mjf-su commited on
Commit
62fdfd5
·
verified ·
1 Parent(s): f0a8749

Upload folder using huggingface_hub

Browse files
.gitattributes CHANGED
@@ -33,3 +33,7 @@ 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
+ checkpoint-1000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
+ checkpoint-2000/tokenizer.json filter=lfs diff=lfs merge=lfs -text
38
+ checkpoint-2901/tokenizer.json filter=lfs diff=lfs merge=lfs -text
39
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
chat_template.jinja ADDED
@@ -0,0 +1,120 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0].role == 'system' %}
4
+ {%- if messages[0].content is string %}
5
+ {{- messages[0].content }}
6
+ {%- else %}
7
+ {%- for content in messages[0].content %}
8
+ {%- if 'text' in content %}
9
+ {{- content.text }}
10
+ {%- endif %}
11
+ {%- endfor %}
12
+ {%- endif %}
13
+ {{- '\n\n' }}
14
+ {%- endif %}
15
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
16
+ {%- for tool in tools %}
17
+ {{- "\n" }}
18
+ {{- tool | tojson }}
19
+ {%- endfor %}
20
+ {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
21
+ {%- else %}
22
+ {%- if messages[0].role == 'system' %}
23
+ {{- '<|im_start|>system\n' }}
24
+ {%- if messages[0].content is string %}
25
+ {{- messages[0].content }}
26
+ {%- else %}
27
+ {%- for content in messages[0].content %}
28
+ {%- if 'text' in content %}
29
+ {{- content.text }}
30
+ {%- endif %}
31
+ {%- endfor %}
32
+ {%- endif %}
33
+ {{- '<|im_end|>\n' }}
34
+ {%- endif %}
35
+ {%- endif %}
36
+ {%- set image_count = namespace(value=0) %}
37
+ {%- set video_count = namespace(value=0) %}
38
+ {%- for message in messages %}
39
+ {%- if message.role == "user" %}
40
+ {{- '<|im_start|>' + message.role + '\n' }}
41
+ {%- if message.content is string %}
42
+ {{- message.content }}
43
+ {%- else %}
44
+ {%- for content in message.content %}
45
+ {%- if content.type == 'image' or 'image' in content or 'image_url' in content %}
46
+ {%- set image_count.value = image_count.value + 1 %}
47
+ {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%}
48
+ <|vision_start|><|image_pad|><|vision_end|>
49
+ {%- elif content.type == 'video' or 'video' in content %}
50
+ {%- set video_count.value = video_count.value + 1 %}
51
+ {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%}
52
+ <|vision_start|><|video_pad|><|vision_end|>
53
+ {%- elif 'text' in content %}
54
+ {{- content.text }}
55
+ {%- endif %}
56
+ {%- endfor %}
57
+ {%- endif %}
58
+ {{- '<|im_end|>\n' }}
59
+ {%- elif message.role == "assistant" %}
60
+ {{- '<|im_start|>' + message.role + '\n' }}
61
+ {%- if message.content is string %}
62
+ {{- message.content }}
63
+ {%- else %}
64
+ {%- for content_item in message.content %}
65
+ {%- if 'text' in content_item %}
66
+ {{- content_item.text }}
67
+ {%- endif %}
68
+ {%- endfor %}
69
+ {%- endif %}
70
+ {%- if message.tool_calls %}
71
+ {%- for tool_call in message.tool_calls %}
72
+ {%- if (loop.first and message.content) or (not loop.first) %}
73
+ {{- '\n' }}
74
+ {%- endif %}
75
+ {%- if tool_call.function %}
76
+ {%- set tool_call = tool_call.function %}
77
+ {%- endif %}
78
+ {{- '<tool_call>\n{"name": "' }}
79
+ {{- tool_call.name }}
80
+ {{- '", "arguments": ' }}
81
+ {%- if tool_call.arguments is string %}
82
+ {{- tool_call.arguments }}
83
+ {%- else %}
84
+ {{- tool_call.arguments | tojson }}
85
+ {%- endif %}
86
+ {{- '}\n</tool_call>' }}
87
+ {%- endfor %}
88
+ {%- endif %}
89
+ {{- '<|im_end|>\n' }}
90
+ {%- elif message.role == "tool" %}
91
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
92
+ {{- '<|im_start|>user' }}
93
+ {%- endif %}
94
+ {{- '\n<tool_response>\n' }}
95
+ {%- if message.content is string %}
96
+ {{- message.content }}
97
+ {%- else %}
98
+ {%- for content in message.content %}
99
+ {%- if content.type == 'image' or 'image' in content or 'image_url' in content %}
100
+ {%- set image_count.value = image_count.value + 1 %}
101
+ {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%}
102
+ <|vision_start|><|image_pad|><|vision_end|>
103
+ {%- elif content.type == 'video' or 'video' in content %}
104
+ {%- set video_count.value = video_count.value + 1 %}
105
+ {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%}
106
+ <|vision_start|><|video_pad|><|vision_end|>
107
+ {%- elif 'text' in content %}
108
+ {{- content.text }}
109
+ {%- endif %}
110
+ {%- endfor %}
111
+ {%- endif %}
112
+ {{- '\n</tool_response>' }}
113
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
114
+ {{- '<|im_end|>\n' }}
115
+ {%- endif %}
116
+ {%- endif %}
117
+ {%- endfor %}
118
+ {%- if add_generation_prompt %}
119
+ {{- '<|im_start|>assistant\n' }}
120
+ {%- endif %}
checkpoint-1000/chat_template.jinja ADDED
@@ -0,0 +1,120 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0].role == 'system' %}
4
+ {%- if messages[0].content is string %}
5
+ {{- messages[0].content }}
6
+ {%- else %}
7
+ {%- for content in messages[0].content %}
8
+ {%- if 'text' in content %}
9
+ {{- content.text }}
10
+ {%- endif %}
11
+ {%- endfor %}
12
+ {%- endif %}
13
+ {{- '\n\n' }}
14
+ {%- endif %}
15
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
16
+ {%- for tool in tools %}
17
+ {{- "\n" }}
18
+ {{- tool | tojson }}
19
+ {%- endfor %}
20
+ {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
21
+ {%- else %}
22
+ {%- if messages[0].role == 'system' %}
23
+ {{- '<|im_start|>system\n' }}
24
+ {%- if messages[0].content is string %}
25
+ {{- messages[0].content }}
26
+ {%- else %}
27
+ {%- for content in messages[0].content %}
28
+ {%- if 'text' in content %}
29
+ {{- content.text }}
30
+ {%- endif %}
31
+ {%- endfor %}
32
+ {%- endif %}
33
+ {{- '<|im_end|>\n' }}
34
+ {%- endif %}
35
+ {%- endif %}
36
+ {%- set image_count = namespace(value=0) %}
37
+ {%- set video_count = namespace(value=0) %}
38
+ {%- for message in messages %}
39
+ {%- if message.role == "user" %}
40
+ {{- '<|im_start|>' + message.role + '\n' }}
41
+ {%- if message.content is string %}
42
+ {{- message.content }}
43
+ {%- else %}
44
+ {%- for content in message.content %}
45
+ {%- if content.type == 'image' or 'image' in content or 'image_url' in content %}
46
+ {%- set image_count.value = image_count.value + 1 %}
47
+ {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%}
48
+ <|vision_start|><|image_pad|><|vision_end|>
49
+ {%- elif content.type == 'video' or 'video' in content %}
50
+ {%- set video_count.value = video_count.value + 1 %}
51
+ {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%}
52
+ <|vision_start|><|video_pad|><|vision_end|>
53
+ {%- elif 'text' in content %}
54
+ {{- content.text }}
55
+ {%- endif %}
56
+ {%- endfor %}
57
+ {%- endif %}
58
+ {{- '<|im_end|>\n' }}
59
+ {%- elif message.role == "assistant" %}
60
+ {{- '<|im_start|>' + message.role + '\n' }}
61
+ {%- if message.content is string %}
62
+ {{- message.content }}
63
+ {%- else %}
64
+ {%- for content_item in message.content %}
65
+ {%- if 'text' in content_item %}
66
+ {{- content_item.text }}
67
+ {%- endif %}
68
+ {%- endfor %}
69
+ {%- endif %}
70
+ {%- if message.tool_calls %}
71
+ {%- for tool_call in message.tool_calls %}
72
+ {%- if (loop.first and message.content) or (not loop.first) %}
73
+ {{- '\n' }}
74
+ {%- endif %}
75
+ {%- if tool_call.function %}
76
+ {%- set tool_call = tool_call.function %}
77
+ {%- endif %}
78
+ {{- '<tool_call>\n{"name": "' }}
79
+ {{- tool_call.name }}
80
+ {{- '", "arguments": ' }}
81
+ {%- if tool_call.arguments is string %}
82
+ {{- tool_call.arguments }}
83
+ {%- else %}
84
+ {{- tool_call.arguments | tojson }}
85
+ {%- endif %}
86
+ {{- '}\n</tool_call>' }}
87
+ {%- endfor %}
88
+ {%- endif %}
89
+ {{- '<|im_end|>\n' }}
90
+ {%- elif message.role == "tool" %}
91
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
92
+ {{- '<|im_start|>user' }}
93
+ {%- endif %}
94
+ {{- '\n<tool_response>\n' }}
95
+ {%- if message.content is string %}
96
+ {{- message.content }}
97
+ {%- else %}
98
+ {%- for content in message.content %}
99
+ {%- if content.type == 'image' or 'image' in content or 'image_url' in content %}
100
+ {%- set image_count.value = image_count.value + 1 %}
101
+ {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%}
102
+ <|vision_start|><|image_pad|><|vision_end|>
103
+ {%- elif content.type == 'video' or 'video' in content %}
104
+ {%- set video_count.value = video_count.value + 1 %}
105
+ {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%}
106
+ <|vision_start|><|video_pad|><|vision_end|>
107
+ {%- elif 'text' in content %}
108
+ {{- content.text }}
109
+ {%- endif %}
110
+ {%- endfor %}
111
+ {%- endif %}
112
+ {{- '\n</tool_response>' }}
113
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
114
+ {{- '<|im_end|>\n' }}
115
+ {%- endif %}
116
+ {%- endif %}
117
+ {%- endfor %}
118
+ {%- if add_generation_prompt %}
119
+ {{- '<|im_start|>assistant\n' }}
120
+ {%- endif %}
checkpoint-1000/config.json ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen3VLForConditionalGeneration"
4
+ ],
5
+ "bos_token_id": null,
6
+ "dtype": "bfloat16",
7
+ "eos_token_id": 151645,
8
+ "image_token_id": 151655,
9
+ "model_type": "qwen3_vl",
10
+ "pad_token_id": 151643,
11
+ "text_config": {
12
+ "attention_bias": false,
13
+ "attention_dropout": 0.0,
14
+ "bos_token_id": 151643,
15
+ "dtype": "bfloat16",
16
+ "eos_token_id": 151645,
17
+ "head_dim": 128,
18
+ "hidden_act": "silu",
19
+ "hidden_size": 2048,
20
+ "initializer_range": 0.02,
21
+ "intermediate_size": 6144,
22
+ "max_position_embeddings": 262144,
23
+ "model_type": "qwen3_vl_text",
24
+ "num_attention_heads": 16,
25
+ "num_hidden_layers": 28,
26
+ "num_key_value_heads": 8,
27
+ "pad_token_id": null,
28
+ "rms_norm_eps": 1e-06,
29
+ "rope_parameters": {
30
+ "mrope_interleaved": true,
31
+ "mrope_section": [
32
+ 24,
33
+ 20,
34
+ 20
35
+ ],
36
+ "rope_theta": 5000000,
37
+ "rope_type": "default"
38
+ },
39
+ "tie_word_embeddings": true,
40
+ "use_cache": true,
41
+ "vocab_size": 151936
42
+ },
43
+ "tie_word_embeddings": true,
44
+ "transformers_version": "5.0.0",
45
+ "use_cache": false,
46
+ "video_token_id": 151656,
47
+ "vision_config": {
48
+ "deepstack_visual_indexes": [
49
+ 5,
50
+ 11,
51
+ 17
52
+ ],
53
+ "depth": 24,
54
+ "dtype": "bfloat16",
55
+ "hidden_act": "gelu_pytorch_tanh",
56
+ "hidden_size": 1024,
57
+ "in_channels": 3,
58
+ "initializer_range": 0.02,
59
+ "intermediate_size": 4096,
60
+ "model_type": "qwen3_vl",
61
+ "num_heads": 16,
62
+ "num_position_embeddings": 2304,
63
+ "out_hidden_size": 2048,
64
+ "patch_size": 16,
65
+ "spatial_merge_size": 2,
66
+ "temporal_patch_size": 2
67
+ },
68
+ "vision_end_token_id": 151653,
69
+ "vision_start_token_id": 151652
70
+ }
checkpoint-1000/generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_sample": true,
3
+ "eos_token_id": [
4
+ 151645,
5
+ 151645,
6
+ 151643
7
+ ],
8
+ "pad_token_id": 151643,
9
+ "repetition_penalty": 1.0,
10
+ "temperature": 0.7,
11
+ "top_k": 20,
12
+ "top_p": 0.8,
13
+ "transformers_version": "5.0.0"
14
+ }
checkpoint-1000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:97a7387c3b66998e069890acce0ec60163ac9a1aa371d96a22d78032e0979f66
3
+ size 4255140312
checkpoint-1000/processor_config.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "image_processor": {
3
+ "data_format": "channels_first",
4
+ "do_convert_rgb": true,
5
+ "do_normalize": true,
6
+ "do_rescale": true,
7
+ "do_resize": true,
8
+ "image_mean": [
9
+ 0.5,
10
+ 0.5,
11
+ 0.5
12
+ ],
13
+ "image_processor_type": "Qwen2VLImageProcessorFast",
14
+ "image_std": [
15
+ 0.5,
16
+ 0.5,
17
+ 0.5
18
+ ],
19
+ "merge_size": 2,
20
+ "patch_size": 16,
21
+ "resample": 3,
22
+ "rescale_factor": 0.00392156862745098,
23
+ "size": {
24
+ "longest_edge": 16777216,
25
+ "shortest_edge": 65536
26
+ },
27
+ "temporal_patch_size": 2
28
+ },
29
+ "processor_class": "Qwen3VLProcessor",
30
+ "video_processor": {
31
+ "data_format": "channels_first",
32
+ "default_to_square": true,
33
+ "do_convert_rgb": true,
34
+ "do_normalize": true,
35
+ "do_rescale": true,
36
+ "do_resize": true,
37
+ "do_sample_frames": true,
38
+ "fps": 2,
39
+ "image_mean": [
40
+ 0.5,
41
+ 0.5,
42
+ 0.5
43
+ ],
44
+ "image_std": [
45
+ 0.5,
46
+ 0.5,
47
+ 0.5
48
+ ],
49
+ "max_frames": 768,
50
+ "merge_size": 2,
51
+ "min_frames": 4,
52
+ "patch_size": 16,
53
+ "resample": 3,
54
+ "rescale_factor": 0.00392156862745098,
55
+ "return_metadata": false,
56
+ "size": {
57
+ "longest_edge": 25165824,
58
+ "shortest_edge": 4096
59
+ },
60
+ "temporal_patch_size": 2,
61
+ "video_processor_type": "Qwen3VLVideoProcessor"
62
+ }
63
+ }
checkpoint-1000/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
3
+ size 11422650
checkpoint-1000/tokenizer_config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "extra_special_tokens": [
9
+ "<|im_start|>",
10
+ "<|im_end|>",
11
+ "<|object_ref_start|>",
12
+ "<|object_ref_end|>",
13
+ "<|box_start|>",
14
+ "<|box_end|>",
15
+ "<|quad_start|>",
16
+ "<|quad_end|>",
17
+ "<|vision_start|>",
18
+ "<|vision_end|>",
19
+ "<|vision_pad|>",
20
+ "<|image_pad|>",
21
+ "<|video_pad|>"
22
+ ],
23
+ "is_local": false,
24
+ "model_max_length": 262144,
25
+ "pad_token": "<|endoftext|>",
26
+ "processor_class": "Qwen3VLProcessor",
27
+ "split_special_tokens": false,
28
+ "tokenizer_class": "Qwen2Tokenizer",
29
+ "unk_token": null
30
+ }
checkpoint-1000/trainer_state.json ADDED
@@ -0,0 +1,190 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.0341497067954466,
6
+ "eval_steps": 500,
7
+ "global_step": 1000,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.05174197999310107,
14
+ "grad_norm": 20.0,
15
+ "learning_rate": 3.356164383561644e-06,
16
+ "loss": 2.9015182495117187,
17
+ "step": 50
18
+ },
19
+ {
20
+ "epoch": 0.10348395998620213,
21
+ "grad_norm": 10.8125,
22
+ "learning_rate": 6.78082191780822e-06,
23
+ "loss": 1.9724142456054687,
24
+ "step": 100
25
+ },
26
+ {
27
+ "epoch": 0.1552259399793032,
28
+ "grad_norm": 12.875,
29
+ "learning_rate": 9.999970742404346e-06,
30
+ "loss": 1.6598092651367187,
31
+ "step": 150
32
+ },
33
+ {
34
+ "epoch": 0.20696791997240427,
35
+ "grad_norm": 8.5,
36
+ "learning_rate": 9.990871149623236e-06,
37
+ "loss": 1.5588894653320313,
38
+ "step": 200
39
+ },
40
+ {
41
+ "epoch": 0.25870989996550536,
42
+ "grad_norm": 8.25,
43
+ "learning_rate": 9.965551392499958e-06,
44
+ "loss": 1.5084381103515625,
45
+ "step": 250
46
+ },
47
+ {
48
+ "epoch": 0.3104518799586064,
49
+ "grad_norm": 8.4375,
50
+ "learning_rate": 9.924093759398476e-06,
51
+ "loss": 1.4763961791992188,
52
+ "step": 300
53
+ },
54
+ {
55
+ "epoch": 0.3621938599517075,
56
+ "grad_norm": 7.5,
57
+ "learning_rate": 9.86663298624003e-06,
58
+ "loss": 1.45826416015625,
59
+ "step": 350
60
+ },
61
+ {
62
+ "epoch": 0.41393583994480854,
63
+ "grad_norm": 6.90625,
64
+ "learning_rate": 9.793355818615905e-06,
65
+ "loss": 1.4410775756835938,
66
+ "step": 400
67
+ },
68
+ {
69
+ "epoch": 0.46567781993790963,
70
+ "grad_norm": 7.09375,
71
+ "learning_rate": 9.7045004048705e-06,
72
+ "loss": 1.4302684020996095,
73
+ "step": 450
74
+ },
75
+ {
76
+ "epoch": 0.5174197999310107,
77
+ "grad_norm": 6.46875,
78
+ "learning_rate": 9.600355522127174e-06,
79
+ "loss": 1.4159922790527344,
80
+ "step": 500
81
+ },
82
+ {
83
+ "epoch": 0.5174197999310107,
84
+ "eval_loss": 1.3360999822616577,
85
+ "eval_runtime": 48.5888,
86
+ "eval_samples_per_second": 51.452,
87
+ "eval_steps_per_second": 0.556,
88
+ "step": 500
89
+ },
90
+ {
91
+ "epoch": 0.5691617799241118,
92
+ "grad_norm": 6.46875,
93
+ "learning_rate": 9.481259637772251e-06,
94
+ "loss": 1.408507080078125,
95
+ "step": 550
96
+ },
97
+ {
98
+ "epoch": 0.6209037599172128,
99
+ "grad_norm": 8.0,
100
+ "learning_rate": 9.347599809447331e-06,
101
+ "loss": 1.406485595703125,
102
+ "step": 600
103
+ },
104
+ {
105
+ "epoch": 0.672645739910314,
106
+ "grad_norm": 6.59375,
107
+ "learning_rate": 9.199810427124904e-06,
108
+ "loss": 1.3980792236328126,
109
+ "step": 650
110
+ },
111
+ {
112
+ "epoch": 0.724387719903415,
113
+ "grad_norm": 6.1875,
114
+ "learning_rate": 9.03837180135546e-06,
115
+ "loss": 1.3900822448730468,
116
+ "step": 700
117
+ },
118
+ {
119
+ "epoch": 0.776129699896516,
120
+ "grad_norm": 6.375,
121
+ "learning_rate": 8.86380860227426e-06,
122
+ "loss": 1.377584228515625,
123
+ "step": 750
124
+ },
125
+ {
126
+ "epoch": 0.8278716798896171,
127
+ "grad_norm": 6.46875,
128
+ "learning_rate": 8.676688154440947e-06,
129
+ "loss": 1.3831480407714845,
130
+ "step": 800
131
+ },
132
+ {
133
+ "epoch": 0.8796136598827182,
134
+ "grad_norm": 6.84375,
135
+ "learning_rate": 8.477618593053693e-06,
136
+ "loss": 1.3737605285644532,
137
+ "step": 850
138
+ },
139
+ {
140
+ "epoch": 0.9313556398758193,
141
+ "grad_norm": 6.5625,
142
+ "learning_rate": 8.26724688753015e-06,
143
+ "loss": 1.3781863403320314,
144
+ "step": 900
145
+ },
146
+ {
147
+ "epoch": 0.9830976198689203,
148
+ "grad_norm": 6.0,
149
+ "learning_rate": 8.04625673887846e-06,
150
+ "loss": 1.3639559936523438,
151
+ "step": 950
152
+ },
153
+ {
154
+ "epoch": 1.0341497067954466,
155
+ "grad_norm": 5.84375,
156
+ "learning_rate": 7.815366357691863e-06,
157
+ "loss": 1.3568319702148437,
158
+ "step": 1000
159
+ },
160
+ {
161
+ "epoch": 1.0341497067954466,
162
+ "eval_loss": 1.2891921997070312,
163
+ "eval_runtime": 36.787,
164
+ "eval_samples_per_second": 67.959,
165
+ "eval_steps_per_second": 0.734,
166
+ "step": 1000
167
+ }
168
+ ],
169
+ "logging_steps": 50,
170
+ "max_steps": 2901,
171
+ "num_input_tokens_seen": 0,
172
+ "num_train_epochs": 3,
173
+ "save_steps": 1000,
174
+ "stateful_callbacks": {
175
+ "TrainerControl": {
176
+ "args": {
177
+ "should_epoch_stop": false,
178
+ "should_evaluate": false,
179
+ "should_log": false,
180
+ "should_save": true,
181
+ "should_training_stop": false
182
+ },
183
+ "attributes": {}
184
+ }
185
+ },
186
+ "total_flos": 2.3557991568129393e+18,
187
+ "train_batch_size": 24,
188
+ "trial_name": null,
189
+ "trial_params": null
190
+ }
checkpoint-1000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d14c537eb87fc9a2e52b16edd6c820da00e0bdfe05198c576d1f6eefa7bcf0d
3
+ size 5201
checkpoint-2000/chat_template.jinja ADDED
@@ -0,0 +1,120 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0].role == 'system' %}
4
+ {%- if messages[0].content is string %}
5
+ {{- messages[0].content }}
6
+ {%- else %}
7
+ {%- for content in messages[0].content %}
8
+ {%- if 'text' in content %}
9
+ {{- content.text }}
10
+ {%- endif %}
11
+ {%- endfor %}
12
+ {%- endif %}
13
+ {{- '\n\n' }}
14
+ {%- endif %}
15
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
16
+ {%- for tool in tools %}
17
+ {{- "\n" }}
18
+ {{- tool | tojson }}
19
+ {%- endfor %}
20
+ {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
21
+ {%- else %}
22
+ {%- if messages[0].role == 'system' %}
23
+ {{- '<|im_start|>system\n' }}
24
+ {%- if messages[0].content is string %}
25
+ {{- messages[0].content }}
26
+ {%- else %}
27
+ {%- for content in messages[0].content %}
28
+ {%- if 'text' in content %}
29
+ {{- content.text }}
30
+ {%- endif %}
31
+ {%- endfor %}
32
+ {%- endif %}
33
+ {{- '<|im_end|>\n' }}
34
+ {%- endif %}
35
+ {%- endif %}
36
+ {%- set image_count = namespace(value=0) %}
37
+ {%- set video_count = namespace(value=0) %}
38
+ {%- for message in messages %}
39
+ {%- if message.role == "user" %}
40
+ {{- '<|im_start|>' + message.role + '\n' }}
41
+ {%- if message.content is string %}
42
+ {{- message.content }}
43
+ {%- else %}
44
+ {%- for content in message.content %}
45
+ {%- if content.type == 'image' or 'image' in content or 'image_url' in content %}
46
+ {%- set image_count.value = image_count.value + 1 %}
47
+ {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%}
48
+ <|vision_start|><|image_pad|><|vision_end|>
49
+ {%- elif content.type == 'video' or 'video' in content %}
50
+ {%- set video_count.value = video_count.value + 1 %}
51
+ {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%}
52
+ <|vision_start|><|video_pad|><|vision_end|>
53
+ {%- elif 'text' in content %}
54
+ {{- content.text }}
55
+ {%- endif %}
56
+ {%- endfor %}
57
+ {%- endif %}
58
+ {{- '<|im_end|>\n' }}
59
+ {%- elif message.role == "assistant" %}
60
+ {{- '<|im_start|>' + message.role + '\n' }}
61
+ {%- if message.content is string %}
62
+ {{- message.content }}
63
+ {%- else %}
64
+ {%- for content_item in message.content %}
65
+ {%- if 'text' in content_item %}
66
+ {{- content_item.text }}
67
+ {%- endif %}
68
+ {%- endfor %}
69
+ {%- endif %}
70
+ {%- if message.tool_calls %}
71
+ {%- for tool_call in message.tool_calls %}
72
+ {%- if (loop.first and message.content) or (not loop.first) %}
73
+ {{- '\n' }}
74
+ {%- endif %}
75
+ {%- if tool_call.function %}
76
+ {%- set tool_call = tool_call.function %}
77
+ {%- endif %}
78
+ {{- '<tool_call>\n{"name": "' }}
79
+ {{- tool_call.name }}
80
+ {{- '", "arguments": ' }}
81
+ {%- if tool_call.arguments is string %}
82
+ {{- tool_call.arguments }}
83
+ {%- else %}
84
+ {{- tool_call.arguments | tojson }}
85
+ {%- endif %}
86
+ {{- '}\n</tool_call>' }}
87
+ {%- endfor %}
88
+ {%- endif %}
89
+ {{- '<|im_end|>\n' }}
90
+ {%- elif message.role == "tool" %}
91
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
92
+ {{- '<|im_start|>user' }}
93
+ {%- endif %}
94
+ {{- '\n<tool_response>\n' }}
95
+ {%- if message.content is string %}
96
+ {{- message.content }}
97
+ {%- else %}
98
+ {%- for content in message.content %}
99
+ {%- if content.type == 'image' or 'image' in content or 'image_url' in content %}
100
+ {%- set image_count.value = image_count.value + 1 %}
101
+ {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%}
102
+ <|vision_start|><|image_pad|><|vision_end|>
103
+ {%- elif content.type == 'video' or 'video' in content %}
104
+ {%- set video_count.value = video_count.value + 1 %}
105
+ {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%}
106
+ <|vision_start|><|video_pad|><|vision_end|>
107
+ {%- elif 'text' in content %}
108
+ {{- content.text }}
109
+ {%- endif %}
110
+ {%- endfor %}
111
+ {%- endif %}
112
+ {{- '\n</tool_response>' }}
113
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
114
+ {{- '<|im_end|>\n' }}
115
+ {%- endif %}
116
+ {%- endif %}
117
+ {%- endfor %}
118
+ {%- if add_generation_prompt %}
119
+ {{- '<|im_start|>assistant\n' }}
120
+ {%- endif %}
checkpoint-2000/config.json ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen3VLForConditionalGeneration"
4
+ ],
5
+ "bos_token_id": null,
6
+ "dtype": "bfloat16",
7
+ "eos_token_id": 151645,
8
+ "image_token_id": 151655,
9
+ "model_type": "qwen3_vl",
10
+ "pad_token_id": 151643,
11
+ "text_config": {
12
+ "attention_bias": false,
13
+ "attention_dropout": 0.0,
14
+ "bos_token_id": 151643,
15
+ "dtype": "bfloat16",
16
+ "eos_token_id": 151645,
17
+ "head_dim": 128,
18
+ "hidden_act": "silu",
19
+ "hidden_size": 2048,
20
+ "initializer_range": 0.02,
21
+ "intermediate_size": 6144,
22
+ "max_position_embeddings": 262144,
23
+ "model_type": "qwen3_vl_text",
24
+ "num_attention_heads": 16,
25
+ "num_hidden_layers": 28,
26
+ "num_key_value_heads": 8,
27
+ "pad_token_id": null,
28
+ "rms_norm_eps": 1e-06,
29
+ "rope_parameters": {
30
+ "mrope_interleaved": true,
31
+ "mrope_section": [
32
+ 24,
33
+ 20,
34
+ 20
35
+ ],
36
+ "rope_theta": 5000000,
37
+ "rope_type": "default"
38
+ },
39
+ "tie_word_embeddings": true,
40
+ "use_cache": true,
41
+ "vocab_size": 151936
42
+ },
43
+ "tie_word_embeddings": true,
44
+ "transformers_version": "5.0.0",
45
+ "use_cache": false,
46
+ "video_token_id": 151656,
47
+ "vision_config": {
48
+ "deepstack_visual_indexes": [
49
+ 5,
50
+ 11,
51
+ 17
52
+ ],
53
+ "depth": 24,
54
+ "dtype": "bfloat16",
55
+ "hidden_act": "gelu_pytorch_tanh",
56
+ "hidden_size": 1024,
57
+ "in_channels": 3,
58
+ "initializer_range": 0.02,
59
+ "intermediate_size": 4096,
60
+ "model_type": "qwen3_vl",
61
+ "num_heads": 16,
62
+ "num_position_embeddings": 2304,
63
+ "out_hidden_size": 2048,
64
+ "patch_size": 16,
65
+ "spatial_merge_size": 2,
66
+ "temporal_patch_size": 2
67
+ },
68
+ "vision_end_token_id": 151653,
69
+ "vision_start_token_id": 151652
70
+ }
checkpoint-2000/generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_sample": true,
3
+ "eos_token_id": [
4
+ 151645,
5
+ 151645,
6
+ 151643
7
+ ],
8
+ "pad_token_id": 151643,
9
+ "repetition_penalty": 1.0,
10
+ "temperature": 0.7,
11
+ "top_k": 20,
12
+ "top_p": 0.8,
13
+ "transformers_version": "5.0.0"
14
+ }
checkpoint-2000/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:087566c6e6d66361a11f580dac123ed321da830ec6dee5fefb7da1c797bb7dfc
3
+ size 4255140312
checkpoint-2000/processor_config.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "image_processor": {
3
+ "data_format": "channels_first",
4
+ "do_convert_rgb": true,
5
+ "do_normalize": true,
6
+ "do_rescale": true,
7
+ "do_resize": true,
8
+ "image_mean": [
9
+ 0.5,
10
+ 0.5,
11
+ 0.5
12
+ ],
13
+ "image_processor_type": "Qwen2VLImageProcessorFast",
14
+ "image_std": [
15
+ 0.5,
16
+ 0.5,
17
+ 0.5
18
+ ],
19
+ "merge_size": 2,
20
+ "patch_size": 16,
21
+ "resample": 3,
22
+ "rescale_factor": 0.00392156862745098,
23
+ "size": {
24
+ "longest_edge": 16777216,
25
+ "shortest_edge": 65536
26
+ },
27
+ "temporal_patch_size": 2
28
+ },
29
+ "processor_class": "Qwen3VLProcessor",
30
+ "video_processor": {
31
+ "data_format": "channels_first",
32
+ "default_to_square": true,
33
+ "do_convert_rgb": true,
34
+ "do_normalize": true,
35
+ "do_rescale": true,
36
+ "do_resize": true,
37
+ "do_sample_frames": true,
38
+ "fps": 2,
39
+ "image_mean": [
40
+ 0.5,
41
+ 0.5,
42
+ 0.5
43
+ ],
44
+ "image_std": [
45
+ 0.5,
46
+ 0.5,
47
+ 0.5
48
+ ],
49
+ "max_frames": 768,
50
+ "merge_size": 2,
51
+ "min_frames": 4,
52
+ "patch_size": 16,
53
+ "resample": 3,
54
+ "rescale_factor": 0.00392156862745098,
55
+ "return_metadata": false,
56
+ "size": {
57
+ "longest_edge": 25165824,
58
+ "shortest_edge": 4096
59
+ },
60
+ "temporal_patch_size": 2,
61
+ "video_processor_type": "Qwen3VLVideoProcessor"
62
+ }
63
+ }
checkpoint-2000/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
3
+ size 11422650
checkpoint-2000/tokenizer_config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "extra_special_tokens": [
9
+ "<|im_start|>",
10
+ "<|im_end|>",
11
+ "<|object_ref_start|>",
12
+ "<|object_ref_end|>",
13
+ "<|box_start|>",
14
+ "<|box_end|>",
15
+ "<|quad_start|>",
16
+ "<|quad_end|>",
17
+ "<|vision_start|>",
18
+ "<|vision_end|>",
19
+ "<|vision_pad|>",
20
+ "<|image_pad|>",
21
+ "<|video_pad|>"
22
+ ],
23
+ "is_local": false,
24
+ "model_max_length": 262144,
25
+ "pad_token": "<|endoftext|>",
26
+ "processor_class": "Qwen3VLProcessor",
27
+ "split_special_tokens": false,
28
+ "tokenizer_class": "Qwen2Tokenizer",
29
+ "unk_token": null
30
+ }
checkpoint-2000/trainer_state.json ADDED
@@ -0,0 +1,346 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 2.068299413590893,
6
+ "eval_steps": 500,
7
+ "global_step": 2000,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.05174197999310107,
14
+ "grad_norm": 20.0,
15
+ "learning_rate": 3.356164383561644e-06,
16
+ "loss": 2.9015182495117187,
17
+ "step": 50
18
+ },
19
+ {
20
+ "epoch": 0.10348395998620213,
21
+ "grad_norm": 10.8125,
22
+ "learning_rate": 6.78082191780822e-06,
23
+ "loss": 1.9724142456054687,
24
+ "step": 100
25
+ },
26
+ {
27
+ "epoch": 0.1552259399793032,
28
+ "grad_norm": 12.875,
29
+ "learning_rate": 9.999970742404346e-06,
30
+ "loss": 1.6598092651367187,
31
+ "step": 150
32
+ },
33
+ {
34
+ "epoch": 0.20696791997240427,
35
+ "grad_norm": 8.5,
36
+ "learning_rate": 9.990871149623236e-06,
37
+ "loss": 1.5588894653320313,
38
+ "step": 200
39
+ },
40
+ {
41
+ "epoch": 0.25870989996550536,
42
+ "grad_norm": 8.25,
43
+ "learning_rate": 9.965551392499958e-06,
44
+ "loss": 1.5084381103515625,
45
+ "step": 250
46
+ },
47
+ {
48
+ "epoch": 0.3104518799586064,
49
+ "grad_norm": 8.4375,
50
+ "learning_rate": 9.924093759398476e-06,
51
+ "loss": 1.4763961791992188,
52
+ "step": 300
53
+ },
54
+ {
55
+ "epoch": 0.3621938599517075,
56
+ "grad_norm": 7.5,
57
+ "learning_rate": 9.86663298624003e-06,
58
+ "loss": 1.45826416015625,
59
+ "step": 350
60
+ },
61
+ {
62
+ "epoch": 0.41393583994480854,
63
+ "grad_norm": 6.90625,
64
+ "learning_rate": 9.793355818615905e-06,
65
+ "loss": 1.4410775756835938,
66
+ "step": 400
67
+ },
68
+ {
69
+ "epoch": 0.46567781993790963,
70
+ "grad_norm": 7.09375,
71
+ "learning_rate": 9.7045004048705e-06,
72
+ "loss": 1.4302684020996095,
73
+ "step": 450
74
+ },
75
+ {
76
+ "epoch": 0.5174197999310107,
77
+ "grad_norm": 6.46875,
78
+ "learning_rate": 9.600355522127174e-06,
79
+ "loss": 1.4159922790527344,
80
+ "step": 500
81
+ },
82
+ {
83
+ "epoch": 0.5174197999310107,
84
+ "eval_loss": 1.3360999822616577,
85
+ "eval_runtime": 48.5888,
86
+ "eval_samples_per_second": 51.452,
87
+ "eval_steps_per_second": 0.556,
88
+ "step": 500
89
+ },
90
+ {
91
+ "epoch": 0.5691617799241118,
92
+ "grad_norm": 6.46875,
93
+ "learning_rate": 9.481259637772251e-06,
94
+ "loss": 1.408507080078125,
95
+ "step": 550
96
+ },
97
+ {
98
+ "epoch": 0.6209037599172128,
99
+ "grad_norm": 8.0,
100
+ "learning_rate": 9.347599809447331e-06,
101
+ "loss": 1.406485595703125,
102
+ "step": 600
103
+ },
104
+ {
105
+ "epoch": 0.672645739910314,
106
+ "grad_norm": 6.59375,
107
+ "learning_rate": 9.199810427124904e-06,
108
+ "loss": 1.3980792236328126,
109
+ "step": 650
110
+ },
111
+ {
112
+ "epoch": 0.724387719903415,
113
+ "grad_norm": 6.1875,
114
+ "learning_rate": 9.03837180135546e-06,
115
+ "loss": 1.3900822448730468,
116
+ "step": 700
117
+ },
118
+ {
119
+ "epoch": 0.776129699896516,
120
+ "grad_norm": 6.375,
121
+ "learning_rate": 8.86380860227426e-06,
122
+ "loss": 1.377584228515625,
123
+ "step": 750
124
+ },
125
+ {
126
+ "epoch": 0.8278716798896171,
127
+ "grad_norm": 6.46875,
128
+ "learning_rate": 8.676688154440947e-06,
129
+ "loss": 1.3831480407714845,
130
+ "step": 800
131
+ },
132
+ {
133
+ "epoch": 0.8796136598827182,
134
+ "grad_norm": 6.84375,
135
+ "learning_rate": 8.477618593053693e-06,
136
+ "loss": 1.3737605285644532,
137
+ "step": 850
138
+ },
139
+ {
140
+ "epoch": 0.9313556398758193,
141
+ "grad_norm": 6.5625,
142
+ "learning_rate": 8.26724688753015e-06,
143
+ "loss": 1.3781863403320314,
144
+ "step": 900
145
+ },
146
+ {
147
+ "epoch": 0.9830976198689203,
148
+ "grad_norm": 6.0,
149
+ "learning_rate": 8.04625673887846e-06,
150
+ "loss": 1.3639559936523438,
151
+ "step": 950
152
+ },
153
+ {
154
+ "epoch": 1.0341497067954466,
155
+ "grad_norm": 5.84375,
156
+ "learning_rate": 7.815366357691863e-06,
157
+ "loss": 1.3568319702148437,
158
+ "step": 1000
159
+ },
160
+ {
161
+ "epoch": 1.0341497067954466,
162
+ "eval_loss": 1.2891921997070312,
163
+ "eval_runtime": 36.787,
164
+ "eval_samples_per_second": 67.959,
165
+ "eval_steps_per_second": 0.734,
166
+ "step": 1000
167
+ },
168
+ {
169
+ "epoch": 1.0858916867885477,
170
+ "grad_norm": 5.84375,
171
+ "learning_rate": 7.575326129988268e-06,
172
+ "loss": 1.3579391479492187,
173
+ "step": 1050
174
+ },
175
+ {
176
+ "epoch": 1.137633666781649,
177
+ "grad_norm": 6.15625,
178
+ "learning_rate": 7.326916178480796e-06,
179
+ "loss": 1.353500213623047,
180
+ "step": 1100
181
+ },
182
+ {
183
+ "epoch": 1.1893756467747498,
184
+ "grad_norm": 5.625,
185
+ "learning_rate": 7.070943827205048e-06,
186
+ "loss": 1.3528330993652344,
187
+ "step": 1150
188
+ },
189
+ {
190
+ "epoch": 1.241117626767851,
191
+ "grad_norm": 5.96875,
192
+ "learning_rate": 6.808240977742987e-06,
193
+ "loss": 1.3461532592773438,
194
+ "step": 1200
195
+ },
196
+ {
197
+ "epoch": 1.2928596067609521,
198
+ "grad_norm": 5.28125,
199
+ "learning_rate": 6.539661405570615e-06,
200
+ "loss": 1.3481027221679687,
201
+ "step": 1250
202
+ },
203
+ {
204
+ "epoch": 1.344601586754053,
205
+ "grad_norm": 6.09375,
206
+ "learning_rate": 6.266077985316224e-06,
207
+ "loss": 1.348444061279297,
208
+ "step": 1300
209
+ },
210
+ {
211
+ "epoch": 1.3963435667471542,
212
+ "grad_norm": 6.1875,
213
+ "learning_rate": 5.988379853947046e-06,
214
+ "loss": 1.3519085693359374,
215
+ "step": 1350
216
+ },
217
+ {
218
+ "epoch": 1.4480855467402551,
219
+ "grad_norm": 5.375,
220
+ "learning_rate": 5.7074695211038385e-06,
221
+ "loss": 1.3478456115722657,
222
+ "step": 1400
223
+ },
224
+ {
225
+ "epoch": 1.4998275267333563,
226
+ "grad_norm": 5.84375,
227
+ "learning_rate": 5.424259935974736e-06,
228
+ "loss": 1.3487643432617187,
229
+ "step": 1450
230
+ },
231
+ {
232
+ "epoch": 1.5515695067264574,
233
+ "grad_norm": 6.09375,
234
+ "learning_rate": 5.139671520240915e-06,
235
+ "loss": 1.3462490844726562,
236
+ "step": 1500
237
+ },
238
+ {
239
+ "epoch": 1.5515695067264574,
240
+ "eval_loss": 1.2737045288085938,
241
+ "eval_runtime": 35.6507,
242
+ "eval_samples_per_second": 70.125,
243
+ "eval_steps_per_second": 0.757,
244
+ "step": 1500
245
+ },
246
+ {
247
+ "epoch": 1.6033114867195586,
248
+ "grad_norm": 10.125,
249
+ "learning_rate": 4.854629176736911e-06,
250
+ "loss": 1.34809326171875,
251
+ "step": 1550
252
+ },
253
+ {
254
+ "epoch": 1.6550534667126595,
255
+ "grad_norm": 5.84375,
256
+ "learning_rate": 4.570059283547328e-06,
257
+ "loss": 1.343378448486328,
258
+ "step": 1600
259
+ },
260
+ {
261
+ "epoch": 1.7067954467057604,
262
+ "grad_norm": 6.0625,
263
+ "learning_rate": 4.286886683309053e-06,
264
+ "loss": 1.341541748046875,
265
+ "step": 1650
266
+ },
267
+ {
268
+ "epoch": 1.7585374266988616,
269
+ "grad_norm": 5.40625,
270
+ "learning_rate": 4.0060316775036025e-06,
271
+ "loss": 1.3389480590820313,
272
+ "step": 1700
273
+ },
274
+ {
275
+ "epoch": 1.8102794066919627,
276
+ "grad_norm": 5.75,
277
+ "learning_rate": 3.728407035508089e-06,
278
+ "loss": 1.3394900512695314,
279
+ "step": 1750
280
+ },
281
+ {
282
+ "epoch": 1.862021386685064,
283
+ "grad_norm": 5.8125,
284
+ "learning_rate": 3.4549150281252635e-06,
285
+ "loss": 1.3372491455078126,
286
+ "step": 1800
287
+ },
288
+ {
289
+ "epoch": 1.913763366678165,
290
+ "grad_norm": 5.71875,
291
+ "learning_rate": 3.1864444952335823e-06,
292
+ "loss": 1.3394888305664063,
293
+ "step": 1850
294
+ },
295
+ {
296
+ "epoch": 1.965505346671266,
297
+ "grad_norm": 5.59375,
298
+ "learning_rate": 2.9238679570873217e-06,
299
+ "loss": 1.34095458984375,
300
+ "step": 1900
301
+ },
302
+ {
303
+ "epoch": 2.0165574335977925,
304
+ "grad_norm": 6.0625,
305
+ "learning_rate": 2.66803877865493e-06,
306
+ "loss": 1.3422109985351562,
307
+ "step": 1950
308
+ },
309
+ {
310
+ "epoch": 2.068299413590893,
311
+ "grad_norm": 5.65625,
312
+ "learning_rate": 2.4197883962114164e-06,
313
+ "loss": 1.3344174194335938,
314
+ "step": 2000
315
+ },
316
+ {
317
+ "epoch": 2.068299413590893,
318
+ "eval_loss": 1.269867181777954,
319
+ "eval_runtime": 37.0763,
320
+ "eval_samples_per_second": 67.429,
321
+ "eval_steps_per_second": 0.728,
322
+ "step": 2000
323
+ }
324
+ ],
325
+ "logging_steps": 50,
326
+ "max_steps": 2901,
327
+ "num_input_tokens_seen": 0,
328
+ "num_train_epochs": 3,
329
+ "save_steps": 1000,
330
+ "stateful_callbacks": {
331
+ "TrainerControl": {
332
+ "args": {
333
+ "should_epoch_stop": false,
334
+ "should_evaluate": false,
335
+ "should_log": false,
336
+ "should_save": true,
337
+ "should_training_stop": false
338
+ },
339
+ "attributes": {}
340
+ }
341
+ },
342
+ "total_flos": 4.711604567098262e+18,
343
+ "train_batch_size": 24,
344
+ "trial_name": null,
345
+ "trial_params": null
346
+ }
checkpoint-2000/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d14c537eb87fc9a2e52b16edd6c820da00e0bdfe05198c576d1f6eefa7bcf0d
3
+ size 5201
checkpoint-2901/chat_template.jinja ADDED
@@ -0,0 +1,120 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0].role == 'system' %}
4
+ {%- if messages[0].content is string %}
5
+ {{- messages[0].content }}
6
+ {%- else %}
7
+ {%- for content in messages[0].content %}
8
+ {%- if 'text' in content %}
9
+ {{- content.text }}
10
+ {%- endif %}
11
+ {%- endfor %}
12
+ {%- endif %}
13
+ {{- '\n\n' }}
14
+ {%- endif %}
15
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
16
+ {%- for tool in tools %}
17
+ {{- "\n" }}
18
+ {{- tool | tojson }}
19
+ {%- endfor %}
20
+ {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
21
+ {%- else %}
22
+ {%- if messages[0].role == 'system' %}
23
+ {{- '<|im_start|>system\n' }}
24
+ {%- if messages[0].content is string %}
25
+ {{- messages[0].content }}
26
+ {%- else %}
27
+ {%- for content in messages[0].content %}
28
+ {%- if 'text' in content %}
29
+ {{- content.text }}
30
+ {%- endif %}
31
+ {%- endfor %}
32
+ {%- endif %}
33
+ {{- '<|im_end|>\n' }}
34
+ {%- endif %}
35
+ {%- endif %}
36
+ {%- set image_count = namespace(value=0) %}
37
+ {%- set video_count = namespace(value=0) %}
38
+ {%- for message in messages %}
39
+ {%- if message.role == "user" %}
40
+ {{- '<|im_start|>' + message.role + '\n' }}
41
+ {%- if message.content is string %}
42
+ {{- message.content }}
43
+ {%- else %}
44
+ {%- for content in message.content %}
45
+ {%- if content.type == 'image' or 'image' in content or 'image_url' in content %}
46
+ {%- set image_count.value = image_count.value + 1 %}
47
+ {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%}
48
+ <|vision_start|><|image_pad|><|vision_end|>
49
+ {%- elif content.type == 'video' or 'video' in content %}
50
+ {%- set video_count.value = video_count.value + 1 %}
51
+ {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%}
52
+ <|vision_start|><|video_pad|><|vision_end|>
53
+ {%- elif 'text' in content %}
54
+ {{- content.text }}
55
+ {%- endif %}
56
+ {%- endfor %}
57
+ {%- endif %}
58
+ {{- '<|im_end|>\n' }}
59
+ {%- elif message.role == "assistant" %}
60
+ {{- '<|im_start|>' + message.role + '\n' }}
61
+ {%- if message.content is string %}
62
+ {{- message.content }}
63
+ {%- else %}
64
+ {%- for content_item in message.content %}
65
+ {%- if 'text' in content_item %}
66
+ {{- content_item.text }}
67
+ {%- endif %}
68
+ {%- endfor %}
69
+ {%- endif %}
70
+ {%- if message.tool_calls %}
71
+ {%- for tool_call in message.tool_calls %}
72
+ {%- if (loop.first and message.content) or (not loop.first) %}
73
+ {{- '\n' }}
74
+ {%- endif %}
75
+ {%- if tool_call.function %}
76
+ {%- set tool_call = tool_call.function %}
77
+ {%- endif %}
78
+ {{- '<tool_call>\n{"name": "' }}
79
+ {{- tool_call.name }}
80
+ {{- '", "arguments": ' }}
81
+ {%- if tool_call.arguments is string %}
82
+ {{- tool_call.arguments }}
83
+ {%- else %}
84
+ {{- tool_call.arguments | tojson }}
85
+ {%- endif %}
86
+ {{- '}\n</tool_call>' }}
87
+ {%- endfor %}
88
+ {%- endif %}
89
+ {{- '<|im_end|>\n' }}
90
+ {%- elif message.role == "tool" %}
91
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
92
+ {{- '<|im_start|>user' }}
93
+ {%- endif %}
94
+ {{- '\n<tool_response>\n' }}
95
+ {%- if message.content is string %}
96
+ {{- message.content }}
97
+ {%- else %}
98
+ {%- for content in message.content %}
99
+ {%- if content.type == 'image' or 'image' in content or 'image_url' in content %}
100
+ {%- set image_count.value = image_count.value + 1 %}
101
+ {%- if add_vision_id %}Picture {{ image_count.value }}: {% endif -%}
102
+ <|vision_start|><|image_pad|><|vision_end|>
103
+ {%- elif content.type == 'video' or 'video' in content %}
104
+ {%- set video_count.value = video_count.value + 1 %}
105
+ {%- if add_vision_id %}Video {{ video_count.value }}: {% endif -%}
106
+ <|vision_start|><|video_pad|><|vision_end|>
107
+ {%- elif 'text' in content %}
108
+ {{- content.text }}
109
+ {%- endif %}
110
+ {%- endfor %}
111
+ {%- endif %}
112
+ {{- '\n</tool_response>' }}
113
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
114
+ {{- '<|im_end|>\n' }}
115
+ {%- endif %}
116
+ {%- endif %}
117
+ {%- endfor %}
118
+ {%- if add_generation_prompt %}
119
+ {{- '<|im_start|>assistant\n' }}
120
+ {%- endif %}
checkpoint-2901/config.json ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen3VLForConditionalGeneration"
4
+ ],
5
+ "bos_token_id": null,
6
+ "dtype": "bfloat16",
7
+ "eos_token_id": 151645,
8
+ "image_token_id": 151655,
9
+ "model_type": "qwen3_vl",
10
+ "pad_token_id": 151643,
11
+ "text_config": {
12
+ "attention_bias": false,
13
+ "attention_dropout": 0.0,
14
+ "bos_token_id": 151643,
15
+ "dtype": "bfloat16",
16
+ "eos_token_id": 151645,
17
+ "head_dim": 128,
18
+ "hidden_act": "silu",
19
+ "hidden_size": 2048,
20
+ "initializer_range": 0.02,
21
+ "intermediate_size": 6144,
22
+ "max_position_embeddings": 262144,
23
+ "model_type": "qwen3_vl_text",
24
+ "num_attention_heads": 16,
25
+ "num_hidden_layers": 28,
26
+ "num_key_value_heads": 8,
27
+ "pad_token_id": null,
28
+ "rms_norm_eps": 1e-06,
29
+ "rope_parameters": {
30
+ "mrope_interleaved": true,
31
+ "mrope_section": [
32
+ 24,
33
+ 20,
34
+ 20
35
+ ],
36
+ "rope_theta": 5000000,
37
+ "rope_type": "default"
38
+ },
39
+ "tie_word_embeddings": true,
40
+ "use_cache": true,
41
+ "vocab_size": 151936
42
+ },
43
+ "tie_word_embeddings": true,
44
+ "transformers_version": "5.0.0",
45
+ "use_cache": false,
46
+ "video_token_id": 151656,
47
+ "vision_config": {
48
+ "deepstack_visual_indexes": [
49
+ 5,
50
+ 11,
51
+ 17
52
+ ],
53
+ "depth": 24,
54
+ "dtype": "bfloat16",
55
+ "hidden_act": "gelu_pytorch_tanh",
56
+ "hidden_size": 1024,
57
+ "in_channels": 3,
58
+ "initializer_range": 0.02,
59
+ "intermediate_size": 4096,
60
+ "model_type": "qwen3_vl",
61
+ "num_heads": 16,
62
+ "num_position_embeddings": 2304,
63
+ "out_hidden_size": 2048,
64
+ "patch_size": 16,
65
+ "spatial_merge_size": 2,
66
+ "temporal_patch_size": 2
67
+ },
68
+ "vision_end_token_id": 151653,
69
+ "vision_start_token_id": 151652
70
+ }
checkpoint-2901/generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_sample": true,
3
+ "eos_token_id": [
4
+ 151645,
5
+ 151645,
6
+ 151643
7
+ ],
8
+ "pad_token_id": 151643,
9
+ "repetition_penalty": 1.0,
10
+ "temperature": 0.7,
11
+ "top_k": 20,
12
+ "top_p": 0.8,
13
+ "transformers_version": "5.0.0"
14
+ }
checkpoint-2901/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:72fee7f8f716003a91ed14481748cd48dc0c469f4f0ecd3f7259d7911c900f23
3
+ size 4255140312
checkpoint-2901/processor_config.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "image_processor": {
3
+ "data_format": "channels_first",
4
+ "do_convert_rgb": true,
5
+ "do_normalize": true,
6
+ "do_rescale": true,
7
+ "do_resize": true,
8
+ "image_mean": [
9
+ 0.5,
10
+ 0.5,
11
+ 0.5
12
+ ],
13
+ "image_processor_type": "Qwen2VLImageProcessorFast",
14
+ "image_std": [
15
+ 0.5,
16
+ 0.5,
17
+ 0.5
18
+ ],
19
+ "merge_size": 2,
20
+ "patch_size": 16,
21
+ "resample": 3,
22
+ "rescale_factor": 0.00392156862745098,
23
+ "size": {
24
+ "longest_edge": 16777216,
25
+ "shortest_edge": 65536
26
+ },
27
+ "temporal_patch_size": 2
28
+ },
29
+ "processor_class": "Qwen3VLProcessor",
30
+ "video_processor": {
31
+ "data_format": "channels_first",
32
+ "default_to_square": true,
33
+ "do_convert_rgb": true,
34
+ "do_normalize": true,
35
+ "do_rescale": true,
36
+ "do_resize": true,
37
+ "do_sample_frames": true,
38
+ "fps": 2,
39
+ "image_mean": [
40
+ 0.5,
41
+ 0.5,
42
+ 0.5
43
+ ],
44
+ "image_std": [
45
+ 0.5,
46
+ 0.5,
47
+ 0.5
48
+ ],
49
+ "max_frames": 768,
50
+ "merge_size": 2,
51
+ "min_frames": 4,
52
+ "patch_size": 16,
53
+ "resample": 3,
54
+ "rescale_factor": 0.00392156862745098,
55
+ "return_metadata": false,
56
+ "size": {
57
+ "longest_edge": 25165824,
58
+ "shortest_edge": 4096
59
+ },
60
+ "temporal_patch_size": 2,
61
+ "video_processor_type": "Qwen3VLVideoProcessor"
62
+ }
63
+ }
checkpoint-2901/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
3
+ size 11422650
checkpoint-2901/tokenizer_config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "extra_special_tokens": [
9
+ "<|im_start|>",
10
+ "<|im_end|>",
11
+ "<|object_ref_start|>",
12
+ "<|object_ref_end|>",
13
+ "<|box_start|>",
14
+ "<|box_end|>",
15
+ "<|quad_start|>",
16
+ "<|quad_end|>",
17
+ "<|vision_start|>",
18
+ "<|vision_end|>",
19
+ "<|vision_pad|>",
20
+ "<|image_pad|>",
21
+ "<|video_pad|>"
22
+ ],
23
+ "is_local": false,
24
+ "model_max_length": 262144,
25
+ "pad_token": "<|endoftext|>",
26
+ "processor_class": "Qwen3VLProcessor",
27
+ "split_special_tokens": false,
28
+ "tokenizer_class": "Qwen2Tokenizer",
29
+ "unk_token": null
30
+ }
checkpoint-2901/trainer_state.json ADDED
@@ -0,0 +1,480 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 3.0,
6
+ "eval_steps": 500,
7
+ "global_step": 2901,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.05174197999310107,
14
+ "grad_norm": 20.0,
15
+ "learning_rate": 3.356164383561644e-06,
16
+ "loss": 2.9015182495117187,
17
+ "step": 50
18
+ },
19
+ {
20
+ "epoch": 0.10348395998620213,
21
+ "grad_norm": 10.8125,
22
+ "learning_rate": 6.78082191780822e-06,
23
+ "loss": 1.9724142456054687,
24
+ "step": 100
25
+ },
26
+ {
27
+ "epoch": 0.1552259399793032,
28
+ "grad_norm": 12.875,
29
+ "learning_rate": 9.999970742404346e-06,
30
+ "loss": 1.6598092651367187,
31
+ "step": 150
32
+ },
33
+ {
34
+ "epoch": 0.20696791997240427,
35
+ "grad_norm": 8.5,
36
+ "learning_rate": 9.990871149623236e-06,
37
+ "loss": 1.5588894653320313,
38
+ "step": 200
39
+ },
40
+ {
41
+ "epoch": 0.25870989996550536,
42
+ "grad_norm": 8.25,
43
+ "learning_rate": 9.965551392499958e-06,
44
+ "loss": 1.5084381103515625,
45
+ "step": 250
46
+ },
47
+ {
48
+ "epoch": 0.3104518799586064,
49
+ "grad_norm": 8.4375,
50
+ "learning_rate": 9.924093759398476e-06,
51
+ "loss": 1.4763961791992188,
52
+ "step": 300
53
+ },
54
+ {
55
+ "epoch": 0.3621938599517075,
56
+ "grad_norm": 7.5,
57
+ "learning_rate": 9.86663298624003e-06,
58
+ "loss": 1.45826416015625,
59
+ "step": 350
60
+ },
61
+ {
62
+ "epoch": 0.41393583994480854,
63
+ "grad_norm": 6.90625,
64
+ "learning_rate": 9.793355818615905e-06,
65
+ "loss": 1.4410775756835938,
66
+ "step": 400
67
+ },
68
+ {
69
+ "epoch": 0.46567781993790963,
70
+ "grad_norm": 7.09375,
71
+ "learning_rate": 9.7045004048705e-06,
72
+ "loss": 1.4302684020996095,
73
+ "step": 450
74
+ },
75
+ {
76
+ "epoch": 0.5174197999310107,
77
+ "grad_norm": 6.46875,
78
+ "learning_rate": 9.600355522127174e-06,
79
+ "loss": 1.4159922790527344,
80
+ "step": 500
81
+ },
82
+ {
83
+ "epoch": 0.5174197999310107,
84
+ "eval_loss": 1.3360999822616577,
85
+ "eval_runtime": 48.5888,
86
+ "eval_samples_per_second": 51.452,
87
+ "eval_steps_per_second": 0.556,
88
+ "step": 500
89
+ },
90
+ {
91
+ "epoch": 0.5691617799241118,
92
+ "grad_norm": 6.46875,
93
+ "learning_rate": 9.481259637772251e-06,
94
+ "loss": 1.408507080078125,
95
+ "step": 550
96
+ },
97
+ {
98
+ "epoch": 0.6209037599172128,
99
+ "grad_norm": 8.0,
100
+ "learning_rate": 9.347599809447331e-06,
101
+ "loss": 1.406485595703125,
102
+ "step": 600
103
+ },
104
+ {
105
+ "epoch": 0.672645739910314,
106
+ "grad_norm": 6.59375,
107
+ "learning_rate": 9.199810427124904e-06,
108
+ "loss": 1.3980792236328126,
109
+ "step": 650
110
+ },
111
+ {
112
+ "epoch": 0.724387719903415,
113
+ "grad_norm": 6.1875,
114
+ "learning_rate": 9.03837180135546e-06,
115
+ "loss": 1.3900822448730468,
116
+ "step": 700
117
+ },
118
+ {
119
+ "epoch": 0.776129699896516,
120
+ "grad_norm": 6.375,
121
+ "learning_rate": 8.86380860227426e-06,
122
+ "loss": 1.377584228515625,
123
+ "step": 750
124
+ },
125
+ {
126
+ "epoch": 0.8278716798896171,
127
+ "grad_norm": 6.46875,
128
+ "learning_rate": 8.676688154440947e-06,
129
+ "loss": 1.3831480407714845,
130
+ "step": 800
131
+ },
132
+ {
133
+ "epoch": 0.8796136598827182,
134
+ "grad_norm": 6.84375,
135
+ "learning_rate": 8.477618593053693e-06,
136
+ "loss": 1.3737605285644532,
137
+ "step": 850
138
+ },
139
+ {
140
+ "epoch": 0.9313556398758193,
141
+ "grad_norm": 6.5625,
142
+ "learning_rate": 8.26724688753015e-06,
143
+ "loss": 1.3781863403320314,
144
+ "step": 900
145
+ },
146
+ {
147
+ "epoch": 0.9830976198689203,
148
+ "grad_norm": 6.0,
149
+ "learning_rate": 8.04625673887846e-06,
150
+ "loss": 1.3639559936523438,
151
+ "step": 950
152
+ },
153
+ {
154
+ "epoch": 1.0341497067954466,
155
+ "grad_norm": 5.84375,
156
+ "learning_rate": 7.815366357691863e-06,
157
+ "loss": 1.3568319702148437,
158
+ "step": 1000
159
+ },
160
+ {
161
+ "epoch": 1.0341497067954466,
162
+ "eval_loss": 1.2891921997070312,
163
+ "eval_runtime": 36.787,
164
+ "eval_samples_per_second": 67.959,
165
+ "eval_steps_per_second": 0.734,
166
+ "step": 1000
167
+ },
168
+ {
169
+ "epoch": 1.0858916867885477,
170
+ "grad_norm": 5.84375,
171
+ "learning_rate": 7.575326129988268e-06,
172
+ "loss": 1.3579391479492187,
173
+ "step": 1050
174
+ },
175
+ {
176
+ "epoch": 1.137633666781649,
177
+ "grad_norm": 6.15625,
178
+ "learning_rate": 7.326916178480796e-06,
179
+ "loss": 1.353500213623047,
180
+ "step": 1100
181
+ },
182
+ {
183
+ "epoch": 1.1893756467747498,
184
+ "grad_norm": 5.625,
185
+ "learning_rate": 7.070943827205048e-06,
186
+ "loss": 1.3528330993652344,
187
+ "step": 1150
188
+ },
189
+ {
190
+ "epoch": 1.241117626767851,
191
+ "grad_norm": 5.96875,
192
+ "learning_rate": 6.808240977742987e-06,
193
+ "loss": 1.3461532592773438,
194
+ "step": 1200
195
+ },
196
+ {
197
+ "epoch": 1.2928596067609521,
198
+ "grad_norm": 5.28125,
199
+ "learning_rate": 6.539661405570615e-06,
200
+ "loss": 1.3481027221679687,
201
+ "step": 1250
202
+ },
203
+ {
204
+ "epoch": 1.344601586754053,
205
+ "grad_norm": 6.09375,
206
+ "learning_rate": 6.266077985316224e-06,
207
+ "loss": 1.348444061279297,
208
+ "step": 1300
209
+ },
210
+ {
211
+ "epoch": 1.3963435667471542,
212
+ "grad_norm": 6.1875,
213
+ "learning_rate": 5.988379853947046e-06,
214
+ "loss": 1.3519085693359374,
215
+ "step": 1350
216
+ },
217
+ {
218
+ "epoch": 1.4480855467402551,
219
+ "grad_norm": 5.375,
220
+ "learning_rate": 5.7074695211038385e-06,
221
+ "loss": 1.3478456115722657,
222
+ "step": 1400
223
+ },
224
+ {
225
+ "epoch": 1.4998275267333563,
226
+ "grad_norm": 5.84375,
227
+ "learning_rate": 5.424259935974736e-06,
228
+ "loss": 1.3487643432617187,
229
+ "step": 1450
230
+ },
231
+ {
232
+ "epoch": 1.5515695067264574,
233
+ "grad_norm": 6.09375,
234
+ "learning_rate": 5.139671520240915e-06,
235
+ "loss": 1.3462490844726562,
236
+ "step": 1500
237
+ },
238
+ {
239
+ "epoch": 1.5515695067264574,
240
+ "eval_loss": 1.2737045288085938,
241
+ "eval_runtime": 35.6507,
242
+ "eval_samples_per_second": 70.125,
243
+ "eval_steps_per_second": 0.757,
244
+ "step": 1500
245
+ },
246
+ {
247
+ "epoch": 1.6033114867195586,
248
+ "grad_norm": 10.125,
249
+ "learning_rate": 4.854629176736911e-06,
250
+ "loss": 1.34809326171875,
251
+ "step": 1550
252
+ },
253
+ {
254
+ "epoch": 1.6550534667126595,
255
+ "grad_norm": 5.84375,
256
+ "learning_rate": 4.570059283547328e-06,
257
+ "loss": 1.343378448486328,
258
+ "step": 1600
259
+ },
260
+ {
261
+ "epoch": 1.7067954467057604,
262
+ "grad_norm": 6.0625,
263
+ "learning_rate": 4.286886683309053e-06,
264
+ "loss": 1.341541748046875,
265
+ "step": 1650
266
+ },
267
+ {
268
+ "epoch": 1.7585374266988616,
269
+ "grad_norm": 5.40625,
270
+ "learning_rate": 4.0060316775036025e-06,
271
+ "loss": 1.3389480590820313,
272
+ "step": 1700
273
+ },
274
+ {
275
+ "epoch": 1.8102794066919627,
276
+ "grad_norm": 5.75,
277
+ "learning_rate": 3.728407035508089e-06,
278
+ "loss": 1.3394900512695314,
279
+ "step": 1750
280
+ },
281
+ {
282
+ "epoch": 1.862021386685064,
283
+ "grad_norm": 5.8125,
284
+ "learning_rate": 3.4549150281252635e-06,
285
+ "loss": 1.3372491455078126,
286
+ "step": 1800
287
+ },
288
+ {
289
+ "epoch": 1.913763366678165,
290
+ "grad_norm": 5.71875,
291
+ "learning_rate": 3.1864444952335823e-06,
292
+ "loss": 1.3394888305664063,
293
+ "step": 1850
294
+ },
295
+ {
296
+ "epoch": 1.965505346671266,
297
+ "grad_norm": 5.59375,
298
+ "learning_rate": 2.9238679570873217e-06,
299
+ "loss": 1.34095458984375,
300
+ "step": 1900
301
+ },
302
+ {
303
+ "epoch": 2.0165574335977925,
304
+ "grad_norm": 6.0625,
305
+ "learning_rate": 2.66803877865493e-06,
306
+ "loss": 1.3422109985351562,
307
+ "step": 1950
308
+ },
309
+ {
310
+ "epoch": 2.068299413590893,
311
+ "grad_norm": 5.65625,
312
+ "learning_rate": 2.4197883962114164e-06,
313
+ "loss": 1.3344174194335938,
314
+ "step": 2000
315
+ },
316
+ {
317
+ "epoch": 2.068299413590893,
318
+ "eval_loss": 1.269867181777954,
319
+ "eval_runtime": 37.0763,
320
+ "eval_samples_per_second": 67.429,
321
+ "eval_steps_per_second": 0.728,
322
+ "step": 2000
323
+ },
324
+ {
325
+ "epoch": 2.1200413935839943,
326
+ "grad_norm": 5.84375,
327
+ "learning_rate": 2.1799236151982623e-06,
328
+ "loss": 1.3353939819335938,
329
+ "step": 2050
330
+ },
331
+ {
332
+ "epoch": 2.1717833735770955,
333
+ "grad_norm": 5.5,
334
+ "learning_rate": 1.9492239881327103e-06,
335
+ "loss": 1.3307119750976562,
336
+ "step": 2100
337
+ },
338
+ {
339
+ "epoch": 2.2235253535701966,
340
+ "grad_norm": 6.65625,
341
+ "learning_rate": 1.7284392810881568e-06,
342
+ "loss": 1.3394575500488282,
343
+ "step": 2150
344
+ },
345
+ {
346
+ "epoch": 2.275267333563298,
347
+ "grad_norm": 6.59375,
348
+ "learning_rate": 1.5182870369794838e-06,
349
+ "loss": 1.3347157287597655,
350
+ "step": 2200
351
+ },
352
+ {
353
+ "epoch": 2.327009313556399,
354
+ "grad_norm": 5.25,
355
+ "learning_rate": 1.3194502435725787e-06,
356
+ "loss": 1.3337785339355468,
357
+ "step": 2250
358
+ },
359
+ {
360
+ "epoch": 2.3787512935494997,
361
+ "grad_norm": 5.78125,
362
+ "learning_rate": 1.1325751137969381e-06,
363
+ "loss": 1.3284304809570313,
364
+ "step": 2300
365
+ },
366
+ {
367
+ "epoch": 2.430493273542601,
368
+ "grad_norm": 6.09375,
369
+ "learning_rate": 9.582689855752515e-07,
370
+ "loss": 1.3337535095214843,
371
+ "step": 2350
372
+ },
373
+ {
374
+ "epoch": 2.482235253535702,
375
+ "grad_norm": 5.875,
376
+ "learning_rate": 7.970983479954631e-07,
377
+ "loss": 1.3316661071777345,
378
+ "step": 2400
379
+ },
380
+ {
381
+ "epoch": 2.533977233528803,
382
+ "grad_norm": 5.6875,
383
+ "learning_rate": 6.495870002401822e-07,
384
+ "loss": 1.3348265075683594,
385
+ "step": 2450
386
+ },
387
+ {
388
+ "epoch": 2.5857192135219043,
389
+ "grad_norm": 5.59375,
390
+ "learning_rate": 5.162143492568511e-07,
391
+ "loss": 1.3390269470214844,
392
+ "step": 2500
393
+ },
394
+ {
395
+ "epoch": 2.5857192135219043,
396
+ "eval_loss": 1.269131064414978,
397
+ "eval_runtime": 36.297,
398
+ "eval_samples_per_second": 68.876,
399
+ "eval_steps_per_second": 0.744,
400
+ "step": 2500
401
+ },
402
+ {
403
+ "epoch": 2.637461193515005,
404
+ "grad_norm": 5.8125,
405
+ "learning_rate": 3.9741385170120273e-07,
406
+ "loss": 1.3350772094726562,
407
+ "step": 2550
408
+ },
409
+ {
410
+ "epoch": 2.689203173508106,
411
+ "grad_norm": 5.75,
412
+ "learning_rate": 2.935716052176163e-07,
413
+ "loss": 1.336797637939453,
414
+ "step": 2600
415
+ },
416
+ {
417
+ "epoch": 2.7409451535012073,
418
+ "grad_norm": 5.84375,
419
+ "learning_rate": 2.050250936346937e-07,
420
+ "loss": 1.336610107421875,
421
+ "step": 2650
422
+ },
423
+ {
424
+ "epoch": 2.7926871334943084,
425
+ "grad_norm": 5.5625,
426
+ "learning_rate": 1.3206209015409532e-07,
427
+ "loss": 1.3384811401367187,
428
+ "step": 2700
429
+ },
430
+ {
431
+ "epoch": 2.8444291134874096,
432
+ "grad_norm": 6.03125,
433
+ "learning_rate": 7.491972209725807e-08,
434
+ "loss": 1.3393882751464843,
435
+ "step": 2750
436
+ },
437
+ {
438
+ "epoch": 2.8961710934805103,
439
+ "grad_norm": 7.21875,
440
+ "learning_rate": 3.3783700249533925e-08,
441
+ "loss": 1.3347865295410157,
442
+ "step": 2800
443
+ },
444
+ {
445
+ "epoch": 2.947913073473612,
446
+ "grad_norm": 5.71875,
447
+ "learning_rate": 8.787715306344258e-09,
448
+ "loss": 1.342543487548828,
449
+ "step": 2850
450
+ },
451
+ {
452
+ "epoch": 2.9996550534667126,
453
+ "grad_norm": 6.15625,
454
+ "learning_rate": 1.3003382891563221e-11,
455
+ "loss": 1.3331295776367187,
456
+ "step": 2900
457
+ }
458
+ ],
459
+ "logging_steps": 50,
460
+ "max_steps": 2901,
461
+ "num_input_tokens_seen": 0,
462
+ "num_train_epochs": 3,
463
+ "save_steps": 1000,
464
+ "stateful_callbacks": {
465
+ "TrainerControl": {
466
+ "args": {
467
+ "should_epoch_stop": false,
468
+ "should_evaluate": false,
469
+ "should_log": false,
470
+ "should_save": true,
471
+ "should_training_stop": true
472
+ },
473
+ "attributes": {}
474
+ }
475
+ },
476
+ "total_flos": 6.833997897629958e+18,
477
+ "train_batch_size": 24,
478
+ "trial_name": null,
479
+ "trial_params": null
480
+ }
checkpoint-2901/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d14c537eb87fc9a2e52b16edd6c820da00e0bdfe05198c576d1f6eefa7bcf0d
3
+ size 5201
config.json ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen3VLForConditionalGeneration"
4
+ ],
5
+ "bos_token_id": null,
6
+ "dtype": "bfloat16",
7
+ "eos_token_id": 151645,
8
+ "image_token_id": 151655,
9
+ "model_type": "qwen3_vl",
10
+ "pad_token_id": 151643,
11
+ "text_config": {
12
+ "attention_bias": false,
13
+ "attention_dropout": 0.0,
14
+ "bos_token_id": 151643,
15
+ "dtype": "bfloat16",
16
+ "eos_token_id": 151645,
17
+ "head_dim": 128,
18
+ "hidden_act": "silu",
19
+ "hidden_size": 2048,
20
+ "initializer_range": 0.02,
21
+ "intermediate_size": 6144,
22
+ "max_position_embeddings": 262144,
23
+ "model_type": "qwen3_vl_text",
24
+ "num_attention_heads": 16,
25
+ "num_hidden_layers": 28,
26
+ "num_key_value_heads": 8,
27
+ "pad_token_id": null,
28
+ "rms_norm_eps": 1e-06,
29
+ "rope_parameters": {
30
+ "mrope_interleaved": true,
31
+ "mrope_section": [
32
+ 24,
33
+ 20,
34
+ 20
35
+ ],
36
+ "rope_theta": 5000000,
37
+ "rope_type": "default"
38
+ },
39
+ "tie_word_embeddings": true,
40
+ "use_cache": true,
41
+ "vocab_size": 151936
42
+ },
43
+ "tie_word_embeddings": true,
44
+ "transformers_version": "5.0.0",
45
+ "use_cache": false,
46
+ "video_token_id": 151656,
47
+ "vision_config": {
48
+ "deepstack_visual_indexes": [
49
+ 5,
50
+ 11,
51
+ 17
52
+ ],
53
+ "depth": 24,
54
+ "dtype": "bfloat16",
55
+ "hidden_act": "gelu_pytorch_tanh",
56
+ "hidden_size": 1024,
57
+ "in_channels": 3,
58
+ "initializer_range": 0.02,
59
+ "intermediate_size": 4096,
60
+ "model_type": "qwen3_vl",
61
+ "num_heads": 16,
62
+ "num_position_embeddings": 2304,
63
+ "out_hidden_size": 2048,
64
+ "patch_size": 16,
65
+ "spatial_merge_size": 2,
66
+ "temporal_patch_size": 2
67
+ },
68
+ "vision_end_token_id": 151653,
69
+ "vision_start_token_id": 151652
70
+ }
generation_config.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "do_sample": true,
3
+ "eos_token_id": [
4
+ 151645,
5
+ 151645,
6
+ 151643
7
+ ],
8
+ "pad_token_id": 151643,
9
+ "repetition_penalty": 1.0,
10
+ "temperature": 0.7,
11
+ "top_k": 20,
12
+ "top_p": 0.8,
13
+ "transformers_version": "5.0.0"
14
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:72fee7f8f716003a91ed14481748cd48dc0c469f4f0ecd3f7259d7911c900f23
3
+ size 4255140312
processor_config.json ADDED
@@ -0,0 +1,63 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "image_processor": {
3
+ "data_format": "channels_first",
4
+ "do_convert_rgb": true,
5
+ "do_normalize": true,
6
+ "do_rescale": true,
7
+ "do_resize": true,
8
+ "image_mean": [
9
+ 0.5,
10
+ 0.5,
11
+ 0.5
12
+ ],
13
+ "image_processor_type": "Qwen2VLImageProcessorFast",
14
+ "image_std": [
15
+ 0.5,
16
+ 0.5,
17
+ 0.5
18
+ ],
19
+ "merge_size": 2,
20
+ "patch_size": 16,
21
+ "resample": 3,
22
+ "rescale_factor": 0.00392156862745098,
23
+ "size": {
24
+ "longest_edge": 16777216,
25
+ "shortest_edge": 65536
26
+ },
27
+ "temporal_patch_size": 2
28
+ },
29
+ "processor_class": "Qwen3VLProcessor",
30
+ "video_processor": {
31
+ "data_format": "channels_first",
32
+ "default_to_square": true,
33
+ "do_convert_rgb": true,
34
+ "do_normalize": true,
35
+ "do_rescale": true,
36
+ "do_resize": true,
37
+ "do_sample_frames": true,
38
+ "fps": 2,
39
+ "image_mean": [
40
+ 0.5,
41
+ 0.5,
42
+ 0.5
43
+ ],
44
+ "image_std": [
45
+ 0.5,
46
+ 0.5,
47
+ 0.5
48
+ ],
49
+ "max_frames": 768,
50
+ "merge_size": 2,
51
+ "min_frames": 4,
52
+ "patch_size": 16,
53
+ "resample": 3,
54
+ "rescale_factor": 0.00392156862745098,
55
+ "return_metadata": false,
56
+ "size": {
57
+ "longest_edge": 25165824,
58
+ "shortest_edge": 4096
59
+ },
60
+ "temporal_patch_size": 2,
61
+ "video_processor_type": "Qwen3VLVideoProcessor"
62
+ }
63
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
3
+ size 11422650
tokenizer_config.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "extra_special_tokens": [
9
+ "<|im_start|>",
10
+ "<|im_end|>",
11
+ "<|object_ref_start|>",
12
+ "<|object_ref_end|>",
13
+ "<|box_start|>",
14
+ "<|box_end|>",
15
+ "<|quad_start|>",
16
+ "<|quad_end|>",
17
+ "<|vision_start|>",
18
+ "<|vision_end|>",
19
+ "<|vision_pad|>",
20
+ "<|image_pad|>",
21
+ "<|video_pad|>"
22
+ ],
23
+ "is_local": false,
24
+ "model_max_length": 262144,
25
+ "pad_token": "<|endoftext|>",
26
+ "processor_class": "Qwen3VLProcessor",
27
+ "split_special_tokens": false,
28
+ "tokenizer_class": "Qwen2Tokenizer",
29
+ "unk_token": null
30
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d14c537eb87fc9a2e52b16edd6c820da00e0bdfe05198c576d1f6eefa7bcf0d
3
+ size 5201