diff --git a/sub791dockerrizhi_clean.txt b/sub791dockerrizhi_clean.txt new file mode 100644 index 00000000..ff342ba7 --- /dev/null +++ b/sub791dockerrizhi_clean.txt @@ -0,0 +1,3344 @@ +2026-08-21T09:01:45.658004498Z /usr/local/corex/lib64/python3/dist-packages/torch/cuda/__init__.py:51: FutureWarning: The pynvml package is deprecated. Please install nvidia-ml-py instead. If you did not install pynvml directly, please report this to the maintainers of the package that installed pynvml for you. +2026-08-21T09:01:45.658066491Z import pynvml # type: ignore[import] +2026-08-21T09:01:47.200984875Z INFO 08-21 09:01:47 importing.py:10] Triton not installed; certain GPU-related functions will not be available. +2026-08-21T09:01:48.689063592Z 2026-08-21 09:01:48.688962: I tensorflow/core/util/port.cc:110] oneDNN custom operations are on. You may see slightly different numerical results due to floating-point round-off errors from different computation orders. To turn them off, set the environment variable `TF_ENABLE_ONEDNN_OPTS=0`. +2026-08-21T09:01:48.742734644Z 2026-08-21 09:01:48.742677: I tensorflow/core/platform/cpu_feature_guard.cc:182] This TensorFlow binary is optimized to use available CPU instructions in performance-critical operations. +2026-08-21T09:01:48.742746164Z To enable the following instructions: SSE3 SSE4.1 SSE4.2 AVX AVX2 AVX512F AVX512_VNNI AVX512_BF16 AVX_VNNI AMX_TILE AMX_INT8 AMX_BF16 FMA, in other operations, rebuild TensorFlow with the appropriate compiler flags. +2026-08-21T09:01:48.773931628Z WARNING:tensorflow:Deprecation warnings have been disabled. Set TF_ENABLE_DEPRECATION_WARNINGS=1 to re-enable them. +2026-08-21T09:01:54.159789381Z INFO 08-21 09:01:54 api_server.py:1006] vLLM API server version 0.6.3 +2026-08-21T09:01:54.160020367Z INFO 08-21 09:01:54 api_server.py:1007] args: Namespace(host=None, port=8000, uvicorn_log_level='info', allow_credentials=False, allowed_origins=['*'], allowed_methods=['*'], allowed_headers=['*'], api_key=None, lora_modules=None, prompt_adapters=None, chat_template=None, response_role='assistant', ssl_keyfile=None, ssl_certfile=None, ssl_ca_certs=None, ssl_cert_reqs=0, root_path=None, middleware=[], return_tokens_as_token_ids=False, disable_frontend_multiprocessing=True, enable_auto_tool_choice=True, tool_call_parser='qwen3_coder', tool_parser_plugin='', reasoning_parser='qwen3', model='/model', tokenizer=None, skip_tokenizer_init=False, revision=None, code_revision=None, tokenizer_revision=None, tokenizer_mode='auto', trust_remote_code=True, download_dir=None, load_format='auto', config_format='auto', dtype='auto', kv_cache_dtype='auto', quantization_param_path=None, max_model_len=262144, guided_decoding_backend='outlines', distributed_executor_backend=None, worker_use_ray=False, pipeline_parallel_size=1, tensor_parallel_size=4, max_parallel_loading_workers=None, ray_workers_use_nsight=False, block_size=16, enable_prefix_caching=True, disable_sliding_window=False, use_v2_block_manager=True, num_lookahead_slots=0, seed=0, swap_space=4, cpu_offload_gb=0, gpu_memory_utilization=0.9, num_gpu_blocks_override=None, max_num_batched_tokens=8192, max_num_seqs=1, max_logprobs=20, disable_log_stats=False, quantization=None, rope_scaling=None, rope_theta=None, enforce_eager=True, max_context_len_to_capture=None, max_seq_len_to_capture=32768, disable_custom_all_reduce=False, tokenizer_pool_size=0, tokenizer_pool_type='ray', tokenizer_pool_extra_config=None, limit_mm_per_prompt=None, mm_processor_kwargs=None, enable_lora=False, max_loras=1, max_lora_rank=16, lora_extra_vocab_size=256, lora_dtype='auto', long_lora_scaling_factors=None, max_cpu_loras=None, fully_sharded_loras=False, enable_prompt_adapter=False, max_prompt_adapters=1, max_prompt_adapter_token=0, device='auto', num_scheduler_steps=1, multi_step_stream_outputs=True, scheduler_delay_factor=0.0, enable_chunked_prefill=True, speculative_model=None, speculative_model_quantization=None, num_speculative_tokens=None, speculative_disable_mqa_scorer=False, speculative_draft_tensor_parallel_size=None, speculative_max_model_len=None, speculative_disable_by_batch_size=None, ngram_prompt_lookup_max=None, ngram_prompt_lookup_min=None, spec_decoding_acceptance_method='rejection_sampler', typical_acceptance_sampler_posterior_threshold=None, typical_acceptance_sampler_posterior_alpha=None, disable_logprobs_during_spec_decoding=None, model_loader_extra_config=None, ignore_patterns=[], preemption_mode=None, served_model_name=['llm'], qlora_adapter_name_or_path=None, otlp_traces_endpoint=None, collect_detailed_traces=None, disable_async_output_proc=False, override_neuron_config=None, scheduling_policy='fcfs', disable_log_requests=True, max_log_len=None, disable_fastapi_docs=False) +2026-08-21T09:01:54.175663565Z INFO 08-21 09:01:54 config.py:1670] Downcasting torch.float32 to torch.float16. +2026-08-21T09:02:04.790340814Z INFO 08-21 09:02:04 arg_utils.py:909] Keeping prefix caching enabled for the Qwen3.6 native vision path. +2026-08-21T09:02:04.790824978Z INFO 08-21 09:02:04 config.py:887] Defaulting to use mp for distributed inference +2026-08-21T09:02:04.791125425Z INFO 08-21 09:02:04 config.py:1005] Chunked prefill is enabled with max_num_batched_tokens=8192. +2026-08-21T09:02:04.791228968Z WARNING 08-21 09:02:04 config.py:380] To see benefits of async output processing, enable CUDA graph. Since, enforce-eager is enabled, async output processor cannot be used +2026-08-21T09:02:04.797776622Z INFO 08-21 09:02:04 llm_engine.py:237] Initializing an LLM engine (v0.6.3) with config: model='/model', speculative_config=None, tokenizer='/model', skip_tokenizer_init=False, tokenizer_mode=auto, revision=None, override_neuron_config=None, rope_scaling=None, rope_theta=None, tokenizer_revision=None, trust_remote_code=True, dtype=torch.float16, max_seq_len=262144, download_dir=None, load_format=LoadFormat.AUTO, tensor_parallel_size=4, pipeline_parallel_size=1, disable_custom_all_reduce=True, quantization=None, enforce_eager=True, kv_cache_dtype=auto, quantization_param_path=None, device_config=cuda, decoding_config=DecodingConfig(guided_decoding_backend='outlines'), observability_config=ObservabilityConfig(otlp_traces_endpoint=None, collect_model_forward_time=False, collect_model_execute_time=False), seed=0, served_model_name=llm, use_v2_block_manager=True, num_scheduler_steps=1, chunked_prefill_enabled=True multi_step_stream_outputs=True, enable_prefix_caching=True, use_async_output_proc=False, use_cached_outputs=False, mm_processor_kwargs=None) +2026-08-21T09:02:05.293565575Z WARNING 08-21 09:02:05 multiproc_gpu_executor.py:58] Reducing Torch parallelism from 64 threads to 1 to avoid unnecessary CPU contention. Set OMP_NUM_THREADS in the external environment to tune this value as needed. +2026-08-21T09:02:05.335736867Z INFO 08-21 09:02:05 custom_cache_manager.py:17] Setting Triton cache manager to: vllm.triton_utils.custom_cache_manager:CustomCacheManager +2026-08-21T09:02:05.393315501Z INFO 08-21 09:02:05 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:05.393341804Z INFO 08-21 09:02:05 selector.py:115] Using XFormers backend. +2026-08-21T09:02:05.807302697Z /usr/local/corex/lib64/python3/dist-packages/torch/cuda/__init__.py:51: FutureWarning: The pynvml package is deprecated. Please install nvidia-ml-py instead. If you did not install pynvml directly, please report this to the maintainers of the package that installed pynvml for you. +2026-08-21T09:02:05.807309638Z import pynvml # type: ignore[import] +2026-08-21T09:02:05.824127325Z /usr/local/corex/lib64/python3/dist-packages/torch/cuda/__init__.py:51: FutureWarning: The pynvml package is deprecated. Please install nvidia-ml-py instead. If you did not install pynvml directly, please report this to the maintainers of the package that installed pynvml for you. +2026-08-21T09:02:05.824137656Z import pynvml # type: ignore[import] +2026-08-21T09:02:05.824256931Z /usr/local/corex/lib64/python3/dist-packages/torch/cuda/__init__.py:51: FutureWarning: The pynvml package is deprecated. Please install nvidia-ml-py instead. If you did not install pynvml directly, please report this to the maintainers of the package that installed pynvml for you. +2026-08-21T09:02:05.824269321Z import pynvml # type: ignore[import] +2026-08-21T09:02:07.329940131Z INFO 08-21 09:02:07 importing.py:10] Triton not installed; certain GPU-related functions will not be available. +2026-08-21T09:02:07.337795612Z INFO 08-21 09:02:07 importing.py:10] Triton not installed; certain GPU-related functions will not be available. +2026-08-21T09:02:07.337940204Z INFO 08-21 09:02:07 importing.py:10] Triton not installed; certain GPU-related functions will not be available. +2026-08-21T09:02:08.918983194Z WARNING:tensorflow:Deprecation warnings have been disabled. Set TF_ENABLE_DEPRECATION_WARNINGS=1 to re-enable them. +2026-08-21T09:02:08.918992985Z WARNING:tensorflow:Deprecation warnings have been disabled. Set TF_ENABLE_DEPRECATION_WARNINGS=1 to re-enable them. +2026-08-21T09:02:08.918996192Z WARNING:tensorflow:Deprecation warnings have been disabled. Set TF_ENABLE_DEPRECATION_WARNINGS=1 to re-enable them. +2026-08-21T09:02:14.460559487Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.460602035Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:14.461994280Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.462040598Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:14.471925854Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.471977270Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:14.482641607Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 multiproc_worker_utils.py:216] Worker ready; awaiting tasks +2026-08-21T09:02:14.483730800Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 multiproc_worker_utils.py:216] Worker ready; awaiting tasks +2026-08-21T09:02:14.493826451Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 multiproc_worker_utils.py:216] Worker ready; awaiting tasks +2026-08-21T09:02:14.817030241Z INFO 08-21 09:02:14 shm_broadcast.py:242] vLLM message queue communication handle: Handle(connect_ip='127.0.0.1', local_reader_ranks=[1, 2, 3], buffer=, local_subscribe_port=59293, remote_subscribe_port=None) +2026-08-21T09:02:14.847415260Z INFO 08-21 09:02:14 model_runner.py:1138] Starting to load model /model... +2026-08-21T09:02:14.847614690Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 model_runner.py:1138] Starting to load model /model... +2026-08-21T09:02:14.847739181Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 model_runner.py:1138] Starting to load model /model... +2026-08-21T09:02:14.848387320Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 model_runner.py:1138] Starting to load model /model... +2026-08-21T09:02:14.858988644Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 qwen3_5.py:2024] [BI100] Qwen hybrid KV accounting; tp_rank=2 env_mode=full_attention config_mode=full_attention configured_kv_layers=10 full_attention_layers=10 full_attention_ordinals=3,7,11,15,19,23,27,31,35,39 +2026-08-21T09:02:14.859467814Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 qwen3_5.py:2024] [BI100] Qwen hybrid KV accounting; tp_rank=1 env_mode=full_attention config_mode=full_attention configured_kv_layers=10 full_attention_layers=10 full_attention_ordinals=3,7,11,15,19,23,27,31,35,39 +2026-08-21T09:02:14.864551807Z INFO 08-21 09:02:14 qwen3_5.py:2024] [BI100] Qwen hybrid KV accounting; tp_rank=0 env_mode=full_attention config_mode=full_attention configured_kv_layers=10 full_attention_layers=10 full_attention_ordinals=3,7,11,15,19,23,27,31,35,39 +2026-08-21T09:02:14.864802679Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 qwen3_5.py:2024] [BI100] Qwen hybrid KV accounting; tp_rank=3 env_mode=full_attention config_mode=full_attention configured_kv_layers=10 full_attention_layers=10 full_attention_ordinals=3,7,11,15,19,23,27,31,35,39 +2026-08-21T09:02:14.903080269Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.903130512Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:14.903137640Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.903191892Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:14.914028211Z INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.914071736Z INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:14.920641847Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.920693615Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:15.030346513Z +Loading safetensors checkpoint shards: 0% Completed | 0/16 [00:00, local_subscribe_port=59293, remote_subscribe_port=None) +2026-08-21T09:02:14.847415260Z INFO 08-21 09:02:14 model_runner.py:1138] Starting to load model /model... +2026-08-21T09:02:14.847614690Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 model_runner.py:1138] Starting to load model /model... +2026-08-21T09:02:14.847739181Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 model_runner.py:1138] Starting to load model /model... +2026-08-21T09:02:14.848387320Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 model_runner.py:1138] Starting to load model /model... +2026-08-21T09:02:14.858988644Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 qwen3_5.py:2024] [BI100] Qwen hybrid KV accounting; tp_rank=2 env_mode=full_attention config_mode=full_attention configured_kv_layers=10 full_attention_layers=10 full_attention_ordinals=3,7,11,15,19,23,27,31,35,39 +2026-08-21T09:02:14.859467814Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 qwen3_5.py:2024] [BI100] Qwen hybrid KV accounting; tp_rank=1 env_mode=full_attention config_mode=full_attention configured_kv_layers=10 full_attention_layers=10 full_attention_ordinals=3,7,11,15,19,23,27,31,35,39 +2026-08-21T09:02:14.864551807Z INFO 08-21 09:02:14 qwen3_5.py:2024] [BI100] Qwen hybrid KV accounting; tp_rank=0 env_mode=full_attention config_mode=full_attention configured_kv_layers=10 full_attention_layers=10 full_attention_ordinals=3,7,11,15,19,23,27,31,35,39 +2026-08-21T09:02:14.864802679Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 qwen3_5.py:2024] [BI100] Qwen hybrid KV accounting; tp_rank=3 env_mode=full_attention config_mode=full_attention configured_kv_layers=10 full_attention_layers=10 full_attention_ordinals=3,7,11,15,19,23,27,31,35,39 +2026-08-21T09:02:14.903080269Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.903130512Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:14.903137640Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.903191892Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:14.914028211Z INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.914071736Z INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:14.920641847Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.920693615Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:15.030346513Z +Loading safetensors checkpoint shards: 0% Completed | 0/16 [00:00, local_subscribe_port=59293, remote_subscribe_port=None) +2026-08-21T09:02:14.847415260Z INFO 08-21 09:02:14 model_runner.py:1138] Starting to load model /model... +2026-08-21T09:02:14.847614690Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 model_runner.py:1138] Starting to load model /model... +2026-08-21T09:02:14.847739181Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 model_runner.py:1138] Starting to load model /model... +2026-08-21T09:02:14.848387320Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 model_runner.py:1138] Starting to load model /model... +2026-08-21T09:02:14.858988644Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 qwen3_5.py:2024] [BI100] Qwen hybrid KV accounting; tp_rank=2 env_mode=full_attention config_mode=full_attention configured_kv_layers=10 full_attention_layers=10 full_attention_ordinals=3,7,11,15,19,23,27,31,35,39 +2026-08-21T09:02:14.859467814Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 qwen3_5.py:2024] [BI100] Qwen hybrid KV accounting; tp_rank=1 env_mode=full_attention config_mode=full_attention configured_kv_layers=10 full_attention_layers=10 full_attention_ordinals=3,7,11,15,19,23,27,31,35,39 +2026-08-21T09:02:14.864551807Z INFO 08-21 09:02:14 qwen3_5.py:2024] [BI100] Qwen hybrid KV accounting; tp_rank=0 env_mode=full_attention config_mode=full_attention configured_kv_layers=10 full_attention_layers=10 full_attention_ordinals=3,7,11,15,19,23,27,31,35,39 +2026-08-21T09:02:14.864802679Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 qwen3_5.py:2024] [BI100] Qwen hybrid KV accounting; tp_rank=3 env_mode=full_attention config_mode=full_attention configured_kv_layers=10 full_attention_layers=10 full_attention_ordinals=3,7,11,15,19,23,27,31,35,39 +2026-08-21T09:02:14.903080269Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.903130512Z (VllmWorkerProcess pid=345) INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:14.903137640Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.903191892Z (VllmWorkerProcess pid=344) INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:14.914028211Z INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.914071736Z INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:14.920641847Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 selector.py:266] Cannot use FlashAttention-2 backend because the vllm.vllm_flash_attn package is not found. Make sure that vllm_flash_attn was built and installed (on by default). +2026-08-21T09:02:14.920693615Z (VllmWorkerProcess pid=346) INFO 08-21 09:02:14 selector.py:115] Using XFormers backend. +2026-08-21T09:02:15.030346513Z +Loading safetensors checkpoint shards: 0% Completed | 0/16 [00:00