[1141]: llama_init_from_model: graph nodes = 1030 3月 13 09:04:00 oldpc ollama[1141]: llama_init_from_model: graph splits = 1 3月 13 09:04:00 oldpc ollama[1141]: time=2025-03-13T09:04:00.341+08:00 level=INFO source=server.go:624 msg="llama runner started in 5.78 seconds" 3月 13 09:04:00 oldpc ollama[1141]: [GIN] 2025/03/13 - 09:04:00 | 200 | 6.816201089s | 127.0.0.1 | POST "/api/generate" ``` ### OS Linux ### GPU Nvidia ### CPU Intel ### Ollama version 0.60
GiteaMirror added the bug label 2026-04-12 17:51:08 -05:00
GiteaMirror commented 2026-04-12 17:51:09 -05:00
Author
Owner
Copy Link

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

3月 13 09:03:54 oldpc ollama[1141]: load_backend: loaded CPU backend from /usr/lib/ollama/libggml-cpu-haswell.so
3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.592+08:00 level=INFO source=ggml.go:109 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.LLAMAFILE=1 CPU.1.LLAMAFILE=1 compiler=cgo(gcc)

The GPU backend wasn't loaded. If you set OLLAMA_DEBUG=1 in the server environment there might be some useful info. Warning, the debug logging is voluminous.

<!-- gh-comment-id:2719556275 --> @rick-github commented on GitHub (Mar 13, 2025): ``` 3月 13 09:03:54 oldpc ollama[1141]: load_backend: loaded CPU backend from /usr/lib/ollama/libggml-cpu-haswell.so 3月 13 09:03:54 oldpc ollama[1141]: time=2025-03-13T09:03:54.592+08:00 level=INFO source=ggml.go:109 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.LLAMAFILE=1 CPU.1.LLAMAFILE=1 compiler=cgo(gcc) ``` The GPU backend wasn't loaded. If you set `OLLAMA_DEBUG=1` in the server environment there might be some useful info. Warning, the debug logging is voluminous.
GiteaMirror commented 2026-04-12 17:51:09 -05:00
Author
Owner
Copy Link

@Jay021 commented on GitHub (Mar 16, 2025):

修改后的中文内容：

我的Ollama在运行Ollama ps命令时，虽然显示GPU使用率为100%，但实际运行大模型时，CPU运转飞快，而NVIDIA MSI显示GPU占用率为0%，显存也未占用。我尝试过卸载并重新安装Ollama，但问题依旧。直到我看到了这个帖子：

#9266
by https://github.com/Hsq12138

我在PATH中添加了以下路径（请根据自己的电脑用户名修改PC部分）：
C:\Users\<你的用户名>\AppData\Local\Programs\Ollama\lib\ollama

完美解决了问题，你可以试试。

修改后的英文翻译：

When running the Ollama ps command, my Ollama shows 100% GPU usage, but in reality, the CPU is running at full speed while the NVIDIA MSI shows 0% GPU usage and no VRAM is being utilized. I tried uninstalling and reinstalling Ollama, but the issue persisted. Until I came across this post:

#9266
by https://github.com/Hsq12138

I added the following path to the PATH environment variable (please replace <YourUsername> with your actual computer username):
C:\Users\<YourUsername>\AppData\Local\Programs\Ollama\lib\ollama

This perfectly resolved the issue. You can give it a try.

<!-- gh-comment-id:2727175562 --> @Jay021 commented on GitHub (Mar 16, 2025): ### 修改后的中文内容： 我的Ollama在运行`Ollama ps`命令时，虽然显示GPU使用率为100%，但实际运行大模型时，CPU运转飞快，而NVIDIA MSI显示GPU占用率为0%，显存也未占用。我尝试过卸载并重新安装Ollama，但问题依旧。直到我看到了这个帖子： #9266 by https://github.com/Hsq12138 我在PATH中添加了以下路径（请根据自己的电脑用户名修改`PC`部分）： `C:\Users\<你的用户名>\AppData\Local\Programs\Ollama\lib\ollama` 完美解决了问题，你可以试试。 --- ### 修改后的英文翻译： When running the `Ollama ps` command, my Ollama shows 100% GPU usage, but in reality, the CPU is running at full speed while the NVIDIA MSI shows 0% GPU usage and no VRAM is being utilized. I tried uninstalling and reinstalling Ollama, but the issue persisted. Until I came across this post: #9266 by https://github.com/Hsq12138 I added the following path to the PATH environment variable (please replace `<YourUsername>` with your actual computer username): `C:\Users\<YourUsername>\AppData\Local\Programs\Ollama\lib\ollama` This perfectly resolved the issue. You can give it a try.
GiteaMirror commented 2026-04-12 17:51:09 -05:00
Author
Owner
Copy Link

@hyperu commented on GitHub (Mar 21, 2025):

OLLAMA_DEBUG=1

21 16:28:30 old-pc ollama[15112]: time=2025-03-21T16:28:30.475+08:00 level=INFO source=server.go:405 msg="starting llama server" cmd
="/usr/bin/ollama runner --model /var/lib/ollama/blobs/sha256-27d438e34ebae3b1c7aa2f43bb1ce1b053a0c039fa1dee927abd147ff4e55c55 --ctx-size 81
92 --batch-size 512 --n-gpu-layers 33 --verbose --threads 20 --parallel 4 --port 32961"
21 16:28:30 old-pc ollama[15112]: time=2025-03-21T16:28:30.475+08:00 level=DEBUG source=server.go:423 msg=subprocess environment="[P
ATH=/usr/local/sbin:/usr/local/bin:/usr/bin LD_LIBRARY_PATH=/usr/lib/ollama CUDA_VISIBLE_DEVICES=GPU-a7e500ed-9c23-3cc4-c694-ca39e481c42a]"
21 16:28:30 old-pc ollama[15112]: time=2025-03-21T16:28:30.476+08:00 level=INFO source=sched.go:450 msg="loaded runners" count=1
21 16:28:30 old-pc ollama[15112]: time=2025-03-21T16:28:30.476+08:00 level=INFO source=server.go:580 msg="waiting for llama runner t
o start responding"
21 16:28:30 old-pc ollama[15112]: time=2025-03-21T16:28:30.477+08:00 level=INFO source=server.go:614 msg="waiting for server to beco
me available" status="llm server error"
21 16:28:30 old-pc ollama[15112]: time=2025-03-21T16:28:30.491+08:00 level=INFO source=runner.go:846 msg="starting go runner"
21 16:28:30 old-pc ollama[15112]: time=2025-03-21T16:28:30.491+08:00 level=DEBUG source=ggml.go:99 msg="ggml backend load all from p
ath" path=/usr/lib/ollama

logs in debug level, no more details for "llm server error"

ollama version is 0.62

<!-- gh-comment-id:2742690811 --> @hyperu commented on GitHub (Mar 21, 2025): > OLLAMA_DEBUG=1 21 16:28:30 old-pc ollama[15112]: time=2025-03-21T16:28:30.475+08:00 level=INFO source=server.go:405 msg="starting llama server" cmd ="/usr/bin/ollama runner --model /var/lib/ollama/blobs/sha256-27d438e34ebae3b1c7aa2f43bb1ce1b053a0c039fa1dee927abd147ff4e55c55 --ctx-size 81 92 --batch-size 512 --n-gpu-layers 33 --verbose --threads 20 --parallel 4 --port 32961" 21 16:28:30 old-pc ollama[15112]: time=2025-03-21T16:28:30.475+08:00 level=DEBUG source=server.go:423 msg=subprocess environment="[P ATH=/usr/local/sbin:/usr/local/bin:/usr/bin LD_LIBRARY_PATH=/usr/lib/ollama CUDA_VISIBLE_DEVICES=GPU-a7e500ed-9c23-3cc4-c694-ca39e481c42a]" 21 16:28:30 old-pc ollama[15112]: time=2025-03-21T16:28:30.476+08:00 level=INFO source=sched.go:450 msg="loaded runners" count=1 21 16:28:30 old-pc ollama[15112]: time=2025-03-21T16:28:30.476+08:00 level=INFO source=server.go:580 msg="waiting for llama runner t o start responding" 21 16:28:30 old-pc ollama[15112]: time=2025-03-21T16:28:30.477+08:00 level=INFO source=server.go:614 msg="waiting for server to beco me available" **status="llm server error"** 21 16:28:30 old-pc ollama[15112]: time=2025-03-21T16:28:30.491+08:00 level=INFO source=runner.go:846 msg="starting go runner" 21 16:28:30 old-pc ollama[15112]: time=2025-03-21T16:28:30.491+08:00 level=DEBUG source=ggml.go:99 msg="ggml backend load all from p ath" path=/usr/lib/ollama logs in debug level, no more details for "llm server error" ollama version is 0.62
GiteaMirror commented 2026-04-12 17:51:10 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Mar 21, 2025):

Full log.

<!-- gh-comment-id:2742703697 --> @rick-github commented on GitHub (Mar 21, 2025): Full log.
GiteaMirror commented 2026-04-12 17:51:11 -05:00
Author
Owner
Copy Link

@hyperu commented on GitHub (Mar 21, 2025):

fixed with:
sudo pacman -S ollama-cuda

so I got:
llama-cuda /usr/lib/ollama/cuda_v12/
ollama-cuda /usr/lib/ollama/cuda_v12/libggml-cuda.so

though I am sure ollama-cuda was not reqired months before

<!-- gh-comment-id:2742824805 --> @hyperu commented on GitHub (Mar 21, 2025): fixed with: **sudo pacman -S ollama-cuda** so I got: llama-cuda /usr/lib/ollama/cuda_v12/ ollama-cuda /usr/lib/ollama/cuda_v12/libggml-cuda.so though I am sure ollama-cuda was not reqired months before
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#6337
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?