W0202 14:14:12.618000 10287 lib/python3.10/dist-packages/torch/distributed/run.py:793] W0202 14:14:12.618000 10287 lib/python3.10/dist-packages/torch/distributed/run.py:793] ***************************************** W0202 14:14:12.618000 10287 lib/python3.10/dist-packages/torch/distributed/run.py:793] Setting OMP_NUM_THREADS environment variable for each process to be 1 in default, to avoid your system being overloaded, please further tune the variable for optimal performance in your application as needed. W0202 14:14:12.618000 10287 lib/python3.10/dist-packages/torch/distributed/run.py:793] ***************************************** Namespace(model='HYVideo-T/2-cfgdistill', latent_channels=16, precision='bf16', rope_theta=256, vae='884-16c-hy', vae_precision='fp16', vae_tiling=True, text_encoder='llm', text_encoder_precision='fp16', text_states_dim=4096, text_len=256, tokenizer='llm', prompt_template='dit-llm-encode', prompt_template_video='dit-llm-encode-video', hidden_state_skip_layer=2, apply_final_norm=False, text_encoder_2='clipL', text_encoder_precision_2='fp16', text_states_dim_2=768, tokenizer_2='clipL', text_len_2=77, denoise_type='flow', flow_shift=7.0, flow_reverse=True, flow_solver='euler', use_linear_quadratic_schedule=False, linear_schedule_end=25, model_base='ckpts', dit_weight='ckpts/hunyuan-video-t2v-720p/transformers/mp_rank_00_model_states.pt', model_resolution='540p', load_key='module', use_cpu_offload=False, batch_size=1, infer_steps=20, disable_autocast=False, save_path='./results', save_path_suffix='', name_suffix='', num_videos=1, video_size=[1280, 720], video_length=33, prompt='A cat walks on the grass, realistic style.', seed_type='auto', seed=42, neg_prompt=None, cfg_scale=1.0, embedded_cfg_scale=6.0, use_fp8=False, reproduce=False, ulysses_degree=2, ring_degree=1) 2026-02-02 14:14:17.479 | INFO | hyvideo.inference:from_pretrained:154 - Got text-to-video model root path: ckpts WARNING: Logging before InitGoogleLogging() is written to STDERR I0202 14:14:17.495100 10321 ProcessGroupNCCL.cpp:934] [PG ID 0 PG GUID 0 Rank 0] ProcessGroupNCCL initialization options: size: 2, global rank: 0, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0, SPLIT_COLOR: 0, PG Name: 0 I0202 14:14:17.495119 10321 ProcessGroupNCCL.cpp:943] [PG ID 0 PG GUID 0 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 DEBUG 02-02 14:14:17 [parallel_state.py:200] world_size=2 rank=0 local_rank=-1 distributed_init_method=env:// backend=nccl I0202 14:14:17.495599 10321 ProcessGroupNCCL.cpp:934] [PG ID 1 PG GUID 1 Rank 0] ProcessGroupNCCL initialization options: size: 2, global rank: 0, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0x560712355bc0, SPLIT_COLOR: 22836467197190088, PG Name: 1 I0202 14:14:17.495608 10321 ProcessGroupNCCL.cpp:943] [PG ID 1 PG GUID 1 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 Namespace(model='HYVideo-T/2-cfgdistill', latent_channels=16, precision='bf16', rope_theta=256, vae='884-16c-hy', vae_precision='fp16', vae_tiling=True, text_encoder='llm', text_encoder_precision='fp16', text_states_dim=4096, text_len=256, tokenizer='llm', prompt_template='dit-llm-encode', prompt_template_video='dit-llm-encode-video', hidden_state_skip_layer=2, apply_final_norm=False, text_encoder_2='clipL', text_encoder_precision_2='fp16', text_states_dim_2=768, tokenizer_2='clipL', text_len_2=77, denoise_type='flow', flow_shift=7.0, flow_reverse=True, flow_solver='euler', use_linear_quadratic_schedule=False, linear_schedule_end=25, model_base='ckpts', dit_weight='ckpts/hunyuan-video-t2v-720p/transformers/mp_rank_00_model_states.pt', model_resolution='540p', load_key='module', use_cpu_offload=False, batch_size=1, infer_steps=20, disable_autocast=False, save_path='./results', save_path_suffix='', name_suffix='', num_videos=1, video_size=[1280, 720], video_length=33, prompt='A cat walks on the grass, realistic style.', seed_type='auto', seed=42, neg_prompt=None, cfg_scale=1.0, embedded_cfg_scale=6.0, use_fp8=False, reproduce=False, ulysses_degree=2, ring_degree=1) 2026-02-02 14:14:17.823 | INFO | hyvideo.inference:from_pretrained:154 - Got text-to-video model root path: ckpts WARNING: Logging before InitGoogleLogging() is written to STDERR I0202 14:14:17.837303 10322 ProcessGroupNCCL.cpp:934] [PG ID 0 PG GUID 0 Rank 1] ProcessGroupNCCL initialization options: size: 2, global rank: 1, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0, SPLIT_COLOR: 0, PG Name: 0 I0202 14:14:17.837325 10322 ProcessGroupNCCL.cpp:943] [PG ID 0 PG GUID 0 Rank 1] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 DEBUG 02-02 14:14:17 [parallel_state.py:200] world_size=2 rank=1 local_rank=-1 distributed_init_method=env:// backend=nccl I0202 14:14:17.837966 10322 ProcessGroupNCCL.cpp:934] [PG ID 1 PG GUID 1 Rank 1] ProcessGroupNCCL initialization options: size: 2, global rank: 1, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0x563b7a34bac0, SPLIT_COLOR: 22836467197190088, PG Name: 1 I0202 14:14:17.837976 10322 ProcessGroupNCCL.cpp:943] [PG ID 1 PG GUID 1 Rank 1] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.844038 10322 ProcessGroupNCCL.cpp:934] [PG ID 2 PG GUID 5 Rank 0] ProcessGroupNCCL initialization options: size: 1, global rank: 1, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0, SPLIT_COLOR: 0, PG Name: 5 I0202 14:14:17.844046 10322 ProcessGroupNCCL.cpp:943] [PG ID 2 PG GUID 5 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.844133 10321 ProcessGroupNCCL.cpp:934] [PG ID 2 PG GUID 3 Rank 0] ProcessGroupNCCL initialization options: size: 1, global rank: 0, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0, SPLIT_COLOR: 0, PG Name: 3 I0202 14:14:17.844147 10321 ProcessGroupNCCL.cpp:943] [PG ID 2 PG GUID 3 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.844980 10322 ProcessGroupNCCL.cpp:934] [PG ID 3 PG GUID 9 Rank 0] ProcessGroupNCCL initialization options: size: 1, global rank: 1, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0, SPLIT_COLOR: 0, PG Name: 9 I0202 14:14:17.844990 10322 ProcessGroupNCCL.cpp:943] [PG ID 3 PG GUID 9 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.845337 10321 ProcessGroupNCCL.cpp:934] [PG ID 3 PG GUID 7 Rank 0] ProcessGroupNCCL initialization options: size: 1, global rank: 0, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0, SPLIT_COLOR: 0, PG Name: 7 I0202 14:14:17.845350 10321 ProcessGroupNCCL.cpp:943] [PG ID 3 PG GUID 7 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.846011 10322 ProcessGroupNCCL.cpp:934] [PG ID 4 PG GUID 13 Rank 0] ProcessGroupNCCL initialization options: size: 1, global rank: 1, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0, SPLIT_COLOR: 0, PG Name: 13 I0202 14:14:17.846024 10322 ProcessGroupNCCL.cpp:943] [PG ID 4 PG GUID 13 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.846212 10321 ProcessGroupNCCL.cpp:934] [PG ID 4 PG GUID 11 Rank 0] ProcessGroupNCCL initialization options: size: 1, global rank: 0, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0, SPLIT_COLOR: 0, PG Name: 11 I0202 14:14:17.846223 10321 ProcessGroupNCCL.cpp:943] [PG ID 4 PG GUID 11 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.847054 10322 ProcessGroupNCCL.cpp:934] [PG ID 5 PG GUID 16 Rank 0] ProcessGroupNCCL initialization options: size: 1, global rank: 1, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0, SPLIT_COLOR: 0, PG Name: 16 I0202 14:14:17.847065 10322 ProcessGroupNCCL.cpp:943] [PG ID 5 PG GUID 16 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.847277 10321 ProcessGroupNCCL.cpp:934] [PG ID 5 PG GUID 15 Rank 0] ProcessGroupNCCL initialization options: size: 1, global rank: 0, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0, SPLIT_COLOR: 0, PG Name: 15 I0202 14:14:17.847280 10322 ProcessGroupNCCL.cpp:934] [PG ID 6 PG GUID 17 Rank 1] ProcessGroupNCCL initialization options: size: 2, global rank: 1, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0x563b7a34bac0, SPLIT_COLOR: 22836467197190088, PG Name: 17 I0202 14:14:17.847290 10321 ProcessGroupNCCL.cpp:943] [PG ID 5 PG GUID 15 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.847291 10322 ProcessGroupNCCL.cpp:943] [PG ID 6 PG GUID 17 Rank 1] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.847499 10322 ProcessGroupNCCL.cpp:934] [PG ID 7 PG GUID 19 Rank 0] ProcessGroupNCCL initialization options: size: 1, global rank: 1, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0, SPLIT_COLOR: 0, PG Name: 19 I0202 14:14:17.847507 10322 ProcessGroupNCCL.cpp:943] [PG ID 7 PG GUID 19 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.847539 10321 ProcessGroupNCCL.cpp:934] [PG ID 6 PG GUID 17 Rank 0] ProcessGroupNCCL initialization options: size: 2, global rank: 0, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0x560712355bc0, SPLIT_COLOR: 22836467197190088, PG Name: 17 I0202 14:14:17.847550 10321 ProcessGroupNCCL.cpp:943] [PG ID 6 PG GUID 17 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.847719 10322 ProcessGroupNCCL.cpp:934] [PG ID 8 PG GUID 20 Rank 1] ProcessGroupNCCL initialization options: size: 2, global rank: 1, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0x563b7a34bac0, SPLIT_COLOR: 22836467197190088, PG Name: 20 I0202 14:14:17.847729 10322 ProcessGroupNCCL.cpp:943] [PG ID 8 PG GUID 20 Rank 1] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.847836 10321 ProcessGroupNCCL.cpp:934] [PG ID 7 PG GUID 18 Rank 0] ProcessGroupNCCL initialization options: size: 1, global rank: 0, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0, SPLIT_COLOR: 0, PG Name: 18 I0202 14:14:17.847846 10321 ProcessGroupNCCL.cpp:943] [PG ID 7 PG GUID 18 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.848086 10321 ProcessGroupNCCL.cpp:934] [PG ID 8 PG GUID 20 Rank 0] ProcessGroupNCCL initialization options: size: 2, global rank: 0, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0x560712355bc0, SPLIT_COLOR: 22836467197190088, PG Name: 20 I0202 14:14:17.848096 10321 ProcessGroupNCCL.cpp:943] [PG ID 8 PG GUID 20 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.849359 10322 ProcessGroupNCCL.cpp:934] [PG ID 9 PG GUID 24 Rank 0] ProcessGroupNCCL initialization options: size: 1, global rank: 1, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0, SPLIT_COLOR: 0, PG Name: 24 I0202 14:14:17.849367 10322 ProcessGroupNCCL.cpp:943] [PG ID 9 PG GUID 24 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.849380 10321 ProcessGroupNCCL.cpp:934] [PG ID 9 PG GUID 22 Rank 0] ProcessGroupNCCL initialization options: size: 1, global rank: 0, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0, SPLIT_COLOR: 0, PG Name: 22 I0202 14:14:17.849390 10321 ProcessGroupNCCL.cpp:943] [PG ID 9 PG GUID 22 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.850406 10321 ProcessGroupNCCL.cpp:934] [PG ID 10 PG GUID 26 Rank 0] ProcessGroupNCCL initialization options: size: 2, global rank: 0, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0x560712355bc0, SPLIT_COLOR: 22836467197190088, PG Name: 26 I0202 14:14:17.850417 10321 ProcessGroupNCCL.cpp:943] [PG ID 10 PG GUID 26 Rank 0] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 I0202 14:14:17.850459 10322 ProcessGroupNCCL.cpp:934] [PG ID 10 PG GUID 26 Rank 1] ProcessGroupNCCL initialization options: size: 2, global rank: 1, TIMEOUT(ms): 600000, USE_HIGH_PRIORITY_STREAM: 0, SPLIT_FROM: 0x563b7a34bac0, SPLIT_COLOR: 22836467197190088, PG Name: 26 I0202 14:14:17.850469 10322 ProcessGroupNCCL.cpp:943] [PG ID 10 PG GUID 26 Rank 1] ProcessGroupNCCL environments: NCCL version: 2.22.3, TORCH_NCCL_ASYNC_ERROR_HANDLING: 1, TORCH_NCCL_DUMP_ON_TIMEOUT: 0, TORCH_NCCL_WAIT_TIMEOUT_DUMP_MILSEC: 60000, TORCH_NCCL_DESYNC_DEBUG: 0, TORCH_NCCL_ENABLE_TIMING: 0, TORCH_NCCL_BLOCKING_WAIT: 0, TORCH_DISTRIBUTED_DEBUG: OFF, TORCH_NCCL_USE_TENSOR_REGISTER_ALLOCATOR_HOOK: 0, TORCH_NCCL_ENABLE_MONITORING: 1, TORCH_NCCL_HEARTBEAT_TIMEOUT_SEC: 480, TORCH_NCCL_TRACE_BUFFER_SIZE: 0, TORCH_NCCL_COORD_CHECK_MILSEC: 1000, TORCH_NCCL_NAN_CHECK: 0, TORCH_NCCL_CUDA_EVENT_CACHE: 0, TORCH_NCCL_LOG_CPP_STACK_ON_UNCLEAN_SHUTDOWN: 1 2026-02-02 14:14:17.850 | INFO | hyvideo.inference:from_pretrained:189 - Building model... 2026-02-02 14:14:17.850 | INFO | hyvideo.inference:from_pretrained:189 - Building model... 2026-02-02 14:14:18.469 | INFO | hyvideo.inference:load_state_dict:340 - Loading torch model ckpts/hunyuan-video-t2v-720p/transformers/mp_rank_00_model_states.pt... /workspace/cicd/HunyuanVideo-t2v/hyvideo/inference.py:341: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. state_dict = torch.load(model_path, map_location=lambda storage, loc: storage) 2026-02-02 14:14:18.566 | INFO | hyvideo.inference:load_state_dict:340 - Loading torch model ckpts/hunyuan-video-t2v-720p/transformers/mp_rank_00_model_states.pt... /workspace/cicd/HunyuanVideo-t2v/hyvideo/inference.py:341: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. state_dict = torch.load(model_path, map_location=lambda storage, loc: storage) 2026-02-02 14:14:32.421 | INFO | hyvideo.vae:load_vae:29 - Loading 3D VAE model (884-16c-hy) from: ./ckpts/hunyuan-video-t2v-720p/vae /workspace/cicd/HunyuanVideo-t2v/hyvideo/vae/__init__.py:39: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature. ckpt = torch.load(vae_ckpt, map_location=vae.device) 2026-02-02 14:14:34.427 | INFO | hyvideo.vae:load_vae:55 - VAE to dtype: torch.float16 2026-02-02 14:14:34.558 | INFO | hyvideo.text_encoder:load_text_encoder:28 - Loading text encoder model (llm) from: ./ckpts/text_encoder Using the `SDPA` attention implementation on multi-gpu setup with ROCM may lead to performance issues due to the FA backend. Disabling it to use alternative backends. Loading checkpoint shards: 0%| | 0/4 [00:00