Hugging Face's logo

shu4dev
/
Qwen2.5_VL_10K

TensorBoard
Safetensors
qwen2_5_vl
Model card Files
xet
 Metrics Community
shu4dev commited on
Commit
4a8f12d
·
verified ·
1 Parent(s): 50c76e8

Upload folder using huggingface_hub

Browse files
Files changed (35) hide show
  1. .gitattributes +1 -0
  2. added_tokens.json +24 -0
  3. chat_template.jinja +1 -0
  4. config.json +132 -0
  5. generation_config.json +11 -0
  6. logs/rank_0000.log +225 -0
  7. logs/rank_0001.log +174 -0
  8. merges.txt +0 -0
  9. model-00001-of-00007.safetensors +3 -0
  10. model-00002-of-00007.safetensors +3 -0
  11. model-00003-of-00007.safetensors +3 -0
  12. model-00004-of-00007.safetensors +3 -0
  13. model-00005-of-00007.safetensors +3 -0
  14. model-00006-of-00007.safetensors +3 -0
  15. model-00007-of-00007.safetensors +3 -0
  16. model.safetensors.index.json +737 -0
  17. preprocessor_config.json +39 -0
  18. runs/Dec08_10-49-54_192-222-53-232/events.out.tfevents.1765191410.192-222-53-232.5450.0 +3 -0
  19. runs/Dec08_11-06-11_192-222-53-232/events.out.tfevents.1765192372.192-222-53-232.6208.0 +3 -0
  20. runs/Dec08_11-19-20_192-222-53-232/events.out.tfevents.1765193149.192-222-53-232.6783.0 +3 -0
  21. runs/Dec08_11-39-42_192-222-53-232/events.out.tfevents.1765194375.192-222-53-232.7928.0 +3 -0
  22. runs/Dec08_12-06-49_192-222-53-232/events.out.tfevents.1765195741.192-222-53-232.9073.0 +3 -0
  23. special_tokens_map.json +31 -0
  24. telemetry/devices_info.txt +3 -0
  25. telemetry/telemetry_callback_metrics_rank0000.json +10 -0
  26. telemetry/telemetry_callback_rank0000.json +39 -0
  27. telemetry/telemetry_callback_wandb_rank0000.json +5 -0
  28. telemetry/training_config.yaml +245 -0
  29. telemetry/world_size.json +4 -0
  30. tokenizer.json +3 -0
  31. tokenizer_config.json +209 -0
  32. trainer_state.json +0 -0
  33. training_args.bin +3 -0
  34. video_preprocessor_config.json +43 -0
  35. vocab.json +0 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
added_tokens.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "</tool_call>": 151658,
3
+ "<tool_call>": 151657,
4
+ "<|box_end|>": 151649,
5
+ "<|box_start|>": 151648,
6
+ "<|endoftext|>": 151643,
7
+ "<|file_sep|>": 151664,
8
+ "<|fim_middle|>": 151660,
9
+ "<|fim_pad|>": 151662,
10
+ "<|fim_prefix|>": 151659,
11
+ "<|fim_suffix|>": 151661,
12
+ "<|im_end|>": 151645,
13
+ "<|im_start|>": 151644,
14
+ "<|image_pad|>": 151655,
15
+ "<|object_ref_end|>": 151647,
16
+ "<|object_ref_start|>": 151646,
17
+ "<|quad_end|>": 151651,
18
+ "<|quad_start|>": 151650,
19
+ "<|repo_name|>": 151663,
20
+ "<|video_pad|>": 151656,
21
+ "<|vision_end|>": 151653,
22
+ "<|vision_pad|>": 151654,
23
+ "<|vision_start|>": 151652
24
+ }
chat_template.jinja ADDED
@@ -0,0 +1 @@
 
 
1
+ {% set image_count = namespace(value=0) %}{% set video_count = namespace(value=0) %}{%- for message in messages -%}{%- if loop.first and message['role'] != 'system' -%}{{ '<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n' }}{%- endif -%}{{ '<|im_start|>' + message['role'] + '\n' }}{%- if message['content'] is string -%}{{- message['content'] -}}{%- elif message['content'] is iterable -%}{%- for item in message['content'] -%}{%- if item['type'].startswith('image') -%}{%- set image_count.value = image_count.value + 1 -%}{%- if add_vision_id -%}{{ 'Picture ' + image_count.value + ': ' }}{%- endif -%}{{ '<|vision_start|><|image_pad|><|vision_end|>' }}{%- elif item['type'].startswith('video') -%}{%- set video_count.value = video_count.value + 1 -%}{%- if add_vision_id -%}{{ 'Video ' + video_count.value + ': ' }}{%- endif -%}{{ '<|vision_start|><|video_pad|><|vision_end|>' }}{%- elif item['type']=='text' -%}{{- item['text'] if 'text' in item else item['content'] -}}{%- endif -%}{%- endfor -%}{%- endif -%}{{ '<|im_end|>\n' }}{%- endfor -%}{%- if add_generation_prompt -%}{{- '<|im_start|>assistant\n' -}}{%- endif -%}
config.json ADDED
@@ -0,0 +1,132 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen2_5_VLForConditionalGeneration"
4
+ ],
5
+ "attention_dropout": 0.0,
6
+ "bos_token_id": 151643,
7
+ "dtype": "float32",
8
+ "eos_token_id": 151645,
9
+ "hidden_act": "silu",
10
+ "hidden_size": 3584,
11
+ "image_token_id": 151655,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 18944,
14
+ "max_position_embeddings": 128000,
15
+ "max_window_layers": 28,
16
+ "model_type": "qwen2_5_vl",
17
+ "num_attention_heads": 28,
18
+ "num_hidden_layers": 28,
19
+ "num_key_value_heads": 4,
20
+ "rms_norm_eps": 1e-06,
21
+ "rope_scaling": {
22
+ "mrope_section": [
23
+ 16,
24
+ 24,
25
+ 24
26
+ ],
27
+ "rope_type": "default",
28
+ "type": "default"
29
+ },
30
+ "rope_theta": 1000000.0,
31
+ "sliding_window": 32768,
32
+ "text_config": {
33
+ "_name_or_path": "Qwen/Qwen2.5-VL-7B-Instruct",
34
+ "architectures": [
35
+ "Qwen2_5_VLForConditionalGeneration"
36
+ ],
37
+ "attention_dropout": 0.0,
38
+ "dtype": "bfloat16",
39
+ "eos_token_id": 151645,
40
+ "hidden_act": "silu",
41
+ "hidden_size": 3584,
42
+ "initializer_range": 0.02,
43
+ "intermediate_size": 18944,
44
+ "layer_types": [
45
+ "full_attention",
46
+ "full_attention",
47
+ "full_attention",
48
+ "full_attention",
49
+ "full_attention",
50
+ "full_attention",
51
+ "full_attention",
52
+ "full_attention",
53
+ "full_attention",
54
+ "full_attention",
55
+ "full_attention",
56
+ "full_attention",
57
+ "full_attention",
58
+ "full_attention",
59
+ "full_attention",
60
+ "full_attention",
61
+ "full_attention",
62
+ "full_attention",
63
+ "full_attention",
64
+ "full_attention",
65
+ "full_attention",
66
+ "full_attention",
67
+ "full_attention",
68
+ "full_attention",
69
+ "full_attention",
70
+ "full_attention",
71
+ "full_attention",
72
+ "full_attention"
73
+ ],
74
+ "max_position_embeddings": 128000,
75
+ "max_window_layers": 28,
76
+ "model_type": "qwen2_5_vl_text",
77
+ "num_attention_heads": 28,
78
+ "num_hidden_layers": 28,
79
+ "num_key_value_heads": 4,
80
+ "pad_token_id": 151643,
81
+ "rms_norm_eps": 1e-06,
82
+ "rope_scaling": {
83
+ "mrope_section": [
84
+ 16,
85
+ 24,
86
+ 24
87
+ ],
88
+ "rope_type": "default",
89
+ "type": "default"
90
+ },
91
+ "rope_theta": 1000000.0,
92
+ "sliding_window": null,
93
+ "use_cache": false,
94
+ "use_sliding_window": false,
95
+ "vision_token_id": 151654,
96
+ "vocab_size": 152064
97
+ },
98
+ "tie_word_embeddings": false,
99
+ "transformers_version": "4.57.3",
100
+ "use_cache": true,
101
+ "use_sliding_window": false,
102
+ "video_token_id": 151656,
103
+ "vision_config": {
104
+ "depth": 32,
105
+ "dtype": "bfloat16",
106
+ "fullatt_block_indexes": [
107
+ 7,
108
+ 15,
109
+ 23,
110
+ 31
111
+ ],
112
+ "hidden_act": "silu",
113
+ "hidden_size": 1280,
114
+ "in_channels": 3,
115
+ "in_chans": 3,
116
+ "initializer_range": 0.02,
117
+ "intermediate_size": 3420,
118
+ "model_type": "qwen2_5_vl",
119
+ "num_heads": 16,
120
+ "out_hidden_size": 3584,
121
+ "patch_size": 14,
122
+ "spatial_merge_size": 2,
123
+ "spatial_patch_size": 14,
124
+ "temporal_patch_size": 2,
125
+ "tokens_per_second": 2,
126
+ "window_size": 112
127
+ },
128
+ "vision_end_token_id": 151653,
129
+ "vision_start_token_id": 151652,
130
+ "vision_token_id": 151654,
131
+ "vocab_size": 152064
132
+ }
generation_config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_sample": true,
3
+ "eos_token_id": [
4
+ 151645,
5
+ 151643
6
+ ],
7
+ "pad_token_id": 151643,
8
+ "repetition_penalty": 1.05,
9
+ "temperature": 1e-06,
10
+ "transformers_version": "4.57.3"
11
+ }
logs/rank_0000.log ADDED
@@ -0,0 +1,225 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2025-12-08 10:48:16,975][oumi][rank0][pid:5193][MainThread][INFO]][train.py:318] Training config saved to output/qwen2_5_vl_7b_10K/telemetry/training_config.yaml
2
+ [2025-12-08 10:48:17,053][oumi][rank0][pid:5193][MainThread][INFO]][models.py:469] Setting tokenizer to use the 'right' padding side for model 'Qwen/Qwen2.5-VL-7B-Instruct'. The 'right' padding side is configured as the default value for this model type.
3
+ [2025-12-08 10:48:19,589][oumi][rank0][pid:5193][MainThread][INFO]][train.py:348] Set `training.trainer_kwargs.remove_unused_columns=False` for VLM training with TRL_SFT trainer.
4
+ [2025-12-08 10:48:19,615][oumi][rank0][pid:5193][MainThread][WARNING]][data.py:66] Using torchdata preprocessing pipeline. This is currently in beta and may not be stable.
5
+ [2025-12-08 10:48:19,616][oumi][rank0][pid:5193][MainThread][INFO]][base_map_dataset.py:91] Creating map dataset (type: HuggingFaceVisionDataset)... dataset_name: 'shu4dev/DCVLR_10K'
6
+ [2025-12-08 10:49:28,515][oumi][rank0][pid:5450][MainThread][INFO]][train.py:318] Training config saved to output/qwen2_5_vl_7b_10K/telemetry/training_config.yaml
7
+ [2025-12-08 10:49:28,588][oumi][rank0][pid:5450][MainThread][INFO]][models.py:469] Setting tokenizer to use the 'right' padding side for model 'Qwen/Qwen2.5-VL-7B-Instruct'. The 'right' padding side is configured as the default value for this model type.
8
+ [2025-12-08 10:49:30,611][oumi][rank0][pid:5450][MainThread][INFO]][train.py:348] Set `training.trainer_kwargs.remove_unused_columns=False` for VLM training with TRL_SFT trainer.
9
+ [2025-12-08 10:49:30,638][oumi][rank0][pid:5450][MainThread][WARNING]][data.py:66] Using torchdata preprocessing pipeline. This is currently in beta and may not be stable.
10
+ [2025-12-08 10:49:30,639][oumi][rank0][pid:5450][MainThread][INFO]][base_map_dataset.py:91] Creating map dataset (type: HuggingFaceVisionDataset)... dataset_name: 'shu4dev/DCVLR_10K'
11
+ [2025-12-08 10:49:45,107][oumi][rank0][pid:5450][MainThread][INFO]][base_map_dataset.py:487] Dataset Info:
12
+ Split: train
13
+ Version: 0.0.0
14
+ Dataset size: 2334594551
15
+ Download size: 2274721765
16
+ Size: 4609316316 bytes
17
+ Rows: 10000
18
+ Columns: ['path', 'question', 'reasoning_trace']
19
+ [2025-12-08 10:49:46,592][oumi][rank0][pid:5450][MainThread][INFO]][base_map_dataset.py:426] Loaded DataFrame with shape: (10000, 3). Columns:
20
+ path object
21
+ question object
22
+ reasoning_trace object
23
+ dtype: object
24
+ [2025-12-08 10:49:48,765][oumi][rank0][pid:5450][MainThread][INFO]][distributed.py:303] Initialized distributed (True): DeviceRankInfo(world_size=2, rank=0, local_world_size=2, local_rank=0)
25
+ [2025-12-08 10:49:48,767][oumi][rank0][pid:5450][MainThread][INFO]][train.py:449] Set Accelerate environment variables for FSDP: {'ACCELERATE_DYNAMO_BACKEND': 'NO', 'ACCELERATE_DYNAMO_MODE': 'default', 'ACCELERATE_DYNAMO_USE_FULLGRAPH': 'False', 'ACCELERATE_DYNAMO_USE_DYNAMIC': 'False', 'FSDP_CPU_RAM_EFFICIENT_LOADING': 'true', 'FSDP_USE_ORIG_PARAMS': 'true', 'ACCELERATE_USE_FSDP': 'true', 'FSDP_SHARDING_STRATEGY': 'HYBRID_SHARD', 'FSDP_OFFLOAD_PARAMS': 'false', 'ACCELERATE_MIXED_PRECISION': 'bf16', 'FSDP_BACKWARD_PREFETCH': 'BACKWARD_PRE', 'FSDP_FORWARD_PREFETCH': 'true', 'FSDP_STATE_DICT_TYPE': 'FULL_STATE_DICT', 'FSDP_AUTO_WRAP_POLICY': 'SIZE_BASED_WRAP', 'FSDP_MIN_NUM_PARAMS': '100000', 'FSDP_SYNC_MODULE_STATES': 'true', 'FSDP_ACTIVATION_CHECKPOINTING': 'true'}
26
+ [2025-12-08 10:49:48,769][oumi][rank0][pid:5450][MainThread][INFO]][models.py:251] Accelerate FSDP run detected! Setting device_map to None.
27
+ [2025-12-08 10:49:48,770][oumi][rank0][pid:5450][MainThread][INFO]][models.py:260] Building model using device_map: None (DeviceRankInfo(world_size=2, rank=0, local_world_size=2, local_rank=0))...
28
+ [2025-12-08 10:49:48,823][oumi][rank0][pid:5450][MainThread][INFO]][models.py:336] Using model class: <class 'transformers.models.auto.modeling_auto.AutoModelForVision2Seq'> to instantiate model.
29
+ [2025-12-08 10:49:54,069][oumi][rank0][pid:5450][MainThread][INFO]][torch_utils.py:288]
30
+ Model Parameters Summary:
31
+ 🔢 Total parameters: 8,292,166,656
32
+ 🔗 Embedding parameters: 544,997,376
33
+ 🎯 Trainable parameters: 8,292,166,656
34
+ 🔒 Frozen parameters: 0 (0.00%)
35
+
36
+ [2025-12-08 10:49:54,520][oumi][rank0][pid:5450][MainThread][INFO]][torch_profiler_utils.py:164] PROF: Torch Profiler disabled!
37
+ [2025-12-08 10:49:54,522][oumi][rank0][pid:5450][MainThread][WARNING]][callbacks.py:72] MFU logging requires packed datasets. Skipping MFU callbacks.
38
+ [2025-12-08 10:49:54,856][oumi][rank0][pid:5450][MainThread][INFO]][device_utils.py:343] GPU Metrics Before Training: GPU runtime info: NVidiaGpuRuntimeInfo(device_index=0, device_count=2, used_memory_mb=1979.0, temperature=28, fan_speed=None, fan_speeds=None, power_usage_watts=115.278, power_limit_watts=700.0, gpu_utilization=0, memory_utilization=0, performance_state=0, clock_speed_graphics=1980, clock_speed_sm=1980, clock_speed_memory=2619).
39
+ [2025-12-08 10:49:54,990][oumi][rank0][pid:5450][MainThread][INFO]][train.py:558] Training init time: 26.645s
40
+ [2025-12-08 10:49:54,992][oumi][rank0][pid:5450][MainThread][INFO]][train.py:559] Starting training... (TrainerType.TRL_SFT, transformers: 4.57.3)
41
+ [2025-12-08 11:06:02,751][oumi][rank0][pid:6208][MainThread][INFO]][train.py:318] Training config saved to output/qwen2_5_vl_7b_10K/telemetry/training_config.yaml
42
+ [2025-12-08 11:06:02,823][oumi][rank0][pid:6208][MainThread][INFO]][models.py:469] Setting tokenizer to use the 'right' padding side for model 'Qwen/Qwen2.5-VL-7B-Instruct'. The 'right' padding side is configured as the default value for this model type.
43
+ [2025-12-08 11:06:04,794][oumi][rank0][pid:6208][MainThread][INFO]][train.py:348] Set `training.trainer_kwargs.remove_unused_columns=False` for VLM training with TRL_SFT trainer.
44
+ [2025-12-08 11:06:04,818][oumi][rank0][pid:6208][MainThread][WARNING]][data.py:66] Using torchdata preprocessing pipeline. This is currently in beta and may not be stable.
45
+ [2025-12-08 11:06:04,819][oumi][rank0][pid:6208][MainThread][INFO]][base_map_dataset.py:91] Creating map dataset (type: HuggingFaceVisionDataset)... dataset_name: 'shu4dev/DCVLR_10K'
46
+ [2025-12-08 11:06:05,847][oumi][rank0][pid:6208][MainThread][INFO]][base_map_dataset.py:487] Dataset Info:
47
+ Split: train
48
+ Version: 0.0.0
49
+ Dataset size: 2334594551
50
+ Download size: 2274721765
51
+ Size: 4609316316 bytes
52
+ Rows: 10000
53
+ Columns: ['path', 'question', 'reasoning_trace']
54
+ [2025-12-08 11:06:07,436][oumi][rank0][pid:6208][MainThread][INFO]][base_map_dataset.py:426] Loaded DataFrame with shape: (10000, 3). Columns:
55
+ path object
56
+ question object
57
+ reasoning_trace object
58
+ dtype: object
59
+ [2025-12-08 11:06:09,484][oumi][rank0][pid:6208][MainThread][INFO]][distributed.py:303] Initialized distributed (True): DeviceRankInfo(world_size=2, rank=0, local_world_size=2, local_rank=0)
60
+ [2025-12-08 11:06:09,486][oumi][rank0][pid:6208][MainThread][INFO]][train.py:449] Set Accelerate environment variables for FSDP: {'ACCELERATE_DYNAMO_BACKEND': 'NO', 'ACCELERATE_DYNAMO_MODE': 'default', 'ACCELERATE_DYNAMO_USE_FULLGRAPH': 'False', 'ACCELERATE_DYNAMO_USE_DYNAMIC': 'False', 'FSDP_CPU_RAM_EFFICIENT_LOADING': 'true', 'FSDP_USE_ORIG_PARAMS': 'false', 'ACCELERATE_USE_FSDP': 'true', 'FSDP_SHARDING_STRATEGY': 'HYBRID_SHARD', 'FSDP_OFFLOAD_PARAMS': 'false', 'ACCELERATE_MIXED_PRECISION': 'bf16', 'FSDP_BACKWARD_PREFETCH': 'BACKWARD_PRE', 'FSDP_FORWARD_PREFETCH': 'true', 'FSDP_STATE_DICT_TYPE': 'FULL_STATE_DICT', 'FSDP_AUTO_WRAP_POLICY': 'SIZE_BASED_WRAP', 'FSDP_MIN_NUM_PARAMS': '100000', 'FSDP_SYNC_MODULE_STATES': 'true', 'FSDP_ACTIVATION_CHECKPOINTING': 'true'}
61
+ [2025-12-08 11:06:09,488][oumi][rank0][pid:6208][MainThread][INFO]][models.py:251] Accelerate FSDP run detected! Setting device_map to None.
62
+ [2025-12-08 11:06:09,488][oumi][rank0][pid:6208][MainThread][INFO]][models.py:260] Building model using device_map: None (DeviceRankInfo(world_size=2, rank=0, local_world_size=2, local_rank=0))...
63
+ [2025-12-08 11:06:09,543][oumi][rank0][pid:6208][MainThread][INFO]][models.py:336] Using model class: <class 'transformers.models.auto.modeling_auto.AutoModelForVision2Seq'> to instantiate model.
64
+ [2025-12-08 11:06:09,880][oumi][rank0][pid:6208][MainThread][INFO]][train.py:463] Building PEFT model...
65
+ [2025-12-08 11:06:10,724][oumi][rank0][pid:6208][MainThread][INFO]][torch_utils.py:288]
66
+ Model Parameters Summary:
67
+ 🔢 Total parameters: 8,315,961,344
68
+ 🔗 Embedding parameters: 544,997,376
69
+ 🎯 Trainable parameters: 23,794,688
70
+ 🔒 Frozen parameters: 8,292,166,656 (99.71%)
71
+
72
+ [2025-12-08 11:06:11,180][oumi][rank0][pid:6208][MainThread][INFO]][torch_profiler_utils.py:164] PROF: Torch Profiler disabled!
73
+ [2025-12-08 11:06:11,181][oumi][rank0][pid:6208][MainThread][WARNING]][callbacks.py:69] MFU logging is not supported for PEFT. Skipping MFU callbacks.
74
+ [2025-12-08 11:06:11,516][oumi][rank0][pid:6208][MainThread][INFO]][device_utils.py:343] GPU Metrics Before Training: GPU runtime info: NVidiaGpuRuntimeInfo(device_index=0, device_count=2, used_memory_mb=1979.0, temperature=27, fan_speed=None, fan_speeds=None, power_usage_watts=115.054, power_limit_watts=700.0, gpu_utilization=0, memory_utilization=0, performance_state=0, clock_speed_graphics=1980, clock_speed_sm=1980, clock_speed_memory=2619).
75
+ [2025-12-08 11:06:11,659][oumi][rank0][pid:6208][MainThread][INFO]][train.py:558] Training init time: 9.088s
76
+ [2025-12-08 11:06:11,660][oumi][rank0][pid:6208][MainThread][INFO]][train.py:559] Starting training... (TrainerType.TRL_SFT, transformers: 4.57.3)
77
+ [2025-12-08 11:19:11,807][oumi][rank0][pid:6783][MainThread][INFO]][train.py:318] Training config saved to output/qwen2_5_vl_7b_10K/telemetry/training_config.yaml
78
+ [2025-12-08 11:19:11,883][oumi][rank0][pid:6783][MainThread][INFO]][models.py:469] Setting tokenizer to use the 'right' padding side for model 'Qwen/Qwen2.5-VL-7B-Instruct'. The 'right' padding side is configured as the default value for this model type.
79
+ [2025-12-08 11:19:14,076][oumi][rank0][pid:6783][MainThread][INFO]][train.py:348] Set `training.trainer_kwargs.remove_unused_columns=False` for VLM training with TRL_SFT trainer.
80
+ [2025-12-08 11:19:14,102][oumi][rank0][pid:6783][MainThread][WARNING]][data.py:66] Using torchdata preprocessing pipeline. This is currently in beta and may not be stable.
81
+ [2025-12-08 11:19:14,104][oumi][rank0][pid:6783][MainThread][INFO]][base_map_dataset.py:91] Creating map dataset (type: HuggingFaceVisionDataset)... dataset_name: 'shu4dev/DCVLR_10K'
82
+ [2025-12-08 11:19:15,047][oumi][rank0][pid:6783][MainThread][INFO]][base_map_dataset.py:487] Dataset Info:
83
+ Split: train
84
+ Version: 0.0.0
85
+ Dataset size: 2334594551
86
+ Download size: 2274721765
87
+ Size: 4609316316 bytes
88
+ Rows: 10000
89
+ Columns: ['path', 'question', 'reasoning_trace']
90
+ [2025-12-08 11:19:16,606][oumi][rank0][pid:6783][MainThread][INFO]][base_map_dataset.py:426] Loaded DataFrame with shape: (10000, 3). Columns:
91
+ path object
92
+ question object
93
+ reasoning_trace object
94
+ dtype: object
95
+ [2025-12-08 11:19:18,709][oumi][rank0][pid:6783][MainThread][INFO]][distributed.py:303] Initialized distributed (True): DeviceRankInfo(world_size=2, rank=0, local_world_size=2, local_rank=0)
96
+ [2025-12-08 11:19:18,710][oumi][rank0][pid:6783][MainThread][INFO]][train.py:449] Set Accelerate environment variables for FSDP: {'ACCELERATE_DYNAMO_BACKEND': 'NO', 'ACCELERATE_DYNAMO_MODE': 'default', 'ACCELERATE_DYNAMO_USE_FULLGRAPH': 'False', 'ACCELERATE_DYNAMO_USE_DYNAMIC': 'False', 'FSDP_CPU_RAM_EFFICIENT_LOADING': 'true', 'FSDP_USE_ORIG_PARAMS': 'false', 'ACCELERATE_USE_FSDP': 'true', 'FSDP_SHARDING_STRATEGY': 'HYBRID_SHARD', 'FSDP_OFFLOAD_PARAMS': 'false', 'ACCELERATE_MIXED_PRECISION': 'bf16', 'FSDP_BACKWARD_PREFETCH': 'BACKWARD_PRE', 'FSDP_FORWARD_PREFETCH': 'true', 'FSDP_STATE_DICT_TYPE': 'FULL_STATE_DICT', 'FSDP_AUTO_WRAP_POLICY': 'SIZE_BASED_WRAP', 'FSDP_MIN_NUM_PARAMS': '100000', 'FSDP_SYNC_MODULE_STATES': 'true', 'FSDP_ACTIVATION_CHECKPOINTING': 'false'}
97
+ [2025-12-08 11:19:18,712][oumi][rank0][pid:6783][MainThread][INFO]][models.py:251] Accelerate FSDP run detected! Setting device_map to None.
98
+ [2025-12-08 11:19:18,713][oumi][rank0][pid:6783][MainThread][INFO]][models.py:260] Building model using device_map: None (DeviceRankInfo(world_size=2, rank=0, local_world_size=2, local_rank=0))...
99
+ [2025-12-08 11:19:18,767][oumi][rank0][pid:6783][MainThread][INFO]][models.py:336] Using model class: <class 'transformers.models.auto.modeling_auto.AutoModelForVision2Seq'> to instantiate model.
100
+ [2025-12-08 11:19:19,035][oumi][rank0][pid:6783][MainThread][INFO]][train.py:463] Building PEFT model...
101
+ [2025-12-08 11:19:19,885][oumi][rank0][pid:6783][MainThread][INFO]][torch_utils.py:288]
102
+ Model Parameters Summary:
103
+ 🔢 Total parameters: 8,315,961,344
104
+ 🔗 Embedding parameters: 544,997,376
105
+ 🎯 Trainable parameters: 23,794,688
106
+ 🔒 Frozen parameters: 8,292,166,656 (99.71%)
107
+
108
+ [2025-12-08 11:19:20,328][oumi][rank0][pid:6783][MainThread][INFO]][torch_profiler_utils.py:164] PROF: Torch Profiler disabled!
109
+ [2025-12-08 11:19:20,330][oumi][rank0][pid:6783][MainThread][WARNING]][callbacks.py:69] MFU logging is not supported for PEFT. Skipping MFU callbacks.
110
+ [2025-12-08 11:19:20,663][oumi][rank0][pid:6783][MainThread][INFO]][device_utils.py:343] GPU Metrics Before Training: GPU runtime info: NVidiaGpuRuntimeInfo(device_index=0, device_count=2, used_memory_mb=1979.0, temperature=28, fan_speed=None, fan_speeds=None, power_usage_watts=116.483, power_limit_watts=700.0, gpu_utilization=0, memory_utilization=0, performance_state=0, clock_speed_graphics=1980, clock_speed_sm=1980, clock_speed_memory=2619).
111
+ [2025-12-08 11:19:20,804][oumi][rank0][pid:6783][MainThread][INFO]][train.py:558] Training init time: 9.189s
112
+ [2025-12-08 11:19:20,805][oumi][rank0][pid:6783][MainThread][INFO]][train.py:559] Starting training... (TrainerType.TRL_SFT, transformers: 4.57.3)
113
+ [2025-12-08 11:39:34,657][oumi][rank0][pid:7928][MainThread][INFO]][train.py:318] Training config saved to output/qwen2_5_vl_7b_10K/telemetry/training_config.yaml
114
+ [2025-12-08 11:39:34,732][oumi][rank0][pid:7928][MainThread][INFO]][models.py:469] Setting tokenizer to use the 'right' padding side for model 'Qwen/Qwen2.5-VL-7B-Instruct'. The 'right' padding side is configured as the default value for this model type.
115
+ [2025-12-08 11:39:36,926][oumi][rank0][pid:7928][MainThread][INFO]][train.py:348] Set `training.trainer_kwargs.remove_unused_columns=False` for VLM training with TRL_SFT trainer.
116
+ [2025-12-08 11:39:36,951][oumi][rank0][pid:7928][MainThread][WARNING]][data.py:66] Using torchdata preprocessing pipeline. This is currently in beta and may not be stable.
117
+ [2025-12-08 11:39:36,952][oumi][rank0][pid:7928][MainThread][INFO]][base_map_dataset.py:91] Creating map dataset (type: HuggingFaceVisionDataset)... dataset_name: 'shu4dev/DCVLR_10K'
118
+ [2025-12-08 11:39:38,292][oumi][rank0][pid:7928][MainThread][INFO]][base_map_dataset.py:487] Dataset Info:
119
+ Split: train
120
+ Version: 0.0.0
121
+ Dataset size: 2334594551
122
+ Download size: 2274721765
123
+ Size: 4609316316 bytes
124
+ Rows: 10000
125
+ Columns: ['path', 'question', 'reasoning_trace']
126
+ [2025-12-08 11:39:39,850][oumi][rank0][pid:7928][MainThread][INFO]][base_map_dataset.py:426] Loaded DataFrame with shape: (10000, 3). Columns:
127
+ path object
128
+ question object
129
+ reasoning_trace object
130
+ dtype: object
131
+ [2025-12-08 11:39:41,902][oumi][rank0][pid:7928][MainThread][INFO]][distributed.py:303] Initialized distributed (True): DeviceRankInfo(world_size=2, rank=0, local_world_size=2, local_rank=0)
132
+ [2025-12-08 11:39:41,903][oumi][rank0][pid:7928][MainThread][INFO]][train.py:449] Set Accelerate environment variables for FSDP: {'ACCELERATE_DYNAMO_BACKEND': 'NO', 'ACCELERATE_DYNAMO_MODE': 'default', 'ACCELERATE_DYNAMO_USE_FULLGRAPH': 'False', 'ACCELERATE_DYNAMO_USE_DYNAMIC': 'False', 'FSDP_CPU_RAM_EFFICIENT_LOADING': 'true', 'FSDP_USE_ORIG_PARAMS': 'true', 'ACCELERATE_USE_FSDP': 'true', 'FSDP_SHARDING_STRATEGY': 'HYBRID_SHARD', 'FSDP_OFFLOAD_PARAMS': 'false', 'ACCELERATE_MIXED_PRECISION': 'bf16', 'FSDP_BACKWARD_PREFETCH': 'BACKWARD_PRE', 'FSDP_FORWARD_PREFETCH': 'true', 'FSDP_STATE_DICT_TYPE': 'FULL_STATE_DICT', 'FSDP_AUTO_WRAP_POLICY': 'SIZE_BASED_WRAP', 'FSDP_MIN_NUM_PARAMS': '100000', 'FSDP_SYNC_MODULE_STATES': 'true', 'FSDP_ACTIVATION_CHECKPOINTING': 'true'}
133
+ [2025-12-08 11:39:41,905][oumi][rank0][pid:7928][MainThread][INFO]][models.py:251] Accelerate FSDP run detected! Setting device_map to None.
134
+ [2025-12-08 11:39:41,906][oumi][rank0][pid:7928][MainThread][INFO]][models.py:260] Building model using device_map: None (DeviceRankInfo(world_size=2, rank=0, local_world_size=2, local_rank=0))...
135
+ [2025-12-08 11:39:41,961][oumi][rank0][pid:7928][MainThread][INFO]][models.py:336] Using model class: <class 'transformers.models.auto.modeling_auto.AutoModelForVision2Seq'> to instantiate model.
136
+ [2025-12-08 11:39:42,232][oumi][rank0][pid:7928][MainThread][INFO]][torch_utils.py:288]
137
+ Model Parameters Summary:
138
+ 🔢 Total parameters: 8,292,166,656
139
+ 🔗 Embedding parameters: 544,997,376
140
+ 🎯 Trainable parameters: 8,292,166,656
141
+ 🔒 Frozen parameters: 0 (0.00%)
142
+
143
+ [2025-12-08 11:39:42,654][oumi][rank0][pid:7928][MainThread][INFO]][torch_profiler_utils.py:164] PROF: Torch Profiler disabled!
144
+ [2025-12-08 11:39:42,655][oumi][rank0][pid:7928][MainThread][WARNING]][callbacks.py:72] MFU logging requires packed datasets. Skipping MFU callbacks.
145
+ [2025-12-08 11:39:42,998][oumi][rank0][pid:7928][MainThread][INFO]][device_utils.py:343] GPU Metrics Before Training: GPU runtime info: NVidiaGpuRuntimeInfo(device_index=0, device_count=2, used_memory_mb=1979.0, temperature=28, fan_speed=None, fan_speeds=None, power_usage_watts=117.138, power_limit_watts=700.0, gpu_utilization=0, memory_utilization=0, performance_state=0, clock_speed_graphics=1980, clock_speed_sm=1980, clock_speed_memory=2619).
146
+ [2025-12-08 11:39:43,144][oumi][rank0][pid:7928][MainThread][INFO]][train.py:558] Training init time: 8.666s
147
+ [2025-12-08 11:39:43,145][oumi][rank0][pid:7928][MainThread][INFO]][train.py:559] Starting training... (TrainerType.TRL_SFT, transformers: 4.57.3)
148
+ [2025-12-08 12:05:31,985][oumi][rank0][pid:8710][MainThread][INFO]][train.py:318] Training config saved to output/qwen2_5_vl_7b_10K/telemetry/training_config.yaml
149
+ [2025-12-08 12:05:32,058][oumi][rank0][pid:8710][MainThread][INFO]][models.py:469] Setting tokenizer to use the 'right' padding side for model 'Qwen/Qwen2.5-VL-7B-Instruct'. The 'right' padding side is configured as the default value for this model type.
150
+ [2025-12-08 12:05:34,107][oumi][rank0][pid:8710][MainThread][INFO]][train.py:348] Set `training.trainer_kwargs.remove_unused_columns=False` for VLM training with TRL_SFT trainer.
151
+ [2025-12-08 12:05:34,133][oumi][rank0][pid:8710][MainThread][WARNING]][data.py:66] Using torchdata preprocessing pipeline. This is currently in beta and may not be stable.
152
+ [2025-12-08 12:05:34,134][oumi][rank0][pid:8710][MainThread][INFO]][base_map_dataset.py:91] Creating map dataset (type: HuggingFaceVisionDataset)... dataset_name: 'shu4dev/DCVLR_10K'
153
+ [2025-12-08 12:05:47,929][oumi][rank0][pid:8710][MainThread][INFO]][base_map_dataset.py:487] Dataset Info:
154
+ Split: train
155
+ Version: 0.0.0
156
+ Dataset size: 1944466265
157
+ Download size: 1910842881
158
+ Size: 3855309146 bytes
159
+ Rows: 10000
160
+ Columns: ['image', 'problem', 'solution']
161
+ [2025-12-08 12:05:49,283][oumi][rank0][pid:8710][MainThread][INFO]][base_map_dataset.py:426] Loaded DataFrame with shape: (10000, 3). Columns:
162
+ image object
163
+ problem object
164
+ solution object
165
+ dtype: object
166
+ [2025-12-08 12:05:51,449][oumi][rank0][pid:8710][MainThread][INFO]][distributed.py:303] Initialized distributed (True): DeviceRankInfo(world_size=2, rank=0, local_world_size=2, local_rank=0)
167
+ [2025-12-08 12:05:51,451][oumi][rank0][pid:8710][MainThread][INFO]][train.py:449] Set Accelerate environment variables for FSDP: {'ACCELERATE_DYNAMO_BACKEND': 'NO', 'ACCELERATE_DYNAMO_MODE': 'default', 'ACCELERATE_DYNAMO_USE_FULLGRAPH': 'False', 'ACCELERATE_DYNAMO_USE_DYNAMIC': 'False', 'FSDP_CPU_RAM_EFFICIENT_LOADING': 'true', 'FSDP_USE_ORIG_PARAMS': 'true', 'ACCELERATE_USE_FSDP': 'true', 'FSDP_SHARDING_STRATEGY': 'HYBRID_SHARD', 'FSDP_OFFLOAD_PARAMS': 'false', 'ACCELERATE_MIXED_PRECISION': 'bf16', 'FSDP_BACKWARD_PREFETCH': 'BACKWARD_PRE', 'FSDP_FORWARD_PREFETCH': 'true', 'FSDP_STATE_DICT_TYPE': 'FULL_STATE_DICT', 'FSDP_AUTO_WRAP_POLICY': 'SIZE_BASED_WRAP', 'FSDP_MIN_NUM_PARAMS': '100000', 'FSDP_SYNC_MODULE_STATES': 'true', 'FSDP_ACTIVATION_CHECKPOINTING': 'true'}
168
+ [2025-12-08 12:05:51,452][oumi][rank0][pid:8710][MainThread][INFO]][models.py:251] Accelerate FSDP run detected! Setting device_map to None.
169
+ [2025-12-08 12:05:51,453][oumi][rank0][pid:8710][MainThread][INFO]][models.py:260] Building model using device_map: None (DeviceRankInfo(world_size=2, rank=0, local_world_size=2, local_rank=0))...
170
+ [2025-12-08 12:05:51,508][oumi][rank0][pid:8710][MainThread][INFO]][models.py:336] Using model class: <class 'transformers.models.auto.modeling_auto.AutoModelForVision2Seq'> to instantiate model.
171
+ [2025-12-08 12:05:51,778][oumi][rank0][pid:8710][MainThread][INFO]][torch_utils.py:288]
172
+ Model Parameters Summary:
173
+ 🔢 Total parameters: 8,292,166,656
174
+ 🔗 Embedding parameters: 544,997,376
175
+ 🎯 Trainable parameters: 8,292,166,656
176
+ 🔒 Frozen parameters: 0 (0.00%)
177
+
178
+ [2025-12-08 12:06:42,507][oumi][rank0][pid:9073][MainThread][INFO]][train.py:318] Training config saved to output/qwen2_5_vl_7b_10K/telemetry/training_config.yaml
179
+ [2025-12-08 12:06:42,582][oumi][rank0][pid:9073][MainThread][INFO]][models.py:469] Setting tokenizer to use the 'right' padding side for model 'Qwen/Qwen2.5-VL-7B-Instruct'. The 'right' padding side is configured as the default value for this model type.
180
+ [2025-12-08 12:06:44,620][oumi][rank0][pid:9073][MainThread][INFO]][train.py:348] Set `training.trainer_kwargs.remove_unused_columns=False` for VLM training with TRL_SFT trainer.
181
+ [2025-12-08 12:06:44,645][oumi][rank0][pid:9073][MainThread][WARNING]][data.py:66] Using torchdata preprocessing pipeline. This is currently in beta and may not be stable.
182
+ [2025-12-08 12:06:44,646][oumi][rank0][pid:9073][MainThread][INFO]][base_map_dataset.py:91] Creating map dataset (type: HuggingFaceVisionDataset)... dataset_name: 'shu4dev/DCVLR_10K'
183
+ [2025-12-08 12:06:45,581][oumi][rank0][pid:9073][MainThread][INFO]][base_map_dataset.py:487] Dataset Info:
184
+ Split: train
185
+ Version: 0.0.0
186
+ Dataset size: 1944466265
187
+ Download size: 1910842881
188
+ Size: 3855309146 bytes
189
+ Rows: 10000
190
+ Columns: ['image', 'problem', 'solution']
191
+ [2025-12-08 12:06:46,873][oumi][rank0][pid:9073][MainThread][INFO]][base_map_dataset.py:426] Loaded DataFrame with shape: (10000, 3). Columns:
192
+ image object
193
+ problem object
194
+ solution object
195
+ dtype: object
196
+ [2025-12-08 12:06:48,690][oumi][rank0][pid:9073][MainThread][INFO]][distributed.py:303] Initialized distributed (True): DeviceRankInfo(world_size=2, rank=0, local_world_size=2, local_rank=0)
197
+ [2025-12-08 12:06:48,691][oumi][rank0][pid:9073][MainThread][INFO]][train.py:449] Set Accelerate environment variables for FSDP: {'ACCELERATE_DYNAMO_BACKEND': 'NO', 'ACCELERATE_DYNAMO_MODE': 'default', 'ACCELERATE_DYNAMO_USE_FULLGRAPH': 'False', 'ACCELERATE_DYNAMO_USE_DYNAMIC': 'False', 'FSDP_CPU_RAM_EFFICIENT_LOADING': 'true', 'FSDP_USE_ORIG_PARAMS': 'true', 'ACCELERATE_USE_FSDP': 'true', 'FSDP_SHARDING_STRATEGY': 'HYBRID_SHARD', 'FSDP_OFFLOAD_PARAMS': 'false', 'ACCELERATE_MIXED_PRECISION': 'bf16', 'FSDP_BACKWARD_PREFETCH': 'BACKWARD_PRE', 'FSDP_FORWARD_PREFETCH': 'true', 'FSDP_STATE_DICT_TYPE': 'FULL_STATE_DICT', 'FSDP_AUTO_WRAP_POLICY': 'SIZE_BASED_WRAP', 'FSDP_MIN_NUM_PARAMS': '100000', 'FSDP_SYNC_MODULE_STATES': 'true', 'FSDP_ACTIVATION_CHECKPOINTING': 'true'}
198
+ [2025-12-08 12:06:48,693][oumi][rank0][pid:9073][MainThread][INFO]][models.py:251] Accelerate FSDP run detected! Setting device_map to None.
199
+ [2025-12-08 12:06:48,694][oumi][rank0][pid:9073][MainThread][INFO]][models.py:260] Building model using device_map: None (DeviceRankInfo(world_size=2, rank=0, local_world_size=2, local_rank=0))...
200
+ [2025-12-08 12:06:48,747][oumi][rank0][pid:9073][MainThread][INFO]][models.py:336] Using model class: <class 'transformers.models.auto.modeling_auto.AutoModelForVision2Seq'> to instantiate model.
201
+ [2025-12-08 12:06:49,017][oumi][rank0][pid:9073][MainThread][INFO]][torch_utils.py:288]
202
+ Model Parameters Summary:
203
+ 🔢 Total parameters: 8,292,166,656
204
+ 🔗 Embedding parameters: 544,997,376
205
+ 🎯 Trainable parameters: 8,292,166,656
206
+ 🔒 Frozen parameters: 0 (0.00%)
207
+
208
+ [2025-12-08 12:06:49,402][oumi][rank0][pid:9073][MainThread][INFO]][torch_profiler_utils.py:164] PROF: Torch Profiler disabled!
209
+ [2025-12-08 12:06:49,403][oumi][rank0][pid:9073][MainThread][WARNING]][callbacks.py:72] MFU logging requires packed datasets. Skipping MFU callbacks.
210
+ [2025-12-08 12:06:49,742][oumi][rank0][pid:9073][MainThread][INFO]][device_utils.py:343] GPU Metrics Before Training: GPU runtime info: NVidiaGpuRuntimeInfo(device_index=0, device_count=2, used_memory_mb=1979.0, temperature=28, fan_speed=None, fan_speeds=None, power_usage_watts=115.258, power_limit_watts=700.0, gpu_utilization=0, memory_utilization=0, performance_state=0, clock_speed_graphics=1980, clock_speed_sm=1980, clock_speed_memory=2619).
211
+ [2025-12-08 12:06:49,885][oumi][rank0][pid:9073][MainThread][INFO]][train.py:558] Training init time: 7.557s
212
+ [2025-12-08 12:06:49,886][oumi][rank0][pid:9073][MainThread][INFO]][train.py:559] Starting training... (TrainerType.TRL_SFT, transformers: 4.57.3)
213
+ [2025-12-08 14:26:24,671][oumi][rank0][pid:9073][MainThread][INFO]][device_utils.py:343] On epoch end: GPU runtime info: NVidiaGpuRuntimeInfo(device_index=0, device_count=2, used_memory_mb=80345.0, temperature=35, fan_speed=None, fan_speeds=None, power_usage_watts=266.502, power_limit_watts=700.0, gpu_utilization=31, memory_utilization=11, performance_state=0, clock_speed_graphics=1980, clock_speed_sm=1980, clock_speed_memory=2619).
214
+ [2025-12-08 14:26:24,706][oumi][rank0][pid:9073][MainThread][INFO]][telemetry_callback.py:242] Saving telemetry callback summary to output/qwen2_5_vl_7b_10K/telemetry/telemetry_callback_rank0000.json...
215
+ [2025-12-08 14:26:24,762][oumi][rank0][pid:9073][MainThread][INFO]][train.py:566] Training is Complete.
216
+ [2025-12-08 14:26:24,765][oumi][rank0][pid:9073][MainThread][INFO]][device_utils.py:343] GPU Metrics After Training: GPU runtime info: NVidiaGpuRuntimeInfo(device_index=0, device_count=2, used_memory_mb=80345.0, temperature=35, fan_speed=None, fan_speeds=None, power_usage_watts=277.043, power_limit_watts=700.0, gpu_utilization=31, memory_utilization=11, performance_state=0, clock_speed_graphics=1980, clock_speed_sm=1980, clock_speed_memory=2619).
217
+ [2025-12-08 14:26:24,767][oumi][rank0][pid:9073][MainThread][INFO]][torch_utils.py:135] Peak GPU memory usage: 67.34 GB
218
+ [2025-12-08 14:26:24,768][oumi][rank0][pid:9073][MainThread][INFO]][train.py:573] Saving final state...
219
+ [2025-12-08 14:26:24,803][oumi][rank0][pid:9073][MainThread][INFO]][train.py:578] Saving final model...
220
+ [2025-12-08 14:26:24,804][oumi][rank0][pid:9073][MainThread][INFO]][hf_trainer.py:145] Saving FULL_STATE_DICT for final model checkpoint.
221
+ [2025-12-08 14:27:56,755][oumi][rank0][pid:9073][MainThread][INFO]][hf_trainer.py:152] Model has been saved at output/qwen2_5_vl_7b_10K
222
+ [2025-12-08 14:27:57,195][oumi][rank0][pid:9073][MainThread][INFO]][hf_trainer.py:156] Processor config has been saved at output/qwen2_5_vl_7b_10K
223
+ [2025-12-08 14:27:57,793][oumi][rank0][pid:9073][MainThread][INFO]][train.py:230]
224
+
225
+ » We're always looking for feedback. What's one thing we can improve? https://oumi.ai/feedback
logs/rank_0001.log ADDED
@@ -0,0 +1,174 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [2025-12-08 10:48:16,879][oumi][rank1][pid:5194][MainThread][INFO]][models.py:469] Setting tokenizer to use the 'right' padding side for model 'Qwen/Qwen2.5-VL-7B-Instruct'. The 'right' padding side is configured as the default value for this model type.
2
+ [2025-12-08 10:48:19,545][oumi][rank1][pid:5194][MainThread][INFO]][train.py:348] Set `training.trainer_kwargs.remove_unused_columns=False` for VLM training with TRL_SFT trainer.
3
+ [2025-12-08 10:48:19,570][oumi][rank1][pid:5194][MainThread][WARNING]][data.py:66] Using torchdata preprocessing pipeline. This is currently in beta and may not be stable.
4
+ [2025-12-08 10:48:19,571][oumi][rank1][pid:5194][MainThread][INFO]][base_map_dataset.py:91] Creating map dataset (type: HuggingFaceVisionDataset)... dataset_name: 'shu4dev/DCVLR_10K'
5
+ [2025-12-08 10:49:28,443][oumi][rank1][pid:5451][MainThread][INFO]][models.py:469] Setting tokenizer to use the 'right' padding side for model 'Qwen/Qwen2.5-VL-7B-Instruct'. The 'right' padding side is configured as the default value for this model type.
6
+ [2025-12-08 10:49:31,021][oumi][rank1][pid:5451][MainThread][INFO]][train.py:348] Set `training.trainer_kwargs.remove_unused_columns=False` for VLM training with TRL_SFT trainer.
7
+ [2025-12-08 10:49:31,047][oumi][rank1][pid:5451][MainThread][WARNING]][data.py:66] Using torchdata preprocessing pipeline. This is currently in beta and may not be stable.
8
+ [2025-12-08 10:49:31,048][oumi][rank1][pid:5451][MainThread][INFO]][base_map_dataset.py:91] Creating map dataset (type: HuggingFaceVisionDataset)... dataset_name: 'shu4dev/DCVLR_10K'
9
+ [2025-12-08 10:49:45,140][oumi][rank1][pid:5451][MainThread][INFO]][base_map_dataset.py:487] Dataset Info:
10
+ Split: train
11
+ Version: 0.0.0
12
+ Dataset size: 2334594551
13
+ Download size: 2274721765
14
+ Size: 4609316316 bytes
15
+ Rows: 10000
16
+ Columns: ['path', 'question', 'reasoning_trace']
17
+ [2025-12-08 10:49:46,621][oumi][rank1][pid:5451][MainThread][INFO]][base_map_dataset.py:426] Loaded DataFrame with shape: (10000, 3). Columns:
18
+ path object
19
+ question object
20
+ reasoning_trace object
21
+ dtype: object
22
+ [2025-12-08 10:49:48,765][oumi][rank1][pid:5451][MainThread][INFO]][distributed.py:303] Initialized distributed (True): DeviceRankInfo(world_size=2, rank=1, local_world_size=2, local_rank=1)
23
+ [2025-12-08 10:49:48,765][oumi][rank1][pid:5451][MainThread][INFO]][train.py:449] Set Accelerate environment variables for FSDP: {'ACCELERATE_DYNAMO_BACKEND': 'NO', 'ACCELERATE_DYNAMO_MODE': 'default', 'ACCELERATE_DYNAMO_USE_FULLGRAPH': 'False', 'ACCELERATE_DYNAMO_USE_DYNAMIC': 'False', 'FSDP_CPU_RAM_EFFICIENT_LOADING': 'true', 'FSDP_USE_ORIG_PARAMS': 'true', 'ACCELERATE_USE_FSDP': 'true', 'FSDP_SHARDING_STRATEGY': 'HYBRID_SHARD', 'FSDP_OFFLOAD_PARAMS': 'false', 'ACCELERATE_MIXED_PRECISION': 'bf16', 'FSDP_BACKWARD_PREFETCH': 'BACKWARD_PRE', 'FSDP_FORWARD_PREFETCH': 'true', 'FSDP_STATE_DICT_TYPE': 'FULL_STATE_DICT', 'FSDP_AUTO_WRAP_POLICY': 'SIZE_BASED_WRAP', 'FSDP_MIN_NUM_PARAMS': '100000', 'FSDP_SYNC_MODULE_STATES': 'true', 'FSDP_ACTIVATION_CHECKPOINTING': 'true'}
24
+ [2025-12-08 10:49:48,765][oumi][rank1][pid:5451][MainThread][INFO]][models.py:251] Accelerate FSDP run detected! Setting device_map to None.
25
+ [2025-12-08 10:49:48,765][oumi][rank1][pid:5451][MainThread][INFO]][models.py:260] Building model using device_map: None (DeviceRankInfo(world_size=2, rank=1, local_world_size=2, local_rank=1))...
26
+ [2025-12-08 10:49:48,818][oumi][rank1][pid:5451][MainThread][INFO]][models.py:336] Using model class: <class 'transformers.models.auto.modeling_auto.AutoModelForVision2Seq'> to instantiate model.
27
+ [2025-12-08 10:49:54,510][oumi][rank1][pid:5451][MainThread][INFO]][torch_profiler_utils.py:164] PROF: Torch Profiler disabled!
28
+ [2025-12-08 10:49:54,510][oumi][rank1][pid:5451][MainThread][WARNING]][callbacks.py:72] MFU logging requires packed datasets. Skipping MFU callbacks.
29
+ [2025-12-08 10:49:54,847][oumi][rank1][pid:5451][MainThread][INFO]][device_utils.py:343] GPU Metrics Before Training: GPU runtime info: NVidiaGpuRuntimeInfo(device_index=0, device_count=2, used_memory_mb=1979.0, temperature=28, fan_speed=None, fan_speeds=None, power_usage_watts=115.278, power_limit_watts=700.0, gpu_utilization=0, memory_utilization=0, performance_state=0, clock_speed_graphics=1980, clock_speed_sm=1980, clock_speed_memory=2619).
30
+ [2025-12-08 10:49:54,990][oumi][rank1][pid:5451][MainThread][INFO]][train.py:558] Training init time: 26.623s
31
+ [2025-12-08 10:49:54,990][oumi][rank1][pid:5451][MainThread][INFO]][train.py:559] Starting training... (TrainerType.TRL_SFT, transformers: 4.57.3)
32
+ [2025-12-08 11:06:02,690][oumi][rank1][pid:6209][MainThread][INFO]][models.py:469] Setting tokenizer to use the 'right' padding side for model 'Qwen/Qwen2.5-VL-7B-Instruct'. The 'right' padding side is configured as the default value for this model type.
33
+ [2025-12-08 11:06:04,795][oumi][rank1][pid:6209][MainThread][INFO]][train.py:348] Set `training.trainer_kwargs.remove_unused_columns=False` for VLM training with TRL_SFT trainer.
34
+ [2025-12-08 11:06:04,817][oumi][rank1][pid:6209][MainThread][WARNING]][data.py:66] Using torchdata preprocessing pipeline. This is currently in beta and may not be stable.
35
+ [2025-12-08 11:06:04,818][oumi][rank1][pid:6209][MainThread][INFO]][base_map_dataset.py:91] Creating map dataset (type: HuggingFaceVisionDataset)... dataset_name: 'shu4dev/DCVLR_10K'
36
+ [2025-12-08 11:06:05,958][oumi][rank1][pid:6209][MainThread][INFO]][base_map_dataset.py:487] Dataset Info:
37
+ Split: train
38
+ Version: 0.0.0
39
+ Dataset size: 2334594551
40
+ Download size: 2274721765
41
+ Size: 4609316316 bytes
42
+ Rows: 10000
43
+ Columns: ['path', 'question', 'reasoning_trace']
44
+ [2025-12-08 11:06:07,495][oumi][rank1][pid:6209][MainThread][INFO]][base_map_dataset.py:426] Loaded DataFrame with shape: (10000, 3). Columns:
45
+ path object
46
+ question object
47
+ reasoning_trace object
48
+ dtype: object
49
+ [2025-12-08 11:06:09,483][oumi][rank1][pid:6209][MainThread][INFO]][distributed.py:303] Initialized distributed (True): DeviceRankInfo(world_size=2, rank=1, local_world_size=2, local_rank=1)
50
+ [2025-12-08 11:06:09,483][oumi][rank1][pid:6209][MainThread][INFO]][train.py:449] Set Accelerate environment variables for FSDP: {'ACCELERATE_DYNAMO_BACKEND': 'NO', 'ACCELERATE_DYNAMO_MODE': 'default', 'ACCELERATE_DYNAMO_USE_FULLGRAPH': 'False', 'ACCELERATE_DYNAMO_USE_DYNAMIC': 'False', 'FSDP_CPU_RAM_EFFICIENT_LOADING': 'true', 'FSDP_USE_ORIG_PARAMS': 'false', 'ACCELERATE_USE_FSDP': 'true', 'FSDP_SHARDING_STRATEGY': 'HYBRID_SHARD', 'FSDP_OFFLOAD_PARAMS': 'false', 'ACCELERATE_MIXED_PRECISION': 'bf16', 'FSDP_BACKWARD_PREFETCH': 'BACKWARD_PRE', 'FSDP_FORWARD_PREFETCH': 'true', 'FSDP_STATE_DICT_TYPE': 'FULL_STATE_DICT', 'FSDP_AUTO_WRAP_POLICY': 'SIZE_BASED_WRAP', 'FSDP_MIN_NUM_PARAMS': '100000', 'FSDP_SYNC_MODULE_STATES': 'true', 'FSDP_ACTIVATION_CHECKPOINTING': 'true'}
51
+ [2025-12-08 11:06:09,483][oumi][rank1][pid:6209][MainThread][INFO]][models.py:251] Accelerate FSDP run detected! Setting device_map to None.
52
+ [2025-12-08 11:06:09,483][oumi][rank1][pid:6209][MainThread][INFO]][models.py:260] Building model using device_map: None (DeviceRankInfo(world_size=2, rank=1, local_world_size=2, local_rank=1))...
53
+ [2025-12-08 11:06:09,537][oumi][rank1][pid:6209][MainThread][INFO]][models.py:336] Using model class: <class 'transformers.models.auto.modeling_auto.AutoModelForVision2Seq'> to instantiate model.
54
+ [2025-12-08 11:06:09,763][oumi][rank1][pid:6209][MainThread][INFO]][train.py:463] Building PEFT model...
55
+ [2025-12-08 11:06:11,018][oumi][rank1][pid:6209][MainThread][INFO]][torch_profiler_utils.py:164] PROF: Torch Profiler disabled!
56
+ [2025-12-08 11:06:11,019][oumi][rank1][pid:6209][MainThread][WARNING]][callbacks.py:69] MFU logging is not supported for PEFT. Skipping MFU callbacks.
57
+ [2025-12-08 11:06:11,358][oumi][rank1][pid:6209][MainThread][INFO]][device_utils.py:343] GPU Metrics Before Training: GPU runtime info: NVidiaGpuRuntimeInfo(device_index=0, device_count=2, used_memory_mb=1979.0, temperature=27, fan_speed=None, fan_speeds=None, power_usage_watts=115.055, power_limit_watts=700.0, gpu_utilization=0, memory_utilization=0, performance_state=0, clock_speed_graphics=1980, clock_speed_sm=1980, clock_speed_memory=2619).
58
+ [2025-12-08 11:06:11,659][oumi][rank1][pid:6209][MainThread][INFO]][train.py:558] Training init time: 9.066s
59
+ [2025-12-08 11:06:11,659][oumi][rank1][pid:6209][MainThread][INFO]][train.py:559] Starting training... (TrainerType.TRL_SFT, transformers: 4.57.3)
60
+ [2025-12-08 11:19:11,684][oumi][rank1][pid:6784][MainThread][INFO]][models.py:469] Setting tokenizer to use the 'right' padding side for model 'Qwen/Qwen2.5-VL-7B-Instruct'. The 'right' padding side is configured as the default value for this model type.
61
+ [2025-12-08 11:19:13,892][oumi][rank1][pid:6784][MainThread][INFO]][train.py:348] Set `training.trainer_kwargs.remove_unused_columns=False` for VLM training with TRL_SFT trainer.
62
+ [2025-12-08 11:19:13,915][oumi][rank1][pid:6784][MainThread][WARNING]][data.py:66] Using torchdata preprocessing pipeline. This is currently in beta and may not be stable.
63
+ [2025-12-08 11:19:13,915][oumi][rank1][pid:6784][MainThread][INFO]][base_map_dataset.py:91] Creating map dataset (type: HuggingFaceVisionDataset)... dataset_name: 'shu4dev/DCVLR_10K'
64
+ [2025-12-08 11:19:14,959][oumi][rank1][pid:6784][MainThread][INFO]][base_map_dataset.py:487] Dataset Info:
65
+ Split: train
66
+ Version: 0.0.0
67
+ Dataset size: 2334594551
68
+ Download size: 2274721765
69
+ Size: 4609316316 bytes
70
+ Rows: 10000
71
+ Columns: ['path', 'question', 'reasoning_trace']
72
+ [2025-12-08 11:19:16,471][oumi][rank1][pid:6784][MainThread][INFO]][base_map_dataset.py:426] Loaded DataFrame with shape: (10000, 3). Columns:
73
+ path object
74
+ question object
75
+ reasoning_trace object
76
+ dtype: object
77
+ [2025-12-08 11:19:18,709][oumi][rank1][pid:6784][MainThread][INFO]][distributed.py:303] Initialized distributed (True): DeviceRankInfo(world_size=2, rank=1, local_world_size=2, local_rank=1)
78
+ [2025-12-08 11:19:18,709][oumi][rank1][pid:6784][MainThread][INFO]][train.py:449] Set Accelerate environment variables for FSDP: {'ACCELERATE_DYNAMO_BACKEND': 'NO', 'ACCELERATE_DYNAMO_MODE': 'default', 'ACCELERATE_DYNAMO_USE_FULLGRAPH': 'False', 'ACCELERATE_DYNAMO_USE_DYNAMIC': 'False', 'FSDP_CPU_RAM_EFFICIENT_LOADING': 'true', 'FSDP_USE_ORIG_PARAMS': 'false', 'ACCELERATE_USE_FSDP': 'true', 'FSDP_SHARDING_STRATEGY': 'HYBRID_SHARD', 'FSDP_OFFLOAD_PARAMS': 'false', 'ACCELERATE_MIXED_PRECISION': 'bf16', 'FSDP_BACKWARD_PREFETCH': 'BACKWARD_PRE', 'FSDP_FORWARD_PREFETCH': 'true', 'FSDP_STATE_DICT_TYPE': 'FULL_STATE_DICT', 'FSDP_AUTO_WRAP_POLICY': 'SIZE_BASED_WRAP', 'FSDP_MIN_NUM_PARAMS': '100000', 'FSDP_SYNC_MODULE_STATES': 'true', 'FSDP_ACTIVATION_CHECKPOINTING': 'false'}
79
+ [2025-12-08 11:19:18,709][oumi][rank1][pid:6784][MainThread][INFO]][models.py:251] Accelerate FSDP run detected! Setting device_map to None.
80
+ [2025-12-08 11:19:18,709][oumi][rank1][pid:6784][MainThread][INFO]][models.py:260] Building model using device_map: None (DeviceRankInfo(world_size=2, rank=1, local_world_size=2, local_rank=1))...
81
+ [2025-12-08 11:19:18,763][oumi][rank1][pid:6784][MainThread][INFO]][models.py:336] Using model class: <class 'transformers.models.auto.modeling_auto.AutoModelForVision2Seq'> to instantiate model.
82
+ [2025-12-08 11:19:18,990][oumi][rank1][pid:6784][MainThread][INFO]][train.py:463] Building PEFT model...
83
+ [2025-12-08 11:19:20,294][oumi][rank1][pid:6784][MainThread][INFO]][torch_profiler_utils.py:164] PROF: Torch Profiler disabled!
84
+ [2025-12-08 11:19:20,294][oumi][rank1][pid:6784][MainThread][WARNING]][callbacks.py:69] MFU logging is not supported for PEFT. Skipping MFU callbacks.
85
+ [2025-12-08 11:19:20,633][oumi][rank1][pid:6784][MainThread][INFO]][device_utils.py:343] GPU Metrics Before Training: GPU runtime info: NVidiaGpuRuntimeInfo(device_index=0, device_count=2, used_memory_mb=1979.0, temperature=28, fan_speed=None, fan_speeds=None, power_usage_watts=116.483, power_limit_watts=700.0, gpu_utilization=0, memory_utilization=0, performance_state=0, clock_speed_graphics=1980, clock_speed_sm=1980, clock_speed_memory=2619).
86
+ [2025-12-08 11:19:20,804][oumi][rank1][pid:6784][MainThread][INFO]][train.py:558] Training init time: 9.198s
87
+ [2025-12-08 11:19:20,804][oumi][rank1][pid:6784][MainThread][INFO]][train.py:559] Starting training... (TrainerType.TRL_SFT, transformers: 4.57.3)
88
+ [2025-12-08 11:39:34,520][oumi][rank1][pid:7929][MainThread][INFO]][models.py:469] Setting tokenizer to use the 'right' padding side for model 'Qwen/Qwen2.5-VL-7B-Instruct'. The 'right' padding side is configured as the default value for this model type.
89
+ [2025-12-08 11:39:36,497][oumi][rank1][pid:7929][MainThread][INFO]][train.py:348] Set `training.trainer_kwargs.remove_unused_columns=False` for VLM training with TRL_SFT trainer.
90
+ [2025-12-08 11:39:36,521][oumi][rank1][pid:7929][MainThread][WARNING]][data.py:66] Using torchdata preprocessing pipeline. This is currently in beta and may not be stable.
91
+ [2025-12-08 11:39:36,521][oumi][rank1][pid:7929][MainThread][INFO]][base_map_dataset.py:91] Creating map dataset (type: HuggingFaceVisionDataset)... dataset_name: 'shu4dev/DCVLR_10K'
92
+ [2025-12-08 11:39:38,401][oumi][rank1][pid:7929][MainThread][INFO]][base_map_dataset.py:487] Dataset Info:
93
+ Split: train
94
+ Version: 0.0.0
95
+ Dataset size: 2334594551
96
+ Download size: 2274721765
97
+ Size: 4609316316 bytes
98
+ Rows: 10000
99
+ Columns: ['path', 'question', 'reasoning_trace']
100
+ [2025-12-08 11:39:39,888][oumi][rank1][pid:7929][MainThread][INFO]][base_map_dataset.py:426] Loaded DataFrame with shape: (10000, 3). Columns:
101
+ path object
102
+ question object
103
+ reasoning_trace object
104
+ dtype: object
105
+ [2025-12-08 11:39:41,901][oumi][rank1][pid:7929][MainThread][INFO]][distributed.py:303] Initialized distributed (True): DeviceRankInfo(world_size=2, rank=1, local_world_size=2, local_rank=1)
106
+ [2025-12-08 11:39:41,901][oumi][rank1][pid:7929][MainThread][INFO]][train.py:449] Set Accelerate environment variables for FSDP: {'ACCELERATE_DYNAMO_BACKEND': 'NO', 'ACCELERATE_DYNAMO_MODE': 'default', 'ACCELERATE_DYNAMO_USE_FULLGRAPH': 'False', 'ACCELERATE_DYNAMO_USE_DYNAMIC': 'False', 'FSDP_CPU_RAM_EFFICIENT_LOADING': 'true', 'FSDP_USE_ORIG_PARAMS': 'true', 'ACCELERATE_USE_FSDP': 'true', 'FSDP_SHARDING_STRATEGY': 'HYBRID_SHARD', 'FSDP_OFFLOAD_PARAMS': 'false', 'ACCELERATE_MIXED_PRECISION': 'bf16', 'FSDP_BACKWARD_PREFETCH': 'BACKWARD_PRE', 'FSDP_FORWARD_PREFETCH': 'true', 'FSDP_STATE_DICT_TYPE': 'FULL_STATE_DICT', 'FSDP_AUTO_WRAP_POLICY': 'SIZE_BASED_WRAP', 'FSDP_MIN_NUM_PARAMS': '100000', 'FSDP_SYNC_MODULE_STATES': 'true', 'FSDP_ACTIVATION_CHECKPOINTING': 'true'}
107
+ [2025-12-08 11:39:41,902][oumi][rank1][pid:7929][MainThread][INFO]][models.py:251] Accelerate FSDP run detected! Setting device_map to None.
108
+ [2025-12-08 11:39:41,902][oumi][rank1][pid:7929][MainThread][INFO]][models.py:260] Building model using device_map: None (DeviceRankInfo(world_size=2, rank=1, local_world_size=2, local_rank=1))...
109
+ [2025-12-08 11:39:41,957][oumi][rank1][pid:7929][MainThread][INFO]][models.py:336] Using model class: <class 'transformers.models.auto.modeling_auto.AutoModelForVision2Seq'> to instantiate model.
110
+ [2025-12-08 11:39:42,601][oumi][rank1][pid:7929][MainThread][INFO]][torch_profiler_utils.py:164] PROF: Torch Profiler disabled!
111
+ [2025-12-08 11:39:42,602][oumi][rank1][pid:7929][MainThread][WARNING]][callbacks.py:72] MFU logging requires packed datasets. Skipping MFU callbacks.
112
+ [2025-12-08 11:39:42,942][oumi][rank1][pid:7929][MainThread][INFO]][device_utils.py:343] GPU Metrics Before Training: GPU runtime info: NVidiaGpuRuntimeInfo(device_index=0, device_count=2, used_memory_mb=1979.0, temperature=28, fan_speed=None, fan_speeds=None, power_usage_watts=117.138, power_limit_watts=700.0, gpu_utilization=0, memory_utilization=0, performance_state=0, clock_speed_graphics=1980, clock_speed_sm=1980, clock_speed_memory=2619).
113
+ [2025-12-08 11:39:43,144][oumi][rank1][pid:7929][MainThread][INFO]][train.py:558] Training init time: 8.700s
114
+ [2025-12-08 11:39:43,144][oumi][rank1][pid:7929][MainThread][INFO]][train.py:559] Starting training... (TrainerType.TRL_SFT, transformers: 4.57.3)
115
+ [2025-12-08 12:05:31,888][oumi][rank1][pid:8711][MainThread][INFO]][models.py:469] Setting tokenizer to use the 'right' padding side for model 'Qwen/Qwen2.5-VL-7B-Instruct'. The 'right' padding side is configured as the default value for this model type.
116
+ [2025-12-08 12:05:34,165][oumi][rank1][pid:8711][MainThread][INFO]][train.py:348] Set `training.trainer_kwargs.remove_unused_columns=False` for VLM training with TRL_SFT trainer.
117
+ [2025-12-08 12:05:34,190][oumi][rank1][pid:8711][MainThread][WARNING]][data.py:66] Using torchdata preprocessing pipeline. This is currently in beta and may not be stable.
118
+ [2025-12-08 12:05:34,191][oumi][rank1][pid:8711][MainThread][INFO]][base_map_dataset.py:91] Creating map dataset (type: HuggingFaceVisionDataset)... dataset_name: 'shu4dev/DCVLR_10K'
119
+ [2025-12-08 12:05:47,891][oumi][rank1][pid:8711][MainThread][INFO]][base_map_dataset.py:487] Dataset Info:
120
+ Split: train
121
+ Version: 0.0.0
122
+ Dataset size: 1944466265
123
+ Download size: 1910842881
124
+ Size: 3855309146 bytes
125
+ Rows: 10000
126
+ Columns: ['image', 'problem', 'solution']
127
+ [2025-12-08 12:05:49,246][oumi][rank1][pid:8711][MainThread][INFO]][base_map_dataset.py:426] Loaded DataFrame with shape: (10000, 3). Columns:
128
+ image object
129
+ problem object
130
+ solution object
131
+ dtype: object
132
+ [2025-12-08 12:05:51,448][oumi][rank1][pid:8711][MainThread][INFO]][distributed.py:303] Initialized distributed (True): DeviceRankInfo(world_size=2, rank=1, local_world_size=2, local_rank=1)
133
+ [2025-12-08 12:05:51,449][oumi][rank1][pid:8711][MainThread][INFO]][train.py:449] Set Accelerate environment variables for FSDP: {'ACCELERATE_DYNAMO_BACKEND': 'NO', 'ACCELERATE_DYNAMO_MODE': 'default', 'ACCELERATE_DYNAMO_USE_FULLGRAPH': 'False', 'ACCELERATE_DYNAMO_USE_DYNAMIC': 'False', 'FSDP_CPU_RAM_EFFICIENT_LOADING': 'true', 'FSDP_USE_ORIG_PARAMS': 'true', 'ACCELERATE_USE_FSDP': 'true', 'FSDP_SHARDING_STRATEGY': 'HYBRID_SHARD', 'FSDP_OFFLOAD_PARAMS': 'false', 'ACCELERATE_MIXED_PRECISION': 'bf16', 'FSDP_BACKWARD_PREFETCH': 'BACKWARD_PRE', 'FSDP_FORWARD_PREFETCH': 'true', 'FSDP_STATE_DICT_TYPE': 'FULL_STATE_DICT', 'FSDP_AUTO_WRAP_POLICY': 'SIZE_BASED_WRAP', 'FSDP_MIN_NUM_PARAMS': '100000', 'FSDP_SYNC_MODULE_STATES': 'true', 'FSDP_ACTIVATION_CHECKPOINTING': 'true'}
134
+ [2025-12-08 12:05:51,449][oumi][rank1][pid:8711][MainThread][INFO]][models.py:251] Accelerate FSDP run detected! Setting device_map to None.
135
+ [2025-12-08 12:05:51,449][oumi][rank1][pid:8711][MainThread][INFO]][models.py:260] Building model using device_map: None (DeviceRankInfo(world_size=2, rank=1, local_world_size=2, local_rank=1))...
136
+ [2025-12-08 12:05:51,502][oumi][rank1][pid:8711][MainThread][INFO]][models.py:336] Using model class: <class 'transformers.models.auto.modeling_auto.AutoModelForVision2Seq'> to instantiate model.
137
+ [2025-12-08 12:06:42,423][oumi][rank1][pid:9074][MainThread][INFO]][models.py:469] Setting tokenizer to use the 'right' padding side for model 'Qwen/Qwen2.5-VL-7B-Instruct'. The 'right' padding side is configured as the default value for this model type.
138
+ [2025-12-08 12:06:44,463][oumi][rank1][pid:9074][MainThread][INFO]][train.py:348] Set `training.trainer_kwargs.remove_unused_columns=False` for VLM training with TRL_SFT trainer.
139
+ [2025-12-08 12:06:44,487][oumi][rank1][pid:9074][MainThread][WARNING]][data.py:66] Using torchdata preprocessing pipeline. This is currently in beta and may not be stable.
140
+ [2025-12-08 12:06:44,487][oumi][rank1][pid:9074][MainThread][INFO]][base_map_dataset.py:91] Creating map dataset (type: HuggingFaceVisionDataset)... dataset_name: 'shu4dev/DCVLR_10K'
141
+ [2025-12-08 12:06:45,414][oumi][rank1][pid:9074][MainThread][INFO]][base_map_dataset.py:487] Dataset Info:
142
+ Split: train
143
+ Version: 0.0.0
144
+ Dataset size: 1944466265
145
+ Download size: 1910842881
146
+ Size: 3855309146 bytes
147
+ Rows: 10000
148
+ Columns: ['image', 'problem', 'solution']
149
+ [2025-12-08 12:06:46,749][oumi][rank1][pid:9074][MainThread][INFO]][base_map_dataset.py:426] Loaded DataFrame with shape: (10000, 3). Columns:
150
+ image object
151
+ problem object
152
+ solution object
153
+ dtype: object
154
+ [2025-12-08 12:06:48,689][oumi][rank1][pid:9074][MainThread][INFO]][distributed.py:303] Initialized distributed (True): DeviceRankInfo(world_size=2, rank=1, local_world_size=2, local_rank=1)
155
+ [2025-12-08 12:06:48,689][oumi][rank1][pid:9074][MainThread][INFO]][train.py:449] Set Accelerate environment variables for FSDP: {'ACCELERATE_DYNAMO_BACKEND': 'NO', 'ACCELERATE_DYNAMO_MODE': 'default', 'ACCELERATE_DYNAMO_USE_FULLGRAPH': 'False', 'ACCELERATE_DYNAMO_USE_DYNAMIC': 'False', 'FSDP_CPU_RAM_EFFICIENT_LOADING': 'true', 'FSDP_USE_ORIG_PARAMS': 'true', 'ACCELERATE_USE_FSDP': 'true', 'FSDP_SHARDING_STRATEGY': 'HYBRID_SHARD', 'FSDP_OFFLOAD_PARAMS': 'false', 'ACCELERATE_MIXED_PRECISION': 'bf16', 'FSDP_BACKWARD_PREFETCH': 'BACKWARD_PRE', 'FSDP_FORWARD_PREFETCH': 'true', 'FSDP_STATE_DICT_TYPE': 'FULL_STATE_DICT', 'FSDP_AUTO_WRAP_POLICY': 'SIZE_BASED_WRAP', 'FSDP_MIN_NUM_PARAMS': '100000', 'FSDP_SYNC_MODULE_STATES': 'true', 'FSDP_ACTIVATION_CHECKPOINTING': 'true'}
156
+ [2025-12-08 12:06:48,690][oumi][rank1][pid:9074][MainThread][INFO]][models.py:251] Accelerate FSDP run detected! Setting device_map to None.
157
+ [2025-12-08 12:06:48,690][oumi][rank1][pid:9074][MainThread][INFO]][models.py:260] Building model using device_map: None (DeviceRankInfo(world_size=2, rank=1, local_world_size=2, local_rank=1))...
158
+ [2025-12-08 12:06:48,741][oumi][rank1][pid:9074][MainThread][INFO]][models.py:336] Using model class: <class 'transformers.models.auto.modeling_auto.AutoModelForVision2Seq'> to instantiate model.
159
+ [2025-12-08 12:06:49,379][oumi][rank1][pid:9074][MainThread][INFO]][torch_profiler_utils.py:164] PROF: Torch Profiler disabled!
160
+ [2025-12-08 12:06:49,379][oumi][rank1][pid:9074][MainThread][WARNING]][callbacks.py:72] MFU logging requires packed datasets. Skipping MFU callbacks.
161
+ [2025-12-08 12:06:49,720][oumi][rank1][pid:9074][MainThread][INFO]][device_utils.py:343] GPU Metrics Before Training: GPU runtime info: NVidiaGpuRuntimeInfo(device_index=0, device_count=2, used_memory_mb=1979.0, temperature=28, fan_speed=None, fan_speeds=None, power_usage_watts=115.258, power_limit_watts=700.0, gpu_utilization=0, memory_utilization=0, performance_state=0, clock_speed_graphics=1980, clock_speed_sm=1980, clock_speed_memory=2619).
162
+ [2025-12-08 12:06:49,885][oumi][rank1][pid:9074][MainThread][INFO]][train.py:558] Training init time: 7.539s
163
+ [2025-12-08 12:06:49,885][oumi][rank1][pid:9074][MainThread][INFO]][train.py:559] Starting training... (TrainerType.TRL_SFT, transformers: 4.57.3)
164
+ [2025-12-08 14:26:24,765][oumi][rank1][pid:9074][MainThread][INFO]][train.py:566] Training is Complete.
165
+ [2025-12-08 14:26:24,766][oumi][rank1][pid:9074][MainThread][INFO]][device_utils.py:343] GPU Metrics After Training: GPU runtime info: NVidiaGpuRuntimeInfo(device_index=0, device_count=2, used_memory_mb=80345.0, temperature=35, fan_speed=None, fan_speeds=None, power_usage_watts=277.043, power_limit_watts=700.0, gpu_utilization=31, memory_utilization=11, performance_state=0, clock_speed_graphics=1980, clock_speed_sm=1980, clock_speed_memory=2619).
166
+ [2025-12-08 14:26:24,766][oumi][rank1][pid:9074][MainThread][INFO]][torch_utils.py:135] Peak GPU memory usage: 67.35 GB
167
+ [2025-12-08 14:26:24,766][oumi][rank1][pid:9074][MainThread][INFO]][train.py:573] Saving final state...
168
+ [2025-12-08 14:26:24,803][oumi][rank1][pid:9074][MainThread][INFO]][train.py:578] Saving final model...
169
+ [2025-12-08 14:26:24,803][oumi][rank1][pid:9074][MainThread][INFO]][hf_trainer.py:145] Saving FULL_STATE_DICT for final model checkpoint.
170
+ [2025-12-08 14:27:04,534][oumi][rank1][pid:9074][MainThread][INFO]][hf_trainer.py:152] Model has been saved at output/qwen2_5_vl_7b_10K
171
+ [2025-12-08 14:27:05,044][oumi][rank1][pid:9074][MainThread][INFO]][hf_trainer.py:156] Processor config has been saved at output/qwen2_5_vl_7b_10K
172
+ [2025-12-08 14:27:57,875][oumi][rank1][pid:9074][MainThread][INFO]][train.py:230]
173
+
174
+ » We're always looking for feedback. What's one thing we can improve? https://oumi.ai/feedback
merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
model-00001-of-00007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d4438d521ee64ea5115acafea069cfd8a1a7e2811cd18b9ada9c79c3ca2da611
3
+ size 4952311608
model-00002-of-00007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:068788428da7e883cd3dcc45d7c742861b141464eb75835db12f755e40d10a4d
3
+ size 4984124272
model-00003-of-00007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:87aaee4de23c1ef6d36986b62701fbdf15311a8ed3223d35a0f907b7c72f5760
3
+ size 4932743936
model-00004-of-00007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26dd22e9bcb2a6fb9c95d24c0590a734f35425d3a2fc6943e1ad6ac7ffc3bdf0
3
+ size 4998852296
model-00005-of-00007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8cd1fa3c3ac94cf7a7774a94dca1ba6e22d11b39b4162f9168bba3b479bbe4b6
3
+ size 4984124336
model-00006-of-00007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:aa41fad05022dd7e4b99d6551ef3930d0daf669b133adc0f3de86a881e2e5795
3
+ size 4932743992
model-00007-of-00007.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:423bf9c27e19e013f79771b7593bf2aee5df245fae9805e6e1b7b4407bfb5114
3
+ size 3383846800
model.safetensors.index.json ADDED
@@ -0,0 +1,737 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "metadata": {
3
+ "total_parameters": 4146083328,
4
+ "total_size": 33168666624
5
+ },
6
+ "weight_map": {
7
+ "lm_head.weight": "model-00007-of-00007.safetensors",
8
+ "model.embed_tokens.weight": "model-00001-of-00007.safetensors",
9
+ "model.layers.0.input_layernorm.weight": "model-00002-of-00007.safetensors",
10
+ "model.layers.0.mlp.down_proj.weight": "model-00002-of-00007.safetensors",
11
+ "model.layers.0.mlp.gate_proj.weight": "model-00002-of-00007.safetensors",
12
+ "model.layers.0.mlp.up_proj.weight": "model-00002-of-00007.safetensors",
13
+ "model.layers.0.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
14
+ "model.layers.0.self_attn.k_proj.bias": "model-00001-of-00007.safetensors",
15
+ "model.layers.0.self_attn.k_proj.weight": "model-00001-of-00007.safetensors",
16
+ "model.layers.0.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
17
+ "model.layers.0.self_attn.q_proj.bias": "model-00001-of-00007.safetensors",
18
+ "model.layers.0.self_attn.q_proj.weight": "model-00001-of-00007.safetensors",
19
+ "model.layers.0.self_attn.v_proj.bias": "model-00001-of-00007.safetensors",
20
+ "model.layers.0.self_attn.v_proj.weight": "model-00001-of-00007.safetensors",
21
+ "model.layers.1.input_layernorm.weight": "model-00002-of-00007.safetensors",
22
+ "model.layers.1.mlp.down_proj.weight": "model-00002-of-00007.safetensors",
23
+ "model.layers.1.mlp.gate_proj.weight": "model-00002-of-00007.safetensors",
24
+ "model.layers.1.mlp.up_proj.weight": "model-00002-of-00007.safetensors",
25
+ "model.layers.1.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
26
+ "model.layers.1.self_attn.k_proj.bias": "model-00002-of-00007.safetensors",
27
+ "model.layers.1.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
28
+ "model.layers.1.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
29
+ "model.layers.1.self_attn.q_proj.bias": "model-00002-of-00007.safetensors",
30
+ "model.layers.1.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
31
+ "model.layers.1.self_attn.v_proj.bias": "model-00002-of-00007.safetensors",
32
+ "model.layers.1.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
33
+ "model.layers.10.input_layernorm.weight": "model-00004-of-00007.safetensors",
34
+ "model.layers.10.mlp.down_proj.weight": "model-00004-of-00007.safetensors",
35
+ "model.layers.10.mlp.gate_proj.weight": "model-00003-of-00007.safetensors",
36
+ "model.layers.10.mlp.up_proj.weight": "model-00003-of-00007.safetensors",
37
+ "model.layers.10.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
38
+ "model.layers.10.self_attn.k_proj.bias": "model-00003-of-00007.safetensors",
39
+ "model.layers.10.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
40
+ "model.layers.10.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
41
+ "model.layers.10.self_attn.q_proj.bias": "model-00003-of-00007.safetensors",
42
+ "model.layers.10.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
43
+ "model.layers.10.self_attn.v_proj.bias": "model-00003-of-00007.safetensors",
44
+ "model.layers.10.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
45
+ "model.layers.11.input_layernorm.weight": "model-00004-of-00007.safetensors",
46
+ "model.layers.11.mlp.down_proj.weight": "model-00004-of-00007.safetensors",
47
+ "model.layers.11.mlp.gate_proj.weight": "model-00004-of-00007.safetensors",
48
+ "model.layers.11.mlp.up_proj.weight": "model-00004-of-00007.safetensors",
49
+ "model.layers.11.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
50
+ "model.layers.11.self_attn.k_proj.bias": "model-00004-of-00007.safetensors",
51
+ "model.layers.11.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
52
+ "model.layers.11.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
53
+ "model.layers.11.self_attn.q_proj.bias": "model-00004-of-00007.safetensors",
54
+ "model.layers.11.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
55
+ "model.layers.11.self_attn.v_proj.bias": "model-00004-of-00007.safetensors",
56
+ "model.layers.11.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
57
+ "model.layers.12.input_layernorm.weight": "model-00004-of-00007.safetensors",
58
+ "model.layers.12.mlp.down_proj.weight": "model-00004-of-00007.safetensors",
59
+ "model.layers.12.mlp.gate_proj.weight": "model-00004-of-00007.safetensors",
60
+ "model.layers.12.mlp.up_proj.weight": "model-00004-of-00007.safetensors",
61
+ "model.layers.12.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
62
+ "model.layers.12.self_attn.k_proj.bias": "model-00004-of-00007.safetensors",
63
+ "model.layers.12.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
64
+ "model.layers.12.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
65
+ "model.layers.12.self_attn.q_proj.bias": "model-00004-of-00007.safetensors",
66
+ "model.layers.12.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
67
+ "model.layers.12.self_attn.v_proj.bias": "model-00004-of-00007.safetensors",
68
+ "model.layers.12.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
69
+ "model.layers.13.input_layernorm.weight": "model-00004-of-00007.safetensors",
70
+ "model.layers.13.mlp.down_proj.weight": "model-00004-of-00007.safetensors",
71
+ "model.layers.13.mlp.gate_proj.weight": "model-00004-of-00007.safetensors",
72
+ "model.layers.13.mlp.up_proj.weight": "model-00004-of-00007.safetensors",
73
+ "model.layers.13.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
74
+ "model.layers.13.self_attn.k_proj.bias": "model-00004-of-00007.safetensors",
75
+ "model.layers.13.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
76
+ "model.layers.13.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
77
+ "model.layers.13.self_attn.q_proj.bias": "model-00004-of-00007.safetensors",
78
+ "model.layers.13.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
79
+ "model.layers.13.self_attn.v_proj.bias": "model-00004-of-00007.safetensors",
80
+ "model.layers.13.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
81
+ "model.layers.14.input_layernorm.weight": "model-00004-of-00007.safetensors",
82
+ "model.layers.14.mlp.down_proj.weight": "model-00004-of-00007.safetensors",
83
+ "model.layers.14.mlp.gate_proj.weight": "model-00004-of-00007.safetensors",
84
+ "model.layers.14.mlp.up_proj.weight": "model-00004-of-00007.safetensors",
85
+ "model.layers.14.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
86
+ "model.layers.14.self_attn.k_proj.bias": "model-00004-of-00007.safetensors",
87
+ "model.layers.14.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
88
+ "model.layers.14.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
89
+ "model.layers.14.self_attn.q_proj.bias": "model-00004-of-00007.safetensors",
90
+ "model.layers.14.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
91
+ "model.layers.14.self_attn.v_proj.bias": "model-00004-of-00007.safetensors",
92
+ "model.layers.14.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
93
+ "model.layers.15.input_layernorm.weight": "model-00004-of-00007.safetensors",
94
+ "model.layers.15.mlp.down_proj.weight": "model-00004-of-00007.safetensors",
95
+ "model.layers.15.mlp.gate_proj.weight": "model-00004-of-00007.safetensors",
96
+ "model.layers.15.mlp.up_proj.weight": "model-00004-of-00007.safetensors",
97
+ "model.layers.15.post_attention_layernorm.weight": "model-00004-of-00007.safetensors",
98
+ "model.layers.15.self_attn.k_proj.bias": "model-00004-of-00007.safetensors",
99
+ "model.layers.15.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
100
+ "model.layers.15.self_attn.o_proj.weight": "model-00004-of-00007.safetensors",
101
+ "model.layers.15.self_attn.q_proj.bias": "model-00004-of-00007.safetensors",
102
+ "model.layers.15.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
103
+ "model.layers.15.self_attn.v_proj.bias": "model-00004-of-00007.safetensors",
104
+ "model.layers.15.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
105
+ "model.layers.16.input_layernorm.weight": "model-00005-of-00007.safetensors",
106
+ "model.layers.16.mlp.down_proj.weight": "model-00005-of-00007.safetensors",
107
+ "model.layers.16.mlp.gate_proj.weight": "model-00005-of-00007.safetensors",
108
+ "model.layers.16.mlp.up_proj.weight": "model-00005-of-00007.safetensors",
109
+ "model.layers.16.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
110
+ "model.layers.16.self_attn.k_proj.bias": "model-00004-of-00007.safetensors",
111
+ "model.layers.16.self_attn.k_proj.weight": "model-00004-of-00007.safetensors",
112
+ "model.layers.16.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
113
+ "model.layers.16.self_attn.q_proj.bias": "model-00004-of-00007.safetensors",
114
+ "model.layers.16.self_attn.q_proj.weight": "model-00004-of-00007.safetensors",
115
+ "model.layers.16.self_attn.v_proj.bias": "model-00004-of-00007.safetensors",
116
+ "model.layers.16.self_attn.v_proj.weight": "model-00004-of-00007.safetensors",
117
+ "model.layers.17.input_layernorm.weight": "model-00005-of-00007.safetensors",
118
+ "model.layers.17.mlp.down_proj.weight": "model-00005-of-00007.safetensors",
119
+ "model.layers.17.mlp.gate_proj.weight": "model-00005-of-00007.safetensors",
120
+ "model.layers.17.mlp.up_proj.weight": "model-00005-of-00007.safetensors",
121
+ "model.layers.17.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
122
+ "model.layers.17.self_attn.k_proj.bias": "model-00005-of-00007.safetensors",
123
+ "model.layers.17.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
124
+ "model.layers.17.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
125
+ "model.layers.17.self_attn.q_proj.bias": "model-00005-of-00007.safetensors",
126
+ "model.layers.17.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
127
+ "model.layers.17.self_attn.v_proj.bias": "model-00005-of-00007.safetensors",
128
+ "model.layers.17.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
129
+ "model.layers.18.input_layernorm.weight": "model-00005-of-00007.safetensors",
130
+ "model.layers.18.mlp.down_proj.weight": "model-00005-of-00007.safetensors",
131
+ "model.layers.18.mlp.gate_proj.weight": "model-00005-of-00007.safetensors",
132
+ "model.layers.18.mlp.up_proj.weight": "model-00005-of-00007.safetensors",
133
+ "model.layers.18.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
134
+ "model.layers.18.self_attn.k_proj.bias": "model-00005-of-00007.safetensors",
135
+ "model.layers.18.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
136
+ "model.layers.18.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
137
+ "model.layers.18.self_attn.q_proj.bias": "model-00005-of-00007.safetensors",
138
+ "model.layers.18.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
139
+ "model.layers.18.self_attn.v_proj.bias": "model-00005-of-00007.safetensors",
140
+ "model.layers.18.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
141
+ "model.layers.19.input_layernorm.weight": "model-00005-of-00007.safetensors",
142
+ "model.layers.19.mlp.down_proj.weight": "model-00005-of-00007.safetensors",
143
+ "model.layers.19.mlp.gate_proj.weight": "model-00005-of-00007.safetensors",
144
+ "model.layers.19.mlp.up_proj.weight": "model-00005-of-00007.safetensors",
145
+ "model.layers.19.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
146
+ "model.layers.19.self_attn.k_proj.bias": "model-00005-of-00007.safetensors",
147
+ "model.layers.19.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
148
+ "model.layers.19.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
149
+ "model.layers.19.self_attn.q_proj.bias": "model-00005-of-00007.safetensors",
150
+ "model.layers.19.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
151
+ "model.layers.19.self_attn.v_proj.bias": "model-00005-of-00007.safetensors",
152
+ "model.layers.19.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
153
+ "model.layers.2.input_layernorm.weight": "model-00002-of-00007.safetensors",
154
+ "model.layers.2.mlp.down_proj.weight": "model-00002-of-00007.safetensors",
155
+ "model.layers.2.mlp.gate_proj.weight": "model-00002-of-00007.safetensors",
156
+ "model.layers.2.mlp.up_proj.weight": "model-00002-of-00007.safetensors",
157
+ "model.layers.2.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
158
+ "model.layers.2.self_attn.k_proj.bias": "model-00002-of-00007.safetensors",
159
+ "model.layers.2.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
160
+ "model.layers.2.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
161
+ "model.layers.2.self_attn.q_proj.bias": "model-00002-of-00007.safetensors",
162
+ "model.layers.2.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
163
+ "model.layers.2.self_attn.v_proj.bias": "model-00002-of-00007.safetensors",
164
+ "model.layers.2.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
165
+ "model.layers.20.input_layernorm.weight": "model-00005-of-00007.safetensors",
166
+ "model.layers.20.mlp.down_proj.weight": "model-00005-of-00007.safetensors",
167
+ "model.layers.20.mlp.gate_proj.weight": "model-00005-of-00007.safetensors",
168
+ "model.layers.20.mlp.up_proj.weight": "model-00005-of-00007.safetensors",
169
+ "model.layers.20.post_attention_layernorm.weight": "model-00005-of-00007.safetensors",
170
+ "model.layers.20.self_attn.k_proj.bias": "model-00005-of-00007.safetensors",
171
+ "model.layers.20.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
172
+ "model.layers.20.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
173
+ "model.layers.20.self_attn.q_proj.bias": "model-00005-of-00007.safetensors",
174
+ "model.layers.20.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
175
+ "model.layers.20.self_attn.v_proj.bias": "model-00005-of-00007.safetensors",
176
+ "model.layers.20.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
177
+ "model.layers.21.input_layernorm.weight": "model-00006-of-00007.safetensors",
178
+ "model.layers.21.mlp.down_proj.weight": "model-00006-of-00007.safetensors",
179
+ "model.layers.21.mlp.gate_proj.weight": "model-00005-of-00007.safetensors",
180
+ "model.layers.21.mlp.up_proj.weight": "model-00006-of-00007.safetensors",
181
+ "model.layers.21.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
182
+ "model.layers.21.self_attn.k_proj.bias": "model-00005-of-00007.safetensors",
183
+ "model.layers.21.self_attn.k_proj.weight": "model-00005-of-00007.safetensors",
184
+ "model.layers.21.self_attn.o_proj.weight": "model-00005-of-00007.safetensors",
185
+ "model.layers.21.self_attn.q_proj.bias": "model-00005-of-00007.safetensors",
186
+ "model.layers.21.self_attn.q_proj.weight": "model-00005-of-00007.safetensors",
187
+ "model.layers.21.self_attn.v_proj.bias": "model-00005-of-00007.safetensors",
188
+ "model.layers.21.self_attn.v_proj.weight": "model-00005-of-00007.safetensors",
189
+ "model.layers.22.input_layernorm.weight": "model-00006-of-00007.safetensors",
190
+ "model.layers.22.mlp.down_proj.weight": "model-00006-of-00007.safetensors",
191
+ "model.layers.22.mlp.gate_proj.weight": "model-00006-of-00007.safetensors",
192
+ "model.layers.22.mlp.up_proj.weight": "model-00006-of-00007.safetensors",
193
+ "model.layers.22.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
194
+ "model.layers.22.self_attn.k_proj.bias": "model-00006-of-00007.safetensors",
195
+ "model.layers.22.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
196
+ "model.layers.22.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
197
+ "model.layers.22.self_attn.q_proj.bias": "model-00006-of-00007.safetensors",
198
+ "model.layers.22.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
199
+ "model.layers.22.self_attn.v_proj.bias": "model-00006-of-00007.safetensors",
200
+ "model.layers.22.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
201
+ "model.layers.23.input_layernorm.weight": "model-00006-of-00007.safetensors",
202
+ "model.layers.23.mlp.down_proj.weight": "model-00006-of-00007.safetensors",
203
+ "model.layers.23.mlp.gate_proj.weight": "model-00006-of-00007.safetensors",
204
+ "model.layers.23.mlp.up_proj.weight": "model-00006-of-00007.safetensors",
205
+ "model.layers.23.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
206
+ "model.layers.23.self_attn.k_proj.bias": "model-00006-of-00007.safetensors",
207
+ "model.layers.23.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
208
+ "model.layers.23.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
209
+ "model.layers.23.self_attn.q_proj.bias": "model-00006-of-00007.safetensors",
210
+ "model.layers.23.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
211
+ "model.layers.23.self_attn.v_proj.bias": "model-00006-of-00007.safetensors",
212
+ "model.layers.23.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
213
+ "model.layers.24.input_layernorm.weight": "model-00006-of-00007.safetensors",
214
+ "model.layers.24.mlp.down_proj.weight": "model-00006-of-00007.safetensors",
215
+ "model.layers.24.mlp.gate_proj.weight": "model-00006-of-00007.safetensors",
216
+ "model.layers.24.mlp.up_proj.weight": "model-00006-of-00007.safetensors",
217
+ "model.layers.24.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
218
+ "model.layers.24.self_attn.k_proj.bias": "model-00006-of-00007.safetensors",
219
+ "model.layers.24.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
220
+ "model.layers.24.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
221
+ "model.layers.24.self_attn.q_proj.bias": "model-00006-of-00007.safetensors",
222
+ "model.layers.24.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
223
+ "model.layers.24.self_attn.v_proj.bias": "model-00006-of-00007.safetensors",
224
+ "model.layers.24.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
225
+ "model.layers.25.input_layernorm.weight": "model-00006-of-00007.safetensors",
226
+ "model.layers.25.mlp.down_proj.weight": "model-00006-of-00007.safetensors",
227
+ "model.layers.25.mlp.gate_proj.weight": "model-00006-of-00007.safetensors",
228
+ "model.layers.25.mlp.up_proj.weight": "model-00006-of-00007.safetensors",
229
+ "model.layers.25.post_attention_layernorm.weight": "model-00006-of-00007.safetensors",
230
+ "model.layers.25.self_attn.k_proj.bias": "model-00006-of-00007.safetensors",
231
+ "model.layers.25.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
232
+ "model.layers.25.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
233
+ "model.layers.25.self_attn.q_proj.bias": "model-00006-of-00007.safetensors",
234
+ "model.layers.25.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
235
+ "model.layers.25.self_attn.v_proj.bias": "model-00006-of-00007.safetensors",
236
+ "model.layers.25.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
237
+ "model.layers.26.input_layernorm.weight": "model-00007-of-00007.safetensors",
238
+ "model.layers.26.mlp.down_proj.weight": "model-00007-of-00007.safetensors",
239
+ "model.layers.26.mlp.gate_proj.weight": "model-00006-of-00007.safetensors",
240
+ "model.layers.26.mlp.up_proj.weight": "model-00006-of-00007.safetensors",
241
+ "model.layers.26.post_attention_layernorm.weight": "model-00007-of-00007.safetensors",
242
+ "model.layers.26.self_attn.k_proj.bias": "model-00006-of-00007.safetensors",
243
+ "model.layers.26.self_attn.k_proj.weight": "model-00006-of-00007.safetensors",
244
+ "model.layers.26.self_attn.o_proj.weight": "model-00006-of-00007.safetensors",
245
+ "model.layers.26.self_attn.q_proj.bias": "model-00006-of-00007.safetensors",
246
+ "model.layers.26.self_attn.q_proj.weight": "model-00006-of-00007.safetensors",
247
+ "model.layers.26.self_attn.v_proj.bias": "model-00006-of-00007.safetensors",
248
+ "model.layers.26.self_attn.v_proj.weight": "model-00006-of-00007.safetensors",
249
+ "model.layers.27.input_layernorm.weight": "model-00007-of-00007.safetensors",
250
+ "model.layers.27.mlp.down_proj.weight": "model-00007-of-00007.safetensors",
251
+ "model.layers.27.mlp.gate_proj.weight": "model-00007-of-00007.safetensors",
252
+ "model.layers.27.mlp.up_proj.weight": "model-00007-of-00007.safetensors",
253
+ "model.layers.27.post_attention_layernorm.weight": "model-00007-of-00007.safetensors",
254
+ "model.layers.27.self_attn.k_proj.bias": "model-00007-of-00007.safetensors",
255
+ "model.layers.27.self_attn.k_proj.weight": "model-00007-of-00007.safetensors",
256
+ "model.layers.27.self_attn.o_proj.weight": "model-00007-of-00007.safetensors",
257
+ "model.layers.27.self_attn.q_proj.bias": "model-00007-of-00007.safetensors",
258
+ "model.layers.27.self_attn.q_proj.weight": "model-00007-of-00007.safetensors",
259
+ "model.layers.27.self_attn.v_proj.bias": "model-00007-of-00007.safetensors",
260
+ "model.layers.27.self_attn.v_proj.weight": "model-00007-of-00007.safetensors",
261
+ "model.layers.3.input_layernorm.weight": "model-00002-of-00007.safetensors",
262
+ "model.layers.3.mlp.down_proj.weight": "model-00002-of-00007.safetensors",
263
+ "model.layers.3.mlp.gate_proj.weight": "model-00002-of-00007.safetensors",
264
+ "model.layers.3.mlp.up_proj.weight": "model-00002-of-00007.safetensors",
265
+ "model.layers.3.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
266
+ "model.layers.3.self_attn.k_proj.bias": "model-00002-of-00007.safetensors",
267
+ "model.layers.3.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
268
+ "model.layers.3.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
269
+ "model.layers.3.self_attn.q_proj.bias": "model-00002-of-00007.safetensors",
270
+ "model.layers.3.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
271
+ "model.layers.3.self_attn.v_proj.bias": "model-00002-of-00007.safetensors",
272
+ "model.layers.3.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
273
+ "model.layers.4.input_layernorm.weight": "model-00002-of-00007.safetensors",
274
+ "model.layers.4.mlp.down_proj.weight": "model-00002-of-00007.safetensors",
275
+ "model.layers.4.mlp.gate_proj.weight": "model-00002-of-00007.safetensors",
276
+ "model.layers.4.mlp.up_proj.weight": "model-00002-of-00007.safetensors",
277
+ "model.layers.4.post_attention_layernorm.weight": "model-00002-of-00007.safetensors",
278
+ "model.layers.4.self_attn.k_proj.bias": "model-00002-of-00007.safetensors",
279
+ "model.layers.4.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
280
+ "model.layers.4.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
281
+ "model.layers.4.self_attn.q_proj.bias": "model-00002-of-00007.safetensors",
282
+ "model.layers.4.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
283
+ "model.layers.4.self_attn.v_proj.bias": "model-00002-of-00007.safetensors",
284
+ "model.layers.4.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
285
+ "model.layers.5.input_layernorm.weight": "model-00003-of-00007.safetensors",
286
+ "model.layers.5.mlp.down_proj.weight": "model-00003-of-00007.safetensors",
287
+ "model.layers.5.mlp.gate_proj.weight": "model-00002-of-00007.safetensors",
288
+ "model.layers.5.mlp.up_proj.weight": "model-00003-of-00007.safetensors",
289
+ "model.layers.5.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
290
+ "model.layers.5.self_attn.k_proj.bias": "model-00002-of-00007.safetensors",
291
+ "model.layers.5.self_attn.k_proj.weight": "model-00002-of-00007.safetensors",
292
+ "model.layers.5.self_attn.o_proj.weight": "model-00002-of-00007.safetensors",
293
+ "model.layers.5.self_attn.q_proj.bias": "model-00002-of-00007.safetensors",
294
+ "model.layers.5.self_attn.q_proj.weight": "model-00002-of-00007.safetensors",
295
+ "model.layers.5.self_attn.v_proj.bias": "model-00002-of-00007.safetensors",
296
+ "model.layers.5.self_attn.v_proj.weight": "model-00002-of-00007.safetensors",
297
+ "model.layers.6.input_layernorm.weight": "model-00003-of-00007.safetensors",
298
+ "model.layers.6.mlp.down_proj.weight": "model-00003-of-00007.safetensors",
299
+ "model.layers.6.mlp.gate_proj.weight": "model-00003-of-00007.safetensors",
300
+ "model.layers.6.mlp.up_proj.weight": "model-00003-of-00007.safetensors",
301
+ "model.layers.6.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
302
+ "model.layers.6.self_attn.k_proj.bias": "model-00003-of-00007.safetensors",
303
+ "model.layers.6.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
304
+ "model.layers.6.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
305
+ "model.layers.6.self_attn.q_proj.bias": "model-00003-of-00007.safetensors",
306
+ "model.layers.6.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
307
+ "model.layers.6.self_attn.v_proj.bias": "model-00003-of-00007.safetensors",
308
+ "model.layers.6.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
309
+ "model.layers.7.input_layernorm.weight": "model-00003-of-00007.safetensors",
310
+ "model.layers.7.mlp.down_proj.weight": "model-00003-of-00007.safetensors",
311
+ "model.layers.7.mlp.gate_proj.weight": "model-00003-of-00007.safetensors",
312
+ "model.layers.7.mlp.up_proj.weight": "model-00003-of-00007.safetensors",
313
+ "model.layers.7.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
314
+ "model.layers.7.self_attn.k_proj.bias": "model-00003-of-00007.safetensors",
315
+ "model.layers.7.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
316
+ "model.layers.7.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
317
+ "model.layers.7.self_attn.q_proj.bias": "model-00003-of-00007.safetensors",
318
+ "model.layers.7.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
319
+ "model.layers.7.self_attn.v_proj.bias": "model-00003-of-00007.safetensors",
320
+ "model.layers.7.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
321
+ "model.layers.8.input_layernorm.weight": "model-00003-of-00007.safetensors",
322
+ "model.layers.8.mlp.down_proj.weight": "model-00003-of-00007.safetensors",
323
+ "model.layers.8.mlp.gate_proj.weight": "model-00003-of-00007.safetensors",
324
+ "model.layers.8.mlp.up_proj.weight": "model-00003-of-00007.safetensors",
325
+ "model.layers.8.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
326
+ "model.layers.8.self_attn.k_proj.bias": "model-00003-of-00007.safetensors",
327
+ "model.layers.8.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
328
+ "model.layers.8.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
329
+ "model.layers.8.self_attn.q_proj.bias": "model-00003-of-00007.safetensors",
330
+ "model.layers.8.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
331
+ "model.layers.8.self_attn.v_proj.bias": "model-00003-of-00007.safetensors",
332
+ "model.layers.8.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
333
+ "model.layers.9.input_layernorm.weight": "model-00003-of-00007.safetensors",
334
+ "model.layers.9.mlp.down_proj.weight": "model-00003-of-00007.safetensors",
335
+ "model.layers.9.mlp.gate_proj.weight": "model-00003-of-00007.safetensors",
336
+ "model.layers.9.mlp.up_proj.weight": "model-00003-of-00007.safetensors",
337
+ "model.layers.9.post_attention_layernorm.weight": "model-00003-of-00007.safetensors",
338
+ "model.layers.9.self_attn.k_proj.bias": "model-00003-of-00007.safetensors",
339
+ "model.layers.9.self_attn.k_proj.weight": "model-00003-of-00007.safetensors",
340
+ "model.layers.9.self_attn.o_proj.weight": "model-00003-of-00007.safetensors",
341
+ "model.layers.9.self_attn.q_proj.bias": "model-00003-of-00007.safetensors",
342
+ "model.layers.9.self_attn.q_proj.weight": "model-00003-of-00007.safetensors",
343
+ "model.layers.9.self_attn.v_proj.bias": "model-00003-of-00007.safetensors",
344
+ "model.layers.9.self_attn.v_proj.weight": "model-00003-of-00007.safetensors",
345
+ "model.norm.weight": "model-00007-of-00007.safetensors",
346
+ "visual.blocks.0.attn.proj.bias": "model-00001-of-00007.safetensors",
347
+ "visual.blocks.0.attn.proj.weight": "model-00001-of-00007.safetensors",
348
+ "visual.blocks.0.attn.qkv.bias": "model-00001-of-00007.safetensors",
349
+ "visual.blocks.0.attn.qkv.weight": "model-00001-of-00007.safetensors",
350
+ "visual.blocks.0.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
351
+ "visual.blocks.0.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
352
+ "visual.blocks.0.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
353
+ "visual.blocks.0.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
354
+ "visual.blocks.0.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
355
+ "visual.blocks.0.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
356
+ "visual.blocks.0.norm1.weight": "model-00001-of-00007.safetensors",
357
+ "visual.blocks.0.norm2.weight": "model-00001-of-00007.safetensors",
358
+ "visual.blocks.1.attn.proj.bias": "model-00001-of-00007.safetensors",
359
+ "visual.blocks.1.attn.proj.weight": "model-00001-of-00007.safetensors",
360
+ "visual.blocks.1.attn.qkv.bias": "model-00001-of-00007.safetensors",
361
+ "visual.blocks.1.attn.qkv.weight": "model-00001-of-00007.safetensors",
362
+ "visual.blocks.1.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
363
+ "visual.blocks.1.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
364
+ "visual.blocks.1.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
365
+ "visual.blocks.1.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
366
+ "visual.blocks.1.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
367
+ "visual.blocks.1.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
368
+ "visual.blocks.1.norm1.weight": "model-00001-of-00007.safetensors",
369
+ "visual.blocks.1.norm2.weight": "model-00001-of-00007.safetensors",
370
+ "visual.blocks.10.attn.proj.bias": "model-00001-of-00007.safetensors",
371
+ "visual.blocks.10.attn.proj.weight": "model-00001-of-00007.safetensors",
372
+ "visual.blocks.10.attn.qkv.bias": "model-00001-of-00007.safetensors",
373
+ "visual.blocks.10.attn.qkv.weight": "model-00001-of-00007.safetensors",
374
+ "visual.blocks.10.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
375
+ "visual.blocks.10.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
376
+ "visual.blocks.10.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
377
+ "visual.blocks.10.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
378
+ "visual.blocks.10.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
379
+ "visual.blocks.10.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
380
+ "visual.blocks.10.norm1.weight": "model-00001-of-00007.safetensors",
381
+ "visual.blocks.10.norm2.weight": "model-00001-of-00007.safetensors",
382
+ "visual.blocks.11.attn.proj.bias": "model-00001-of-00007.safetensors",
383
+ "visual.blocks.11.attn.proj.weight": "model-00001-of-00007.safetensors",
384
+ "visual.blocks.11.attn.qkv.bias": "model-00001-of-00007.safetensors",
385
+ "visual.blocks.11.attn.qkv.weight": "model-00001-of-00007.safetensors",
386
+ "visual.blocks.11.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
387
+ "visual.blocks.11.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
388
+ "visual.blocks.11.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
389
+ "visual.blocks.11.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
390
+ "visual.blocks.11.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
391
+ "visual.blocks.11.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
392
+ "visual.blocks.11.norm1.weight": "model-00001-of-00007.safetensors",
393
+ "visual.blocks.11.norm2.weight": "model-00001-of-00007.safetensors",
394
+ "visual.blocks.12.attn.proj.bias": "model-00001-of-00007.safetensors",
395
+ "visual.blocks.12.attn.proj.weight": "model-00001-of-00007.safetensors",
396
+ "visual.blocks.12.attn.qkv.bias": "model-00001-of-00007.safetensors",
397
+ "visual.blocks.12.attn.qkv.weight": "model-00001-of-00007.safetensors",
398
+ "visual.blocks.12.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
399
+ "visual.blocks.12.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
400
+ "visual.blocks.12.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
401
+ "visual.blocks.12.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
402
+ "visual.blocks.12.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
403
+ "visual.blocks.12.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
404
+ "visual.blocks.12.norm1.weight": "model-00001-of-00007.safetensors",
405
+ "visual.blocks.12.norm2.weight": "model-00001-of-00007.safetensors",
406
+ "visual.blocks.13.attn.proj.bias": "model-00001-of-00007.safetensors",
407
+ "visual.blocks.13.attn.proj.weight": "model-00001-of-00007.safetensors",
408
+ "visual.blocks.13.attn.qkv.bias": "model-00001-of-00007.safetensors",
409
+ "visual.blocks.13.attn.qkv.weight": "model-00001-of-00007.safetensors",
410
+ "visual.blocks.13.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
411
+ "visual.blocks.13.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
412
+ "visual.blocks.13.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
413
+ "visual.blocks.13.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
414
+ "visual.blocks.13.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
415
+ "visual.blocks.13.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
416
+ "visual.blocks.13.norm1.weight": "model-00001-of-00007.safetensors",
417
+ "visual.blocks.13.norm2.weight": "model-00001-of-00007.safetensors",
418
+ "visual.blocks.14.attn.proj.bias": "model-00001-of-00007.safetensors",
419
+ "visual.blocks.14.attn.proj.weight": "model-00001-of-00007.safetensors",
420
+ "visual.blocks.14.attn.qkv.bias": "model-00001-of-00007.safetensors",
421
+ "visual.blocks.14.attn.qkv.weight": "model-00001-of-00007.safetensors",
422
+ "visual.blocks.14.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
423
+ "visual.blocks.14.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
424
+ "visual.blocks.14.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
425
+ "visual.blocks.14.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
426
+ "visual.blocks.14.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
427
+ "visual.blocks.14.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
428
+ "visual.blocks.14.norm1.weight": "model-00001-of-00007.safetensors",
429
+ "visual.blocks.14.norm2.weight": "model-00001-of-00007.safetensors",
430
+ "visual.blocks.15.attn.proj.bias": "model-00001-of-00007.safetensors",
431
+ "visual.blocks.15.attn.proj.weight": "model-00001-of-00007.safetensors",
432
+ "visual.blocks.15.attn.qkv.bias": "model-00001-of-00007.safetensors",
433
+ "visual.blocks.15.attn.qkv.weight": "model-00001-of-00007.safetensors",
434
+ "visual.blocks.15.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
435
+ "visual.blocks.15.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
436
+ "visual.blocks.15.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
437
+ "visual.blocks.15.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
438
+ "visual.blocks.15.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
439
+ "visual.blocks.15.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
440
+ "visual.blocks.15.norm1.weight": "model-00001-of-00007.safetensors",
441
+ "visual.blocks.15.norm2.weight": "model-00001-of-00007.safetensors",
442
+ "visual.blocks.16.attn.proj.bias": "model-00001-of-00007.safetensors",
443
+ "visual.blocks.16.attn.proj.weight": "model-00001-of-00007.safetensors",
444
+ "visual.blocks.16.attn.qkv.bias": "model-00001-of-00007.safetensors",
445
+ "visual.blocks.16.attn.qkv.weight": "model-00001-of-00007.safetensors",
446
+ "visual.blocks.16.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
447
+ "visual.blocks.16.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
448
+ "visual.blocks.16.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
449
+ "visual.blocks.16.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
450
+ "visual.blocks.16.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
451
+ "visual.blocks.16.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
452
+ "visual.blocks.16.norm1.weight": "model-00001-of-00007.safetensors",
453
+ "visual.blocks.16.norm2.weight": "model-00001-of-00007.safetensors",
454
+ "visual.blocks.17.attn.proj.bias": "model-00001-of-00007.safetensors",
455
+ "visual.blocks.17.attn.proj.weight": "model-00001-of-00007.safetensors",
456
+ "visual.blocks.17.attn.qkv.bias": "model-00001-of-00007.safetensors",
457
+ "visual.blocks.17.attn.qkv.weight": "model-00001-of-00007.safetensors",
458
+ "visual.blocks.17.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
459
+ "visual.blocks.17.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
460
+ "visual.blocks.17.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
461
+ "visual.blocks.17.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
462
+ "visual.blocks.17.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
463
+ "visual.blocks.17.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
464
+ "visual.blocks.17.norm1.weight": "model-00001-of-00007.safetensors",
465
+ "visual.blocks.17.norm2.weight": "model-00001-of-00007.safetensors",
466
+ "visual.blocks.18.attn.proj.bias": "model-00001-of-00007.safetensors",
467
+ "visual.blocks.18.attn.proj.weight": "model-00001-of-00007.safetensors",
468
+ "visual.blocks.18.attn.qkv.bias": "model-00001-of-00007.safetensors",
469
+ "visual.blocks.18.attn.qkv.weight": "model-00001-of-00007.safetensors",
470
+ "visual.blocks.18.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
471
+ "visual.blocks.18.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
472
+ "visual.blocks.18.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
473
+ "visual.blocks.18.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
474
+ "visual.blocks.18.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
475
+ "visual.blocks.18.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
476
+ "visual.blocks.18.norm1.weight": "model-00001-of-00007.safetensors",
477
+ "visual.blocks.18.norm2.weight": "model-00001-of-00007.safetensors",
478
+ "visual.blocks.19.attn.proj.bias": "model-00001-of-00007.safetensors",
479
+ "visual.blocks.19.attn.proj.weight": "model-00001-of-00007.safetensors",
480
+ "visual.blocks.19.attn.qkv.bias": "model-00001-of-00007.safetensors",
481
+ "visual.blocks.19.attn.qkv.weight": "model-00001-of-00007.safetensors",
482
+ "visual.blocks.19.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
483
+ "visual.blocks.19.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
484
+ "visual.blocks.19.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
485
+ "visual.blocks.19.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
486
+ "visual.blocks.19.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
487
+ "visual.blocks.19.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
488
+ "visual.blocks.19.norm1.weight": "model-00001-of-00007.safetensors",
489
+ "visual.blocks.19.norm2.weight": "model-00001-of-00007.safetensors",
490
+ "visual.blocks.2.attn.proj.bias": "model-00001-of-00007.safetensors",
491
+ "visual.blocks.2.attn.proj.weight": "model-00001-of-00007.safetensors",
492
+ "visual.blocks.2.attn.qkv.bias": "model-00001-of-00007.safetensors",
493
+ "visual.blocks.2.attn.qkv.weight": "model-00001-of-00007.safetensors",
494
+ "visual.blocks.2.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
495
+ "visual.blocks.2.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
496
+ "visual.blocks.2.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
497
+ "visual.blocks.2.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
498
+ "visual.blocks.2.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
499
+ "visual.blocks.2.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
500
+ "visual.blocks.2.norm1.weight": "model-00001-of-00007.safetensors",
501
+ "visual.blocks.2.norm2.weight": "model-00001-of-00007.safetensors",
502
+ "visual.blocks.20.attn.proj.bias": "model-00001-of-00007.safetensors",
503
+ "visual.blocks.20.attn.proj.weight": "model-00001-of-00007.safetensors",
504
+ "visual.blocks.20.attn.qkv.bias": "model-00001-of-00007.safetensors",
505
+ "visual.blocks.20.attn.qkv.weight": "model-00001-of-00007.safetensors",
506
+ "visual.blocks.20.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
507
+ "visual.blocks.20.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
508
+ "visual.blocks.20.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
509
+ "visual.blocks.20.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
510
+ "visual.blocks.20.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
511
+ "visual.blocks.20.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
512
+ "visual.blocks.20.norm1.weight": "model-00001-of-00007.safetensors",
513
+ "visual.blocks.20.norm2.weight": "model-00001-of-00007.safetensors",
514
+ "visual.blocks.21.attn.proj.bias": "model-00001-of-00007.safetensors",
515
+ "visual.blocks.21.attn.proj.weight": "model-00001-of-00007.safetensors",
516
+ "visual.blocks.21.attn.qkv.bias": "model-00001-of-00007.safetensors",
517
+ "visual.blocks.21.attn.qkv.weight": "model-00001-of-00007.safetensors",
518
+ "visual.blocks.21.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
519
+ "visual.blocks.21.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
520
+ "visual.blocks.21.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
521
+ "visual.blocks.21.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
522
+ "visual.blocks.21.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
523
+ "visual.blocks.21.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
524
+ "visual.blocks.21.norm1.weight": "model-00001-of-00007.safetensors",
525
+ "visual.blocks.21.norm2.weight": "model-00001-of-00007.safetensors",
526
+ "visual.blocks.22.attn.proj.bias": "model-00001-of-00007.safetensors",
527
+ "visual.blocks.22.attn.proj.weight": "model-00001-of-00007.safetensors",
528
+ "visual.blocks.22.attn.qkv.bias": "model-00001-of-00007.safetensors",
529
+ "visual.blocks.22.attn.qkv.weight": "model-00001-of-00007.safetensors",
530
+ "visual.blocks.22.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
531
+ "visual.blocks.22.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
532
+ "visual.blocks.22.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
533
+ "visual.blocks.22.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
534
+ "visual.blocks.22.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
535
+ "visual.blocks.22.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
536
+ "visual.blocks.22.norm1.weight": "model-00001-of-00007.safetensors",
537
+ "visual.blocks.22.norm2.weight": "model-00001-of-00007.safetensors",
538
+ "visual.blocks.23.attn.proj.bias": "model-00001-of-00007.safetensors",
539
+ "visual.blocks.23.attn.proj.weight": "model-00001-of-00007.safetensors",
540
+ "visual.blocks.23.attn.qkv.bias": "model-00001-of-00007.safetensors",
541
+ "visual.blocks.23.attn.qkv.weight": "model-00001-of-00007.safetensors",
542
+ "visual.blocks.23.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
543
+ "visual.blocks.23.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
544
+ "visual.blocks.23.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
545
+ "visual.blocks.23.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
546
+ "visual.blocks.23.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
547
+ "visual.blocks.23.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
548
+ "visual.blocks.23.norm1.weight": "model-00001-of-00007.safetensors",
549
+ "visual.blocks.23.norm2.weight": "model-00001-of-00007.safetensors",
550
+ "visual.blocks.24.attn.proj.bias": "model-00001-of-00007.safetensors",
551
+ "visual.blocks.24.attn.proj.weight": "model-00001-of-00007.safetensors",
552
+ "visual.blocks.24.attn.qkv.bias": "model-00001-of-00007.safetensors",
553
+ "visual.blocks.24.attn.qkv.weight": "model-00001-of-00007.safetensors",
554
+ "visual.blocks.24.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
555
+ "visual.blocks.24.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
556
+ "visual.blocks.24.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
557
+ "visual.blocks.24.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
558
+ "visual.blocks.24.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
559
+ "visual.blocks.24.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
560
+ "visual.blocks.24.norm1.weight": "model-00001-of-00007.safetensors",
561
+ "visual.blocks.24.norm2.weight": "model-00001-of-00007.safetensors",
562
+ "visual.blocks.25.attn.proj.bias": "model-00001-of-00007.safetensors",
563
+ "visual.blocks.25.attn.proj.weight": "model-00001-of-00007.safetensors",
564
+ "visual.blocks.25.attn.qkv.bias": "model-00001-of-00007.safetensors",
565
+ "visual.blocks.25.attn.qkv.weight": "model-00001-of-00007.safetensors",
566
+ "visual.blocks.25.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
567
+ "visual.blocks.25.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
568
+ "visual.blocks.25.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
569
+ "visual.blocks.25.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
570
+ "visual.blocks.25.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
571
+ "visual.blocks.25.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
572
+ "visual.blocks.25.norm1.weight": "model-00001-of-00007.safetensors",
573
+ "visual.blocks.25.norm2.weight": "model-00001-of-00007.safetensors",
574
+ "visual.blocks.26.attn.proj.bias": "model-00001-of-00007.safetensors",
575
+ "visual.blocks.26.attn.proj.weight": "model-00001-of-00007.safetensors",
576
+ "visual.blocks.26.attn.qkv.bias": "model-00001-of-00007.safetensors",
577
+ "visual.blocks.26.attn.qkv.weight": "model-00001-of-00007.safetensors",
578
+ "visual.blocks.26.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
579
+ "visual.blocks.26.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
580
+ "visual.blocks.26.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
581
+ "visual.blocks.26.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
582
+ "visual.blocks.26.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
583
+ "visual.blocks.26.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
584
+ "visual.blocks.26.norm1.weight": "model-00001-of-00007.safetensors",
585
+ "visual.blocks.26.norm2.weight": "model-00001-of-00007.safetensors",
586
+ "visual.blocks.27.attn.proj.bias": "model-00001-of-00007.safetensors",
587
+ "visual.blocks.27.attn.proj.weight": "model-00001-of-00007.safetensors",
588
+ "visual.blocks.27.attn.qkv.bias": "model-00001-of-00007.safetensors",
589
+ "visual.blocks.27.attn.qkv.weight": "model-00001-of-00007.safetensors",
590
+ "visual.blocks.27.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
591
+ "visual.blocks.27.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
592
+ "visual.blocks.27.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
593
+ "visual.blocks.27.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
594
+ "visual.blocks.27.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
595
+ "visual.blocks.27.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
596
+ "visual.blocks.27.norm1.weight": "model-00001-of-00007.safetensors",
597
+ "visual.blocks.27.norm2.weight": "model-00001-of-00007.safetensors",
598
+ "visual.blocks.28.attn.proj.bias": "model-00001-of-00007.safetensors",
599
+ "visual.blocks.28.attn.proj.weight": "model-00001-of-00007.safetensors",
600
+ "visual.blocks.28.attn.qkv.bias": "model-00001-of-00007.safetensors",
601
+ "visual.blocks.28.attn.qkv.weight": "model-00001-of-00007.safetensors",
602
+ "visual.blocks.28.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
603
+ "visual.blocks.28.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
604
+ "visual.blocks.28.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
605
+ "visual.blocks.28.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
606
+ "visual.blocks.28.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
607
+ "visual.blocks.28.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
608
+ "visual.blocks.28.norm1.weight": "model-00001-of-00007.safetensors",
609
+ "visual.blocks.28.norm2.weight": "model-00001-of-00007.safetensors",
610
+ "visual.blocks.29.attn.proj.bias": "model-00001-of-00007.safetensors",
611
+ "visual.blocks.29.attn.proj.weight": "model-00001-of-00007.safetensors",
612
+ "visual.blocks.29.attn.qkv.bias": "model-00001-of-00007.safetensors",
613
+ "visual.blocks.29.attn.qkv.weight": "model-00001-of-00007.safetensors",
614
+ "visual.blocks.29.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
615
+ "visual.blocks.29.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
616
+ "visual.blocks.29.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
617
+ "visual.blocks.29.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
618
+ "visual.blocks.29.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
619
+ "visual.blocks.29.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
620
+ "visual.blocks.29.norm1.weight": "model-00001-of-00007.safetensors",
621
+ "visual.blocks.29.norm2.weight": "model-00001-of-00007.safetensors",
622
+ "visual.blocks.3.attn.proj.bias": "model-00001-of-00007.safetensors",
623
+ "visual.blocks.3.attn.proj.weight": "model-00001-of-00007.safetensors",
624
+ "visual.blocks.3.attn.qkv.bias": "model-00001-of-00007.safetensors",
625
+ "visual.blocks.3.attn.qkv.weight": "model-00001-of-00007.safetensors",
626
+ "visual.blocks.3.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
627
+ "visual.blocks.3.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
628
+ "visual.blocks.3.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
629
+ "visual.blocks.3.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
630
+ "visual.blocks.3.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
631
+ "visual.blocks.3.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
632
+ "visual.blocks.3.norm1.weight": "model-00001-of-00007.safetensors",
633
+ "visual.blocks.3.norm2.weight": "model-00001-of-00007.safetensors",
634
+ "visual.blocks.30.attn.proj.bias": "model-00001-of-00007.safetensors",
635
+ "visual.blocks.30.attn.proj.weight": "model-00001-of-00007.safetensors",
636
+ "visual.blocks.30.attn.qkv.bias": "model-00001-of-00007.safetensors",
637
+ "visual.blocks.30.attn.qkv.weight": "model-00001-of-00007.safetensors",
638
+ "visual.blocks.30.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
639
+ "visual.blocks.30.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
640
+ "visual.blocks.30.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
641
+ "visual.blocks.30.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
642
+ "visual.blocks.30.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
643
+ "visual.blocks.30.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
644
+ "visual.blocks.30.norm1.weight": "model-00001-of-00007.safetensors",
645
+ "visual.blocks.30.norm2.weight": "model-00001-of-00007.safetensors",
646
+ "visual.blocks.31.attn.proj.bias": "model-00001-of-00007.safetensors",
647
+ "visual.blocks.31.attn.proj.weight": "model-00001-of-00007.safetensors",
648
+ "visual.blocks.31.attn.qkv.bias": "model-00001-of-00007.safetensors",
649
+ "visual.blocks.31.attn.qkv.weight": "model-00001-of-00007.safetensors",
650
+ "visual.blocks.31.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
651
+ "visual.blocks.31.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
652
+ "visual.blocks.31.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
653
+ "visual.blocks.31.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
654
+ "visual.blocks.31.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
655
+ "visual.blocks.31.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
656
+ "visual.blocks.31.norm1.weight": "model-00001-of-00007.safetensors",
657
+ "visual.blocks.31.norm2.weight": "model-00001-of-00007.safetensors",
658
+ "visual.blocks.4.attn.proj.bias": "model-00001-of-00007.safetensors",
659
+ "visual.blocks.4.attn.proj.weight": "model-00001-of-00007.safetensors",
660
+ "visual.blocks.4.attn.qkv.bias": "model-00001-of-00007.safetensors",
661
+ "visual.blocks.4.attn.qkv.weight": "model-00001-of-00007.safetensors",
662
+ "visual.blocks.4.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
663
+ "visual.blocks.4.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
664
+ "visual.blocks.4.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
665
+ "visual.blocks.4.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
666
+ "visual.blocks.4.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
667
+ "visual.blocks.4.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
668
+ "visual.blocks.4.norm1.weight": "model-00001-of-00007.safetensors",
669
+ "visual.blocks.4.norm2.weight": "model-00001-of-00007.safetensors",
670
+ "visual.blocks.5.attn.proj.bias": "model-00001-of-00007.safetensors",
671
+ "visual.blocks.5.attn.proj.weight": "model-00001-of-00007.safetensors",
672
+ "visual.blocks.5.attn.qkv.bias": "model-00001-of-00007.safetensors",
673
+ "visual.blocks.5.attn.qkv.weight": "model-00001-of-00007.safetensors",
674
+ "visual.blocks.5.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
675
+ "visual.blocks.5.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
676
+ "visual.blocks.5.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
677
+ "visual.blocks.5.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
678
+ "visual.blocks.5.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
679
+ "visual.blocks.5.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
680
+ "visual.blocks.5.norm1.weight": "model-00001-of-00007.safetensors",
681
+ "visual.blocks.5.norm2.weight": "model-00001-of-00007.safetensors",
682
+ "visual.blocks.6.attn.proj.bias": "model-00001-of-00007.safetensors",
683
+ "visual.blocks.6.attn.proj.weight": "model-00001-of-00007.safetensors",
684
+ "visual.blocks.6.attn.qkv.bias": "model-00001-of-00007.safetensors",
685
+ "visual.blocks.6.attn.qkv.weight": "model-00001-of-00007.safetensors",
686
+ "visual.blocks.6.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
687
+ "visual.blocks.6.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
688
+ "visual.blocks.6.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
689
+ "visual.blocks.6.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
690
+ "visual.blocks.6.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
691
+ "visual.blocks.6.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
692
+ "visual.blocks.6.norm1.weight": "model-00001-of-00007.safetensors",
693
+ "visual.blocks.6.norm2.weight": "model-00001-of-00007.safetensors",
694
+ "visual.blocks.7.attn.proj.bias": "model-00001-of-00007.safetensors",
695
+ "visual.blocks.7.attn.proj.weight": "model-00001-of-00007.safetensors",
696
+ "visual.blocks.7.attn.qkv.bias": "model-00001-of-00007.safetensors",
697
+ "visual.blocks.7.attn.qkv.weight": "model-00001-of-00007.safetensors",
698
+ "visual.blocks.7.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
699
+ "visual.blocks.7.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
700
+ "visual.blocks.7.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
701
+ "visual.blocks.7.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
702
+ "visual.blocks.7.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
703
+ "visual.blocks.7.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
704
+ "visual.blocks.7.norm1.weight": "model-00001-of-00007.safetensors",
705
+ "visual.blocks.7.norm2.weight": "model-00001-of-00007.safetensors",
706
+ "visual.blocks.8.attn.proj.bias": "model-00001-of-00007.safetensors",
707
+ "visual.blocks.8.attn.proj.weight": "model-00001-of-00007.safetensors",
708
+ "visual.blocks.8.attn.qkv.bias": "model-00001-of-00007.safetensors",
709
+ "visual.blocks.8.attn.qkv.weight": "model-00001-of-00007.safetensors",
710
+ "visual.blocks.8.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
711
+ "visual.blocks.8.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
712
+ "visual.blocks.8.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
713
+ "visual.blocks.8.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
714
+ "visual.blocks.8.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
715
+ "visual.blocks.8.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
716
+ "visual.blocks.8.norm1.weight": "model-00001-of-00007.safetensors",
717
+ "visual.blocks.8.norm2.weight": "model-00001-of-00007.safetensors",
718
+ "visual.blocks.9.attn.proj.bias": "model-00001-of-00007.safetensors",
719
+ "visual.blocks.9.attn.proj.weight": "model-00001-of-00007.safetensors",
720
+ "visual.blocks.9.attn.qkv.bias": "model-00001-of-00007.safetensors",
721
+ "visual.blocks.9.attn.qkv.weight": "model-00001-of-00007.safetensors",
722
+ "visual.blocks.9.mlp.down_proj.bias": "model-00001-of-00007.safetensors",
723
+ "visual.blocks.9.mlp.down_proj.weight": "model-00001-of-00007.safetensors",
724
+ "visual.blocks.9.mlp.gate_proj.bias": "model-00001-of-00007.safetensors",
725
+ "visual.blocks.9.mlp.gate_proj.weight": "model-00001-of-00007.safetensors",
726
+ "visual.blocks.9.mlp.up_proj.bias": "model-00001-of-00007.safetensors",
727
+ "visual.blocks.9.mlp.up_proj.weight": "model-00001-of-00007.safetensors",
728
+ "visual.blocks.9.norm1.weight": "model-00001-of-00007.safetensors",
729
+ "visual.blocks.9.norm2.weight": "model-00001-of-00007.safetensors",
730
+ "visual.merger.ln_q.weight": "model-00001-of-00007.safetensors",
731
+ "visual.merger.mlp.0.bias": "model-00001-of-00007.safetensors",
732
+ "visual.merger.mlp.0.weight": "model-00001-of-00007.safetensors",
733
+ "visual.merger.mlp.2.bias": "model-00001-of-00007.safetensors",
734
+ "visual.merger.mlp.2.weight": "model-00001-of-00007.safetensors",
735
+ "visual.patch_embed.proj.weight": "model-00001-of-00007.safetensors"
736
+ }
737
+ }
preprocessor_config.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "crop_size": null,
3
+ "data_format": "channels_first",
4
+ "default_to_square": true,
5
+ "device": null,
6
+ "disable_grouping": null,
7
+ "do_center_crop": null,
8
+ "do_convert_rgb": true,
9
+ "do_normalize": true,
10
+ "do_pad": null,
11
+ "do_rescale": true,
12
+ "do_resize": true,
13
+ "image_mean": [
14
+ 0.48145466,
15
+ 0.4578275,
16
+ 0.40821073
17
+ ],
18
+ "image_processor_type": "Qwen2VLImageProcessorFast",
19
+ "image_std": [
20
+ 0.26862954,
21
+ 0.26130258,
22
+ 0.27577711
23
+ ],
24
+ "input_data_format": null,
25
+ "max_pixels": 12845056,
26
+ "merge_size": 2,
27
+ "min_pixels": 3136,
28
+ "pad_size": null,
29
+ "patch_size": 14,
30
+ "processor_class": "Qwen2_5_VLProcessor",
31
+ "resample": 3,
32
+ "rescale_factor": 0.00392156862745098,
33
+ "return_tensors": null,
34
+ "size": {
35
+ "longest_edge": 12845056,
36
+ "shortest_edge": 3136
37
+ },
38
+ "temporal_patch_size": 2
39
+ }
runs/Dec08_10-49-54_192-222-53-232/events.out.tfevents.1765191410.192-222-53-232.5450.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8bd7422f6695e8e2db0c3a6d3d931bae3de5e5171a9dae88d267bdb2667513f3
3
+ size 8632
runs/Dec08_11-06-11_192-222-53-232/events.out.tfevents.1765192372.192-222-53-232.6208.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e11746416d0849a11d28fd14ea1dae108b7b7e4dfe2258f178ec935b391cbf5c
3
+ size 8068
runs/Dec08_11-19-20_192-222-53-232/events.out.tfevents.1765193149.192-222-53-232.6783.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:821dcf6a82d3f76b6878f66d0736181d3f8628f28568fecd5585c9ba66d88dd3
3
+ size 8602
runs/Dec08_11-39-42_192-222-53-232/events.out.tfevents.1765194375.192-222-53-232.7928.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bcf67bf966fd740aee93f35d0af1c30f76f78229e661c09e03df159628253e89
3
+ size 8632
runs/Dec08_12-06-49_192-222-53-232/events.out.tfevents.1765195741.192-222-53-232.9073.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f3c6237f7d1ceffa6eaefb79f15ce6fa482abb5a392163f4d0f6a02390088d5b
3
+ size 582306
special_tokens_map.json ADDED
@@ -0,0 +1,31 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "additional_special_tokens": [
3
+ "<|im_start|>",
4
+ "<|im_end|>",
5
+ "<|object_ref_start|>",
6
+ "<|object_ref_end|>",
7
+ "<|box_start|>",
8
+ "<|box_end|>",
9
+ "<|quad_start|>",
10
+ "<|quad_end|>",
11
+ "<|vision_start|>",
12
+ "<|vision_end|>",
13
+ "<|vision_pad|>",
14
+ "<|image_pad|>",
15
+ "<|video_pad|>"
16
+ ],
17
+ "eos_token": {
18
+ "content": "<|im_end|>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ },
24
+ "pad_token": {
25
+ "content": "<|endoftext|>",
26
+ "lstrip": false,
27
+ "normalized": false,
28
+ "rstrip": false,
29
+ "single_word": false
30
+ }
31
+ }
telemetry/devices_info.txt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ CPU cores: 52 CUDA devices: 2
2
+ device(0)='NVIDIA H100 80GB HBM3' Capability: (9, 0) Memory: [Total: 79.19GiB Free: 78.17GiB Allocated: 0.0GiB Cached: 0.0GiB]
3
+ device(1)='NVIDIA H100 80GB HBM3' Capability: (9, 0) Memory: [Total: 79.19GiB Free: 78.68GiB Allocated: 0.0GiB Cached: 0.0GiB]
telemetry/telemetry_callback_metrics_rank0000.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train_runtime": 8244.7275,
3
+ "train_samples_per_second": 0.606,
4
+ "train_steps_per_second": 0.606,
5
+ "train_tokens_per_second": 740.833,
6
+ "total_flos": 1.4909655028544307e+17,
7
+ "train_loss": 1.2858075158119202,
8
+ "epoch": 1.0,
9
+ "num_input_tokens_seen": 6415098
10
+ }
telemetry/telemetry_callback_rank0000.json ADDED
@@ -0,0 +1,39 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "hostname": "192-222-53-232",
3
+ "total_time": 8375.288021429,
4
+ "timers": {
5
+ "epochs": {
6
+ "count": 1.0,
7
+ "mean": 8243.57664895,
8
+ "median": 8243.57664895,
9
+ "std_dev": 0,
10
+ "min": 8243.57664895,
11
+ "max": 8243.57664895,
12
+ "total": 8243.57664895,
13
+ "percentage": 98.42738097911375
14
+ },
15
+ "microsteps": {
16
+ "count": 4998.0,
17
+ "mean": 1.5826767139225664,
18
+ "median": 1.487843046999842,
19
+ "std_dev": 0.1933355445553847,
20
+ "min": 1.3639970860003814,
21
+ "max": 2.346771123000508,
22
+ "total": 7910.218216184987,
23
+ "percentage": 94.44711866560186
24
+ },
25
+ "steps": {
26
+ "count": 4998.0,
27
+ "mean": 1.582678092825712,
28
+ "median": 1.4878443304996836,
29
+ "std_dev": 0.19333556450996478,
30
+ "min": 1.363998167000318,
31
+ "max": 2.3467728750001697,
32
+ "total": 7910.225107942909,
33
+ "percentage": 94.44720095241881
34
+ }
35
+ },
36
+ "cuda_timers": {},
37
+ "gpu_memory": [],
38
+ "gpu_temperature": {}
39
+ }
telemetry/telemetry_callback_wandb_rank0000.json ADDED
@@ -0,0 +1,5 @@
 
 
 
 
 
 
1
+ {
2
+ "id": "u1u8gi5o",
3
+ "name": "still-fire-12",
4
+ "url": "https://wandb.ai/shu4-university-of-hawaii-system/huggingface/runs/u1u8gi5o"
5
+ }
telemetry/training_config.yaml ADDED
@@ -0,0 +1,245 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ data:
2
+ train:
3
+ datasets:
4
+ - dataset_name: hf_vision
5
+ dataset_path: null
6
+ subset: null
7
+ split: train
8
+ dataset_kwargs:
9
+ hf_dataset_path: shu4dev/DCVLR_10K
10
+ image_column: image
11
+ question_column: problem
12
+ answer_column: solution
13
+ return_tensors: true
14
+ processor_name: Qwen/Qwen2.5-VL-7B-Instruct
15
+ return_conversations: true
16
+ sample_count: null
17
+ mixture_proportion: null
18
+ shuffle: true
19
+ seed: 42
20
+ shuffle_buffer_size: 1000
21
+ trust_remote_code: true
22
+ transform_num_workers: auto
23
+ collator_name: vision_language_sft
24
+ collator_kwargs:
25
+ process_individually: true
26
+ pack: false
27
+ stream: false
28
+ target_col: null
29
+ mixture_strategy: first_exhausted
30
+ seed: null
31
+ use_torchdata: true
32
+ test:
33
+ datasets: []
34
+ collator_name: null
35
+ collator_kwargs: {}
36
+ pack: false
37
+ stream: false
38
+ target_col: null
39
+ mixture_strategy: first_exhausted
40
+ seed: null
41
+ use_torchdata: null
42
+ validation:
43
+ datasets: []
44
+ collator_name: null
45
+ collator_kwargs: {}
46
+ pack: false
47
+ stream: false
48
+ target_col: null
49
+ mixture_strategy: first_exhausted
50
+ seed: null
51
+ use_torchdata: null
52
+ model:
53
+ model_name: Qwen/Qwen2.5-VL-7B-Instruct
54
+ adapter_model: null
55
+ tokenizer_name: null
56
+ tokenizer_pad_token: null
57
+ tokenizer_kwargs: {}
58
+ processor_kwargs: {}
59
+ model_max_length: 10000
60
+ load_pretrained_weights: true
61
+ trust_remote_code: true
62
+ torch_dtype_str: bfloat16
63
+ compile: false
64
+ chat_template: qwen2-vl-instruct
65
+ chat_template_kwargs: null
66
+ attn_implementation: flash_attention_2
67
+ device_map: auto
68
+ model_kwargs: {}
69
+ enable_liger_kernel: false
70
+ shard_for_eval: false
71
+ freeze_layers: []
72
+ model_revision: null
73
+ training:
74
+ use_peft: false
75
+ trainer_type: TRL_SFT
76
+ enable_gradient_checkpointing: true
77
+ gradient_checkpointing_kwargs:
78
+ use_reentrant: false
79
+ output_dir: output/qwen2_5_vl_7b_10K
80
+ per_device_train_batch_size: 1
81
+ per_device_eval_batch_size: 8
82
+ gradient_accumulation_steps: 1
83
+ max_steps: -1
84
+ num_train_epochs: 1
85
+ save_epoch: false
86
+ save_steps: 0
87
+ save_final_model: true
88
+ seed: 42
89
+ data_seed: 42
90
+ use_deterministic: false
91
+ full_determinism: false
92
+ run_name: null
93
+ metrics_function: null
94
+ reward_functions: null
95
+ grpo:
96
+ model_init_kwargs: {}
97
+ max_prompt_length: null
98
+ max_completion_length: null
99
+ num_generations: null
100
+ temperature: 0.9
101
+ remove_unused_columns: false
102
+ repetition_penalty: 1.0
103
+ use_vllm: false
104
+ vllm_mode: null
105
+ vllm_gpu_memory_utilization: 0.9
106
+ epsilon: 0.2
107
+ log_completions: false
108
+ rollout_function: null
109
+ gkd:
110
+ teacher_model_name_or_path: null
111
+ teacher_model_init_kwargs:
112
+ dtype: auto
113
+ temperature: 0.9
114
+ lmbda: 0.5
115
+ beta: 0.5
116
+ max_new_tokens: 128
117
+ disable_dropout: true
118
+ seq_kd: false
119
+ log_level: info
120
+ dep_log_level: warning
121
+ log_examples: false
122
+ enable_wandb: true
123
+ enable_mlflow: false
124
+ enable_tensorboard: true
125
+ logging_strategy: steps
126
+ logging_dir: null
127
+ logging_steps: 5
128
+ logging_first_step: false
129
+ eval_strategy: 'no'
130
+ eval_steps: 500
131
+ learning_rate: 2.0e-05
132
+ lr_scheduler_type: cosine
133
+ lr_scheduler_kwargs: {}
134
+ warmup_ratio: 0.03
135
+ warmup_steps: null
136
+ optimizer: adamw_torch_fused
137
+ weight_decay: 0.01
138
+ adam_beta1: 0.9
139
+ adam_beta2: 0.999
140
+ adam_epsilon: 1.0e-08
141
+ sgd_momentum: 0.0
142
+ mixed_precision_dtype: NONE
143
+ compile: false
144
+ include_performance_metrics: true
145
+ include_alternative_mfu_metrics: false
146
+ log_model_summary: false
147
+ resume_from_checkpoint: null
148
+ try_resume_from_last_checkpoint: false
149
+ dataloader_num_workers: 2
150
+ dataloader_persistent_workers: false
151
+ dataloader_prefetch_factor: 8
152
+ dataloader_main_process_only: false
153
+ ddp_find_unused_parameters: false
154
+ max_grad_norm: 1.0
155
+ trainer_kwargs:
156
+ max_length: 10000
157
+ remove_unused_columns: false
158
+ dataset_kwargs:
159
+ skip_prepare_dataset: true
160
+ verl_config_overrides: {}
161
+ profiler:
162
+ save_dir: null
163
+ enable_cpu_profiling: false
164
+ enable_cuda_profiling: false
165
+ record_shapes: false
166
+ profile_memory: false
167
+ with_stack: false
168
+ with_flops: false
169
+ with_modules: false
170
+ row_limit: 50
171
+ schedule:
172
+ enable_schedule: false
173
+ wait: 0
174
+ warmup: 1
175
+ active: 3
176
+ repeat: 1
177
+ skip_first: 1
178
+ telemetry:
179
+ telemetry_dir: telemetry
180
+ collect_telemetry_for_all_ranks: false
181
+ track_gpu_temperature: false
182
+ empty_device_cache_steps: 1
183
+ nccl_default_timeout_minutes: null
184
+ label_ignore_index: null
185
+ peft:
186
+ lora_r: 8
187
+ lora_alpha: 8
188
+ lora_dropout: 0.0
189
+ lora_target_modules: null
190
+ lora_target_parameters: null
191
+ lora_modules_to_save: null
192
+ lora_bias: none
193
+ lora_init_weights: DEFAULT
194
+ lora_task_type: CAUSAL_LM
195
+ q_lora: false
196
+ q_lora_bits: 4
197
+ bnb_4bit_quant_type: fp4
198
+ llm_int8_skip_modules: null
199
+ use_bnb_nested_quant: false
200
+ bnb_4bit_quant_storage: uint8
201
+ bnb_4bit_compute_dtype: float32
202
+ peft_save_mode: ADAPTER_ONLY
203
+ fsdp:
204
+ enable_fsdp: true
205
+ sharding_strategy: HYBRID_SHARD
206
+ cpu_offload: false
207
+ mixed_precision: bf16
208
+ backward_prefetch: BACKWARD_PRE
209
+ forward_prefetch: true
210
+ use_orig_params: null
211
+ state_dict_type: FULL_STATE_DICT
212
+ auto_wrap_policy: SIZE_BASED_WRAP
213
+ min_num_params: 100000
214
+ transformer_layer_cls: null
215
+ sync_module_states: true
216
+ deepspeed:
217
+ enable_deepspeed: false
218
+ deepspeed_config_path: null
219
+ zero_stage: ZERO_0
220
+ offload_optimizer: null
221
+ offload_param: null
222
+ precision: null
223
+ overlap_comm: false
224
+ contiguous_gradients: true
225
+ reduce_bucket_size: 500000000
226
+ allgather_bucket_size: 500000000
227
+ allgather_partitions: true
228
+ reduce_scatter: true
229
+ round_robin_gradients: false
230
+ stage3_prefetch_bucket_size: 50000000
231
+ stage3_param_persistence_threshold: 100000
232
+ stage3_max_live_parameters: 1000000000
233
+ stage3_max_reuse_distance: 1000000000
234
+ stage3_gather_16bit_weights_on_model_save: false
235
+ sub_group_size: 1000000000
236
+ train_batch_size: auto
237
+ train_micro_batch_size_per_gpu: auto
238
+ gradient_accumulation_steps: auto
239
+ gradient_clipping: auto
240
+ zero_allow_untested_optimizer: true
241
+ zero_force_ds_cpu_optimizer: true
242
+ activation_checkpointing: {}
243
+ memory_efficient_linear: false
244
+ steps_per_print: 10
245
+ wall_clock_breakdown: false
telemetry/world_size.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "LOCAL_WORLD_SIZE": 2,
3
+ "WORLD_SIZE": 2
4
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
3
+ size 11421896
tokenizer_config.json ADDED
@@ -0,0 +1,209 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": false,
3
+ "add_prefix_space": false,
4
+ "added_tokens_decoder": {
5
+ "151643": {
6
+ "content": "<|endoftext|>",
7
+ "lstrip": false,
8
+ "normalized": false,
9
+ "rstrip": false,
10
+ "single_word": false,
11
+ "special": true
12
+ },
13
+ "151644": {
14
+ "content": "<|im_start|>",
15
+ "lstrip": false,
16
+ "normalized": false,
17
+ "rstrip": false,
18
+ "single_word": false,
19
+ "special": true
20
+ },
21
+ "151645": {
22
+ "content": "<|im_end|>",
23
+ "lstrip": false,
24
+ "normalized": false,
25
+ "rstrip": false,
26
+ "single_word": false,
27
+ "special": true
28
+ },
29
+ "151646": {
30
+ "content": "<|object_ref_start|>",
31
+ "lstrip": false,
32
+ "normalized": false,
33
+ "rstrip": false,
34
+ "single_word": false,
35
+ "special": true
36
+ },
37
+ "151647": {
38
+ "content": "<|object_ref_end|>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false,
43
+ "special": true
44
+ },
45
+ "151648": {
46
+ "content": "<|box_start|>",
47
+ "lstrip": false,
48
+ "normalized": false,
49
+ "rstrip": false,
50
+ "single_word": false,
51
+ "special": true
52
+ },
53
+ "151649": {
54
+ "content": "<|box_end|>",
55
+ "lstrip": false,
56
+ "normalized": false,
57
+ "rstrip": false,
58
+ "single_word": false,
59
+ "special": true
60
+ },
61
+ "151650": {
62
+ "content": "<|quad_start|>",
63
+ "lstrip": false,
64
+ "normalized": false,
65
+ "rstrip": false,
66
+ "single_word": false,
67
+ "special": true
68
+ },
69
+ "151651": {
70
+ "content": "<|quad_end|>",
71
+ "lstrip": false,
72
+ "normalized": false,
73
+ "rstrip": false,
74
+ "single_word": false,
75
+ "special": true
76
+ },
77
+ "151652": {
78
+ "content": "<|vision_start|>",
79
+ "lstrip": false,
80
+ "normalized": false,
81
+ "rstrip": false,
82
+ "single_word": false,
83
+ "special": true
84
+ },
85
+ "151653": {
86
+ "content": "<|vision_end|>",
87
+ "lstrip": false,
88
+ "normalized": false,
89
+ "rstrip": false,
90
+ "single_word": false,
91
+ "special": true
92
+ },
93
+ "151654": {
94
+ "content": "<|vision_pad|>",
95
+ "lstrip": false,
96
+ "normalized": false,
97
+ "rstrip": false,
98
+ "single_word": false,
99
+ "special": true
100
+ },
101
+ "151655": {
102
+ "content": "<|image_pad|>",
103
+ "lstrip": false,
104
+ "normalized": false,
105
+ "rstrip": false,
106
+ "single_word": false,
107
+ "special": true
108
+ },
109
+ "151656": {
110
+ "content": "<|video_pad|>",
111
+ "lstrip": false,
112
+ "normalized": false,
113
+ "rstrip": false,
114
+ "single_word": false,
115
+ "special": true
116
+ },
117
+ "151657": {
118
+ "content": "<tool_call>",
119
+ "lstrip": false,
120
+ "normalized": false,
121
+ "rstrip": false,
122
+ "single_word": false,
123
+ "special": false
124
+ },
125
+ "151658": {
126
+ "content": "</tool_call>",
127
+ "lstrip": false,
128
+ "normalized": false,
129
+ "rstrip": false,
130
+ "single_word": false,
131
+ "special": false
132
+ },
133
+ "151659": {
134
+ "content": "<|fim_prefix|>",
135
+ "lstrip": false,
136
+ "normalized": false,
137
+ "rstrip": false,
138
+ "single_word": false,
139
+ "special": false
140
+ },
141
+ "151660": {
142
+ "content": "<|fim_middle|>",
143
+ "lstrip": false,
144
+ "normalized": false,
145
+ "rstrip": false,
146
+ "single_word": false,
147
+ "special": false
148
+ },
149
+ "151661": {
150
+ "content": "<|fim_suffix|>",
151
+ "lstrip": false,
152
+ "normalized": false,
153
+ "rstrip": false,
154
+ "single_word": false,
155
+ "special": false
156
+ },
157
+ "151662": {
158
+ "content": "<|fim_pad|>",
159
+ "lstrip": false,
160
+ "normalized": false,
161
+ "rstrip": false,
162
+ "single_word": false,
163
+ "special": false
164
+ },
165
+ "151663": {
166
+ "content": "<|repo_name|>",
167
+ "lstrip": false,
168
+ "normalized": false,
169
+ "rstrip": false,
170
+ "single_word": false,
171
+ "special": false
172
+ },
173
+ "151664": {
174
+ "content": "<|file_sep|>",
175
+ "lstrip": false,
176
+ "normalized": false,
177
+ "rstrip": false,
178
+ "single_word": false,
179
+ "special": false
180
+ }
181
+ },
182
+ "additional_special_tokens": [
183
+ "<|im_start|>",
184
+ "<|im_end|>",
185
+ "<|object_ref_start|>",
186
+ "<|object_ref_end|>",
187
+ "<|box_start|>",
188
+ "<|box_end|>",
189
+ "<|quad_start|>",
190
+ "<|quad_end|>",
191
+ "<|vision_start|>",
192
+ "<|vision_end|>",
193
+ "<|vision_pad|>",
194
+ "<|image_pad|>",
195
+ "<|video_pad|>"
196
+ ],
197
+ "bos_token": null,
198
+ "clean_up_tokenization_spaces": false,
199
+ "eos_token": "<|im_end|>",
200
+ "errors": "replace",
201
+ "extra_special_tokens": {},
202
+ "model_max_length": 10000,
203
+ "pad_token": "<|endoftext|>",
204
+ "padding_side": "right",
205
+ "processor_class": "Qwen2_5_VLProcessor",
206
+ "split_special_tokens": false,
207
+ "tokenizer_class": "Qwen2Tokenizer",
208
+ "unk_token": null
209
+ }
trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4ffe38194e08d34516b37e5543c51da13be638b68ec8da9e7d332d8bec924c97
3
+ size 6289
video_preprocessor_config.json ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "crop_size": null,
3
+ "data_format": "channels_first",
4
+ "default_to_square": true,
5
+ "device": null,
6
+ "do_center_crop": null,
7
+ "do_convert_rgb": true,
8
+ "do_normalize": true,
9
+ "do_rescale": true,
10
+ "do_resize": true,
11
+ "do_sample_frames": false,
12
+ "fps": null,
13
+ "image_mean": [
14
+ 0.48145466,
15
+ 0.4578275,
16
+ 0.40821073
17
+ ],
18
+ "image_std": [
19
+ 0.26862954,
20
+ 0.26130258,
21
+ 0.27577711
22
+ ],
23
+ "input_data_format": null,
24
+ "max_frames": 768,
25
+ "max_pixels": 12845056,
26
+ "merge_size": 2,
27
+ "min_frames": 4,
28
+ "min_pixels": 3136,
29
+ "num_frames": null,
30
+ "pad_size": null,
31
+ "patch_size": 14,
32
+ "processor_class": "Qwen2_5_VLProcessor",
33
+ "resample": 3,
34
+ "rescale_factor": 0.00392156862745098,
35
+ "return_metadata": false,
36
+ "size": {
37
+ "longest_edge": 12845056,
38
+ "shortest_edge": 3136
39
+ },
40
+ "temporal_patch_size": 2,
41
+ "video_metadata": null,
42
+ "video_processor_type": "Qwen2VLVideoProcessor"
43
+ }
vocab.json ADDED
The diff for this file is too large to render. See raw diff