github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-08 17:49:24 -05:00
Code Issues 14.6k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #10980] Ollama Ignores System Prompts When Used with Qwen-Agent RAG Example #53750

New Issue
Closed
opened 2026-04-29 04:39:58 -05:00 by GiteaMirror · 4 comments
GiteaMirror commented 2026-04-29 04:39:58 -05:00
Owner
Copy Link

Originally created by @Harry-Up on GitHub (Jun 5, 2025).
Original GitHub issue: https://github.com/ollama/ollama/issues/10980

What is the issue?

Environment

  • ​​OS​​: Windows 11
  • ​​Ollama Version​​: 0.7.0
  • ​​Python​​: 3.11.11
  • ​​Qwen-Agent​​: 0.0.24

​​Tested Models​​:

  • qwen3:30b (local via Ollama)
  • qwen3:14b (local via Ollama)
  • qwen3:14b-q8_0 (local via Ollama)
  • qwq (local via Ollama)

Problem Description
When running the examples/assistant_rag.py from Qwen-Agent with ​​Ollama-served Qwen models​​, the models consistently ​​fail to process RAG tasks​​ (file content is ignored). This occurs ​​only when system prompts are used​​ in their standard position.

Key observations:

​​- Workaround​​: Moving the system prompt to the user prompt position resolves the issue.
​​- Comparison Tests​​:
✅ ​​Aliyun Bailian​​ models (qwen-latest-plus, qwen3:30b, qwen3:8b) handle RAG correctly with system prompts.
✅ ​​vLLM Docker​​ (using unsloth/Qwen3-14B-unsloth-bnb-4bit) also processes RAG successfully.

  • ​​Suspected Root Cause​​: Ollama's template engine ​​may not inject system prompts correctly​​ for Qwen models.

Relevant log output

OS

Windows

GPU

Nvidia

CPU

AMD

Ollama version

0.7.0

Originally created by @Harry-Up on GitHub (Jun 5, 2025). Original GitHub issue: https://github.com/ollama/ollama/issues/10980 ### What is the issue? **Environment** - ​​OS​​: Windows 11 - ​​Ollama Version​​: 0.7.0 - ​​Python​​: 3.11.11 - ​​Qwen-Agent​​: 0.0.24 **​​Tested Models**​​: - qwen3:30b (local via Ollama) - qwen3:14b (local via Ollama) - qwen3:14b-q8_0 (local via Ollama) - qwq (local via Ollama) **Problem Description** When running the examples/assistant_rag.py from [Qwen-Agent](https://github.com/QwenLM/qwen-agent) with ​​Ollama-served Qwen models​​, the models consistently ​​fail to process RAG tasks​​ (file content is ignored). This occurs ​​only when system prompts are used​​ in their standard position. **Key observations**: ​​- **Workaround​​**: Moving the system prompt to the user prompt position resolves the issue. ​​- **Comparison Tests​​**: ✅ ​​Aliyun Bailian​​ models (qwen-latest-plus, qwen3:30b, qwen3:8b) handle RAG correctly with system prompts. ✅ ​​vLLM Docker​​ (using unsloth/Qwen3-14B-unsloth-bnb-4bit) also processes RAG successfully. - ​​**Suspected Root Cause**​​: Ollama's template engine ​​may not inject system prompts correctly​​ for Qwen models. ### Relevant log output ```shell ``` ### OS Windows ### GPU Nvidia ### CPU AMD ### Ollama version 0.7.0
GiteaMirror added the bug label 2026-04-29 04:39:58 -05:00
GiteaMirror commented 2026-04-29 04:40:06 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Jun 5, 2025):

​​Suspected Root Cause​​: Ollama's template engine ​​may not inject system prompts correctly​​ for Qwen models.

More likely is that the context is too small. Server logs with OLLAMA_DEBUG=1 will aid in debugging.

<!-- gh-comment-id:2944086771 --> @rick-github commented on GitHub (Jun 5, 2025): > ​​Suspected Root Cause​​: Ollama's template engine ​​may not inject system prompts correctly​​ for Qwen models. More likely is that the context is too small. [Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) with `OLLAMA_DEBUG=1` will aid in debugging.
GiteaMirror commented 2026-04-29 04:40:07 -05:00
Author
Owner
Copy Link

@Harry-Up commented on GitHub (Jun 6, 2025):

​​Suspected Root Cause​​: Ollama's template engine ​​may not inject system prompts correctly​​ for Qwen models.

More likely is that the context is too small. Server logs with OLLAMA_DEBUG=1 will aid in debugging.

Yeah, I update Ollama and try it again with OLLAMA_DEBUG=1. The problem keeps as well. Here is the server log.

time=2025-06-06T21:47:24.165+08:00 level=INFO source=routes.go:1234 msg="server config" env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_CONTEXT_LENGTH:4096 OLLAMA_DEBUG:DEBUG OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:C:\Users\weihb\.ollama\models 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:]"
time=2025-06-06T21:47:24.169+08:00 level=INFO source=images.go:479 msg="total blobs: 13"
time=2025-06-06T21:47:24.170+08:00 level=INFO source=images.go:486 msg="total unused blobs removed: 0"
time=2025-06-06T21:47:24.170+08:00 level=INFO source=routes.go:1287 msg="Listening on [::]:11434 (version 0.9.0)"
time=2025-06-06T21:47:24.170+08:00 level=DEBUG source=sched.go:108 msg="starting llm scheduler"
time=2025-06-06T21:47:24.170+08:00 level=INFO source=gpu.go:217 msg="looking for compatible GPUs"
time=2025-06-06T21:47:24.170+08:00 level=INFO source=gpu_windows.go:167 msg=packages count=1
time=2025-06-06T21:47:24.170+08:00 level=INFO source=gpu_windows.go:214 msg="" package=0 cores=8 efficiency=0 threads=16
time=2025-06-06T21:47:24.170+08:00 level=DEBUG source=gpu.go:98 msg="searching for GPU discovery libraries for NVIDIA"
time=2025-06-06T21:47:24.170+08:00 level=DEBUG source=gpu.go:501 msg="Searching for GPU library" name=nvml.dll
time=2025-06-06T21:47:24.170+08:00 level=DEBUG source=gpu.go:525 msg="gpu library search" globs="[C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\nvml.dll C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\bin\nvml.dll C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\libnvvp\nvml.dll C:\WINDOWS\system32\nvml.dll C:\WINDOWS\nvml.dll C:\WINDOWS\System32\Wbem\nvml.dll C:\WINDOWS\System32\WindowsPowerShell\v1.0\nvml.dll C:\WINDOWS\System32\OpenSSH\nvml.dll C:\Program Files (x86)\Windows Kits\10\Windows Performance Toolkit\nvml.dll C:\Program Files\NVIDIA Corporation\NVIDIA App\NvDLISR\nvml.dll C:\Program Files (x86)\NVIDIA Corporation\PhysX\Common\nvml.dll C:\Program Files\Git\cmd\nvml.dll C:\Program Files\Docker\Docker\resources\bin\nvml.dll C:\Program Files\Microsoft Visual Studio\2022\Community\VC\Tools\MSVC\14.43.34808\bin\Hostx64\x64\nvml.dll C:\Program Files (x86)\Windows Kits\10\bin\10.0.22621.0\x64\nvml.dll C:\MinGW\bin\nvml.dll C:\Program Files\NVIDIA Corporation\Nsight Compute 2022.3.0\nvml.dll C:\Program Files\dotnet\nvml.dll C:\Users\weihb\AppData\Local\Microsoft\WindowsApps\nvml.dll C:\Users\weihb\AppData\Local\Programs\Microsoft VS Code\bin\nvml.dll C:\Users\weihb\AppData\Local\Programs\Ollama\nvml.dll c:\Windows\System32\nvml.dll]"
time=2025-06-06T21:47:24.170+08:00 level=DEBUG source=gpu.go:529 msg="skipping PhysX cuda library path" path="C:\Program Files (x86)\NVIDIA Corporation\PhysX\Common\nvml.dll"
time=2025-06-06T21:47:24.171+08:00 level=DEBUG source=gpu.go:558 msg="discovered GPU libraries" paths="[C:\WINDOWS\system32\nvml.dll c:\Windows\System32\nvml.dll]"
time=2025-06-06T21:47:24.180+08:00 level=DEBUG source=gpu.go:111 msg="nvidia-ml loaded" library=C:\WINDOWS\system32\nvml.dll
time=2025-06-06T21:47:24.180+08:00 level=DEBUG source=gpu.go:501 msg="Searching for GPU library" name=nvcuda.dll
time=2025-06-06T21:47:24.180+08:00 level=DEBUG source=gpu.go:525 msg="gpu library search" globs="[C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\nvcuda.dll C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\bin\nvcuda.dll C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\libnvvp\nvcuda.dll C:\WINDOWS\system32\nvcuda.dll C:\WINDOWS\nvcuda.dll C:\WINDOWS\System32\Wbem\nvcuda.dll C:\WINDOWS\System32\WindowsPowerShell\v1.0\nvcuda.dll C:\WINDOWS\System32\OpenSSH\nvcuda.dll C:\Program Files (x86)\Windows Kits\10\Windows Performance Toolkit\nvcuda.dll C:\Program Files\NVIDIA Corporation\NVIDIA App\NvDLISR\nvcuda.dll C:\Program Files (x86)\NVIDIA Corporation\PhysX\Common\nvcuda.dll C:\Program Files\Git\cmd\nvcuda.dll C:\Program Files\Docker\Docker\resources\bin\nvcuda.dll C:\Program Files\Microsoft Visual Studio\2022\Community\VC\Tools\MSVC\14.43.34808\bin\Hostx64\x64\nvcuda.dll C:\Program Files (x86)\Windows Kits\10\bin\10.0.22621.0\x64\nvcuda.dll C:\MinGW\bin\nvcuda.dll C:\Program Files\NVIDIA Corporation\Nsight Compute 2022.3.0\nvcuda.dll C:\Program Files\dotnet\nvcuda.dll C:\Users\weihb\AppData\Local\Microsoft\WindowsApps\nvcuda.dll C:\Users\weihb\AppData\Local\Programs\Microsoft VS Code\bin\nvcuda.dll C:\Users\weihb\AppData\Local\Programs\Ollama\nvcuda.dll c:\windows\system\nvcuda.dll]"
time=2025-06-06T21:47:24.180+08:00 level=DEBUG source=gpu.go:529 msg="skipping PhysX cuda library path" path="C:\Program Files (x86)\NVIDIA Corporation\PhysX\Common\nvcuda.dll"
time=2025-06-06T21:47:24.180+08:00 level=DEBUG source=gpu.go:558 msg="discovered GPU libraries" paths=[C:\WINDOWS\system32\nvcuda.dll]
initializing C:\WINDOWS\system32\nvcuda.dll
dlsym: cuInit - 00007FFE357E5F80
dlsym: cuDriverGetVersion - 00007FFE357E6020
dlsym: cuDeviceGetCount - 00007FFE357E6816
dlsym: cuDeviceGet - 00007FFE357E6810
dlsym: cuDeviceGetAttribute - 00007FFE357E6170
dlsym: cuDeviceGetUuid - 00007FFE357E6822
dlsym: cuDeviceGetName - 00007FFE357E681C
dlsym: cuCtxCreate_v3 - 00007FFE357E6894
dlsym: cuMemGetInfo_v2 - 00007FFE357E6996
dlsym: cuCtxDestroy - 00007FFE357E68A6
calling cuInit
calling cuDriverGetVersion
raw version 0x2f30
CUDA driver version: 12.8
calling cuDeviceGetCount
device count 1
time=2025-06-06T21:47:24.188+08:00 level=DEBUG source=gpu.go:125 msg="detected GPUs" count=1 library=C:\WINDOWS\system32\nvcuda.dll
[GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f] CUDA totalMem 32606mb
[GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f] CUDA freeMem 30843mb
[GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f] Compute Capability 12.0
time=2025-06-06T21:47:24.283+08:00 level=DEBUG source=amd_hip_windows.go:88 msg=hipDriverGetVersion version=60342560
time=2025-06-06T21:47:24.283+08:00 level=INFO source=amd_hip_windows.go:103 msg="AMD ROCm reports no devices found"
time=2025-06-06T21:47:24.283+08:00 level=INFO source=amd_windows.go:49 msg="no compatible amdgpu devices detected"
releasing cuda driver library
releasing nvml library
time=2025-06-06T21:47:24.283+08:00 level=INFO source=types.go:130 msg="inference compute" id=GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f library=cuda variant=v12 compute=12.0 driver=12.8 name="NVIDIA GeForce RTX 5090 D" total="31.8 GiB" available="30.1 GiB"
[GIN] 2025/06/06 - 21:47:31 | 200 | 582.8µs | 127.0.0.1 | GET "/api/version"
[GIN] 2025/06/06 - 21:47:35 | 200 | 0s | 127.0.0.1 | HEAD "/"
[GIN] 2025/06/06 - 21:47:35 | 200 | 25.7005ms | 127.0.0.1 | GET "/api/tags"
time=2025-06-06T21:48:23.357+08:00 level=DEBUG source=ggml.go:155 msg="key not found" key=general.alignment default=32
time=2025-06-06T21:48:23.359+08:00 level=DEBUG source=gpu.go:391 msg="updating system memory data" before.total="31.6 GiB" before.free="15.5 GiB" before.free_swap="24.0 GiB" now.total="31.6 GiB" now.free="15.0 GiB" now.free_swap="22.9 GiB"
time=2025-06-06T21:48:23.375+08:00 level=DEBUG source=gpu.go:441 msg="updating cuda memory data" gpu=GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f name="NVIDIA GeForce RTX 5090 D" overhead="0 B" before.total="31.8 GiB" before.free="30.1 GiB" now.total="31.8 GiB" now.free="29.7 GiB" now.used="2.2 GiB"
releasing nvml library
time=2025-06-06T21:48:23.376+08:00 level=DEBUG source=sched.go:185 msg="updating default concurrency" OLLAMA_MAX_LOADED_MODELS=3 gpu_count=1
time=2025-06-06T21:48:23.383+08:00 level=DEBUG source=ggml.go:155 msg="key not found" key=general.alignment default=32
time=2025-06-06T21:48:23.391+08:00 level=DEBUG source=ggml.go:155 msg="key not found" key=general.alignment default=32
time=2025-06-06T21:48:23.392+08:00 level=DEBUG source=sched.go:228 msg="loading first model" model=C:\Users\weihb.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac
time=2025-06-06T21:48:23.393+08:00 level=DEBUG source=memory.go:111 msg=evaluating library=cuda gpu_count=1 available="[29.7 GiB]"
time=2025-06-06T21:48:23.393+08:00 level=DEBUG source=ggml.go:155 msg="key not found" key=qwen3moe.vision.block_count default=0
time=2025-06-06T21:48:23.393+08:00 level=INFO source=sched.go:788 msg="new model will fit in available VRAM in single GPU, loading" model=C:\Users\weihb.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac gpu=GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f parallel=2 available=31845679104 required="19.8 GiB"
time=2025-06-06T21:48:23.393+08:00 level=DEBUG source=gpu.go:391 msg="updating system memory data" before.total="31.6 GiB" before.free="15.0 GiB" before.free_swap="22.9 GiB" now.total="31.6 GiB" now.free="15.0 GiB" now.free_swap="22.9 GiB"
time=2025-06-06T21:48:23.405+08:00 level=DEBUG source=gpu.go:441 msg="updating cuda memory data" gpu=GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f name="NVIDIA GeForce RTX 5090 D" overhead="0 B" before.total="31.8 GiB" before.free="29.7 GiB" now.total="31.8 GiB" now.free="29.7 GiB" now.used="2.2 GiB"
releasing nvml library
time=2025-06-06T21:48:23.406+08:00 level=INFO source=server.go:135 msg="system memory" total="31.6 GiB" free="15.0 GiB" free_swap="22.9 GiB"
time=2025-06-06T21:48:23.406+08:00 level=DEBUG source=memory.go:111 msg=evaluating library=cuda gpu_count=1 available="[29.7 GiB]"
time=2025-06-06T21:48:23.406+08:00 level=DEBUG source=ggml.go:155 msg="key not found" key=qwen3moe.vision.block_count default=0
time=2025-06-06T21:48:23.406+08:00 level=INFO source=server.go:168 msg=offload library=cuda layers.requested=-1 layers.model=49 layers.offload=49 layers.split="" memory.available="[29.7 GiB]" memory.gpu_overhead="0 B" memory.required.full="19.8 GiB" memory.required.partial="19.8 GiB" memory.required.kv="768.0 MiB" memory.required.allocations="[19.8 GiB]" memory.weights.total="17.2 GiB" memory.weights.repeating="16.9 GiB" memory.weights.nonrepeating="243.4 MiB" memory.graph.full="1.0 GiB" memory.graph.partial="1.0 GiB"
time=2025-06-06T21:48:23.406+08:00 level=DEBUG source=server.go:284 msg="compatible gpu libraries" compatible="[cuda_v12 cuda_v11]"
llama_model_loader: loaded meta data with 31 key-value pairs and 579 tensors from C:\Users\weihb.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = qwen3moe
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = Qwen3 30B A3B
llama_model_loader: - kv 3: general.basename str = Qwen3
llama_model_loader: - kv 4: general.size_label str = 30B-A3B
llama_model_loader: - kv 5: general.license str = apache-2.0
llama_model_loader: - kv 6: qwen3moe.block_count u32 = 48
llama_model_loader: - kv 7: qwen3moe.context_length u32 = 40960
llama_model_loader: - kv 8: qwen3moe.embedding_length u32 = 2048
llama_model_loader: - kv 9: qwen3moe.feed_forward_length u32 = 6144
llama_model_loader: - kv 10: qwen3moe.attention.head_count u32 = 32
llama_model_loader: - kv 11: qwen3moe.attention.head_count_kv u32 = 4
llama_model_loader: - kv 12: qwen3moe.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 13: qwen3moe.attention.layer_norm_rms_epsilon f32 = 0.000001
llama_model_loader: - kv 14: qwen3moe.expert_used_count u32 = 8
llama_model_loader: - kv 15: qwen3moe.attention.key_length u32 = 128
llama_model_loader: - kv 16: qwen3moe.attention.value_length u32 = 128
llama_model_loader: - kv 17: qwen3moe.expert_count u32 = 128
llama_model_loader: - kv 18: qwen3moe.expert_feed_forward_length u32 = 768
llama_model_loader: - kv 19: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 20: tokenizer.ggml.pre str = qwen2
llama_model_loader: - kv 21: tokenizer.ggml.tokens arr[str,151936] = ["!", """, "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 22: tokenizer.ggml.token_type arr[i32,151936] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 23: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
llama_model_loader: - kv 24: tokenizer.ggml.eos_token_id u32 = 151645
llama_model_loader: - kv 25: tokenizer.ggml.padding_token_id u32 = 151643
llama_model_loader: - kv 26: tokenizer.ggml.bos_token_id u32 = 151643
llama_model_loader: - kv 27: tokenizer.ggml.add_bos_token bool = false
llama_model_loader: - kv 28: tokenizer.chat_template str = {%- if tools %}\n {{- '<|im_start|>...
llama_model_loader: - kv 29: general.quantization_version u32 = 2
llama_model_loader: - kv 30: general.file_type u32 = 15
llama_model_loader: - type f32: 241 tensors
llama_model_loader: - type f16: 48 tensors
llama_model_loader: - type q4_K: 265 tensors
llama_model_loader: - type q6_K: 25 tensors
print_info: file format = GGUF V3 (latest)
print_info: file type = Q4_K - Medium
print_info: file size = 17.34 GiB (4.88 BPW)
init_tokenizer: initializing tokenizer for type 2
load: control token: 151659 '<|fim_prefix|>' is not marked as EOG
load: control token: 151656 '<|video_pad|>' is not marked as EOG
load: control token: 151655 '<|image_pad|>' is not marked as EOG
load: control token: 151653 '<|vision_end|>' is not marked as EOG
load: control token: 151652 '<|vision_start|>' is not marked as EOG
load: control token: 151651 '<|quad_end|>' is not marked as EOG
load: control token: 151649 '<|box_end|>' is not marked as EOG
load: control token: 151648 '<|box_start|>' is not marked as EOG
load: control token: 151646 '<|object_ref_start|>' is not marked as EOG
load: control token: 151644 '<|im_start|>' is not marked as EOG
load: control token: 151661 '<|fim_suffix|>' is not marked as EOG
load: control token: 151647 '<|object_ref_end|>' is not marked as EOG
load: control token: 151660 '<|fim_middle|>' is not marked as EOG
load: control token: 151654 '<|vision_pad|>' is not marked as EOG
load: control token: 151650 '<|quad_start|>' is not marked as EOG
load: special tokens cache size = 26
load: token to piece cache size = 0.9311 MB
print_info: arch = qwen3moe
print_info: vocab_only = 1
print_info: model type = ?B
print_info: model params = 30.53 B
print_info: general.name = Qwen3 30B A3B
print_info: n_ff_exp = 0
print_info: vocab type = BPE
print_info: n_vocab = 151936
print_info: n_merges = 151387
print_info: BOS token = 151643 '<|endoftext|>'
print_info: EOS token = 151645 '<|im_end|>'
print_info: EOT token = 151645 '<|im_end|>'
print_info: PAD token = 151643 '<|endoftext|>'
print_info: LF token = 198 'Ċ'
print_info: FIM PRE token = 151659 '<|fim_prefix|>'
print_info: FIM SUF token = 151661 '<|fim_suffix|>'
print_info: FIM MID token = 151660 '<|fim_middle|>'
print_info: FIM PAD token = 151662 '<|fim_pad|>'
print_info: FIM REP token = 151663 '<|repo_name|>'
print_info: FIM SEP token = 151664 '<|file_sep|>'
print_info: EOG token = 151643 '<|endoftext|>'
print_info: EOG token = 151645 '<|im_end|>'
print_info: EOG token = 151662 '<|fim_pad|>'
print_info: EOG token = 151663 '<|repo_name|>'
print_info: EOG token = 151664 '<|file_sep|>'
print_info: max token length = 256
llama_model_load: vocab only - skipping tensors
time=2025-06-06T21:48:23.529+08:00 level=DEBUG source=server.go:360 msg="adding gpu library" path=C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\cuda_v12
time=2025-06-06T21:48:23.529+08:00 level=DEBUG source=server.go:367 msg="adding gpu dependency paths" paths=[C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\cuda_v12]
time=2025-06-06T21:48:23.529+08:00 level=INFO source=server.go:431 msg="starting llama server" cmd="C:\Users\weihb\AppData\Local\Programs\Ollama\ollama.exe runner --model C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac --ctx-size 8192 --batch-size 512 --n-gpu-layers 49 --threads 8 --no-mmap --parallel 2 --port 53373"
time=2025-06-06T21:48:23.529+08:00 level=DEBUG source=server.go:432 msg=subprocess CUDA_PATH="C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8" OLLAMA_API_KEY=!@#$1234qaZ OLLAMA_DEBUG=1 OLLAMA_HOST=0.0.0.0:11434 OLLAMA_MAX_LOADED_MODELS=3 PATH="C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\cuda_v12;C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\cuda_v12;C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama;C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\bin;C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v12.8\libnvvp;C:\WINDOWS\system32;C:\WINDOWS;C:\WINDOWS\System32\Wbem;C:\WINDOWS\System32\WindowsPowerShell\v1.0\;C:\WINDOWS\System32\OpenSSH\;C:\Program Files (x86)\Windows Kits\10\Windows Performance Toolkit\;C:\Program Files\NVIDIA Corporation\NVIDIA App\NvDLISR;C:\Program Files (x86)\NVIDIA Corporation\PhysX\Common;C:\Program Files\Git\cmd;C:\Program Files\Docker\Docker\resources\bin;C:\Program Files\Microsoft Visual Studio\2022\Community\VC\Tools\MSVC\14.43.34808\bin\Hostx64\x64;C:\Program Files (x86)\Windows Kits\10\bin\10.0.22621.0\x64;C:\MinGW\bin;C:\Program Files\NVIDIA Corporation\Nsight Compute 2022.3.0\;C:\Program Files\dotnet\;C:\Users\weihb\AppData\Local\Microsoft\WindowsApps;C:\Users\weihb\AppData\Local\Programs\Microsoft VS Code\bin;C:\Users\weihb\AppData\Local\Programs\Ollama;C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama" OLLAMA_LIBRARY_PATH=C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama;C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\cuda_v12 CUDA_VISIBLE_DEVICES=GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f
time=2025-06-06T21:48:23.532+08:00 level=INFO source=sched.go:483 msg="loaded runners" count=1
time=2025-06-06T21:48:23.532+08:00 level=INFO source=server.go:591 msg="waiting for llama runner to start responding"
time=2025-06-06T21:48:23.533+08:00 level=INFO source=server.go:625 msg="waiting for server to become available" status="llm server error"
time=2025-06-06T21:48:23.556+08:00 level=INFO source=runner.go:815 msg="starting go runner"
time=2025-06-06T21:48:23.559+08:00 level=DEBUG source=ggml.go:94 msg="ggml backend load all from path" path=C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama
load_backend: loaded CPU backend from C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\ggml-cpu-icelake.dll
time=2025-06-06T21:48:23.661+08:00 level=DEBUG source=ggml.go:94 msg="ggml backend load all from path" path=C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\cuda_v12
ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 1 CUDA devices:
Device 0: NVIDIA GeForce RTX 5090 D, compute capability 12.0, VMM: yes
load_backend: loaded CUDA backend from C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\cuda_v12\ggml-cuda.dll
time=2025-06-06T21:48:39.340+08:00 level=INFO source=ggml.go:104 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.BMI2=1 CPU.0.AVX512=1 CPU.0.AVX512_VBMI=1 CPU.0.AVX512_VNNI=1 CPU.0.LLAMAFILE=1 CPU.1.LLAMAFILE=1 CUDA.0.ARCHS=500,600,610,700,750,800,860,870,890,900,1200 CUDA.0.USE_GRAPHS=1 CUDA.0.PEER_MAX_BATCH_SIZE=128 compiler=cgo(clang)
time=2025-06-06T21:48:39.341+08:00 level=INFO source=runner.go:874 msg="Server listening on 127.0.0.1:53373"
llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce RTX 5090 D) - 30843 MiB free
llama_model_loader: loaded meta data with 31 key-value pairs and 579 tensors from C:\Users\weihb.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac (version GGUF V3 (latest))
llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output.
llama_model_loader: - kv 0: general.architecture str = qwen3moe
llama_model_loader: - kv 1: general.type str = model
llama_model_loader: - kv 2: general.name str = Qwen3 30B A3B
llama_model_loader: - kv 3: general.basename str = Qwen3
llama_model_loader: - kv 4: general.size_label str = 30B-A3B
llama_model_loader: - kv 5: general.license str = apache-2.0
llama_model_loader: - kv 6: qwen3moe.block_count u32 = 48
llama_model_loader: - kv 7: qwen3moe.context_length u32 = 40960
llama_model_loader: - kv 8: qwen3moe.embedding_length u32 = 2048
llama_model_loader: - kv 9: qwen3moe.feed_forward_length u32 = 6144
llama_model_loader: - kv 10: qwen3moe.attention.head_count u32 = 32
llama_model_loader: - kv 11: qwen3moe.attention.head_count_kv u32 = 4
llama_model_loader: - kv 12: qwen3moe.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 13: qwen3moe.attention.layer_norm_rms_epsilon f32 = 0.000001
llama_model_loader: - kv 14: qwen3moe.expert_used_count u32 = 8
llama_model_loader: - kv 15: qwen3moe.attention.key_length u32 = 128
llama_model_loader: - kv 16: qwen3moe.attention.value_length u32 = 128
llama_model_loader: - kv 17: qwen3moe.expert_count u32 = 128
llama_model_loader: - kv 18: qwen3moe.expert_feed_forward_length u32 = 768
llama_model_loader: - kv 19: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 20: tokenizer.ggml.pre str = qwen2
llama_model_loader: - kv 21: tokenizer.ggml.tokens arr[str,151936] = ["!", """, "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 22: tokenizer.ggml.token_type arr[i32,151936] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 23: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",...
llama_model_loader: - kv 24: tokenizer.ggml.eos_token_id u32 = 151645
llama_model_loader: - kv 25: tokenizer.ggml.padding_token_id u32 = 151643
llama_model_loader: - kv 26: tokenizer.ggml.bos_token_id u32 = 151643
llama_model_loader: - kv 27: tokenizer.ggml.add_bos_token bool = false
llama_model_loader: - kv 28: tokenizer.chat_template str = {%- if tools %}\n {{- '<|im_start|>...
llama_model_loader: - kv 29: general.quantization_version u32 = 2
llama_model_loader: - kv 30: general.file_type u32 = 15
llama_model_loader: - type f32: 241 tensors
llama_model_loader: - type f16: 48 tensors
llama_model_loader: - type q4_K: 265 tensors
llama_model_loader: - type q6_K: 25 tensors
print_info: file format = GGUF V3 (latest)
print_info: file type = Q4_K - Medium
print_info: file size = 17.34 GiB (4.88 BPW)
init_tokenizer: initializing tokenizer for type 2
load: control token: 151659 '<|fim_prefix|>' is not marked as EOG
load: control token: 151656 '<|video_pad|>' is not marked as EOG
load: control token: 151655 '<|image_pad|>' is not marked as EOG
load: control token: 151653 '<|vision_end|>' is not marked as EOG
load: control token: 151652 '<|vision_start|>' is not marked as EOG
load: control token: 151651 '<|quad_end|>' is not marked as EOG
load: control token: 151649 '<|box_end|>' is not marked as EOG
load: control token: 151648 '<|box_start|>' is not marked as EOG
load: control token: 151646 '<|object_ref_start|>' is not marked as EOG
load: control token: 151644 '<|im_start|>' is not marked as EOG
load: control token: 151661 '<|fim_suffix|>' is not marked as EOG
load: control token: 151647 '<|object_ref_end|>' is not marked as EOG
load: control token: 151660 '<|fim_middle|>' is not marked as EOG
load: control token: 151654 '<|vision_pad|>' is not marked as EOG
load: control token: 151650 '<|quad_start|>' is not marked as EOG
load: special tokens cache size = 26
time=2025-06-06T21:48:39.553+08:00 level=INFO source=server.go:625 msg="waiting for server to become available" status="llm server loading model"
load: token to piece cache size = 0.9311 MB
print_info: arch = qwen3moe
print_info: vocab_only = 0
print_info: n_ctx_train = 40960
print_info: n_embd = 2048
print_info: n_layer = 48
print_info: n_head = 32
print_info: n_head_kv = 4
print_info: n_rot = 128
print_info: n_swa = 0
print_info: n_swa_pattern = 1
print_info: n_embd_head_k = 128
print_info: n_embd_head_v = 128
print_info: n_gqa = 8
print_info: n_embd_k_gqa = 512
print_info: n_embd_v_gqa = 512
print_info: f_norm_eps = 0.0e+00
print_info: f_norm_rms_eps = 1.0e-06
print_info: f_clamp_kqv = 0.0e+00
print_info: f_max_alibi_bias = 0.0e+00
print_info: f_logit_scale = 0.0e+00
print_info: f_attn_scale = 0.0e+00
print_info: n_ff = 6144
print_info: n_expert = 128
print_info: n_expert_used = 8
print_info: causal attn = 1
print_info: pooling type = 0
print_info: rope type = 2
print_info: rope scaling = linear
print_info: freq_base_train = 1000000.0
print_info: freq_scale_train = 1
print_info: n_ctx_orig_yarn = 40960
print_info: rope_finetuned = unknown
print_info: ssm_d_conv = 0
print_info: ssm_d_inner = 0
print_info: ssm_d_state = 0
print_info: ssm_dt_rank = 0
print_info: ssm_dt_b_c_rms = 0
print_info: model type = 30B.A3B
print_info: model params = 30.53 B
print_info: general.name = Qwen3 30B A3B
print_info: n_ff_exp = 768
print_info: vocab type = BPE
print_info: n_vocab = 151936
print_info: n_merges = 151387
print_info: BOS token = 151643 '<|endoftext|>'
print_info: EOS token = 151645 '<|im_end|>'
print_info: EOT token = 151645 '<|im_end|>'
print_info: PAD token = 151643 '<|endoftext|>'
print_info: LF token = 198 'Ċ'
print_info: FIM PRE token = 151659 '<|fim_prefix|>'
print_info: FIM SUF token = 151661 '<|fim_suffix|>'
print_info: FIM MID token = 151660 '<|fim_middle|>'
print_info: FIM PAD token = 151662 '<|fim_pad|>'
print_info: FIM REP token = 151663 '<|repo_name|>'
print_info: FIM SEP token = 151664 '<|file_sep|>'
print_info: EOG token = 151643 '<|endoftext|>'
print_info: EOG token = 151645 '<|im_end|>'
print_info: EOG token = 151662 '<|fim_pad|>'
print_info: EOG token = 151663 '<|repo_name|>'
print_info: EOG token = 151664 '<|file_sep|>'
print_info: max token length = 256
load_tensors: loading model tensors, this can take a while... (mmap = false)
load_tensors: layer 0 assigned to device CUDA0, is_swa = 0
load_tensors: layer 1 assigned to device CUDA0, is_swa = 0
load_tensors: layer 2 assigned to device CUDA0, is_swa = 0
load_tensors: layer 3 assigned to device CUDA0, is_swa = 0
load_tensors: layer 4 assigned to device CUDA0, is_swa = 0
load_tensors: layer 5 assigned to device CUDA0, is_swa = 0
load_tensors: layer 6 assigned to device CUDA0, is_swa = 0
load_tensors: layer 7 assigned to device CUDA0, is_swa = 0
load_tensors: layer 8 assigned to device CUDA0, is_swa = 0
load_tensors: layer 9 assigned to device CUDA0, is_swa = 0
load_tensors: layer 10 assigned to device CUDA0, is_swa = 0
load_tensors: layer 11 assigned to device CUDA0, is_swa = 0
load_tensors: layer 12 assigned to device CUDA0, is_swa = 0
load_tensors: layer 13 assigned to device CUDA0, is_swa = 0
load_tensors: layer 14 assigned to device CUDA0, is_swa = 0
load_tensors: layer 15 assigned to device CUDA0, is_swa = 0
load_tensors: layer 16 assigned to device CUDA0, is_swa = 0
load_tensors: layer 17 assigned to device CUDA0, is_swa = 0
load_tensors: layer 18 assigned to device CUDA0, is_swa = 0
load_tensors: layer 19 assigned to device CUDA0, is_swa = 0
load_tensors: layer 20 assigned to device CUDA0, is_swa = 0
load_tensors: layer 21 assigned to device CUDA0, is_swa = 0
load_tensors: layer 22 assigned to device CUDA0, is_swa = 0
load_tensors: layer 23 assigned to device CUDA0, is_swa = 0
load_tensors: layer 24 assigned to device CUDA0, is_swa = 0
load_tensors: layer 25 assigned to device CUDA0, is_swa = 0
load_tensors: layer 26 assigned to device CUDA0, is_swa = 0
load_tensors: layer 27 assigned to device CUDA0, is_swa = 0
load_tensors: layer 28 assigned to device CUDA0, is_swa = 0
load_tensors: layer 29 assigned to device CUDA0, is_swa = 0
load_tensors: layer 30 assigned to device CUDA0, is_swa = 0
load_tensors: layer 31 assigned to device CUDA0, is_swa = 0
load_tensors: layer 32 assigned to device CUDA0, is_swa = 0
load_tensors: layer 33 assigned to device CUDA0, is_swa = 0
load_tensors: layer 34 assigned to device CUDA0, is_swa = 0
load_tensors: layer 35 assigned to device CUDA0, is_swa = 0
load_tensors: layer 36 assigned to device CUDA0, is_swa = 0
load_tensors: layer 37 assigned to device CUDA0, is_swa = 0
load_tensors: layer 38 assigned to device CUDA0, is_swa = 0
load_tensors: layer 39 assigned to device CUDA0, is_swa = 0
load_tensors: layer 40 assigned to device CUDA0, is_swa = 0
load_tensors: layer 41 assigned to device CUDA0, is_swa = 0
load_tensors: layer 42 assigned to device CUDA0, is_swa = 0
load_tensors: layer 43 assigned to device CUDA0, is_swa = 0
load_tensors: layer 44 assigned to device CUDA0, is_swa = 0
load_tensors: layer 45 assigned to device CUDA0, is_swa = 0
load_tensors: layer 46 assigned to device CUDA0, is_swa = 0
load_tensors: layer 47 assigned to device CUDA0, is_swa = 0
load_tensors: layer 48 assigned to device CUDA0, is_swa = 0
load_tensors: tensor 'token_embd.weight' (q4_K) (and 0 others) cannot be used with preferred buffer type CUDA_Host, using CPU instead
load_tensors: offloading 48 repeating layers to GPU
load_tensors: offloading output layer to GPU
load_tensors: offloaded 49/49 layers to GPU
load_tensors: CUDA0 model buffer size = 17587.24 MiB
load_tensors: CPU model buffer size = 166.92 MiB
load_all_data: using async uploads for device CUDA0, buffer type CUDA0, backend CUDA0
time=2025-06-06T21:48:39.803+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.00"
time=2025-06-06T21:48:40.053+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.03"
time=2025-06-06T21:48:40.304+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.06"
time=2025-06-06T21:48:40.554+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.09"
time=2025-06-06T21:48:40.804+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.11"
time=2025-06-06T21:48:41.054+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.14"
time=2025-06-06T21:48:41.304+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.18"
time=2025-06-06T21:48:41.555+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.20"
time=2025-06-06T21:48:41.805+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.22"
time=2025-06-06T21:48:42.056+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.25"
time=2025-06-06T21:48:42.306+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.28"
time=2025-06-06T21:48:42.556+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.31"
time=2025-06-06T21:48:42.806+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.33"
time=2025-06-06T21:48:43.056+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.35"
time=2025-06-06T21:48:43.307+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.38"
time=2025-06-06T21:48:43.557+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.40"
time=2025-06-06T21:48:43.807+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.43"
time=2025-06-06T21:48:44.057+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.46"
time=2025-06-06T21:48:44.308+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.49"
time=2025-06-06T21:48:44.558+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.52"
time=2025-06-06T21:48:44.809+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.55"
time=2025-06-06T21:48:45.059+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.58"
time=2025-06-06T21:48:45.309+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.61"
time=2025-06-06T21:48:45.559+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.63"
time=2025-06-06T21:48:45.810+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.67"
time=2025-06-06T21:48:46.060+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.69"
time=2025-06-06T21:48:46.311+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.72"
time=2025-06-06T21:48:46.561+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.75"
time=2025-06-06T21:48:46.811+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.78"
time=2025-06-06T21:48:47.062+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.80"
time=2025-06-06T21:48:47.312+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.82"
time=2025-06-06T21:48:47.563+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.85"
time=2025-06-06T21:48:47.813+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.88"
time=2025-06-06T21:48:48.063+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.91"
time=2025-06-06T21:48:48.314+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.93"
time=2025-06-06T21:48:48.564+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.97"
time=2025-06-06T21:48:48.814+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.98"
load_all_data: no device found for buffer type CPU for async uploads
llama_context: constructing llama_context
llama_context: n_seq_max = 2
llama_context: n_ctx = 8192
llama_context: n_ctx_per_seq = 4096
llama_context: n_batch = 1024
llama_context: n_ubatch = 512
llama_context: causal_attn = 1
llama_context: flash_attn = 0
llama_context: freq_base = 1000000.0
llama_context: freq_scale = 1
llama_context: n_ctx_per_seq (4096) < n_ctx_train (40960) -- the full capacity of the model will not be utilized
set_abort_callback: call
llama_context: CUDA_Host output buffer size = 1.17 MiB
create_memory: n_ctx = 8192 (padded)
llama_kv_cache_unified: kv_size = 8192, type_k = 'f16', type_v = 'f16', n_layer = 48, can_shift = 1, padding = 32
llama_kv_cache_unified: layer 0: dev = CUDA0
llama_kv_cache_unified: layer 1: dev = CUDA0
llama_kv_cache_unified: layer 2: dev = CUDA0
llama_kv_cache_unified: layer 3: dev = CUDA0
llama_kv_cache_unified: layer 4: dev = CUDA0
llama_kv_cache_unified: layer 5: dev = CUDA0
llama_kv_cache_unified: layer 6: dev = CUDA0
llama_kv_cache_unified: layer 7: dev = CUDA0
llama_kv_cache_unified: layer 8: dev = CUDA0
llama_kv_cache_unified: layer 9: dev = CUDA0
llama_kv_cache_unified: layer 10: dev = CUDA0
llama_kv_cache_unified: layer 11: dev = CUDA0
llama_kv_cache_unified: layer 12: dev = CUDA0
llama_kv_cache_unified: layer 13: dev = CUDA0
llama_kv_cache_unified: layer 14: dev = CUDA0
llama_kv_cache_unified: layer 15: dev = CUDA0
llama_kv_cache_unified: layer 16: dev = CUDA0
llama_kv_cache_unified: layer 17: dev = CUDA0
llama_kv_cache_unified: layer 18: dev = CUDA0
llama_kv_cache_unified: layer 19: dev = CUDA0
llama_kv_cache_unified: layer 20: dev = CUDA0
llama_kv_cache_unified: layer 21: dev = CUDA0
llama_kv_cache_unified: layer 22: dev = CUDA0
llama_kv_cache_unified: layer 23: dev = CUDA0
llama_kv_cache_unified: layer 24: dev = CUDA0
llama_kv_cache_unified: layer 25: dev = CUDA0
llama_kv_cache_unified: layer 26: dev = CUDA0
llama_kv_cache_unified: layer 27: dev = CUDA0
llama_kv_cache_unified: layer 28: dev = CUDA0
llama_kv_cache_unified: layer 29: dev = CUDA0
llama_kv_cache_unified: layer 30: dev = CUDA0
llama_kv_cache_unified: layer 31: dev = CUDA0
llama_kv_cache_unified: layer 32: dev = CUDA0
llama_kv_cache_unified: layer 33: dev = CUDA0
llama_kv_cache_unified: layer 34: dev = CUDA0
llama_kv_cache_unified: layer 35: dev = CUDA0
llama_kv_cache_unified: layer 36: dev = CUDA0
llama_kv_cache_unified: layer 37: dev = CUDA0
llama_kv_cache_unified: layer 38: dev = CUDA0
llama_kv_cache_unified: layer 39: dev = CUDA0
llama_kv_cache_unified: layer 40: dev = CUDA0
llama_kv_cache_unified: layer 41: dev = CUDA0
llama_kv_cache_unified: layer 42: dev = CUDA0
llama_kv_cache_unified: layer 43: dev = CUDA0
llama_kv_cache_unified: layer 44: dev = CUDA0
llama_kv_cache_unified: layer 45: dev = CUDA0
llama_kv_cache_unified: layer 46: dev = CUDA0
llama_kv_cache_unified: layer 47: dev = CUDA0
llama_kv_cache_unified: CUDA0 KV buffer size = 768.00 MiB
llama_kv_cache_unified: KV self size = 768.00 MiB, K (f16): 384.00 MiB, V (f16): 384.00 MiB
llama_context: enumerating backends
llama_context: backend_ptrs.size() = 2
llama_context: max_nodes = 65536
llama_context: worst-case: n_tokens = 512, n_seqs = 1, n_outputs = 0
llama_context: reserving graph for n_tokens = 512, n_seqs = 1
llama_context: reserving graph for n_tokens = 1, n_seqs = 1
llama_context: reserving graph for n_tokens = 512, n_seqs = 1
llama_context: CUDA0 compute buffer size = 552.00 MiB
llama_context: CUDA_Host compute buffer size = 20.01 MiB
llama_context: graph nodes = 3126
llama_context: graph splits = 2
time=2025-06-06T21:48:49.065+08:00 level=INFO source=server.go:630 msg="llama runner started in 25.53 seconds"
time=2025-06-06T21:48:49.065+08:00 level=DEBUG source=sched.go:495 msg="finished setting up" runner.name=registry.ollama.ai/library/qwen3:30b runner.inference=cuda runner.devices=1 runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac runner.num_ctx=8192
time=2025-06-06T21:48:49.088+08:00 level=DEBUG source=prompt.go:66 msg="truncating input messages which exceed context length" truncated=2
time=2025-06-06T21:48:49.088+08:00 level=DEBUG source=server.go:729 msg="completion request" images=0 prompt=104 format=""
time=2025-06-06T21:48:49.090+08:00 level=DEBUG source=cache.go:104 msg="loading cache slot" id=0 cache=0 prompt=30 used=0 remaining=30
[GIN] 2025/06/06 - 21:48:51 | 200 | 28.6746475s | 127.0.0.1 | POST "/v1/chat/completions"
time=2025-06-06T21:48:51.995+08:00 level=DEBUG source=sched.go:503 msg="context for request finished"
time=2025-06-06T21:48:51.996+08:00 level=DEBUG source=sched.go:343 msg="runner with non-zero duration has gone idle, adding timer" runner.name=registry.ollama.ai/library/qwen3:30b runner.inference=cuda runner.devices=1 runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac runner.num_ctx=8192 duration=5m0s
time=2025-06-06T21:48:51.996+08:00 level=DEBUG source=sched.go:361 msg="after processing request finished event" runner.name=registry.ollama.ai/library/qwen3:30b runner.inference=cuda runner.devices=1 runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac runner.num_ctx=8192 refCount=0
time=2025-06-06T21:53:52.004+08:00 level=DEBUG source=sched.go:345 msg="timer expired, expiring to unload" runner.name=registry.ollama.ai/library/qwen3:30b runner.inference=cuda runner.devices=1 runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac runner.num_ctx=8192
time=2025-06-06T21:53:52.004+08:00 level=DEBUG source=sched.go:364 msg="runner expired event received" runner.name=registry.ollama.ai/library/qwen3:30b runner.inference=cuda runner.devices=1 runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac runner.num_ctx=8192
time=2025-06-06T21:53:52.005+08:00 level=DEBUG source=sched.go:379 msg="got lock to unload expired event" runner.name=registry.ollama.ai/library/qwen3:30b runner.inference=cuda runner.devices=1 runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac runner.num_ctx=8192
time=2025-06-06T21:53:52.005+08:00 level=DEBUG source=sched.go:402 msg="starting background wait for VRAM recovery" runner.name=registry.ollama.ai/library/qwen3:30b runner.inference=cuda runner.devices=1 runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac runner.num_ctx=8192
time=2025-06-06T21:53:52.005+08:00 level=DEBUG source=gpu.go:391 msg="updating system memory data" before.total="31.6 GiB" before.free="15.0 GiB" before.free_swap="22.9 GiB" now.total="31.6 GiB" now.free="16.8 GiB" now.free_swap="3.0 GiB"
time=2025-06-06T21:53:52.021+08:00 level=DEBUG source=gpu.go:441 msg="updating cuda memory data" gpu=GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f name="NVIDIA GeForce RTX 5090 D" overhead="0 B" before.total="31.8 GiB" before.free="29.7 GiB" now.total="31.8 GiB" now.free="10.6 GiB" now.used="21.3 GiB"
releasing nvml library
time=2025-06-06T21:53:52.039+08:00 level=DEBUG source=server.go:1023 msg="stopping llama server" pid=27056
time=2025-06-06T21:53:52.039+08:00 level=DEBUG source=server.go:1029 msg="waiting for llama server to exit" pid=27056
time=2025-06-06T21:53:52.164+08:00 level=DEBUG source=server.go:1033 msg="llama server stopped" pid=27056
time=2025-06-06T21:53:52.164+08:00 level=DEBUG source=sched.go:407 msg="runner terminated and removed from list, blocking for VRAM recovery" runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac
time=2025-06-06T21:53:52.272+08:00 level=DEBUG source=gpu.go:391 msg="updating system memory data" before.total="31.6 GiB" before.free="16.8 GiB" before.free_swap="3.0 GiB" now.total="31.6 GiB" now.free="17.6 GiB" now.free_swap="23.5 GiB"
time=2025-06-06T21:53:52.285+08:00 level=DEBUG source=gpu.go:441 msg="updating cuda memory data" gpu=GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f name="NVIDIA GeForce RTX 5090 D" overhead="0 B" before.total="31.8 GiB" before.free="10.6 GiB" now.total="31.8 GiB" now.free="29.6 GiB" now.used="2.2 GiB"
releasing nvml library
time=2025-06-06T21:53:52.286+08:00 level=DEBUG source=sched.go:700 msg="gpu VRAM free memory converged after 0.28 seconds" runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac
time=2025-06-06T21:53:52.286+08:00 level=DEBUG source=sched.go:410 msg="sending an unloaded event" runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac
time=2025-06-06T21:53:52.286+08:00 level=DEBUG source=sched.go:312 msg="ignoring unload event with no pending requests"

<!-- gh-comment-id:2949354865 --> @Harry-Up commented on GitHub (Jun 6, 2025): > > ​​Suspected Root Cause​​: Ollama's template engine ​​may not inject system prompts correctly​​ for Qwen models. > > More likely is that the context is too small. [Server logs](https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md#how-to-troubleshoot-issues) with `OLLAMA_DEBUG=1` will aid in debugging. Yeah, I update Ollama and try it again with `OLLAMA_DEBUG=1`. The problem keeps as well. Here is the server log. time=2025-06-06T21:47:24.165+08:00 level=INFO source=routes.go:1234 msg="server config" env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: HTTPS_PROXY: HTTP_PROXY: NO_PROXY: OLLAMA_CONTEXT_LENGTH:4096 OLLAMA_DEBUG:DEBUG OLLAMA_FLASH_ATTENTION:false OLLAMA_GPU_OVERHEAD:0 OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE:5m0s OLLAMA_KV_CACHE_TYPE: OLLAMA_LLM_LIBRARY: OLLAMA_LOAD_TIMEOUT:5m0s OLLAMA_MAX_LOADED_MODELS:0 OLLAMA_MAX_QUEUE:512 OLLAMA_MODELS:C:\\Users\\weihb\\.ollama\\models 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:]" time=2025-06-06T21:47:24.169+08:00 level=INFO source=images.go:479 msg="total blobs: 13" time=2025-06-06T21:47:24.170+08:00 level=INFO source=images.go:486 msg="total unused blobs removed: 0" time=2025-06-06T21:47:24.170+08:00 level=INFO source=routes.go:1287 msg="Listening on [::]:11434 (version 0.9.0)" time=2025-06-06T21:47:24.170+08:00 level=DEBUG source=sched.go:108 msg="starting llm scheduler" time=2025-06-06T21:47:24.170+08:00 level=INFO source=gpu.go:217 msg="looking for compatible GPUs" time=2025-06-06T21:47:24.170+08:00 level=INFO source=gpu_windows.go:167 msg=packages count=1 time=2025-06-06T21:47:24.170+08:00 level=INFO source=gpu_windows.go:214 msg="" package=0 cores=8 efficiency=0 threads=16 time=2025-06-06T21:47:24.170+08:00 level=DEBUG source=gpu.go:98 msg="searching for GPU discovery libraries for NVIDIA" time=2025-06-06T21:47:24.170+08:00 level=DEBUG source=gpu.go:501 msg="Searching for GPU library" name=nvml.dll time=2025-06-06T21:47:24.170+08:00 level=DEBUG source=gpu.go:525 msg="gpu library search" globs="[C:\\Users\\weihb\\AppData\\Local\\Programs\\Ollama\\lib\\ollama\\nvml.dll C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v12.8\\bin\\nvml.dll C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v12.8\\libnvvp\\nvml.dll C:\\WINDOWS\\system32\\nvml.dll C:\\WINDOWS\\nvml.dll C:\\WINDOWS\\System32\\Wbem\\nvml.dll C:\\WINDOWS\\System32\\WindowsPowerShell\\v1.0\\nvml.dll C:\\WINDOWS\\System32\\OpenSSH\\nvml.dll C:\\Program Files (x86)\\Windows Kits\\10\\Windows Performance Toolkit\\nvml.dll C:\\Program Files\\NVIDIA Corporation\\NVIDIA App\\NvDLISR\\nvml.dll C:\\Program Files (x86)\\NVIDIA Corporation\\PhysX\\Common\\nvml.dll C:\\Program Files\\Git\\cmd\\nvml.dll C:\\Program Files\\Docker\\Docker\\resources\\bin\\nvml.dll C:\\Program Files\\Microsoft Visual Studio\\2022\\Community\\VC\\Tools\\MSVC\\14.43.34808\\bin\\Hostx64\\x64\\nvml.dll C:\\Program Files (x86)\\Windows Kits\\10\\bin\\10.0.22621.0\\x64\\nvml.dll C:\\MinGW\\bin\\nvml.dll C:\\Program Files\\NVIDIA Corporation\\Nsight Compute 2022.3.0\\nvml.dll C:\\Program Files\\dotnet\\nvml.dll C:\\Users\\weihb\\AppData\\Local\\Microsoft\\WindowsApps\\nvml.dll C:\\Users\\weihb\\AppData\\Local\\Programs\\Microsoft VS Code\\bin\\nvml.dll C:\\Users\\weihb\\AppData\\Local\\Programs\\Ollama\\nvml.dll c:\\Windows\\System32\\nvml.dll]" time=2025-06-06T21:47:24.170+08:00 level=DEBUG source=gpu.go:529 msg="skipping PhysX cuda library path" path="C:\\Program Files (x86)\\NVIDIA Corporation\\PhysX\\Common\\nvml.dll" time=2025-06-06T21:47:24.171+08:00 level=DEBUG source=gpu.go:558 msg="discovered GPU libraries" paths="[C:\\WINDOWS\\system32\\nvml.dll c:\\Windows\\System32\\nvml.dll]" time=2025-06-06T21:47:24.180+08:00 level=DEBUG source=gpu.go:111 msg="nvidia-ml loaded" library=C:\WINDOWS\system32\nvml.dll time=2025-06-06T21:47:24.180+08:00 level=DEBUG source=gpu.go:501 msg="Searching for GPU library" name=nvcuda.dll time=2025-06-06T21:47:24.180+08:00 level=DEBUG source=gpu.go:525 msg="gpu library search" globs="[C:\\Users\\weihb\\AppData\\Local\\Programs\\Ollama\\lib\\ollama\\nvcuda.dll C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v12.8\\bin\\nvcuda.dll C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v12.8\\libnvvp\\nvcuda.dll C:\\WINDOWS\\system32\\nvcuda.dll C:\\WINDOWS\\nvcuda.dll C:\\WINDOWS\\System32\\Wbem\\nvcuda.dll C:\\WINDOWS\\System32\\WindowsPowerShell\\v1.0\\nvcuda.dll C:\\WINDOWS\\System32\\OpenSSH\\nvcuda.dll C:\\Program Files (x86)\\Windows Kits\\10\\Windows Performance Toolkit\\nvcuda.dll C:\\Program Files\\NVIDIA Corporation\\NVIDIA App\\NvDLISR\\nvcuda.dll C:\\Program Files (x86)\\NVIDIA Corporation\\PhysX\\Common\\nvcuda.dll C:\\Program Files\\Git\\cmd\\nvcuda.dll C:\\Program Files\\Docker\\Docker\\resources\\bin\\nvcuda.dll C:\\Program Files\\Microsoft Visual Studio\\2022\\Community\\VC\\Tools\\MSVC\\14.43.34808\\bin\\Hostx64\\x64\\nvcuda.dll C:\\Program Files (x86)\\Windows Kits\\10\\bin\\10.0.22621.0\\x64\\nvcuda.dll C:\\MinGW\\bin\\nvcuda.dll C:\\Program Files\\NVIDIA Corporation\\Nsight Compute 2022.3.0\\nvcuda.dll C:\\Program Files\\dotnet\\nvcuda.dll C:\\Users\\weihb\\AppData\\Local\\Microsoft\\WindowsApps\\nvcuda.dll C:\\Users\\weihb\\AppData\\Local\\Programs\\Microsoft VS Code\\bin\\nvcuda.dll C:\\Users\\weihb\\AppData\\Local\\Programs\\Ollama\\nvcuda.dll c:\\windows\\system*\\nvcuda.dll]" time=2025-06-06T21:47:24.180+08:00 level=DEBUG source=gpu.go:529 msg="skipping PhysX cuda library path" path="C:\\Program Files (x86)\\NVIDIA Corporation\\PhysX\\Common\\nvcuda.dll" time=2025-06-06T21:47:24.180+08:00 level=DEBUG source=gpu.go:558 msg="discovered GPU libraries" paths=[C:\WINDOWS\system32\nvcuda.dll] initializing C:\WINDOWS\system32\nvcuda.dll dlsym: cuInit - 00007FFE357E5F80 dlsym: cuDriverGetVersion - 00007FFE357E6020 dlsym: cuDeviceGetCount - 00007FFE357E6816 dlsym: cuDeviceGet - 00007FFE357E6810 dlsym: cuDeviceGetAttribute - 00007FFE357E6170 dlsym: cuDeviceGetUuid - 00007FFE357E6822 dlsym: cuDeviceGetName - 00007FFE357E681C dlsym: cuCtxCreate_v3 - 00007FFE357E6894 dlsym: cuMemGetInfo_v2 - 00007FFE357E6996 dlsym: cuCtxDestroy - 00007FFE357E68A6 calling cuInit calling cuDriverGetVersion raw version 0x2f30 CUDA driver version: 12.8 calling cuDeviceGetCount device count 1 time=2025-06-06T21:47:24.188+08:00 level=DEBUG source=gpu.go:125 msg="detected GPUs" count=1 library=C:\WINDOWS\system32\nvcuda.dll [GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f] CUDA totalMem 32606mb [GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f] CUDA freeMem 30843mb [GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f] Compute Capability 12.0 time=2025-06-06T21:47:24.283+08:00 level=DEBUG source=amd_hip_windows.go:88 msg=hipDriverGetVersion version=60342560 time=2025-06-06T21:47:24.283+08:00 level=INFO source=amd_hip_windows.go:103 msg="AMD ROCm reports no devices found" time=2025-06-06T21:47:24.283+08:00 level=INFO source=amd_windows.go:49 msg="no compatible amdgpu devices detected" releasing cuda driver library releasing nvml library time=2025-06-06T21:47:24.283+08:00 level=INFO source=types.go:130 msg="inference compute" id=GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f library=cuda variant=v12 compute=12.0 driver=12.8 name="NVIDIA GeForce RTX 5090 D" total="31.8 GiB" available="30.1 GiB" [GIN] 2025/06/06 - 21:47:31 | 200 | 582.8µs | 127.0.0.1 | GET "/api/version" [GIN] 2025/06/06 - 21:47:35 | 200 | 0s | 127.0.0.1 | HEAD "/" [GIN] 2025/06/06 - 21:47:35 | 200 | 25.7005ms | 127.0.0.1 | GET "/api/tags" time=2025-06-06T21:48:23.357+08:00 level=DEBUG source=ggml.go:155 msg="key not found" key=general.alignment default=32 time=2025-06-06T21:48:23.359+08:00 level=DEBUG source=gpu.go:391 msg="updating system memory data" before.total="31.6 GiB" before.free="15.5 GiB" before.free_swap="24.0 GiB" now.total="31.6 GiB" now.free="15.0 GiB" now.free_swap="22.9 GiB" time=2025-06-06T21:48:23.375+08:00 level=DEBUG source=gpu.go:441 msg="updating cuda memory data" gpu=GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f name="NVIDIA GeForce RTX 5090 D" overhead="0 B" before.total="31.8 GiB" before.free="30.1 GiB" now.total="31.8 GiB" now.free="29.7 GiB" now.used="2.2 GiB" releasing nvml library time=2025-06-06T21:48:23.376+08:00 level=DEBUG source=sched.go:185 msg="updating default concurrency" OLLAMA_MAX_LOADED_MODELS=3 gpu_count=1 time=2025-06-06T21:48:23.383+08:00 level=DEBUG source=ggml.go:155 msg="key not found" key=general.alignment default=32 time=2025-06-06T21:48:23.391+08:00 level=DEBUG source=ggml.go:155 msg="key not found" key=general.alignment default=32 time=2025-06-06T21:48:23.392+08:00 level=DEBUG source=sched.go:228 msg="loading first model" model=C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac time=2025-06-06T21:48:23.393+08:00 level=DEBUG source=memory.go:111 msg=evaluating library=cuda gpu_count=1 available="[29.7 GiB]" time=2025-06-06T21:48:23.393+08:00 level=DEBUG source=ggml.go:155 msg="key not found" key=qwen3moe.vision.block_count default=0 time=2025-06-06T21:48:23.393+08:00 level=INFO source=sched.go:788 msg="new model will fit in available VRAM in single GPU, loading" model=C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac gpu=GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f parallel=2 available=31845679104 required="19.8 GiB" time=2025-06-06T21:48:23.393+08:00 level=DEBUG source=gpu.go:391 msg="updating system memory data" before.total="31.6 GiB" before.free="15.0 GiB" before.free_swap="22.9 GiB" now.total="31.6 GiB" now.free="15.0 GiB" now.free_swap="22.9 GiB" time=2025-06-06T21:48:23.405+08:00 level=DEBUG source=gpu.go:441 msg="updating cuda memory data" gpu=GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f name="NVIDIA GeForce RTX 5090 D" overhead="0 B" before.total="31.8 GiB" before.free="29.7 GiB" now.total="31.8 GiB" now.free="29.7 GiB" now.used="2.2 GiB" releasing nvml library time=2025-06-06T21:48:23.406+08:00 level=INFO source=server.go:135 msg="system memory" total="31.6 GiB" free="15.0 GiB" free_swap="22.9 GiB" time=2025-06-06T21:48:23.406+08:00 level=DEBUG source=memory.go:111 msg=evaluating library=cuda gpu_count=1 available="[29.7 GiB]" time=2025-06-06T21:48:23.406+08:00 level=DEBUG source=ggml.go:155 msg="key not found" key=qwen3moe.vision.block_count default=0 time=2025-06-06T21:48:23.406+08:00 level=INFO source=server.go:168 msg=offload library=cuda layers.requested=-1 layers.model=49 layers.offload=49 layers.split="" memory.available="[29.7 GiB]" memory.gpu_overhead="0 B" memory.required.full="19.8 GiB" memory.required.partial="19.8 GiB" memory.required.kv="768.0 MiB" memory.required.allocations="[19.8 GiB]" memory.weights.total="17.2 GiB" memory.weights.repeating="16.9 GiB" memory.weights.nonrepeating="243.4 MiB" memory.graph.full="1.0 GiB" memory.graph.partial="1.0 GiB" time=2025-06-06T21:48:23.406+08:00 level=DEBUG source=server.go:284 msg="compatible gpu libraries" compatible="[cuda_v12 cuda_v11]" llama_model_loader: loaded meta data with 31 key-value pairs and 579 tensors from C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac (version GGUF V3 (latest)) llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. llama_model_loader: - kv 0: general.architecture str = qwen3moe llama_model_loader: - kv 1: general.type str = model llama_model_loader: - kv 2: general.name str = Qwen3 30B A3B llama_model_loader: - kv 3: general.basename str = Qwen3 llama_model_loader: - kv 4: general.size_label str = 30B-A3B llama_model_loader: - kv 5: general.license str = apache-2.0 llama_model_loader: - kv 6: qwen3moe.block_count u32 = 48 llama_model_loader: - kv 7: qwen3moe.context_length u32 = 40960 llama_model_loader: - kv 8: qwen3moe.embedding_length u32 = 2048 llama_model_loader: - kv 9: qwen3moe.feed_forward_length u32 = 6144 llama_model_loader: - kv 10: qwen3moe.attention.head_count u32 = 32 llama_model_loader: - kv 11: qwen3moe.attention.head_count_kv u32 = 4 llama_model_loader: - kv 12: qwen3moe.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 13: qwen3moe.attention.layer_norm_rms_epsilon f32 = 0.000001 llama_model_loader: - kv 14: qwen3moe.expert_used_count u32 = 8 llama_model_loader: - kv 15: qwen3moe.attention.key_length u32 = 128 llama_model_loader: - kv 16: qwen3moe.attention.value_length u32 = 128 llama_model_loader: - kv 17: qwen3moe.expert_count u32 = 128 llama_model_loader: - kv 18: qwen3moe.expert_feed_forward_length u32 = 768 llama_model_loader: - kv 19: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 20: tokenizer.ggml.pre str = qwen2 llama_model_loader: - kv 21: tokenizer.ggml.tokens arr[str,151936] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 22: tokenizer.ggml.token_type arr[i32,151936] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 23: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 24: tokenizer.ggml.eos_token_id u32 = 151645 llama_model_loader: - kv 25: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 26: tokenizer.ggml.bos_token_id u32 = 151643 llama_model_loader: - kv 27: tokenizer.ggml.add_bos_token bool = false llama_model_loader: - kv 28: tokenizer.chat_template str = {%- if tools %}\n {{- '<|im_start|>... llama_model_loader: - kv 29: general.quantization_version u32 = 2 llama_model_loader: - kv 30: general.file_type u32 = 15 llama_model_loader: - type f32: 241 tensors llama_model_loader: - type f16: 48 tensors llama_model_loader: - type q4_K: 265 tensors llama_model_loader: - type q6_K: 25 tensors print_info: file format = GGUF V3 (latest) print_info: file type = Q4_K - Medium print_info: file size = 17.34 GiB (4.88 BPW) init_tokenizer: initializing tokenizer for type 2 load: control token: 151659 '<|fim_prefix|>' is not marked as EOG load: control token: 151656 '<|video_pad|>' is not marked as EOG load: control token: 151655 '<|image_pad|>' is not marked as EOG load: control token: 151653 '<|vision_end|>' is not marked as EOG load: control token: 151652 '<|vision_start|>' is not marked as EOG load: control token: 151651 '<|quad_end|>' is not marked as EOG load: control token: 151649 '<|box_end|>' is not marked as EOG load: control token: 151648 '<|box_start|>' is not marked as EOG load: control token: 151646 '<|object_ref_start|>' is not marked as EOG load: control token: 151644 '<|im_start|>' is not marked as EOG load: control token: 151661 '<|fim_suffix|>' is not marked as EOG load: control token: 151647 '<|object_ref_end|>' is not marked as EOG load: control token: 151660 '<|fim_middle|>' is not marked as EOG load: control token: 151654 '<|vision_pad|>' is not marked as EOG load: control token: 151650 '<|quad_start|>' is not marked as EOG load: special tokens cache size = 26 load: token to piece cache size = 0.9311 MB print_info: arch = qwen3moe print_info: vocab_only = 1 print_info: model type = ?B print_info: model params = 30.53 B print_info: general.name = Qwen3 30B A3B print_info: n_ff_exp = 0 print_info: vocab type = BPE print_info: n_vocab = 151936 print_info: n_merges = 151387 print_info: BOS token = 151643 '<|endoftext|>' print_info: EOS token = 151645 '<|im_end|>' print_info: EOT token = 151645 '<|im_end|>' print_info: PAD token = 151643 '<|endoftext|>' print_info: LF token = 198 'Ċ' print_info: FIM PRE token = 151659 '<|fim_prefix|>' print_info: FIM SUF token = 151661 '<|fim_suffix|>' print_info: FIM MID token = 151660 '<|fim_middle|>' print_info: FIM PAD token = 151662 '<|fim_pad|>' print_info: FIM REP token = 151663 '<|repo_name|>' print_info: FIM SEP token = 151664 '<|file_sep|>' print_info: EOG token = 151643 '<|endoftext|>' print_info: EOG token = 151645 '<|im_end|>' print_info: EOG token = 151662 '<|fim_pad|>' print_info: EOG token = 151663 '<|repo_name|>' print_info: EOG token = 151664 '<|file_sep|>' print_info: max token length = 256 llama_model_load: vocab only - skipping tensors time=2025-06-06T21:48:23.529+08:00 level=DEBUG source=server.go:360 msg="adding gpu library" path=C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\cuda_v12 time=2025-06-06T21:48:23.529+08:00 level=DEBUG source=server.go:367 msg="adding gpu dependency paths" paths=[C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\cuda_v12] time=2025-06-06T21:48:23.529+08:00 level=INFO source=server.go:431 msg="starting llama server" cmd="C:\\Users\\weihb\\AppData\\Local\\Programs\\Ollama\\ollama.exe runner --model C:\\Users\\weihb\\.ollama\\models\\blobs\\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac --ctx-size 8192 --batch-size 512 --n-gpu-layers 49 --threads 8 --no-mmap --parallel 2 --port 53373" time=2025-06-06T21:48:23.529+08:00 level=DEBUG source=server.go:432 msg=subprocess CUDA_PATH="C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v12.8" OLLAMA_API_KEY=!@#$1234qaZ OLLAMA_DEBUG=1 OLLAMA_HOST=0.0.0.0:11434 OLLAMA_MAX_LOADED_MODELS=3 PATH="C:\\Users\\weihb\\AppData\\Local\\Programs\\Ollama\\lib\\ollama\\cuda_v12;C:\\Users\\weihb\\AppData\\Local\\Programs\\Ollama\\lib\\ollama\\cuda_v12;C:\\Users\\weihb\\AppData\\Local\\Programs\\Ollama\\lib\\ollama;C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v12.8\\bin;C:\\Program Files\\NVIDIA GPU Computing Toolkit\\CUDA\\v12.8\\libnvvp;C:\\WINDOWS\\system32;C:\\WINDOWS;C:\\WINDOWS\\System32\\Wbem;C:\\WINDOWS\\System32\\WindowsPowerShell\\v1.0\\;C:\\WINDOWS\\System32\\OpenSSH\\;C:\\Program Files (x86)\\Windows Kits\\10\\Windows Performance Toolkit\\;C:\\Program Files\\NVIDIA Corporation\\NVIDIA App\\NvDLISR;C:\\Program Files (x86)\\NVIDIA Corporation\\PhysX\\Common;C:\\Program Files\\Git\\cmd;C:\\Program Files\\Docker\\Docker\\resources\\bin;C:\\Program Files\\Microsoft Visual Studio\\2022\\Community\\VC\\Tools\\MSVC\\14.43.34808\\bin\\Hostx64\\x64;C:\\Program Files (x86)\\Windows Kits\\10\\bin\\10.0.22621.0\\x64;C:\\MinGW\\bin;C:\\Program Files\\NVIDIA Corporation\\Nsight Compute 2022.3.0\\;C:\\Program Files\\dotnet\\;C:\\Users\\weihb\\AppData\\Local\\Microsoft\\WindowsApps;C:\\Users\\weihb\\AppData\\Local\\Programs\\Microsoft VS Code\\bin;C:\\Users\\weihb\\AppData\\Local\\Programs\\Ollama;C:\\Users\\weihb\\AppData\\Local\\Programs\\Ollama\\lib\\ollama" OLLAMA_LIBRARY_PATH=C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama;C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\cuda_v12 CUDA_VISIBLE_DEVICES=GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f time=2025-06-06T21:48:23.532+08:00 level=INFO source=sched.go:483 msg="loaded runners" count=1 time=2025-06-06T21:48:23.532+08:00 level=INFO source=server.go:591 msg="waiting for llama runner to start responding" time=2025-06-06T21:48:23.533+08:00 level=INFO source=server.go:625 msg="waiting for server to become available" status="llm server error" time=2025-06-06T21:48:23.556+08:00 level=INFO source=runner.go:815 msg="starting go runner" time=2025-06-06T21:48:23.559+08:00 level=DEBUG source=ggml.go:94 msg="ggml backend load all from path" path=C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama load_backend: loaded CPU backend from C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\ggml-cpu-icelake.dll time=2025-06-06T21:48:23.661+08:00 level=DEBUG source=ggml.go:94 msg="ggml backend load all from path" path=C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\cuda_v12 ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no ggml_cuda_init: found 1 CUDA devices: Device 0: NVIDIA GeForce RTX 5090 D, compute capability 12.0, VMM: yes load_backend: loaded CUDA backend from C:\Users\weihb\AppData\Local\Programs\Ollama\lib\ollama\cuda_v12\ggml-cuda.dll time=2025-06-06T21:48:39.340+08:00 level=INFO source=ggml.go:104 msg=system CPU.0.SSE3=1 CPU.0.SSSE3=1 CPU.0.AVX=1 CPU.0.AVX2=1 CPU.0.F16C=1 CPU.0.FMA=1 CPU.0.BMI2=1 CPU.0.AVX512=1 CPU.0.AVX512_VBMI=1 CPU.0.AVX512_VNNI=1 CPU.0.LLAMAFILE=1 CPU.1.LLAMAFILE=1 CUDA.0.ARCHS=500,600,610,700,750,800,860,870,890,900,1200 CUDA.0.USE_GRAPHS=1 CUDA.0.PEER_MAX_BATCH_SIZE=128 compiler=cgo(clang) time=2025-06-06T21:48:39.341+08:00 level=INFO source=runner.go:874 msg="Server listening on 127.0.0.1:53373" llama_model_load_from_file_impl: using device CUDA0 (NVIDIA GeForce RTX 5090 D) - 30843 MiB free llama_model_loader: loaded meta data with 31 key-value pairs and 579 tensors from C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac (version GGUF V3 (latest)) llama_model_loader: Dumping metadata keys/values. Note: KV overrides do not apply in this output. llama_model_loader: - kv 0: general.architecture str = qwen3moe llama_model_loader: - kv 1: general.type str = model llama_model_loader: - kv 2: general.name str = Qwen3 30B A3B llama_model_loader: - kv 3: general.basename str = Qwen3 llama_model_loader: - kv 4: general.size_label str = 30B-A3B llama_model_loader: - kv 5: general.license str = apache-2.0 llama_model_loader: - kv 6: qwen3moe.block_count u32 = 48 llama_model_loader: - kv 7: qwen3moe.context_length u32 = 40960 llama_model_loader: - kv 8: qwen3moe.embedding_length u32 = 2048 llama_model_loader: - kv 9: qwen3moe.feed_forward_length u32 = 6144 llama_model_loader: - kv 10: qwen3moe.attention.head_count u32 = 32 llama_model_loader: - kv 11: qwen3moe.attention.head_count_kv u32 = 4 llama_model_loader: - kv 12: qwen3moe.rope.freq_base f32 = 1000000.000000 llama_model_loader: - kv 13: qwen3moe.attention.layer_norm_rms_epsilon f32 = 0.000001 llama_model_loader: - kv 14: qwen3moe.expert_used_count u32 = 8 llama_model_loader: - kv 15: qwen3moe.attention.key_length u32 = 128 llama_model_loader: - kv 16: qwen3moe.attention.value_length u32 = 128 llama_model_loader: - kv 17: qwen3moe.expert_count u32 = 128 llama_model_loader: - kv 18: qwen3moe.expert_feed_forward_length u32 = 768 llama_model_loader: - kv 19: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 20: tokenizer.ggml.pre str = qwen2 llama_model_loader: - kv 21: tokenizer.ggml.tokens arr[str,151936] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 22: tokenizer.ggml.token_type arr[i32,151936] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 23: tokenizer.ggml.merges arr[str,151387] = ["Ġ Ġ", "ĠĠ ĠĠ", "i n", "Ġ t",... llama_model_loader: - kv 24: tokenizer.ggml.eos_token_id u32 = 151645 llama_model_loader: - kv 25: tokenizer.ggml.padding_token_id u32 = 151643 llama_model_loader: - kv 26: tokenizer.ggml.bos_token_id u32 = 151643 llama_model_loader: - kv 27: tokenizer.ggml.add_bos_token bool = false llama_model_loader: - kv 28: tokenizer.chat_template str = {%- if tools %}\n {{- '<|im_start|>... llama_model_loader: - kv 29: general.quantization_version u32 = 2 llama_model_loader: - kv 30: general.file_type u32 = 15 llama_model_loader: - type f32: 241 tensors llama_model_loader: - type f16: 48 tensors llama_model_loader: - type q4_K: 265 tensors llama_model_loader: - type q6_K: 25 tensors print_info: file format = GGUF V3 (latest) print_info: file type = Q4_K - Medium print_info: file size = 17.34 GiB (4.88 BPW) init_tokenizer: initializing tokenizer for type 2 load: control token: 151659 '<|fim_prefix|>' is not marked as EOG load: control token: 151656 '<|video_pad|>' is not marked as EOG load: control token: 151655 '<|image_pad|>' is not marked as EOG load: control token: 151653 '<|vision_end|>' is not marked as EOG load: control token: 151652 '<|vision_start|>' is not marked as EOG load: control token: 151651 '<|quad_end|>' is not marked as EOG load: control token: 151649 '<|box_end|>' is not marked as EOG load: control token: 151648 '<|box_start|>' is not marked as EOG load: control token: 151646 '<|object_ref_start|>' is not marked as EOG load: control token: 151644 '<|im_start|>' is not marked as EOG load: control token: 151661 '<|fim_suffix|>' is not marked as EOG load: control token: 151647 '<|object_ref_end|>' is not marked as EOG load: control token: 151660 '<|fim_middle|>' is not marked as EOG load: control token: 151654 '<|vision_pad|>' is not marked as EOG load: control token: 151650 '<|quad_start|>' is not marked as EOG load: special tokens cache size = 26 time=2025-06-06T21:48:39.553+08:00 level=INFO source=server.go:625 msg="waiting for server to become available" status="llm server loading model" load: token to piece cache size = 0.9311 MB print_info: arch = qwen3moe print_info: vocab_only = 0 print_info: n_ctx_train = 40960 print_info: n_embd = 2048 print_info: n_layer = 48 print_info: n_head = 32 print_info: n_head_kv = 4 print_info: n_rot = 128 print_info: n_swa = 0 print_info: n_swa_pattern = 1 print_info: n_embd_head_k = 128 print_info: n_embd_head_v = 128 print_info: n_gqa = 8 print_info: n_embd_k_gqa = 512 print_info: n_embd_v_gqa = 512 print_info: f_norm_eps = 0.0e+00 print_info: f_norm_rms_eps = 1.0e-06 print_info: f_clamp_kqv = 0.0e+00 print_info: f_max_alibi_bias = 0.0e+00 print_info: f_logit_scale = 0.0e+00 print_info: f_attn_scale = 0.0e+00 print_info: n_ff = 6144 print_info: n_expert = 128 print_info: n_expert_used = 8 print_info: causal attn = 1 print_info: pooling type = 0 print_info: rope type = 2 print_info: rope scaling = linear print_info: freq_base_train = 1000000.0 print_info: freq_scale_train = 1 print_info: n_ctx_orig_yarn = 40960 print_info: rope_finetuned = unknown print_info: ssm_d_conv = 0 print_info: ssm_d_inner = 0 print_info: ssm_d_state = 0 print_info: ssm_dt_rank = 0 print_info: ssm_dt_b_c_rms = 0 print_info: model type = 30B.A3B print_info: model params = 30.53 B print_info: general.name = Qwen3 30B A3B print_info: n_ff_exp = 768 print_info: vocab type = BPE print_info: n_vocab = 151936 print_info: n_merges = 151387 print_info: BOS token = 151643 '<|endoftext|>' print_info: EOS token = 151645 '<|im_end|>' print_info: EOT token = 151645 '<|im_end|>' print_info: PAD token = 151643 '<|endoftext|>' print_info: LF token = 198 'Ċ' print_info: FIM PRE token = 151659 '<|fim_prefix|>' print_info: FIM SUF token = 151661 '<|fim_suffix|>' print_info: FIM MID token = 151660 '<|fim_middle|>' print_info: FIM PAD token = 151662 '<|fim_pad|>' print_info: FIM REP token = 151663 '<|repo_name|>' print_info: FIM SEP token = 151664 '<|file_sep|>' print_info: EOG token = 151643 '<|endoftext|>' print_info: EOG token = 151645 '<|im_end|>' print_info: EOG token = 151662 '<|fim_pad|>' print_info: EOG token = 151663 '<|repo_name|>' print_info: EOG token = 151664 '<|file_sep|>' print_info: max token length = 256 load_tensors: loading model tensors, this can take a while... (mmap = false) load_tensors: layer 0 assigned to device CUDA0, is_swa = 0 load_tensors: layer 1 assigned to device CUDA0, is_swa = 0 load_tensors: layer 2 assigned to device CUDA0, is_swa = 0 load_tensors: layer 3 assigned to device CUDA0, is_swa = 0 load_tensors: layer 4 assigned to device CUDA0, is_swa = 0 load_tensors: layer 5 assigned to device CUDA0, is_swa = 0 load_tensors: layer 6 assigned to device CUDA0, is_swa = 0 load_tensors: layer 7 assigned to device CUDA0, is_swa = 0 load_tensors: layer 8 assigned to device CUDA0, is_swa = 0 load_tensors: layer 9 assigned to device CUDA0, is_swa = 0 load_tensors: layer 10 assigned to device CUDA0, is_swa = 0 load_tensors: layer 11 assigned to device CUDA0, is_swa = 0 load_tensors: layer 12 assigned to device CUDA0, is_swa = 0 load_tensors: layer 13 assigned to device CUDA0, is_swa = 0 load_tensors: layer 14 assigned to device CUDA0, is_swa = 0 load_tensors: layer 15 assigned to device CUDA0, is_swa = 0 load_tensors: layer 16 assigned to device CUDA0, is_swa = 0 load_tensors: layer 17 assigned to device CUDA0, is_swa = 0 load_tensors: layer 18 assigned to device CUDA0, is_swa = 0 load_tensors: layer 19 assigned to device CUDA0, is_swa = 0 load_tensors: layer 20 assigned to device CUDA0, is_swa = 0 load_tensors: layer 21 assigned to device CUDA0, is_swa = 0 load_tensors: layer 22 assigned to device CUDA0, is_swa = 0 load_tensors: layer 23 assigned to device CUDA0, is_swa = 0 load_tensors: layer 24 assigned to device CUDA0, is_swa = 0 load_tensors: layer 25 assigned to device CUDA0, is_swa = 0 load_tensors: layer 26 assigned to device CUDA0, is_swa = 0 load_tensors: layer 27 assigned to device CUDA0, is_swa = 0 load_tensors: layer 28 assigned to device CUDA0, is_swa = 0 load_tensors: layer 29 assigned to device CUDA0, is_swa = 0 load_tensors: layer 30 assigned to device CUDA0, is_swa = 0 load_tensors: layer 31 assigned to device CUDA0, is_swa = 0 load_tensors: layer 32 assigned to device CUDA0, is_swa = 0 load_tensors: layer 33 assigned to device CUDA0, is_swa = 0 load_tensors: layer 34 assigned to device CUDA0, is_swa = 0 load_tensors: layer 35 assigned to device CUDA0, is_swa = 0 load_tensors: layer 36 assigned to device CUDA0, is_swa = 0 load_tensors: layer 37 assigned to device CUDA0, is_swa = 0 load_tensors: layer 38 assigned to device CUDA0, is_swa = 0 load_tensors: layer 39 assigned to device CUDA0, is_swa = 0 load_tensors: layer 40 assigned to device CUDA0, is_swa = 0 load_tensors: layer 41 assigned to device CUDA0, is_swa = 0 load_tensors: layer 42 assigned to device CUDA0, is_swa = 0 load_tensors: layer 43 assigned to device CUDA0, is_swa = 0 load_tensors: layer 44 assigned to device CUDA0, is_swa = 0 load_tensors: layer 45 assigned to device CUDA0, is_swa = 0 load_tensors: layer 46 assigned to device CUDA0, is_swa = 0 load_tensors: layer 47 assigned to device CUDA0, is_swa = 0 load_tensors: layer 48 assigned to device CUDA0, is_swa = 0 load_tensors: tensor 'token_embd.weight' (q4_K) (and 0 others) cannot be used with preferred buffer type CUDA_Host, using CPU instead load_tensors: offloading 48 repeating layers to GPU load_tensors: offloading output layer to GPU load_tensors: offloaded 49/49 layers to GPU load_tensors: CUDA0 model buffer size = 17587.24 MiB load_tensors: CPU model buffer size = 166.92 MiB load_all_data: using async uploads for device CUDA0, buffer type CUDA0, backend CUDA0 time=2025-06-06T21:48:39.803+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.00" time=2025-06-06T21:48:40.053+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.03" time=2025-06-06T21:48:40.304+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.06" time=2025-06-06T21:48:40.554+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.09" time=2025-06-06T21:48:40.804+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.11" time=2025-06-06T21:48:41.054+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.14" time=2025-06-06T21:48:41.304+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.18" time=2025-06-06T21:48:41.555+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.20" time=2025-06-06T21:48:41.805+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.22" time=2025-06-06T21:48:42.056+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.25" time=2025-06-06T21:48:42.306+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.28" time=2025-06-06T21:48:42.556+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.31" time=2025-06-06T21:48:42.806+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.33" time=2025-06-06T21:48:43.056+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.35" time=2025-06-06T21:48:43.307+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.38" time=2025-06-06T21:48:43.557+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.40" time=2025-06-06T21:48:43.807+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.43" time=2025-06-06T21:48:44.057+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.46" time=2025-06-06T21:48:44.308+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.49" time=2025-06-06T21:48:44.558+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.52" time=2025-06-06T21:48:44.809+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.55" time=2025-06-06T21:48:45.059+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.58" time=2025-06-06T21:48:45.309+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.61" time=2025-06-06T21:48:45.559+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.63" time=2025-06-06T21:48:45.810+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.67" time=2025-06-06T21:48:46.060+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.69" time=2025-06-06T21:48:46.311+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.72" time=2025-06-06T21:48:46.561+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.75" time=2025-06-06T21:48:46.811+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.78" time=2025-06-06T21:48:47.062+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.80" time=2025-06-06T21:48:47.312+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.82" time=2025-06-06T21:48:47.563+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.85" time=2025-06-06T21:48:47.813+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.88" time=2025-06-06T21:48:48.063+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.91" time=2025-06-06T21:48:48.314+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.93" time=2025-06-06T21:48:48.564+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.97" time=2025-06-06T21:48:48.814+08:00 level=DEBUG source=server.go:636 msg="model load progress 0.98" load_all_data: no device found for buffer type CPU for async uploads llama_context: constructing llama_context llama_context: n_seq_max = 2 llama_context: n_ctx = 8192 llama_context: n_ctx_per_seq = 4096 llama_context: n_batch = 1024 llama_context: n_ubatch = 512 llama_context: causal_attn = 1 llama_context: flash_attn = 0 llama_context: freq_base = 1000000.0 llama_context: freq_scale = 1 llama_context: n_ctx_per_seq (4096) < n_ctx_train (40960) -- the full capacity of the model will not be utilized set_abort_callback: call llama_context: CUDA_Host output buffer size = 1.17 MiB create_memory: n_ctx = 8192 (padded) llama_kv_cache_unified: kv_size = 8192, type_k = 'f16', type_v = 'f16', n_layer = 48, can_shift = 1, padding = 32 llama_kv_cache_unified: layer 0: dev = CUDA0 llama_kv_cache_unified: layer 1: dev = CUDA0 llama_kv_cache_unified: layer 2: dev = CUDA0 llama_kv_cache_unified: layer 3: dev = CUDA0 llama_kv_cache_unified: layer 4: dev = CUDA0 llama_kv_cache_unified: layer 5: dev = CUDA0 llama_kv_cache_unified: layer 6: dev = CUDA0 llama_kv_cache_unified: layer 7: dev = CUDA0 llama_kv_cache_unified: layer 8: dev = CUDA0 llama_kv_cache_unified: layer 9: dev = CUDA0 llama_kv_cache_unified: layer 10: dev = CUDA0 llama_kv_cache_unified: layer 11: dev = CUDA0 llama_kv_cache_unified: layer 12: dev = CUDA0 llama_kv_cache_unified: layer 13: dev = CUDA0 llama_kv_cache_unified: layer 14: dev = CUDA0 llama_kv_cache_unified: layer 15: dev = CUDA0 llama_kv_cache_unified: layer 16: dev = CUDA0 llama_kv_cache_unified: layer 17: dev = CUDA0 llama_kv_cache_unified: layer 18: dev = CUDA0 llama_kv_cache_unified: layer 19: dev = CUDA0 llama_kv_cache_unified: layer 20: dev = CUDA0 llama_kv_cache_unified: layer 21: dev = CUDA0 llama_kv_cache_unified: layer 22: dev = CUDA0 llama_kv_cache_unified: layer 23: dev = CUDA0 llama_kv_cache_unified: layer 24: dev = CUDA0 llama_kv_cache_unified: layer 25: dev = CUDA0 llama_kv_cache_unified: layer 26: dev = CUDA0 llama_kv_cache_unified: layer 27: dev = CUDA0 llama_kv_cache_unified: layer 28: dev = CUDA0 llama_kv_cache_unified: layer 29: dev = CUDA0 llama_kv_cache_unified: layer 30: dev = CUDA0 llama_kv_cache_unified: layer 31: dev = CUDA0 llama_kv_cache_unified: layer 32: dev = CUDA0 llama_kv_cache_unified: layer 33: dev = CUDA0 llama_kv_cache_unified: layer 34: dev = CUDA0 llama_kv_cache_unified: layer 35: dev = CUDA0 llama_kv_cache_unified: layer 36: dev = CUDA0 llama_kv_cache_unified: layer 37: dev = CUDA0 llama_kv_cache_unified: layer 38: dev = CUDA0 llama_kv_cache_unified: layer 39: dev = CUDA0 llama_kv_cache_unified: layer 40: dev = CUDA0 llama_kv_cache_unified: layer 41: dev = CUDA0 llama_kv_cache_unified: layer 42: dev = CUDA0 llama_kv_cache_unified: layer 43: dev = CUDA0 llama_kv_cache_unified: layer 44: dev = CUDA0 llama_kv_cache_unified: layer 45: dev = CUDA0 llama_kv_cache_unified: layer 46: dev = CUDA0 llama_kv_cache_unified: layer 47: dev = CUDA0 llama_kv_cache_unified: CUDA0 KV buffer size = 768.00 MiB llama_kv_cache_unified: KV self size = 768.00 MiB, K (f16): 384.00 MiB, V (f16): 384.00 MiB llama_context: enumerating backends llama_context: backend_ptrs.size() = 2 llama_context: max_nodes = 65536 llama_context: worst-case: n_tokens = 512, n_seqs = 1, n_outputs = 0 llama_context: reserving graph for n_tokens = 512, n_seqs = 1 llama_context: reserving graph for n_tokens = 1, n_seqs = 1 llama_context: reserving graph for n_tokens = 512, n_seqs = 1 llama_context: CUDA0 compute buffer size = 552.00 MiB llama_context: CUDA_Host compute buffer size = 20.01 MiB llama_context: graph nodes = 3126 llama_context: graph splits = 2 time=2025-06-06T21:48:49.065+08:00 level=INFO source=server.go:630 msg="llama runner started in 25.53 seconds" time=2025-06-06T21:48:49.065+08:00 level=DEBUG source=sched.go:495 msg="finished setting up" runner.name=registry.ollama.ai/library/qwen3:30b runner.inference=cuda runner.devices=1 runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac runner.num_ctx=8192 time=2025-06-06T21:48:49.088+08:00 level=DEBUG source=prompt.go:66 msg="truncating input messages which exceed context length" truncated=2 time=2025-06-06T21:48:49.088+08:00 level=DEBUG source=server.go:729 msg="completion request" images=0 prompt=104 format="" time=2025-06-06T21:48:49.090+08:00 level=DEBUG source=cache.go:104 msg="loading cache slot" id=0 cache=0 prompt=30 used=0 remaining=30 [GIN] 2025/06/06 - 21:48:51 | 200 | 28.6746475s | 127.0.0.1 | POST "/v1/chat/completions" time=2025-06-06T21:48:51.995+08:00 level=DEBUG source=sched.go:503 msg="context for request finished" time=2025-06-06T21:48:51.996+08:00 level=DEBUG source=sched.go:343 msg="runner with non-zero duration has gone idle, adding timer" runner.name=registry.ollama.ai/library/qwen3:30b runner.inference=cuda runner.devices=1 runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac runner.num_ctx=8192 duration=5m0s time=2025-06-06T21:48:51.996+08:00 level=DEBUG source=sched.go:361 msg="after processing request finished event" runner.name=registry.ollama.ai/library/qwen3:30b runner.inference=cuda runner.devices=1 runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac runner.num_ctx=8192 refCount=0 time=2025-06-06T21:53:52.004+08:00 level=DEBUG source=sched.go:345 msg="timer expired, expiring to unload" runner.name=registry.ollama.ai/library/qwen3:30b runner.inference=cuda runner.devices=1 runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac runner.num_ctx=8192 time=2025-06-06T21:53:52.004+08:00 level=DEBUG source=sched.go:364 msg="runner expired event received" runner.name=registry.ollama.ai/library/qwen3:30b runner.inference=cuda runner.devices=1 runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac runner.num_ctx=8192 time=2025-06-06T21:53:52.005+08:00 level=DEBUG source=sched.go:379 msg="got lock to unload expired event" runner.name=registry.ollama.ai/library/qwen3:30b runner.inference=cuda runner.devices=1 runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac runner.num_ctx=8192 time=2025-06-06T21:53:52.005+08:00 level=DEBUG source=sched.go:402 msg="starting background wait for VRAM recovery" runner.name=registry.ollama.ai/library/qwen3:30b runner.inference=cuda runner.devices=1 runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac runner.num_ctx=8192 time=2025-06-06T21:53:52.005+08:00 level=DEBUG source=gpu.go:391 msg="updating system memory data" before.total="31.6 GiB" before.free="15.0 GiB" before.free_swap="22.9 GiB" now.total="31.6 GiB" now.free="16.8 GiB" now.free_swap="3.0 GiB" time=2025-06-06T21:53:52.021+08:00 level=DEBUG source=gpu.go:441 msg="updating cuda memory data" gpu=GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f name="NVIDIA GeForce RTX 5090 D" overhead="0 B" before.total="31.8 GiB" before.free="29.7 GiB" now.total="31.8 GiB" now.free="10.6 GiB" now.used="21.3 GiB" releasing nvml library time=2025-06-06T21:53:52.039+08:00 level=DEBUG source=server.go:1023 msg="stopping llama server" pid=27056 time=2025-06-06T21:53:52.039+08:00 level=DEBUG source=server.go:1029 msg="waiting for llama server to exit" pid=27056 time=2025-06-06T21:53:52.164+08:00 level=DEBUG source=server.go:1033 msg="llama server stopped" pid=27056 time=2025-06-06T21:53:52.164+08:00 level=DEBUG source=sched.go:407 msg="runner terminated and removed from list, blocking for VRAM recovery" runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac time=2025-06-06T21:53:52.272+08:00 level=DEBUG source=gpu.go:391 msg="updating system memory data" before.total="31.6 GiB" before.free="16.8 GiB" before.free_swap="3.0 GiB" now.total="31.6 GiB" now.free="17.6 GiB" now.free_swap="23.5 GiB" time=2025-06-06T21:53:52.285+08:00 level=DEBUG source=gpu.go:441 msg="updating cuda memory data" gpu=GPU-d2bc81d1-bd18-101d-0f21-85f4096fdc9f name="NVIDIA GeForce RTX 5090 D" overhead="0 B" before.total="31.8 GiB" before.free="10.6 GiB" now.total="31.8 GiB" now.free="29.6 GiB" now.used="2.2 GiB" releasing nvml library time=2025-06-06T21:53:52.286+08:00 level=DEBUG source=sched.go:700 msg="gpu VRAM free memory converged after 0.28 seconds" runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac time=2025-06-06T21:53:52.286+08:00 level=DEBUG source=sched.go:410 msg="sending an unloaded event" runner.size="19.8 GiB" runner.vram="19.8 GiB" runner.parallel=2 runner.pid=27056 runner.model=C:\Users\weihb\.ollama\models\blobs\sha256-e9183b5c18a0cf736578c1e3d1cbd4b7e98e3ad3be6176b68c20f156d54a07ac time=2025-06-06T21:53:52.286+08:00 level=DEBUG source=sched.go:312 msg="ignoring unload event with no pending requests"
GiteaMirror commented 2026-04-29 04:40:10 -05:00
Author
Owner
Copy Link

@rick-github commented on GitHub (Jun 6, 2025):

You have the default context length of 4096 tokens: OLLAMA_CONTEXT_LENGTH:4096.

time=2025-06-06T21:48:49.088+08:00 level=DEBUG source=prompt.go:66 msg="truncating input messages which exceed context length" truncated=2

The prompt is truncated, removing the system prompt.

Increase the context length.

<!-- gh-comment-id:2949368097 --> @rick-github commented on GitHub (Jun 6, 2025): You have the default context length of 4096 tokens: `OLLAMA_CONTEXT_LENGTH:4096`. ``` time=2025-06-06T21:48:49.088+08:00 level=DEBUG source=prompt.go:66 msg="truncating input messages which exceed context length" truncated=2 ``` The prompt is truncated, removing the system prompt. Increase the [context length](https://github.com/ollama/ollama/blob/main/docs/faq.md#how-can-i-specify-the-context-window-size).
GiteaMirror commented 2026-04-29 04:40:12 -05:00
Author
Owner
Copy Link

@Harry-Up commented on GitHub (Jun 7, 2025):

You have the default context length of 4096 tokens: OLLAMA_CONTEXT_LENGTH:4096.

time=2025-06-06T21:48:49.088+08:00 level=DEBUG source=prompt.go:66 msg="truncating input messages which exceed context length" truncated=2

The prompt is truncated, removing the system prompt.

Increase the context length.

Thank you for your patient help. By increasing the context length, this problem was fixed. That helped me a lot.

<!-- gh-comment-id:2951409864 --> @Harry-Up commented on GitHub (Jun 7, 2025): > You have the default context length of 4096 tokens: `OLLAMA_CONTEXT_LENGTH:4096`. > > ``` > time=2025-06-06T21:48:49.088+08:00 level=DEBUG source=prompt.go:66 msg="truncating input messages which exceed context length" truncated=2 > ``` > > The prompt is truncated, removing the system prompt. > > Increase the [context length](https://github.com/ollama/ollama/blob/main/docs/faq.md#how-can-i-specify-the-context-window-size). Thank you for your patient help. By increasing the context length, this problem was fixed. That helped me a lot.
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#53750
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?