github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-07-16 13:19:40 -05:00
Code Issues 16.9k Packages Projects Releases 122 Wiki Activity

[GH-ISSUE #4780] Radeon VII gfx906:sramecc-:xnack- windows support #65047

New Issue
Closed
opened 2026-05-03 19:37:54 -05:00 by GiteaMirror · 11 comments
GiteaMirror commented 2026-05-03 19:37:54 -05:00
Owner
Copy Link

Originally created by @MrSteelRat on GitHub (Jun 2, 2024).
Original GitHub issue: https://github.com/ollama/ollama/issues/4780

Originally assigned to: @dhiltgen on GitHub.

Hello, please add support for this GPU Radeon VII to the regular version, I can help with testing if necessary

Originally created by @MrSteelRat on GitHub (Jun 2, 2024). Original GitHub issue: https://github.com/ollama/ollama/issues/4780 Originally assigned to: @dhiltgen on GitHub. Hello, please add support for this GPU Radeon VII to the regular version, I can help with testing if necessary
GiteaMirror added the gpuamdfeature requestwindows labels 2026-05-03 19:37:55 -05:00
GiteaMirror commented 2026-05-03 19:38:00 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Jun 13, 2024):

I believe this GPU should be supported. Have you tried? If it didn't work, can you share the server log?

<!-- gh-comment-id:2166495078 --> @dhiltgen commented on GitHub (Jun 13, 2024): I believe this GPU should be supported. Have you tried? If it didn't work, can you share the server log?
GiteaMirror commented 2026-05-03 19:38:01 -05:00
Author
Owner
Copy Link

@MrSteelRat commented on GitHub (Jun 13, 2024):

Hello, I looked at the news about AMD GPU support and there is only the Pro version of my video card, I have a regular one without the Pro prefix

<!-- gh-comment-id:2166498743 --> @MrSteelRat commented on GitHub (Jun 13, 2024): Hello, I looked at the news about AMD GPU support and there is only the Pro version of my video card, I have a regular one without the Pro prefix
GiteaMirror commented 2026-05-03 19:38:03 -05:00
Author
Owner
Copy Link

@MrSteelRat commented on GitHub (Jun 13, 2024):

image
image

<!-- gh-comment-id:2166506555 --> @MrSteelRat commented on GitHub (Jun 13, 2024): ![image](https://github.com/ollama/ollama/assets/31157848/c11347b2-06fa-4b70-873e-2f32dc48b6a6) ![image](https://github.com/ollama/ollama/assets/31157848/fb0758a4-36e8-4366-847d-f9de7fac2dae)
GiteaMirror commented 2026-05-03 19:38:04 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Jun 18, 2024):

If the GPU isn't working, please share your server log so we can see what went wrong.

https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md

<!-- gh-comment-id:2177142231 --> @dhiltgen commented on GitHub (Jun 18, 2024): If the GPU isn't working, please share your server log so we can see what went wrong. https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md
GiteaMirror commented 2026-05-03 19:38:06 -05:00
Author
Owner
Copy Link

@fofocx commented on GitHub (Jun 26, 2024):

Same problem.
GPU: AMD Radeon VII
OLLAMA version: 0.1.46

server.log:
2024/06/26 15:17:40 routes.go:1064: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE: OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:1 OLLAMA_MAX_QUEUE:512 OLLAMA_MAX_VRAM:0 OLLAMA_MODELS:C:\Users\Administrator\.ollama\models OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:1 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://*] OLLAMA_RUNNERS_DIR:C:\Users\Administrator\AppData\Local\Programs\Ollama\ollama_runners OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
time=2024-06-26T15:17:40.405+08:00 level=INFO source=images.go:730 msg="total blobs: 0"
time=2024-06-26T15:17:40.405+08:00 level=INFO source=images.go:737 msg="total unused blobs removed: 0"
time=2024-06-26T15:17:40.407+08:00 level=INFO source=routes.go:1111 msg="Listening on [::]:11434 (version 0.1.46)"
time=2024-06-26T15:17:40.409+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11.3 rocm_v5.7]"
time=2024-06-26T15:17:40.545+08:00 level=WARN source=amd_windows.go:97 msg="amdgpu is not supported" gpu=0 gpu_type=gfx906:sramecc-:xnack- library=C:\Users\Administrator\AppData\Local\Programs\Ollama\rocm supported_types="[gfx1030 gfx1100 gfx1101 gfx1102 gfx906]"
time=2024-06-26T15:17:40.545+08:00 level=WARN source=amd_windows.go:99 msg="See https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md for HSA_OVERRIDE_GFX_VERSION usage"
time=2024-06-26T15:17:40.608+08:00 level=INFO source=types.go:98 msg="inference compute" id=0 library=cpu compute="" driver=0.0 name="" total="32.0 GiB" available="15.4 GiB"

<!-- gh-comment-id:2191000645 --> @fofocx commented on GitHub (Jun 26, 2024): Same problem. GPU: AMD Radeon VII OLLAMA version: 0.1.46 server.log: 2024/06/26 15:17:40 routes.go:1064: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE: OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:1 OLLAMA_MAX_QUEUE:512 OLLAMA_MAX_VRAM:0 OLLAMA_MODELS:C:\\Users\\Administrator\\.ollama\\models OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:1 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://*] OLLAMA_RUNNERS_DIR:C:\\Users\\Administrator\\AppData\\Local\\Programs\\Ollama\\ollama_runners OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" time=2024-06-26T15:17:40.405+08:00 level=INFO source=images.go:730 msg="total blobs: 0" time=2024-06-26T15:17:40.405+08:00 level=INFO source=images.go:737 msg="total unused blobs removed: 0" time=2024-06-26T15:17:40.407+08:00 level=INFO source=routes.go:1111 msg="Listening on [::]:11434 (version 0.1.46)" time=2024-06-26T15:17:40.409+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11.3 rocm_v5.7]" time=2024-06-26T15:17:40.545+08:00 level=WARN source=amd_windows.go:97 msg="amdgpu is not supported" gpu=0 gpu_type=gfx906:sramecc-:xnack- library=C:\Users\Administrator\AppData\Local\Programs\Ollama\rocm supported_types="[gfx1030 gfx1100 gfx1101 gfx1102 gfx906]" time=2024-06-26T15:17:40.545+08:00 level=WARN source=amd_windows.go:99 msg="See https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md for HSA_OVERRIDE_GFX_VERSION usage" time=2024-06-26T15:17:40.608+08:00 level=INFO source=types.go:98 msg="inference compute" id=0 library=cpu compute="" driver=0.0 name="" total="32.0 GiB" available="15.4 GiB"
GiteaMirror commented 2026-05-03 19:38:07 -05:00
Author
Owner
Copy Link

@fofocx commented on GitHub (Jun 26, 2024):

Ollama will throw "Error: llama runner process has terminated: exit status 0xc0000005" and crash when execute "ollama run .." if set Environment variable "HSA_OVERRIDE_GFX_VERSION=9.0.6" .

server.log:
2024/06/26 15:31:15 routes.go:1064: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION:9.0.6 OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE: OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:1 OLLAMA_MAX_QUEUE:512 OLLAMA_MAX_VRAM:0 OLLAMA_MODELS:C:\Users\Administrator\.ollama\models OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:1 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://*] OLLAMA_RUNNERS_DIR:C:\Users\Administrator\AppData\Local\Programs\Ollama\ollama_runners OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
time=2024-06-26T15:31:15.361+08:00 level=INFO source=images.go:730 msg="total blobs: 0"
time=2024-06-26T15:31:15.362+08:00 level=INFO source=images.go:737 msg="total unused blobs removed: 0"
time=2024-06-26T15:31:15.363+08:00 level=INFO source=routes.go:1111 msg="Listening on [::]:11434 (version 0.1.46)"
time=2024-06-26T15:31:15.365+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11.3 rocm_v5.7]"
time=2024-06-26T15:31:15.484+08:00 level=INFO source=amd_windows.go:65 msg="skipping rocm gfx compatibility check" HSA_OVERRIDE_GFX_VERSION=9.0.6
time=2024-06-26T15:31:16.212+08:00 level=INFO source=types.go:98 msg="inference compute" id=0 library=rocm compute=gfx906:sramecc-:xnack- driver=0.0 name="AMD Radeon VII" total="16.0 GiB" available="15.9 GiB"
[GIN] 2024/06/26 - 15:31:59 | 200 | 0s | 127.0.0.1 | HEAD "/"
[GIN] 2024/06/26 - 15:32:34 | 201 | 20.6674284s | 127.0.0.1 | POST "/api/blobs/sha256:08d6043f9b965d26fe8568c7b791d5b312eaaf09ce912d36defaca55f5cd711c"
[GIN] 2024/06/26 - 15:32:52 | 200 | 18.0231413s | 127.0.0.1 | POST "/api/create"
[GIN] 2024/06/26 - 15:33:52 | 200 | 0s | 127.0.0.1 | HEAD "/"
[GIN] 2024/06/26 - 15:33:52 | 200 | 22.6084ms | 127.0.0.1 | POST "/api/show"
time=2024-06-26T15:33:53.561+08:00 level=INFO source=memory.go:309 msg="offload to rocm" layers.requested=-1 layers.model=33 layers.offload=33 layers.split="" memory.available="[15.9 GiB]" memory.required.full="6.0 GiB" memory.required.partial="6.0 GiB" memory.required.kv="256.0 MiB" memory.required.allocations="[6.0 GiB]" memory.weights.total="4.8 GiB" memory.weights.repeating="4.4 GiB" memory.weights.nonrepeating="411.0 MiB" memory.graph.full="164.0 MiB" memory.graph.partial="677.5 MiB"
time=2024-06-26T15:33:53.579+08:00 level=INFO source=server.go:368 msg="starting llama server" cmd="C:\Users\Administrator\AppData\Local\Programs\Ollama\ollama_runners\rocm_v5.7\ollama_llama_server.exe --model C:\Users\Administrator\.ollama\models\blobs\sha256-08d6043f9b965d26fe8568c7b791d5b312eaaf09ce912d36defaca55f5cd711c --ctx-size 2048 --batch-size 512 --embedding --log-disable --n-gpu-layers 33 --parallel 1 --port 51536"
time=2024-06-26T15:33:53.646+08:00 level=INFO source=sched.go:382 msg="loaded runners" count=1
time=2024-06-26T15:33:53.647+08:00 level=INFO source=server.go:556 msg="waiting for llama runner to start responding"
time=2024-06-26T15:33:53.648+08:00 level=INFO source=server.go:594 msg="waiting for server to become available" status="llm server error"
INFO [wmain] build info | build=3171 commit="7c26775a" tid="28472" timestamp=1719387233
INFO [wmain] system info | n_threads=8 n_threads_batch=-1 system_info="AVX = 1 | AVX_VNNI = 0 | AVX2 = 0 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | AVX512_BF16 = 0 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | " tid="28472" timestamp=1719387233 total_threads=16
INFO [wmain] HTTP server listening | hostname="127.0.0.1" n_threads_http="15" port="51536" tid="28472" timestamp=1719387233
llama_model_loader: loaded meta data with 22 key-value pairs and 291 tensors from C:\Users\Administrator.ollama\models\blobs\sha256-08d6043f9b965d26fe8568c7b791d5b312eaaf09ce912d36defaca55f5cd711c (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 = llama
llama_model_loader: - kv 1: general.name str = Llama-3
llama_model_loader: - kv 2: llama.block_count u32 = 32
llama_model_loader: - kv 3: llama.context_length u32 = 8192
llama_model_loader: - kv 4: llama.embedding_length u32 = 4096
llama_model_loader: - kv 5: llama.feed_forward_length u32 = 14336
llama_model_loader: - kv 6: llama.attention.head_count u32 = 32
llama_model_loader: - kv 7: llama.attention.head_count_kv u32 = 8
llama_model_loader: - kv 8: llama.rope.freq_base f32 = 500000.000000
llama_model_loader: - kv 9: llama.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 10: general.file_type u32 = 17
llama_model_loader: - kv 11: llama.vocab_size u32 = 128256
llama_model_loader: - kv 12: llama.rope.dimension_count u32 = 128
llama_model_loader: - kv 13: tokenizer.ggml.model str = gpt2
llama_model_loader: - kv 14: tokenizer.ggml.pre str = llama-bpe
llama_model_loader: - kv 15: tokenizer.ggml.tokens arr[str,128256] = ["!", """, "#", "$", "%", "&", "'", ...
llama_model_loader: - kv 16: tokenizer.ggml.token_type arr[i32,128256] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ...
llama_model_loader: - kv 17: tokenizer.ggml.merges arr[str,280147] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "...
llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 128000
llama_model_loader: - kv 19: tokenizer.ggml.eos_token_id u32 = 128001
llama_model_loader: - kv 20: tokenizer.chat_template str = {% set loop_messages = messages %}{% ...
llama_model_loader: - kv 21: general.quantization_version u32 = 2
llama_model_loader: - type f32: 65 tensors
llama_model_loader: - type q5_K: 193 tensors
llama_model_loader: - type q6_K: 33 tensors
time=2024-06-26T15:33:53.905+08:00 level=INFO source=server.go:594 msg="waiting for server to become available" status="llm server loading model"
llm_load_vocab: special tokens cache size = 256
llm_load_vocab: token to piece cache size = 0.8000 MB
llm_load_print_meta: format = GGUF V3 (latest)
llm_load_print_meta: arch = llama
llm_load_print_meta: vocab type = BPE
llm_load_print_meta: n_vocab = 128256
llm_load_print_meta: n_merges = 280147
llm_load_print_meta: n_ctx_train = 8192
llm_load_print_meta: n_embd = 4096
llm_load_print_meta: n_head = 32
llm_load_print_meta: n_head_kv = 8
llm_load_print_meta: n_layer = 32
llm_load_print_meta: n_rot = 128
llm_load_print_meta: n_embd_head_k = 128
llm_load_print_meta: n_embd_head_v = 128
llm_load_print_meta: n_gqa = 4
llm_load_print_meta: n_embd_k_gqa = 1024
llm_load_print_meta: n_embd_v_gqa = 1024
llm_load_print_meta: f_norm_eps = 0.0e+00
llm_load_print_meta: f_norm_rms_eps = 1.0e-05
llm_load_print_meta: f_clamp_kqv = 0.0e+00
llm_load_print_meta: f_max_alibi_bias = 0.0e+00
llm_load_print_meta: f_logit_scale = 0.0e+00
llm_load_print_meta: n_ff = 14336
llm_load_print_meta: n_expert = 0
llm_load_print_meta: n_expert_used = 0
llm_load_print_meta: causal attn = 1
llm_load_print_meta: pooling type = 0
llm_load_print_meta: rope type = 0
llm_load_print_meta: rope scaling = linear
llm_load_print_meta: freq_base_train = 500000.0
llm_load_print_meta: freq_scale_train = 1
llm_load_print_meta: n_ctx_orig_yarn = 8192
llm_load_print_meta: rope_finetuned = unknown
llm_load_print_meta: ssm_d_conv = 0
llm_load_print_meta: ssm_d_inner = 0
llm_load_print_meta: ssm_d_state = 0
llm_load_print_meta: ssm_dt_rank = 0
llm_load_print_meta: model type = 8B
llm_load_print_meta: model ftype = Q5_K - Medium
llm_load_print_meta: model params = 8.03 B
llm_load_print_meta: model size = 5.33 GiB (5.70 BPW)
llm_load_print_meta: general.name = Llama-3
llm_load_print_meta: BOS token = 128000 '<|begin_of_text|>'
llm_load_print_meta: EOS token = 128001 '<|end_of_text|>'
llm_load_print_meta: LF token = 128 'Ä'
llm_load_print_meta: EOT token = 128009 '<|eot_id|>'
ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
ggml_cuda_init: CUDA_USE_TENSOR_CORES: yes
ggml_cuda_init: found 1 ROCm devices:
Device 0: AMD Radeon VII, compute capability 9.0, VMM: no
llm_load_tensors: ggml ctx size = 0.30 MiB
llm_load_tensors: offloading 32 repeating layers to GPU
llm_load_tensors: offloading non-repeating layers to GPU
llm_load_tensors: offloaded 33/33 layers to GPU
llm_load_tensors: ROCm0 buffer size = 5115.49 MiB
llm_load_tensors: CPU buffer size = 344.44 MiB
llama_new_context_with_model: n_ctx = 2048
llama_new_context_with_model: n_batch = 512
llama_new_context_with_model: n_ubatch = 512
llama_new_context_with_model: flash_attn = 0
llama_new_context_with_model: freq_base = 500000.0
llama_new_context_with_model: freq_scale = 1
llama_kv_cache_init: ROCm0 KV buffer size = 256.00 MiB
llama_new_context_with_model: KV self size = 256.00 MiB, K (f16): 128.00 MiB, V (f16): 128.00 MiB
llama_new_context_with_model: ROCm_Host output buffer size = 0.50 MiB
llama_new_context_with_model: ROCm0 compute buffer size = 258.50 MiB
llama_new_context_with_model: ROCm_Host compute buffer size = 12.01 MiB
llama_new_context_with_model: graph nodes = 1030
llama_new_context_with_model: graph splits = 2
time=2024-06-26T15:34:01.844+08:00 level=INFO source=server.go:594 msg="waiting for server to become available" status="llm server error"
time=2024-06-26T15:34:02.350+08:00 level=ERROR source=sched.go:388 msg="error loading llama server" error="llama runner process has terminated: exit status 0xc0000005 "
[GIN] 2024/06/26 - 15:34:02 | 500 | 9.5332306s | 127.0.0.1 | POST "/api/chat"

<!-- gh-comment-id:2191052721 --> @fofocx commented on GitHub (Jun 26, 2024): Ollama will throw "Error: llama runner process has terminated: exit status 0xc0000005" and crash when execute "ollama run .." if set Environment variable "HSA_OVERRIDE_GFX_VERSION=9.0.6" . server.log: 2024/06/26 15:31:15 routes.go:1064: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION:9.0.6 OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE: OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:1 OLLAMA_MAX_QUEUE:512 OLLAMA_MAX_VRAM:0 OLLAMA_MODELS:C:\\Users\\Administrator\\.ollama\\models OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:1 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://*] OLLAMA_RUNNERS_DIR:C:\\Users\\Administrator\\AppData\\Local\\Programs\\Ollama\\ollama_runners OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" time=2024-06-26T15:31:15.361+08:00 level=INFO source=images.go:730 msg="total blobs: 0" time=2024-06-26T15:31:15.362+08:00 level=INFO source=images.go:737 msg="total unused blobs removed: 0" time=2024-06-26T15:31:15.363+08:00 level=INFO source=routes.go:1111 msg="Listening on [::]:11434 (version 0.1.46)" time=2024-06-26T15:31:15.365+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11.3 rocm_v5.7]" time=2024-06-26T15:31:15.484+08:00 level=INFO source=amd_windows.go:65 msg="skipping rocm gfx compatibility check" HSA_OVERRIDE_GFX_VERSION=9.0.6 time=2024-06-26T15:31:16.212+08:00 level=INFO source=types.go:98 msg="inference compute" id=0 library=rocm compute=gfx906:sramecc-:xnack- driver=0.0 name="AMD Radeon VII" total="16.0 GiB" available="15.9 GiB" [GIN] 2024/06/26 - 15:31:59 | 200 | 0s | 127.0.0.1 | HEAD "/" [GIN] 2024/06/26 - 15:32:34 | 201 | 20.6674284s | 127.0.0.1 | POST "/api/blobs/sha256:08d6043f9b965d26fe8568c7b791d5b312eaaf09ce912d36defaca55f5cd711c" [GIN] 2024/06/26 - 15:32:52 | 200 | 18.0231413s | 127.0.0.1 | POST "/api/create" [GIN] 2024/06/26 - 15:33:52 | 200 | 0s | 127.0.0.1 | HEAD "/" [GIN] 2024/06/26 - 15:33:52 | 200 | 22.6084ms | 127.0.0.1 | POST "/api/show" time=2024-06-26T15:33:53.561+08:00 level=INFO source=memory.go:309 msg="offload to rocm" layers.requested=-1 layers.model=33 layers.offload=33 layers.split="" memory.available="[15.9 GiB]" memory.required.full="6.0 GiB" memory.required.partial="6.0 GiB" memory.required.kv="256.0 MiB" memory.required.allocations="[6.0 GiB]" memory.weights.total="4.8 GiB" memory.weights.repeating="4.4 GiB" memory.weights.nonrepeating="411.0 MiB" memory.graph.full="164.0 MiB" memory.graph.partial="677.5 MiB" time=2024-06-26T15:33:53.579+08:00 level=INFO source=server.go:368 msg="starting llama server" cmd="C:\\Users\\Administrator\\AppData\\Local\\Programs\\Ollama\\ollama_runners\\rocm_v5.7\\ollama_llama_server.exe --model C:\\Users\\Administrator\\.ollama\\models\\blobs\\sha256-08d6043f9b965d26fe8568c7b791d5b312eaaf09ce912d36defaca55f5cd711c --ctx-size 2048 --batch-size 512 --embedding --log-disable --n-gpu-layers 33 --parallel 1 --port 51536" time=2024-06-26T15:33:53.646+08:00 level=INFO source=sched.go:382 msg="loaded runners" count=1 time=2024-06-26T15:33:53.647+08:00 level=INFO source=server.go:556 msg="waiting for llama runner to start responding" time=2024-06-26T15:33:53.648+08:00 level=INFO source=server.go:594 msg="waiting for server to become available" status="llm server error" INFO [wmain] build info | build=3171 commit="7c26775a" tid="28472" timestamp=1719387233 INFO [wmain] system info | n_threads=8 n_threads_batch=-1 system_info="AVX = 1 | AVX_VNNI = 0 | AVX2 = 0 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | AVX512_BF16 = 0 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1 | " tid="28472" timestamp=1719387233 total_threads=16 INFO [wmain] HTTP server listening | hostname="127.0.0.1" n_threads_http="15" port="51536" tid="28472" timestamp=1719387233 llama_model_loader: loaded meta data with 22 key-value pairs and 291 tensors from C:\Users\Administrator\.ollama\models\blobs\sha256-08d6043f9b965d26fe8568c7b791d5b312eaaf09ce912d36defaca55f5cd711c (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 = llama llama_model_loader: - kv 1: general.name str = Llama-3 llama_model_loader: - kv 2: llama.block_count u32 = 32 llama_model_loader: - kv 3: llama.context_length u32 = 8192 llama_model_loader: - kv 4: llama.embedding_length u32 = 4096 llama_model_loader: - kv 5: llama.feed_forward_length u32 = 14336 llama_model_loader: - kv 6: llama.attention.head_count u32 = 32 llama_model_loader: - kv 7: llama.attention.head_count_kv u32 = 8 llama_model_loader: - kv 8: llama.rope.freq_base f32 = 500000.000000 llama_model_loader: - kv 9: llama.attention.layer_norm_rms_epsilon f32 = 0.000010 llama_model_loader: - kv 10: general.file_type u32 = 17 llama_model_loader: - kv 11: llama.vocab_size u32 = 128256 llama_model_loader: - kv 12: llama.rope.dimension_count u32 = 128 llama_model_loader: - kv 13: tokenizer.ggml.model str = gpt2 llama_model_loader: - kv 14: tokenizer.ggml.pre str = llama-bpe llama_model_loader: - kv 15: tokenizer.ggml.tokens arr[str,128256] = ["!", "\"", "#", "$", "%", "&", "'", ... llama_model_loader: - kv 16: tokenizer.ggml.token_type arr[i32,128256] = [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, ... llama_model_loader: - kv 17: tokenizer.ggml.merges arr[str,280147] = ["Ġ Ġ", "Ġ ĠĠĠ", "ĠĠ ĠĠ", "... llama_model_loader: - kv 18: tokenizer.ggml.bos_token_id u32 = 128000 llama_model_loader: - kv 19: tokenizer.ggml.eos_token_id u32 = 128001 llama_model_loader: - kv 20: tokenizer.chat_template str = {% set loop_messages = messages %}{% ... llama_model_loader: - kv 21: general.quantization_version u32 = 2 llama_model_loader: - type f32: 65 tensors llama_model_loader: - type q5_K: 193 tensors llama_model_loader: - type q6_K: 33 tensors time=2024-06-26T15:33:53.905+08:00 level=INFO source=server.go:594 msg="waiting for server to become available" status="llm server loading model" llm_load_vocab: special tokens cache size = 256 llm_load_vocab: token to piece cache size = 0.8000 MB llm_load_print_meta: format = GGUF V3 (latest) llm_load_print_meta: arch = llama llm_load_print_meta: vocab type = BPE llm_load_print_meta: n_vocab = 128256 llm_load_print_meta: n_merges = 280147 llm_load_print_meta: n_ctx_train = 8192 llm_load_print_meta: n_embd = 4096 llm_load_print_meta: n_head = 32 llm_load_print_meta: n_head_kv = 8 llm_load_print_meta: n_layer = 32 llm_load_print_meta: n_rot = 128 llm_load_print_meta: n_embd_head_k = 128 llm_load_print_meta: n_embd_head_v = 128 llm_load_print_meta: n_gqa = 4 llm_load_print_meta: n_embd_k_gqa = 1024 llm_load_print_meta: n_embd_v_gqa = 1024 llm_load_print_meta: f_norm_eps = 0.0e+00 llm_load_print_meta: f_norm_rms_eps = 1.0e-05 llm_load_print_meta: f_clamp_kqv = 0.0e+00 llm_load_print_meta: f_max_alibi_bias = 0.0e+00 llm_load_print_meta: f_logit_scale = 0.0e+00 llm_load_print_meta: n_ff = 14336 llm_load_print_meta: n_expert = 0 llm_load_print_meta: n_expert_used = 0 llm_load_print_meta: causal attn = 1 llm_load_print_meta: pooling type = 0 llm_load_print_meta: rope type = 0 llm_load_print_meta: rope scaling = linear llm_load_print_meta: freq_base_train = 500000.0 llm_load_print_meta: freq_scale_train = 1 llm_load_print_meta: n_ctx_orig_yarn = 8192 llm_load_print_meta: rope_finetuned = unknown llm_load_print_meta: ssm_d_conv = 0 llm_load_print_meta: ssm_d_inner = 0 llm_load_print_meta: ssm_d_state = 0 llm_load_print_meta: ssm_dt_rank = 0 llm_load_print_meta: model type = 8B llm_load_print_meta: model ftype = Q5_K - Medium llm_load_print_meta: model params = 8.03 B llm_load_print_meta: model size = 5.33 GiB (5.70 BPW) llm_load_print_meta: general.name = Llama-3 llm_load_print_meta: BOS token = 128000 '<|begin_of_text|>' llm_load_print_meta: EOS token = 128001 '<|end_of_text|>' llm_load_print_meta: LF token = 128 'Ä' llm_load_print_meta: EOT token = 128009 '<|eot_id|>' ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no ggml_cuda_init: CUDA_USE_TENSOR_CORES: yes ggml_cuda_init: found 1 ROCm devices: Device 0: AMD Radeon VII, compute capability 9.0, VMM: no llm_load_tensors: ggml ctx size = 0.30 MiB llm_load_tensors: offloading 32 repeating layers to GPU llm_load_tensors: offloading non-repeating layers to GPU llm_load_tensors: offloaded 33/33 layers to GPU llm_load_tensors: ROCm0 buffer size = 5115.49 MiB llm_load_tensors: CPU buffer size = 344.44 MiB llama_new_context_with_model: n_ctx = 2048 llama_new_context_with_model: n_batch = 512 llama_new_context_with_model: n_ubatch = 512 llama_new_context_with_model: flash_attn = 0 llama_new_context_with_model: freq_base = 500000.0 llama_new_context_with_model: freq_scale = 1 llama_kv_cache_init: ROCm0 KV buffer size = 256.00 MiB llama_new_context_with_model: KV self size = 256.00 MiB, K (f16): 128.00 MiB, V (f16): 128.00 MiB llama_new_context_with_model: ROCm_Host output buffer size = 0.50 MiB llama_new_context_with_model: ROCm0 compute buffer size = 258.50 MiB llama_new_context_with_model: ROCm_Host compute buffer size = 12.01 MiB llama_new_context_with_model: graph nodes = 1030 llama_new_context_with_model: graph splits = 2 time=2024-06-26T15:34:01.844+08:00 level=INFO source=server.go:594 msg="waiting for server to become available" status="llm server error" time=2024-06-26T15:34:02.350+08:00 level=ERROR source=sched.go:388 msg="error loading llama server" error="llama runner process has terminated: exit status 0xc0000005 " [GIN] 2024/06/26 - 15:34:02 | 500 | 9.5332306s | 127.0.0.1 | POST "/api/chat"
GiteaMirror commented 2026-05-03 19:38:09 -05:00
Author
Owner
Copy Link

@MrSteelRat commented on GitHub (Jun 27, 2024):

Full log for experince https://pastebin.com/sB5LdD9h

<!-- gh-comment-id:2195242523 --> @MrSteelRat commented on GitHub (Jun 27, 2024): Full log for experince https://pastebin.com/sB5LdD9h
GiteaMirror commented 2026-05-03 19:38:11 -05:00
Author
Owner
Copy Link

@fofocx commented on GitHub (Jun 28, 2024):

Another value "gfx906:sramecc-:xnack+" after installed AMD HIP SDK.
Is there a way to enable ecc?😂
server.log:
2024/06/28 19:33:47 routes.go:1064: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE: OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:1 OLLAMA_MAX_QUEUE:512 OLLAMA_MAX_VRAM:0 OLLAMA_MODELS:H:\models OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:1 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://*] OLLAMA_RUNNERS_DIR:C:\Users\Administrator\AppData\Local\Programs\Ollama\ollama_runners OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]"
time=2024-06-28T19:33:47.704+08:00 level=INFO source=images.go:730 msg="total blobs: 7"
time=2024-06-28T19:33:47.705+08:00 level=INFO source=images.go:737 msg="total unused blobs removed: 0"
time=2024-06-28T19:33:47.705+08:00 level=INFO source=routes.go:1111 msg="Listening on [::]:11434 (version 0.1.47)"
time=2024-06-28T19:33:47.707+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11.3 rocm_v5.7]"
time=2024-06-28T19:33:47.838+08:00 level=WARN source=amd_windows.go:97 msg="amdgpu is not supported" gpu=0 gpu_type=gfx906:sramecc-:xnack+ library="C:\Program Files\AMD\ROCm\5.7\bin" supported_types="[gfx1030 gfx1100 gfx1101 gfx1102 gfx906]"
time=2024-06-28T19:33:47.838+08:00 level=WARN source=amd_windows.go:99 msg="See https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md for HSA_OVERRIDE_GFX_VERSION usage"
time=2024-06-28T19:33:47.897+08:00 level=INFO source=types.go:98 msg="inference compute" id=0 library=cpu compute="" driver=0.0 name="" total="31.9 GiB" available="20.6 GiB"

<!-- gh-comment-id:2196717435 --> @fofocx commented on GitHub (Jun 28, 2024): Another value "gfx906:sramecc-:xnack+" after installed AMD HIP SDK. Is there a way to enable ecc?😂 server.log: 2024/06/28 19:33:47 routes.go:1064: INFO server config env="map[CUDA_VISIBLE_DEVICES: GPU_DEVICE_ORDINAL: HIP_VISIBLE_DEVICES: HSA_OVERRIDE_GFX_VERSION: OLLAMA_DEBUG:false OLLAMA_FLASH_ATTENTION:false OLLAMA_HOST:http://0.0.0.0:11434 OLLAMA_INTEL_GPU:false OLLAMA_KEEP_ALIVE: OLLAMA_LLM_LIBRARY: OLLAMA_MAX_LOADED_MODELS:1 OLLAMA_MAX_QUEUE:512 OLLAMA_MAX_VRAM:0 OLLAMA_MODELS:H:\\models OLLAMA_NOHISTORY:false OLLAMA_NOPRUNE:false OLLAMA_NUM_PARALLEL:1 OLLAMA_ORIGINS:[http://localhost https://localhost http://localhost:* https://localhost:* http://127.0.0.1 https://127.0.0.1 http://127.0.0.1:* https://127.0.0.1:* http://0.0.0.0 https://0.0.0.0 http://0.0.0.0:* https://0.0.0.0:* app://* file://* tauri://*] OLLAMA_RUNNERS_DIR:C:\\Users\\Administrator\\AppData\\Local\\Programs\\Ollama\\ollama_runners OLLAMA_SCHED_SPREAD:false OLLAMA_TMPDIR: ROCR_VISIBLE_DEVICES:]" time=2024-06-28T19:33:47.704+08:00 level=INFO source=images.go:730 msg="total blobs: 7" time=2024-06-28T19:33:47.705+08:00 level=INFO source=images.go:737 msg="total unused blobs removed: 0" time=2024-06-28T19:33:47.705+08:00 level=INFO source=routes.go:1111 msg="Listening on [::]:11434 (version 0.1.47)" time=2024-06-28T19:33:47.707+08:00 level=INFO source=payload.go:44 msg="Dynamic LLM libraries [cpu cpu_avx cpu_avx2 cuda_v11.3 rocm_v5.7]" time=2024-06-28T19:33:47.838+08:00 level=WARN source=amd_windows.go:97 msg="amdgpu is not supported" gpu=0 gpu_type=gfx906:sramecc-:xnack+ library="C:\\Program Files\\AMD\\ROCm\\5.7\\bin" supported_types="[gfx1030 gfx1100 gfx1101 gfx1102 gfx906]" time=2024-06-28T19:33:47.838+08:00 level=WARN source=amd_windows.go:99 msg="See https://github.com/ollama/ollama/blob/main/docs/troubleshooting.md for HSA_OVERRIDE_GFX_VERSION usage" time=2024-06-28T19:33:47.897+08:00 level=INFO source=types.go:98 msg="inference compute" id=0 library=cpu compute="" driver=0.0 name="" total="31.9 GiB" available="20.6 GiB"
GiteaMirror commented 2026-05-03 19:38:13 -05:00
Author
Owner
Copy Link

@dhiltgen commented on GitHub (Jul 22, 2024):

Unfortunately with the latest ROCm v6 these cards are no longer supported on Windows. I've updated our docs to reflect this in #5815

<!-- gh-comment-id:2243399488 --> @dhiltgen commented on GitHub (Jul 22, 2024): Unfortunately with the latest ROCm v6 these cards are no longer supported on Windows. I've updated our docs to reflect this in #5815
GiteaMirror commented 2026-05-03 19:38:14 -05:00
Author
Owner
Copy Link

@MrSteelRat commented on GitHub (Aug 3, 2024):

So, support for Radeon VII in ollama is not worth waiting for?

<!-- gh-comment-id:2266674335 --> @MrSteelRat commented on GitHub (Aug 3, 2024): So, support for Radeon VII in ollama is not worth waiting for?
GiteaMirror commented 2026-05-03 19:38:16 -05:00
Author
Owner
Copy Link

@miningnome commented on GitHub (Feb 1, 2025):

Hi,

It's supposed the version 24.12.1 can be used for ROCm with Radeon VII. If you are on windows you shall do the wsl process.

<!-- gh-comment-id:2628655208 --> @miningnome commented on GitHub (Feb 1, 2025): Hi, It's supposed the version 24.12.1 can be used for ROCm with Radeon VII. If you are on windows you shall do the wsl process.
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 amd feature request gpu windows
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#65047
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?