Upload LLaVA-Next-3D/17652629.out with huggingface_hub

LLaVA-Next-3D/17652629.out

@@ -0,0 +1,2023 @@
+PREV_STAGE_CHECKPOINT: data/model_2/LLaVA-Video-7B-Qwen2
+MID_RUN_NAME: try
+SH-IDC1-10-140-0-209
+22006
+localhost:
+[2025-04-02 01:07:52,632] [INFO] [real_accelerator.py:203:get_accelerator] Setting ds_accelerator to cuda (auto detect)
+ [WARNING]  async_io requires the dev libaio .so object and headers but these were not found.
+ [WARNING]  async_io: please install the libaio-devel package with yum
+ [WARNING]  If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found.
+ [WARNING]  Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH
+ [WARNING]  sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.1
+ [WARNING]  using untested triton version (2.1.0), only 1.0.0 is known to be compatible
+[2025-04-02 01:07:55,852] [INFO] [runner.py:568:main] cmd = srun -n 8 --nodes 1 --export=ALL,PYTHONPATH=/mnt/petrelfs/wangzehan/LLaVA-Next-3D:/var/spool/slurmd/job17652629/..:,NCCL_SOCKET_IFNAME=eth0,NCCL_IB_HCA=mlx5_0 /mnt/petrelfs/wangzehan/miniconda3/envs/llava-3d/bin/python -u llava/train/train_3d_slurm.py --deepspeed scripts/zero3.json --model_name_or_path data/model_2/LLaVA-Video-7B-Qwen2 --version qwen_1_5 --data_path scripts/3d/train/only_grounding.yaml --image_folder data --video_folder data --embodiedscan_folder data/embodiedscan_2/ --mm_tunable_parts mm_vision_tower,mm_mlp_adapter,mm_language_model --mm_vision_tower_lr 2e-6 --vision_tower google/siglip-so400m-patch14-384 --mm_projector_type mlp2x_gelu --mm_vision_select_layer -2 --mm_use_im_start_end False --mm_use_im_patch_token False --image_aspect_ratio anyres_max_9 --image_grid_pinpoints (1x1),...,(6x6) --mm_patch_merge_type spatial_unpad --add_frame_id True --bf16 True --run_name try --output_dir ./ckpt/try --num_train_epochs 1 --per_device_train_batch_size 1 --per_device_eval_batch_size 4 --gradient_accumulation_steps 2 --evaluation_strategy no --save_strategy steps --save_steps 500 --save_total_limit 1 --learning_rate 1e-5 --weight_decay 0. --warmup_ratio 0.03 --lr_scheduler_type cosine --logging_steps 1 --tf32 True --model_max_length 32768 --gradient_checkpointing True --dataloader_num_workers 1 --lazy_preprocess True --torch_compile True --torch_compile_backend inductor --dataloader_drop_last True --mm_newline_position grid --add_spatial_instruction True --force_sample True --mm_spatial_pool_stride 2 --world_position_embedding_type avg-discrete-sin3d --object_feature_type patch14-pe --ground_head_type infonce --group_by_task_length True --frame_sampling_strategy uniform --frames_upbound 32
+[2025-04-02 01:08:02,172] [INFO] [real_accelerator.py:203:get_accelerator] Setting ds_accelerator to cuda (auto detect)
+[2025-04-02 01:08:02,174] [INFO] [real_accelerator.py:203:get_accelerator] Setting ds_accelerator to cuda (auto detect)
+[2025-04-02 01:08:02,188] [INFO] [real_accelerator.py:203:get_accelerator] Setting ds_accelerator to cuda (auto detect)
+[2025-04-02 01:08:02,194] [INFO] [real_accelerator.py:203:get_accelerator] Setting ds_accelerator to cuda (auto detect)
+[2025-04-02 01:08:02,202] [INFO] [real_accelerator.py:203:get_accelerator] Setting ds_accelerator to cuda (auto detect)
+[2025-04-02 01:08:02,203] [INFO] [real_accelerator.py:203:get_accelerator] Setting ds_accelerator to cuda (auto detect)
+[2025-04-02 01:08:02,207] [INFO] [real_accelerator.py:203:get_accelerator] Setting ds_accelerator to cuda (auto detect)
+[2025-04-02 01:08:02,208] [INFO] [real_accelerator.py:203:get_accelerator] Setting ds_accelerator to cuda (auto detect)
+ [WARNING]  async_io requires the dev libaio .so object and headers but these were not found.
+ [WARNING]  async_io requires the dev libaio .so object and headers but these were not found.
+ [WARNING]  async_io requires the dev libaio .so object and headers but these were not found.
+ [WARNING]  async_io requires the dev libaio .so object and headers but these were not found.
+ [WARNING]  async_io requires the dev libaio .so object and headers but these were not found.
+ [WARNING]  async_io requires the dev libaio .so object and headers but these were not found.
+ [WARNING]  async_io requires the dev libaio .so object and headers but these were not found.
+ [WARNING]  async_io requires the dev libaio .so object and headers but these were not found.
+ [WARNING]  async_io: please install the libaio-devel package with yum
+ [WARNING]  async_io: please install the libaio-devel package with yum
+ [WARNING]  async_io: please install the libaio-devel package with yum
+ [WARNING]  If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found.
+ [WARNING]  If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found.
+ [WARNING]  If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found.
+ [WARNING]  Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH
+ [WARNING]  async_io: please install the libaio-devel package with yum
+ [WARNING]  Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH
+ [WARNING]  Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH
+ [WARNING]  If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found.
+ [WARNING]  Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH
+ [WARNING]  async_io: please install the libaio-devel package with yum
+ [WARNING]  async_io: please install the libaio-devel package with yum
+ [WARNING]  async_io: please install the libaio-devel package with yum
+ [WARNING]  If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found.
+ [WARNING]  If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found.
+ [WARNING]  Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH
+ [WARNING]  Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH
+ [WARNING]  If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found.
+ [WARNING]  Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH
+ [WARNING]  async_io: please install the libaio-devel package with yum
+ [WARNING]  If libaio is already installed (perhaps from source), try setting the CFLAGS and LDFLAGS environment variables to where it can be found.
+ [WARNING]  Please specify the CUTLASS repo directory as environment variable $CUTLASS_PATH
+ [WARNING]  sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.1
+ [WARNING]  using untested triton version (2.1.0), only 1.0.0 is known to be compatible
+ [WARNING]  sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.1
+ [WARNING]  using untested triton version (2.1.0), only 1.0.0 is known to be compatible
+ [WARNING]  sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.1
+ [WARNING]  using untested triton version (2.1.0), only 1.0.0 is known to be compatible
+ [WARNING]  sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.1
+ [WARNING]  sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.1
+ [WARNING]  using untested triton version (2.1.0), only 1.0.0 is known to be compatible
+ [WARNING]  using untested triton version (2.1.0), only 1.0.0 is known to be compatible
+ [WARNING]  sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.1
+ [WARNING]  sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.1
+ [WARNING]  sparse_attn requires a torch version >= 1.5 and < 2.0 but detected 2.1
+ [WARNING]  using untested triton version (2.1.0), only 1.0.0 is known to be compatible
+ [WARNING]  using untested triton version (2.1.0), only 1.0.0 is known to be compatible
+ [WARNING]  using untested triton version (2.1.0), only 1.0.0 is known to be compatible
+[2025-04-02 01:08:12,372] [INFO] [comm.py:637:init_distributed] cdb=None
+[2025-04-02 01:08:12,372] [INFO] [comm.py:637:init_distributed] cdb=None
+[2025-04-02 01:08:12,372] [INFO] [comm.py:668:init_distributed] Initializing TorchBackend in DeepSpeed with backend nccl
+[2025-04-02 01:08:12,372] [INFO] [comm.py:637:init_distributed] cdb=None
+[2025-04-02 01:08:12,372] [INFO] [comm.py:637:init_distributed] cdb=None
+[2025-04-02 01:08:12,372] [INFO] [comm.py:637:init_distributed] cdb=None
+[2025-04-02 01:08:12,503] [INFO] [comm.py:637:init_distributed] cdb=None
+[2025-04-02 01:08:12,631] [INFO] [comm.py:637:init_distributed] cdb=None
+[2025-04-02 01:08:12,631] [INFO] [comm.py:637:init_distributed] cdb=None
+Rank 0:  Overwriting config with {'use_pos_skipping': False, 'pos_skipping_range': 4096, 'mm_spatial_pool_mode': 'bilinear', 'object_feature_type': 'patch14-pe', 'world_position_embedding_type': 'avg-discrete-sin3d', 'voxel_size': 0.1, 'min_xyz_range': [-15, -15, -5], 'max_xyz_range': [15, 15, 5], 'ground_head_type': 'infonce', 'ground_head_hidden_size': 512, 'ground_loss_scale': 1, 'ground_head_temperature': 0.07}
+{'config': LlavaConfig {
+  "_name_or_path": "data/model_2/LLaVA-Video-7B-Qwen2",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "force_sample": true,
+  "ground_head_hidden_size": 512,
+  "ground_head_temperature": 0.07,
+  "ground_head_type": "infonce",
+  "ground_loss_scale": 1,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "max_xyz_range": [
+    15,
+    15,
+    5
+  ],
+  "min_xyz_range": [
+    -15,
+    -15,
+    -5
+  ],
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "/mnt/hwfile/openmmlab/zhuchenming/vision_encoder/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": 2e-06,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "object_feature_type": "patch14-pe",
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.1",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null,
+  "voxel_size": 0.1,
+  "world_position_embedding_type": "avg-discrete-sin3d"
+}
+}
+ModelArguments(model_name_or_path='data/model_2/LLaVA-Video-7B-Qwen2', model_class_name=None, mm_tunable_parts='mm_vision_tower,mm_mlp_adapter,mm_language_model', version='qwen_1_5', freeze_backbone=False, tune_mm_mlp_adapter=False, tune_mm_vision_resampler=False, vision_tower='google/siglip-so400m-patch14-384', vision_tower_pretrained=None, unfreeze_mm_vision_tower=False, unfreeze_language_model=False, mm_vision_select_layer=-2, pretrain_mm_mlp_adapter=None, mm_projector_type='mlp2x_gelu', mm_use_im_start_end=False, mm_use_im_patch_token=False, mm_patch_merge_type='spatial_unpad', mm_vision_select_feature='patch', mm_resampler_type=None, mm_mask_drop_mode='fixed', mm_mask_drop_skip_percentage=0.0, mm_mask_drop_ratio=0.25, mm_mask_drop_ratio_upper=None, mm_mask_drop_ratio_lower=None, mm_spatial_pool_stride=2, mm_spatial_pool_mode='bilinear', mm_spatial_pool_out_channels=None, mm_perceiver_depth=3, mm_perceiver_latents=32, mm_perceiver_ff_mult=4, mm_perceiver_pretrained=None, mm_qformer_depth=3, mm_qformer_latents=32, mm_qformer_pretrained=None, rope_scaling_factor=None, rope_scaling_type=None, s2=False, s2_scales='336,672,1008', use_pos_skipping=False, pos_skipping_range=4096, mm_newline_position='grid', delay_load=True, add_frame_id=True, add_faster_video=False, faster_token_stride=10, world_position_embedding_type='avg-discrete-sin3d', object_feature_type='patch14-pe', ground_head_type='infonce', ground_head_hidden_size=512, ground_loss_scale=1, ground_head_temperature=0.07)
+{'config': LlavaConfig {
+  "_name_or_path": "data/model_2/LLaVA-Video-7B-Qwen2",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "force_sample": true,
+  "ground_head_hidden_size": 512,
+  "ground_head_temperature": 0.07,
+  "ground_head_type": "infonce",
+  "ground_loss_scale": 1,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "max_xyz_range": [
+    15,
+    15,
+    5
+  ],
+  "min_xyz_range": [
+    -15,
+    -15,
+    -5
+  ],
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "/mnt/hwfile/openmmlab/zhuchenming/vision_encoder/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": 2e-06,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "object_feature_type": "patch14-pe",
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.1",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null,
+  "voxel_size": 0.1,
+  "world_position_embedding_type": "avg-discrete-sin3d"
+}
+}
+ModelArguments(model_name_or_path='data/model_2/LLaVA-Video-7B-Qwen2', model_class_name=None, mm_tunable_parts='mm_vision_tower,mm_mlp_adapter,mm_language_model', version='qwen_1_5', freeze_backbone=False, tune_mm_mlp_adapter=False, tune_mm_vision_resampler=False, vision_tower='google/siglip-so400m-patch14-384', vision_tower_pretrained=None, unfreeze_mm_vision_tower=False, unfreeze_language_model=False, mm_vision_select_layer=-2, pretrain_mm_mlp_adapter=None, mm_projector_type='mlp2x_gelu', mm_use_im_start_end=False, mm_use_im_patch_token=False, mm_patch_merge_type='spatial_unpad', mm_vision_select_feature='patch', mm_resampler_type=None, mm_mask_drop_mode='fixed', mm_mask_drop_skip_percentage=0.0, mm_mask_drop_ratio=0.25, mm_mask_drop_ratio_upper=None, mm_mask_drop_ratio_lower=None, mm_spatial_pool_stride=2, mm_spatial_pool_mode='bilinear', mm_spatial_pool_out_channels=None, mm_perceiver_depth=3, mm_perceiver_latents=32, mm_perceiver_ff_mult=4, mm_perceiver_pretrained=None, mm_qformer_depth=3, mm_qformer_latents=32, mm_qformer_pretrained=None, rope_scaling_factor=None, rope_scaling_type=None, s2=False, s2_scales='336,672,1008', use_pos_skipping=False, pos_skipping_range=4096, mm_newline_position='grid', delay_load=True, add_frame_id=True, add_faster_video=False, faster_token_stride=10, world_position_embedding_type='avg-discrete-sin3d', object_feature_type='patch14-pe', ground_head_type='infonce', ground_head_hidden_size=512, ground_loss_scale=1, ground_head_temperature=0.07)
+{'config': LlavaConfig {
+  "_name_or_path": "data/model_2/LLaVA-Video-7B-Qwen2",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "force_sample": true,
+  "ground_head_hidden_size": 512,
+  "ground_head_temperature": 0.07,
+  "ground_head_type": "infonce",
+  "ground_loss_scale": 1,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "max_xyz_range": [
+    15,
+    15,
+    5
+  ],
+  "min_xyz_range": [
+    -15,
+    -15,
+    -5
+  ],
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "/mnt/hwfile/openmmlab/zhuchenming/vision_encoder/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": 2e-06,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "object_feature_type": "patch14-pe",
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.1",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null,
+  "voxel_size": 0.1,
+  "world_position_embedding_type": "avg-discrete-sin3d"
+}
+}
+ModelArguments(model_name_or_path='data/model_2/LLaVA-Video-7B-Qwen2', model_class_name=None, mm_tunable_parts='mm_vision_tower,mm_mlp_adapter,mm_language_model', version='qwen_1_5', freeze_backbone=False, tune_mm_mlp_adapter=False, tune_mm_vision_resampler=False, vision_tower='google/siglip-so400m-patch14-384', vision_tower_pretrained=None, unfreeze_mm_vision_tower=False, unfreeze_language_model=False, mm_vision_select_layer=-2, pretrain_mm_mlp_adapter=None, mm_projector_type='mlp2x_gelu', mm_use_im_start_end=False, mm_use_im_patch_token=False, mm_patch_merge_type='spatial_unpad', mm_vision_select_feature='patch', mm_resampler_type=None, mm_mask_drop_mode='fixed', mm_mask_drop_skip_percentage=0.0, mm_mask_drop_ratio=0.25, mm_mask_drop_ratio_upper=None, mm_mask_drop_ratio_lower=None, mm_spatial_pool_stride=2, mm_spatial_pool_mode='bilinear', mm_spatial_pool_out_channels=None, mm_perceiver_depth=3, mm_perceiver_latents=32, mm_perceiver_ff_mult=4, mm_perceiver_pretrained=None, mm_qformer_depth=3, mm_qformer_latents=32, mm_qformer_pretrained=None, rope_scaling_factor=None, rope_scaling_type=None, s2=False, s2_scales='336,672,1008', use_pos_skipping=False, pos_skipping_range=4096, mm_newline_position='grid', delay_load=True, add_frame_id=True, add_faster_video=False, faster_token_stride=10, world_position_embedding_type='avg-discrete-sin3d', object_feature_type='patch14-pe', ground_head_type='infonce', ground_head_hidden_size=512, ground_loss_scale=1, ground_head_temperature=0.07)
+{'config': LlavaConfig {
+  "_name_or_path": "data/model_2/LLaVA-Video-7B-Qwen2",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "force_sample": true,
+  "ground_head_hidden_size": 512,
+  "ground_head_temperature": 0.07,
+  "ground_head_type": "infonce",
+  "ground_loss_scale": 1,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "max_xyz_range": [
+    15,
+    15,
+    5
+  ],
+  "min_xyz_range": [
+    -15,
+    -15,
+    -5
+  ],
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "/mnt/hwfile/openmmlab/zhuchenming/vision_encoder/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": 2e-06,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "object_feature_type": "patch14-pe",
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.1",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null,
+  "voxel_size": 0.1,
+  "world_position_embedding_type": "avg-discrete-sin3d"
+}
+}
+ModelArguments(model_name_or_path='data/model_2/LLaVA-Video-7B-Qwen2', model_class_name=None, mm_tunable_parts='mm_vision_tower,mm_mlp_adapter,mm_language_model', version='qwen_1_5', freeze_backbone=False, tune_mm_mlp_adapter=False, tune_mm_vision_resampler=False, vision_tower='google/siglip-so400m-patch14-384', vision_tower_pretrained=None, unfreeze_mm_vision_tower=False, unfreeze_language_model=False, mm_vision_select_layer=-2, pretrain_mm_mlp_adapter=None, mm_projector_type='mlp2x_gelu', mm_use_im_start_end=False, mm_use_im_patch_token=False, mm_patch_merge_type='spatial_unpad', mm_vision_select_feature='patch', mm_resampler_type=None, mm_mask_drop_mode='fixed', mm_mask_drop_skip_percentage=0.0, mm_mask_drop_ratio=0.25, mm_mask_drop_ratio_upper=None, mm_mask_drop_ratio_lower=None, mm_spatial_pool_stride=2, mm_spatial_pool_mode='bilinear', mm_spatial_pool_out_channels=None, mm_perceiver_depth=3, mm_perceiver_latents=32, mm_perceiver_ff_mult=4, mm_perceiver_pretrained=None, mm_qformer_depth=3, mm_qformer_latents=32, mm_qformer_pretrained=None, rope_scaling_factor=None, rope_scaling_type=None, s2=False, s2_scales='336,672,1008', use_pos_skipping=False, pos_skipping_range=4096, mm_newline_position='grid', delay_load=True, add_frame_id=True, add_faster_video=False, faster_token_stride=10, world_position_embedding_type='avg-discrete-sin3d', object_feature_type='patch14-pe', ground_head_type='infonce', ground_head_hidden_size=512, ground_loss_scale=1, ground_head_temperature=0.07)
+{'config': LlavaConfig {
+  "_name_or_path": "data/model_2/LLaVA-Video-7B-Qwen2",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "force_sample": true,
+  "ground_head_hidden_size": 512,
+  "ground_head_temperature": 0.07,
+  "ground_head_type": "infonce",
+  "ground_loss_scale": 1,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "max_xyz_range": [
+    15,
+    15,
+    5
+  ],
+  "min_xyz_range": [
+    -15,
+    -15,
+    -5
+  ],
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "/mnt/hwfile/openmmlab/zhuchenming/vision_encoder/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": 2e-06,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "object_feature_type": "patch14-pe",
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.1",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null,
+  "voxel_size": 0.1,
+  "world_position_embedding_type": "avg-discrete-sin3d"
+}
+}
+ModelArguments(model_name_or_path='data/model_2/LLaVA-Video-7B-Qwen2', model_class_name=None, mm_tunable_parts='mm_vision_tower,mm_mlp_adapter,mm_language_model', version='qwen_1_5', freeze_backbone=False, tune_mm_mlp_adapter=False, tune_mm_vision_resampler=False, vision_tower='google/siglip-so400m-patch14-384', vision_tower_pretrained=None, unfreeze_mm_vision_tower=False, unfreeze_language_model=False, mm_vision_select_layer=-2, pretrain_mm_mlp_adapter=None, mm_projector_type='mlp2x_gelu', mm_use_im_start_end=False, mm_use_im_patch_token=False, mm_patch_merge_type='spatial_unpad', mm_vision_select_feature='patch', mm_resampler_type=None, mm_mask_drop_mode='fixed', mm_mask_drop_skip_percentage=0.0, mm_mask_drop_ratio=0.25, mm_mask_drop_ratio_upper=None, mm_mask_drop_ratio_lower=None, mm_spatial_pool_stride=2, mm_spatial_pool_mode='bilinear', mm_spatial_pool_out_channels=None, mm_perceiver_depth=3, mm_perceiver_latents=32, mm_perceiver_ff_mult=4, mm_perceiver_pretrained=None, mm_qformer_depth=3, mm_qformer_latents=32, mm_qformer_pretrained=None, rope_scaling_factor=None, rope_scaling_type=None, s2=False, s2_scales='336,672,1008', use_pos_skipping=False, pos_skipping_range=4096, mm_newline_position='grid', delay_load=True, add_frame_id=True, add_faster_video=False, faster_token_stride=10, world_position_embedding_type='avg-discrete-sin3d', object_feature_type='patch14-pe', ground_head_type='infonce', ground_head_hidden_size=512, ground_loss_scale=1, ground_head_temperature=0.07)
+{'config': LlavaConfig {
+  "_name_or_path": "data/model_2/LLaVA-Video-7B-Qwen2",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "force_sample": true,
+  "ground_head_hidden_size": 512,
+  "ground_head_temperature": 0.07,
+  "ground_head_type": "infonce",
+  "ground_loss_scale": 1,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "max_xyz_range": [
+    15,
+    15,
+    5
+  ],
+  "min_xyz_range": [
+    -15,
+    -15,
+    -5
+  ],
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "/mnt/hwfile/openmmlab/zhuchenming/vision_encoder/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": 2e-06,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "object_feature_type": "patch14-pe",
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.1",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null,
+  "voxel_size": 0.1,
+  "world_position_embedding_type": "avg-discrete-sin3d"
+}
+}
+ModelArguments(model_name_or_path='data/model_2/LLaVA-Video-7B-Qwen2', model_class_name=None, mm_tunable_parts='mm_vision_tower,mm_mlp_adapter,mm_language_model', version='qwen_1_5', freeze_backbone=False, tune_mm_mlp_adapter=False, tune_mm_vision_resampler=False, vision_tower='google/siglip-so400m-patch14-384', vision_tower_pretrained=None, unfreeze_mm_vision_tower=False, unfreeze_language_model=False, mm_vision_select_layer=-2, pretrain_mm_mlp_adapter=None, mm_projector_type='mlp2x_gelu', mm_use_im_start_end=False, mm_use_im_patch_token=False, mm_patch_merge_type='spatial_unpad', mm_vision_select_feature='patch', mm_resampler_type=None, mm_mask_drop_mode='fixed', mm_mask_drop_skip_percentage=0.0, mm_mask_drop_ratio=0.25, mm_mask_drop_ratio_upper=None, mm_mask_drop_ratio_lower=None, mm_spatial_pool_stride=2, mm_spatial_pool_mode='bilinear', mm_spatial_pool_out_channels=None, mm_perceiver_depth=3, mm_perceiver_latents=32, mm_perceiver_ff_mult=4, mm_perceiver_pretrained=None, mm_qformer_depth=3, mm_qformer_latents=32, mm_qformer_pretrained=None, rope_scaling_factor=None, rope_scaling_type=None, s2=False, s2_scales='336,672,1008', use_pos_skipping=False, pos_skipping_range=4096, mm_newline_position='grid', delay_load=True, add_frame_id=True, add_faster_video=False, faster_token_stride=10, world_position_embedding_type='avg-discrete-sin3d', object_feature_type='patch14-pe', ground_head_type='infonce', ground_head_hidden_size=512, ground_loss_scale=1, ground_head_temperature=0.07)
+{'config': LlavaConfig {
+  "_name_or_path": "data/model_2/LLaVA-Video-7B-Qwen2",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "force_sample": true,
+  "ground_head_hidden_size": 512,
+  "ground_head_temperature": 0.07,
+  "ground_head_type": "infonce",
+  "ground_loss_scale": 1,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "max_xyz_range": [
+    15,
+    15,
+    5
+  ],
+  "min_xyz_range": [
+    -15,
+    -15,
+    -5
+  ],
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "/mnt/hwfile/openmmlab/zhuchenming/vision_encoder/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": 2e-06,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "object_feature_type": "patch14-pe",
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.1",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null,
+  "voxel_size": 0.1,
+  "world_position_embedding_type": "avg-discrete-sin3d"
+}
+}
+ModelArguments(model_name_or_path='data/model_2/LLaVA-Video-7B-Qwen2', model_class_name=None, mm_tunable_parts='mm_vision_tower,mm_mlp_adapter,mm_language_model', version='qwen_1_5', freeze_backbone=False, tune_mm_mlp_adapter=False, tune_mm_vision_resampler=False, vision_tower='google/siglip-so400m-patch14-384', vision_tower_pretrained=None, unfreeze_mm_vision_tower=False, unfreeze_language_model=False, mm_vision_select_layer=-2, pretrain_mm_mlp_adapter=None, mm_projector_type='mlp2x_gelu', mm_use_im_start_end=False, mm_use_im_patch_token=False, mm_patch_merge_type='spatial_unpad', mm_vision_select_feature='patch', mm_resampler_type=None, mm_mask_drop_mode='fixed', mm_mask_drop_skip_percentage=0.0, mm_mask_drop_ratio=0.25, mm_mask_drop_ratio_upper=None, mm_mask_drop_ratio_lower=None, mm_spatial_pool_stride=2, mm_spatial_pool_mode='bilinear', mm_spatial_pool_out_channels=None, mm_perceiver_depth=3, mm_perceiver_latents=32, mm_perceiver_ff_mult=4, mm_perceiver_pretrained=None, mm_qformer_depth=3, mm_qformer_latents=32, mm_qformer_pretrained=None, rope_scaling_factor=None, rope_scaling_type=None, s2=False, s2_scales='336,672,1008', use_pos_skipping=False, pos_skipping_range=4096, mm_newline_position='grid', delay_load=True, add_frame_id=True, add_faster_video=False, faster_token_stride=10, world_position_embedding_type='avg-discrete-sin3d', object_feature_type='patch14-pe', ground_head_type='infonce', ground_head_hidden_size=512, ground_loss_scale=1, ground_head_temperature=0.07)
+{'config': LlavaConfig {
+  "_name_or_path": "data/model_2/LLaVA-Video-7B-Qwen2",
+  "add_faster_video": false,
+  "add_time_instruction": true,
+  "architectures": [
+    "LlavaQwenForCausalLM"
+  ],
+  "attention_dropout": 0.0,
+  "bos_token_id": 151643,
+  "eos_token_id": 151645,
+  "force_sample": true,
+  "ground_head_hidden_size": 512,
+  "ground_head_temperature": 0.07,
+  "ground_head_type": "infonce",
+  "ground_loss_scale": 1,
+  "hidden_act": "silu",
+  "hidden_size": 3584,
+  "ignore_index": -100,
+  "image_aspect_ratio": "anyres_max_9",
+  "image_crop_resolution": null,
+  "image_grid_pinpoints": [
+    [
+      384,
+      384
+    ],
+    [
+      384,
+      768
+    ],
+    [
+      384,
+      1152
+    ],
+    [
+      384,
+      1536
+    ],
+    [
+      384,
+      1920
+    ],
+    [
+      384,
+      2304
+    ],
+    [
+      768,
+      384
+    ],
+    [
+      768,
+      768
+    ],
+    [
+      768,
+      1152
+    ],
+    [
+      768,
+      1536
+    ],
+    [
+      768,
+      1920
+    ],
+    [
+      768,
+      2304
+    ],
+    [
+      1152,
+      384
+    ],
+    [
+      1152,
+      768
+    ],
+    [
+      1152,
+      1152
+    ],
+    [
+      1152,
+      1536
+    ],
+    [
+      1152,
+      1920
+    ],
+    [
+      1152,
+      2304
+    ],
+    [
+      1536,
+      384
+    ],
+    [
+      1536,
+      768
+    ],
+    [
+      1536,
+      1152
+    ],
+    [
+      1536,
+      1536
+    ],
+    [
+      1536,
+      1920
+    ],
+    [
+      1536,
+      2304
+    ],
+    [
+      1920,
+      384
+    ],
+    [
+      1920,
+      768
+    ],
+    [
+      1920,
+      1152
+    ],
+    [
+      1920,
+      1536
+    ],
+    [
+      1920,
+      1920
+    ],
+    [
+      1920,
+      2304
+    ],
+    [
+      2304,
+      384
+    ],
+    [
+      2304,
+      768
+    ],
+    [
+      2304,
+      1152
+    ],
+    [
+      2304,
+      1536
+    ],
+    [
+      2304,
+      1920
+    ],
+    [
+      2304,
+      2304
+    ]
+  ],
+  "image_split_resolution": null,
+  "image_token_index": 151646,
+  "initializer_range": 0.02,
+  "intermediate_size": 18944,
+  "max_position_embeddings": 32768,
+  "max_window_layers": 28,
+  "max_xyz_range": [
+    15,
+    15,
+    5
+  ],
+  "min_xyz_range": [
+    -15,
+    -15,
+    -5
+  ],
+  "mm_hidden_size": 1152,
+  "mm_newline_position": "grid",
+  "mm_patch_merge_type": "spatial_unpad",
+  "mm_projector_lr": null,
+  "mm_projector_type": "mlp2x_gelu",
+  "mm_resampler_type": null,
+  "mm_spatial_pool_mode": "bilinear",
+  "mm_tunable_parts": "mm_vision_tower,mm_mlp_adapter,mm_language_model",
+  "mm_use_im_patch_token": false,
+  "mm_use_im_start_end": false,
+  "mm_vision_select_feature": "patch",
+  "mm_vision_select_layer": -2,
+  "mm_vision_tower": "/mnt/hwfile/openmmlab/zhuchenming/vision_encoder/siglip-so400m-patch14-384",
+  "mm_vision_tower_lr": 2e-06,
+  "model_type": "llava",
+  "num_attention_heads": 28,
+  "num_hidden_layers": 28,
+  "num_key_value_heads": 4,
+  "object_feature_type": "patch14-pe",
+  "pos_skipping_range": 4096,
+  "projector_hidden_act": "gelu",
+  "rms_norm_eps": 1e-06,
+  "rope_scaling": null,
+  "rope_theta": 1000000.0,
+  "sliding_window": 131072,
+  "text_config": {
+    "model_type": "llama"
+  },
+  "tie_word_embeddings": false,
+  "tokenizer_model_max_length": 32768,
+  "tokenizer_padding_side": "right",
+  "torch_dtype": "bfloat16",
+  "transformers_version": "4.40.1",
+  "use_cache": true,
+  "use_mm_proj": true,
+  "use_pos_skipping": false,
+  "use_sliding_window": false,
+  "vision_config": {
+    "hidden_size": 1024,
+    "image_size": 336,
+    "intermediate_size": 4096,
+    "model_type": "clip_vision_model",
+    "num_attention_heads": 16,
+    "num_hidden_layers": 24,
+    "patch_size": 14,
+    "projection_dim": 768,
+    "vocab_size": 32000
+  },
+  "vision_feature_layer": -2,
+  "vision_feature_select_strategy": "default",
+  "vision_tower_pretrained": null,
+  "voxel_size": 0.1,
+  "world_position_embedding_type": "avg-discrete-sin3d"
+}
+}
+ModelArguments(model_name_or_path='data/model_2/LLaVA-Video-7B-Qwen2', model_class_name=None, mm_tunable_parts='mm_vision_tower,mm_mlp_adapter,mm_language_model', version='qwen_1_5', freeze_backbone=False, tune_mm_mlp_adapter=False, tune_mm_vision_resampler=False, vision_tower='google/siglip-so400m-patch14-384', vision_tower_pretrained=None, unfreeze_mm_vision_tower=False, unfreeze_language_model=False, mm_vision_select_layer=-2, pretrain_mm_mlp_adapter=None, mm_projector_type='mlp2x_gelu', mm_use_im_start_end=False, mm_use_im_patch_token=False, mm_patch_merge_type='spatial_unpad', mm_vision_select_feature='patch', mm_resampler_type=None, mm_mask_drop_mode='fixed', mm_mask_drop_skip_percentage=0.0, mm_mask_drop_ratio=0.25, mm_mask_drop_ratio_upper=None, mm_mask_drop_ratio_lower=None, mm_spatial_pool_stride=2, mm_spatial_pool_mode='bilinear', mm_spatial_pool_out_channels=None, mm_perceiver_depth=3, mm_perceiver_latents=32, mm_perceiver_ff_mult=4, mm_perceiver_pretrained=None, mm_qformer_depth=3, mm_qformer_latents=32, mm_qformer_pretrained=None, rope_scaling_factor=None, rope_scaling_type=None, s2=False, s2_scales='336,672,1008', use_pos_skipping=False, pos_skipping_range=4096, mm_newline_position='grid', delay_load=True, add_frame_id=True, add_faster_video=False, faster_token_stride=10, world_position_embedding_type='avg-discrete-sin3d', object_feature_type='patch14-pe', ground_head_type='infonce', ground_head_hidden_size=512, ground_loss_scale=1, ground_head_temperature=0.07)
+Rank 0:  Loading vision tower: /mnt/hwfile/openmmlab/zhuchenming/vision_encoder/siglip-so400m-patch14-384
+[2025-04-02 01:08:29,228] [INFO] [partition_parameters.py:345:__exit__] finished initializing model - num_params = 1145, num_elems = 15.73B
+Rank 0:  Prompt version: qwen_1_5
+Rank 0:  /mnt/hwfile/openmmlab/zhuchenming/vision_encoder/siglip-so400m-patch14-384 is already loaded, `load_model` called again, skipping.
+Rank 0:  Using mm_tunable_parts: mm_vision_tower,mm_mlp_adapter,mm_language_model
+Rank 0:  Total parameters: ~8078.90 MB)
+Rank 0:  Trainable parameters: ~8078.90 MB)
+============frame sampling strategy: uniform=============
+============frame sampling strategy: uniform=============
+============frame sampling strategy: uniform=============
+============frame sampling strategy: uniform=============
+============frame sampling strategy: uniform=============
+============frame sampling strategy: uniform=============
+============frame sampling strategy: uniform=============
+============frame sampling strategy: uniform=============
+Rank 0:  Loading data/processed/scanrefer_vg_train_llava_style.json with all sampling strategy
+Rank 0:  Loaded 36665 samples from data/processed/scanrefer_vg_train_llava_style.json
+Rank 0:  Loading data/processed/multi3drefer_train_llava_style.json with all sampling strategy
+Rank 0:  Loaded 43838 samples from data/processed/multi3drefer_train_llava_style.json
+Rank 0:  Loaded 80503 samples from scripts/3d/train/only_grounding.yaml
+Rank 0:  Formatting inputs...Skip in lazy mode
+Rank 0:  Setting NCCL timeout to INF to avoid running errors.
+Parameter Offload: Total persistent parameters: 761248 in 416 params
+{'loss': 0.8253, 'grad_norm': 7.206215437962158, 'learning_rate': 6.622516556291392e-08, 'epoch': 0.0}
+{'loss': 0.839, 'grad_norm': 6.697667683778133, 'learning_rate': 1.3245033112582784e-07, 'epoch': 0.0}
+{'loss': 0.8101, 'grad_norm': 7.566465534749536, 'learning_rate': 1.9867549668874176e-07, 'epoch': 0.0}
+{'loss': 0.8521, 'grad_norm': 7.239312387581277, 'learning_rate': 2.649006622516557e-07, 'epoch': 0.0}
+{'loss': 0.8466, 'grad_norm': 10.250733376067924, 'learning_rate': 3.311258278145696e-07, 'epoch': 0.0}
+{'loss': 0.8853, 'grad_norm': 11.867459772505295, 'learning_rate': 3.973509933774835e-07, 'epoch': 0.0}
+{'loss': 0.8648, 'grad_norm': 9.360830663018339, 'learning_rate': 4.635761589403974e-07, 'epoch': 0.0}
+{'loss': 0.9201, 'grad_norm': 10.1279766832907, 'learning_rate': 5.298013245033113e-07, 'epoch': 0.0}
+{'loss': 0.7776, 'grad_norm': 6.999117411387415, 'learning_rate': 5.960264900662252e-07, 'epoch': 0.0}
+{'loss': 0.8714, 'grad_norm': 7.65101698397561, 'learning_rate': 6.622516556291392e-07, 'epoch': 0.0}
+{'loss': 0.8287, 'grad_norm': 7.826364624523219, 'learning_rate': 7.28476821192053e-07, 'epoch': 0.0}
+{'loss': 0.8802, 'grad_norm': 7.846013587076487, 'learning_rate': 7.94701986754967e-07, 'epoch': 0.0}
+{'loss': 0.7485, 'grad_norm': 6.016615408277487, 'learning_rate': 8.609271523178808e-07, 'epoch': 0.0}
+{'loss': 0.7837, 'grad_norm': 5.383529318043835, 'learning_rate': 9.271523178807948e-07, 'epoch': 0.0}
+{'loss': 0.791, 'grad_norm': 6.747653092149375, 'learning_rate': 9.933774834437087e-07, 'epoch': 0.0}
+{'loss': 0.8047, 'grad_norm': 6.685831215166226, 'learning_rate': 1.0596026490066227e-06, 'epoch': 0.0}
+{'loss': 0.6973, 'grad_norm': 8.035813606112747, 'learning_rate': 1.1258278145695367e-06, 'epoch': 0.0}