问题是什么?
当我在物理机器上运行mixtral:8x7b-instruct-v0.1-q4_K_M时,遇到了这个错误:
[root@5dc6ecf27031 /]# ollama run mixtral:8x7b-instruct-v0.1-q4_K_M
Error: llama runner process has terminated: signal: segmentation fault (core dumped)
[root@5dc6ecf27031 /]#
日志:
[GIN] 2024/07/11 - 13:22:44 | 200 | 16.23µs | 127.0.0.1 | HEAD "/"
[GIN] 2024/07/11 - 13:22:44 | 200 | 7.724554ms | 127.0.0.1 | POST "/api/show"
time=2024-07-11T13:22:44.297Z level=INFO source=sched.go:754 msg="new model will fit in available VRAM, loading" model=/root/.ollama/models/blobs/sha256-3a17f7cde150070bbc815645693fb93c311cc42e7deaf198364acadcf08458f8 library=rocm parallel=4 required="33.2 GiB"
time=2024-07-11T13:22:44.298Z level=INFO source=memory.go:309 msg="offload to rocm" layers.requested=-1 layers.model=33 layers.offload=33 layers.split=11,11,11 memory.available="[24.0 GiB 24.0 GiB 24.0 GiB]" memory.required.full="33.2 GiB" memory.required.partial="33.2 GiB" memory.required.kv="1.0 GiB" memory.required.allocations="[11.3 GiB 11.3 GiB 10.6 GiB]" memory.weights.total="25.5 GiB" memory.weights.repeating="25.4 GiB" memory.weights.nonrepeating="102.6 MiB" memory.graph.full="1.3 GiB" memory.graph.partial="1.3 GiB"
time=2024-07-11T13:22:44.299Z level=INFO source=server.go:375 msg="starting llama server" cmd="/tmp/ollama1419561683/runners/rocm_v60101/ollama_llama_server --model /root/.ollama/models/blobs/sha256-3a17f7cde150070bbc815645693fb93c311cc42e7deaf198364acadcf08458f8 --ctx-size 8192 --batch-size 512 --embedding --log-disable --n-gpu-layers 33 --parallel 4 --tensor-split 11,11,11 --tensor-split 11,11,11 --port 41695"
time=2024-07-11T13:22:44.299Z level=INFO source=sched.go:474 msg="loaded runners" count=1
time=2024-07-11T13:22:44.299Z level=INFO source=server.go:563 msg="waiting for llama runner to start responding"
time=2024-07-11T13:22:44.299Z level=INFO source=server.go:604 msg="waiting for server to become available" status="llm server error"
INFO [main] build info | build=1 commit="a8db2a9" tid="140134008951616" timestamp=1720704164
INFO [main] system info | n_threads=16 n_threads_batch=-1 system_info="AVX = 1 | AVX_VNNI = 0 | AVX2 = 0 | AVX512 = 0 | AVX512_VBMI = 0 | AVX512_VNNI = 0 | AVX512_BF16 = 0 | FMA = 0 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 0 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 1 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 0 | " tid="140134008951616" timestamp=1720704164 total_threads=32
INFO [main] HTTP server listening | hostname="127.0.0.1" n_threads_http="31" port="41695" tid="140134008951616" timestamp=1720704164
llama_model_loader: loaded meta data with 26 key-value pairs and 995 tensors from /root/.ollama/models/blobs/sha256-3a17f7cde150070bbc815645693fb93c311cc42e7deaf198364acadcf08458f8 (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 = mistralai
llama_model_loader: - kv 2: llama.context_length u32 = 32768
llama_model_loader: - kv 3: llama.embedding_length u32 = 4096
llama_model_loader: - kv 4: llama.block_count u32 = 32
llama_model_loader: - kv 5: llama.feed_forward_length u32 = 14336
llama_model_loader: - kv 6: llama.rope.dimension_count u32 = 128
llama_model_loader: - kv 7: llama.attention.head_count u32 = 32
llama_model_loader: - kv 8: llama.attention.head_count_kv u32 = 8
llama_model_loader: - kv 9: llama.expert_count u32 = 8
llama_model_loader: - kv 10: llama.expert_used_count u32 = 2
llama_model_loader: - kv 11: llama.attention.layer_norm_rms_epsilon f32 = 0.000010
llama_model_loader: - kv 12: llama.rope.freq_base f32 = 1000000.000000
llama_model_loader: - kv 13: general.file_type u32 = 15
llama_model_loader: - kv 14: tokenizer.ggml.model str = llama
llama_model_loader: - kv 15: tokenizer.ggml.tokens arr[str,32000] = ["<unk>", "<s>", "</s>", "<0x00>", "<...
llama_model_loader: - kv 16: tokenizer.ggml.scores arr[f32,32000] = [0.000000, 0.000000, 0.000000, 0.0000...
llama_model_loader: - kv 17: tokenizer.ggml.token_type arr[i32,32000] = [2, 3, 3, 6, 6, 6, 6, 6, 6, 6, 6, 6, ...
llama_model_loader: - kv 18: tokenizer.ggml.merges arr[str,58980] = ["▁ t", "i n", "e r", "▁ a", "h e...
llama_model_loader: - kv 19: tokenizer.ggml.bos_token_id u32 = 1
llama_model_loader: - kv 20: tokenizer.ggml.eos_token_id u32 = 2
llama_model_loader: - kv 21: tokenizer.ggml.unknown_token_id u32 = 0
llama_model_loader: - kv 22: tokenizer.ggml.add_bos_token bool = true
llama_model_loader: - kv 23: tokenizer.ggml.add_eos_token bool = false
llama_model_loader: - kv 24: tokenizer.chat_template str = {{ bos_token }}{% for message in mess...
llama_model_loader: - kv 25: general.quantization_version u32 = 2
llama_model_loader: - type f32: 65 tensors
llama_model_loader: - type f16: 32 tensors
llama_model_loader: - type q8_0: 64 tensors
llama_model_loader: - type q4_K: 833 tensors
llama_model_loader: - type q6_K: 1 tensors
llm_load_vocab: special tokens cache size = 259
llm_load_vocab: token to piece cache size = 0.1637 MB
llm_load_print_meta: format = GGUF V3 (latest)
llm_load_print_meta: arch = llama
llm_load_print_meta: vocab type = SPM
llm_load_print_meta: n_vocab = 32000
llm_load_print_meta: n_merges = 0
llm_load_print_meta: vocab_only = 0
llm_load_print_meta: n_ctx_train = 32768
llm_load_print_meta: n_embd = 4096
llm_load_print_meta: n_layer = 32
llm_load_print_meta: n_head = 32
llm_load_print_meta: n_head_kv = 8
llm_load_print_meta: n_rot = 128
llm_load_print_meta: n_swa = 0
llm_load_print_meta: n_embd_head_k = 128
llm_load_print_meta: n_embd_head_v = 128
llm_load_print_meta: n_gqa = 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 = 8
llm_load_print_meta: n_expert_used = 2
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 = 1000000.0
llm_load_print_meta: freq_scale_train = 1
llm_load_print_meta: n_ctx_orig_yarn = 32768
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 = 8x7B
llm_load_print_meta: model ftype = Q4_K - Medium
llm_load_print_meta: model params = 46.70 B
llm_load_print_meta: model size = 24.62 GiB (4.53 BPW)
llm_load_print_meta: general.name = mistralai
llm_load_print_meta: BOS token = 1 '<s>'
llm_load_print_meta: EOS token = 2 '</s>'
llm_load_print_meta: UNK token = 0 '<unk>'
llm_load_print_meta: LF token = 13 '<0x0A>'
llm_load_print_meta: max token length = 48
time=2024-07-11T13:22:44.549Z level=INFO source=server.go:604 msg="waiting for server to become available" status="llm server loading model"
ggml_cuda_init: GGML_CUDA_FORCE_MMQ: no
ggml_cuda_init: GGML_CUDA_FORCE_CUBLAS: no
ggml_cuda_init: found 3 ROCm devices:
Device 0: Radeon RX 7900 XTX, compute capability 11.0, VMM: no
Device 1: Radeon RX 7900 XTX, compute capability 11.0, VMM: no
Device 2: Radeon RX 7900 XTX, compute capability 11.0, VMM: no
llm_load_tensors: ggml ctx size = 1.53 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 = 8608.53 MiB
llm_load_tensors: ROCm1 buffer size = 8608.53 MiB
llm_load_tensors: ROCm2 buffer size = 7928.49 MiB
llm_load_tensors: ROCm_Host buffer size = 70.31 MiB
time=2024-07-11T13:23:03.566Z level=INFO source=server.go:604 msg="waiting for server to become available" status="llm server not responding"
llama_new_context_with_model: n_ctx = 8192
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 = 1000000.0
llama_new_context_with_model: freq_scale = 1
time=2024-07-11T13:23:04.460Z level=INFO source=server.go:604 msg="waiting for server to become available" status="llm server loading model"
llama_kv_cache_init: ROCm0 KV buffer size = 352.00 MiB
llama_kv_cache_init: ROCm1 KV buffer size = 352.00 MiB
llama_kv_cache_init: ROCm2 KV buffer size = 320.00 MiB
llama_new_context_with_model: KV self size = 1024.00 MiB, K (f16): 512.00 MiB, V (f16): 512.00 MiB
llama_new_context_with_model: ROCm_Host output buffer size = 0.55 MiB
llama_new_context_with_model: pipeline parallelism enabled (n_copies=4)
llama_new_context_with_model: ROCm0 compute buffer size = 640.01 MiB
llama_new_context_with_model: ROCm1 compute buffer size = 640.01 MiB
llama_new_context_with_model: ROCm2 compute buffer size = 640.02 MiB
llama_new_context_with_model: ROCm_Host compute buffer size = 72.02 MiB
llama_new_context_with_model: graph nodes = 1510
llama_new_context_with_model: graph splits = 4
time=2024-07-11T13:23:06.864Z level=INFO source=server.go:604 msg="waiting for server to become available" status="llm server error"
[GIN] 2024/07/11 - 13:23:07 | 500 | 22.834580361s | 127.0.0.1 | POST "/api/chat"
time=2024-07-11T13:23:07.115Z level=ERROR source=sched.go:480 msg="error loading llama server" error="llama runner process has terminated: signal: segmentation fault (core dumped) "
time=2024-07-11T13:23:12.116Z level=WARN source=sched.go:671 msg="gpu VRAM usage didn't recover within timeout" seconds=5.001085328 model=/root/.ollama/models/blobs/sha256-3a17f7cde150070bbc815645693fb93c311cc42e7deaf198364acadcf08458f8
time=2024-07-11T13:23:12.366Z level=WARN source=sched.go:671 msg="gpu VRAM usage didn't recover within timeout" seconds=5.251122065 model=/root/.ollama/models/blobs/sha256-3a17f7cde150070bbc815645693fb93c311cc42e7deaf198364acadcf08458f8
time=2024-07-11T13:23:12.616Z level=WARN source=sched.go:671 msg="gpu VRAM usage didn't recover within timeout" seconds=5.500799906 model=/root/.ollama/models/blobs/sha256-3a17f7cde150070bbc815645693fb93c311cc42e7deaf198364acadcf08458f8
我正在运行这个Docker版本
docker run -d --restart unless-stopped --device /dev/kfd --device /dev/dri -v ollama:/root/.ollama -p 11442:11434 --name dvz3 ollama/ollama:0.2.1-rocm
OS
Linux
GPU
AMD
CPU
AMD
Ollama版本
0.2.1-rocm
6条答案
按热度按时间atmip9wb1#
对于混合8x7b <= q3,运行完美,但对于q4+,错误始终是:错误:llama runner进程已终止:信号:段错误(核心转储)。这很奇怪,因为我在GPU上有72gb内存。
MIXTRAL Q3 日志
Q4 日志:
hsgswve42#
错误是因为它只尝试适应一个GPU吗?
6qqygrtg3#
我目前也在AMD GPU(2 x 16GB 7800 XTs)上遇到了类似的内存不足错误,出现在
ollama 0.2.1
上。加载的模型使用了双GPU和CPU RAM的混合(有64GB的RAM可用,但只使用了很小一部分)。系统是Ubuntu 22.04。vh0rcniy4#
这个有什么更新吗?
wvmv3b1j5#
我无法复现这个问题,但我没有3x Radeon的设置 - 我的双Radeon测试盒似乎表现正常。
这可能是ROCm回归,或者在llama.cpp中b3051和b3171之间的回归。
ryhaxcpt6#
@dhiltgen ,如果这有帮助的话,我们可以为您提供3张卡的设置。请告诉我。