Escapist-X commited on
Commit
02660d7
·
verified ·
1 Parent(s): 5bb8934

Add files using upload-large-folder tool

Browse files
Files changed (50) hide show
  1. README.md +67 -0
  2. checkpoint-1050/chat_template.jinja +7 -0
  3. checkpoint-1050/config.json +143 -0
  4. checkpoint-1050/generation_config.json +11 -0
  5. checkpoint-1050/model.safetensors.index.json +0 -0
  6. checkpoint-1050/trainer_state.json +0 -0
  7. checkpoint-1150/vocab.json +0 -0
  8. checkpoint-1200/added_tokens.json +24 -0
  9. checkpoint-1200/chat_template.jinja +7 -0
  10. checkpoint-1200/generation_config.json +11 -0
  11. checkpoint-1200/merges.txt +0 -0
  12. checkpoint-1200/special_tokens_map.json +31 -0
  13. checkpoint-1200/trainer_state.json +0 -0
  14. checkpoint-1200/video_preprocessor_config.json +44 -0
  15. checkpoint-1200/vocab.json +0 -0
  16. checkpoint-800/added_tokens.json +24 -0
  17. checkpoint-800/chat_template.jinja +7 -0
  18. checkpoint-800/generation_config.json +11 -0
  19. checkpoint-800/merges.txt +0 -0
  20. checkpoint-800/model.safetensors.index.json +0 -0
  21. checkpoint-800/preprocessor_config.json +37 -0
  22. checkpoint-800/special_tokens_map.json +31 -0
  23. checkpoint-800/tokenizer_config.json +212 -0
  24. checkpoint-800/trainer_state.json +2334 -0
  25. checkpoint-800/video_preprocessor_config.json +44 -0
  26. checkpoint-800/vocab.json +0 -0
  27. checkpoint-850/added_tokens.json +24 -0
  28. checkpoint-850/chat_template.jinja +7 -0
  29. checkpoint-850/config.json +143 -0
  30. checkpoint-850/generation_config.json +11 -0
  31. checkpoint-850/merges.txt +0 -0
  32. checkpoint-850/model.safetensors.index.json +0 -0
  33. checkpoint-850/preprocessor_config.json +37 -0
  34. checkpoint-850/special_tokens_map.json +31 -0
  35. checkpoint-850/tokenizer_config.json +212 -0
  36. checkpoint-850/trainer_state.json +2472 -0
  37. checkpoint-850/video_preprocessor_config.json +44 -0
  38. checkpoint-850/vocab.json +0 -0
  39. checkpoint-950/added_tokens.json +24 -0
  40. checkpoint-950/chat_template.jinja +7 -0
  41. checkpoint-950/config.json +143 -0
  42. checkpoint-950/generation_config.json +11 -0
  43. checkpoint-950/merges.txt +0 -0
  44. checkpoint-950/model.safetensors.index.json +0 -0
  45. checkpoint-950/preprocessor_config.json +37 -0
  46. checkpoint-950/special_tokens_map.json +31 -0
  47. checkpoint-950/tokenizer_config.json +212 -0
  48. checkpoint-950/trainer_state.json +2748 -0
  49. checkpoint-950/video_preprocessor_config.json +44 -0
  50. checkpoint-950/vocab.json +0 -0
README.md ADDED
@@ -0,0 +1,67 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ model_name: out_dir_grpo_dr_geo_plus
4
+ tags:
5
+ - generated_from_trainer
6
+ - trl
7
+ - grpo
8
+ licence: license
9
+ ---
10
+
11
+ # Model Card for out_dir_grpo_dr_geo_plus
12
+
13
+ This model is a fine-tuned version of [None](https://huggingface.co/None).
14
+ It has been trained using [TRL](https://github.com/huggingface/trl).
15
+
16
+ ## Quick start
17
+
18
+ ```python
19
+ from transformers import pipeline
20
+
21
+ question = "If you had a time machine, but could only go to the past or the future once and never return, which would you choose and why?"
22
+ generator = pipeline("text-generation", model="None", device="cuda")
23
+ output = generator([{"role": "user", "content": question}], max_new_tokens=128, return_full_text=False)[0]
24
+ print(output["generated_text"])
25
+ ```
26
+
27
+ ## Training procedure
28
+
29
+
30
+
31
+
32
+ This model was trained with GRPO, a method introduced in [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://huggingface.co/papers/2402.03300).
33
+
34
+ ### Framework versions
35
+
36
+ - TRL: 0.24.0.dev0
37
+ - Transformers: 4.56.2
38
+ - Pytorch: 2.8.0
39
+ - Datasets: 4.0.0
40
+ - Tokenizers: 0.22.0
41
+
42
+ ## Citations
43
+
44
+ Cite GRPO as:
45
+
46
+ ```bibtex
47
+ @article{shao2024deepseekmath,
48
+ title = {{DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models}},
49
+ author = {Zhihong Shao and Peiyi Wang and Qihao Zhu and Runxin Xu and Junxiao Song and Mingchuan Zhang and Y. K. Li and Y. Wu and Daya Guo},
50
+ year = 2024,
51
+ eprint = {arXiv:2402.03300},
52
+ }
53
+
54
+ ```
55
+
56
+ Cite TRL as:
57
+
58
+ ```bibtex
59
+ @misc{vonwerra2022trl,
60
+ title = {{TRL: Transformer Reinforcement Learning}},
61
+ author = {Leandro von Werra and Younes Belkada and Lewis Tunstall and Edward Beeching and Tristan Thrush and Nathan Lambert and Shengyi Huang and Kashif Rasul and Quentin Gallou{\'e}dec},
62
+ year = 2020,
63
+ journal = {GitHub repository},
64
+ publisher = {GitHub},
65
+ howpublished = {\url{https://github.com/huggingface/trl}}
66
+ }
67
+ ```
checkpoint-1050/chat_template.jinja ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
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
2
+ You are a helpful assistant.<|im_end|>
3
+ {% endif %}<|im_start|>{{ message['role'] }}
4
+ {% if message['content'] is string %}{{ message['content'] }}<|im_end|>
5
+ {% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>
6
+ {% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant
7
+ {% endif %}
checkpoint-1050/config.json ADDED
@@ -0,0 +1,143 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen2_5_PVLForConditionalGeneration"
4
+ ],
5
+ "attention_dropout": 0.0,
6
+ "dtype": "bfloat16",
7
+ "eos_token_id": 151645,
8
+ "hidden_act": "silu",
9
+ "hidden_size": 2048,
10
+ "image_token_id": 151655,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 11008,
13
+ "max_position_embeddings": 128000,
14
+ "max_window_layers": 70,
15
+ "model_type": "qwen2_5_vl",
16
+ "num_attention_heads": 16,
17
+ "num_hidden_layers": 36,
18
+ "num_key_value_heads": 2,
19
+ "pad_token_id": 151643,
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
+ "architectures": [
34
+ "Qwen2_5_VLForConditionalGeneration"
35
+ ],
36
+ "attention_dropout": 0.0,
37
+ "bos_token_id": 151643,
38
+ "dtype": "bfloat16",
39
+ "eos_token_id": 151645,
40
+ "hidden_act": "silu",
41
+ "hidden_size": 2048,
42
+ "image_token_id": null,
43
+ "initializer_range": 0.02,
44
+ "intermediate_size": 11008,
45
+ "layer_types": [
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
+ "full_attention",
74
+ "full_attention",
75
+ "full_attention",
76
+ "full_attention",
77
+ "full_attention",
78
+ "full_attention",
79
+ "full_attention",
80
+ "full_attention",
81
+ "full_attention"
82
+ ],
83
+ "max_position_embeddings": 128000,
84
+ "max_window_layers": 70,
85
+ "model_type": "qwen2_5_vl_text",
86
+ "num_attention_heads": 16,
87
+ "num_hidden_layers": 36,
88
+ "num_key_value_heads": 2,
89
+ "rms_norm_eps": 1e-06,
90
+ "rope_scaling": {
91
+ "mrope_section": [
92
+ 16,
93
+ 24,
94
+ 24
95
+ ],
96
+ "rope_type": "default",
97
+ "type": "default"
98
+ },
99
+ "rope_theta": 1000000.0,
100
+ "sliding_window": null,
101
+ "tie_word_embeddings": true,
102
+ "use_cache": true,
103
+ "use_sliding_window": false,
104
+ "video_token_id": null,
105
+ "vision_end_token_id": 151653,
106
+ "vision_start_token_id": 151652,
107
+ "vision_token_id": 151654,
108
+ "vocab_size": 151936
109
+ },
110
+ "transformers_version": "4.56.2",
111
+ "use_cache": true,
112
+ "use_sliding_window": false,
113
+ "video_token_id": 151656,
114
+ "vision_config": {
115
+ "depth": 32,
116
+ "dtype": "bfloat16",
117
+ "fullatt_block_indexes": [
118
+ 7,
119
+ 15,
120
+ 23,
121
+ 31
122
+ ],
123
+ "hidden_act": "silu",
124
+ "hidden_size": 1280,
125
+ "in_channels": 3,
126
+ "in_chans": 3,
127
+ "initializer_range": 0.02,
128
+ "intermediate_size": 3420,
129
+ "model_type": "qwen2_5_vl",
130
+ "num_heads": 16,
131
+ "out_hidden_size": 2048,
132
+ "patch_size": 14,
133
+ "spatial_merge_size": 2,
134
+ "spatial_patch_size": 14,
135
+ "temporal_patch_size": 2,
136
+ "tokens_per_second": 2,
137
+ "window_size": 112
138
+ },
139
+ "vision_end_token_id": 151653,
140
+ "vision_start_token_id": 151652,
141
+ "vision_token_id": 151654,
142
+ "vocab_size": 151936
143
+ }
checkpoint-1050/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": "4.56.2"
11
+ }
checkpoint-1050/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1050/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1150/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1200/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
+ }
checkpoint-1200/chat_template.jinja ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
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
2
+ You are a helpful assistant.<|im_end|>
3
+ {% endif %}<|im_start|>{{ message['role'] }}
4
+ {% if message['content'] is string %}{{ message['content'] }}<|im_end|>
5
+ {% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>
6
+ {% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant
7
+ {% endif %}
checkpoint-1200/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": "4.56.2"
11
+ }
checkpoint-1200/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1200/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
+ }
checkpoint-1200/trainer_state.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-1200/video_preprocessor_config.json ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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_pad": null,
10
+ "do_rescale": true,
11
+ "do_resize": true,
12
+ "do_sample_frames": false,
13
+ "fps": null,
14
+ "image_mean": [
15
+ 0.48145466,
16
+ 0.4578275,
17
+ 0.40821073
18
+ ],
19
+ "image_std": [
20
+ 0.26862954,
21
+ 0.26130258,
22
+ 0.27577711
23
+ ],
24
+ "input_data_format": null,
25
+ "max_frames": 768,
26
+ "max_pixels": 12845056,
27
+ "merge_size": 2,
28
+ "min_frames": 4,
29
+ "min_pixels": 3136,
30
+ "num_frames": 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
+ "size_divisor": null,
41
+ "temporal_patch_size": 2,
42
+ "video_metadata": null,
43
+ "video_processor_type": "Qwen2VLVideoProcessor"
44
+ }
checkpoint-1200/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-800/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
+ }
checkpoint-800/chat_template.jinja ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
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
2
+ You are a helpful assistant.<|im_end|>
3
+ {% endif %}<|im_start|>{{ message['role'] }}
4
+ {% if message['content'] is string %}{{ message['content'] }}<|im_end|>
5
+ {% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>
6
+ {% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant
7
+ {% endif %}
checkpoint-800/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": "4.56.2"
11
+ }
checkpoint-800/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-800/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-800/preprocessor_config.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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_rescale": true,
11
+ "do_resize": true,
12
+ "image_mean": [
13
+ 0.48145466,
14
+ 0.4578275,
15
+ 0.40821073
16
+ ],
17
+ "image_processor_type": "Qwen2VLImageProcessorFast",
18
+ "image_std": [
19
+ 0.26862954,
20
+ 0.26130258,
21
+ 0.27577711
22
+ ],
23
+ "input_data_format": null,
24
+ "max_pixels": 12845056,
25
+ "merge_size": 2,
26
+ "min_pixels": 3136,
27
+ "patch_size": 14,
28
+ "processor_class": "Qwen2_5_VLProcessor",
29
+ "resample": 3,
30
+ "rescale_factor": 0.00392156862745098,
31
+ "return_tensors": null,
32
+ "size": {
33
+ "longest_edge": 12845056,
34
+ "shortest_edge": 3136
35
+ },
36
+ "temporal_patch_size": 2
37
+ }
checkpoint-800/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
+ }
checkpoint-800/tokenizer_config.json ADDED
@@ -0,0 +1,212 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "max_length": null,
203
+ "model_max_length": 131072,
204
+ "pad_to_multiple_of": null,
205
+ "pad_token": "<|endoftext|>",
206
+ "pad_token_type_id": 0,
207
+ "padding_side": "left",
208
+ "processor_class": "Qwen2_5_VLProcessor",
209
+ "split_special_tokens": false,
210
+ "tokenizer_class": "Qwen2Tokenizer",
211
+ "unk_token": null
212
+ }
checkpoint-800/trainer_state.json ADDED
@@ -0,0 +1,2334 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.2822865208186309,
6
+ "eval_steps": 500,
7
+ "global_step": 800,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "3d_bbox": 0.6610370799899101,
14
+ "adv": 0.000926484264937244,
15
+ "bbox": 0.8075725510716438,
16
+ "completions/clipped_ratio": 0.0,
17
+ "completions/max_length": 1592.875,
18
+ "completions/max_terminated_length": 1592.875,
19
+ "completions/mean_length": 1573.1640625,
20
+ "completions/mean_terminated_length": 1573.1640625,
21
+ "completions/min_length": 1554.875,
22
+ "completions/min_terminated_length": 1554.875,
23
+ "entropy": 0.2564338077791035,
24
+ "epoch": 0.0028228652081863093,
25
+ "geo": 0.7611365988850594,
26
+ "grad_norm": 0.439453125,
27
+ "learning_rate": 7e-06,
28
+ "loss": -0.0002,
29
+ "num_tokens": 943810.0,
30
+ "point_2d": 0.875,
31
+ "point_3d": 0.638671875,
32
+ "reward": 0.7455703690648079,
33
+ "step": 8
34
+ },
35
+ {
36
+ "3d_bbox": 0.6985021606087685,
37
+ "adv": 0.0008333708265126916,
38
+ "bbox": 0.9951911941170692,
39
+ "completions/clipped_ratio": 0.0,
40
+ "completions/max_length": 1594.25,
41
+ "completions/max_terminated_length": 1594.25,
42
+ "completions/mean_length": 1573.64453125,
43
+ "completions/mean_terminated_length": 1573.64453125,
44
+ "completions/min_length": 1552.375,
45
+ "completions/min_terminated_length": 1552.375,
46
+ "entropy": 0.2595143502112478,
47
+ "epoch": 0.0056457304163726185,
48
+ "geo": 0.8901385143399239,
49
+ "grad_norm": 0.341796875,
50
+ "learning_rate": 9.974874371859297e-06,
51
+ "loss": 0.0564,
52
+ "num_tokens": 1888015.0,
53
+ "point_2d": 1.0,
54
+ "point_3d": 0.5947265625,
55
+ "reward": 0.8221049904823303,
56
+ "step": 16
57
+ },
58
+ {
59
+ "3d_bbox": 0.6706115230917931,
60
+ "adv": -0.11667590066201683,
61
+ "bbox": 0.9641365110874176,
62
+ "completions/clipped_ratio": 0.0,
63
+ "completions/max_length": 1588.75,
64
+ "completions/max_terminated_length": 1588.75,
65
+ "completions/mean_length": 1568.69921875,
66
+ "completions/mean_terminated_length": 1568.69921875,
67
+ "completions/min_length": 1549.625,
68
+ "completions/min_terminated_length": 1549.625,
69
+ "entropy": 0.2620172018650919,
70
+ "epoch": 0.008468595624558928,
71
+ "geo": 0.8719085603952408,
72
+ "grad_norm": 0.31640625,
73
+ "learning_rate": 9.934673366834172e-06,
74
+ "loss": 0.0434,
75
+ "num_tokens": 2830754.0,
76
+ "point_2d": 0.96484375,
77
+ "point_3d": 0.7060546875,
78
+ "reward": 0.8264116197824478,
79
+ "step": 24
80
+ },
81
+ {
82
+ "3d_bbox": 0.6768141128122807,
83
+ "adv": 0.00011848201484099263,
84
+ "bbox": 0.8640391379594803,
85
+ "completions/clipped_ratio": 0.0,
86
+ "completions/max_length": 1599.875,
87
+ "completions/max_terminated_length": 1599.875,
88
+ "completions/mean_length": 1579.40625,
89
+ "completions/mean_terminated_length": 1579.40625,
90
+ "completions/min_length": 1563.0,
91
+ "completions/min_terminated_length": 1563.0,
92
+ "entropy": 0.26498199813067913,
93
+ "epoch": 0.011291460832745237,
94
+ "geo": 0.8661364912986755,
95
+ "grad_norm": 0.36328125,
96
+ "learning_rate": 9.894472361809046e-06,
97
+ "loss": -0.0005,
98
+ "num_tokens": 3776674.0,
99
+ "point_2d": 0.8642578125,
100
+ "point_3d": 0.603515625,
101
+ "reward": 0.752156674861908,
102
+ "step": 32
103
+ },
104
+ {
105
+ "3d_bbox": 0.6714089289307594,
106
+ "adv": -0.000722882674608627,
107
+ "bbox": 0.9624817371368408,
108
+ "completions/clipped_ratio": 0.0,
109
+ "completions/max_length": 1593.75,
110
+ "completions/max_terminated_length": 1593.75,
111
+ "completions/mean_length": 1575.52734375,
112
+ "completions/mean_terminated_length": 1575.52734375,
113
+ "completions/min_length": 1559.875,
114
+ "completions/min_terminated_length": 1559.875,
115
+ "entropy": 0.2738626003265381,
116
+ "epoch": 0.014114326040931546,
117
+ "geo": 0.8932898789644241,
118
+ "grad_norm": 0.341796875,
119
+ "learning_rate": 9.85427135678392e-06,
120
+ "loss": 0.0002,
121
+ "num_tokens": 4721169.0,
122
+ "point_2d": 0.99609375,
123
+ "point_3d": 0.6064453125,
124
+ "reward": 0.8091074377298355,
125
+ "step": 40
126
+ },
127
+ {
128
+ "3d_bbox": 0.6575617864727974,
129
+ "adv": -0.0007460190936399158,
130
+ "bbox": 0.9782034233212471,
131
+ "completions/clipped_ratio": 0.0,
132
+ "completions/max_length": 1598.75,
133
+ "completions/max_terminated_length": 1598.75,
134
+ "completions/mean_length": 1581.515625,
135
+ "completions/mean_terminated_length": 1581.515625,
136
+ "completions/min_length": 1563.875,
137
+ "completions/min_terminated_length": 1563.875,
138
+ "entropy": 0.26874606031924486,
139
+ "epoch": 0.016937191249117856,
140
+ "geo": 0.922828696668148,
141
+ "grad_norm": 0.384765625,
142
+ "learning_rate": 9.814070351758794e-06,
143
+ "loss": 0.0001,
144
+ "num_tokens": 5667501.0,
145
+ "point_2d": 0.9404296875,
146
+ "point_3d": 0.6171875,
147
+ "reward": 0.7983455955982208,
148
+ "step": 48
149
+ },
150
+ {
151
+ "3d_bbox": 0.7516707628965378,
152
+ "adv": 0.0002944665375252953,
153
+ "bbox": 0.870270783547312,
154
+ "completions/clipped_ratio": 0.0,
155
+ "completions/max_length": 1605.5,
156
+ "completions/max_terminated_length": 1605.5,
157
+ "completions/mean_length": 1583.109375,
158
+ "completions/mean_terminated_length": 1583.109375,
159
+ "completions/min_length": 1563.125,
160
+ "completions/min_terminated_length": 1563.125,
161
+ "entropy": 0.27441642340272665,
162
+ "epoch": 0.019760056457304165,
163
+ "geo": 0.8707154765725136,
164
+ "grad_norm": 0.35546875,
165
+ "learning_rate": 9.773869346733669e-06,
166
+ "loss": 0.0002,
167
+ "num_tokens": 6614089.0,
168
+ "point_2d": 0.875,
169
+ "point_3d": 0.728515625,
170
+ "reward": 0.8063642866909504,
171
+ "step": 56
172
+ },
173
+ {
174
+ "3d_bbox": 0.7704295367002487,
175
+ "adv": 0.0006564971263287589,
176
+ "bbox": 0.8711516009643674,
177
+ "completions/clipped_ratio": 0.0,
178
+ "completions/max_length": 1604.75,
179
+ "completions/max_terminated_length": 1604.75,
180
+ "completions/mean_length": 1582.09375,
181
+ "completions/mean_terminated_length": 1582.09375,
182
+ "completions/min_length": 1562.25,
183
+ "completions/min_terminated_length": 1562.25,
184
+ "entropy": 0.2721484359353781,
185
+ "epoch": 0.022582921665490474,
186
+ "geo": 0.8998556286096573,
187
+ "grad_norm": 0.373046875,
188
+ "learning_rate": 9.733668341708543e-06,
189
+ "loss": 0.0156,
190
+ "num_tokens": 7560513.0,
191
+ "point_2d": 0.875,
192
+ "point_3d": 0.7880859375,
193
+ "reward": 0.8261667564511299,
194
+ "step": 64
195
+ },
196
+ {
197
+ "3d_bbox": 0.6965150758624077,
198
+ "adv": 5.7399442994210403e-05,
199
+ "bbox": 0.9924400746822357,
200
+ "completions/clipped_ratio": 0.0,
201
+ "completions/max_length": 1603.875,
202
+ "completions/max_terminated_length": 1603.875,
203
+ "completions/mean_length": 1580.8203125,
204
+ "completions/mean_terminated_length": 1580.8203125,
205
+ "completions/min_length": 1562.0,
206
+ "completions/min_terminated_length": 1562.0,
207
+ "entropy": 0.2694994225166738,
208
+ "epoch": 0.025405786873676783,
209
+ "geo": 0.7992948293685913,
210
+ "grad_norm": 0.40625,
211
+ "learning_rate": 9.693467336683417e-06,
212
+ "loss": -0.0004,
213
+ "num_tokens": 8506787.0,
214
+ "point_2d": 0.875,
215
+ "point_3d": 0.6201171875,
216
+ "reward": 0.796018086373806,
217
+ "step": 72
218
+ },
219
+ {
220
+ "3d_bbox": 0.9022366181015968,
221
+ "adv": 0.000732647989934776,
222
+ "bbox": 0.9670402258634567,
223
+ "completions/clipped_ratio": 0.0,
224
+ "completions/max_length": 1598.5,
225
+ "completions/max_terminated_length": 1598.5,
226
+ "completions/mean_length": 1577.83203125,
227
+ "completions/mean_terminated_length": 1577.83203125,
228
+ "completions/min_length": 1560.0,
229
+ "completions/min_terminated_length": 1560.0,
230
+ "entropy": 0.2729479926638305,
231
+ "epoch": 0.028228652081863093,
232
+ "geo": 0.770210200920701,
233
+ "grad_norm": 0.28515625,
234
+ "learning_rate": 9.653266331658291e-06,
235
+ "loss": 0.0285,
236
+ "num_tokens": 9452032.0,
237
+ "point_2d": 0.7578125,
238
+ "point_3d": 0.8740234375,
239
+ "reward": 0.8752781972289085,
240
+ "step": 80
241
+ },
242
+ {
243
+ "3d_bbox": 0.8243749290704727,
244
+ "adv": 0.0004942354327113208,
245
+ "bbox": 0.9483200311660767,
246
+ "completions/clipped_ratio": 0.0,
247
+ "completions/max_length": 1599.0,
248
+ "completions/max_terminated_length": 1599.0,
249
+ "completions/mean_length": 1578.5703125,
250
+ "completions/mean_terminated_length": 1578.5703125,
251
+ "completions/min_length": 1564.875,
252
+ "completions/min_terminated_length": 1564.875,
253
+ "entropy": 0.27196697471663356,
254
+ "epoch": 0.031051517290049402,
255
+ "geo": 0.8520620986819267,
256
+ "grad_norm": 0.337890625,
257
+ "learning_rate": 9.613065326633166e-06,
258
+ "loss": 0.0,
259
+ "num_tokens": 10397602.0,
260
+ "point_2d": 0.84375,
261
+ "point_3d": 0.7431640625,
262
+ "reward": 0.8399022594094276,
263
+ "step": 88
264
+ },
265
+ {
266
+ "3d_bbox": 0.7930564805865288,
267
+ "adv": -0.05789744130743202,
268
+ "bbox": 0.9452124014496803,
269
+ "completions/clipped_ratio": 0.0,
270
+ "completions/max_length": 1600.125,
271
+ "completions/max_terminated_length": 1600.125,
272
+ "completions/mean_length": 1579.25390625,
273
+ "completions/mean_terminated_length": 1579.25390625,
274
+ "completions/min_length": 1560.25,
275
+ "completions/min_terminated_length": 1560.25,
276
+ "entropy": 0.2732193539850414,
277
+ "epoch": 0.03387438249823571,
278
+ "geo": 0.8466203399002552,
279
+ "grad_norm": 0.35546875,
280
+ "learning_rate": 9.57286432160804e-06,
281
+ "loss": 0.0143,
282
+ "num_tokens": 11343267.0,
283
+ "point_2d": 0.7138671875,
284
+ "point_3d": 0.8203125,
285
+ "reward": 0.8181121572852135,
286
+ "step": 96
287
+ },
288
+ {
289
+ "3d_bbox": 0.6107643470168114,
290
+ "adv": -0.0019174332555849105,
291
+ "bbox": 0.9693493247032166,
292
+ "completions/clipped_ratio": 0.0,
293
+ "completions/max_length": 1605.375,
294
+ "completions/max_terminated_length": 1605.375,
295
+ "completions/mean_length": 1580.6875,
296
+ "completions/mean_terminated_length": 1580.6875,
297
+ "completions/min_length": 1559.75,
298
+ "completions/min_terminated_length": 1559.75,
299
+ "entropy": 0.2712484383955598,
300
+ "epoch": 0.03669724770642202,
301
+ "geo": 0.760684534907341,
302
+ "grad_norm": 0.392578125,
303
+ "learning_rate": 9.532663316582916e-06,
304
+ "loss": 0.0006,
305
+ "num_tokens": 12289259.0,
306
+ "point_2d": 0.849609375,
307
+ "point_3d": 0.7109375,
308
+ "reward": 0.7851651459932327,
309
+ "step": 104
310
+ },
311
+ {
312
+ "3d_bbox": 0.7011105790734291,
313
+ "adv": 0.000332610031364311,
314
+ "bbox": 0.997136726975441,
315
+ "completions/clipped_ratio": 0.0,
316
+ "completions/max_length": 1601.5,
317
+ "completions/max_terminated_length": 1601.5,
318
+ "completions/mean_length": 1578.38671875,
319
+ "completions/mean_terminated_length": 1578.38671875,
320
+ "completions/min_length": 1560.0,
321
+ "completions/min_terminated_length": 1560.0,
322
+ "entropy": 0.2741631781682372,
323
+ "epoch": 0.03952011291460833,
324
+ "geo": 0.868082121014595,
325
+ "grad_norm": 0.349609375,
326
+ "learning_rate": 9.49246231155779e-06,
327
+ "loss": -0.0005,
328
+ "num_tokens": 13234494.0,
329
+ "point_2d": 0.875,
330
+ "point_3d": 0.759765625,
331
+ "reward": 0.8332532122731209,
332
+ "step": 112
333
+ },
334
+ {
335
+ "3d_bbox": 0.6414853371679783,
336
+ "adv": 6.49127487122314e-05,
337
+ "bbox": 0.9946233779191971,
338
+ "completions/clipped_ratio": 0.0,
339
+ "completions/max_length": 1599.875,
340
+ "completions/max_terminated_length": 1599.875,
341
+ "completions/mean_length": 1580.48828125,
342
+ "completions/mean_terminated_length": 1580.48828125,
343
+ "completions/min_length": 1562.5,
344
+ "completions/min_terminated_length": 1562.5,
345
+ "entropy": 0.2765682805329561,
346
+ "epoch": 0.04234297812279464,
347
+ "geo": 0.7812720276415348,
348
+ "grad_norm": 0.392578125,
349
+ "learning_rate": 9.452261306532664e-06,
350
+ "loss": -0.0002,
351
+ "num_tokens": 14180739.0,
352
+ "point_2d": 0.875,
353
+ "point_3d": 0.7275390625,
354
+ "reward": 0.8096619322896004,
355
+ "step": 120
356
+ },
357
+ {
358
+ "3d_bbox": 0.696478821337223,
359
+ "adv": 0.00018893864194069465,
360
+ "bbox": 0.9915411844849586,
361
+ "completions/clipped_ratio": 0.0,
362
+ "completions/max_length": 1600.75,
363
+ "completions/max_terminated_length": 1600.75,
364
+ "completions/mean_length": 1581.48828125,
365
+ "completions/mean_terminated_length": 1581.48828125,
366
+ "completions/min_length": 1566.25,
367
+ "completions/min_terminated_length": 1566.25,
368
+ "entropy": 0.27939224475994706,
369
+ "epoch": 0.04516584333098095,
370
+ "geo": 0.9225678443908691,
371
+ "grad_norm": 0.359375,
372
+ "learning_rate": 9.412060301507539e-06,
373
+ "loss": 0.0004,
374
+ "num_tokens": 15126808.0,
375
+ "point_2d": 1.0,
376
+ "point_3d": 0.8037109375,
377
+ "reward": 0.8729327321052551,
378
+ "step": 128
379
+ },
380
+ {
381
+ "3d_bbox": 0.7811713069677353,
382
+ "adv": -0.00017097587488024146,
383
+ "bbox": 0.983598880469799,
384
+ "completions/clipped_ratio": 0.0,
385
+ "completions/max_length": 1599.0,
386
+ "completions/max_terminated_length": 1599.0,
387
+ "completions/mean_length": 1575.9765625,
388
+ "completions/mean_terminated_length": 1575.9765625,
389
+ "completions/min_length": 1560.875,
390
+ "completions/min_terminated_length": 1560.875,
391
+ "entropy": 0.2712345658801496,
392
+ "epoch": 0.04798870853916726,
393
+ "geo": 0.829373586922884,
394
+ "grad_norm": 0.365234375,
395
+ "learning_rate": 9.371859296482413e-06,
396
+ "loss": 0.0002,
397
+ "num_tokens": 16071682.0,
398
+ "point_2d": 0.75,
399
+ "point_3d": 0.76171875,
400
+ "reward": 0.819122239947319,
401
+ "step": 136
402
+ },
403
+ {
404
+ "3d_bbox": 0.6566238105297089,
405
+ "adv": 5.832723485355018e-05,
406
+ "bbox": 0.9859664067626,
407
+ "completions/clipped_ratio": 0.0,
408
+ "completions/max_length": 1597.5,
409
+ "completions/max_terminated_length": 1597.5,
410
+ "completions/mean_length": 1580.12890625,
411
+ "completions/mean_terminated_length": 1580.12890625,
412
+ "completions/min_length": 1565.0,
413
+ "completions/min_terminated_length": 1565.0,
414
+ "entropy": 0.2754409732297063,
415
+ "epoch": 0.05081157374735357,
416
+ "geo": 0.8895541727542877,
417
+ "grad_norm": 0.318359375,
418
+ "learning_rate": 9.331658291457287e-06,
419
+ "loss": 0.0218,
420
+ "num_tokens": 17017499.0,
421
+ "point_2d": 1.0,
422
+ "point_3d": 0.6328125,
423
+ "reward": 0.8188506811857224,
424
+ "step": 144
425
+ },
426
+ {
427
+ "3d_bbox": 0.655942264944315,
428
+ "adv": 0.00024080171715468168,
429
+ "bbox": 0.9830196872353554,
430
+ "completions/clipped_ratio": 0.0,
431
+ "completions/max_length": 1600.125,
432
+ "completions/max_terminated_length": 1600.125,
433
+ "completions/mean_length": 1581.3828125,
434
+ "completions/mean_terminated_length": 1581.3828125,
435
+ "completions/min_length": 1564.0,
436
+ "completions/min_terminated_length": 1564.0,
437
+ "entropy": 0.2764031286351383,
438
+ "epoch": 0.053634438955539876,
439
+ "geo": 0.8772727139294147,
440
+ "grad_norm": 0.294921875,
441
+ "learning_rate": 9.291457286432161e-06,
442
+ "loss": 0.015,
443
+ "num_tokens": 17963757.0,
444
+ "point_2d": 0.875,
445
+ "point_3d": 0.7880859375,
446
+ "reward": 0.8255119547247887,
447
+ "step": 152
448
+ },
449
+ {
450
+ "3d_bbox": 0.8087940737605095,
451
+ "adv": -0.062093503400319605,
452
+ "bbox": 0.9799230694770813,
453
+ "completions/clipped_ratio": 0.0,
454
+ "completions/max_length": 1598.5,
455
+ "completions/max_terminated_length": 1598.5,
456
+ "completions/mean_length": 1579.2421875,
457
+ "completions/mean_terminated_length": 1579.2421875,
458
+ "completions/min_length": 1560.25,
459
+ "completions/min_terminated_length": 1560.25,
460
+ "entropy": 0.27893486013635993,
461
+ "epoch": 0.056457304163726185,
462
+ "geo": 0.8533996716141701,
463
+ "grad_norm": 0.314453125,
464
+ "learning_rate": 9.251256281407036e-06,
465
+ "loss": 0.0157,
466
+ "num_tokens": 18909307.0,
467
+ "point_2d": 0.859375,
468
+ "point_3d": 0.814453125,
469
+ "reward": 0.8656363189220428,
470
+ "step": 160
471
+ },
472
+ {
473
+ "3d_bbox": 0.7417622283101082,
474
+ "adv": 0.00038278038937278325,
475
+ "bbox": 0.9947961643338203,
476
+ "completions/clipped_ratio": 0.0,
477
+ "completions/max_length": 1596.875,
478
+ "completions/max_terminated_length": 1596.875,
479
+ "completions/mean_length": 1578.4765625,
480
+ "completions/mean_terminated_length": 1578.4765625,
481
+ "completions/min_length": 1562.375,
482
+ "completions/min_terminated_length": 1562.375,
483
+ "entropy": 0.2772809020243585,
484
+ "epoch": 0.059280169371912494,
485
+ "geo": 0.8446554020047188,
486
+ "grad_norm": 0.337890625,
487
+ "learning_rate": 9.21105527638191e-06,
488
+ "loss": -0.0001,
489
+ "num_tokens": 19854893.0,
490
+ "point_2d": 0.875,
491
+ "point_3d": 0.7666015625,
492
+ "reward": 0.844539999961853,
493
+ "step": 168
494
+ },
495
+ {
496
+ "3d_bbox": 0.7199448123574257,
497
+ "adv": -0.00025279237161157653,
498
+ "bbox": 0.9939833134412766,
499
+ "completions/clipped_ratio": 0.0,
500
+ "completions/max_length": 1599.375,
501
+ "completions/max_terminated_length": 1599.375,
502
+ "completions/mean_length": 1577.875,
503
+ "completions/mean_terminated_length": 1577.875,
504
+ "completions/min_length": 1560.875,
505
+ "completions/min_terminated_length": 1560.875,
506
+ "entropy": 0.2748410953208804,
507
+ "epoch": 0.062103034580098804,
508
+ "geo": 0.8426482602953911,
509
+ "grad_norm": 0.353515625,
510
+ "learning_rate": 9.170854271356784e-06,
511
+ "loss": 0.0001,
512
+ "num_tokens": 20800093.0,
513
+ "point_2d": 0.8935546875,
514
+ "point_3d": 0.7666015625,
515
+ "reward": 0.8435210958123207,
516
+ "step": 176
517
+ },
518
+ {
519
+ "3d_bbox": 0.8168098330497742,
520
+ "adv": 0.0003567840885807527,
521
+ "bbox": 0.8971930183470249,
522
+ "completions/clipped_ratio": 0.0,
523
+ "completions/max_length": 1601.5,
524
+ "completions/max_terminated_length": 1601.5,
525
+ "completions/mean_length": 1581.01171875,
526
+ "completions/mean_terminated_length": 1581.01171875,
527
+ "completions/min_length": 1562.625,
528
+ "completions/min_terminated_length": 1562.625,
529
+ "entropy": 0.27716846112161875,
530
+ "epoch": 0.06492589978828511,
531
+ "geo": 0.8458116948604584,
532
+ "grad_norm": 0.376953125,
533
+ "learning_rate": 9.13065326633166e-06,
534
+ "loss": -0.0,
535
+ "num_tokens": 21746232.0,
536
+ "point_2d": 0.736328125,
537
+ "point_3d": 0.859375,
538
+ "reward": 0.8274264857172966,
539
+ "step": 184
540
+ },
541
+ {
542
+ "3d_bbox": 0.7658436857163906,
543
+ "adv": -0.00043575108065851964,
544
+ "bbox": 0.994478665292263,
545
+ "completions/clipped_ratio": 0.0,
546
+ "completions/max_length": 1599.5,
547
+ "completions/max_terminated_length": 1599.5,
548
+ "completions/mean_length": 1577.2265625,
549
+ "completions/mean_terminated_length": 1577.2265625,
550
+ "completions/min_length": 1558.25,
551
+ "completions/min_terminated_length": 1558.25,
552
+ "entropy": 0.28411462996155024,
553
+ "epoch": 0.06774876499647142,
554
+ "geo": 0.8858354687690735,
555
+ "grad_norm": 0.404296875,
556
+ "learning_rate": 9.090452261306534e-06,
557
+ "loss": -0.0,
558
+ "num_tokens": 22691162.0,
559
+ "point_2d": 1.0,
560
+ "point_3d": 0.83203125,
561
+ "reward": 0.8980883881449699,
562
+ "step": 192
563
+ },
564
+ {
565
+ "3d_bbox": 0.689879797399044,
566
+ "adv": 0.0008139211470279406,
567
+ "bbox": 0.8756355233490467,
568
+ "completions/clipped_ratio": 0.0,
569
+ "completions/max_length": 1595.25,
570
+ "completions/max_terminated_length": 1595.25,
571
+ "completions/mean_length": 1579.328125,
572
+ "completions/mean_terminated_length": 1579.328125,
573
+ "completions/min_length": 1562.125,
574
+ "completions/min_terminated_length": 1562.125,
575
+ "entropy": 0.27792059257626534,
576
+ "epoch": 0.07057163020465773,
577
+ "geo": 0.8776071295142174,
578
+ "grad_norm": 0.28125,
579
+ "learning_rate": 9.050251256281409e-06,
580
+ "loss": -0.0006,
581
+ "num_tokens": 23637118.0,
582
+ "point_2d": 0.875,
583
+ "point_3d": 0.765625,
584
+ "reward": 0.8015350699424744,
585
+ "step": 200
586
+ },
587
+ {
588
+ "3d_bbox": 0.7628588750958443,
589
+ "adv": 0.0005178588535272866,
590
+ "bbox": 0.9862369671463966,
591
+ "completions/clipped_ratio": 0.0,
592
+ "completions/max_length": 1603.5,
593
+ "completions/max_terminated_length": 1603.5,
594
+ "completions/mean_length": 1582.5,
595
+ "completions/mean_terminated_length": 1582.5,
596
+ "completions/min_length": 1561.75,
597
+ "completions/min_terminated_length": 1561.75,
598
+ "entropy": 0.2777437614277005,
599
+ "epoch": 0.07339449541284404,
600
+ "geo": 0.7896297052502632,
601
+ "grad_norm": 0.318359375,
602
+ "learning_rate": 9.010050251256283e-06,
603
+ "loss": -0.0003,
604
+ "num_tokens": 24583790.0,
605
+ "point_2d": 0.75,
606
+ "point_3d": 0.822265625,
607
+ "reward": 0.8303403630852699,
608
+ "step": 208
609
+ },
610
+ {
611
+ "3d_bbox": 0.6901299897581339,
612
+ "adv": 1.327232166659087e-05,
613
+ "bbox": 0.9742533341050148,
614
+ "completions/clipped_ratio": 0.0,
615
+ "completions/max_length": 1595.375,
616
+ "completions/max_terminated_length": 1595.375,
617
+ "completions/mean_length": 1580.3828125,
618
+ "completions/mean_terminated_length": 1580.3828125,
619
+ "completions/min_length": 1566.125,
620
+ "completions/min_terminated_length": 1566.125,
621
+ "entropy": 0.2789337821304798,
622
+ "epoch": 0.07621736062103035,
623
+ "geo": 0.9093400537967682,
624
+ "grad_norm": 0.33984375,
625
+ "learning_rate": 8.969849246231157e-06,
626
+ "loss": 0.0001,
627
+ "num_tokens": 25529904.0,
628
+ "point_2d": 1.0,
629
+ "point_3d": 0.724609375,
630
+ "reward": 0.8472481593489647,
631
+ "step": 216
632
+ },
633
+ {
634
+ "3d_bbox": 0.8317120671272278,
635
+ "adv": 0.0007813970478309784,
636
+ "bbox": 0.99237409979105,
637
+ "completions/clipped_ratio": 0.0,
638
+ "completions/max_length": 1592.375,
639
+ "completions/max_terminated_length": 1592.375,
640
+ "completions/mean_length": 1576.484375,
641
+ "completions/mean_terminated_length": 1576.484375,
642
+ "completions/min_length": 1559.125,
643
+ "completions/min_terminated_length": 1559.125,
644
+ "entropy": 0.281503742095083,
645
+ "epoch": 0.07904022582921666,
646
+ "geo": 0.9148535653948784,
647
+ "grad_norm": 0.33203125,
648
+ "learning_rate": 8.92964824120603e-06,
649
+ "loss": -0.0005,
650
+ "num_tokens": 26474884.0,
651
+ "point_2d": 1.0,
652
+ "point_3d": 0.859375,
653
+ "reward": 0.9208653047680855,
654
+ "step": 224
655
+ },
656
+ {
657
+ "3d_bbox": 0.8539343103766441,
658
+ "adv": 0.0006083497164581786,
659
+ "bbox": 0.9947174340486526,
660
+ "completions/clipped_ratio": 0.0,
661
+ "completions/max_length": 1607.25,
662
+ "completions/max_terminated_length": 1607.25,
663
+ "completions/mean_length": 1582.21875,
664
+ "completions/mean_terminated_length": 1582.21875,
665
+ "completions/min_length": 1562.75,
666
+ "completions/min_terminated_length": 1562.75,
667
+ "entropy": 0.28075447864830494,
668
+ "epoch": 0.08186309103740297,
669
+ "geo": 0.9205377101898193,
670
+ "grad_norm": 0.3828125,
671
+ "learning_rate": 8.889447236180904e-06,
672
+ "loss": 0.0151,
673
+ "num_tokens": 27421356.0,
674
+ "point_2d": 0.9423828125,
675
+ "point_3d": 0.919921875,
676
+ "reward": 0.9277391210198402,
677
+ "step": 232
678
+ },
679
+ {
680
+ "3d_bbox": 0.8143627047538757,
681
+ "adv": 0.0002584877729532309,
682
+ "bbox": 0.9961168393492699,
683
+ "completions/clipped_ratio": 0.0,
684
+ "completions/max_length": 1604.375,
685
+ "completions/max_terminated_length": 1604.375,
686
+ "completions/mean_length": 1581.84375,
687
+ "completions/mean_terminated_length": 1581.84375,
688
+ "completions/min_length": 1562.0,
689
+ "completions/min_terminated_length": 1562.0,
690
+ "entropy": 0.2808487247675657,
691
+ "epoch": 0.08468595624558928,
692
+ "geo": 0.9482192173600197,
693
+ "grad_norm": 0.3125,
694
+ "learning_rate": 8.84924623115578e-06,
695
+ "loss": 0.015,
696
+ "num_tokens": 28367588.0,
697
+ "point_2d": 1.0,
698
+ "point_3d": 0.8505859375,
699
+ "reward": 0.9152663722634315,
700
+ "step": 240
701
+ },
702
+ {
703
+ "3d_bbox": 0.7562066540122032,
704
+ "adv": 0.0011895421193912625,
705
+ "bbox": 0.9870459213852882,
706
+ "completions/clipped_ratio": 0.0,
707
+ "completions/max_length": 1605.625,
708
+ "completions/max_terminated_length": 1605.625,
709
+ "completions/mean_length": 1582.22265625,
710
+ "completions/mean_terminated_length": 1582.22265625,
711
+ "completions/min_length": 1560.125,
712
+ "completions/min_terminated_length": 1560.125,
713
+ "entropy": 0.2828774293884635,
714
+ "epoch": 0.08750882145377559,
715
+ "geo": 0.8906307443976402,
716
+ "grad_norm": 0.310546875,
717
+ "learning_rate": 8.809045226130654e-06,
718
+ "loss": -0.0001,
719
+ "num_tokens": 29314053.0,
720
+ "point_2d": 1.0,
721
+ "point_3d": 0.84765625,
722
+ "reward": 0.8977271988987923,
723
+ "step": 248
724
+ },
725
+ {
726
+ "3d_bbox": 0.662892609834671,
727
+ "adv": -0.0004179688154977157,
728
+ "bbox": 0.9923923686146736,
729
+ "completions/clipped_ratio": 0.0,
730
+ "completions/max_length": 1598.875,
731
+ "completions/max_terminated_length": 1598.875,
732
+ "completions/mean_length": 1579.796875,
733
+ "completions/mean_terminated_length": 1579.796875,
734
+ "completions/min_length": 1563.125,
735
+ "completions/min_terminated_length": 1563.125,
736
+ "entropy": 0.27959624165669084,
737
+ "epoch": 0.0903316866619619,
738
+ "geo": 0.9055434167385101,
739
+ "grad_norm": 0.259765625,
740
+ "learning_rate": 8.768844221105528e-06,
741
+ "loss": -0.0002,
742
+ "num_tokens": 30259921.0,
743
+ "point_2d": 0.7509765625,
744
+ "point_3d": 0.8154296875,
745
+ "reward": 0.805422805249691,
746
+ "step": 256
747
+ },
748
+ {
749
+ "3d_bbox": 0.7727803811430931,
750
+ "adv": -0.0010380800276834634,
751
+ "bbox": 0.9807735681533813,
752
+ "completions/clipped_ratio": 0.0,
753
+ "completions/max_length": 1604.375,
754
+ "completions/max_terminated_length": 1604.375,
755
+ "completions/mean_length": 1581.1796875,
756
+ "completions/mean_terminated_length": 1581.1796875,
757
+ "completions/min_length": 1564.375,
758
+ "completions/min_terminated_length": 1564.375,
759
+ "entropy": 0.28243178501725197,
760
+ "epoch": 0.0931545518701482,
761
+ "geo": 0.783092018100433,
762
+ "grad_norm": 0.3203125,
763
+ "learning_rate": 8.728643216080403e-06,
764
+ "loss": 0.0003,
765
+ "num_tokens": 31205887.0,
766
+ "point_2d": 0.75,
767
+ "point_3d": 0.8544921875,
768
+ "reward": 0.8395115286111832,
769
+ "step": 264
770
+ },
771
+ {
772
+ "3d_bbox": 0.7379852011799812,
773
+ "adv": -0.0005326499958755448,
774
+ "bbox": 0.9934581220149994,
775
+ "completions/clipped_ratio": 0.0,
776
+ "completions/max_length": 1598.875,
777
+ "completions/max_terminated_length": 1598.875,
778
+ "completions/mean_length": 1579.44921875,
779
+ "completions/mean_terminated_length": 1579.44921875,
780
+ "completions/min_length": 1560.375,
781
+ "completions/min_terminated_length": 1560.375,
782
+ "entropy": 0.2767978576011956,
783
+ "epoch": 0.09597741707833451,
784
+ "geo": 0.9199443385004997,
785
+ "grad_norm": 0.40625,
786
+ "learning_rate": 8.688442211055277e-06,
787
+ "loss": 0.0002,
788
+ "num_tokens": 32151650.0,
789
+ "point_2d": 0.9990234375,
790
+ "point_3d": 0.8896484375,
791
+ "reward": 0.9050287976861,
792
+ "step": 272
793
+ },
794
+ {
795
+ "3d_bbox": 0.6362542062997818,
796
+ "adv": 0.0003683161558001302,
797
+ "bbox": 0.9273895137012005,
798
+ "completions/clipped_ratio": 0.0,
799
+ "completions/max_length": 1603.375,
800
+ "completions/max_terminated_length": 1603.375,
801
+ "completions/mean_length": 1583.1875,
802
+ "completions/mean_terminated_length": 1583.1875,
803
+ "completions/min_length": 1566.25,
804
+ "completions/min_terminated_length": 1566.25,
805
+ "entropy": 0.28121166303753853,
806
+ "epoch": 0.09880028228652082,
807
+ "geo": 0.6698016822338104,
808
+ "grad_norm": 0.419921875,
809
+ "learning_rate": 8.648241206030151e-06,
810
+ "loss": 0.0104,
811
+ "num_tokens": 33098514.0,
812
+ "point_2d": 0.748046875,
813
+ "point_3d": 0.8564453125,
814
+ "reward": 0.7920339703559875,
815
+ "step": 280
816
+ },
817
+ {
818
+ "3d_bbox": 0.8045329824090004,
819
+ "adv": 0.0005392557122831931,
820
+ "bbox": 0.9957676902413368,
821
+ "completions/clipped_ratio": 0.0,
822
+ "completions/max_length": 1598.75,
823
+ "completions/max_terminated_length": 1598.75,
824
+ "completions/mean_length": 1582.8515625,
825
+ "completions/mean_terminated_length": 1582.8515625,
826
+ "completions/min_length": 1564.75,
827
+ "completions/min_terminated_length": 1564.75,
828
+ "entropy": 0.2840573275461793,
829
+ "epoch": 0.10162314749470713,
830
+ "geo": 0.894830085337162,
831
+ "grad_norm": 0.263671875,
832
+ "learning_rate": 8.608040201005025e-06,
833
+ "loss": -0.0003,
834
+ "num_tokens": 34045156.0,
835
+ "point_2d": 1.0,
836
+ "point_3d": 0.8974609375,
837
+ "reward": 0.9244403913617134,
838
+ "step": 288
839
+ },
840
+ {
841
+ "3d_bbox": 0.6487490832805634,
842
+ "adv": -0.036328162997961044,
843
+ "bbox": 0.9537674784660339,
844
+ "completions/clipped_ratio": 0.0,
845
+ "completions/max_length": 1606.125,
846
+ "completions/max_terminated_length": 1606.125,
847
+ "completions/mean_length": 1583.43359375,
848
+ "completions/mean_terminated_length": 1583.43359375,
849
+ "completions/min_length": 1564.25,
850
+ "completions/min_terminated_length": 1564.25,
851
+ "entropy": 0.2789871282875538,
852
+ "epoch": 0.10444601270289344,
853
+ "geo": 0.8219972774386406,
854
+ "grad_norm": 0.3671875,
855
+ "learning_rate": 8.5678391959799e-06,
856
+ "loss": 0.0091,
857
+ "num_tokens": 34991979.0,
858
+ "point_2d": 0.875,
859
+ "point_3d": 0.8125,
860
+ "reward": 0.8225041329860687,
861
+ "step": 296
862
+ },
863
+ {
864
+ "3d_bbox": 0.7852863818407059,
865
+ "adv": -2.922765997936949e-05,
866
+ "bbox": 0.9355472847819328,
867
+ "completions/clipped_ratio": 0.0,
868
+ "completions/max_length": 1601.5,
869
+ "completions/max_terminated_length": 1601.5,
870
+ "completions/mean_length": 1579.7890625,
871
+ "completions/mean_terminated_length": 1579.7890625,
872
+ "completions/min_length": 1561.5,
873
+ "completions/min_terminated_length": 1561.5,
874
+ "entropy": 0.2808959139510989,
875
+ "epoch": 0.10726887791107975,
876
+ "geo": 0.7100357487797737,
877
+ "grad_norm": 0.333984375,
878
+ "learning_rate": 8.527638190954774e-06,
879
+ "loss": 0.0003,
880
+ "num_tokens": 35937853.0,
881
+ "point_2d": 0.6728515625,
882
+ "point_3d": 0.798828125,
883
+ "reward": 0.7981283441185951,
884
+ "step": 304
885
+ },
886
+ {
887
+ "3d_bbox": 0.6897341758012772,
888
+ "adv": -0.0010775219143397408,
889
+ "bbox": 0.9957167953252792,
890
+ "completions/clipped_ratio": 0.0,
891
+ "completions/max_length": 1598.0,
892
+ "completions/max_terminated_length": 1598.0,
893
+ "completions/mean_length": 1578.40234375,
894
+ "completions/mean_terminated_length": 1578.40234375,
895
+ "completions/min_length": 1560.625,
896
+ "completions/min_terminated_length": 1560.625,
897
+ "entropy": 0.28013237891718745,
898
+ "epoch": 0.11009174311926606,
899
+ "geo": 0.8601852916181087,
900
+ "grad_norm": 0.322265625,
901
+ "learning_rate": 8.487437185929648e-06,
902
+ "loss": 0.0003,
903
+ "num_tokens": 36883412.0,
904
+ "point_2d": 1.0,
905
+ "point_3d": 0.8154296875,
906
+ "reward": 0.8752201721072197,
907
+ "step": 312
908
+ },
909
+ {
910
+ "3d_bbox": 0.7732156217098236,
911
+ "adv": 0.00015621728016412817,
912
+ "bbox": 0.9900884553790092,
913
+ "completions/clipped_ratio": 0.0,
914
+ "completions/max_length": 1595.25,
915
+ "completions/max_terminated_length": 1595.25,
916
+ "completions/mean_length": 1574.06640625,
917
+ "completions/mean_terminated_length": 1574.06640625,
918
+ "completions/min_length": 1556.5,
919
+ "completions/min_terminated_length": 1556.5,
920
+ "entropy": 0.2842056662775576,
921
+ "epoch": 0.11291460832745237,
922
+ "geo": 0.8583031129091978,
923
+ "grad_norm": 0.359375,
924
+ "learning_rate": 8.447236180904524e-06,
925
+ "loss": -0.0003,
926
+ "num_tokens": 37827701.0,
927
+ "point_2d": 0.8583984375,
928
+ "point_3d": 0.8681640625,
929
+ "reward": 0.8724666386842728,
930
+ "step": 320
931
+ },
932
+ {
933
+ "3d_bbox": 0.7427574843168259,
934
+ "adv": -0.00022968700068304315,
935
+ "bbox": 0.9458357989788055,
936
+ "completions/clipped_ratio": 0.0,
937
+ "completions/max_length": 1599.25,
938
+ "completions/max_terminated_length": 1599.25,
939
+ "completions/mean_length": 1576.78125,
940
+ "completions/mean_terminated_length": 1576.78125,
941
+ "completions/min_length": 1555.625,
942
+ "completions/min_terminated_length": 1555.625,
943
+ "entropy": 0.28632913110777736,
944
+ "epoch": 0.11573747353563868,
945
+ "geo": 0.9527824521064758,
946
+ "grad_norm": 0.2578125,
947
+ "learning_rate": 8.407035175879398e-06,
948
+ "loss": -0.0,
949
+ "num_tokens": 38772853.0,
950
+ "point_2d": 1.0,
951
+ "point_3d": 0.8623046875,
952
+ "reward": 0.8877245113253593,
953
+ "step": 328
954
+ },
955
+ {
956
+ "3d_bbox": 0.7519282810389996,
957
+ "adv": 0.0003111484527380526,
958
+ "bbox": 0.9336327686905861,
959
+ "completions/clipped_ratio": 0.0,
960
+ "completions/max_length": 1598.0,
961
+ "completions/max_terminated_length": 1598.0,
962
+ "completions/mean_length": 1578.54296875,
963
+ "completions/mean_terminated_length": 1578.54296875,
964
+ "completions/min_length": 1560.0,
965
+ "completions/min_terminated_length": 1560.0,
966
+ "entropy": 0.2819040766917169,
967
+ "epoch": 0.11856033874382499,
968
+ "geo": 0.9259950742125511,
969
+ "grad_norm": 0.306640625,
970
+ "learning_rate": 8.366834170854273e-06,
971
+ "loss": 0.0202,
972
+ "num_tokens": 39718400.0,
973
+ "point_2d": 0.7451171875,
974
+ "point_3d": 0.748046875,
975
+ "reward": 0.7946812808513641,
976
+ "step": 336
977
+ },
978
+ {
979
+ "3d_bbox": 0.6988782063126564,
980
+ "adv": 8.153711894226534e-05,
981
+ "bbox": 0.9918540865182877,
982
+ "completions/clipped_ratio": 0.0,
983
+ "completions/max_length": 1602.0,
984
+ "completions/max_terminated_length": 1602.0,
985
+ "completions/mean_length": 1578.3984375,
986
+ "completions/mean_terminated_length": 1578.3984375,
987
+ "completions/min_length": 1557.0,
988
+ "completions/min_terminated_length": 1557.0,
989
+ "entropy": 0.2852953290566802,
990
+ "epoch": 0.1213832039520113,
991
+ "geo": 0.824953593313694,
992
+ "grad_norm": 0.33984375,
993
+ "learning_rate": 8.326633165829147e-06,
994
+ "loss": 0.0003,
995
+ "num_tokens": 40663774.0,
996
+ "point_2d": 0.875,
997
+ "point_3d": 0.771484375,
998
+ "reward": 0.8343041799962521,
999
+ "step": 344
1000
+ },
1001
+ {
1002
+ "3d_bbox": 0.7085578441619873,
1003
+ "adv": 0.000511349891894497,
1004
+ "bbox": 0.9954689294099808,
1005
+ "completions/clipped_ratio": 0.0,
1006
+ "completions/max_length": 1601.625,
1007
+ "completions/max_terminated_length": 1601.625,
1008
+ "completions/mean_length": 1578.21484375,
1009
+ "completions/mean_terminated_length": 1578.21484375,
1010
+ "completions/min_length": 1557.5,
1011
+ "completions/min_terminated_length": 1557.5,
1012
+ "entropy": 0.2835490182042122,
1013
+ "epoch": 0.12420606916019761,
1014
+ "geo": 0.8287867829203606,
1015
+ "grad_norm": 0.31640625,
1016
+ "learning_rate": 8.286432160804021e-06,
1017
+ "loss": 0.0,
1018
+ "num_tokens": 41609261.0,
1019
+ "point_2d": 0.875,
1020
+ "point_3d": 0.8564453125,
1021
+ "reward": 0.8588679879903793,
1022
+ "step": 352
1023
+ },
1024
+ {
1025
+ "3d_bbox": 0.7324622049927711,
1026
+ "adv": -0.06052318853471661,
1027
+ "bbox": 0.9785870239138603,
1028
+ "completions/clipped_ratio": 0.0,
1029
+ "completions/max_length": 1599.75,
1030
+ "completions/max_terminated_length": 1599.75,
1031
+ "completions/mean_length": 1580.8671875,
1032
+ "completions/mean_terminated_length": 1580.8671875,
1033
+ "completions/min_length": 1562.875,
1034
+ "completions/min_terminated_length": 1562.875,
1035
+ "entropy": 0.2788377907127142,
1036
+ "epoch": 0.12702893436838392,
1037
+ "geo": 0.8886202499270439,
1038
+ "grad_norm": 0.765625,
1039
+ "learning_rate": 8.246231155778895e-06,
1040
+ "loss": 0.0295,
1041
+ "num_tokens": 42555355.0,
1042
+ "point_2d": 0.984375,
1043
+ "point_3d": 0.80078125,
1044
+ "reward": 0.8740513622760773,
1045
+ "step": 360
1046
+ },
1047
+ {
1048
+ "3d_bbox": 0.6365351751446724,
1049
+ "adv": 0.0003275554918218404,
1050
+ "bbox": 0.8720735665410757,
1051
+ "completions/clipped_ratio": 0.0,
1052
+ "completions/max_length": 1601.875,
1053
+ "completions/max_terminated_length": 1601.875,
1054
+ "completions/mean_length": 1577.73046875,
1055
+ "completions/mean_terminated_length": 1577.73046875,
1056
+ "completions/min_length": 1555.75,
1057
+ "completions/min_terminated_length": 1555.75,
1058
+ "entropy": 0.28446380188688636,
1059
+ "epoch": 0.12985179957657023,
1060
+ "geo": 0.8144352469826117,
1061
+ "grad_norm": 0.34375,
1062
+ "learning_rate": 8.20603015075377e-06,
1063
+ "loss": -0.0001,
1064
+ "num_tokens": 43500534.0,
1065
+ "point_2d": 0.875,
1066
+ "point_3d": 0.7666015625,
1067
+ "reward": 0.787552572786808,
1068
+ "step": 368
1069
+ },
1070
+ {
1071
+ "3d_bbox": 0.7198036424815655,
1072
+ "adv": -9.503715409664437e-05,
1073
+ "bbox": 0.9939679279923439,
1074
+ "completions/clipped_ratio": 0.0,
1075
+ "completions/max_length": 1592.75,
1076
+ "completions/max_terminated_length": 1592.75,
1077
+ "completions/mean_length": 1573.19140625,
1078
+ "completions/mean_terminated_length": 1573.19140625,
1079
+ "completions/min_length": 1557.0,
1080
+ "completions/min_terminated_length": 1557.0,
1081
+ "entropy": 0.28550330083817244,
1082
+ "epoch": 0.13267466478475654,
1083
+ "geo": 0.9650512784719467,
1084
+ "grad_norm": 0.359375,
1085
+ "learning_rate": 8.165829145728644e-06,
1086
+ "loss": 0.0004,
1087
+ "num_tokens": 44444583.0,
1088
+ "point_2d": 0.9892578125,
1089
+ "point_3d": 0.83984375,
1090
+ "reward": 0.8857182934880257,
1091
+ "step": 376
1092
+ },
1093
+ {
1094
+ "3d_bbox": 0.7569727748632431,
1095
+ "adv": 0.0004161505894444417,
1096
+ "bbox": 0.9042535312473774,
1097
+ "completions/clipped_ratio": 0.0,
1098
+ "completions/max_length": 1599.5,
1099
+ "completions/max_terminated_length": 1599.5,
1100
+ "completions/mean_length": 1577.55859375,
1101
+ "completions/mean_terminated_length": 1577.55859375,
1102
+ "completions/min_length": 1561.375,
1103
+ "completions/min_terminated_length": 1561.375,
1104
+ "entropy": 0.2907078079879284,
1105
+ "epoch": 0.13549752999294284,
1106
+ "geo": 0.6764855124056339,
1107
+ "grad_norm": 0.345703125,
1108
+ "learning_rate": 8.125628140703518e-06,
1109
+ "loss": 0.0001,
1110
+ "num_tokens": 45389846.0,
1111
+ "point_2d": 0.69140625,
1112
+ "point_3d": 0.8193359375,
1113
+ "reward": 0.7929921373724937,
1114
+ "step": 384
1115
+ },
1116
+ {
1117
+ "3d_bbox": 0.6764840334653854,
1118
+ "adv": -1.4531655324390158e-05,
1119
+ "bbox": 0.8646975085139275,
1120
+ "completions/clipped_ratio": 0.0,
1121
+ "completions/max_length": 1600.5,
1122
+ "completions/max_terminated_length": 1600.5,
1123
+ "completions/mean_length": 1579.08984375,
1124
+ "completions/mean_terminated_length": 1579.08984375,
1125
+ "completions/min_length": 1558.375,
1126
+ "completions/min_terminated_length": 1558.375,
1127
+ "entropy": 0.2863780874758959,
1128
+ "epoch": 0.13832039520112915,
1129
+ "geo": 0.5404472178779542,
1130
+ "grad_norm": 0.3125,
1131
+ "learning_rate": 8.085427135678392e-06,
1132
+ "loss": 0.0,
1133
+ "num_tokens": 46335629.0,
1134
+ "point_2d": 0.5048828125,
1135
+ "point_3d": 0.806640625,
1136
+ "reward": 0.713176254183054,
1137
+ "step": 392
1138
+ },
1139
+ {
1140
+ "3d_bbox": 0.7255214750766754,
1141
+ "adv": -0.0006330321593850385,
1142
+ "bbox": 0.891293590888381,
1143
+ "completions/clipped_ratio": 0.0,
1144
+ "completions/max_length": 1602.375,
1145
+ "completions/max_terminated_length": 1602.375,
1146
+ "completions/mean_length": 1576.9921875,
1147
+ "completions/mean_terminated_length": 1576.9921875,
1148
+ "completions/min_length": 1558.125,
1149
+ "completions/min_terminated_length": 1558.125,
1150
+ "entropy": 0.28716066014021635,
1151
+ "epoch": 0.14114326040931546,
1152
+ "geo": 0.9741705134510994,
1153
+ "grad_norm": 0.314453125,
1154
+ "learning_rate": 8.045226130653267e-06,
1155
+ "loss": 0.0004,
1156
+ "num_tokens": 47280507.0,
1157
+ "point_2d": 0.8916015625,
1158
+ "point_3d": 0.859375,
1159
+ "reward": 0.8419479243457317,
1160
+ "step": 400
1161
+ },
1162
+ {
1163
+ "3d_bbox": 0.7029357459396124,
1164
+ "adv": 0.0008838025460136123,
1165
+ "bbox": 0.9950641989707947,
1166
+ "completions/clipped_ratio": 0.0,
1167
+ "completions/max_length": 1598.125,
1168
+ "completions/max_terminated_length": 1598.125,
1169
+ "completions/mean_length": 1575.92578125,
1170
+ "completions/mean_terminated_length": 1575.92578125,
1171
+ "completions/min_length": 1557.75,
1172
+ "completions/min_terminated_length": 1557.75,
1173
+ "entropy": 0.2837083409540355,
1174
+ "epoch": 0.14396612561750177,
1175
+ "geo": 0.864502564072609,
1176
+ "grad_norm": 0.345703125,
1177
+ "learning_rate": 8.005025125628141e-06,
1178
+ "loss": -0.0003,
1179
+ "num_tokens": 48225240.0,
1180
+ "point_2d": 1.0,
1181
+ "point_3d": 0.8310546875,
1182
+ "reward": 0.8822636753320694,
1183
+ "step": 408
1184
+ },
1185
+ {
1186
+ "3d_bbox": 0.6913789249956608,
1187
+ "adv": -0.0002196824352722615,
1188
+ "bbox": 0.9765095412731171,
1189
+ "completions/clipped_ratio": 0.0,
1190
+ "completions/max_length": 1598.125,
1191
+ "completions/max_terminated_length": 1598.125,
1192
+ "completions/mean_length": 1579.28515625,
1193
+ "completions/mean_terminated_length": 1579.28515625,
1194
+ "completions/min_length": 1559.25,
1195
+ "completions/min_terminated_length": 1559.25,
1196
+ "entropy": 0.28565866220742464,
1197
+ "epoch": 0.14678899082568808,
1198
+ "geo": 0.8599895760416985,
1199
+ "grad_norm": 0.30859375,
1200
+ "learning_rate": 7.964824120603015e-06,
1201
+ "loss": 0.0001,
1202
+ "num_tokens": 49171057.0,
1203
+ "point_2d": 1.0,
1204
+ "point_3d": 0.904296875,
1205
+ "reward": 0.8930463194847107,
1206
+ "step": 416
1207
+ },
1208
+ {
1209
+ "3d_bbox": 0.6788053698837757,
1210
+ "adv": -0.00020617676273104735,
1211
+ "bbox": 0.9748031497001648,
1212
+ "completions/clipped_ratio": 0.0,
1213
+ "completions/max_length": 1595.25,
1214
+ "completions/max_terminated_length": 1595.25,
1215
+ "completions/mean_length": 1577.51953125,
1216
+ "completions/mean_terminated_length": 1577.51953125,
1217
+ "completions/min_length": 1559.625,
1218
+ "completions/min_terminated_length": 1559.625,
1219
+ "entropy": 0.28440852370113134,
1220
+ "epoch": 0.1496118560338744,
1221
+ "geo": 0.822394497692585,
1222
+ "grad_norm": 0.40234375,
1223
+ "learning_rate": 7.92462311557789e-06,
1224
+ "loss": 0.0006,
1225
+ "num_tokens": 50116302.0,
1226
+ "point_2d": 0.875,
1227
+ "point_3d": 0.8310546875,
1228
+ "reward": 0.8399158045649529,
1229
+ "step": 424
1230
+ },
1231
+ {
1232
+ "3d_bbox": 0.6794650703668594,
1233
+ "adv": 0.0011559298304746335,
1234
+ "bbox": 0.9097459949553013,
1235
+ "completions/clipped_ratio": 0.0,
1236
+ "completions/max_length": 1596.625,
1237
+ "completions/max_terminated_length": 1596.625,
1238
+ "completions/mean_length": 1573.80859375,
1239
+ "completions/mean_terminated_length": 1573.80859375,
1240
+ "completions/min_length": 1552.625,
1241
+ "completions/min_terminated_length": 1552.625,
1242
+ "entropy": 0.28376144589856267,
1243
+ "epoch": 0.1524347212420607,
1244
+ "geo": 0.7893084269016981,
1245
+ "grad_norm": 0.3203125,
1246
+ "learning_rate": 7.884422110552764e-06,
1247
+ "loss": 0.0,
1248
+ "num_tokens": 51060365.0,
1249
+ "point_2d": 1.0,
1250
+ "point_3d": 0.8681640625,
1251
+ "reward": 0.8643437847495079,
1252
+ "step": 432
1253
+ },
1254
+ {
1255
+ "3d_bbox": 0.7942164540290833,
1256
+ "adv": -3.183347871527076e-05,
1257
+ "bbox": 0.9926317036151886,
1258
+ "completions/clipped_ratio": 0.0,
1259
+ "completions/max_length": 1598.75,
1260
+ "completions/max_terminated_length": 1598.75,
1261
+ "completions/mean_length": 1577.796875,
1262
+ "completions/mean_terminated_length": 1577.796875,
1263
+ "completions/min_length": 1553.125,
1264
+ "completions/min_terminated_length": 1553.125,
1265
+ "entropy": 0.2821194725111127,
1266
+ "epoch": 0.155257586450247,
1267
+ "geo": 0.9503362327814102,
1268
+ "grad_norm": 0.3984375,
1269
+ "learning_rate": 7.844221105527638e-06,
1270
+ "loss": 0.0002,
1271
+ "num_tokens": 52005633.0,
1272
+ "point_2d": 1.0,
1273
+ "point_3d": 0.869140625,
1274
+ "reward": 0.9139972031116486,
1275
+ "step": 440
1276
+ },
1277
+ {
1278
+ "3d_bbox": 0.6752692721784115,
1279
+ "adv": -9.695853077573702e-05,
1280
+ "bbox": 0.982423759996891,
1281
+ "completions/clipped_ratio": 0.0,
1282
+ "completions/max_length": 1596.25,
1283
+ "completions/max_terminated_length": 1596.25,
1284
+ "completions/mean_length": 1574.5546875,
1285
+ "completions/mean_terminated_length": 1574.5546875,
1286
+ "completions/min_length": 1555.125,
1287
+ "completions/min_terminated_length": 1555.125,
1288
+ "entropy": 0.2852726257406175,
1289
+ "epoch": 0.15808045165843332,
1290
+ "geo": 0.5512092066928744,
1291
+ "grad_norm": 0.322265625,
1292
+ "learning_rate": 7.804020100502512e-06,
1293
+ "loss": -0.0004,
1294
+ "num_tokens": 52949951.0,
1295
+ "point_2d": 0.75,
1296
+ "point_3d": 0.7998046875,
1297
+ "reward": 0.801874428987503,
1298
+ "step": 448
1299
+ },
1300
+ {
1301
+ "3d_bbox": 0.7081510461866856,
1302
+ "adv": 0.0003901398231391795,
1303
+ "bbox": 0.9892760217189789,
1304
+ "completions/clipped_ratio": 0.0,
1305
+ "completions/max_length": 1596.875,
1306
+ "completions/max_terminated_length": 1596.875,
1307
+ "completions/mean_length": 1573.0390625,
1308
+ "completions/mean_terminated_length": 1573.0390625,
1309
+ "completions/min_length": 1551.875,
1310
+ "completions/min_terminated_length": 1551.875,
1311
+ "entropy": 0.2837893716059625,
1312
+ "epoch": 0.16090331686661963,
1313
+ "geo": 0.9603376165032387,
1314
+ "grad_norm": 0.3125,
1315
+ "learning_rate": 7.763819095477387e-06,
1316
+ "loss": -0.0007,
1317
+ "num_tokens": 53894057.0,
1318
+ "point_2d": 1.0,
1319
+ "point_3d": 0.794921875,
1320
+ "reward": 0.8730872422456741,
1321
+ "step": 456
1322
+ },
1323
+ {
1324
+ "3d_bbox": 0.7504811957478523,
1325
+ "adv": 0.0007141208916436881,
1326
+ "bbox": 0.9912561923265457,
1327
+ "completions/clipped_ratio": 0.0,
1328
+ "completions/max_length": 1594.75,
1329
+ "completions/max_terminated_length": 1594.75,
1330
+ "completions/mean_length": 1576.1640625,
1331
+ "completions/mean_terminated_length": 1576.1640625,
1332
+ "completions/min_length": 1555.75,
1333
+ "completions/min_terminated_length": 1555.75,
1334
+ "entropy": 0.28930201614275575,
1335
+ "epoch": 0.16372618207480594,
1336
+ "geo": 0.8148366361856461,
1337
+ "grad_norm": 0.275390625,
1338
+ "learning_rate": 7.723618090452262e-06,
1339
+ "loss": 0.0003,
1340
+ "num_tokens": 54839099.0,
1341
+ "point_2d": 0.875,
1342
+ "point_3d": 0.8896484375,
1343
+ "reward": 0.8765964508056641,
1344
+ "step": 464
1345
+ },
1346
+ {
1347
+ "3d_bbox": 0.819125022739172,
1348
+ "adv": 0.00010633431156747974,
1349
+ "bbox": 0.9906315729022026,
1350
+ "completions/clipped_ratio": 0.0,
1351
+ "completions/max_length": 1603.0,
1352
+ "completions/max_terminated_length": 1603.0,
1353
+ "completions/mean_length": 1578.3203125,
1354
+ "completions/mean_terminated_length": 1578.3203125,
1355
+ "completions/min_length": 1556.375,
1356
+ "completions/min_terminated_length": 1556.375,
1357
+ "entropy": 0.28396829310804605,
1358
+ "epoch": 0.16654904728299225,
1359
+ "geo": 0.8597357729449868,
1360
+ "grad_norm": 0.30078125,
1361
+ "learning_rate": 7.683417085427137e-06,
1362
+ "loss": 0.0004,
1363
+ "num_tokens": 55784557.0,
1364
+ "point_2d": 0.8916015625,
1365
+ "point_3d": 0.9345703125,
1366
+ "reward": 0.9089821055531502,
1367
+ "step": 472
1368
+ },
1369
+ {
1370
+ "3d_bbox": 0.7331585213541985,
1371
+ "adv": -0.00033648866519797593,
1372
+ "bbox": 0.8809062782675028,
1373
+ "completions/clipped_ratio": 0.0,
1374
+ "completions/max_length": 1594.625,
1375
+ "completions/max_terminated_length": 1594.625,
1376
+ "completions/mean_length": 1576.11328125,
1377
+ "completions/mean_terminated_length": 1576.11328125,
1378
+ "completions/min_length": 1559.25,
1379
+ "completions/min_terminated_length": 1559.25,
1380
+ "entropy": 0.28669640561565757,
1381
+ "epoch": 0.16937191249117856,
1382
+ "geo": 0.9506581574678421,
1383
+ "grad_norm": 0.2890625,
1384
+ "learning_rate": 7.643216080402011e-06,
1385
+ "loss": 0.0132,
1386
+ "num_tokens": 56729514.0,
1387
+ "point_2d": 0.8759765625,
1388
+ "point_3d": 0.880859375,
1389
+ "reward": 0.8427251875400543,
1390
+ "step": 480
1391
+ },
1392
+ {
1393
+ "3d_bbox": 0.7568283602595329,
1394
+ "adv": -0.0008519564944435842,
1395
+ "bbox": 0.9879242181777954,
1396
+ "completions/clipped_ratio": 0.0,
1397
+ "completions/max_length": 1600.875,
1398
+ "completions/max_terminated_length": 1600.875,
1399
+ "completions/mean_length": 1574.87890625,
1400
+ "completions/mean_terminated_length": 1574.87890625,
1401
+ "completions/min_length": 1554.25,
1402
+ "completions/min_terminated_length": 1554.25,
1403
+ "entropy": 0.28737954096868634,
1404
+ "epoch": 0.17219477769936486,
1405
+ "geo": 0.7571075335144997,
1406
+ "grad_norm": 0.400390625,
1407
+ "learning_rate": 7.603015075376885e-06,
1408
+ "loss": 0.0105,
1409
+ "num_tokens": 57674099.0,
1410
+ "point_2d": 0.7998046875,
1411
+ "point_3d": 0.9228515625,
1412
+ "reward": 0.8668522015213966,
1413
+ "step": 488
1414
+ },
1415
+ {
1416
+ "3d_bbox": 0.7658256888389587,
1417
+ "adv": 0.0011402575073589105,
1418
+ "bbox": 0.9932135492563248,
1419
+ "completions/clipped_ratio": 0.0,
1420
+ "completions/max_length": 1596.25,
1421
+ "completions/max_terminated_length": 1596.25,
1422
+ "completions/mean_length": 1574.265625,
1423
+ "completions/mean_terminated_length": 1574.265625,
1424
+ "completions/min_length": 1554.375,
1425
+ "completions/min_terminated_length": 1554.375,
1426
+ "entropy": 0.2865150640718639,
1427
+ "epoch": 0.17501764290755117,
1428
+ "geo": 0.8641151636838913,
1429
+ "grad_norm": 0.37890625,
1430
+ "learning_rate": 7.5628140703517595e-06,
1431
+ "loss": -0.0004,
1432
+ "num_tokens": 58618575.0,
1433
+ "point_2d": 0.91015625,
1434
+ "point_3d": 0.904296875,
1435
+ "reward": 0.8933730870485306,
1436
+ "step": 496
1437
+ },
1438
+ {
1439
+ "3d_bbox": 0.7145381905138493,
1440
+ "adv": 0.0003104716488451231,
1441
+ "bbox": 0.9926531538367271,
1442
+ "completions/clipped_ratio": 0.0,
1443
+ "completions/max_length": 1594.75,
1444
+ "completions/max_terminated_length": 1594.75,
1445
+ "completions/mean_length": 1573.12890625,
1446
+ "completions/mean_terminated_length": 1573.12890625,
1447
+ "completions/min_length": 1553.375,
1448
+ "completions/min_terminated_length": 1553.375,
1449
+ "entropy": 0.28603550279513,
1450
+ "epoch": 0.17784050811573748,
1451
+ "geo": 0.7219396531581879,
1452
+ "grad_norm": 0.310546875,
1453
+ "learning_rate": 7.522613065326634e-06,
1454
+ "loss": -0.0,
1455
+ "num_tokens": 59562552.0,
1456
+ "point_2d": 0.7509765625,
1457
+ "point_3d": 0.9365234375,
1458
+ "reward": 0.8486728444695473,
1459
+ "step": 504
1460
+ },
1461
+ {
1462
+ "3d_bbox": 0.7242622300982475,
1463
+ "adv": -0.0005710294717573561,
1464
+ "bbox": 0.9246156923472881,
1465
+ "completions/clipped_ratio": 0.0,
1466
+ "completions/max_length": 1591.75,
1467
+ "completions/max_terminated_length": 1591.75,
1468
+ "completions/mean_length": 1571.015625,
1469
+ "completions/mean_terminated_length": 1571.015625,
1470
+ "completions/min_length": 1551.625,
1471
+ "completions/min_terminated_length": 1551.625,
1472
+ "entropy": 0.2915982282720506,
1473
+ "epoch": 0.1806633733239238,
1474
+ "geo": 0.7993985265493393,
1475
+ "grad_norm": 0.408203125,
1476
+ "learning_rate": 7.482412060301508e-06,
1477
+ "loss": 0.0001,
1478
+ "num_tokens": 60505948.0,
1479
+ "point_2d": 0.7919921875,
1480
+ "point_3d": 0.8671875,
1481
+ "reward": 0.8270144015550613,
1482
+ "step": 512
1483
+ },
1484
+ {
1485
+ "3d_bbox": 0.7831042557954788,
1486
+ "adv": -0.000159245344093506,
1487
+ "bbox": 0.9960433393716812,
1488
+ "completions/clipped_ratio": 0.0,
1489
+ "completions/max_length": 1592.375,
1490
+ "completions/max_terminated_length": 1592.375,
1491
+ "completions/mean_length": 1574.84375,
1492
+ "completions/mean_terminated_length": 1574.84375,
1493
+ "completions/min_length": 1559.125,
1494
+ "completions/min_terminated_length": 1559.125,
1495
+ "entropy": 0.2845937805250287,
1496
+ "epoch": 0.1834862385321101,
1497
+ "geo": 0.9170719608664513,
1498
+ "grad_norm": 0.345703125,
1499
+ "learning_rate": 7.442211055276382e-06,
1500
+ "loss": -0.0,
1501
+ "num_tokens": 61450564.0,
1502
+ "point_2d": 1.0,
1503
+ "point_3d": 0.8408203125,
1504
+ "reward": 0.90499197691679,
1505
+ "step": 520
1506
+ },
1507
+ {
1508
+ "3d_bbox": 0.7821980714797974,
1509
+ "adv": -0.000531579293010509,
1510
+ "bbox": 0.9920122027397156,
1511
+ "completions/clipped_ratio": 0.0,
1512
+ "completions/max_length": 1595.375,
1513
+ "completions/max_terminated_length": 1595.375,
1514
+ "completions/mean_length": 1571.765625,
1515
+ "completions/mean_terminated_length": 1571.765625,
1516
+ "completions/min_length": 1551.5,
1517
+ "completions/min_terminated_length": 1551.5,
1518
+ "entropy": 0.2882155287079513,
1519
+ "epoch": 0.1863091037402964,
1520
+ "geo": 0.901853509247303,
1521
+ "grad_norm": 0.365234375,
1522
+ "learning_rate": 7.402010050251257e-06,
1523
+ "loss": 0.0002,
1524
+ "num_tokens": 62394336.0,
1525
+ "point_2d": 0.888671875,
1526
+ "point_3d": 0.9052734375,
1527
+ "reward": 0.8920388966798782,
1528
+ "step": 528
1529
+ },
1530
+ {
1531
+ "3d_bbox": 0.7537402808666229,
1532
+ "adv": -0.0004438982141437009,
1533
+ "bbox": 0.9946581870317459,
1534
+ "completions/clipped_ratio": 0.0,
1535
+ "completions/max_length": 1594.375,
1536
+ "completions/max_terminated_length": 1594.375,
1537
+ "completions/mean_length": 1574.3203125,
1538
+ "completions/mean_terminated_length": 1574.3203125,
1539
+ "completions/min_length": 1554.75,
1540
+ "completions/min_terminated_length": 1554.75,
1541
+ "entropy": 0.2886669235303998,
1542
+ "epoch": 0.18913196894848272,
1543
+ "geo": 0.8310712724924088,
1544
+ "grad_norm": 0.36328125,
1545
+ "learning_rate": 7.361809045226132e-06,
1546
+ "loss": 0.0002,
1547
+ "num_tokens": 63338834.0,
1548
+ "point_2d": 0.873046875,
1549
+ "point_3d": 0.904296875,
1550
+ "reward": 0.8814355507493019,
1551
+ "step": 536
1552
+ },
1553
+ {
1554
+ "3d_bbox": 0.6960996612906456,
1555
+ "adv": 0.0006180387658787367,
1556
+ "bbox": 0.869446316966787,
1557
+ "completions/clipped_ratio": 0.0,
1558
+ "completions/max_length": 1597.25,
1559
+ "completions/max_terminated_length": 1597.25,
1560
+ "completions/mean_length": 1571.88671875,
1561
+ "completions/mean_terminated_length": 1571.88671875,
1562
+ "completions/min_length": 1552.25,
1563
+ "completions/min_terminated_length": 1552.25,
1564
+ "entropy": 0.2887332271784544,
1565
+ "epoch": 0.19195483415666903,
1566
+ "geo": 0.826970137655735,
1567
+ "grad_norm": 0.267578125,
1568
+ "learning_rate": 7.321608040201006e-06,
1569
+ "loss": -0.0,
1570
+ "num_tokens": 64282685.0,
1571
+ "point_2d": 0.875,
1572
+ "point_3d": 0.8662109375,
1573
+ "reward": 0.8266892246901989,
1574
+ "step": 544
1575
+ },
1576
+ {
1577
+ "3d_bbox": 0.8654715269804001,
1578
+ "adv": -0.0008252508814621251,
1579
+ "bbox": 0.9932455867528915,
1580
+ "completions/clipped_ratio": 0.0,
1581
+ "completions/max_length": 1596.125,
1582
+ "completions/max_terminated_length": 1596.125,
1583
+ "completions/mean_length": 1572.1953125,
1584
+ "completions/mean_terminated_length": 1572.1953125,
1585
+ "completions/min_length": 1553.625,
1586
+ "completions/min_terminated_length": 1553.625,
1587
+ "entropy": 0.2873402568511665,
1588
+ "epoch": 0.19477769936485534,
1589
+ "geo": 0.872327484190464,
1590
+ "grad_norm": 0.2890625,
1591
+ "learning_rate": 7.28140703517588e-06,
1592
+ "loss": 0.0004,
1593
+ "num_tokens": 65226639.0,
1594
+ "point_2d": 0.8935546875,
1595
+ "point_3d": 0.9482421875,
1596
+ "reward": 0.9251285046339035,
1597
+ "step": 552
1598
+ },
1599
+ {
1600
+ "3d_bbox": 0.7851101011037827,
1601
+ "adv": -0.0011301148442726117,
1602
+ "bbox": 0.9379025921225548,
1603
+ "completions/clipped_ratio": 0.0,
1604
+ "completions/max_length": 1588.5,
1605
+ "completions/max_terminated_length": 1588.5,
1606
+ "completions/mean_length": 1571.4765625,
1607
+ "completions/mean_terminated_length": 1571.4765625,
1608
+ "completions/min_length": 1555.0,
1609
+ "completions/min_terminated_length": 1555.0,
1610
+ "entropy": 0.2852159277535975,
1611
+ "epoch": 0.19760056457304165,
1612
+ "geo": 0.690475556999445,
1613
+ "grad_norm": 0.3125,
1614
+ "learning_rate": 7.241206030150754e-06,
1615
+ "loss": 0.0006,
1616
+ "num_tokens": 66170305.0,
1617
+ "point_2d": 0.625,
1618
+ "point_3d": 0.837890625,
1619
+ "reward": 0.7964758351445198,
1620
+ "step": 560
1621
+ },
1622
+ {
1623
+ "3d_bbox": 0.7313521131873131,
1624
+ "adv": -0.000422738248744281,
1625
+ "bbox": 0.8720479012117721,
1626
+ "completions/clipped_ratio": 0.0,
1627
+ "completions/max_length": 1601.875,
1628
+ "completions/max_terminated_length": 1601.875,
1629
+ "completions/mean_length": 1574.91015625,
1630
+ "completions/mean_terminated_length": 1574.91015625,
1631
+ "completions/min_length": 1551.25,
1632
+ "completions/min_terminated_length": 1551.25,
1633
+ "entropy": 0.2880417066626251,
1634
+ "epoch": 0.20042342978122796,
1635
+ "geo": 0.8383861493784934,
1636
+ "grad_norm": 0.412109375,
1637
+ "learning_rate": 7.2010050251256295e-06,
1638
+ "loss": 0.0004,
1639
+ "num_tokens": 67114842.0,
1640
+ "point_2d": 0.8740234375,
1641
+ "point_3d": 0.84765625,
1642
+ "reward": 0.8312699124217033,
1643
+ "step": 568
1644
+ },
1645
+ {
1646
+ "3d_bbox": 0.6841592267155647,
1647
+ "adv": -0.0008141829575833981,
1648
+ "bbox": 0.9068506434559822,
1649
+ "completions/clipped_ratio": 0.0,
1650
+ "completions/max_length": 1597.5,
1651
+ "completions/max_terminated_length": 1597.5,
1652
+ "completions/mean_length": 1571.79296875,
1653
+ "completions/mean_terminated_length": 1571.79296875,
1654
+ "completions/min_length": 1546.75,
1655
+ "completions/min_terminated_length": 1546.75,
1656
+ "entropy": 0.2889366364106536,
1657
+ "epoch": 0.20324629498941427,
1658
+ "geo": 0.7308010272681713,
1659
+ "grad_norm": 0.2578125,
1660
+ "learning_rate": 7.160804020100504e-06,
1661
+ "loss": 0.0002,
1662
+ "num_tokens": 68058501.0,
1663
+ "point_2d": 0.8525390625,
1664
+ "point_3d": 0.794921875,
1665
+ "reward": 0.8096176832914352,
1666
+ "step": 576
1667
+ },
1668
+ {
1669
+ "3d_bbox": 0.5905975513160229,
1670
+ "adv": -0.00022596626513404772,
1671
+ "bbox": 0.9184868931770325,
1672
+ "completions/clipped_ratio": 0.0,
1673
+ "completions/max_length": 1597.5,
1674
+ "completions/max_terminated_length": 1597.5,
1675
+ "completions/mean_length": 1572.45703125,
1676
+ "completions/mean_terminated_length": 1572.45703125,
1677
+ "completions/min_length": 1551.125,
1678
+ "completions/min_terminated_length": 1551.125,
1679
+ "entropy": 0.2837987565435469,
1680
+ "epoch": 0.20606916019760058,
1681
+ "geo": 0.8267702832818031,
1682
+ "grad_norm": 0.314453125,
1683
+ "learning_rate": 7.120603015075378e-06,
1684
+ "loss": -0.0003,
1685
+ "num_tokens": 69002250.0,
1686
+ "point_2d": 0.8681640625,
1687
+ "point_3d": 0.7445197626948357,
1688
+ "reward": 0.7804420664906502,
1689
+ "step": 584
1690
+ },
1691
+ {
1692
+ "3d_bbox": 0.6897582598030567,
1693
+ "adv": -0.0004831653022847604,
1694
+ "bbox": 0.9742325842380524,
1695
+ "completions/clipped_ratio": 0.0,
1696
+ "completions/max_length": 1595.25,
1697
+ "completions/max_terminated_length": 1595.25,
1698
+ "completions/mean_length": 1572.9296875,
1699
+ "completions/mean_terminated_length": 1572.9296875,
1700
+ "completions/min_length": 1553.5,
1701
+ "completions/min_terminated_length": 1553.5,
1702
+ "entropy": 0.2835669592022896,
1703
+ "epoch": 0.20889202540578689,
1704
+ "geo": 0.8361655548214912,
1705
+ "grad_norm": 0.330078125,
1706
+ "learning_rate": 7.080402010050251e-06,
1707
+ "loss": 0.0004,
1708
+ "num_tokens": 69946400.0,
1709
+ "point_2d": 0.875,
1710
+ "point_3d": 0.8948184698820114,
1711
+ "reward": 0.8584523350000381,
1712
+ "step": 592
1713
+ },
1714
+ {
1715
+ "3d_bbox": 0.785271767526865,
1716
+ "adv": -0.00028875649149995297,
1717
+ "bbox": 0.9752914533019066,
1718
+ "completions/clipped_ratio": 0.0,
1719
+ "completions/max_length": 1599.0,
1720
+ "completions/max_terminated_length": 1599.0,
1721
+ "completions/mean_length": 1572.3671875,
1722
+ "completions/mean_terminated_length": 1572.3671875,
1723
+ "completions/min_length": 1549.875,
1724
+ "completions/min_terminated_length": 1549.875,
1725
+ "entropy": 0.29118923749774694,
1726
+ "epoch": 0.2117148906139732,
1727
+ "geo": 0.9338327869772911,
1728
+ "grad_norm": 0.3671875,
1729
+ "learning_rate": 7.040201005025126e-06,
1730
+ "loss": 0.0007,
1731
+ "num_tokens": 70890342.0,
1732
+ "point_2d": 0.998046875,
1733
+ "point_3d": 0.94140625,
1734
+ "reward": 0.9250040724873543,
1735
+ "step": 600
1736
+ },
1737
+ {
1738
+ "3d_bbox": 0.8766504302620888,
1739
+ "adv": -0.002171848598663928,
1740
+ "bbox": 0.9872715324163437,
1741
+ "completions/clipped_ratio": 0.0,
1742
+ "completions/max_length": 1601.125,
1743
+ "completions/max_terminated_length": 1601.125,
1744
+ "completions/mean_length": 1576.87890625,
1745
+ "completions/mean_terminated_length": 1576.87890625,
1746
+ "completions/min_length": 1556.375,
1747
+ "completions/min_terminated_length": 1556.375,
1748
+ "entropy": 0.28933983854949474,
1749
+ "epoch": 0.2145377558221595,
1750
+ "geo": 0.845995269715786,
1751
+ "grad_norm": 0.33203125,
1752
+ "learning_rate": 7e-06,
1753
+ "loss": 0.0003,
1754
+ "num_tokens": 71835359.0,
1755
+ "point_2d": 0.875,
1756
+ "point_3d": 0.9454848319292068,
1757
+ "reward": 0.9211016893386841,
1758
+ "step": 608
1759
+ },
1760
+ {
1761
+ "3d_bbox": 0.8059130385518074,
1762
+ "adv": 0.0005476666337926872,
1763
+ "bbox": 0.9907352328300476,
1764
+ "completions/clipped_ratio": 0.0,
1765
+ "completions/max_length": 1593.625,
1766
+ "completions/max_terminated_length": 1593.625,
1767
+ "completions/mean_length": 1570.9921875,
1768
+ "completions/mean_terminated_length": 1570.9921875,
1769
+ "completions/min_length": 1554.125,
1770
+ "completions/min_terminated_length": 1554.125,
1771
+ "entropy": 0.28401868836954236,
1772
+ "epoch": 0.2173606210303458,
1773
+ "geo": 0.8720198646187782,
1774
+ "grad_norm": 0.384765625,
1775
+ "learning_rate": 6.959798994974874e-06,
1776
+ "loss": -0.0004,
1777
+ "num_tokens": 72778933.0,
1778
+ "point_2d": 0.9248046875,
1779
+ "point_3d": 0.9135454967617989,
1780
+ "reward": 0.9087496176362038,
1781
+ "step": 616
1782
+ },
1783
+ {
1784
+ "3d_bbox": 0.8549554497003555,
1785
+ "adv": -0.0005420902471087174,
1786
+ "bbox": 0.9758080616593361,
1787
+ "completions/clipped_ratio": 0.0,
1788
+ "completions/max_length": 1593.0,
1789
+ "completions/max_terminated_length": 1593.0,
1790
+ "completions/mean_length": 1570.1640625,
1791
+ "completions/mean_terminated_length": 1570.1640625,
1792
+ "completions/min_length": 1549.0,
1793
+ "completions/min_terminated_length": 1549.0,
1794
+ "entropy": 0.29045473085716367,
1795
+ "epoch": 0.22018348623853212,
1796
+ "geo": 0.8916067183017731,
1797
+ "grad_norm": 0.3828125,
1798
+ "learning_rate": 6.919597989949749e-06,
1799
+ "loss": 0.0156,
1800
+ "num_tokens": 73722191.0,
1801
+ "point_2d": 1.0,
1802
+ "point_3d": 0.9639246314764023,
1803
+ "reward": 0.9486720338463783,
1804
+ "step": 624
1805
+ },
1806
+ {
1807
+ "3d_bbox": 0.7975610345602036,
1808
+ "adv": 0.0014872762099003012,
1809
+ "bbox": 0.9193930514156818,
1810
+ "completions/clipped_ratio": 0.0,
1811
+ "completions/max_length": 1600.625,
1812
+ "completions/max_terminated_length": 1600.625,
1813
+ "completions/mean_length": 1573.29296875,
1814
+ "completions/mean_terminated_length": 1573.29296875,
1815
+ "completions/min_length": 1553.0,
1816
+ "completions/min_terminated_length": 1553.0,
1817
+ "entropy": 0.28542913123965263,
1818
+ "epoch": 0.22300635144671843,
1819
+ "geo": 0.936805784702301,
1820
+ "grad_norm": 0.2734375,
1821
+ "learning_rate": 6.8793969849246235e-06,
1822
+ "loss": 0.0152,
1823
+ "num_tokens": 74666354.0,
1824
+ "point_2d": 0.9990234375,
1825
+ "point_3d": 0.93359375,
1826
+ "reward": 0.9123928248882294,
1827
+ "step": 632
1828
+ },
1829
+ {
1830
+ "3d_bbox": 0.6216618679463863,
1831
+ "adv": -0.0013801682944176719,
1832
+ "bbox": 0.9884917959570885,
1833
+ "completions/clipped_ratio": 0.0,
1834
+ "completions/max_length": 1601.375,
1835
+ "completions/max_terminated_length": 1601.375,
1836
+ "completions/mean_length": 1577.00390625,
1837
+ "completions/mean_terminated_length": 1577.00390625,
1838
+ "completions/min_length": 1554.0,
1839
+ "completions/min_terminated_length": 1554.0,
1840
+ "entropy": 0.28678226098418236,
1841
+ "epoch": 0.22582921665490474,
1842
+ "geo": 0.9096385985612869,
1843
+ "grad_norm": 0.42578125,
1844
+ "learning_rate": 6.839195979899498e-06,
1845
+ "loss": 0.0005,
1846
+ "num_tokens": 75611339.0,
1847
+ "point_2d": 0.8291015625,
1848
+ "point_3d": 0.7913028448820114,
1849
+ "reward": 0.8076395243406296,
1850
+ "step": 640
1851
+ },
1852
+ {
1853
+ "3d_bbox": 0.7582510337233543,
1854
+ "adv": -0.00019565537664334443,
1855
+ "bbox": 0.9911796301603317,
1856
+ "completions/clipped_ratio": 0.0,
1857
+ "completions/max_length": 1596.625,
1858
+ "completions/max_terminated_length": 1596.625,
1859
+ "completions/mean_length": 1575.71484375,
1860
+ "completions/mean_terminated_length": 1575.71484375,
1861
+ "completions/min_length": 1556.25,
1862
+ "completions/min_terminated_length": 1556.25,
1863
+ "entropy": 0.2863654876127839,
1864
+ "epoch": 0.22865208186309105,
1865
+ "geo": 0.9025658443570137,
1866
+ "grad_norm": 0.451171875,
1867
+ "learning_rate": 6.798994974874372e-06,
1868
+ "loss": -0.0001,
1869
+ "num_tokens": 76556042.0,
1870
+ "point_2d": 0.95703125,
1871
+ "point_3d": 0.8994140625,
1872
+ "reward": 0.9014689922332764,
1873
+ "step": 648
1874
+ },
1875
+ {
1876
+ "3d_bbox": 0.753610834479332,
1877
+ "adv": 0.00039293405916396296,
1878
+ "bbox": 0.9021090529859066,
1879
+ "completions/clipped_ratio": 0.0,
1880
+ "completions/max_length": 1592.625,
1881
+ "completions/max_terminated_length": 1592.625,
1882
+ "completions/mean_length": 1572.203125,
1883
+ "completions/mean_terminated_length": 1572.203125,
1884
+ "completions/min_length": 1552.25,
1885
+ "completions/min_terminated_length": 1552.25,
1886
+ "entropy": 0.286099414806813,
1887
+ "epoch": 0.23147494707127736,
1888
+ "geo": 0.9557241797447205,
1889
+ "grad_norm": 0.455078125,
1890
+ "learning_rate": 6.758793969849246e-06,
1891
+ "loss": 0.0002,
1892
+ "num_tokens": 77499886.0,
1893
+ "point_2d": 1.0,
1894
+ "point_3d": 0.91015625,
1895
+ "reward": 0.8914690390229225,
1896
+ "step": 656
1897
+ },
1898
+ {
1899
+ "3d_bbox": 0.831758588552475,
1900
+ "adv": -7.158980952226557e-05,
1901
+ "bbox": 0.9794979766011238,
1902
+ "completions/clipped_ratio": 0.0,
1903
+ "completions/max_length": 1595.5,
1904
+ "completions/max_terminated_length": 1595.5,
1905
+ "completions/mean_length": 1574.5,
1906
+ "completions/mean_terminated_length": 1574.5,
1907
+ "completions/min_length": 1558.75,
1908
+ "completions/min_terminated_length": 1558.75,
1909
+ "entropy": 0.2819104827940464,
1910
+ "epoch": 0.23429781227946367,
1911
+ "geo": 0.8940749987959862,
1912
+ "grad_norm": 0.294921875,
1913
+ "learning_rate": 6.718592964824121e-06,
1914
+ "loss": 0.0001,
1915
+ "num_tokens": 78444526.0,
1916
+ "point_2d": 1.0,
1917
+ "point_3d": 0.939453125,
1918
+ "reward": 0.9376774355769157,
1919
+ "step": 664
1920
+ },
1921
+ {
1922
+ "3d_bbox": 0.8210256397724152,
1923
+ "adv": 0.000540634344361024,
1924
+ "bbox": 0.9576755836606026,
1925
+ "completions/clipped_ratio": 0.0,
1926
+ "completions/max_length": 1594.125,
1927
+ "completions/max_terminated_length": 1594.125,
1928
+ "completions/mean_length": 1573.80859375,
1929
+ "completions/mean_terminated_length": 1573.80859375,
1930
+ "completions/min_length": 1553.25,
1931
+ "completions/min_terminated_length": 1553.25,
1932
+ "entropy": 0.2863646815530956,
1933
+ "epoch": 0.23712067748764998,
1934
+ "geo": 0.9599328562617302,
1935
+ "grad_norm": 0.28125,
1936
+ "learning_rate": 6.678391959798996e-06,
1937
+ "loss": -0.0001,
1938
+ "num_tokens": 79388949.0,
1939
+ "point_2d": 1.0,
1940
+ "point_3d": 0.92578125,
1941
+ "reward": 0.9261206015944481,
1942
+ "step": 672
1943
+ },
1944
+ {
1945
+ "3d_bbox": 0.8162731677293777,
1946
+ "adv": -0.0006006620242260396,
1947
+ "bbox": 0.9639560282230377,
1948
+ "completions/clipped_ratio": 0.0,
1949
+ "completions/max_length": 1593.75,
1950
+ "completions/max_terminated_length": 1593.75,
1951
+ "completions/mean_length": 1573.62890625,
1952
+ "completions/mean_terminated_length": 1573.62890625,
1953
+ "completions/min_length": 1557.125,
1954
+ "completions/min_terminated_length": 1557.125,
1955
+ "entropy": 0.2868205299600959,
1956
+ "epoch": 0.2399435426958363,
1957
+ "geo": 0.9127583503723145,
1958
+ "grad_norm": 0.44921875,
1959
+ "learning_rate": 6.63819095477387e-06,
1960
+ "loss": -0.0,
1961
+ "num_tokens": 80333190.0,
1962
+ "point_2d": 1.0,
1963
+ "point_3d": 0.9267578125,
1964
+ "reward": 0.926746778190136,
1965
+ "step": 680
1966
+ },
1967
+ {
1968
+ "3d_bbox": 0.7580128684639931,
1969
+ "adv": 0.0009847661577850886,
1970
+ "bbox": 0.9946200400590897,
1971
+ "completions/clipped_ratio": 0.0,
1972
+ "completions/max_length": 1598.0,
1973
+ "completions/max_terminated_length": 1598.0,
1974
+ "completions/mean_length": 1572.2109375,
1975
+ "completions/mean_terminated_length": 1572.2109375,
1976
+ "completions/min_length": 1552.0,
1977
+ "completions/min_terminated_length": 1552.0,
1978
+ "entropy": 0.286160247400403,
1979
+ "epoch": 0.2427664079040226,
1980
+ "geo": 0.9191197454929352,
1981
+ "grad_norm": 0.361328125,
1982
+ "learning_rate": 6.597989949748744e-06,
1983
+ "loss": 0.0001,
1984
+ "num_tokens": 81277036.0,
1985
+ "point_2d": 1.0,
1986
+ "point_3d": 0.8818359375,
1987
+ "reward": 0.9086171984672546,
1988
+ "step": 688
1989
+ },
1990
+ {
1991
+ "3d_bbox": 0.729152224957943,
1992
+ "adv": -0.0009101523137360346,
1993
+ "bbox": 0.9927223324775696,
1994
+ "completions/clipped_ratio": 0.0,
1995
+ "completions/max_length": 1592.875,
1996
+ "completions/max_terminated_length": 1592.875,
1997
+ "completions/mean_length": 1571.8515625,
1998
+ "completions/mean_terminated_length": 1571.8515625,
1999
+ "completions/min_length": 1553.375,
2000
+ "completions/min_terminated_length": 1553.375,
2001
+ "entropy": 0.28340971609577537,
2002
+ "epoch": 0.2455892731122089,
2003
+ "geo": 0.9836970791220665,
2004
+ "grad_norm": 0.34765625,
2005
+ "learning_rate": 6.557788944723618e-06,
2006
+ "loss": 0.0005,
2007
+ "num_tokens": 82220822.0,
2008
+ "point_2d": 1.0,
2009
+ "point_3d": 0.9014246314764023,
2010
+ "reward": 0.9058247655630112,
2011
+ "step": 696
2012
+ },
2013
+ {
2014
+ "3d_bbox": 0.7259362824261189,
2015
+ "adv": 0.0003162266129947966,
2016
+ "bbox": 0.922937985509634,
2017
+ "completions/clipped_ratio": 0.0,
2018
+ "completions/max_length": 1596.0,
2019
+ "completions/max_terminated_length": 1596.0,
2020
+ "completions/mean_length": 1575.2265625,
2021
+ "completions/mean_terminated_length": 1575.2265625,
2022
+ "completions/min_length": 1555.625,
2023
+ "completions/min_terminated_length": 1555.625,
2024
+ "entropy": 0.286905855871737,
2025
+ "epoch": 0.24841213832039521,
2026
+ "geo": 0.8376269713044167,
2027
+ "grad_norm": 0.30078125,
2028
+ "learning_rate": 6.5175879396984935e-06,
2029
+ "loss": 0.0,
2030
+ "num_tokens": 83165600.0,
2031
+ "point_2d": 0.9541015625,
2032
+ "point_3d": 0.8759765625,
2033
+ "reward": 0.8697381019592285,
2034
+ "step": 704
2035
+ },
2036
+ {
2037
+ "3d_bbox": 0.7864346951246262,
2038
+ "adv": -0.0007161469466154813,
2039
+ "bbox": 0.9958559349179268,
2040
+ "completions/clipped_ratio": 0.0,
2041
+ "completions/max_length": 1595.0,
2042
+ "completions/max_terminated_length": 1595.0,
2043
+ "completions/mean_length": 1572.4921875,
2044
+ "completions/mean_terminated_length": 1572.4921875,
2045
+ "completions/min_length": 1545.125,
2046
+ "completions/min_terminated_length": 1545.125,
2047
+ "entropy": 0.2886675829067826,
2048
+ "epoch": 0.2512350035285815,
2049
+ "geo": 0.8416303154081106,
2050
+ "grad_norm": 0.333984375,
2051
+ "learning_rate": 6.477386934673368e-06,
2052
+ "loss": 0.0008,
2053
+ "num_tokens": 84109606.0,
2054
+ "point_2d": 0.75,
2055
+ "point_3d": 0.947265625,
2056
+ "reward": 0.8698890581727028,
2057
+ "step": 712
2058
+ },
2059
+ {
2060
+ "3d_bbox": 0.657585222274065,
2061
+ "adv": -0.0004964938416378573,
2062
+ "bbox": 0.9940767288208008,
2063
+ "completions/clipped_ratio": 0.0,
2064
+ "completions/max_length": 1589.75,
2065
+ "completions/max_terminated_length": 1589.75,
2066
+ "completions/mean_length": 1571.515625,
2067
+ "completions/mean_terminated_length": 1571.515625,
2068
+ "completions/min_length": 1553.5,
2069
+ "completions/min_terminated_length": 1553.5,
2070
+ "entropy": 0.2854380081407726,
2071
+ "epoch": 0.25405786873676783,
2072
+ "geo": 0.8985359221696854,
2073
+ "grad_norm": 0.275390625,
2074
+ "learning_rate": 6.437185929648242e-06,
2075
+ "loss": 0.0,
2076
+ "num_tokens": 85053346.0,
2077
+ "point_2d": 1.0,
2078
+ "point_3d": 0.8623046875,
2079
+ "reward": 0.8784916624426842,
2080
+ "step": 720
2081
+ },
2082
+ {
2083
+ "3d_bbox": 0.8299823924899101,
2084
+ "adv": -0.0006786447083868552,
2085
+ "bbox": 0.9957724660634995,
2086
+ "completions/clipped_ratio": 0.0,
2087
+ "completions/max_length": 1599.75,
2088
+ "completions/max_terminated_length": 1599.75,
2089
+ "completions/mean_length": 1575.05078125,
2090
+ "completions/mean_terminated_length": 1575.05078125,
2091
+ "completions/min_length": 1553.625,
2092
+ "completions/min_terminated_length": 1553.625,
2093
+ "entropy": 0.2834048862569034,
2094
+ "epoch": 0.25688073394495414,
2095
+ "geo": 0.7720311819575727,
2096
+ "grad_norm": 0.353515625,
2097
+ "learning_rate": 6.396984924623116e-06,
2098
+ "loss": 0.0008,
2099
+ "num_tokens": 85997743.0,
2100
+ "point_2d": 0.875,
2101
+ "point_3d": 0.8583984375,
2102
+ "reward": 0.8897883221507072,
2103
+ "step": 728
2104
+ },
2105
+ {
2106
+ "3d_bbox": 0.814798042178154,
2107
+ "adv": -0.00043147145015609567,
2108
+ "bbox": 0.994331993162632,
2109
+ "completions/clipped_ratio": 0.0,
2110
+ "completions/max_length": 1603.0,
2111
+ "completions/max_terminated_length": 1603.0,
2112
+ "completions/mean_length": 1578.44921875,
2113
+ "completions/mean_terminated_length": 1578.44921875,
2114
+ "completions/min_length": 1556.25,
2115
+ "completions/min_terminated_length": 1556.25,
2116
+ "entropy": 0.2801376706920564,
2117
+ "epoch": 0.25970359915314045,
2118
+ "geo": 0.9146922156214714,
2119
+ "grad_norm": 0.25,
2120
+ "learning_rate": 6.3567839195979905e-06,
2121
+ "loss": 0.0001,
2122
+ "num_tokens": 86943082.0,
2123
+ "point_2d": 0.8798828125,
2124
+ "point_3d": 0.94921875,
2125
+ "reward": 0.909557893872261,
2126
+ "step": 736
2127
+ },
2128
+ {
2129
+ "3d_bbox": 0.7457513771951199,
2130
+ "adv": -1.4886696476423822e-05,
2131
+ "bbox": 0.9952592924237251,
2132
+ "completions/clipped_ratio": 0.0,
2133
+ "completions/max_length": 1592.875,
2134
+ "completions/max_terminated_length": 1592.875,
2135
+ "completions/mean_length": 1572.21484375,
2136
+ "completions/mean_terminated_length": 1572.21484375,
2137
+ "completions/min_length": 1553.625,
2138
+ "completions/min_terminated_length": 1553.625,
2139
+ "entropy": 0.28585455380380154,
2140
+ "epoch": 0.26252646436132676,
2141
+ "geo": 0.9565036073327065,
2142
+ "grad_norm": 0.345703125,
2143
+ "learning_rate": 6.316582914572866e-06,
2144
+ "loss": -0.0003,
2145
+ "num_tokens": 87886913.0,
2146
+ "point_2d": 1.0,
2147
+ "point_3d": 0.8095703125,
2148
+ "reward": 0.8876452520489693,
2149
+ "step": 744
2150
+ },
2151
+ {
2152
+ "3d_bbox": 0.6948670484125614,
2153
+ "adv": -0.00016846299786266172,
2154
+ "bbox": 0.8553073853254318,
2155
+ "completions/clipped_ratio": 0.0,
2156
+ "completions/max_length": 1594.375,
2157
+ "completions/max_terminated_length": 1594.375,
2158
+ "completions/mean_length": 1571.93359375,
2159
+ "completions/mean_terminated_length": 1571.93359375,
2160
+ "completions/min_length": 1552.0,
2161
+ "completions/min_terminated_length": 1552.0,
2162
+ "entropy": 0.28335432009771466,
2163
+ "epoch": 0.26534932956951307,
2164
+ "geo": 0.7740265503525734,
2165
+ "grad_norm": 0.421875,
2166
+ "learning_rate": 6.27638190954774e-06,
2167
+ "loss": 0.0162,
2168
+ "num_tokens": 88830776.0,
2169
+ "point_2d": 0.8701171875,
2170
+ "point_3d": 0.9082605689764023,
2171
+ "reward": 0.8321380317211151,
2172
+ "step": 752
2173
+ },
2174
+ {
2175
+ "3d_bbox": 0.7499241754412651,
2176
+ "adv": 0.0011103415663455962,
2177
+ "bbox": 0.9951865822076797,
2178
+ "completions/clipped_ratio": 0.0,
2179
+ "completions/max_length": 1592.75,
2180
+ "completions/max_terminated_length": 1592.75,
2181
+ "completions/mean_length": 1570.93359375,
2182
+ "completions/mean_terminated_length": 1570.93359375,
2183
+ "completions/min_length": 1552.25,
2184
+ "completions/min_terminated_length": 1552.25,
2185
+ "entropy": 0.28731182869523764,
2186
+ "epoch": 0.2681721947776994,
2187
+ "geo": 0.620879240334034,
2188
+ "grad_norm": 0.2197265625,
2189
+ "learning_rate": 6.236180904522614e-06,
2190
+ "loss": 0.0001,
2191
+ "num_tokens": 89774255.0,
2192
+ "point_2d": 0.7119140625,
2193
+ "point_3d": 0.8916015625,
2194
+ "reward": 0.8371566012501717,
2195
+ "step": 760
2196
+ },
2197
+ {
2198
+ "3d_bbox": 0.7572767026722431,
2199
+ "adv": 0.0006080263119656593,
2200
+ "bbox": 0.9954995736479759,
2201
+ "completions/clipped_ratio": 0.0,
2202
+ "completions/max_length": 1589.0,
2203
+ "completions/max_terminated_length": 1589.0,
2204
+ "completions/mean_length": 1569.015625,
2205
+ "completions/mean_terminated_length": 1569.015625,
2206
+ "completions/min_length": 1549.75,
2207
+ "completions/min_terminated_length": 1549.75,
2208
+ "entropy": 0.2817346737720072,
2209
+ "epoch": 0.2709950599858857,
2210
+ "geo": 0.8461784943938255,
2211
+ "grad_norm": 0.41015625,
2212
+ "learning_rate": 6.195979899497488e-06,
2213
+ "loss": -0.0,
2214
+ "num_tokens": 90717243.0,
2215
+ "point_2d": 1.0,
2216
+ "point_3d": 0.892578125,
2217
+ "reward": 0.9113386049866676,
2218
+ "step": 768
2219
+ },
2220
+ {
2221
+ "3d_bbox": 0.8124237954616547,
2222
+ "adv": -0.00010392009426141158,
2223
+ "bbox": 0.9695260003209114,
2224
+ "completions/clipped_ratio": 0.0,
2225
+ "completions/max_length": 1590.0,
2226
+ "completions/max_terminated_length": 1590.0,
2227
+ "completions/mean_length": 1571.890625,
2228
+ "completions/mean_terminated_length": 1571.890625,
2229
+ "completions/min_length": 1553.75,
2230
+ "completions/min_terminated_length": 1553.75,
2231
+ "entropy": 0.2867649751715362,
2232
+ "epoch": 0.273817925194072,
2233
+ "geo": 0.7172370981425047,
2234
+ "grad_norm": 0.29296875,
2235
+ "learning_rate": 6.155778894472362e-06,
2236
+ "loss": -0.0,
2237
+ "num_tokens": 91660903.0,
2238
+ "point_2d": 0.75,
2239
+ "point_3d": 0.9462890625,
2240
+ "reward": 0.8695597127079964,
2241
+ "step": 776
2242
+ },
2243
+ {
2244
+ "3d_bbox": 0.6247975081205368,
2245
+ "adv": 0.00014501793702947907,
2246
+ "bbox": 0.9793029651045799,
2247
+ "completions/clipped_ratio": 0.0,
2248
+ "completions/max_length": 1592.125,
2249
+ "completions/max_terminated_length": 1592.125,
2250
+ "completions/mean_length": 1571.03515625,
2251
+ "completions/mean_terminated_length": 1571.03515625,
2252
+ "completions/min_length": 1550.875,
2253
+ "completions/min_terminated_length": 1550.875,
2254
+ "entropy": 0.2792881946079433,
2255
+ "epoch": 0.2766407904022583,
2256
+ "geo": 0.8185150865465403,
2257
+ "grad_norm": 0.2890625,
2258
+ "learning_rate": 6.115577889447236e-06,
2259
+ "loss": 0.0,
2260
+ "num_tokens": 92604352.0,
2261
+ "point_2d": 0.8876953125,
2262
+ "point_3d": 0.8994140625,
2263
+ "reward": 0.8478024452924728,
2264
+ "step": 784
2265
+ },
2266
+ {
2267
+ "3d_bbox": 0.7522557079792023,
2268
+ "adv": 0.00011037173771910602,
2269
+ "bbox": 0.8692253082990646,
2270
+ "completions/clipped_ratio": 0.0,
2271
+ "completions/max_length": 1594.625,
2272
+ "completions/max_terminated_length": 1594.625,
2273
+ "completions/mean_length": 1572.27734375,
2274
+ "completions/mean_terminated_length": 1572.27734375,
2275
+ "completions/min_length": 1552.5,
2276
+ "completions/min_terminated_length": 1552.5,
2277
+ "entropy": 0.2874673060141504,
2278
+ "epoch": 0.2794636556104446,
2279
+ "geo": 0.9408122524619102,
2280
+ "grad_norm": 0.31640625,
2281
+ "learning_rate": 6.07537688442211e-06,
2282
+ "loss": 0.0144,
2283
+ "num_tokens": 93548159.0,
2284
+ "point_2d": 0.7587890625,
2285
+ "point_3d": 0.91796875,
2286
+ "reward": 0.8245597034692764,
2287
+ "step": 792
2288
+ },
2289
+ {
2290
+ "3d_bbox": 0.8205922544002533,
2291
+ "adv": 0.0001977139909286052,
2292
+ "bbox": 0.9971112534403801,
2293
+ "completions/clipped_ratio": 0.0,
2294
+ "completions/max_length": 1589.25,
2295
+ "completions/max_terminated_length": 1589.25,
2296
+ "completions/mean_length": 1571.43359375,
2297
+ "completions/mean_terminated_length": 1571.43359375,
2298
+ "completions/min_length": 1552.625,
2299
+ "completions/min_terminated_length": 1552.625,
2300
+ "entropy": 0.28211602522060275,
2301
+ "epoch": 0.2822865208186309,
2302
+ "geo": 0.8445243984460831,
2303
+ "grad_norm": 0.3203125,
2304
+ "learning_rate": 6.035175879396985e-06,
2305
+ "loss": 0.0007,
2306
+ "num_tokens": 94491662.0,
2307
+ "point_2d": 0.884765625,
2308
+ "point_3d": 0.9365234375,
2309
+ "reward": 0.9097481444478035,
2310
+ "step": 800
2311
+ }
2312
+ ],
2313
+ "logging_steps": 8,
2314
+ "max_steps": 2000,
2315
+ "num_input_tokens_seen": 94491662,
2316
+ "num_train_epochs": 1,
2317
+ "save_steps": 50,
2318
+ "stateful_callbacks": {
2319
+ "TrainerControl": {
2320
+ "args": {
2321
+ "should_epoch_stop": false,
2322
+ "should_evaluate": false,
2323
+ "should_log": false,
2324
+ "should_save": true,
2325
+ "should_training_stop": false
2326
+ },
2327
+ "attributes": {}
2328
+ }
2329
+ },
2330
+ "total_flos": 0.0,
2331
+ "train_batch_size": 1,
2332
+ "trial_name": null,
2333
+ "trial_params": null
2334
+ }
checkpoint-800/video_preprocessor_config.json ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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_pad": null,
10
+ "do_rescale": true,
11
+ "do_resize": true,
12
+ "do_sample_frames": false,
13
+ "fps": null,
14
+ "image_mean": [
15
+ 0.48145466,
16
+ 0.4578275,
17
+ 0.40821073
18
+ ],
19
+ "image_std": [
20
+ 0.26862954,
21
+ 0.26130258,
22
+ 0.27577711
23
+ ],
24
+ "input_data_format": null,
25
+ "max_frames": 768,
26
+ "max_pixels": 12845056,
27
+ "merge_size": 2,
28
+ "min_frames": 4,
29
+ "min_pixels": 3136,
30
+ "num_frames": 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
+ "size_divisor": null,
41
+ "temporal_patch_size": 2,
42
+ "video_metadata": null,
43
+ "video_processor_type": "Qwen2VLVideoProcessor"
44
+ }
checkpoint-800/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-850/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
+ }
checkpoint-850/chat_template.jinja ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
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
2
+ You are a helpful assistant.<|im_end|>
3
+ {% endif %}<|im_start|>{{ message['role'] }}
4
+ {% if message['content'] is string %}{{ message['content'] }}<|im_end|>
5
+ {% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>
6
+ {% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant
7
+ {% endif %}
checkpoint-850/config.json ADDED
@@ -0,0 +1,143 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen2_5_PVLForConditionalGeneration"
4
+ ],
5
+ "attention_dropout": 0.0,
6
+ "dtype": "bfloat16",
7
+ "eos_token_id": 151645,
8
+ "hidden_act": "silu",
9
+ "hidden_size": 2048,
10
+ "image_token_id": 151655,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 11008,
13
+ "max_position_embeddings": 128000,
14
+ "max_window_layers": 70,
15
+ "model_type": "qwen2_5_vl",
16
+ "num_attention_heads": 16,
17
+ "num_hidden_layers": 36,
18
+ "num_key_value_heads": 2,
19
+ "pad_token_id": 151643,
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
+ "architectures": [
34
+ "Qwen2_5_VLForConditionalGeneration"
35
+ ],
36
+ "attention_dropout": 0.0,
37
+ "bos_token_id": 151643,
38
+ "dtype": "bfloat16",
39
+ "eos_token_id": 151645,
40
+ "hidden_act": "silu",
41
+ "hidden_size": 2048,
42
+ "image_token_id": null,
43
+ "initializer_range": 0.02,
44
+ "intermediate_size": 11008,
45
+ "layer_types": [
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
+ "full_attention",
74
+ "full_attention",
75
+ "full_attention",
76
+ "full_attention",
77
+ "full_attention",
78
+ "full_attention",
79
+ "full_attention",
80
+ "full_attention",
81
+ "full_attention"
82
+ ],
83
+ "max_position_embeddings": 128000,
84
+ "max_window_layers": 70,
85
+ "model_type": "qwen2_5_vl_text",
86
+ "num_attention_heads": 16,
87
+ "num_hidden_layers": 36,
88
+ "num_key_value_heads": 2,
89
+ "rms_norm_eps": 1e-06,
90
+ "rope_scaling": {
91
+ "mrope_section": [
92
+ 16,
93
+ 24,
94
+ 24
95
+ ],
96
+ "rope_type": "default",
97
+ "type": "default"
98
+ },
99
+ "rope_theta": 1000000.0,
100
+ "sliding_window": null,
101
+ "tie_word_embeddings": true,
102
+ "use_cache": true,
103
+ "use_sliding_window": false,
104
+ "video_token_id": null,
105
+ "vision_end_token_id": 151653,
106
+ "vision_start_token_id": 151652,
107
+ "vision_token_id": 151654,
108
+ "vocab_size": 151936
109
+ },
110
+ "transformers_version": "4.56.2",
111
+ "use_cache": true,
112
+ "use_sliding_window": false,
113
+ "video_token_id": 151656,
114
+ "vision_config": {
115
+ "depth": 32,
116
+ "dtype": "bfloat16",
117
+ "fullatt_block_indexes": [
118
+ 7,
119
+ 15,
120
+ 23,
121
+ 31
122
+ ],
123
+ "hidden_act": "silu",
124
+ "hidden_size": 1280,
125
+ "in_channels": 3,
126
+ "in_chans": 3,
127
+ "initializer_range": 0.02,
128
+ "intermediate_size": 3420,
129
+ "model_type": "qwen2_5_vl",
130
+ "num_heads": 16,
131
+ "out_hidden_size": 2048,
132
+ "patch_size": 14,
133
+ "spatial_merge_size": 2,
134
+ "spatial_patch_size": 14,
135
+ "temporal_patch_size": 2,
136
+ "tokens_per_second": 2,
137
+ "window_size": 112
138
+ },
139
+ "vision_end_token_id": 151653,
140
+ "vision_start_token_id": 151652,
141
+ "vision_token_id": 151654,
142
+ "vocab_size": 151936
143
+ }
checkpoint-850/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": "4.56.2"
11
+ }
checkpoint-850/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-850/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-850/preprocessor_config.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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_rescale": true,
11
+ "do_resize": true,
12
+ "image_mean": [
13
+ 0.48145466,
14
+ 0.4578275,
15
+ 0.40821073
16
+ ],
17
+ "image_processor_type": "Qwen2VLImageProcessorFast",
18
+ "image_std": [
19
+ 0.26862954,
20
+ 0.26130258,
21
+ 0.27577711
22
+ ],
23
+ "input_data_format": null,
24
+ "max_pixels": 12845056,
25
+ "merge_size": 2,
26
+ "min_pixels": 3136,
27
+ "patch_size": 14,
28
+ "processor_class": "Qwen2_5_VLProcessor",
29
+ "resample": 3,
30
+ "rescale_factor": 0.00392156862745098,
31
+ "return_tensors": null,
32
+ "size": {
33
+ "longest_edge": 12845056,
34
+ "shortest_edge": 3136
35
+ },
36
+ "temporal_patch_size": 2
37
+ }
checkpoint-850/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
+ }
checkpoint-850/tokenizer_config.json ADDED
@@ -0,0 +1,212 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "max_length": null,
203
+ "model_max_length": 131072,
204
+ "pad_to_multiple_of": null,
205
+ "pad_token": "<|endoftext|>",
206
+ "pad_token_type_id": 0,
207
+ "padding_side": "left",
208
+ "processor_class": "Qwen2_5_VLProcessor",
209
+ "split_special_tokens": false,
210
+ "tokenizer_class": "Qwen2Tokenizer",
211
+ "unk_token": null
212
+ }
checkpoint-850/trainer_state.json ADDED
@@ -0,0 +1,2472 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.29992942836979536,
6
+ "eval_steps": 500,
7
+ "global_step": 850,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "3d_bbox": 0.6610370799899101,
14
+ "adv": 0.000926484264937244,
15
+ "bbox": 0.8075725510716438,
16
+ "completions/clipped_ratio": 0.0,
17
+ "completions/max_length": 1592.875,
18
+ "completions/max_terminated_length": 1592.875,
19
+ "completions/mean_length": 1573.1640625,
20
+ "completions/mean_terminated_length": 1573.1640625,
21
+ "completions/min_length": 1554.875,
22
+ "completions/min_terminated_length": 1554.875,
23
+ "entropy": 0.2564338077791035,
24
+ "epoch": 0.0028228652081863093,
25
+ "geo": 0.7611365988850594,
26
+ "grad_norm": 0.439453125,
27
+ "learning_rate": 7e-06,
28
+ "loss": -0.0002,
29
+ "num_tokens": 943810.0,
30
+ "point_2d": 0.875,
31
+ "point_3d": 0.638671875,
32
+ "reward": 0.7455703690648079,
33
+ "step": 8
34
+ },
35
+ {
36
+ "3d_bbox": 0.6985021606087685,
37
+ "adv": 0.0008333708265126916,
38
+ "bbox": 0.9951911941170692,
39
+ "completions/clipped_ratio": 0.0,
40
+ "completions/max_length": 1594.25,
41
+ "completions/max_terminated_length": 1594.25,
42
+ "completions/mean_length": 1573.64453125,
43
+ "completions/mean_terminated_length": 1573.64453125,
44
+ "completions/min_length": 1552.375,
45
+ "completions/min_terminated_length": 1552.375,
46
+ "entropy": 0.2595143502112478,
47
+ "epoch": 0.0056457304163726185,
48
+ "geo": 0.8901385143399239,
49
+ "grad_norm": 0.341796875,
50
+ "learning_rate": 9.974874371859297e-06,
51
+ "loss": 0.0564,
52
+ "num_tokens": 1888015.0,
53
+ "point_2d": 1.0,
54
+ "point_3d": 0.5947265625,
55
+ "reward": 0.8221049904823303,
56
+ "step": 16
57
+ },
58
+ {
59
+ "3d_bbox": 0.6706115230917931,
60
+ "adv": -0.11667590066201683,
61
+ "bbox": 0.9641365110874176,
62
+ "completions/clipped_ratio": 0.0,
63
+ "completions/max_length": 1588.75,
64
+ "completions/max_terminated_length": 1588.75,
65
+ "completions/mean_length": 1568.69921875,
66
+ "completions/mean_terminated_length": 1568.69921875,
67
+ "completions/min_length": 1549.625,
68
+ "completions/min_terminated_length": 1549.625,
69
+ "entropy": 0.2620172018650919,
70
+ "epoch": 0.008468595624558928,
71
+ "geo": 0.8719085603952408,
72
+ "grad_norm": 0.31640625,
73
+ "learning_rate": 9.934673366834172e-06,
74
+ "loss": 0.0434,
75
+ "num_tokens": 2830754.0,
76
+ "point_2d": 0.96484375,
77
+ "point_3d": 0.7060546875,
78
+ "reward": 0.8264116197824478,
79
+ "step": 24
80
+ },
81
+ {
82
+ "3d_bbox": 0.6768141128122807,
83
+ "adv": 0.00011848201484099263,
84
+ "bbox": 0.8640391379594803,
85
+ "completions/clipped_ratio": 0.0,
86
+ "completions/max_length": 1599.875,
87
+ "completions/max_terminated_length": 1599.875,
88
+ "completions/mean_length": 1579.40625,
89
+ "completions/mean_terminated_length": 1579.40625,
90
+ "completions/min_length": 1563.0,
91
+ "completions/min_terminated_length": 1563.0,
92
+ "entropy": 0.26498199813067913,
93
+ "epoch": 0.011291460832745237,
94
+ "geo": 0.8661364912986755,
95
+ "grad_norm": 0.36328125,
96
+ "learning_rate": 9.894472361809046e-06,
97
+ "loss": -0.0005,
98
+ "num_tokens": 3776674.0,
99
+ "point_2d": 0.8642578125,
100
+ "point_3d": 0.603515625,
101
+ "reward": 0.752156674861908,
102
+ "step": 32
103
+ },
104
+ {
105
+ "3d_bbox": 0.6714089289307594,
106
+ "adv": -0.000722882674608627,
107
+ "bbox": 0.9624817371368408,
108
+ "completions/clipped_ratio": 0.0,
109
+ "completions/max_length": 1593.75,
110
+ "completions/max_terminated_length": 1593.75,
111
+ "completions/mean_length": 1575.52734375,
112
+ "completions/mean_terminated_length": 1575.52734375,
113
+ "completions/min_length": 1559.875,
114
+ "completions/min_terminated_length": 1559.875,
115
+ "entropy": 0.2738626003265381,
116
+ "epoch": 0.014114326040931546,
117
+ "geo": 0.8932898789644241,
118
+ "grad_norm": 0.341796875,
119
+ "learning_rate": 9.85427135678392e-06,
120
+ "loss": 0.0002,
121
+ "num_tokens": 4721169.0,
122
+ "point_2d": 0.99609375,
123
+ "point_3d": 0.6064453125,
124
+ "reward": 0.8091074377298355,
125
+ "step": 40
126
+ },
127
+ {
128
+ "3d_bbox": 0.6575617864727974,
129
+ "adv": -0.0007460190936399158,
130
+ "bbox": 0.9782034233212471,
131
+ "completions/clipped_ratio": 0.0,
132
+ "completions/max_length": 1598.75,
133
+ "completions/max_terminated_length": 1598.75,
134
+ "completions/mean_length": 1581.515625,
135
+ "completions/mean_terminated_length": 1581.515625,
136
+ "completions/min_length": 1563.875,
137
+ "completions/min_terminated_length": 1563.875,
138
+ "entropy": 0.26874606031924486,
139
+ "epoch": 0.016937191249117856,
140
+ "geo": 0.922828696668148,
141
+ "grad_norm": 0.384765625,
142
+ "learning_rate": 9.814070351758794e-06,
143
+ "loss": 0.0001,
144
+ "num_tokens": 5667501.0,
145
+ "point_2d": 0.9404296875,
146
+ "point_3d": 0.6171875,
147
+ "reward": 0.7983455955982208,
148
+ "step": 48
149
+ },
150
+ {
151
+ "3d_bbox": 0.7516707628965378,
152
+ "adv": 0.0002944665375252953,
153
+ "bbox": 0.870270783547312,
154
+ "completions/clipped_ratio": 0.0,
155
+ "completions/max_length": 1605.5,
156
+ "completions/max_terminated_length": 1605.5,
157
+ "completions/mean_length": 1583.109375,
158
+ "completions/mean_terminated_length": 1583.109375,
159
+ "completions/min_length": 1563.125,
160
+ "completions/min_terminated_length": 1563.125,
161
+ "entropy": 0.27441642340272665,
162
+ "epoch": 0.019760056457304165,
163
+ "geo": 0.8707154765725136,
164
+ "grad_norm": 0.35546875,
165
+ "learning_rate": 9.773869346733669e-06,
166
+ "loss": 0.0002,
167
+ "num_tokens": 6614089.0,
168
+ "point_2d": 0.875,
169
+ "point_3d": 0.728515625,
170
+ "reward": 0.8063642866909504,
171
+ "step": 56
172
+ },
173
+ {
174
+ "3d_bbox": 0.7704295367002487,
175
+ "adv": 0.0006564971263287589,
176
+ "bbox": 0.8711516009643674,
177
+ "completions/clipped_ratio": 0.0,
178
+ "completions/max_length": 1604.75,
179
+ "completions/max_terminated_length": 1604.75,
180
+ "completions/mean_length": 1582.09375,
181
+ "completions/mean_terminated_length": 1582.09375,
182
+ "completions/min_length": 1562.25,
183
+ "completions/min_terminated_length": 1562.25,
184
+ "entropy": 0.2721484359353781,
185
+ "epoch": 0.022582921665490474,
186
+ "geo": 0.8998556286096573,
187
+ "grad_norm": 0.373046875,
188
+ "learning_rate": 9.733668341708543e-06,
189
+ "loss": 0.0156,
190
+ "num_tokens": 7560513.0,
191
+ "point_2d": 0.875,
192
+ "point_3d": 0.7880859375,
193
+ "reward": 0.8261667564511299,
194
+ "step": 64
195
+ },
196
+ {
197
+ "3d_bbox": 0.6965150758624077,
198
+ "adv": 5.7399442994210403e-05,
199
+ "bbox": 0.9924400746822357,
200
+ "completions/clipped_ratio": 0.0,
201
+ "completions/max_length": 1603.875,
202
+ "completions/max_terminated_length": 1603.875,
203
+ "completions/mean_length": 1580.8203125,
204
+ "completions/mean_terminated_length": 1580.8203125,
205
+ "completions/min_length": 1562.0,
206
+ "completions/min_terminated_length": 1562.0,
207
+ "entropy": 0.2694994225166738,
208
+ "epoch": 0.025405786873676783,
209
+ "geo": 0.7992948293685913,
210
+ "grad_norm": 0.40625,
211
+ "learning_rate": 9.693467336683417e-06,
212
+ "loss": -0.0004,
213
+ "num_tokens": 8506787.0,
214
+ "point_2d": 0.875,
215
+ "point_3d": 0.6201171875,
216
+ "reward": 0.796018086373806,
217
+ "step": 72
218
+ },
219
+ {
220
+ "3d_bbox": 0.9022366181015968,
221
+ "adv": 0.000732647989934776,
222
+ "bbox": 0.9670402258634567,
223
+ "completions/clipped_ratio": 0.0,
224
+ "completions/max_length": 1598.5,
225
+ "completions/max_terminated_length": 1598.5,
226
+ "completions/mean_length": 1577.83203125,
227
+ "completions/mean_terminated_length": 1577.83203125,
228
+ "completions/min_length": 1560.0,
229
+ "completions/min_terminated_length": 1560.0,
230
+ "entropy": 0.2729479926638305,
231
+ "epoch": 0.028228652081863093,
232
+ "geo": 0.770210200920701,
233
+ "grad_norm": 0.28515625,
234
+ "learning_rate": 9.653266331658291e-06,
235
+ "loss": 0.0285,
236
+ "num_tokens": 9452032.0,
237
+ "point_2d": 0.7578125,
238
+ "point_3d": 0.8740234375,
239
+ "reward": 0.8752781972289085,
240
+ "step": 80
241
+ },
242
+ {
243
+ "3d_bbox": 0.8243749290704727,
244
+ "adv": 0.0004942354327113208,
245
+ "bbox": 0.9483200311660767,
246
+ "completions/clipped_ratio": 0.0,
247
+ "completions/max_length": 1599.0,
248
+ "completions/max_terminated_length": 1599.0,
249
+ "completions/mean_length": 1578.5703125,
250
+ "completions/mean_terminated_length": 1578.5703125,
251
+ "completions/min_length": 1564.875,
252
+ "completions/min_terminated_length": 1564.875,
253
+ "entropy": 0.27196697471663356,
254
+ "epoch": 0.031051517290049402,
255
+ "geo": 0.8520620986819267,
256
+ "grad_norm": 0.337890625,
257
+ "learning_rate": 9.613065326633166e-06,
258
+ "loss": 0.0,
259
+ "num_tokens": 10397602.0,
260
+ "point_2d": 0.84375,
261
+ "point_3d": 0.7431640625,
262
+ "reward": 0.8399022594094276,
263
+ "step": 88
264
+ },
265
+ {
266
+ "3d_bbox": 0.7930564805865288,
267
+ "adv": -0.05789744130743202,
268
+ "bbox": 0.9452124014496803,
269
+ "completions/clipped_ratio": 0.0,
270
+ "completions/max_length": 1600.125,
271
+ "completions/max_terminated_length": 1600.125,
272
+ "completions/mean_length": 1579.25390625,
273
+ "completions/mean_terminated_length": 1579.25390625,
274
+ "completions/min_length": 1560.25,
275
+ "completions/min_terminated_length": 1560.25,
276
+ "entropy": 0.2732193539850414,
277
+ "epoch": 0.03387438249823571,
278
+ "geo": 0.8466203399002552,
279
+ "grad_norm": 0.35546875,
280
+ "learning_rate": 9.57286432160804e-06,
281
+ "loss": 0.0143,
282
+ "num_tokens": 11343267.0,
283
+ "point_2d": 0.7138671875,
284
+ "point_3d": 0.8203125,
285
+ "reward": 0.8181121572852135,
286
+ "step": 96
287
+ },
288
+ {
289
+ "3d_bbox": 0.6107643470168114,
290
+ "adv": -0.0019174332555849105,
291
+ "bbox": 0.9693493247032166,
292
+ "completions/clipped_ratio": 0.0,
293
+ "completions/max_length": 1605.375,
294
+ "completions/max_terminated_length": 1605.375,
295
+ "completions/mean_length": 1580.6875,
296
+ "completions/mean_terminated_length": 1580.6875,
297
+ "completions/min_length": 1559.75,
298
+ "completions/min_terminated_length": 1559.75,
299
+ "entropy": 0.2712484383955598,
300
+ "epoch": 0.03669724770642202,
301
+ "geo": 0.760684534907341,
302
+ "grad_norm": 0.392578125,
303
+ "learning_rate": 9.532663316582916e-06,
304
+ "loss": 0.0006,
305
+ "num_tokens": 12289259.0,
306
+ "point_2d": 0.849609375,
307
+ "point_3d": 0.7109375,
308
+ "reward": 0.7851651459932327,
309
+ "step": 104
310
+ },
311
+ {
312
+ "3d_bbox": 0.7011105790734291,
313
+ "adv": 0.000332610031364311,
314
+ "bbox": 0.997136726975441,
315
+ "completions/clipped_ratio": 0.0,
316
+ "completions/max_length": 1601.5,
317
+ "completions/max_terminated_length": 1601.5,
318
+ "completions/mean_length": 1578.38671875,
319
+ "completions/mean_terminated_length": 1578.38671875,
320
+ "completions/min_length": 1560.0,
321
+ "completions/min_terminated_length": 1560.0,
322
+ "entropy": 0.2741631781682372,
323
+ "epoch": 0.03952011291460833,
324
+ "geo": 0.868082121014595,
325
+ "grad_norm": 0.349609375,
326
+ "learning_rate": 9.49246231155779e-06,
327
+ "loss": -0.0005,
328
+ "num_tokens": 13234494.0,
329
+ "point_2d": 0.875,
330
+ "point_3d": 0.759765625,
331
+ "reward": 0.8332532122731209,
332
+ "step": 112
333
+ },
334
+ {
335
+ "3d_bbox": 0.6414853371679783,
336
+ "adv": 6.49127487122314e-05,
337
+ "bbox": 0.9946233779191971,
338
+ "completions/clipped_ratio": 0.0,
339
+ "completions/max_length": 1599.875,
340
+ "completions/max_terminated_length": 1599.875,
341
+ "completions/mean_length": 1580.48828125,
342
+ "completions/mean_terminated_length": 1580.48828125,
343
+ "completions/min_length": 1562.5,
344
+ "completions/min_terminated_length": 1562.5,
345
+ "entropy": 0.2765682805329561,
346
+ "epoch": 0.04234297812279464,
347
+ "geo": 0.7812720276415348,
348
+ "grad_norm": 0.392578125,
349
+ "learning_rate": 9.452261306532664e-06,
350
+ "loss": -0.0002,
351
+ "num_tokens": 14180739.0,
352
+ "point_2d": 0.875,
353
+ "point_3d": 0.7275390625,
354
+ "reward": 0.8096619322896004,
355
+ "step": 120
356
+ },
357
+ {
358
+ "3d_bbox": 0.696478821337223,
359
+ "adv": 0.00018893864194069465,
360
+ "bbox": 0.9915411844849586,
361
+ "completions/clipped_ratio": 0.0,
362
+ "completions/max_length": 1600.75,
363
+ "completions/max_terminated_length": 1600.75,
364
+ "completions/mean_length": 1581.48828125,
365
+ "completions/mean_terminated_length": 1581.48828125,
366
+ "completions/min_length": 1566.25,
367
+ "completions/min_terminated_length": 1566.25,
368
+ "entropy": 0.27939224475994706,
369
+ "epoch": 0.04516584333098095,
370
+ "geo": 0.9225678443908691,
371
+ "grad_norm": 0.359375,
372
+ "learning_rate": 9.412060301507539e-06,
373
+ "loss": 0.0004,
374
+ "num_tokens": 15126808.0,
375
+ "point_2d": 1.0,
376
+ "point_3d": 0.8037109375,
377
+ "reward": 0.8729327321052551,
378
+ "step": 128
379
+ },
380
+ {
381
+ "3d_bbox": 0.7811713069677353,
382
+ "adv": -0.00017097587488024146,
383
+ "bbox": 0.983598880469799,
384
+ "completions/clipped_ratio": 0.0,
385
+ "completions/max_length": 1599.0,
386
+ "completions/max_terminated_length": 1599.0,
387
+ "completions/mean_length": 1575.9765625,
388
+ "completions/mean_terminated_length": 1575.9765625,
389
+ "completions/min_length": 1560.875,
390
+ "completions/min_terminated_length": 1560.875,
391
+ "entropy": 0.2712345658801496,
392
+ "epoch": 0.04798870853916726,
393
+ "geo": 0.829373586922884,
394
+ "grad_norm": 0.365234375,
395
+ "learning_rate": 9.371859296482413e-06,
396
+ "loss": 0.0002,
397
+ "num_tokens": 16071682.0,
398
+ "point_2d": 0.75,
399
+ "point_3d": 0.76171875,
400
+ "reward": 0.819122239947319,
401
+ "step": 136
402
+ },
403
+ {
404
+ "3d_bbox": 0.6566238105297089,
405
+ "adv": 5.832723485355018e-05,
406
+ "bbox": 0.9859664067626,
407
+ "completions/clipped_ratio": 0.0,
408
+ "completions/max_length": 1597.5,
409
+ "completions/max_terminated_length": 1597.5,
410
+ "completions/mean_length": 1580.12890625,
411
+ "completions/mean_terminated_length": 1580.12890625,
412
+ "completions/min_length": 1565.0,
413
+ "completions/min_terminated_length": 1565.0,
414
+ "entropy": 0.2754409732297063,
415
+ "epoch": 0.05081157374735357,
416
+ "geo": 0.8895541727542877,
417
+ "grad_norm": 0.318359375,
418
+ "learning_rate": 9.331658291457287e-06,
419
+ "loss": 0.0218,
420
+ "num_tokens": 17017499.0,
421
+ "point_2d": 1.0,
422
+ "point_3d": 0.6328125,
423
+ "reward": 0.8188506811857224,
424
+ "step": 144
425
+ },
426
+ {
427
+ "3d_bbox": 0.655942264944315,
428
+ "adv": 0.00024080171715468168,
429
+ "bbox": 0.9830196872353554,
430
+ "completions/clipped_ratio": 0.0,
431
+ "completions/max_length": 1600.125,
432
+ "completions/max_terminated_length": 1600.125,
433
+ "completions/mean_length": 1581.3828125,
434
+ "completions/mean_terminated_length": 1581.3828125,
435
+ "completions/min_length": 1564.0,
436
+ "completions/min_terminated_length": 1564.0,
437
+ "entropy": 0.2764031286351383,
438
+ "epoch": 0.053634438955539876,
439
+ "geo": 0.8772727139294147,
440
+ "grad_norm": 0.294921875,
441
+ "learning_rate": 9.291457286432161e-06,
442
+ "loss": 0.015,
443
+ "num_tokens": 17963757.0,
444
+ "point_2d": 0.875,
445
+ "point_3d": 0.7880859375,
446
+ "reward": 0.8255119547247887,
447
+ "step": 152
448
+ },
449
+ {
450
+ "3d_bbox": 0.8087940737605095,
451
+ "adv": -0.062093503400319605,
452
+ "bbox": 0.9799230694770813,
453
+ "completions/clipped_ratio": 0.0,
454
+ "completions/max_length": 1598.5,
455
+ "completions/max_terminated_length": 1598.5,
456
+ "completions/mean_length": 1579.2421875,
457
+ "completions/mean_terminated_length": 1579.2421875,
458
+ "completions/min_length": 1560.25,
459
+ "completions/min_terminated_length": 1560.25,
460
+ "entropy": 0.27893486013635993,
461
+ "epoch": 0.056457304163726185,
462
+ "geo": 0.8533996716141701,
463
+ "grad_norm": 0.314453125,
464
+ "learning_rate": 9.251256281407036e-06,
465
+ "loss": 0.0157,
466
+ "num_tokens": 18909307.0,
467
+ "point_2d": 0.859375,
468
+ "point_3d": 0.814453125,
469
+ "reward": 0.8656363189220428,
470
+ "step": 160
471
+ },
472
+ {
473
+ "3d_bbox": 0.7417622283101082,
474
+ "adv": 0.00038278038937278325,
475
+ "bbox": 0.9947961643338203,
476
+ "completions/clipped_ratio": 0.0,
477
+ "completions/max_length": 1596.875,
478
+ "completions/max_terminated_length": 1596.875,
479
+ "completions/mean_length": 1578.4765625,
480
+ "completions/mean_terminated_length": 1578.4765625,
481
+ "completions/min_length": 1562.375,
482
+ "completions/min_terminated_length": 1562.375,
483
+ "entropy": 0.2772809020243585,
484
+ "epoch": 0.059280169371912494,
485
+ "geo": 0.8446554020047188,
486
+ "grad_norm": 0.337890625,
487
+ "learning_rate": 9.21105527638191e-06,
488
+ "loss": -0.0001,
489
+ "num_tokens": 19854893.0,
490
+ "point_2d": 0.875,
491
+ "point_3d": 0.7666015625,
492
+ "reward": 0.844539999961853,
493
+ "step": 168
494
+ },
495
+ {
496
+ "3d_bbox": 0.7199448123574257,
497
+ "adv": -0.00025279237161157653,
498
+ "bbox": 0.9939833134412766,
499
+ "completions/clipped_ratio": 0.0,
500
+ "completions/max_length": 1599.375,
501
+ "completions/max_terminated_length": 1599.375,
502
+ "completions/mean_length": 1577.875,
503
+ "completions/mean_terminated_length": 1577.875,
504
+ "completions/min_length": 1560.875,
505
+ "completions/min_terminated_length": 1560.875,
506
+ "entropy": 0.2748410953208804,
507
+ "epoch": 0.062103034580098804,
508
+ "geo": 0.8426482602953911,
509
+ "grad_norm": 0.353515625,
510
+ "learning_rate": 9.170854271356784e-06,
511
+ "loss": 0.0001,
512
+ "num_tokens": 20800093.0,
513
+ "point_2d": 0.8935546875,
514
+ "point_3d": 0.7666015625,
515
+ "reward": 0.8435210958123207,
516
+ "step": 176
517
+ },
518
+ {
519
+ "3d_bbox": 0.8168098330497742,
520
+ "adv": 0.0003567840885807527,
521
+ "bbox": 0.8971930183470249,
522
+ "completions/clipped_ratio": 0.0,
523
+ "completions/max_length": 1601.5,
524
+ "completions/max_terminated_length": 1601.5,
525
+ "completions/mean_length": 1581.01171875,
526
+ "completions/mean_terminated_length": 1581.01171875,
527
+ "completions/min_length": 1562.625,
528
+ "completions/min_terminated_length": 1562.625,
529
+ "entropy": 0.27716846112161875,
530
+ "epoch": 0.06492589978828511,
531
+ "geo": 0.8458116948604584,
532
+ "grad_norm": 0.376953125,
533
+ "learning_rate": 9.13065326633166e-06,
534
+ "loss": -0.0,
535
+ "num_tokens": 21746232.0,
536
+ "point_2d": 0.736328125,
537
+ "point_3d": 0.859375,
538
+ "reward": 0.8274264857172966,
539
+ "step": 184
540
+ },
541
+ {
542
+ "3d_bbox": 0.7658436857163906,
543
+ "adv": -0.00043575108065851964,
544
+ "bbox": 0.994478665292263,
545
+ "completions/clipped_ratio": 0.0,
546
+ "completions/max_length": 1599.5,
547
+ "completions/max_terminated_length": 1599.5,
548
+ "completions/mean_length": 1577.2265625,
549
+ "completions/mean_terminated_length": 1577.2265625,
550
+ "completions/min_length": 1558.25,
551
+ "completions/min_terminated_length": 1558.25,
552
+ "entropy": 0.28411462996155024,
553
+ "epoch": 0.06774876499647142,
554
+ "geo": 0.8858354687690735,
555
+ "grad_norm": 0.404296875,
556
+ "learning_rate": 9.090452261306534e-06,
557
+ "loss": -0.0,
558
+ "num_tokens": 22691162.0,
559
+ "point_2d": 1.0,
560
+ "point_3d": 0.83203125,
561
+ "reward": 0.8980883881449699,
562
+ "step": 192
563
+ },
564
+ {
565
+ "3d_bbox": 0.689879797399044,
566
+ "adv": 0.0008139211470279406,
567
+ "bbox": 0.8756355233490467,
568
+ "completions/clipped_ratio": 0.0,
569
+ "completions/max_length": 1595.25,
570
+ "completions/max_terminated_length": 1595.25,
571
+ "completions/mean_length": 1579.328125,
572
+ "completions/mean_terminated_length": 1579.328125,
573
+ "completions/min_length": 1562.125,
574
+ "completions/min_terminated_length": 1562.125,
575
+ "entropy": 0.27792059257626534,
576
+ "epoch": 0.07057163020465773,
577
+ "geo": 0.8776071295142174,
578
+ "grad_norm": 0.28125,
579
+ "learning_rate": 9.050251256281409e-06,
580
+ "loss": -0.0006,
581
+ "num_tokens": 23637118.0,
582
+ "point_2d": 0.875,
583
+ "point_3d": 0.765625,
584
+ "reward": 0.8015350699424744,
585
+ "step": 200
586
+ },
587
+ {
588
+ "3d_bbox": 0.7628588750958443,
589
+ "adv": 0.0005178588535272866,
590
+ "bbox": 0.9862369671463966,
591
+ "completions/clipped_ratio": 0.0,
592
+ "completions/max_length": 1603.5,
593
+ "completions/max_terminated_length": 1603.5,
594
+ "completions/mean_length": 1582.5,
595
+ "completions/mean_terminated_length": 1582.5,
596
+ "completions/min_length": 1561.75,
597
+ "completions/min_terminated_length": 1561.75,
598
+ "entropy": 0.2777437614277005,
599
+ "epoch": 0.07339449541284404,
600
+ "geo": 0.7896297052502632,
601
+ "grad_norm": 0.318359375,
602
+ "learning_rate": 9.010050251256283e-06,
603
+ "loss": -0.0003,
604
+ "num_tokens": 24583790.0,
605
+ "point_2d": 0.75,
606
+ "point_3d": 0.822265625,
607
+ "reward": 0.8303403630852699,
608
+ "step": 208
609
+ },
610
+ {
611
+ "3d_bbox": 0.6901299897581339,
612
+ "adv": 1.327232166659087e-05,
613
+ "bbox": 0.9742533341050148,
614
+ "completions/clipped_ratio": 0.0,
615
+ "completions/max_length": 1595.375,
616
+ "completions/max_terminated_length": 1595.375,
617
+ "completions/mean_length": 1580.3828125,
618
+ "completions/mean_terminated_length": 1580.3828125,
619
+ "completions/min_length": 1566.125,
620
+ "completions/min_terminated_length": 1566.125,
621
+ "entropy": 0.2789337821304798,
622
+ "epoch": 0.07621736062103035,
623
+ "geo": 0.9093400537967682,
624
+ "grad_norm": 0.33984375,
625
+ "learning_rate": 8.969849246231157e-06,
626
+ "loss": 0.0001,
627
+ "num_tokens": 25529904.0,
628
+ "point_2d": 1.0,
629
+ "point_3d": 0.724609375,
630
+ "reward": 0.8472481593489647,
631
+ "step": 216
632
+ },
633
+ {
634
+ "3d_bbox": 0.8317120671272278,
635
+ "adv": 0.0007813970478309784,
636
+ "bbox": 0.99237409979105,
637
+ "completions/clipped_ratio": 0.0,
638
+ "completions/max_length": 1592.375,
639
+ "completions/max_terminated_length": 1592.375,
640
+ "completions/mean_length": 1576.484375,
641
+ "completions/mean_terminated_length": 1576.484375,
642
+ "completions/min_length": 1559.125,
643
+ "completions/min_terminated_length": 1559.125,
644
+ "entropy": 0.281503742095083,
645
+ "epoch": 0.07904022582921666,
646
+ "geo": 0.9148535653948784,
647
+ "grad_norm": 0.33203125,
648
+ "learning_rate": 8.92964824120603e-06,
649
+ "loss": -0.0005,
650
+ "num_tokens": 26474884.0,
651
+ "point_2d": 1.0,
652
+ "point_3d": 0.859375,
653
+ "reward": 0.9208653047680855,
654
+ "step": 224
655
+ },
656
+ {
657
+ "3d_bbox": 0.8539343103766441,
658
+ "adv": 0.0006083497164581786,
659
+ "bbox": 0.9947174340486526,
660
+ "completions/clipped_ratio": 0.0,
661
+ "completions/max_length": 1607.25,
662
+ "completions/max_terminated_length": 1607.25,
663
+ "completions/mean_length": 1582.21875,
664
+ "completions/mean_terminated_length": 1582.21875,
665
+ "completions/min_length": 1562.75,
666
+ "completions/min_terminated_length": 1562.75,
667
+ "entropy": 0.28075447864830494,
668
+ "epoch": 0.08186309103740297,
669
+ "geo": 0.9205377101898193,
670
+ "grad_norm": 0.3828125,
671
+ "learning_rate": 8.889447236180904e-06,
672
+ "loss": 0.0151,
673
+ "num_tokens": 27421356.0,
674
+ "point_2d": 0.9423828125,
675
+ "point_3d": 0.919921875,
676
+ "reward": 0.9277391210198402,
677
+ "step": 232
678
+ },
679
+ {
680
+ "3d_bbox": 0.8143627047538757,
681
+ "adv": 0.0002584877729532309,
682
+ "bbox": 0.9961168393492699,
683
+ "completions/clipped_ratio": 0.0,
684
+ "completions/max_length": 1604.375,
685
+ "completions/max_terminated_length": 1604.375,
686
+ "completions/mean_length": 1581.84375,
687
+ "completions/mean_terminated_length": 1581.84375,
688
+ "completions/min_length": 1562.0,
689
+ "completions/min_terminated_length": 1562.0,
690
+ "entropy": 0.2808487247675657,
691
+ "epoch": 0.08468595624558928,
692
+ "geo": 0.9482192173600197,
693
+ "grad_norm": 0.3125,
694
+ "learning_rate": 8.84924623115578e-06,
695
+ "loss": 0.015,
696
+ "num_tokens": 28367588.0,
697
+ "point_2d": 1.0,
698
+ "point_3d": 0.8505859375,
699
+ "reward": 0.9152663722634315,
700
+ "step": 240
701
+ },
702
+ {
703
+ "3d_bbox": 0.7562066540122032,
704
+ "adv": 0.0011895421193912625,
705
+ "bbox": 0.9870459213852882,
706
+ "completions/clipped_ratio": 0.0,
707
+ "completions/max_length": 1605.625,
708
+ "completions/max_terminated_length": 1605.625,
709
+ "completions/mean_length": 1582.22265625,
710
+ "completions/mean_terminated_length": 1582.22265625,
711
+ "completions/min_length": 1560.125,
712
+ "completions/min_terminated_length": 1560.125,
713
+ "entropy": 0.2828774293884635,
714
+ "epoch": 0.08750882145377559,
715
+ "geo": 0.8906307443976402,
716
+ "grad_norm": 0.310546875,
717
+ "learning_rate": 8.809045226130654e-06,
718
+ "loss": -0.0001,
719
+ "num_tokens": 29314053.0,
720
+ "point_2d": 1.0,
721
+ "point_3d": 0.84765625,
722
+ "reward": 0.8977271988987923,
723
+ "step": 248
724
+ },
725
+ {
726
+ "3d_bbox": 0.662892609834671,
727
+ "adv": -0.0004179688154977157,
728
+ "bbox": 0.9923923686146736,
729
+ "completions/clipped_ratio": 0.0,
730
+ "completions/max_length": 1598.875,
731
+ "completions/max_terminated_length": 1598.875,
732
+ "completions/mean_length": 1579.796875,
733
+ "completions/mean_terminated_length": 1579.796875,
734
+ "completions/min_length": 1563.125,
735
+ "completions/min_terminated_length": 1563.125,
736
+ "entropy": 0.27959624165669084,
737
+ "epoch": 0.0903316866619619,
738
+ "geo": 0.9055434167385101,
739
+ "grad_norm": 0.259765625,
740
+ "learning_rate": 8.768844221105528e-06,
741
+ "loss": -0.0002,
742
+ "num_tokens": 30259921.0,
743
+ "point_2d": 0.7509765625,
744
+ "point_3d": 0.8154296875,
745
+ "reward": 0.805422805249691,
746
+ "step": 256
747
+ },
748
+ {
749
+ "3d_bbox": 0.7727803811430931,
750
+ "adv": -0.0010380800276834634,
751
+ "bbox": 0.9807735681533813,
752
+ "completions/clipped_ratio": 0.0,
753
+ "completions/max_length": 1604.375,
754
+ "completions/max_terminated_length": 1604.375,
755
+ "completions/mean_length": 1581.1796875,
756
+ "completions/mean_terminated_length": 1581.1796875,
757
+ "completions/min_length": 1564.375,
758
+ "completions/min_terminated_length": 1564.375,
759
+ "entropy": 0.28243178501725197,
760
+ "epoch": 0.0931545518701482,
761
+ "geo": 0.783092018100433,
762
+ "grad_norm": 0.3203125,
763
+ "learning_rate": 8.728643216080403e-06,
764
+ "loss": 0.0003,
765
+ "num_tokens": 31205887.0,
766
+ "point_2d": 0.75,
767
+ "point_3d": 0.8544921875,
768
+ "reward": 0.8395115286111832,
769
+ "step": 264
770
+ },
771
+ {
772
+ "3d_bbox": 0.7379852011799812,
773
+ "adv": -0.0005326499958755448,
774
+ "bbox": 0.9934581220149994,
775
+ "completions/clipped_ratio": 0.0,
776
+ "completions/max_length": 1598.875,
777
+ "completions/max_terminated_length": 1598.875,
778
+ "completions/mean_length": 1579.44921875,
779
+ "completions/mean_terminated_length": 1579.44921875,
780
+ "completions/min_length": 1560.375,
781
+ "completions/min_terminated_length": 1560.375,
782
+ "entropy": 0.2767978576011956,
783
+ "epoch": 0.09597741707833451,
784
+ "geo": 0.9199443385004997,
785
+ "grad_norm": 0.40625,
786
+ "learning_rate": 8.688442211055277e-06,
787
+ "loss": 0.0002,
788
+ "num_tokens": 32151650.0,
789
+ "point_2d": 0.9990234375,
790
+ "point_3d": 0.8896484375,
791
+ "reward": 0.9050287976861,
792
+ "step": 272
793
+ },
794
+ {
795
+ "3d_bbox": 0.6362542062997818,
796
+ "adv": 0.0003683161558001302,
797
+ "bbox": 0.9273895137012005,
798
+ "completions/clipped_ratio": 0.0,
799
+ "completions/max_length": 1603.375,
800
+ "completions/max_terminated_length": 1603.375,
801
+ "completions/mean_length": 1583.1875,
802
+ "completions/mean_terminated_length": 1583.1875,
803
+ "completions/min_length": 1566.25,
804
+ "completions/min_terminated_length": 1566.25,
805
+ "entropy": 0.28121166303753853,
806
+ "epoch": 0.09880028228652082,
807
+ "geo": 0.6698016822338104,
808
+ "grad_norm": 0.419921875,
809
+ "learning_rate": 8.648241206030151e-06,
810
+ "loss": 0.0104,
811
+ "num_tokens": 33098514.0,
812
+ "point_2d": 0.748046875,
813
+ "point_3d": 0.8564453125,
814
+ "reward": 0.7920339703559875,
815
+ "step": 280
816
+ },
817
+ {
818
+ "3d_bbox": 0.8045329824090004,
819
+ "adv": 0.0005392557122831931,
820
+ "bbox": 0.9957676902413368,
821
+ "completions/clipped_ratio": 0.0,
822
+ "completions/max_length": 1598.75,
823
+ "completions/max_terminated_length": 1598.75,
824
+ "completions/mean_length": 1582.8515625,
825
+ "completions/mean_terminated_length": 1582.8515625,
826
+ "completions/min_length": 1564.75,
827
+ "completions/min_terminated_length": 1564.75,
828
+ "entropy": 0.2840573275461793,
829
+ "epoch": 0.10162314749470713,
830
+ "geo": 0.894830085337162,
831
+ "grad_norm": 0.263671875,
832
+ "learning_rate": 8.608040201005025e-06,
833
+ "loss": -0.0003,
834
+ "num_tokens": 34045156.0,
835
+ "point_2d": 1.0,
836
+ "point_3d": 0.8974609375,
837
+ "reward": 0.9244403913617134,
838
+ "step": 288
839
+ },
840
+ {
841
+ "3d_bbox": 0.6487490832805634,
842
+ "adv": -0.036328162997961044,
843
+ "bbox": 0.9537674784660339,
844
+ "completions/clipped_ratio": 0.0,
845
+ "completions/max_length": 1606.125,
846
+ "completions/max_terminated_length": 1606.125,
847
+ "completions/mean_length": 1583.43359375,
848
+ "completions/mean_terminated_length": 1583.43359375,
849
+ "completions/min_length": 1564.25,
850
+ "completions/min_terminated_length": 1564.25,
851
+ "entropy": 0.2789871282875538,
852
+ "epoch": 0.10444601270289344,
853
+ "geo": 0.8219972774386406,
854
+ "grad_norm": 0.3671875,
855
+ "learning_rate": 8.5678391959799e-06,
856
+ "loss": 0.0091,
857
+ "num_tokens": 34991979.0,
858
+ "point_2d": 0.875,
859
+ "point_3d": 0.8125,
860
+ "reward": 0.8225041329860687,
861
+ "step": 296
862
+ },
863
+ {
864
+ "3d_bbox": 0.7852863818407059,
865
+ "adv": -2.922765997936949e-05,
866
+ "bbox": 0.9355472847819328,
867
+ "completions/clipped_ratio": 0.0,
868
+ "completions/max_length": 1601.5,
869
+ "completions/max_terminated_length": 1601.5,
870
+ "completions/mean_length": 1579.7890625,
871
+ "completions/mean_terminated_length": 1579.7890625,
872
+ "completions/min_length": 1561.5,
873
+ "completions/min_terminated_length": 1561.5,
874
+ "entropy": 0.2808959139510989,
875
+ "epoch": 0.10726887791107975,
876
+ "geo": 0.7100357487797737,
877
+ "grad_norm": 0.333984375,
878
+ "learning_rate": 8.527638190954774e-06,
879
+ "loss": 0.0003,
880
+ "num_tokens": 35937853.0,
881
+ "point_2d": 0.6728515625,
882
+ "point_3d": 0.798828125,
883
+ "reward": 0.7981283441185951,
884
+ "step": 304
885
+ },
886
+ {
887
+ "3d_bbox": 0.6897341758012772,
888
+ "adv": -0.0010775219143397408,
889
+ "bbox": 0.9957167953252792,
890
+ "completions/clipped_ratio": 0.0,
891
+ "completions/max_length": 1598.0,
892
+ "completions/max_terminated_length": 1598.0,
893
+ "completions/mean_length": 1578.40234375,
894
+ "completions/mean_terminated_length": 1578.40234375,
895
+ "completions/min_length": 1560.625,
896
+ "completions/min_terminated_length": 1560.625,
897
+ "entropy": 0.28013237891718745,
898
+ "epoch": 0.11009174311926606,
899
+ "geo": 0.8601852916181087,
900
+ "grad_norm": 0.322265625,
901
+ "learning_rate": 8.487437185929648e-06,
902
+ "loss": 0.0003,
903
+ "num_tokens": 36883412.0,
904
+ "point_2d": 1.0,
905
+ "point_3d": 0.8154296875,
906
+ "reward": 0.8752201721072197,
907
+ "step": 312
908
+ },
909
+ {
910
+ "3d_bbox": 0.7732156217098236,
911
+ "adv": 0.00015621728016412817,
912
+ "bbox": 0.9900884553790092,
913
+ "completions/clipped_ratio": 0.0,
914
+ "completions/max_length": 1595.25,
915
+ "completions/max_terminated_length": 1595.25,
916
+ "completions/mean_length": 1574.06640625,
917
+ "completions/mean_terminated_length": 1574.06640625,
918
+ "completions/min_length": 1556.5,
919
+ "completions/min_terminated_length": 1556.5,
920
+ "entropy": 0.2842056662775576,
921
+ "epoch": 0.11291460832745237,
922
+ "geo": 0.8583031129091978,
923
+ "grad_norm": 0.359375,
924
+ "learning_rate": 8.447236180904524e-06,
925
+ "loss": -0.0003,
926
+ "num_tokens": 37827701.0,
927
+ "point_2d": 0.8583984375,
928
+ "point_3d": 0.8681640625,
929
+ "reward": 0.8724666386842728,
930
+ "step": 320
931
+ },
932
+ {
933
+ "3d_bbox": 0.7427574843168259,
934
+ "adv": -0.00022968700068304315,
935
+ "bbox": 0.9458357989788055,
936
+ "completions/clipped_ratio": 0.0,
937
+ "completions/max_length": 1599.25,
938
+ "completions/max_terminated_length": 1599.25,
939
+ "completions/mean_length": 1576.78125,
940
+ "completions/mean_terminated_length": 1576.78125,
941
+ "completions/min_length": 1555.625,
942
+ "completions/min_terminated_length": 1555.625,
943
+ "entropy": 0.28632913110777736,
944
+ "epoch": 0.11573747353563868,
945
+ "geo": 0.9527824521064758,
946
+ "grad_norm": 0.2578125,
947
+ "learning_rate": 8.407035175879398e-06,
948
+ "loss": -0.0,
949
+ "num_tokens": 38772853.0,
950
+ "point_2d": 1.0,
951
+ "point_3d": 0.8623046875,
952
+ "reward": 0.8877245113253593,
953
+ "step": 328
954
+ },
955
+ {
956
+ "3d_bbox": 0.7519282810389996,
957
+ "adv": 0.0003111484527380526,
958
+ "bbox": 0.9336327686905861,
959
+ "completions/clipped_ratio": 0.0,
960
+ "completions/max_length": 1598.0,
961
+ "completions/max_terminated_length": 1598.0,
962
+ "completions/mean_length": 1578.54296875,
963
+ "completions/mean_terminated_length": 1578.54296875,
964
+ "completions/min_length": 1560.0,
965
+ "completions/min_terminated_length": 1560.0,
966
+ "entropy": 0.2819040766917169,
967
+ "epoch": 0.11856033874382499,
968
+ "geo": 0.9259950742125511,
969
+ "grad_norm": 0.306640625,
970
+ "learning_rate": 8.366834170854273e-06,
971
+ "loss": 0.0202,
972
+ "num_tokens": 39718400.0,
973
+ "point_2d": 0.7451171875,
974
+ "point_3d": 0.748046875,
975
+ "reward": 0.7946812808513641,
976
+ "step": 336
977
+ },
978
+ {
979
+ "3d_bbox": 0.6988782063126564,
980
+ "adv": 8.153711894226534e-05,
981
+ "bbox": 0.9918540865182877,
982
+ "completions/clipped_ratio": 0.0,
983
+ "completions/max_length": 1602.0,
984
+ "completions/max_terminated_length": 1602.0,
985
+ "completions/mean_length": 1578.3984375,
986
+ "completions/mean_terminated_length": 1578.3984375,
987
+ "completions/min_length": 1557.0,
988
+ "completions/min_terminated_length": 1557.0,
989
+ "entropy": 0.2852953290566802,
990
+ "epoch": 0.1213832039520113,
991
+ "geo": 0.824953593313694,
992
+ "grad_norm": 0.33984375,
993
+ "learning_rate": 8.326633165829147e-06,
994
+ "loss": 0.0003,
995
+ "num_tokens": 40663774.0,
996
+ "point_2d": 0.875,
997
+ "point_3d": 0.771484375,
998
+ "reward": 0.8343041799962521,
999
+ "step": 344
1000
+ },
1001
+ {
1002
+ "3d_bbox": 0.7085578441619873,
1003
+ "adv": 0.000511349891894497,
1004
+ "bbox": 0.9954689294099808,
1005
+ "completions/clipped_ratio": 0.0,
1006
+ "completions/max_length": 1601.625,
1007
+ "completions/max_terminated_length": 1601.625,
1008
+ "completions/mean_length": 1578.21484375,
1009
+ "completions/mean_terminated_length": 1578.21484375,
1010
+ "completions/min_length": 1557.5,
1011
+ "completions/min_terminated_length": 1557.5,
1012
+ "entropy": 0.2835490182042122,
1013
+ "epoch": 0.12420606916019761,
1014
+ "geo": 0.8287867829203606,
1015
+ "grad_norm": 0.31640625,
1016
+ "learning_rate": 8.286432160804021e-06,
1017
+ "loss": 0.0,
1018
+ "num_tokens": 41609261.0,
1019
+ "point_2d": 0.875,
1020
+ "point_3d": 0.8564453125,
1021
+ "reward": 0.8588679879903793,
1022
+ "step": 352
1023
+ },
1024
+ {
1025
+ "3d_bbox": 0.7324622049927711,
1026
+ "adv": -0.06052318853471661,
1027
+ "bbox": 0.9785870239138603,
1028
+ "completions/clipped_ratio": 0.0,
1029
+ "completions/max_length": 1599.75,
1030
+ "completions/max_terminated_length": 1599.75,
1031
+ "completions/mean_length": 1580.8671875,
1032
+ "completions/mean_terminated_length": 1580.8671875,
1033
+ "completions/min_length": 1562.875,
1034
+ "completions/min_terminated_length": 1562.875,
1035
+ "entropy": 0.2788377907127142,
1036
+ "epoch": 0.12702893436838392,
1037
+ "geo": 0.8886202499270439,
1038
+ "grad_norm": 0.765625,
1039
+ "learning_rate": 8.246231155778895e-06,
1040
+ "loss": 0.0295,
1041
+ "num_tokens": 42555355.0,
1042
+ "point_2d": 0.984375,
1043
+ "point_3d": 0.80078125,
1044
+ "reward": 0.8740513622760773,
1045
+ "step": 360
1046
+ },
1047
+ {
1048
+ "3d_bbox": 0.6365351751446724,
1049
+ "adv": 0.0003275554918218404,
1050
+ "bbox": 0.8720735665410757,
1051
+ "completions/clipped_ratio": 0.0,
1052
+ "completions/max_length": 1601.875,
1053
+ "completions/max_terminated_length": 1601.875,
1054
+ "completions/mean_length": 1577.73046875,
1055
+ "completions/mean_terminated_length": 1577.73046875,
1056
+ "completions/min_length": 1555.75,
1057
+ "completions/min_terminated_length": 1555.75,
1058
+ "entropy": 0.28446380188688636,
1059
+ "epoch": 0.12985179957657023,
1060
+ "geo": 0.8144352469826117,
1061
+ "grad_norm": 0.34375,
1062
+ "learning_rate": 8.20603015075377e-06,
1063
+ "loss": -0.0001,
1064
+ "num_tokens": 43500534.0,
1065
+ "point_2d": 0.875,
1066
+ "point_3d": 0.7666015625,
1067
+ "reward": 0.787552572786808,
1068
+ "step": 368
1069
+ },
1070
+ {
1071
+ "3d_bbox": 0.7198036424815655,
1072
+ "adv": -9.503715409664437e-05,
1073
+ "bbox": 0.9939679279923439,
1074
+ "completions/clipped_ratio": 0.0,
1075
+ "completions/max_length": 1592.75,
1076
+ "completions/max_terminated_length": 1592.75,
1077
+ "completions/mean_length": 1573.19140625,
1078
+ "completions/mean_terminated_length": 1573.19140625,
1079
+ "completions/min_length": 1557.0,
1080
+ "completions/min_terminated_length": 1557.0,
1081
+ "entropy": 0.28550330083817244,
1082
+ "epoch": 0.13267466478475654,
1083
+ "geo": 0.9650512784719467,
1084
+ "grad_norm": 0.359375,
1085
+ "learning_rate": 8.165829145728644e-06,
1086
+ "loss": 0.0004,
1087
+ "num_tokens": 44444583.0,
1088
+ "point_2d": 0.9892578125,
1089
+ "point_3d": 0.83984375,
1090
+ "reward": 0.8857182934880257,
1091
+ "step": 376
1092
+ },
1093
+ {
1094
+ "3d_bbox": 0.7569727748632431,
1095
+ "adv": 0.0004161505894444417,
1096
+ "bbox": 0.9042535312473774,
1097
+ "completions/clipped_ratio": 0.0,
1098
+ "completions/max_length": 1599.5,
1099
+ "completions/max_terminated_length": 1599.5,
1100
+ "completions/mean_length": 1577.55859375,
1101
+ "completions/mean_terminated_length": 1577.55859375,
1102
+ "completions/min_length": 1561.375,
1103
+ "completions/min_terminated_length": 1561.375,
1104
+ "entropy": 0.2907078079879284,
1105
+ "epoch": 0.13549752999294284,
1106
+ "geo": 0.6764855124056339,
1107
+ "grad_norm": 0.345703125,
1108
+ "learning_rate": 8.125628140703518e-06,
1109
+ "loss": 0.0001,
1110
+ "num_tokens": 45389846.0,
1111
+ "point_2d": 0.69140625,
1112
+ "point_3d": 0.8193359375,
1113
+ "reward": 0.7929921373724937,
1114
+ "step": 384
1115
+ },
1116
+ {
1117
+ "3d_bbox": 0.6764840334653854,
1118
+ "adv": -1.4531655324390158e-05,
1119
+ "bbox": 0.8646975085139275,
1120
+ "completions/clipped_ratio": 0.0,
1121
+ "completions/max_length": 1600.5,
1122
+ "completions/max_terminated_length": 1600.5,
1123
+ "completions/mean_length": 1579.08984375,
1124
+ "completions/mean_terminated_length": 1579.08984375,
1125
+ "completions/min_length": 1558.375,
1126
+ "completions/min_terminated_length": 1558.375,
1127
+ "entropy": 0.2863780874758959,
1128
+ "epoch": 0.13832039520112915,
1129
+ "geo": 0.5404472178779542,
1130
+ "grad_norm": 0.3125,
1131
+ "learning_rate": 8.085427135678392e-06,
1132
+ "loss": 0.0,
1133
+ "num_tokens": 46335629.0,
1134
+ "point_2d": 0.5048828125,
1135
+ "point_3d": 0.806640625,
1136
+ "reward": 0.713176254183054,
1137
+ "step": 392
1138
+ },
1139
+ {
1140
+ "3d_bbox": 0.7255214750766754,
1141
+ "adv": -0.0006330321593850385,
1142
+ "bbox": 0.891293590888381,
1143
+ "completions/clipped_ratio": 0.0,
1144
+ "completions/max_length": 1602.375,
1145
+ "completions/max_terminated_length": 1602.375,
1146
+ "completions/mean_length": 1576.9921875,
1147
+ "completions/mean_terminated_length": 1576.9921875,
1148
+ "completions/min_length": 1558.125,
1149
+ "completions/min_terminated_length": 1558.125,
1150
+ "entropy": 0.28716066014021635,
1151
+ "epoch": 0.14114326040931546,
1152
+ "geo": 0.9741705134510994,
1153
+ "grad_norm": 0.314453125,
1154
+ "learning_rate": 8.045226130653267e-06,
1155
+ "loss": 0.0004,
1156
+ "num_tokens": 47280507.0,
1157
+ "point_2d": 0.8916015625,
1158
+ "point_3d": 0.859375,
1159
+ "reward": 0.8419479243457317,
1160
+ "step": 400
1161
+ },
1162
+ {
1163
+ "3d_bbox": 0.7029357459396124,
1164
+ "adv": 0.0008838025460136123,
1165
+ "bbox": 0.9950641989707947,
1166
+ "completions/clipped_ratio": 0.0,
1167
+ "completions/max_length": 1598.125,
1168
+ "completions/max_terminated_length": 1598.125,
1169
+ "completions/mean_length": 1575.92578125,
1170
+ "completions/mean_terminated_length": 1575.92578125,
1171
+ "completions/min_length": 1557.75,
1172
+ "completions/min_terminated_length": 1557.75,
1173
+ "entropy": 0.2837083409540355,
1174
+ "epoch": 0.14396612561750177,
1175
+ "geo": 0.864502564072609,
1176
+ "grad_norm": 0.345703125,
1177
+ "learning_rate": 8.005025125628141e-06,
1178
+ "loss": -0.0003,
1179
+ "num_tokens": 48225240.0,
1180
+ "point_2d": 1.0,
1181
+ "point_3d": 0.8310546875,
1182
+ "reward": 0.8822636753320694,
1183
+ "step": 408
1184
+ },
1185
+ {
1186
+ "3d_bbox": 0.6913789249956608,
1187
+ "adv": -0.0002196824352722615,
1188
+ "bbox": 0.9765095412731171,
1189
+ "completions/clipped_ratio": 0.0,
1190
+ "completions/max_length": 1598.125,
1191
+ "completions/max_terminated_length": 1598.125,
1192
+ "completions/mean_length": 1579.28515625,
1193
+ "completions/mean_terminated_length": 1579.28515625,
1194
+ "completions/min_length": 1559.25,
1195
+ "completions/min_terminated_length": 1559.25,
1196
+ "entropy": 0.28565866220742464,
1197
+ "epoch": 0.14678899082568808,
1198
+ "geo": 0.8599895760416985,
1199
+ "grad_norm": 0.30859375,
1200
+ "learning_rate": 7.964824120603015e-06,
1201
+ "loss": 0.0001,
1202
+ "num_tokens": 49171057.0,
1203
+ "point_2d": 1.0,
1204
+ "point_3d": 0.904296875,
1205
+ "reward": 0.8930463194847107,
1206
+ "step": 416
1207
+ },
1208
+ {
1209
+ "3d_bbox": 0.6788053698837757,
1210
+ "adv": -0.00020617676273104735,
1211
+ "bbox": 0.9748031497001648,
1212
+ "completions/clipped_ratio": 0.0,
1213
+ "completions/max_length": 1595.25,
1214
+ "completions/max_terminated_length": 1595.25,
1215
+ "completions/mean_length": 1577.51953125,
1216
+ "completions/mean_terminated_length": 1577.51953125,
1217
+ "completions/min_length": 1559.625,
1218
+ "completions/min_terminated_length": 1559.625,
1219
+ "entropy": 0.28440852370113134,
1220
+ "epoch": 0.1496118560338744,
1221
+ "geo": 0.822394497692585,
1222
+ "grad_norm": 0.40234375,
1223
+ "learning_rate": 7.92462311557789e-06,
1224
+ "loss": 0.0006,
1225
+ "num_tokens": 50116302.0,
1226
+ "point_2d": 0.875,
1227
+ "point_3d": 0.8310546875,
1228
+ "reward": 0.8399158045649529,
1229
+ "step": 424
1230
+ },
1231
+ {
1232
+ "3d_bbox": 0.6794650703668594,
1233
+ "adv": 0.0011559298304746335,
1234
+ "bbox": 0.9097459949553013,
1235
+ "completions/clipped_ratio": 0.0,
1236
+ "completions/max_length": 1596.625,
1237
+ "completions/max_terminated_length": 1596.625,
1238
+ "completions/mean_length": 1573.80859375,
1239
+ "completions/mean_terminated_length": 1573.80859375,
1240
+ "completions/min_length": 1552.625,
1241
+ "completions/min_terminated_length": 1552.625,
1242
+ "entropy": 0.28376144589856267,
1243
+ "epoch": 0.1524347212420607,
1244
+ "geo": 0.7893084269016981,
1245
+ "grad_norm": 0.3203125,
1246
+ "learning_rate": 7.884422110552764e-06,
1247
+ "loss": 0.0,
1248
+ "num_tokens": 51060365.0,
1249
+ "point_2d": 1.0,
1250
+ "point_3d": 0.8681640625,
1251
+ "reward": 0.8643437847495079,
1252
+ "step": 432
1253
+ },
1254
+ {
1255
+ "3d_bbox": 0.7942164540290833,
1256
+ "adv": -3.183347871527076e-05,
1257
+ "bbox": 0.9926317036151886,
1258
+ "completions/clipped_ratio": 0.0,
1259
+ "completions/max_length": 1598.75,
1260
+ "completions/max_terminated_length": 1598.75,
1261
+ "completions/mean_length": 1577.796875,
1262
+ "completions/mean_terminated_length": 1577.796875,
1263
+ "completions/min_length": 1553.125,
1264
+ "completions/min_terminated_length": 1553.125,
1265
+ "entropy": 0.2821194725111127,
1266
+ "epoch": 0.155257586450247,
1267
+ "geo": 0.9503362327814102,
1268
+ "grad_norm": 0.3984375,
1269
+ "learning_rate": 7.844221105527638e-06,
1270
+ "loss": 0.0002,
1271
+ "num_tokens": 52005633.0,
1272
+ "point_2d": 1.0,
1273
+ "point_3d": 0.869140625,
1274
+ "reward": 0.9139972031116486,
1275
+ "step": 440
1276
+ },
1277
+ {
1278
+ "3d_bbox": 0.6752692721784115,
1279
+ "adv": -9.695853077573702e-05,
1280
+ "bbox": 0.982423759996891,
1281
+ "completions/clipped_ratio": 0.0,
1282
+ "completions/max_length": 1596.25,
1283
+ "completions/max_terminated_length": 1596.25,
1284
+ "completions/mean_length": 1574.5546875,
1285
+ "completions/mean_terminated_length": 1574.5546875,
1286
+ "completions/min_length": 1555.125,
1287
+ "completions/min_terminated_length": 1555.125,
1288
+ "entropy": 0.2852726257406175,
1289
+ "epoch": 0.15808045165843332,
1290
+ "geo": 0.5512092066928744,
1291
+ "grad_norm": 0.322265625,
1292
+ "learning_rate": 7.804020100502512e-06,
1293
+ "loss": -0.0004,
1294
+ "num_tokens": 52949951.0,
1295
+ "point_2d": 0.75,
1296
+ "point_3d": 0.7998046875,
1297
+ "reward": 0.801874428987503,
1298
+ "step": 448
1299
+ },
1300
+ {
1301
+ "3d_bbox": 0.7081510461866856,
1302
+ "adv": 0.0003901398231391795,
1303
+ "bbox": 0.9892760217189789,
1304
+ "completions/clipped_ratio": 0.0,
1305
+ "completions/max_length": 1596.875,
1306
+ "completions/max_terminated_length": 1596.875,
1307
+ "completions/mean_length": 1573.0390625,
1308
+ "completions/mean_terminated_length": 1573.0390625,
1309
+ "completions/min_length": 1551.875,
1310
+ "completions/min_terminated_length": 1551.875,
1311
+ "entropy": 0.2837893716059625,
1312
+ "epoch": 0.16090331686661963,
1313
+ "geo": 0.9603376165032387,
1314
+ "grad_norm": 0.3125,
1315
+ "learning_rate": 7.763819095477387e-06,
1316
+ "loss": -0.0007,
1317
+ "num_tokens": 53894057.0,
1318
+ "point_2d": 1.0,
1319
+ "point_3d": 0.794921875,
1320
+ "reward": 0.8730872422456741,
1321
+ "step": 456
1322
+ },
1323
+ {
1324
+ "3d_bbox": 0.7504811957478523,
1325
+ "adv": 0.0007141208916436881,
1326
+ "bbox": 0.9912561923265457,
1327
+ "completions/clipped_ratio": 0.0,
1328
+ "completions/max_length": 1594.75,
1329
+ "completions/max_terminated_length": 1594.75,
1330
+ "completions/mean_length": 1576.1640625,
1331
+ "completions/mean_terminated_length": 1576.1640625,
1332
+ "completions/min_length": 1555.75,
1333
+ "completions/min_terminated_length": 1555.75,
1334
+ "entropy": 0.28930201614275575,
1335
+ "epoch": 0.16372618207480594,
1336
+ "geo": 0.8148366361856461,
1337
+ "grad_norm": 0.275390625,
1338
+ "learning_rate": 7.723618090452262e-06,
1339
+ "loss": 0.0003,
1340
+ "num_tokens": 54839099.0,
1341
+ "point_2d": 0.875,
1342
+ "point_3d": 0.8896484375,
1343
+ "reward": 0.8765964508056641,
1344
+ "step": 464
1345
+ },
1346
+ {
1347
+ "3d_bbox": 0.819125022739172,
1348
+ "adv": 0.00010633431156747974,
1349
+ "bbox": 0.9906315729022026,
1350
+ "completions/clipped_ratio": 0.0,
1351
+ "completions/max_length": 1603.0,
1352
+ "completions/max_terminated_length": 1603.0,
1353
+ "completions/mean_length": 1578.3203125,
1354
+ "completions/mean_terminated_length": 1578.3203125,
1355
+ "completions/min_length": 1556.375,
1356
+ "completions/min_terminated_length": 1556.375,
1357
+ "entropy": 0.28396829310804605,
1358
+ "epoch": 0.16654904728299225,
1359
+ "geo": 0.8597357729449868,
1360
+ "grad_norm": 0.30078125,
1361
+ "learning_rate": 7.683417085427137e-06,
1362
+ "loss": 0.0004,
1363
+ "num_tokens": 55784557.0,
1364
+ "point_2d": 0.8916015625,
1365
+ "point_3d": 0.9345703125,
1366
+ "reward": 0.9089821055531502,
1367
+ "step": 472
1368
+ },
1369
+ {
1370
+ "3d_bbox": 0.7331585213541985,
1371
+ "adv": -0.00033648866519797593,
1372
+ "bbox": 0.8809062782675028,
1373
+ "completions/clipped_ratio": 0.0,
1374
+ "completions/max_length": 1594.625,
1375
+ "completions/max_terminated_length": 1594.625,
1376
+ "completions/mean_length": 1576.11328125,
1377
+ "completions/mean_terminated_length": 1576.11328125,
1378
+ "completions/min_length": 1559.25,
1379
+ "completions/min_terminated_length": 1559.25,
1380
+ "entropy": 0.28669640561565757,
1381
+ "epoch": 0.16937191249117856,
1382
+ "geo": 0.9506581574678421,
1383
+ "grad_norm": 0.2890625,
1384
+ "learning_rate": 7.643216080402011e-06,
1385
+ "loss": 0.0132,
1386
+ "num_tokens": 56729514.0,
1387
+ "point_2d": 0.8759765625,
1388
+ "point_3d": 0.880859375,
1389
+ "reward": 0.8427251875400543,
1390
+ "step": 480
1391
+ },
1392
+ {
1393
+ "3d_bbox": 0.7568283602595329,
1394
+ "adv": -0.0008519564944435842,
1395
+ "bbox": 0.9879242181777954,
1396
+ "completions/clipped_ratio": 0.0,
1397
+ "completions/max_length": 1600.875,
1398
+ "completions/max_terminated_length": 1600.875,
1399
+ "completions/mean_length": 1574.87890625,
1400
+ "completions/mean_terminated_length": 1574.87890625,
1401
+ "completions/min_length": 1554.25,
1402
+ "completions/min_terminated_length": 1554.25,
1403
+ "entropy": 0.28737954096868634,
1404
+ "epoch": 0.17219477769936486,
1405
+ "geo": 0.7571075335144997,
1406
+ "grad_norm": 0.400390625,
1407
+ "learning_rate": 7.603015075376885e-06,
1408
+ "loss": 0.0105,
1409
+ "num_tokens": 57674099.0,
1410
+ "point_2d": 0.7998046875,
1411
+ "point_3d": 0.9228515625,
1412
+ "reward": 0.8668522015213966,
1413
+ "step": 488
1414
+ },
1415
+ {
1416
+ "3d_bbox": 0.7658256888389587,
1417
+ "adv": 0.0011402575073589105,
1418
+ "bbox": 0.9932135492563248,
1419
+ "completions/clipped_ratio": 0.0,
1420
+ "completions/max_length": 1596.25,
1421
+ "completions/max_terminated_length": 1596.25,
1422
+ "completions/mean_length": 1574.265625,
1423
+ "completions/mean_terminated_length": 1574.265625,
1424
+ "completions/min_length": 1554.375,
1425
+ "completions/min_terminated_length": 1554.375,
1426
+ "entropy": 0.2865150640718639,
1427
+ "epoch": 0.17501764290755117,
1428
+ "geo": 0.8641151636838913,
1429
+ "grad_norm": 0.37890625,
1430
+ "learning_rate": 7.5628140703517595e-06,
1431
+ "loss": -0.0004,
1432
+ "num_tokens": 58618575.0,
1433
+ "point_2d": 0.91015625,
1434
+ "point_3d": 0.904296875,
1435
+ "reward": 0.8933730870485306,
1436
+ "step": 496
1437
+ },
1438
+ {
1439
+ "3d_bbox": 0.7145381905138493,
1440
+ "adv": 0.0003104716488451231,
1441
+ "bbox": 0.9926531538367271,
1442
+ "completions/clipped_ratio": 0.0,
1443
+ "completions/max_length": 1594.75,
1444
+ "completions/max_terminated_length": 1594.75,
1445
+ "completions/mean_length": 1573.12890625,
1446
+ "completions/mean_terminated_length": 1573.12890625,
1447
+ "completions/min_length": 1553.375,
1448
+ "completions/min_terminated_length": 1553.375,
1449
+ "entropy": 0.28603550279513,
1450
+ "epoch": 0.17784050811573748,
1451
+ "geo": 0.7219396531581879,
1452
+ "grad_norm": 0.310546875,
1453
+ "learning_rate": 7.522613065326634e-06,
1454
+ "loss": -0.0,
1455
+ "num_tokens": 59562552.0,
1456
+ "point_2d": 0.7509765625,
1457
+ "point_3d": 0.9365234375,
1458
+ "reward": 0.8486728444695473,
1459
+ "step": 504
1460
+ },
1461
+ {
1462
+ "3d_bbox": 0.7242622300982475,
1463
+ "adv": -0.0005710294717573561,
1464
+ "bbox": 0.9246156923472881,
1465
+ "completions/clipped_ratio": 0.0,
1466
+ "completions/max_length": 1591.75,
1467
+ "completions/max_terminated_length": 1591.75,
1468
+ "completions/mean_length": 1571.015625,
1469
+ "completions/mean_terminated_length": 1571.015625,
1470
+ "completions/min_length": 1551.625,
1471
+ "completions/min_terminated_length": 1551.625,
1472
+ "entropy": 0.2915982282720506,
1473
+ "epoch": 0.1806633733239238,
1474
+ "geo": 0.7993985265493393,
1475
+ "grad_norm": 0.408203125,
1476
+ "learning_rate": 7.482412060301508e-06,
1477
+ "loss": 0.0001,
1478
+ "num_tokens": 60505948.0,
1479
+ "point_2d": 0.7919921875,
1480
+ "point_3d": 0.8671875,
1481
+ "reward": 0.8270144015550613,
1482
+ "step": 512
1483
+ },
1484
+ {
1485
+ "3d_bbox": 0.7831042557954788,
1486
+ "adv": -0.000159245344093506,
1487
+ "bbox": 0.9960433393716812,
1488
+ "completions/clipped_ratio": 0.0,
1489
+ "completions/max_length": 1592.375,
1490
+ "completions/max_terminated_length": 1592.375,
1491
+ "completions/mean_length": 1574.84375,
1492
+ "completions/mean_terminated_length": 1574.84375,
1493
+ "completions/min_length": 1559.125,
1494
+ "completions/min_terminated_length": 1559.125,
1495
+ "entropy": 0.2845937805250287,
1496
+ "epoch": 0.1834862385321101,
1497
+ "geo": 0.9170719608664513,
1498
+ "grad_norm": 0.345703125,
1499
+ "learning_rate": 7.442211055276382e-06,
1500
+ "loss": -0.0,
1501
+ "num_tokens": 61450564.0,
1502
+ "point_2d": 1.0,
1503
+ "point_3d": 0.8408203125,
1504
+ "reward": 0.90499197691679,
1505
+ "step": 520
1506
+ },
1507
+ {
1508
+ "3d_bbox": 0.7821980714797974,
1509
+ "adv": -0.000531579293010509,
1510
+ "bbox": 0.9920122027397156,
1511
+ "completions/clipped_ratio": 0.0,
1512
+ "completions/max_length": 1595.375,
1513
+ "completions/max_terminated_length": 1595.375,
1514
+ "completions/mean_length": 1571.765625,
1515
+ "completions/mean_terminated_length": 1571.765625,
1516
+ "completions/min_length": 1551.5,
1517
+ "completions/min_terminated_length": 1551.5,
1518
+ "entropy": 0.2882155287079513,
1519
+ "epoch": 0.1863091037402964,
1520
+ "geo": 0.901853509247303,
1521
+ "grad_norm": 0.365234375,
1522
+ "learning_rate": 7.402010050251257e-06,
1523
+ "loss": 0.0002,
1524
+ "num_tokens": 62394336.0,
1525
+ "point_2d": 0.888671875,
1526
+ "point_3d": 0.9052734375,
1527
+ "reward": 0.8920388966798782,
1528
+ "step": 528
1529
+ },
1530
+ {
1531
+ "3d_bbox": 0.7537402808666229,
1532
+ "adv": -0.0004438982141437009,
1533
+ "bbox": 0.9946581870317459,
1534
+ "completions/clipped_ratio": 0.0,
1535
+ "completions/max_length": 1594.375,
1536
+ "completions/max_terminated_length": 1594.375,
1537
+ "completions/mean_length": 1574.3203125,
1538
+ "completions/mean_terminated_length": 1574.3203125,
1539
+ "completions/min_length": 1554.75,
1540
+ "completions/min_terminated_length": 1554.75,
1541
+ "entropy": 0.2886669235303998,
1542
+ "epoch": 0.18913196894848272,
1543
+ "geo": 0.8310712724924088,
1544
+ "grad_norm": 0.36328125,
1545
+ "learning_rate": 7.361809045226132e-06,
1546
+ "loss": 0.0002,
1547
+ "num_tokens": 63338834.0,
1548
+ "point_2d": 0.873046875,
1549
+ "point_3d": 0.904296875,
1550
+ "reward": 0.8814355507493019,
1551
+ "step": 536
1552
+ },
1553
+ {
1554
+ "3d_bbox": 0.6960996612906456,
1555
+ "adv": 0.0006180387658787367,
1556
+ "bbox": 0.869446316966787,
1557
+ "completions/clipped_ratio": 0.0,
1558
+ "completions/max_length": 1597.25,
1559
+ "completions/max_terminated_length": 1597.25,
1560
+ "completions/mean_length": 1571.88671875,
1561
+ "completions/mean_terminated_length": 1571.88671875,
1562
+ "completions/min_length": 1552.25,
1563
+ "completions/min_terminated_length": 1552.25,
1564
+ "entropy": 0.2887332271784544,
1565
+ "epoch": 0.19195483415666903,
1566
+ "geo": 0.826970137655735,
1567
+ "grad_norm": 0.267578125,
1568
+ "learning_rate": 7.321608040201006e-06,
1569
+ "loss": -0.0,
1570
+ "num_tokens": 64282685.0,
1571
+ "point_2d": 0.875,
1572
+ "point_3d": 0.8662109375,
1573
+ "reward": 0.8266892246901989,
1574
+ "step": 544
1575
+ },
1576
+ {
1577
+ "3d_bbox": 0.8654715269804001,
1578
+ "adv": -0.0008252508814621251,
1579
+ "bbox": 0.9932455867528915,
1580
+ "completions/clipped_ratio": 0.0,
1581
+ "completions/max_length": 1596.125,
1582
+ "completions/max_terminated_length": 1596.125,
1583
+ "completions/mean_length": 1572.1953125,
1584
+ "completions/mean_terminated_length": 1572.1953125,
1585
+ "completions/min_length": 1553.625,
1586
+ "completions/min_terminated_length": 1553.625,
1587
+ "entropy": 0.2873402568511665,
1588
+ "epoch": 0.19477769936485534,
1589
+ "geo": 0.872327484190464,
1590
+ "grad_norm": 0.2890625,
1591
+ "learning_rate": 7.28140703517588e-06,
1592
+ "loss": 0.0004,
1593
+ "num_tokens": 65226639.0,
1594
+ "point_2d": 0.8935546875,
1595
+ "point_3d": 0.9482421875,
1596
+ "reward": 0.9251285046339035,
1597
+ "step": 552
1598
+ },
1599
+ {
1600
+ "3d_bbox": 0.7851101011037827,
1601
+ "adv": -0.0011301148442726117,
1602
+ "bbox": 0.9379025921225548,
1603
+ "completions/clipped_ratio": 0.0,
1604
+ "completions/max_length": 1588.5,
1605
+ "completions/max_terminated_length": 1588.5,
1606
+ "completions/mean_length": 1571.4765625,
1607
+ "completions/mean_terminated_length": 1571.4765625,
1608
+ "completions/min_length": 1555.0,
1609
+ "completions/min_terminated_length": 1555.0,
1610
+ "entropy": 0.2852159277535975,
1611
+ "epoch": 0.19760056457304165,
1612
+ "geo": 0.690475556999445,
1613
+ "grad_norm": 0.3125,
1614
+ "learning_rate": 7.241206030150754e-06,
1615
+ "loss": 0.0006,
1616
+ "num_tokens": 66170305.0,
1617
+ "point_2d": 0.625,
1618
+ "point_3d": 0.837890625,
1619
+ "reward": 0.7964758351445198,
1620
+ "step": 560
1621
+ },
1622
+ {
1623
+ "3d_bbox": 0.7313521131873131,
1624
+ "adv": -0.000422738248744281,
1625
+ "bbox": 0.8720479012117721,
1626
+ "completions/clipped_ratio": 0.0,
1627
+ "completions/max_length": 1601.875,
1628
+ "completions/max_terminated_length": 1601.875,
1629
+ "completions/mean_length": 1574.91015625,
1630
+ "completions/mean_terminated_length": 1574.91015625,
1631
+ "completions/min_length": 1551.25,
1632
+ "completions/min_terminated_length": 1551.25,
1633
+ "entropy": 0.2880417066626251,
1634
+ "epoch": 0.20042342978122796,
1635
+ "geo": 0.8383861493784934,
1636
+ "grad_norm": 0.412109375,
1637
+ "learning_rate": 7.2010050251256295e-06,
1638
+ "loss": 0.0004,
1639
+ "num_tokens": 67114842.0,
1640
+ "point_2d": 0.8740234375,
1641
+ "point_3d": 0.84765625,
1642
+ "reward": 0.8312699124217033,
1643
+ "step": 568
1644
+ },
1645
+ {
1646
+ "3d_bbox": 0.6841592267155647,
1647
+ "adv": -0.0008141829575833981,
1648
+ "bbox": 0.9068506434559822,
1649
+ "completions/clipped_ratio": 0.0,
1650
+ "completions/max_length": 1597.5,
1651
+ "completions/max_terminated_length": 1597.5,
1652
+ "completions/mean_length": 1571.79296875,
1653
+ "completions/mean_terminated_length": 1571.79296875,
1654
+ "completions/min_length": 1546.75,
1655
+ "completions/min_terminated_length": 1546.75,
1656
+ "entropy": 0.2889366364106536,
1657
+ "epoch": 0.20324629498941427,
1658
+ "geo": 0.7308010272681713,
1659
+ "grad_norm": 0.2578125,
1660
+ "learning_rate": 7.160804020100504e-06,
1661
+ "loss": 0.0002,
1662
+ "num_tokens": 68058501.0,
1663
+ "point_2d": 0.8525390625,
1664
+ "point_3d": 0.794921875,
1665
+ "reward": 0.8096176832914352,
1666
+ "step": 576
1667
+ },
1668
+ {
1669
+ "3d_bbox": 0.5905975513160229,
1670
+ "adv": -0.00022596626513404772,
1671
+ "bbox": 0.9184868931770325,
1672
+ "completions/clipped_ratio": 0.0,
1673
+ "completions/max_length": 1597.5,
1674
+ "completions/max_terminated_length": 1597.5,
1675
+ "completions/mean_length": 1572.45703125,
1676
+ "completions/mean_terminated_length": 1572.45703125,
1677
+ "completions/min_length": 1551.125,
1678
+ "completions/min_terminated_length": 1551.125,
1679
+ "entropy": 0.2837987565435469,
1680
+ "epoch": 0.20606916019760058,
1681
+ "geo": 0.8267702832818031,
1682
+ "grad_norm": 0.314453125,
1683
+ "learning_rate": 7.120603015075378e-06,
1684
+ "loss": -0.0003,
1685
+ "num_tokens": 69002250.0,
1686
+ "point_2d": 0.8681640625,
1687
+ "point_3d": 0.7445197626948357,
1688
+ "reward": 0.7804420664906502,
1689
+ "step": 584
1690
+ },
1691
+ {
1692
+ "3d_bbox": 0.6897582598030567,
1693
+ "adv": -0.0004831653022847604,
1694
+ "bbox": 0.9742325842380524,
1695
+ "completions/clipped_ratio": 0.0,
1696
+ "completions/max_length": 1595.25,
1697
+ "completions/max_terminated_length": 1595.25,
1698
+ "completions/mean_length": 1572.9296875,
1699
+ "completions/mean_terminated_length": 1572.9296875,
1700
+ "completions/min_length": 1553.5,
1701
+ "completions/min_terminated_length": 1553.5,
1702
+ "entropy": 0.2835669592022896,
1703
+ "epoch": 0.20889202540578689,
1704
+ "geo": 0.8361655548214912,
1705
+ "grad_norm": 0.330078125,
1706
+ "learning_rate": 7.080402010050251e-06,
1707
+ "loss": 0.0004,
1708
+ "num_tokens": 69946400.0,
1709
+ "point_2d": 0.875,
1710
+ "point_3d": 0.8948184698820114,
1711
+ "reward": 0.8584523350000381,
1712
+ "step": 592
1713
+ },
1714
+ {
1715
+ "3d_bbox": 0.785271767526865,
1716
+ "adv": -0.00028875649149995297,
1717
+ "bbox": 0.9752914533019066,
1718
+ "completions/clipped_ratio": 0.0,
1719
+ "completions/max_length": 1599.0,
1720
+ "completions/max_terminated_length": 1599.0,
1721
+ "completions/mean_length": 1572.3671875,
1722
+ "completions/mean_terminated_length": 1572.3671875,
1723
+ "completions/min_length": 1549.875,
1724
+ "completions/min_terminated_length": 1549.875,
1725
+ "entropy": 0.29118923749774694,
1726
+ "epoch": 0.2117148906139732,
1727
+ "geo": 0.9338327869772911,
1728
+ "grad_norm": 0.3671875,
1729
+ "learning_rate": 7.040201005025126e-06,
1730
+ "loss": 0.0007,
1731
+ "num_tokens": 70890342.0,
1732
+ "point_2d": 0.998046875,
1733
+ "point_3d": 0.94140625,
1734
+ "reward": 0.9250040724873543,
1735
+ "step": 600
1736
+ },
1737
+ {
1738
+ "3d_bbox": 0.8766504302620888,
1739
+ "adv": -0.002171848598663928,
1740
+ "bbox": 0.9872715324163437,
1741
+ "completions/clipped_ratio": 0.0,
1742
+ "completions/max_length": 1601.125,
1743
+ "completions/max_terminated_length": 1601.125,
1744
+ "completions/mean_length": 1576.87890625,
1745
+ "completions/mean_terminated_length": 1576.87890625,
1746
+ "completions/min_length": 1556.375,
1747
+ "completions/min_terminated_length": 1556.375,
1748
+ "entropy": 0.28933983854949474,
1749
+ "epoch": 0.2145377558221595,
1750
+ "geo": 0.845995269715786,
1751
+ "grad_norm": 0.33203125,
1752
+ "learning_rate": 7e-06,
1753
+ "loss": 0.0003,
1754
+ "num_tokens": 71835359.0,
1755
+ "point_2d": 0.875,
1756
+ "point_3d": 0.9454848319292068,
1757
+ "reward": 0.9211016893386841,
1758
+ "step": 608
1759
+ },
1760
+ {
1761
+ "3d_bbox": 0.8059130385518074,
1762
+ "adv": 0.0005476666337926872,
1763
+ "bbox": 0.9907352328300476,
1764
+ "completions/clipped_ratio": 0.0,
1765
+ "completions/max_length": 1593.625,
1766
+ "completions/max_terminated_length": 1593.625,
1767
+ "completions/mean_length": 1570.9921875,
1768
+ "completions/mean_terminated_length": 1570.9921875,
1769
+ "completions/min_length": 1554.125,
1770
+ "completions/min_terminated_length": 1554.125,
1771
+ "entropy": 0.28401868836954236,
1772
+ "epoch": 0.2173606210303458,
1773
+ "geo": 0.8720198646187782,
1774
+ "grad_norm": 0.384765625,
1775
+ "learning_rate": 6.959798994974874e-06,
1776
+ "loss": -0.0004,
1777
+ "num_tokens": 72778933.0,
1778
+ "point_2d": 0.9248046875,
1779
+ "point_3d": 0.9135454967617989,
1780
+ "reward": 0.9087496176362038,
1781
+ "step": 616
1782
+ },
1783
+ {
1784
+ "3d_bbox": 0.8549554497003555,
1785
+ "adv": -0.0005420902471087174,
1786
+ "bbox": 0.9758080616593361,
1787
+ "completions/clipped_ratio": 0.0,
1788
+ "completions/max_length": 1593.0,
1789
+ "completions/max_terminated_length": 1593.0,
1790
+ "completions/mean_length": 1570.1640625,
1791
+ "completions/mean_terminated_length": 1570.1640625,
1792
+ "completions/min_length": 1549.0,
1793
+ "completions/min_terminated_length": 1549.0,
1794
+ "entropy": 0.29045473085716367,
1795
+ "epoch": 0.22018348623853212,
1796
+ "geo": 0.8916067183017731,
1797
+ "grad_norm": 0.3828125,
1798
+ "learning_rate": 6.919597989949749e-06,
1799
+ "loss": 0.0156,
1800
+ "num_tokens": 73722191.0,
1801
+ "point_2d": 1.0,
1802
+ "point_3d": 0.9639246314764023,
1803
+ "reward": 0.9486720338463783,
1804
+ "step": 624
1805
+ },
1806
+ {
1807
+ "3d_bbox": 0.7975610345602036,
1808
+ "adv": 0.0014872762099003012,
1809
+ "bbox": 0.9193930514156818,
1810
+ "completions/clipped_ratio": 0.0,
1811
+ "completions/max_length": 1600.625,
1812
+ "completions/max_terminated_length": 1600.625,
1813
+ "completions/mean_length": 1573.29296875,
1814
+ "completions/mean_terminated_length": 1573.29296875,
1815
+ "completions/min_length": 1553.0,
1816
+ "completions/min_terminated_length": 1553.0,
1817
+ "entropy": 0.28542913123965263,
1818
+ "epoch": 0.22300635144671843,
1819
+ "geo": 0.936805784702301,
1820
+ "grad_norm": 0.2734375,
1821
+ "learning_rate": 6.8793969849246235e-06,
1822
+ "loss": 0.0152,
1823
+ "num_tokens": 74666354.0,
1824
+ "point_2d": 0.9990234375,
1825
+ "point_3d": 0.93359375,
1826
+ "reward": 0.9123928248882294,
1827
+ "step": 632
1828
+ },
1829
+ {
1830
+ "3d_bbox": 0.6216618679463863,
1831
+ "adv": -0.0013801682944176719,
1832
+ "bbox": 0.9884917959570885,
1833
+ "completions/clipped_ratio": 0.0,
1834
+ "completions/max_length": 1601.375,
1835
+ "completions/max_terminated_length": 1601.375,
1836
+ "completions/mean_length": 1577.00390625,
1837
+ "completions/mean_terminated_length": 1577.00390625,
1838
+ "completions/min_length": 1554.0,
1839
+ "completions/min_terminated_length": 1554.0,
1840
+ "entropy": 0.28678226098418236,
1841
+ "epoch": 0.22582921665490474,
1842
+ "geo": 0.9096385985612869,
1843
+ "grad_norm": 0.42578125,
1844
+ "learning_rate": 6.839195979899498e-06,
1845
+ "loss": 0.0005,
1846
+ "num_tokens": 75611339.0,
1847
+ "point_2d": 0.8291015625,
1848
+ "point_3d": 0.7913028448820114,
1849
+ "reward": 0.8076395243406296,
1850
+ "step": 640
1851
+ },
1852
+ {
1853
+ "3d_bbox": 0.7582510337233543,
1854
+ "adv": -0.00019565537664334443,
1855
+ "bbox": 0.9911796301603317,
1856
+ "completions/clipped_ratio": 0.0,
1857
+ "completions/max_length": 1596.625,
1858
+ "completions/max_terminated_length": 1596.625,
1859
+ "completions/mean_length": 1575.71484375,
1860
+ "completions/mean_terminated_length": 1575.71484375,
1861
+ "completions/min_length": 1556.25,
1862
+ "completions/min_terminated_length": 1556.25,
1863
+ "entropy": 0.2863654876127839,
1864
+ "epoch": 0.22865208186309105,
1865
+ "geo": 0.9025658443570137,
1866
+ "grad_norm": 0.451171875,
1867
+ "learning_rate": 6.798994974874372e-06,
1868
+ "loss": -0.0001,
1869
+ "num_tokens": 76556042.0,
1870
+ "point_2d": 0.95703125,
1871
+ "point_3d": 0.8994140625,
1872
+ "reward": 0.9014689922332764,
1873
+ "step": 648
1874
+ },
1875
+ {
1876
+ "3d_bbox": 0.753610834479332,
1877
+ "adv": 0.00039293405916396296,
1878
+ "bbox": 0.9021090529859066,
1879
+ "completions/clipped_ratio": 0.0,
1880
+ "completions/max_length": 1592.625,
1881
+ "completions/max_terminated_length": 1592.625,
1882
+ "completions/mean_length": 1572.203125,
1883
+ "completions/mean_terminated_length": 1572.203125,
1884
+ "completions/min_length": 1552.25,
1885
+ "completions/min_terminated_length": 1552.25,
1886
+ "entropy": 0.286099414806813,
1887
+ "epoch": 0.23147494707127736,
1888
+ "geo": 0.9557241797447205,
1889
+ "grad_norm": 0.455078125,
1890
+ "learning_rate": 6.758793969849246e-06,
1891
+ "loss": 0.0002,
1892
+ "num_tokens": 77499886.0,
1893
+ "point_2d": 1.0,
1894
+ "point_3d": 0.91015625,
1895
+ "reward": 0.8914690390229225,
1896
+ "step": 656
1897
+ },
1898
+ {
1899
+ "3d_bbox": 0.831758588552475,
1900
+ "adv": -7.158980952226557e-05,
1901
+ "bbox": 0.9794979766011238,
1902
+ "completions/clipped_ratio": 0.0,
1903
+ "completions/max_length": 1595.5,
1904
+ "completions/max_terminated_length": 1595.5,
1905
+ "completions/mean_length": 1574.5,
1906
+ "completions/mean_terminated_length": 1574.5,
1907
+ "completions/min_length": 1558.75,
1908
+ "completions/min_terminated_length": 1558.75,
1909
+ "entropy": 0.2819104827940464,
1910
+ "epoch": 0.23429781227946367,
1911
+ "geo": 0.8940749987959862,
1912
+ "grad_norm": 0.294921875,
1913
+ "learning_rate": 6.718592964824121e-06,
1914
+ "loss": 0.0001,
1915
+ "num_tokens": 78444526.0,
1916
+ "point_2d": 1.0,
1917
+ "point_3d": 0.939453125,
1918
+ "reward": 0.9376774355769157,
1919
+ "step": 664
1920
+ },
1921
+ {
1922
+ "3d_bbox": 0.8210256397724152,
1923
+ "adv": 0.000540634344361024,
1924
+ "bbox": 0.9576755836606026,
1925
+ "completions/clipped_ratio": 0.0,
1926
+ "completions/max_length": 1594.125,
1927
+ "completions/max_terminated_length": 1594.125,
1928
+ "completions/mean_length": 1573.80859375,
1929
+ "completions/mean_terminated_length": 1573.80859375,
1930
+ "completions/min_length": 1553.25,
1931
+ "completions/min_terminated_length": 1553.25,
1932
+ "entropy": 0.2863646815530956,
1933
+ "epoch": 0.23712067748764998,
1934
+ "geo": 0.9599328562617302,
1935
+ "grad_norm": 0.28125,
1936
+ "learning_rate": 6.678391959798996e-06,
1937
+ "loss": -0.0001,
1938
+ "num_tokens": 79388949.0,
1939
+ "point_2d": 1.0,
1940
+ "point_3d": 0.92578125,
1941
+ "reward": 0.9261206015944481,
1942
+ "step": 672
1943
+ },
1944
+ {
1945
+ "3d_bbox": 0.8162731677293777,
1946
+ "adv": -0.0006006620242260396,
1947
+ "bbox": 0.9639560282230377,
1948
+ "completions/clipped_ratio": 0.0,
1949
+ "completions/max_length": 1593.75,
1950
+ "completions/max_terminated_length": 1593.75,
1951
+ "completions/mean_length": 1573.62890625,
1952
+ "completions/mean_terminated_length": 1573.62890625,
1953
+ "completions/min_length": 1557.125,
1954
+ "completions/min_terminated_length": 1557.125,
1955
+ "entropy": 0.2868205299600959,
1956
+ "epoch": 0.2399435426958363,
1957
+ "geo": 0.9127583503723145,
1958
+ "grad_norm": 0.44921875,
1959
+ "learning_rate": 6.63819095477387e-06,
1960
+ "loss": -0.0,
1961
+ "num_tokens": 80333190.0,
1962
+ "point_2d": 1.0,
1963
+ "point_3d": 0.9267578125,
1964
+ "reward": 0.926746778190136,
1965
+ "step": 680
1966
+ },
1967
+ {
1968
+ "3d_bbox": 0.7580128684639931,
1969
+ "adv": 0.0009847661577850886,
1970
+ "bbox": 0.9946200400590897,
1971
+ "completions/clipped_ratio": 0.0,
1972
+ "completions/max_length": 1598.0,
1973
+ "completions/max_terminated_length": 1598.0,
1974
+ "completions/mean_length": 1572.2109375,
1975
+ "completions/mean_terminated_length": 1572.2109375,
1976
+ "completions/min_length": 1552.0,
1977
+ "completions/min_terminated_length": 1552.0,
1978
+ "entropy": 0.286160247400403,
1979
+ "epoch": 0.2427664079040226,
1980
+ "geo": 0.9191197454929352,
1981
+ "grad_norm": 0.361328125,
1982
+ "learning_rate": 6.597989949748744e-06,
1983
+ "loss": 0.0001,
1984
+ "num_tokens": 81277036.0,
1985
+ "point_2d": 1.0,
1986
+ "point_3d": 0.8818359375,
1987
+ "reward": 0.9086171984672546,
1988
+ "step": 688
1989
+ },
1990
+ {
1991
+ "3d_bbox": 0.729152224957943,
1992
+ "adv": -0.0009101523137360346,
1993
+ "bbox": 0.9927223324775696,
1994
+ "completions/clipped_ratio": 0.0,
1995
+ "completions/max_length": 1592.875,
1996
+ "completions/max_terminated_length": 1592.875,
1997
+ "completions/mean_length": 1571.8515625,
1998
+ "completions/mean_terminated_length": 1571.8515625,
1999
+ "completions/min_length": 1553.375,
2000
+ "completions/min_terminated_length": 1553.375,
2001
+ "entropy": 0.28340971609577537,
2002
+ "epoch": 0.2455892731122089,
2003
+ "geo": 0.9836970791220665,
2004
+ "grad_norm": 0.34765625,
2005
+ "learning_rate": 6.557788944723618e-06,
2006
+ "loss": 0.0005,
2007
+ "num_tokens": 82220822.0,
2008
+ "point_2d": 1.0,
2009
+ "point_3d": 0.9014246314764023,
2010
+ "reward": 0.9058247655630112,
2011
+ "step": 696
2012
+ },
2013
+ {
2014
+ "3d_bbox": 0.7259362824261189,
2015
+ "adv": 0.0003162266129947966,
2016
+ "bbox": 0.922937985509634,
2017
+ "completions/clipped_ratio": 0.0,
2018
+ "completions/max_length": 1596.0,
2019
+ "completions/max_terminated_length": 1596.0,
2020
+ "completions/mean_length": 1575.2265625,
2021
+ "completions/mean_terminated_length": 1575.2265625,
2022
+ "completions/min_length": 1555.625,
2023
+ "completions/min_terminated_length": 1555.625,
2024
+ "entropy": 0.286905855871737,
2025
+ "epoch": 0.24841213832039521,
2026
+ "geo": 0.8376269713044167,
2027
+ "grad_norm": 0.30078125,
2028
+ "learning_rate": 6.5175879396984935e-06,
2029
+ "loss": 0.0,
2030
+ "num_tokens": 83165600.0,
2031
+ "point_2d": 0.9541015625,
2032
+ "point_3d": 0.8759765625,
2033
+ "reward": 0.8697381019592285,
2034
+ "step": 704
2035
+ },
2036
+ {
2037
+ "3d_bbox": 0.7864346951246262,
2038
+ "adv": -0.0007161469466154813,
2039
+ "bbox": 0.9958559349179268,
2040
+ "completions/clipped_ratio": 0.0,
2041
+ "completions/max_length": 1595.0,
2042
+ "completions/max_terminated_length": 1595.0,
2043
+ "completions/mean_length": 1572.4921875,
2044
+ "completions/mean_terminated_length": 1572.4921875,
2045
+ "completions/min_length": 1545.125,
2046
+ "completions/min_terminated_length": 1545.125,
2047
+ "entropy": 0.2886675829067826,
2048
+ "epoch": 0.2512350035285815,
2049
+ "geo": 0.8416303154081106,
2050
+ "grad_norm": 0.333984375,
2051
+ "learning_rate": 6.477386934673368e-06,
2052
+ "loss": 0.0008,
2053
+ "num_tokens": 84109606.0,
2054
+ "point_2d": 0.75,
2055
+ "point_3d": 0.947265625,
2056
+ "reward": 0.8698890581727028,
2057
+ "step": 712
2058
+ },
2059
+ {
2060
+ "3d_bbox": 0.657585222274065,
2061
+ "adv": -0.0004964938416378573,
2062
+ "bbox": 0.9940767288208008,
2063
+ "completions/clipped_ratio": 0.0,
2064
+ "completions/max_length": 1589.75,
2065
+ "completions/max_terminated_length": 1589.75,
2066
+ "completions/mean_length": 1571.515625,
2067
+ "completions/mean_terminated_length": 1571.515625,
2068
+ "completions/min_length": 1553.5,
2069
+ "completions/min_terminated_length": 1553.5,
2070
+ "entropy": 0.2854380081407726,
2071
+ "epoch": 0.25405786873676783,
2072
+ "geo": 0.8985359221696854,
2073
+ "grad_norm": 0.275390625,
2074
+ "learning_rate": 6.437185929648242e-06,
2075
+ "loss": 0.0,
2076
+ "num_tokens": 85053346.0,
2077
+ "point_2d": 1.0,
2078
+ "point_3d": 0.8623046875,
2079
+ "reward": 0.8784916624426842,
2080
+ "step": 720
2081
+ },
2082
+ {
2083
+ "3d_bbox": 0.8299823924899101,
2084
+ "adv": -0.0006786447083868552,
2085
+ "bbox": 0.9957724660634995,
2086
+ "completions/clipped_ratio": 0.0,
2087
+ "completions/max_length": 1599.75,
2088
+ "completions/max_terminated_length": 1599.75,
2089
+ "completions/mean_length": 1575.05078125,
2090
+ "completions/mean_terminated_length": 1575.05078125,
2091
+ "completions/min_length": 1553.625,
2092
+ "completions/min_terminated_length": 1553.625,
2093
+ "entropy": 0.2834048862569034,
2094
+ "epoch": 0.25688073394495414,
2095
+ "geo": 0.7720311819575727,
2096
+ "grad_norm": 0.353515625,
2097
+ "learning_rate": 6.396984924623116e-06,
2098
+ "loss": 0.0008,
2099
+ "num_tokens": 85997743.0,
2100
+ "point_2d": 0.875,
2101
+ "point_3d": 0.8583984375,
2102
+ "reward": 0.8897883221507072,
2103
+ "step": 728
2104
+ },
2105
+ {
2106
+ "3d_bbox": 0.814798042178154,
2107
+ "adv": -0.00043147145015609567,
2108
+ "bbox": 0.994331993162632,
2109
+ "completions/clipped_ratio": 0.0,
2110
+ "completions/max_length": 1603.0,
2111
+ "completions/max_terminated_length": 1603.0,
2112
+ "completions/mean_length": 1578.44921875,
2113
+ "completions/mean_terminated_length": 1578.44921875,
2114
+ "completions/min_length": 1556.25,
2115
+ "completions/min_terminated_length": 1556.25,
2116
+ "entropy": 0.2801376706920564,
2117
+ "epoch": 0.25970359915314045,
2118
+ "geo": 0.9146922156214714,
2119
+ "grad_norm": 0.25,
2120
+ "learning_rate": 6.3567839195979905e-06,
2121
+ "loss": 0.0001,
2122
+ "num_tokens": 86943082.0,
2123
+ "point_2d": 0.8798828125,
2124
+ "point_3d": 0.94921875,
2125
+ "reward": 0.909557893872261,
2126
+ "step": 736
2127
+ },
2128
+ {
2129
+ "3d_bbox": 0.7457513771951199,
2130
+ "adv": -1.4886696476423822e-05,
2131
+ "bbox": 0.9952592924237251,
2132
+ "completions/clipped_ratio": 0.0,
2133
+ "completions/max_length": 1592.875,
2134
+ "completions/max_terminated_length": 1592.875,
2135
+ "completions/mean_length": 1572.21484375,
2136
+ "completions/mean_terminated_length": 1572.21484375,
2137
+ "completions/min_length": 1553.625,
2138
+ "completions/min_terminated_length": 1553.625,
2139
+ "entropy": 0.28585455380380154,
2140
+ "epoch": 0.26252646436132676,
2141
+ "geo": 0.9565036073327065,
2142
+ "grad_norm": 0.345703125,
2143
+ "learning_rate": 6.316582914572866e-06,
2144
+ "loss": -0.0003,
2145
+ "num_tokens": 87886913.0,
2146
+ "point_2d": 1.0,
2147
+ "point_3d": 0.8095703125,
2148
+ "reward": 0.8876452520489693,
2149
+ "step": 744
2150
+ },
2151
+ {
2152
+ "3d_bbox": 0.6948670484125614,
2153
+ "adv": -0.00016846299786266172,
2154
+ "bbox": 0.8553073853254318,
2155
+ "completions/clipped_ratio": 0.0,
2156
+ "completions/max_length": 1594.375,
2157
+ "completions/max_terminated_length": 1594.375,
2158
+ "completions/mean_length": 1571.93359375,
2159
+ "completions/mean_terminated_length": 1571.93359375,
2160
+ "completions/min_length": 1552.0,
2161
+ "completions/min_terminated_length": 1552.0,
2162
+ "entropy": 0.28335432009771466,
2163
+ "epoch": 0.26534932956951307,
2164
+ "geo": 0.7740265503525734,
2165
+ "grad_norm": 0.421875,
2166
+ "learning_rate": 6.27638190954774e-06,
2167
+ "loss": 0.0162,
2168
+ "num_tokens": 88830776.0,
2169
+ "point_2d": 0.8701171875,
2170
+ "point_3d": 0.9082605689764023,
2171
+ "reward": 0.8321380317211151,
2172
+ "step": 752
2173
+ },
2174
+ {
2175
+ "3d_bbox": 0.7499241754412651,
2176
+ "adv": 0.0011103415663455962,
2177
+ "bbox": 0.9951865822076797,
2178
+ "completions/clipped_ratio": 0.0,
2179
+ "completions/max_length": 1592.75,
2180
+ "completions/max_terminated_length": 1592.75,
2181
+ "completions/mean_length": 1570.93359375,
2182
+ "completions/mean_terminated_length": 1570.93359375,
2183
+ "completions/min_length": 1552.25,
2184
+ "completions/min_terminated_length": 1552.25,
2185
+ "entropy": 0.28731182869523764,
2186
+ "epoch": 0.2681721947776994,
2187
+ "geo": 0.620879240334034,
2188
+ "grad_norm": 0.2197265625,
2189
+ "learning_rate": 6.236180904522614e-06,
2190
+ "loss": 0.0001,
2191
+ "num_tokens": 89774255.0,
2192
+ "point_2d": 0.7119140625,
2193
+ "point_3d": 0.8916015625,
2194
+ "reward": 0.8371566012501717,
2195
+ "step": 760
2196
+ },
2197
+ {
2198
+ "3d_bbox": 0.7572767026722431,
2199
+ "adv": 0.0006080263119656593,
2200
+ "bbox": 0.9954995736479759,
2201
+ "completions/clipped_ratio": 0.0,
2202
+ "completions/max_length": 1589.0,
2203
+ "completions/max_terminated_length": 1589.0,
2204
+ "completions/mean_length": 1569.015625,
2205
+ "completions/mean_terminated_length": 1569.015625,
2206
+ "completions/min_length": 1549.75,
2207
+ "completions/min_terminated_length": 1549.75,
2208
+ "entropy": 0.2817346737720072,
2209
+ "epoch": 0.2709950599858857,
2210
+ "geo": 0.8461784943938255,
2211
+ "grad_norm": 0.41015625,
2212
+ "learning_rate": 6.195979899497488e-06,
2213
+ "loss": -0.0,
2214
+ "num_tokens": 90717243.0,
2215
+ "point_2d": 1.0,
2216
+ "point_3d": 0.892578125,
2217
+ "reward": 0.9113386049866676,
2218
+ "step": 768
2219
+ },
2220
+ {
2221
+ "3d_bbox": 0.8124237954616547,
2222
+ "adv": -0.00010392009426141158,
2223
+ "bbox": 0.9695260003209114,
2224
+ "completions/clipped_ratio": 0.0,
2225
+ "completions/max_length": 1590.0,
2226
+ "completions/max_terminated_length": 1590.0,
2227
+ "completions/mean_length": 1571.890625,
2228
+ "completions/mean_terminated_length": 1571.890625,
2229
+ "completions/min_length": 1553.75,
2230
+ "completions/min_terminated_length": 1553.75,
2231
+ "entropy": 0.2867649751715362,
2232
+ "epoch": 0.273817925194072,
2233
+ "geo": 0.7172370981425047,
2234
+ "grad_norm": 0.29296875,
2235
+ "learning_rate": 6.155778894472362e-06,
2236
+ "loss": -0.0,
2237
+ "num_tokens": 91660903.0,
2238
+ "point_2d": 0.75,
2239
+ "point_3d": 0.9462890625,
2240
+ "reward": 0.8695597127079964,
2241
+ "step": 776
2242
+ },
2243
+ {
2244
+ "3d_bbox": 0.6247975081205368,
2245
+ "adv": 0.00014501793702947907,
2246
+ "bbox": 0.9793029651045799,
2247
+ "completions/clipped_ratio": 0.0,
2248
+ "completions/max_length": 1592.125,
2249
+ "completions/max_terminated_length": 1592.125,
2250
+ "completions/mean_length": 1571.03515625,
2251
+ "completions/mean_terminated_length": 1571.03515625,
2252
+ "completions/min_length": 1550.875,
2253
+ "completions/min_terminated_length": 1550.875,
2254
+ "entropy": 0.2792881946079433,
2255
+ "epoch": 0.2766407904022583,
2256
+ "geo": 0.8185150865465403,
2257
+ "grad_norm": 0.2890625,
2258
+ "learning_rate": 6.115577889447236e-06,
2259
+ "loss": 0.0,
2260
+ "num_tokens": 92604352.0,
2261
+ "point_2d": 0.8876953125,
2262
+ "point_3d": 0.8994140625,
2263
+ "reward": 0.8478024452924728,
2264
+ "step": 784
2265
+ },
2266
+ {
2267
+ "3d_bbox": 0.7522557079792023,
2268
+ "adv": 0.00011037173771910602,
2269
+ "bbox": 0.8692253082990646,
2270
+ "completions/clipped_ratio": 0.0,
2271
+ "completions/max_length": 1594.625,
2272
+ "completions/max_terminated_length": 1594.625,
2273
+ "completions/mean_length": 1572.27734375,
2274
+ "completions/mean_terminated_length": 1572.27734375,
2275
+ "completions/min_length": 1552.5,
2276
+ "completions/min_terminated_length": 1552.5,
2277
+ "entropy": 0.2874673060141504,
2278
+ "epoch": 0.2794636556104446,
2279
+ "geo": 0.9408122524619102,
2280
+ "grad_norm": 0.31640625,
2281
+ "learning_rate": 6.07537688442211e-06,
2282
+ "loss": 0.0144,
2283
+ "num_tokens": 93548159.0,
2284
+ "point_2d": 0.7587890625,
2285
+ "point_3d": 0.91796875,
2286
+ "reward": 0.8245597034692764,
2287
+ "step": 792
2288
+ },
2289
+ {
2290
+ "3d_bbox": 0.8205922544002533,
2291
+ "adv": 0.0001977139909286052,
2292
+ "bbox": 0.9971112534403801,
2293
+ "completions/clipped_ratio": 0.0,
2294
+ "completions/max_length": 1589.25,
2295
+ "completions/max_terminated_length": 1589.25,
2296
+ "completions/mean_length": 1571.43359375,
2297
+ "completions/mean_terminated_length": 1571.43359375,
2298
+ "completions/min_length": 1552.625,
2299
+ "completions/min_terminated_length": 1552.625,
2300
+ "entropy": 0.28211602522060275,
2301
+ "epoch": 0.2822865208186309,
2302
+ "geo": 0.8445243984460831,
2303
+ "grad_norm": 0.3203125,
2304
+ "learning_rate": 6.035175879396985e-06,
2305
+ "loss": 0.0007,
2306
+ "num_tokens": 94491662.0,
2307
+ "point_2d": 0.884765625,
2308
+ "point_3d": 0.9365234375,
2309
+ "reward": 0.9097481444478035,
2310
+ "step": 800
2311
+ },
2312
+ {
2313
+ "3d_bbox": 0.794718824326992,
2314
+ "adv": -0.00023952151968842372,
2315
+ "bbox": 0.9944768622517586,
2316
+ "completions/clipped_ratio": 0.0,
2317
+ "completions/max_length": 1601.375,
2318
+ "completions/max_terminated_length": 1601.375,
2319
+ "completions/mean_length": 1571.67578125,
2320
+ "completions/mean_terminated_length": 1571.67578125,
2321
+ "completions/min_length": 1552.125,
2322
+ "completions/min_terminated_length": 1552.125,
2323
+ "entropy": 0.2849376816302538,
2324
+ "epoch": 0.28510938602681724,
2325
+ "geo": 0.9431384205818176,
2326
+ "grad_norm": 0.34375,
2327
+ "learning_rate": 5.99497487437186e-06,
2328
+ "loss": 0.0001,
2329
+ "num_tokens": 95435451.0,
2330
+ "point_2d": 0.9990234375,
2331
+ "point_3d": 0.8720703125,
2332
+ "reward": 0.9150723442435265,
2333
+ "step": 808
2334
+ },
2335
+ {
2336
+ "3d_bbox": 0.7696447446942329,
2337
+ "adv": 0.0008425721316598356,
2338
+ "bbox": 0.9485463574528694,
2339
+ "completions/clipped_ratio": 0.0,
2340
+ "completions/max_length": 1593.875,
2341
+ "completions/max_terminated_length": 1593.875,
2342
+ "completions/mean_length": 1570.11328125,
2343
+ "completions/mean_terminated_length": 1570.11328125,
2344
+ "completions/min_length": 1553.875,
2345
+ "completions/min_terminated_length": 1553.875,
2346
+ "entropy": 0.28682070411741734,
2347
+ "epoch": 0.28793225123500354,
2348
+ "geo": 0.9097637981176376,
2349
+ "grad_norm": 0.2353515625,
2350
+ "learning_rate": 5.954773869346734e-06,
2351
+ "loss": -0.0006,
2352
+ "num_tokens": 96378680.0,
2353
+ "point_2d": 0.9990234375,
2354
+ "point_3d": 0.9013671875,
2355
+ "reward": 0.9046454280614853,
2356
+ "step": 816
2357
+ },
2358
+ {
2359
+ "3d_bbox": 0.6897615268826485,
2360
+ "adv": -0.000497654465107189,
2361
+ "bbox": 0.8617971315979958,
2362
+ "completions/clipped_ratio": 0.0,
2363
+ "completions/max_length": 1602.125,
2364
+ "completions/max_terminated_length": 1602.125,
2365
+ "completions/mean_length": 1577.7734375,
2366
+ "completions/mean_terminated_length": 1577.7734375,
2367
+ "completions/min_length": 1558.125,
2368
+ "completions/min_terminated_length": 1558.125,
2369
+ "entropy": 0.28228796226903796,
2370
+ "epoch": 0.29075511644318985,
2371
+ "geo": 0.7427511867135763,
2372
+ "grad_norm": 0.326171875,
2373
+ "learning_rate": 5.914572864321608e-06,
2374
+ "loss": 0.0001,
2375
+ "num_tokens": 97324310.0,
2376
+ "point_2d": 0.7529296875,
2377
+ "point_3d": 0.919921875,
2378
+ "reward": 0.8061025515198708,
2379
+ "step": 824
2380
+ },
2381
+ {
2382
+ "3d_bbox": 0.6537492573261261,
2383
+ "adv": -0.00023956477116371389,
2384
+ "bbox": 0.9781710132956505,
2385
+ "completions/clipped_ratio": 0.0,
2386
+ "completions/max_length": 1596.0,
2387
+ "completions/max_terminated_length": 1596.0,
2388
+ "completions/mean_length": 1574.296875,
2389
+ "completions/mean_terminated_length": 1574.296875,
2390
+ "completions/min_length": 1554.75,
2391
+ "completions/min_terminated_length": 1554.75,
2392
+ "entropy": 0.28043507505208254,
2393
+ "epoch": 0.29357798165137616,
2394
+ "geo": 0.8402111828327179,
2395
+ "grad_norm": 0.291015625,
2396
+ "learning_rate": 5.874371859296482e-06,
2397
+ "loss": -0.0001,
2398
+ "num_tokens": 98268682.0,
2399
+ "point_2d": 1.0,
2400
+ "point_3d": 0.9551355689764023,
2401
+ "reward": 0.8967639729380608,
2402
+ "step": 832
2403
+ },
2404
+ {
2405
+ "3d_bbox": 0.755732849240303,
2406
+ "adv": -0.000661049871268915,
2407
+ "bbox": 0.9913836792111397,
2408
+ "completions/clipped_ratio": 0.0,
2409
+ "completions/max_length": 1598.125,
2410
+ "completions/max_terminated_length": 1598.125,
2411
+ "completions/mean_length": 1575.3125,
2412
+ "completions/mean_terminated_length": 1575.3125,
2413
+ "completions/min_length": 1555.625,
2414
+ "completions/min_terminated_length": 1555.625,
2415
+ "entropy": 0.2857117219828069,
2416
+ "epoch": 0.2964008468595625,
2417
+ "geo": 0.8524684980511665,
2418
+ "grad_norm": 0.380859375,
2419
+ "learning_rate": 5.8341708542713575e-06,
2420
+ "loss": 0.0143,
2421
+ "num_tokens": 99213386.0,
2422
+ "point_2d": 0.875,
2423
+ "point_3d": 0.876953125,
2424
+ "reward": 0.8747674003243446,
2425
+ "step": 840
2426
+ },
2427
+ {
2428
+ "3d_bbox": 0.7413047328591347,
2429
+ "adv": 0.0002993446832988411,
2430
+ "bbox": 0.8264872687868774,
2431
+ "completions/clipped_ratio": 0.0,
2432
+ "completions/max_length": 1594.125,
2433
+ "completions/max_terminated_length": 1594.125,
2434
+ "completions/mean_length": 1573.984375,
2435
+ "completions/mean_terminated_length": 1573.984375,
2436
+ "completions/min_length": 1556.75,
2437
+ "completions/min_terminated_length": 1556.75,
2438
+ "entropy": 0.28746917750686407,
2439
+ "epoch": 0.2992237120677488,
2440
+ "geo": 0.9443965330719948,
2441
+ "grad_norm": 0.3359375,
2442
+ "learning_rate": 5.793969849246232e-06,
2443
+ "loss": 0.0195,
2444
+ "num_tokens": 100157950.0,
2445
+ "point_2d": 1.0,
2446
+ "point_3d": 0.9140625,
2447
+ "reward": 0.8704636245965958,
2448
+ "step": 848
2449
+ }
2450
+ ],
2451
+ "logging_steps": 8,
2452
+ "max_steps": 2000,
2453
+ "num_input_tokens_seen": 100393671,
2454
+ "num_train_epochs": 1,
2455
+ "save_steps": 50,
2456
+ "stateful_callbacks": {
2457
+ "TrainerControl": {
2458
+ "args": {
2459
+ "should_epoch_stop": false,
2460
+ "should_evaluate": false,
2461
+ "should_log": false,
2462
+ "should_save": true,
2463
+ "should_training_stop": false
2464
+ },
2465
+ "attributes": {}
2466
+ }
2467
+ },
2468
+ "total_flos": 0.0,
2469
+ "train_batch_size": 1,
2470
+ "trial_name": null,
2471
+ "trial_params": null
2472
+ }
checkpoint-850/video_preprocessor_config.json ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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_pad": null,
10
+ "do_rescale": true,
11
+ "do_resize": true,
12
+ "do_sample_frames": false,
13
+ "fps": null,
14
+ "image_mean": [
15
+ 0.48145466,
16
+ 0.4578275,
17
+ 0.40821073
18
+ ],
19
+ "image_std": [
20
+ 0.26862954,
21
+ 0.26130258,
22
+ 0.27577711
23
+ ],
24
+ "input_data_format": null,
25
+ "max_frames": 768,
26
+ "max_pixels": 12845056,
27
+ "merge_size": 2,
28
+ "min_frames": 4,
29
+ "min_pixels": 3136,
30
+ "num_frames": 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
+ "size_divisor": null,
41
+ "temporal_patch_size": 2,
42
+ "video_metadata": null,
43
+ "video_processor_type": "Qwen2VLVideoProcessor"
44
+ }
checkpoint-850/vocab.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-950/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
+ }
checkpoint-950/chat_template.jinja ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
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
2
+ You are a helpful assistant.<|im_end|>
3
+ {% endif %}<|im_start|>{{ message['role'] }}
4
+ {% if message['content'] is string %}{{ message['content'] }}<|im_end|>
5
+ {% else %}{% for content in message['content'] %}{% if content['type'] == 'image' or 'image' in content or 'image_url' in content %}{% set image_count.value = image_count.value + 1 %}{% if add_vision_id %}Picture {{ image_count.value }}: {% endif %}<|vision_start|><|image_pad|><|vision_end|>{% elif content['type'] == 'video' or 'video' in content %}{% set video_count.value = video_count.value + 1 %}{% if add_vision_id %}Video {{ video_count.value }}: {% endif %}<|vision_start|><|video_pad|><|vision_end|>{% elif 'text' in content %}{{ content['text'] }}{% endif %}{% endfor %}<|im_end|>
6
+ {% endif %}{% endfor %}{% if add_generation_prompt %}<|im_start|>assistant
7
+ {% endif %}
checkpoint-950/config.json ADDED
@@ -0,0 +1,143 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen2_5_PVLForConditionalGeneration"
4
+ ],
5
+ "attention_dropout": 0.0,
6
+ "dtype": "bfloat16",
7
+ "eos_token_id": 151645,
8
+ "hidden_act": "silu",
9
+ "hidden_size": 2048,
10
+ "image_token_id": 151655,
11
+ "initializer_range": 0.02,
12
+ "intermediate_size": 11008,
13
+ "max_position_embeddings": 128000,
14
+ "max_window_layers": 70,
15
+ "model_type": "qwen2_5_vl",
16
+ "num_attention_heads": 16,
17
+ "num_hidden_layers": 36,
18
+ "num_key_value_heads": 2,
19
+ "pad_token_id": 151643,
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
+ "architectures": [
34
+ "Qwen2_5_VLForConditionalGeneration"
35
+ ],
36
+ "attention_dropout": 0.0,
37
+ "bos_token_id": 151643,
38
+ "dtype": "bfloat16",
39
+ "eos_token_id": 151645,
40
+ "hidden_act": "silu",
41
+ "hidden_size": 2048,
42
+ "image_token_id": null,
43
+ "initializer_range": 0.02,
44
+ "intermediate_size": 11008,
45
+ "layer_types": [
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
+ "full_attention",
74
+ "full_attention",
75
+ "full_attention",
76
+ "full_attention",
77
+ "full_attention",
78
+ "full_attention",
79
+ "full_attention",
80
+ "full_attention",
81
+ "full_attention"
82
+ ],
83
+ "max_position_embeddings": 128000,
84
+ "max_window_layers": 70,
85
+ "model_type": "qwen2_5_vl_text",
86
+ "num_attention_heads": 16,
87
+ "num_hidden_layers": 36,
88
+ "num_key_value_heads": 2,
89
+ "rms_norm_eps": 1e-06,
90
+ "rope_scaling": {
91
+ "mrope_section": [
92
+ 16,
93
+ 24,
94
+ 24
95
+ ],
96
+ "rope_type": "default",
97
+ "type": "default"
98
+ },
99
+ "rope_theta": 1000000.0,
100
+ "sliding_window": null,
101
+ "tie_word_embeddings": true,
102
+ "use_cache": true,
103
+ "use_sliding_window": false,
104
+ "video_token_id": null,
105
+ "vision_end_token_id": 151653,
106
+ "vision_start_token_id": 151652,
107
+ "vision_token_id": 151654,
108
+ "vocab_size": 151936
109
+ },
110
+ "transformers_version": "4.56.2",
111
+ "use_cache": true,
112
+ "use_sliding_window": false,
113
+ "video_token_id": 151656,
114
+ "vision_config": {
115
+ "depth": 32,
116
+ "dtype": "bfloat16",
117
+ "fullatt_block_indexes": [
118
+ 7,
119
+ 15,
120
+ 23,
121
+ 31
122
+ ],
123
+ "hidden_act": "silu",
124
+ "hidden_size": 1280,
125
+ "in_channels": 3,
126
+ "in_chans": 3,
127
+ "initializer_range": 0.02,
128
+ "intermediate_size": 3420,
129
+ "model_type": "qwen2_5_vl",
130
+ "num_heads": 16,
131
+ "out_hidden_size": 2048,
132
+ "patch_size": 14,
133
+ "spatial_merge_size": 2,
134
+ "spatial_patch_size": 14,
135
+ "temporal_patch_size": 2,
136
+ "tokens_per_second": 2,
137
+ "window_size": 112
138
+ },
139
+ "vision_end_token_id": 151653,
140
+ "vision_start_token_id": 151652,
141
+ "vision_token_id": 151654,
142
+ "vocab_size": 151936
143
+ }
checkpoint-950/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": "4.56.2"
11
+ }
checkpoint-950/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-950/model.safetensors.index.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-950/preprocessor_config.json ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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_rescale": true,
11
+ "do_resize": true,
12
+ "image_mean": [
13
+ 0.48145466,
14
+ 0.4578275,
15
+ 0.40821073
16
+ ],
17
+ "image_processor_type": "Qwen2VLImageProcessorFast",
18
+ "image_std": [
19
+ 0.26862954,
20
+ 0.26130258,
21
+ 0.27577711
22
+ ],
23
+ "input_data_format": null,
24
+ "max_pixels": 12845056,
25
+ "merge_size": 2,
26
+ "min_pixels": 3136,
27
+ "patch_size": 14,
28
+ "processor_class": "Qwen2_5_VLProcessor",
29
+ "resample": 3,
30
+ "rescale_factor": 0.00392156862745098,
31
+ "return_tensors": null,
32
+ "size": {
33
+ "longest_edge": 12845056,
34
+ "shortest_edge": 3136
35
+ },
36
+ "temporal_patch_size": 2
37
+ }
checkpoint-950/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
+ }
checkpoint-950/tokenizer_config.json ADDED
@@ -0,0 +1,212 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
+ "max_length": null,
203
+ "model_max_length": 131072,
204
+ "pad_to_multiple_of": null,
205
+ "pad_token": "<|endoftext|>",
206
+ "pad_token_type_id": 0,
207
+ "padding_side": "left",
208
+ "processor_class": "Qwen2_5_VLProcessor",
209
+ "split_special_tokens": false,
210
+ "tokenizer_class": "Qwen2Tokenizer",
211
+ "unk_token": null
212
+ }
checkpoint-950/trainer_state.json ADDED
@@ -0,0 +1,2748 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.3352152434721242,
6
+ "eval_steps": 500,
7
+ "global_step": 950,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "3d_bbox": 0.6610370799899101,
14
+ "adv": 0.000926484264937244,
15
+ "bbox": 0.8075725510716438,
16
+ "completions/clipped_ratio": 0.0,
17
+ "completions/max_length": 1592.875,
18
+ "completions/max_terminated_length": 1592.875,
19
+ "completions/mean_length": 1573.1640625,
20
+ "completions/mean_terminated_length": 1573.1640625,
21
+ "completions/min_length": 1554.875,
22
+ "completions/min_terminated_length": 1554.875,
23
+ "entropy": 0.2564338077791035,
24
+ "epoch": 0.0028228652081863093,
25
+ "geo": 0.7611365988850594,
26
+ "grad_norm": 0.439453125,
27
+ "learning_rate": 7e-06,
28
+ "loss": -0.0002,
29
+ "num_tokens": 943810.0,
30
+ "point_2d": 0.875,
31
+ "point_3d": 0.638671875,
32
+ "reward": 0.7455703690648079,
33
+ "step": 8
34
+ },
35
+ {
36
+ "3d_bbox": 0.6985021606087685,
37
+ "adv": 0.0008333708265126916,
38
+ "bbox": 0.9951911941170692,
39
+ "completions/clipped_ratio": 0.0,
40
+ "completions/max_length": 1594.25,
41
+ "completions/max_terminated_length": 1594.25,
42
+ "completions/mean_length": 1573.64453125,
43
+ "completions/mean_terminated_length": 1573.64453125,
44
+ "completions/min_length": 1552.375,
45
+ "completions/min_terminated_length": 1552.375,
46
+ "entropy": 0.2595143502112478,
47
+ "epoch": 0.0056457304163726185,
48
+ "geo": 0.8901385143399239,
49
+ "grad_norm": 0.341796875,
50
+ "learning_rate": 9.974874371859297e-06,
51
+ "loss": 0.0564,
52
+ "num_tokens": 1888015.0,
53
+ "point_2d": 1.0,
54
+ "point_3d": 0.5947265625,
55
+ "reward": 0.8221049904823303,
56
+ "step": 16
57
+ },
58
+ {
59
+ "3d_bbox": 0.6706115230917931,
60
+ "adv": -0.11667590066201683,
61
+ "bbox": 0.9641365110874176,
62
+ "completions/clipped_ratio": 0.0,
63
+ "completions/max_length": 1588.75,
64
+ "completions/max_terminated_length": 1588.75,
65
+ "completions/mean_length": 1568.69921875,
66
+ "completions/mean_terminated_length": 1568.69921875,
67
+ "completions/min_length": 1549.625,
68
+ "completions/min_terminated_length": 1549.625,
69
+ "entropy": 0.2620172018650919,
70
+ "epoch": 0.008468595624558928,
71
+ "geo": 0.8719085603952408,
72
+ "grad_norm": 0.31640625,
73
+ "learning_rate": 9.934673366834172e-06,
74
+ "loss": 0.0434,
75
+ "num_tokens": 2830754.0,
76
+ "point_2d": 0.96484375,
77
+ "point_3d": 0.7060546875,
78
+ "reward": 0.8264116197824478,
79
+ "step": 24
80
+ },
81
+ {
82
+ "3d_bbox": 0.6768141128122807,
83
+ "adv": 0.00011848201484099263,
84
+ "bbox": 0.8640391379594803,
85
+ "completions/clipped_ratio": 0.0,
86
+ "completions/max_length": 1599.875,
87
+ "completions/max_terminated_length": 1599.875,
88
+ "completions/mean_length": 1579.40625,
89
+ "completions/mean_terminated_length": 1579.40625,
90
+ "completions/min_length": 1563.0,
91
+ "completions/min_terminated_length": 1563.0,
92
+ "entropy": 0.26498199813067913,
93
+ "epoch": 0.011291460832745237,
94
+ "geo": 0.8661364912986755,
95
+ "grad_norm": 0.36328125,
96
+ "learning_rate": 9.894472361809046e-06,
97
+ "loss": -0.0005,
98
+ "num_tokens": 3776674.0,
99
+ "point_2d": 0.8642578125,
100
+ "point_3d": 0.603515625,
101
+ "reward": 0.752156674861908,
102
+ "step": 32
103
+ },
104
+ {
105
+ "3d_bbox": 0.6714089289307594,
106
+ "adv": -0.000722882674608627,
107
+ "bbox": 0.9624817371368408,
108
+ "completions/clipped_ratio": 0.0,
109
+ "completions/max_length": 1593.75,
110
+ "completions/max_terminated_length": 1593.75,
111
+ "completions/mean_length": 1575.52734375,
112
+ "completions/mean_terminated_length": 1575.52734375,
113
+ "completions/min_length": 1559.875,
114
+ "completions/min_terminated_length": 1559.875,
115
+ "entropy": 0.2738626003265381,
116
+ "epoch": 0.014114326040931546,
117
+ "geo": 0.8932898789644241,
118
+ "grad_norm": 0.341796875,
119
+ "learning_rate": 9.85427135678392e-06,
120
+ "loss": 0.0002,
121
+ "num_tokens": 4721169.0,
122
+ "point_2d": 0.99609375,
123
+ "point_3d": 0.6064453125,
124
+ "reward": 0.8091074377298355,
125
+ "step": 40
126
+ },
127
+ {
128
+ "3d_bbox": 0.6575617864727974,
129
+ "adv": -0.0007460190936399158,
130
+ "bbox": 0.9782034233212471,
131
+ "completions/clipped_ratio": 0.0,
132
+ "completions/max_length": 1598.75,
133
+ "completions/max_terminated_length": 1598.75,
134
+ "completions/mean_length": 1581.515625,
135
+ "completions/mean_terminated_length": 1581.515625,
136
+ "completions/min_length": 1563.875,
137
+ "completions/min_terminated_length": 1563.875,
138
+ "entropy": 0.26874606031924486,
139
+ "epoch": 0.016937191249117856,
140
+ "geo": 0.922828696668148,
141
+ "grad_norm": 0.384765625,
142
+ "learning_rate": 9.814070351758794e-06,
143
+ "loss": 0.0001,
144
+ "num_tokens": 5667501.0,
145
+ "point_2d": 0.9404296875,
146
+ "point_3d": 0.6171875,
147
+ "reward": 0.7983455955982208,
148
+ "step": 48
149
+ },
150
+ {
151
+ "3d_bbox": 0.7516707628965378,
152
+ "adv": 0.0002944665375252953,
153
+ "bbox": 0.870270783547312,
154
+ "completions/clipped_ratio": 0.0,
155
+ "completions/max_length": 1605.5,
156
+ "completions/max_terminated_length": 1605.5,
157
+ "completions/mean_length": 1583.109375,
158
+ "completions/mean_terminated_length": 1583.109375,
159
+ "completions/min_length": 1563.125,
160
+ "completions/min_terminated_length": 1563.125,
161
+ "entropy": 0.27441642340272665,
162
+ "epoch": 0.019760056457304165,
163
+ "geo": 0.8707154765725136,
164
+ "grad_norm": 0.35546875,
165
+ "learning_rate": 9.773869346733669e-06,
166
+ "loss": 0.0002,
167
+ "num_tokens": 6614089.0,
168
+ "point_2d": 0.875,
169
+ "point_3d": 0.728515625,
170
+ "reward": 0.8063642866909504,
171
+ "step": 56
172
+ },
173
+ {
174
+ "3d_bbox": 0.7704295367002487,
175
+ "adv": 0.0006564971263287589,
176
+ "bbox": 0.8711516009643674,
177
+ "completions/clipped_ratio": 0.0,
178
+ "completions/max_length": 1604.75,
179
+ "completions/max_terminated_length": 1604.75,
180
+ "completions/mean_length": 1582.09375,
181
+ "completions/mean_terminated_length": 1582.09375,
182
+ "completions/min_length": 1562.25,
183
+ "completions/min_terminated_length": 1562.25,
184
+ "entropy": 0.2721484359353781,
185
+ "epoch": 0.022582921665490474,
186
+ "geo": 0.8998556286096573,
187
+ "grad_norm": 0.373046875,
188
+ "learning_rate": 9.733668341708543e-06,
189
+ "loss": 0.0156,
190
+ "num_tokens": 7560513.0,
191
+ "point_2d": 0.875,
192
+ "point_3d": 0.7880859375,
193
+ "reward": 0.8261667564511299,
194
+ "step": 64
195
+ },
196
+ {
197
+ "3d_bbox": 0.6965150758624077,
198
+ "adv": 5.7399442994210403e-05,
199
+ "bbox": 0.9924400746822357,
200
+ "completions/clipped_ratio": 0.0,
201
+ "completions/max_length": 1603.875,
202
+ "completions/max_terminated_length": 1603.875,
203
+ "completions/mean_length": 1580.8203125,
204
+ "completions/mean_terminated_length": 1580.8203125,
205
+ "completions/min_length": 1562.0,
206
+ "completions/min_terminated_length": 1562.0,
207
+ "entropy": 0.2694994225166738,
208
+ "epoch": 0.025405786873676783,
209
+ "geo": 0.7992948293685913,
210
+ "grad_norm": 0.40625,
211
+ "learning_rate": 9.693467336683417e-06,
212
+ "loss": -0.0004,
213
+ "num_tokens": 8506787.0,
214
+ "point_2d": 0.875,
215
+ "point_3d": 0.6201171875,
216
+ "reward": 0.796018086373806,
217
+ "step": 72
218
+ },
219
+ {
220
+ "3d_bbox": 0.9022366181015968,
221
+ "adv": 0.000732647989934776,
222
+ "bbox": 0.9670402258634567,
223
+ "completions/clipped_ratio": 0.0,
224
+ "completions/max_length": 1598.5,
225
+ "completions/max_terminated_length": 1598.5,
226
+ "completions/mean_length": 1577.83203125,
227
+ "completions/mean_terminated_length": 1577.83203125,
228
+ "completions/min_length": 1560.0,
229
+ "completions/min_terminated_length": 1560.0,
230
+ "entropy": 0.2729479926638305,
231
+ "epoch": 0.028228652081863093,
232
+ "geo": 0.770210200920701,
233
+ "grad_norm": 0.28515625,
234
+ "learning_rate": 9.653266331658291e-06,
235
+ "loss": 0.0285,
236
+ "num_tokens": 9452032.0,
237
+ "point_2d": 0.7578125,
238
+ "point_3d": 0.8740234375,
239
+ "reward": 0.8752781972289085,
240
+ "step": 80
241
+ },
242
+ {
243
+ "3d_bbox": 0.8243749290704727,
244
+ "adv": 0.0004942354327113208,
245
+ "bbox": 0.9483200311660767,
246
+ "completions/clipped_ratio": 0.0,
247
+ "completions/max_length": 1599.0,
248
+ "completions/max_terminated_length": 1599.0,
249
+ "completions/mean_length": 1578.5703125,
250
+ "completions/mean_terminated_length": 1578.5703125,
251
+ "completions/min_length": 1564.875,
252
+ "completions/min_terminated_length": 1564.875,
253
+ "entropy": 0.27196697471663356,
254
+ "epoch": 0.031051517290049402,
255
+ "geo": 0.8520620986819267,
256
+ "grad_norm": 0.337890625,
257
+ "learning_rate": 9.613065326633166e-06,
258
+ "loss": 0.0,
259
+ "num_tokens": 10397602.0,
260
+ "point_2d": 0.84375,
261
+ "point_3d": 0.7431640625,
262
+ "reward": 0.8399022594094276,
263
+ "step": 88
264
+ },
265
+ {
266
+ "3d_bbox": 0.7930564805865288,
267
+ "adv": -0.05789744130743202,
268
+ "bbox": 0.9452124014496803,
269
+ "completions/clipped_ratio": 0.0,
270
+ "completions/max_length": 1600.125,
271
+ "completions/max_terminated_length": 1600.125,
272
+ "completions/mean_length": 1579.25390625,
273
+ "completions/mean_terminated_length": 1579.25390625,
274
+ "completions/min_length": 1560.25,
275
+ "completions/min_terminated_length": 1560.25,
276
+ "entropy": 0.2732193539850414,
277
+ "epoch": 0.03387438249823571,
278
+ "geo": 0.8466203399002552,
279
+ "grad_norm": 0.35546875,
280
+ "learning_rate": 9.57286432160804e-06,
281
+ "loss": 0.0143,
282
+ "num_tokens": 11343267.0,
283
+ "point_2d": 0.7138671875,
284
+ "point_3d": 0.8203125,
285
+ "reward": 0.8181121572852135,
286
+ "step": 96
287
+ },
288
+ {
289
+ "3d_bbox": 0.6107643470168114,
290
+ "adv": -0.0019174332555849105,
291
+ "bbox": 0.9693493247032166,
292
+ "completions/clipped_ratio": 0.0,
293
+ "completions/max_length": 1605.375,
294
+ "completions/max_terminated_length": 1605.375,
295
+ "completions/mean_length": 1580.6875,
296
+ "completions/mean_terminated_length": 1580.6875,
297
+ "completions/min_length": 1559.75,
298
+ "completions/min_terminated_length": 1559.75,
299
+ "entropy": 0.2712484383955598,
300
+ "epoch": 0.03669724770642202,
301
+ "geo": 0.760684534907341,
302
+ "grad_norm": 0.392578125,
303
+ "learning_rate": 9.532663316582916e-06,
304
+ "loss": 0.0006,
305
+ "num_tokens": 12289259.0,
306
+ "point_2d": 0.849609375,
307
+ "point_3d": 0.7109375,
308
+ "reward": 0.7851651459932327,
309
+ "step": 104
310
+ },
311
+ {
312
+ "3d_bbox": 0.7011105790734291,
313
+ "adv": 0.000332610031364311,
314
+ "bbox": 0.997136726975441,
315
+ "completions/clipped_ratio": 0.0,
316
+ "completions/max_length": 1601.5,
317
+ "completions/max_terminated_length": 1601.5,
318
+ "completions/mean_length": 1578.38671875,
319
+ "completions/mean_terminated_length": 1578.38671875,
320
+ "completions/min_length": 1560.0,
321
+ "completions/min_terminated_length": 1560.0,
322
+ "entropy": 0.2741631781682372,
323
+ "epoch": 0.03952011291460833,
324
+ "geo": 0.868082121014595,
325
+ "grad_norm": 0.349609375,
326
+ "learning_rate": 9.49246231155779e-06,
327
+ "loss": -0.0005,
328
+ "num_tokens": 13234494.0,
329
+ "point_2d": 0.875,
330
+ "point_3d": 0.759765625,
331
+ "reward": 0.8332532122731209,
332
+ "step": 112
333
+ },
334
+ {
335
+ "3d_bbox": 0.6414853371679783,
336
+ "adv": 6.49127487122314e-05,
337
+ "bbox": 0.9946233779191971,
338
+ "completions/clipped_ratio": 0.0,
339
+ "completions/max_length": 1599.875,
340
+ "completions/max_terminated_length": 1599.875,
341
+ "completions/mean_length": 1580.48828125,
342
+ "completions/mean_terminated_length": 1580.48828125,
343
+ "completions/min_length": 1562.5,
344
+ "completions/min_terminated_length": 1562.5,
345
+ "entropy": 0.2765682805329561,
346
+ "epoch": 0.04234297812279464,
347
+ "geo": 0.7812720276415348,
348
+ "grad_norm": 0.392578125,
349
+ "learning_rate": 9.452261306532664e-06,
350
+ "loss": -0.0002,
351
+ "num_tokens": 14180739.0,
352
+ "point_2d": 0.875,
353
+ "point_3d": 0.7275390625,
354
+ "reward": 0.8096619322896004,
355
+ "step": 120
356
+ },
357
+ {
358
+ "3d_bbox": 0.696478821337223,
359
+ "adv": 0.00018893864194069465,
360
+ "bbox": 0.9915411844849586,
361
+ "completions/clipped_ratio": 0.0,
362
+ "completions/max_length": 1600.75,
363
+ "completions/max_terminated_length": 1600.75,
364
+ "completions/mean_length": 1581.48828125,
365
+ "completions/mean_terminated_length": 1581.48828125,
366
+ "completions/min_length": 1566.25,
367
+ "completions/min_terminated_length": 1566.25,
368
+ "entropy": 0.27939224475994706,
369
+ "epoch": 0.04516584333098095,
370
+ "geo": 0.9225678443908691,
371
+ "grad_norm": 0.359375,
372
+ "learning_rate": 9.412060301507539e-06,
373
+ "loss": 0.0004,
374
+ "num_tokens": 15126808.0,
375
+ "point_2d": 1.0,
376
+ "point_3d": 0.8037109375,
377
+ "reward": 0.8729327321052551,
378
+ "step": 128
379
+ },
380
+ {
381
+ "3d_bbox": 0.7811713069677353,
382
+ "adv": -0.00017097587488024146,
383
+ "bbox": 0.983598880469799,
384
+ "completions/clipped_ratio": 0.0,
385
+ "completions/max_length": 1599.0,
386
+ "completions/max_terminated_length": 1599.0,
387
+ "completions/mean_length": 1575.9765625,
388
+ "completions/mean_terminated_length": 1575.9765625,
389
+ "completions/min_length": 1560.875,
390
+ "completions/min_terminated_length": 1560.875,
391
+ "entropy": 0.2712345658801496,
392
+ "epoch": 0.04798870853916726,
393
+ "geo": 0.829373586922884,
394
+ "grad_norm": 0.365234375,
395
+ "learning_rate": 9.371859296482413e-06,
396
+ "loss": 0.0002,
397
+ "num_tokens": 16071682.0,
398
+ "point_2d": 0.75,
399
+ "point_3d": 0.76171875,
400
+ "reward": 0.819122239947319,
401
+ "step": 136
402
+ },
403
+ {
404
+ "3d_bbox": 0.6566238105297089,
405
+ "adv": 5.832723485355018e-05,
406
+ "bbox": 0.9859664067626,
407
+ "completions/clipped_ratio": 0.0,
408
+ "completions/max_length": 1597.5,
409
+ "completions/max_terminated_length": 1597.5,
410
+ "completions/mean_length": 1580.12890625,
411
+ "completions/mean_terminated_length": 1580.12890625,
412
+ "completions/min_length": 1565.0,
413
+ "completions/min_terminated_length": 1565.0,
414
+ "entropy": 0.2754409732297063,
415
+ "epoch": 0.05081157374735357,
416
+ "geo": 0.8895541727542877,
417
+ "grad_norm": 0.318359375,
418
+ "learning_rate": 9.331658291457287e-06,
419
+ "loss": 0.0218,
420
+ "num_tokens": 17017499.0,
421
+ "point_2d": 1.0,
422
+ "point_3d": 0.6328125,
423
+ "reward": 0.8188506811857224,
424
+ "step": 144
425
+ },
426
+ {
427
+ "3d_bbox": 0.655942264944315,
428
+ "adv": 0.00024080171715468168,
429
+ "bbox": 0.9830196872353554,
430
+ "completions/clipped_ratio": 0.0,
431
+ "completions/max_length": 1600.125,
432
+ "completions/max_terminated_length": 1600.125,
433
+ "completions/mean_length": 1581.3828125,
434
+ "completions/mean_terminated_length": 1581.3828125,
435
+ "completions/min_length": 1564.0,
436
+ "completions/min_terminated_length": 1564.0,
437
+ "entropy": 0.2764031286351383,
438
+ "epoch": 0.053634438955539876,
439
+ "geo": 0.8772727139294147,
440
+ "grad_norm": 0.294921875,
441
+ "learning_rate": 9.291457286432161e-06,
442
+ "loss": 0.015,
443
+ "num_tokens": 17963757.0,
444
+ "point_2d": 0.875,
445
+ "point_3d": 0.7880859375,
446
+ "reward": 0.8255119547247887,
447
+ "step": 152
448
+ },
449
+ {
450
+ "3d_bbox": 0.8087940737605095,
451
+ "adv": -0.062093503400319605,
452
+ "bbox": 0.9799230694770813,
453
+ "completions/clipped_ratio": 0.0,
454
+ "completions/max_length": 1598.5,
455
+ "completions/max_terminated_length": 1598.5,
456
+ "completions/mean_length": 1579.2421875,
457
+ "completions/mean_terminated_length": 1579.2421875,
458
+ "completions/min_length": 1560.25,
459
+ "completions/min_terminated_length": 1560.25,
460
+ "entropy": 0.27893486013635993,
461
+ "epoch": 0.056457304163726185,
462
+ "geo": 0.8533996716141701,
463
+ "grad_norm": 0.314453125,
464
+ "learning_rate": 9.251256281407036e-06,
465
+ "loss": 0.0157,
466
+ "num_tokens": 18909307.0,
467
+ "point_2d": 0.859375,
468
+ "point_3d": 0.814453125,
469
+ "reward": 0.8656363189220428,
470
+ "step": 160
471
+ },
472
+ {
473
+ "3d_bbox": 0.7417622283101082,
474
+ "adv": 0.00038278038937278325,
475
+ "bbox": 0.9947961643338203,
476
+ "completions/clipped_ratio": 0.0,
477
+ "completions/max_length": 1596.875,
478
+ "completions/max_terminated_length": 1596.875,
479
+ "completions/mean_length": 1578.4765625,
480
+ "completions/mean_terminated_length": 1578.4765625,
481
+ "completions/min_length": 1562.375,
482
+ "completions/min_terminated_length": 1562.375,
483
+ "entropy": 0.2772809020243585,
484
+ "epoch": 0.059280169371912494,
485
+ "geo": 0.8446554020047188,
486
+ "grad_norm": 0.337890625,
487
+ "learning_rate": 9.21105527638191e-06,
488
+ "loss": -0.0001,
489
+ "num_tokens": 19854893.0,
490
+ "point_2d": 0.875,
491
+ "point_3d": 0.7666015625,
492
+ "reward": 0.844539999961853,
493
+ "step": 168
494
+ },
495
+ {
496
+ "3d_bbox": 0.7199448123574257,
497
+ "adv": -0.00025279237161157653,
498
+ "bbox": 0.9939833134412766,
499
+ "completions/clipped_ratio": 0.0,
500
+ "completions/max_length": 1599.375,
501
+ "completions/max_terminated_length": 1599.375,
502
+ "completions/mean_length": 1577.875,
503
+ "completions/mean_terminated_length": 1577.875,
504
+ "completions/min_length": 1560.875,
505
+ "completions/min_terminated_length": 1560.875,
506
+ "entropy": 0.2748410953208804,
507
+ "epoch": 0.062103034580098804,
508
+ "geo": 0.8426482602953911,
509
+ "grad_norm": 0.353515625,
510
+ "learning_rate": 9.170854271356784e-06,
511
+ "loss": 0.0001,
512
+ "num_tokens": 20800093.0,
513
+ "point_2d": 0.8935546875,
514
+ "point_3d": 0.7666015625,
515
+ "reward": 0.8435210958123207,
516
+ "step": 176
517
+ },
518
+ {
519
+ "3d_bbox": 0.8168098330497742,
520
+ "adv": 0.0003567840885807527,
521
+ "bbox": 0.8971930183470249,
522
+ "completions/clipped_ratio": 0.0,
523
+ "completions/max_length": 1601.5,
524
+ "completions/max_terminated_length": 1601.5,
525
+ "completions/mean_length": 1581.01171875,
526
+ "completions/mean_terminated_length": 1581.01171875,
527
+ "completions/min_length": 1562.625,
528
+ "completions/min_terminated_length": 1562.625,
529
+ "entropy": 0.27716846112161875,
530
+ "epoch": 0.06492589978828511,
531
+ "geo": 0.8458116948604584,
532
+ "grad_norm": 0.376953125,
533
+ "learning_rate": 9.13065326633166e-06,
534
+ "loss": -0.0,
535
+ "num_tokens": 21746232.0,
536
+ "point_2d": 0.736328125,
537
+ "point_3d": 0.859375,
538
+ "reward": 0.8274264857172966,
539
+ "step": 184
540
+ },
541
+ {
542
+ "3d_bbox": 0.7658436857163906,
543
+ "adv": -0.00043575108065851964,
544
+ "bbox": 0.994478665292263,
545
+ "completions/clipped_ratio": 0.0,
546
+ "completions/max_length": 1599.5,
547
+ "completions/max_terminated_length": 1599.5,
548
+ "completions/mean_length": 1577.2265625,
549
+ "completions/mean_terminated_length": 1577.2265625,
550
+ "completions/min_length": 1558.25,
551
+ "completions/min_terminated_length": 1558.25,
552
+ "entropy": 0.28411462996155024,
553
+ "epoch": 0.06774876499647142,
554
+ "geo": 0.8858354687690735,
555
+ "grad_norm": 0.404296875,
556
+ "learning_rate": 9.090452261306534e-06,
557
+ "loss": -0.0,
558
+ "num_tokens": 22691162.0,
559
+ "point_2d": 1.0,
560
+ "point_3d": 0.83203125,
561
+ "reward": 0.8980883881449699,
562
+ "step": 192
563
+ },
564
+ {
565
+ "3d_bbox": 0.689879797399044,
566
+ "adv": 0.0008139211470279406,
567
+ "bbox": 0.8756355233490467,
568
+ "completions/clipped_ratio": 0.0,
569
+ "completions/max_length": 1595.25,
570
+ "completions/max_terminated_length": 1595.25,
571
+ "completions/mean_length": 1579.328125,
572
+ "completions/mean_terminated_length": 1579.328125,
573
+ "completions/min_length": 1562.125,
574
+ "completions/min_terminated_length": 1562.125,
575
+ "entropy": 0.27792059257626534,
576
+ "epoch": 0.07057163020465773,
577
+ "geo": 0.8776071295142174,
578
+ "grad_norm": 0.28125,
579
+ "learning_rate": 9.050251256281409e-06,
580
+ "loss": -0.0006,
581
+ "num_tokens": 23637118.0,
582
+ "point_2d": 0.875,
583
+ "point_3d": 0.765625,
584
+ "reward": 0.8015350699424744,
585
+ "step": 200
586
+ },
587
+ {
588
+ "3d_bbox": 0.7628588750958443,
589
+ "adv": 0.0005178588535272866,
590
+ "bbox": 0.9862369671463966,
591
+ "completions/clipped_ratio": 0.0,
592
+ "completions/max_length": 1603.5,
593
+ "completions/max_terminated_length": 1603.5,
594
+ "completions/mean_length": 1582.5,
595
+ "completions/mean_terminated_length": 1582.5,
596
+ "completions/min_length": 1561.75,
597
+ "completions/min_terminated_length": 1561.75,
598
+ "entropy": 0.2777437614277005,
599
+ "epoch": 0.07339449541284404,
600
+ "geo": 0.7896297052502632,
601
+ "grad_norm": 0.318359375,
602
+ "learning_rate": 9.010050251256283e-06,
603
+ "loss": -0.0003,
604
+ "num_tokens": 24583790.0,
605
+ "point_2d": 0.75,
606
+ "point_3d": 0.822265625,
607
+ "reward": 0.8303403630852699,
608
+ "step": 208
609
+ },
610
+ {
611
+ "3d_bbox": 0.6901299897581339,
612
+ "adv": 1.327232166659087e-05,
613
+ "bbox": 0.9742533341050148,
614
+ "completions/clipped_ratio": 0.0,
615
+ "completions/max_length": 1595.375,
616
+ "completions/max_terminated_length": 1595.375,
617
+ "completions/mean_length": 1580.3828125,
618
+ "completions/mean_terminated_length": 1580.3828125,
619
+ "completions/min_length": 1566.125,
620
+ "completions/min_terminated_length": 1566.125,
621
+ "entropy": 0.2789337821304798,
622
+ "epoch": 0.07621736062103035,
623
+ "geo": 0.9093400537967682,
624
+ "grad_norm": 0.33984375,
625
+ "learning_rate": 8.969849246231157e-06,
626
+ "loss": 0.0001,
627
+ "num_tokens": 25529904.0,
628
+ "point_2d": 1.0,
629
+ "point_3d": 0.724609375,
630
+ "reward": 0.8472481593489647,
631
+ "step": 216
632
+ },
633
+ {
634
+ "3d_bbox": 0.8317120671272278,
635
+ "adv": 0.0007813970478309784,
636
+ "bbox": 0.99237409979105,
637
+ "completions/clipped_ratio": 0.0,
638
+ "completions/max_length": 1592.375,
639
+ "completions/max_terminated_length": 1592.375,
640
+ "completions/mean_length": 1576.484375,
641
+ "completions/mean_terminated_length": 1576.484375,
642
+ "completions/min_length": 1559.125,
643
+ "completions/min_terminated_length": 1559.125,
644
+ "entropy": 0.281503742095083,
645
+ "epoch": 0.07904022582921666,
646
+ "geo": 0.9148535653948784,
647
+ "grad_norm": 0.33203125,
648
+ "learning_rate": 8.92964824120603e-06,
649
+ "loss": -0.0005,
650
+ "num_tokens": 26474884.0,
651
+ "point_2d": 1.0,
652
+ "point_3d": 0.859375,
653
+ "reward": 0.9208653047680855,
654
+ "step": 224
655
+ },
656
+ {
657
+ "3d_bbox": 0.8539343103766441,
658
+ "adv": 0.0006083497164581786,
659
+ "bbox": 0.9947174340486526,
660
+ "completions/clipped_ratio": 0.0,
661
+ "completions/max_length": 1607.25,
662
+ "completions/max_terminated_length": 1607.25,
663
+ "completions/mean_length": 1582.21875,
664
+ "completions/mean_terminated_length": 1582.21875,
665
+ "completions/min_length": 1562.75,
666
+ "completions/min_terminated_length": 1562.75,
667
+ "entropy": 0.28075447864830494,
668
+ "epoch": 0.08186309103740297,
669
+ "geo": 0.9205377101898193,
670
+ "grad_norm": 0.3828125,
671
+ "learning_rate": 8.889447236180904e-06,
672
+ "loss": 0.0151,
673
+ "num_tokens": 27421356.0,
674
+ "point_2d": 0.9423828125,
675
+ "point_3d": 0.919921875,
676
+ "reward": 0.9277391210198402,
677
+ "step": 232
678
+ },
679
+ {
680
+ "3d_bbox": 0.8143627047538757,
681
+ "adv": 0.0002584877729532309,
682
+ "bbox": 0.9961168393492699,
683
+ "completions/clipped_ratio": 0.0,
684
+ "completions/max_length": 1604.375,
685
+ "completions/max_terminated_length": 1604.375,
686
+ "completions/mean_length": 1581.84375,
687
+ "completions/mean_terminated_length": 1581.84375,
688
+ "completions/min_length": 1562.0,
689
+ "completions/min_terminated_length": 1562.0,
690
+ "entropy": 0.2808487247675657,
691
+ "epoch": 0.08468595624558928,
692
+ "geo": 0.9482192173600197,
693
+ "grad_norm": 0.3125,
694
+ "learning_rate": 8.84924623115578e-06,
695
+ "loss": 0.015,
696
+ "num_tokens": 28367588.0,
697
+ "point_2d": 1.0,
698
+ "point_3d": 0.8505859375,
699
+ "reward": 0.9152663722634315,
700
+ "step": 240
701
+ },
702
+ {
703
+ "3d_bbox": 0.7562066540122032,
704
+ "adv": 0.0011895421193912625,
705
+ "bbox": 0.9870459213852882,
706
+ "completions/clipped_ratio": 0.0,
707
+ "completions/max_length": 1605.625,
708
+ "completions/max_terminated_length": 1605.625,
709
+ "completions/mean_length": 1582.22265625,
710
+ "completions/mean_terminated_length": 1582.22265625,
711
+ "completions/min_length": 1560.125,
712
+ "completions/min_terminated_length": 1560.125,
713
+ "entropy": 0.2828774293884635,
714
+ "epoch": 0.08750882145377559,
715
+ "geo": 0.8906307443976402,
716
+ "grad_norm": 0.310546875,
717
+ "learning_rate": 8.809045226130654e-06,
718
+ "loss": -0.0001,
719
+ "num_tokens": 29314053.0,
720
+ "point_2d": 1.0,
721
+ "point_3d": 0.84765625,
722
+ "reward": 0.8977271988987923,
723
+ "step": 248
724
+ },
725
+ {
726
+ "3d_bbox": 0.662892609834671,
727
+ "adv": -0.0004179688154977157,
728
+ "bbox": 0.9923923686146736,
729
+ "completions/clipped_ratio": 0.0,
730
+ "completions/max_length": 1598.875,
731
+ "completions/max_terminated_length": 1598.875,
732
+ "completions/mean_length": 1579.796875,
733
+ "completions/mean_terminated_length": 1579.796875,
734
+ "completions/min_length": 1563.125,
735
+ "completions/min_terminated_length": 1563.125,
736
+ "entropy": 0.27959624165669084,
737
+ "epoch": 0.0903316866619619,
738
+ "geo": 0.9055434167385101,
739
+ "grad_norm": 0.259765625,
740
+ "learning_rate": 8.768844221105528e-06,
741
+ "loss": -0.0002,
742
+ "num_tokens": 30259921.0,
743
+ "point_2d": 0.7509765625,
744
+ "point_3d": 0.8154296875,
745
+ "reward": 0.805422805249691,
746
+ "step": 256
747
+ },
748
+ {
749
+ "3d_bbox": 0.7727803811430931,
750
+ "adv": -0.0010380800276834634,
751
+ "bbox": 0.9807735681533813,
752
+ "completions/clipped_ratio": 0.0,
753
+ "completions/max_length": 1604.375,
754
+ "completions/max_terminated_length": 1604.375,
755
+ "completions/mean_length": 1581.1796875,
756
+ "completions/mean_terminated_length": 1581.1796875,
757
+ "completions/min_length": 1564.375,
758
+ "completions/min_terminated_length": 1564.375,
759
+ "entropy": 0.28243178501725197,
760
+ "epoch": 0.0931545518701482,
761
+ "geo": 0.783092018100433,
762
+ "grad_norm": 0.3203125,
763
+ "learning_rate": 8.728643216080403e-06,
764
+ "loss": 0.0003,
765
+ "num_tokens": 31205887.0,
766
+ "point_2d": 0.75,
767
+ "point_3d": 0.8544921875,
768
+ "reward": 0.8395115286111832,
769
+ "step": 264
770
+ },
771
+ {
772
+ "3d_bbox": 0.7379852011799812,
773
+ "adv": -0.0005326499958755448,
774
+ "bbox": 0.9934581220149994,
775
+ "completions/clipped_ratio": 0.0,
776
+ "completions/max_length": 1598.875,
777
+ "completions/max_terminated_length": 1598.875,
778
+ "completions/mean_length": 1579.44921875,
779
+ "completions/mean_terminated_length": 1579.44921875,
780
+ "completions/min_length": 1560.375,
781
+ "completions/min_terminated_length": 1560.375,
782
+ "entropy": 0.2767978576011956,
783
+ "epoch": 0.09597741707833451,
784
+ "geo": 0.9199443385004997,
785
+ "grad_norm": 0.40625,
786
+ "learning_rate": 8.688442211055277e-06,
787
+ "loss": 0.0002,
788
+ "num_tokens": 32151650.0,
789
+ "point_2d": 0.9990234375,
790
+ "point_3d": 0.8896484375,
791
+ "reward": 0.9050287976861,
792
+ "step": 272
793
+ },
794
+ {
795
+ "3d_bbox": 0.6362542062997818,
796
+ "adv": 0.0003683161558001302,
797
+ "bbox": 0.9273895137012005,
798
+ "completions/clipped_ratio": 0.0,
799
+ "completions/max_length": 1603.375,
800
+ "completions/max_terminated_length": 1603.375,
801
+ "completions/mean_length": 1583.1875,
802
+ "completions/mean_terminated_length": 1583.1875,
803
+ "completions/min_length": 1566.25,
804
+ "completions/min_terminated_length": 1566.25,
805
+ "entropy": 0.28121166303753853,
806
+ "epoch": 0.09880028228652082,
807
+ "geo": 0.6698016822338104,
808
+ "grad_norm": 0.419921875,
809
+ "learning_rate": 8.648241206030151e-06,
810
+ "loss": 0.0104,
811
+ "num_tokens": 33098514.0,
812
+ "point_2d": 0.748046875,
813
+ "point_3d": 0.8564453125,
814
+ "reward": 0.7920339703559875,
815
+ "step": 280
816
+ },
817
+ {
818
+ "3d_bbox": 0.8045329824090004,
819
+ "adv": 0.0005392557122831931,
820
+ "bbox": 0.9957676902413368,
821
+ "completions/clipped_ratio": 0.0,
822
+ "completions/max_length": 1598.75,
823
+ "completions/max_terminated_length": 1598.75,
824
+ "completions/mean_length": 1582.8515625,
825
+ "completions/mean_terminated_length": 1582.8515625,
826
+ "completions/min_length": 1564.75,
827
+ "completions/min_terminated_length": 1564.75,
828
+ "entropy": 0.2840573275461793,
829
+ "epoch": 0.10162314749470713,
830
+ "geo": 0.894830085337162,
831
+ "grad_norm": 0.263671875,
832
+ "learning_rate": 8.608040201005025e-06,
833
+ "loss": -0.0003,
834
+ "num_tokens": 34045156.0,
835
+ "point_2d": 1.0,
836
+ "point_3d": 0.8974609375,
837
+ "reward": 0.9244403913617134,
838
+ "step": 288
839
+ },
840
+ {
841
+ "3d_bbox": 0.6487490832805634,
842
+ "adv": -0.036328162997961044,
843
+ "bbox": 0.9537674784660339,
844
+ "completions/clipped_ratio": 0.0,
845
+ "completions/max_length": 1606.125,
846
+ "completions/max_terminated_length": 1606.125,
847
+ "completions/mean_length": 1583.43359375,
848
+ "completions/mean_terminated_length": 1583.43359375,
849
+ "completions/min_length": 1564.25,
850
+ "completions/min_terminated_length": 1564.25,
851
+ "entropy": 0.2789871282875538,
852
+ "epoch": 0.10444601270289344,
853
+ "geo": 0.8219972774386406,
854
+ "grad_norm": 0.3671875,
855
+ "learning_rate": 8.5678391959799e-06,
856
+ "loss": 0.0091,
857
+ "num_tokens": 34991979.0,
858
+ "point_2d": 0.875,
859
+ "point_3d": 0.8125,
860
+ "reward": 0.8225041329860687,
861
+ "step": 296
862
+ },
863
+ {
864
+ "3d_bbox": 0.7852863818407059,
865
+ "adv": -2.922765997936949e-05,
866
+ "bbox": 0.9355472847819328,
867
+ "completions/clipped_ratio": 0.0,
868
+ "completions/max_length": 1601.5,
869
+ "completions/max_terminated_length": 1601.5,
870
+ "completions/mean_length": 1579.7890625,
871
+ "completions/mean_terminated_length": 1579.7890625,
872
+ "completions/min_length": 1561.5,
873
+ "completions/min_terminated_length": 1561.5,
874
+ "entropy": 0.2808959139510989,
875
+ "epoch": 0.10726887791107975,
876
+ "geo": 0.7100357487797737,
877
+ "grad_norm": 0.333984375,
878
+ "learning_rate": 8.527638190954774e-06,
879
+ "loss": 0.0003,
880
+ "num_tokens": 35937853.0,
881
+ "point_2d": 0.6728515625,
882
+ "point_3d": 0.798828125,
883
+ "reward": 0.7981283441185951,
884
+ "step": 304
885
+ },
886
+ {
887
+ "3d_bbox": 0.6897341758012772,
888
+ "adv": -0.0010775219143397408,
889
+ "bbox": 0.9957167953252792,
890
+ "completions/clipped_ratio": 0.0,
891
+ "completions/max_length": 1598.0,
892
+ "completions/max_terminated_length": 1598.0,
893
+ "completions/mean_length": 1578.40234375,
894
+ "completions/mean_terminated_length": 1578.40234375,
895
+ "completions/min_length": 1560.625,
896
+ "completions/min_terminated_length": 1560.625,
897
+ "entropy": 0.28013237891718745,
898
+ "epoch": 0.11009174311926606,
899
+ "geo": 0.8601852916181087,
900
+ "grad_norm": 0.322265625,
901
+ "learning_rate": 8.487437185929648e-06,
902
+ "loss": 0.0003,
903
+ "num_tokens": 36883412.0,
904
+ "point_2d": 1.0,
905
+ "point_3d": 0.8154296875,
906
+ "reward": 0.8752201721072197,
907
+ "step": 312
908
+ },
909
+ {
910
+ "3d_bbox": 0.7732156217098236,
911
+ "adv": 0.00015621728016412817,
912
+ "bbox": 0.9900884553790092,
913
+ "completions/clipped_ratio": 0.0,
914
+ "completions/max_length": 1595.25,
915
+ "completions/max_terminated_length": 1595.25,
916
+ "completions/mean_length": 1574.06640625,
917
+ "completions/mean_terminated_length": 1574.06640625,
918
+ "completions/min_length": 1556.5,
919
+ "completions/min_terminated_length": 1556.5,
920
+ "entropy": 0.2842056662775576,
921
+ "epoch": 0.11291460832745237,
922
+ "geo": 0.8583031129091978,
923
+ "grad_norm": 0.359375,
924
+ "learning_rate": 8.447236180904524e-06,
925
+ "loss": -0.0003,
926
+ "num_tokens": 37827701.0,
927
+ "point_2d": 0.8583984375,
928
+ "point_3d": 0.8681640625,
929
+ "reward": 0.8724666386842728,
930
+ "step": 320
931
+ },
932
+ {
933
+ "3d_bbox": 0.7427574843168259,
934
+ "adv": -0.00022968700068304315,
935
+ "bbox": 0.9458357989788055,
936
+ "completions/clipped_ratio": 0.0,
937
+ "completions/max_length": 1599.25,
938
+ "completions/max_terminated_length": 1599.25,
939
+ "completions/mean_length": 1576.78125,
940
+ "completions/mean_terminated_length": 1576.78125,
941
+ "completions/min_length": 1555.625,
942
+ "completions/min_terminated_length": 1555.625,
943
+ "entropy": 0.28632913110777736,
944
+ "epoch": 0.11573747353563868,
945
+ "geo": 0.9527824521064758,
946
+ "grad_norm": 0.2578125,
947
+ "learning_rate": 8.407035175879398e-06,
948
+ "loss": -0.0,
949
+ "num_tokens": 38772853.0,
950
+ "point_2d": 1.0,
951
+ "point_3d": 0.8623046875,
952
+ "reward": 0.8877245113253593,
953
+ "step": 328
954
+ },
955
+ {
956
+ "3d_bbox": 0.7519282810389996,
957
+ "adv": 0.0003111484527380526,
958
+ "bbox": 0.9336327686905861,
959
+ "completions/clipped_ratio": 0.0,
960
+ "completions/max_length": 1598.0,
961
+ "completions/max_terminated_length": 1598.0,
962
+ "completions/mean_length": 1578.54296875,
963
+ "completions/mean_terminated_length": 1578.54296875,
964
+ "completions/min_length": 1560.0,
965
+ "completions/min_terminated_length": 1560.0,
966
+ "entropy": 0.2819040766917169,
967
+ "epoch": 0.11856033874382499,
968
+ "geo": 0.9259950742125511,
969
+ "grad_norm": 0.306640625,
970
+ "learning_rate": 8.366834170854273e-06,
971
+ "loss": 0.0202,
972
+ "num_tokens": 39718400.0,
973
+ "point_2d": 0.7451171875,
974
+ "point_3d": 0.748046875,
975
+ "reward": 0.7946812808513641,
976
+ "step": 336
977
+ },
978
+ {
979
+ "3d_bbox": 0.6988782063126564,
980
+ "adv": 8.153711894226534e-05,
981
+ "bbox": 0.9918540865182877,
982
+ "completions/clipped_ratio": 0.0,
983
+ "completions/max_length": 1602.0,
984
+ "completions/max_terminated_length": 1602.0,
985
+ "completions/mean_length": 1578.3984375,
986
+ "completions/mean_terminated_length": 1578.3984375,
987
+ "completions/min_length": 1557.0,
988
+ "completions/min_terminated_length": 1557.0,
989
+ "entropy": 0.2852953290566802,
990
+ "epoch": 0.1213832039520113,
991
+ "geo": 0.824953593313694,
992
+ "grad_norm": 0.33984375,
993
+ "learning_rate": 8.326633165829147e-06,
994
+ "loss": 0.0003,
995
+ "num_tokens": 40663774.0,
996
+ "point_2d": 0.875,
997
+ "point_3d": 0.771484375,
998
+ "reward": 0.8343041799962521,
999
+ "step": 344
1000
+ },
1001
+ {
1002
+ "3d_bbox": 0.7085578441619873,
1003
+ "adv": 0.000511349891894497,
1004
+ "bbox": 0.9954689294099808,
1005
+ "completions/clipped_ratio": 0.0,
1006
+ "completions/max_length": 1601.625,
1007
+ "completions/max_terminated_length": 1601.625,
1008
+ "completions/mean_length": 1578.21484375,
1009
+ "completions/mean_terminated_length": 1578.21484375,
1010
+ "completions/min_length": 1557.5,
1011
+ "completions/min_terminated_length": 1557.5,
1012
+ "entropy": 0.2835490182042122,
1013
+ "epoch": 0.12420606916019761,
1014
+ "geo": 0.8287867829203606,
1015
+ "grad_norm": 0.31640625,
1016
+ "learning_rate": 8.286432160804021e-06,
1017
+ "loss": 0.0,
1018
+ "num_tokens": 41609261.0,
1019
+ "point_2d": 0.875,
1020
+ "point_3d": 0.8564453125,
1021
+ "reward": 0.8588679879903793,
1022
+ "step": 352
1023
+ },
1024
+ {
1025
+ "3d_bbox": 0.7324622049927711,
1026
+ "adv": -0.06052318853471661,
1027
+ "bbox": 0.9785870239138603,
1028
+ "completions/clipped_ratio": 0.0,
1029
+ "completions/max_length": 1599.75,
1030
+ "completions/max_terminated_length": 1599.75,
1031
+ "completions/mean_length": 1580.8671875,
1032
+ "completions/mean_terminated_length": 1580.8671875,
1033
+ "completions/min_length": 1562.875,
1034
+ "completions/min_terminated_length": 1562.875,
1035
+ "entropy": 0.2788377907127142,
1036
+ "epoch": 0.12702893436838392,
1037
+ "geo": 0.8886202499270439,
1038
+ "grad_norm": 0.765625,
1039
+ "learning_rate": 8.246231155778895e-06,
1040
+ "loss": 0.0295,
1041
+ "num_tokens": 42555355.0,
1042
+ "point_2d": 0.984375,
1043
+ "point_3d": 0.80078125,
1044
+ "reward": 0.8740513622760773,
1045
+ "step": 360
1046
+ },
1047
+ {
1048
+ "3d_bbox": 0.6365351751446724,
1049
+ "adv": 0.0003275554918218404,
1050
+ "bbox": 0.8720735665410757,
1051
+ "completions/clipped_ratio": 0.0,
1052
+ "completions/max_length": 1601.875,
1053
+ "completions/max_terminated_length": 1601.875,
1054
+ "completions/mean_length": 1577.73046875,
1055
+ "completions/mean_terminated_length": 1577.73046875,
1056
+ "completions/min_length": 1555.75,
1057
+ "completions/min_terminated_length": 1555.75,
1058
+ "entropy": 0.28446380188688636,
1059
+ "epoch": 0.12985179957657023,
1060
+ "geo": 0.8144352469826117,
1061
+ "grad_norm": 0.34375,
1062
+ "learning_rate": 8.20603015075377e-06,
1063
+ "loss": -0.0001,
1064
+ "num_tokens": 43500534.0,
1065
+ "point_2d": 0.875,
1066
+ "point_3d": 0.7666015625,
1067
+ "reward": 0.787552572786808,
1068
+ "step": 368
1069
+ },
1070
+ {
1071
+ "3d_bbox": 0.7198036424815655,
1072
+ "adv": -9.503715409664437e-05,
1073
+ "bbox": 0.9939679279923439,
1074
+ "completions/clipped_ratio": 0.0,
1075
+ "completions/max_length": 1592.75,
1076
+ "completions/max_terminated_length": 1592.75,
1077
+ "completions/mean_length": 1573.19140625,
1078
+ "completions/mean_terminated_length": 1573.19140625,
1079
+ "completions/min_length": 1557.0,
1080
+ "completions/min_terminated_length": 1557.0,
1081
+ "entropy": 0.28550330083817244,
1082
+ "epoch": 0.13267466478475654,
1083
+ "geo": 0.9650512784719467,
1084
+ "grad_norm": 0.359375,
1085
+ "learning_rate": 8.165829145728644e-06,
1086
+ "loss": 0.0004,
1087
+ "num_tokens": 44444583.0,
1088
+ "point_2d": 0.9892578125,
1089
+ "point_3d": 0.83984375,
1090
+ "reward": 0.8857182934880257,
1091
+ "step": 376
1092
+ },
1093
+ {
1094
+ "3d_bbox": 0.7569727748632431,
1095
+ "adv": 0.0004161505894444417,
1096
+ "bbox": 0.9042535312473774,
1097
+ "completions/clipped_ratio": 0.0,
1098
+ "completions/max_length": 1599.5,
1099
+ "completions/max_terminated_length": 1599.5,
1100
+ "completions/mean_length": 1577.55859375,
1101
+ "completions/mean_terminated_length": 1577.55859375,
1102
+ "completions/min_length": 1561.375,
1103
+ "completions/min_terminated_length": 1561.375,
1104
+ "entropy": 0.2907078079879284,
1105
+ "epoch": 0.13549752999294284,
1106
+ "geo": 0.6764855124056339,
1107
+ "grad_norm": 0.345703125,
1108
+ "learning_rate": 8.125628140703518e-06,
1109
+ "loss": 0.0001,
1110
+ "num_tokens": 45389846.0,
1111
+ "point_2d": 0.69140625,
1112
+ "point_3d": 0.8193359375,
1113
+ "reward": 0.7929921373724937,
1114
+ "step": 384
1115
+ },
1116
+ {
1117
+ "3d_bbox": 0.6764840334653854,
1118
+ "adv": -1.4531655324390158e-05,
1119
+ "bbox": 0.8646975085139275,
1120
+ "completions/clipped_ratio": 0.0,
1121
+ "completions/max_length": 1600.5,
1122
+ "completions/max_terminated_length": 1600.5,
1123
+ "completions/mean_length": 1579.08984375,
1124
+ "completions/mean_terminated_length": 1579.08984375,
1125
+ "completions/min_length": 1558.375,
1126
+ "completions/min_terminated_length": 1558.375,
1127
+ "entropy": 0.2863780874758959,
1128
+ "epoch": 0.13832039520112915,
1129
+ "geo": 0.5404472178779542,
1130
+ "grad_norm": 0.3125,
1131
+ "learning_rate": 8.085427135678392e-06,
1132
+ "loss": 0.0,
1133
+ "num_tokens": 46335629.0,
1134
+ "point_2d": 0.5048828125,
1135
+ "point_3d": 0.806640625,
1136
+ "reward": 0.713176254183054,
1137
+ "step": 392
1138
+ },
1139
+ {
1140
+ "3d_bbox": 0.7255214750766754,
1141
+ "adv": -0.0006330321593850385,
1142
+ "bbox": 0.891293590888381,
1143
+ "completions/clipped_ratio": 0.0,
1144
+ "completions/max_length": 1602.375,
1145
+ "completions/max_terminated_length": 1602.375,
1146
+ "completions/mean_length": 1576.9921875,
1147
+ "completions/mean_terminated_length": 1576.9921875,
1148
+ "completions/min_length": 1558.125,
1149
+ "completions/min_terminated_length": 1558.125,
1150
+ "entropy": 0.28716066014021635,
1151
+ "epoch": 0.14114326040931546,
1152
+ "geo": 0.9741705134510994,
1153
+ "grad_norm": 0.314453125,
1154
+ "learning_rate": 8.045226130653267e-06,
1155
+ "loss": 0.0004,
1156
+ "num_tokens": 47280507.0,
1157
+ "point_2d": 0.8916015625,
1158
+ "point_3d": 0.859375,
1159
+ "reward": 0.8419479243457317,
1160
+ "step": 400
1161
+ },
1162
+ {
1163
+ "3d_bbox": 0.7029357459396124,
1164
+ "adv": 0.0008838025460136123,
1165
+ "bbox": 0.9950641989707947,
1166
+ "completions/clipped_ratio": 0.0,
1167
+ "completions/max_length": 1598.125,
1168
+ "completions/max_terminated_length": 1598.125,
1169
+ "completions/mean_length": 1575.92578125,
1170
+ "completions/mean_terminated_length": 1575.92578125,
1171
+ "completions/min_length": 1557.75,
1172
+ "completions/min_terminated_length": 1557.75,
1173
+ "entropy": 0.2837083409540355,
1174
+ "epoch": 0.14396612561750177,
1175
+ "geo": 0.864502564072609,
1176
+ "grad_norm": 0.345703125,
1177
+ "learning_rate": 8.005025125628141e-06,
1178
+ "loss": -0.0003,
1179
+ "num_tokens": 48225240.0,
1180
+ "point_2d": 1.0,
1181
+ "point_3d": 0.8310546875,
1182
+ "reward": 0.8822636753320694,
1183
+ "step": 408
1184
+ },
1185
+ {
1186
+ "3d_bbox": 0.6913789249956608,
1187
+ "adv": -0.0002196824352722615,
1188
+ "bbox": 0.9765095412731171,
1189
+ "completions/clipped_ratio": 0.0,
1190
+ "completions/max_length": 1598.125,
1191
+ "completions/max_terminated_length": 1598.125,
1192
+ "completions/mean_length": 1579.28515625,
1193
+ "completions/mean_terminated_length": 1579.28515625,
1194
+ "completions/min_length": 1559.25,
1195
+ "completions/min_terminated_length": 1559.25,
1196
+ "entropy": 0.28565866220742464,
1197
+ "epoch": 0.14678899082568808,
1198
+ "geo": 0.8599895760416985,
1199
+ "grad_norm": 0.30859375,
1200
+ "learning_rate": 7.964824120603015e-06,
1201
+ "loss": 0.0001,
1202
+ "num_tokens": 49171057.0,
1203
+ "point_2d": 1.0,
1204
+ "point_3d": 0.904296875,
1205
+ "reward": 0.8930463194847107,
1206
+ "step": 416
1207
+ },
1208
+ {
1209
+ "3d_bbox": 0.6788053698837757,
1210
+ "adv": -0.00020617676273104735,
1211
+ "bbox": 0.9748031497001648,
1212
+ "completions/clipped_ratio": 0.0,
1213
+ "completions/max_length": 1595.25,
1214
+ "completions/max_terminated_length": 1595.25,
1215
+ "completions/mean_length": 1577.51953125,
1216
+ "completions/mean_terminated_length": 1577.51953125,
1217
+ "completions/min_length": 1559.625,
1218
+ "completions/min_terminated_length": 1559.625,
1219
+ "entropy": 0.28440852370113134,
1220
+ "epoch": 0.1496118560338744,
1221
+ "geo": 0.822394497692585,
1222
+ "grad_norm": 0.40234375,
1223
+ "learning_rate": 7.92462311557789e-06,
1224
+ "loss": 0.0006,
1225
+ "num_tokens": 50116302.0,
1226
+ "point_2d": 0.875,
1227
+ "point_3d": 0.8310546875,
1228
+ "reward": 0.8399158045649529,
1229
+ "step": 424
1230
+ },
1231
+ {
1232
+ "3d_bbox": 0.6794650703668594,
1233
+ "adv": 0.0011559298304746335,
1234
+ "bbox": 0.9097459949553013,
1235
+ "completions/clipped_ratio": 0.0,
1236
+ "completions/max_length": 1596.625,
1237
+ "completions/max_terminated_length": 1596.625,
1238
+ "completions/mean_length": 1573.80859375,
1239
+ "completions/mean_terminated_length": 1573.80859375,
1240
+ "completions/min_length": 1552.625,
1241
+ "completions/min_terminated_length": 1552.625,
1242
+ "entropy": 0.28376144589856267,
1243
+ "epoch": 0.1524347212420607,
1244
+ "geo": 0.7893084269016981,
1245
+ "grad_norm": 0.3203125,
1246
+ "learning_rate": 7.884422110552764e-06,
1247
+ "loss": 0.0,
1248
+ "num_tokens": 51060365.0,
1249
+ "point_2d": 1.0,
1250
+ "point_3d": 0.8681640625,
1251
+ "reward": 0.8643437847495079,
1252
+ "step": 432
1253
+ },
1254
+ {
1255
+ "3d_bbox": 0.7942164540290833,
1256
+ "adv": -3.183347871527076e-05,
1257
+ "bbox": 0.9926317036151886,
1258
+ "completions/clipped_ratio": 0.0,
1259
+ "completions/max_length": 1598.75,
1260
+ "completions/max_terminated_length": 1598.75,
1261
+ "completions/mean_length": 1577.796875,
1262
+ "completions/mean_terminated_length": 1577.796875,
1263
+ "completions/min_length": 1553.125,
1264
+ "completions/min_terminated_length": 1553.125,
1265
+ "entropy": 0.2821194725111127,
1266
+ "epoch": 0.155257586450247,
1267
+ "geo": 0.9503362327814102,
1268
+ "grad_norm": 0.3984375,
1269
+ "learning_rate": 7.844221105527638e-06,
1270
+ "loss": 0.0002,
1271
+ "num_tokens": 52005633.0,
1272
+ "point_2d": 1.0,
1273
+ "point_3d": 0.869140625,
1274
+ "reward": 0.9139972031116486,
1275
+ "step": 440
1276
+ },
1277
+ {
1278
+ "3d_bbox": 0.6752692721784115,
1279
+ "adv": -9.695853077573702e-05,
1280
+ "bbox": 0.982423759996891,
1281
+ "completions/clipped_ratio": 0.0,
1282
+ "completions/max_length": 1596.25,
1283
+ "completions/max_terminated_length": 1596.25,
1284
+ "completions/mean_length": 1574.5546875,
1285
+ "completions/mean_terminated_length": 1574.5546875,
1286
+ "completions/min_length": 1555.125,
1287
+ "completions/min_terminated_length": 1555.125,
1288
+ "entropy": 0.2852726257406175,
1289
+ "epoch": 0.15808045165843332,
1290
+ "geo": 0.5512092066928744,
1291
+ "grad_norm": 0.322265625,
1292
+ "learning_rate": 7.804020100502512e-06,
1293
+ "loss": -0.0004,
1294
+ "num_tokens": 52949951.0,
1295
+ "point_2d": 0.75,
1296
+ "point_3d": 0.7998046875,
1297
+ "reward": 0.801874428987503,
1298
+ "step": 448
1299
+ },
1300
+ {
1301
+ "3d_bbox": 0.7081510461866856,
1302
+ "adv": 0.0003901398231391795,
1303
+ "bbox": 0.9892760217189789,
1304
+ "completions/clipped_ratio": 0.0,
1305
+ "completions/max_length": 1596.875,
1306
+ "completions/max_terminated_length": 1596.875,
1307
+ "completions/mean_length": 1573.0390625,
1308
+ "completions/mean_terminated_length": 1573.0390625,
1309
+ "completions/min_length": 1551.875,
1310
+ "completions/min_terminated_length": 1551.875,
1311
+ "entropy": 0.2837893716059625,
1312
+ "epoch": 0.16090331686661963,
1313
+ "geo": 0.9603376165032387,
1314
+ "grad_norm": 0.3125,
1315
+ "learning_rate": 7.763819095477387e-06,
1316
+ "loss": -0.0007,
1317
+ "num_tokens": 53894057.0,
1318
+ "point_2d": 1.0,
1319
+ "point_3d": 0.794921875,
1320
+ "reward": 0.8730872422456741,
1321
+ "step": 456
1322
+ },
1323
+ {
1324
+ "3d_bbox": 0.7504811957478523,
1325
+ "adv": 0.0007141208916436881,
1326
+ "bbox": 0.9912561923265457,
1327
+ "completions/clipped_ratio": 0.0,
1328
+ "completions/max_length": 1594.75,
1329
+ "completions/max_terminated_length": 1594.75,
1330
+ "completions/mean_length": 1576.1640625,
1331
+ "completions/mean_terminated_length": 1576.1640625,
1332
+ "completions/min_length": 1555.75,
1333
+ "completions/min_terminated_length": 1555.75,
1334
+ "entropy": 0.28930201614275575,
1335
+ "epoch": 0.16372618207480594,
1336
+ "geo": 0.8148366361856461,
1337
+ "grad_norm": 0.275390625,
1338
+ "learning_rate": 7.723618090452262e-06,
1339
+ "loss": 0.0003,
1340
+ "num_tokens": 54839099.0,
1341
+ "point_2d": 0.875,
1342
+ "point_3d": 0.8896484375,
1343
+ "reward": 0.8765964508056641,
1344
+ "step": 464
1345
+ },
1346
+ {
1347
+ "3d_bbox": 0.819125022739172,
1348
+ "adv": 0.00010633431156747974,
1349
+ "bbox": 0.9906315729022026,
1350
+ "completions/clipped_ratio": 0.0,
1351
+ "completions/max_length": 1603.0,
1352
+ "completions/max_terminated_length": 1603.0,
1353
+ "completions/mean_length": 1578.3203125,
1354
+ "completions/mean_terminated_length": 1578.3203125,
1355
+ "completions/min_length": 1556.375,
1356
+ "completions/min_terminated_length": 1556.375,
1357
+ "entropy": 0.28396829310804605,
1358
+ "epoch": 0.16654904728299225,
1359
+ "geo": 0.8597357729449868,
1360
+ "grad_norm": 0.30078125,
1361
+ "learning_rate": 7.683417085427137e-06,
1362
+ "loss": 0.0004,
1363
+ "num_tokens": 55784557.0,
1364
+ "point_2d": 0.8916015625,
1365
+ "point_3d": 0.9345703125,
1366
+ "reward": 0.9089821055531502,
1367
+ "step": 472
1368
+ },
1369
+ {
1370
+ "3d_bbox": 0.7331585213541985,
1371
+ "adv": -0.00033648866519797593,
1372
+ "bbox": 0.8809062782675028,
1373
+ "completions/clipped_ratio": 0.0,
1374
+ "completions/max_length": 1594.625,
1375
+ "completions/max_terminated_length": 1594.625,
1376
+ "completions/mean_length": 1576.11328125,
1377
+ "completions/mean_terminated_length": 1576.11328125,
1378
+ "completions/min_length": 1559.25,
1379
+ "completions/min_terminated_length": 1559.25,
1380
+ "entropy": 0.28669640561565757,
1381
+ "epoch": 0.16937191249117856,
1382
+ "geo": 0.9506581574678421,
1383
+ "grad_norm": 0.2890625,
1384
+ "learning_rate": 7.643216080402011e-06,
1385
+ "loss": 0.0132,
1386
+ "num_tokens": 56729514.0,
1387
+ "point_2d": 0.8759765625,
1388
+ "point_3d": 0.880859375,
1389
+ "reward": 0.8427251875400543,
1390
+ "step": 480
1391
+ },
1392
+ {
1393
+ "3d_bbox": 0.7568283602595329,
1394
+ "adv": -0.0008519564944435842,
1395
+ "bbox": 0.9879242181777954,
1396
+ "completions/clipped_ratio": 0.0,
1397
+ "completions/max_length": 1600.875,
1398
+ "completions/max_terminated_length": 1600.875,
1399
+ "completions/mean_length": 1574.87890625,
1400
+ "completions/mean_terminated_length": 1574.87890625,
1401
+ "completions/min_length": 1554.25,
1402
+ "completions/min_terminated_length": 1554.25,
1403
+ "entropy": 0.28737954096868634,
1404
+ "epoch": 0.17219477769936486,
1405
+ "geo": 0.7571075335144997,
1406
+ "grad_norm": 0.400390625,
1407
+ "learning_rate": 7.603015075376885e-06,
1408
+ "loss": 0.0105,
1409
+ "num_tokens": 57674099.0,
1410
+ "point_2d": 0.7998046875,
1411
+ "point_3d": 0.9228515625,
1412
+ "reward": 0.8668522015213966,
1413
+ "step": 488
1414
+ },
1415
+ {
1416
+ "3d_bbox": 0.7658256888389587,
1417
+ "adv": 0.0011402575073589105,
1418
+ "bbox": 0.9932135492563248,
1419
+ "completions/clipped_ratio": 0.0,
1420
+ "completions/max_length": 1596.25,
1421
+ "completions/max_terminated_length": 1596.25,
1422
+ "completions/mean_length": 1574.265625,
1423
+ "completions/mean_terminated_length": 1574.265625,
1424
+ "completions/min_length": 1554.375,
1425
+ "completions/min_terminated_length": 1554.375,
1426
+ "entropy": 0.2865150640718639,
1427
+ "epoch": 0.17501764290755117,
1428
+ "geo": 0.8641151636838913,
1429
+ "grad_norm": 0.37890625,
1430
+ "learning_rate": 7.5628140703517595e-06,
1431
+ "loss": -0.0004,
1432
+ "num_tokens": 58618575.0,
1433
+ "point_2d": 0.91015625,
1434
+ "point_3d": 0.904296875,
1435
+ "reward": 0.8933730870485306,
1436
+ "step": 496
1437
+ },
1438
+ {
1439
+ "3d_bbox": 0.7145381905138493,
1440
+ "adv": 0.0003104716488451231,
1441
+ "bbox": 0.9926531538367271,
1442
+ "completions/clipped_ratio": 0.0,
1443
+ "completions/max_length": 1594.75,
1444
+ "completions/max_terminated_length": 1594.75,
1445
+ "completions/mean_length": 1573.12890625,
1446
+ "completions/mean_terminated_length": 1573.12890625,
1447
+ "completions/min_length": 1553.375,
1448
+ "completions/min_terminated_length": 1553.375,
1449
+ "entropy": 0.28603550279513,
1450
+ "epoch": 0.17784050811573748,
1451
+ "geo": 0.7219396531581879,
1452
+ "grad_norm": 0.310546875,
1453
+ "learning_rate": 7.522613065326634e-06,
1454
+ "loss": -0.0,
1455
+ "num_tokens": 59562552.0,
1456
+ "point_2d": 0.7509765625,
1457
+ "point_3d": 0.9365234375,
1458
+ "reward": 0.8486728444695473,
1459
+ "step": 504
1460
+ },
1461
+ {
1462
+ "3d_bbox": 0.7242622300982475,
1463
+ "adv": -0.0005710294717573561,
1464
+ "bbox": 0.9246156923472881,
1465
+ "completions/clipped_ratio": 0.0,
1466
+ "completions/max_length": 1591.75,
1467
+ "completions/max_terminated_length": 1591.75,
1468
+ "completions/mean_length": 1571.015625,
1469
+ "completions/mean_terminated_length": 1571.015625,
1470
+ "completions/min_length": 1551.625,
1471
+ "completions/min_terminated_length": 1551.625,
1472
+ "entropy": 0.2915982282720506,
1473
+ "epoch": 0.1806633733239238,
1474
+ "geo": 0.7993985265493393,
1475
+ "grad_norm": 0.408203125,
1476
+ "learning_rate": 7.482412060301508e-06,
1477
+ "loss": 0.0001,
1478
+ "num_tokens": 60505948.0,
1479
+ "point_2d": 0.7919921875,
1480
+ "point_3d": 0.8671875,
1481
+ "reward": 0.8270144015550613,
1482
+ "step": 512
1483
+ },
1484
+ {
1485
+ "3d_bbox": 0.7831042557954788,
1486
+ "adv": -0.000159245344093506,
1487
+ "bbox": 0.9960433393716812,
1488
+ "completions/clipped_ratio": 0.0,
1489
+ "completions/max_length": 1592.375,
1490
+ "completions/max_terminated_length": 1592.375,
1491
+ "completions/mean_length": 1574.84375,
1492
+ "completions/mean_terminated_length": 1574.84375,
1493
+ "completions/min_length": 1559.125,
1494
+ "completions/min_terminated_length": 1559.125,
1495
+ "entropy": 0.2845937805250287,
1496
+ "epoch": 0.1834862385321101,
1497
+ "geo": 0.9170719608664513,
1498
+ "grad_norm": 0.345703125,
1499
+ "learning_rate": 7.442211055276382e-06,
1500
+ "loss": -0.0,
1501
+ "num_tokens": 61450564.0,
1502
+ "point_2d": 1.0,
1503
+ "point_3d": 0.8408203125,
1504
+ "reward": 0.90499197691679,
1505
+ "step": 520
1506
+ },
1507
+ {
1508
+ "3d_bbox": 0.7821980714797974,
1509
+ "adv": -0.000531579293010509,
1510
+ "bbox": 0.9920122027397156,
1511
+ "completions/clipped_ratio": 0.0,
1512
+ "completions/max_length": 1595.375,
1513
+ "completions/max_terminated_length": 1595.375,
1514
+ "completions/mean_length": 1571.765625,
1515
+ "completions/mean_terminated_length": 1571.765625,
1516
+ "completions/min_length": 1551.5,
1517
+ "completions/min_terminated_length": 1551.5,
1518
+ "entropy": 0.2882155287079513,
1519
+ "epoch": 0.1863091037402964,
1520
+ "geo": 0.901853509247303,
1521
+ "grad_norm": 0.365234375,
1522
+ "learning_rate": 7.402010050251257e-06,
1523
+ "loss": 0.0002,
1524
+ "num_tokens": 62394336.0,
1525
+ "point_2d": 0.888671875,
1526
+ "point_3d": 0.9052734375,
1527
+ "reward": 0.8920388966798782,
1528
+ "step": 528
1529
+ },
1530
+ {
1531
+ "3d_bbox": 0.7537402808666229,
1532
+ "adv": -0.0004438982141437009,
1533
+ "bbox": 0.9946581870317459,
1534
+ "completions/clipped_ratio": 0.0,
1535
+ "completions/max_length": 1594.375,
1536
+ "completions/max_terminated_length": 1594.375,
1537
+ "completions/mean_length": 1574.3203125,
1538
+ "completions/mean_terminated_length": 1574.3203125,
1539
+ "completions/min_length": 1554.75,
1540
+ "completions/min_terminated_length": 1554.75,
1541
+ "entropy": 0.2886669235303998,
1542
+ "epoch": 0.18913196894848272,
1543
+ "geo": 0.8310712724924088,
1544
+ "grad_norm": 0.36328125,
1545
+ "learning_rate": 7.361809045226132e-06,
1546
+ "loss": 0.0002,
1547
+ "num_tokens": 63338834.0,
1548
+ "point_2d": 0.873046875,
1549
+ "point_3d": 0.904296875,
1550
+ "reward": 0.8814355507493019,
1551
+ "step": 536
1552
+ },
1553
+ {
1554
+ "3d_bbox": 0.6960996612906456,
1555
+ "adv": 0.0006180387658787367,
1556
+ "bbox": 0.869446316966787,
1557
+ "completions/clipped_ratio": 0.0,
1558
+ "completions/max_length": 1597.25,
1559
+ "completions/max_terminated_length": 1597.25,
1560
+ "completions/mean_length": 1571.88671875,
1561
+ "completions/mean_terminated_length": 1571.88671875,
1562
+ "completions/min_length": 1552.25,
1563
+ "completions/min_terminated_length": 1552.25,
1564
+ "entropy": 0.2887332271784544,
1565
+ "epoch": 0.19195483415666903,
1566
+ "geo": 0.826970137655735,
1567
+ "grad_norm": 0.267578125,
1568
+ "learning_rate": 7.321608040201006e-06,
1569
+ "loss": -0.0,
1570
+ "num_tokens": 64282685.0,
1571
+ "point_2d": 0.875,
1572
+ "point_3d": 0.8662109375,
1573
+ "reward": 0.8266892246901989,
1574
+ "step": 544
1575
+ },
1576
+ {
1577
+ "3d_bbox": 0.8654715269804001,
1578
+ "adv": -0.0008252508814621251,
1579
+ "bbox": 0.9932455867528915,
1580
+ "completions/clipped_ratio": 0.0,
1581
+ "completions/max_length": 1596.125,
1582
+ "completions/max_terminated_length": 1596.125,
1583
+ "completions/mean_length": 1572.1953125,
1584
+ "completions/mean_terminated_length": 1572.1953125,
1585
+ "completions/min_length": 1553.625,
1586
+ "completions/min_terminated_length": 1553.625,
1587
+ "entropy": 0.2873402568511665,
1588
+ "epoch": 0.19477769936485534,
1589
+ "geo": 0.872327484190464,
1590
+ "grad_norm": 0.2890625,
1591
+ "learning_rate": 7.28140703517588e-06,
1592
+ "loss": 0.0004,
1593
+ "num_tokens": 65226639.0,
1594
+ "point_2d": 0.8935546875,
1595
+ "point_3d": 0.9482421875,
1596
+ "reward": 0.9251285046339035,
1597
+ "step": 552
1598
+ },
1599
+ {
1600
+ "3d_bbox": 0.7851101011037827,
1601
+ "adv": -0.0011301148442726117,
1602
+ "bbox": 0.9379025921225548,
1603
+ "completions/clipped_ratio": 0.0,
1604
+ "completions/max_length": 1588.5,
1605
+ "completions/max_terminated_length": 1588.5,
1606
+ "completions/mean_length": 1571.4765625,
1607
+ "completions/mean_terminated_length": 1571.4765625,
1608
+ "completions/min_length": 1555.0,
1609
+ "completions/min_terminated_length": 1555.0,
1610
+ "entropy": 0.2852159277535975,
1611
+ "epoch": 0.19760056457304165,
1612
+ "geo": 0.690475556999445,
1613
+ "grad_norm": 0.3125,
1614
+ "learning_rate": 7.241206030150754e-06,
1615
+ "loss": 0.0006,
1616
+ "num_tokens": 66170305.0,
1617
+ "point_2d": 0.625,
1618
+ "point_3d": 0.837890625,
1619
+ "reward": 0.7964758351445198,
1620
+ "step": 560
1621
+ },
1622
+ {
1623
+ "3d_bbox": 0.7313521131873131,
1624
+ "adv": -0.000422738248744281,
1625
+ "bbox": 0.8720479012117721,
1626
+ "completions/clipped_ratio": 0.0,
1627
+ "completions/max_length": 1601.875,
1628
+ "completions/max_terminated_length": 1601.875,
1629
+ "completions/mean_length": 1574.91015625,
1630
+ "completions/mean_terminated_length": 1574.91015625,
1631
+ "completions/min_length": 1551.25,
1632
+ "completions/min_terminated_length": 1551.25,
1633
+ "entropy": 0.2880417066626251,
1634
+ "epoch": 0.20042342978122796,
1635
+ "geo": 0.8383861493784934,
1636
+ "grad_norm": 0.412109375,
1637
+ "learning_rate": 7.2010050251256295e-06,
1638
+ "loss": 0.0004,
1639
+ "num_tokens": 67114842.0,
1640
+ "point_2d": 0.8740234375,
1641
+ "point_3d": 0.84765625,
1642
+ "reward": 0.8312699124217033,
1643
+ "step": 568
1644
+ },
1645
+ {
1646
+ "3d_bbox": 0.6841592267155647,
1647
+ "adv": -0.0008141829575833981,
1648
+ "bbox": 0.9068506434559822,
1649
+ "completions/clipped_ratio": 0.0,
1650
+ "completions/max_length": 1597.5,
1651
+ "completions/max_terminated_length": 1597.5,
1652
+ "completions/mean_length": 1571.79296875,
1653
+ "completions/mean_terminated_length": 1571.79296875,
1654
+ "completions/min_length": 1546.75,
1655
+ "completions/min_terminated_length": 1546.75,
1656
+ "entropy": 0.2889366364106536,
1657
+ "epoch": 0.20324629498941427,
1658
+ "geo": 0.7308010272681713,
1659
+ "grad_norm": 0.2578125,
1660
+ "learning_rate": 7.160804020100504e-06,
1661
+ "loss": 0.0002,
1662
+ "num_tokens": 68058501.0,
1663
+ "point_2d": 0.8525390625,
1664
+ "point_3d": 0.794921875,
1665
+ "reward": 0.8096176832914352,
1666
+ "step": 576
1667
+ },
1668
+ {
1669
+ "3d_bbox": 0.5905975513160229,
1670
+ "adv": -0.00022596626513404772,
1671
+ "bbox": 0.9184868931770325,
1672
+ "completions/clipped_ratio": 0.0,
1673
+ "completions/max_length": 1597.5,
1674
+ "completions/max_terminated_length": 1597.5,
1675
+ "completions/mean_length": 1572.45703125,
1676
+ "completions/mean_terminated_length": 1572.45703125,
1677
+ "completions/min_length": 1551.125,
1678
+ "completions/min_terminated_length": 1551.125,
1679
+ "entropy": 0.2837987565435469,
1680
+ "epoch": 0.20606916019760058,
1681
+ "geo": 0.8267702832818031,
1682
+ "grad_norm": 0.314453125,
1683
+ "learning_rate": 7.120603015075378e-06,
1684
+ "loss": -0.0003,
1685
+ "num_tokens": 69002250.0,
1686
+ "point_2d": 0.8681640625,
1687
+ "point_3d": 0.7445197626948357,
1688
+ "reward": 0.7804420664906502,
1689
+ "step": 584
1690
+ },
1691
+ {
1692
+ "3d_bbox": 0.6897582598030567,
1693
+ "adv": -0.0004831653022847604,
1694
+ "bbox": 0.9742325842380524,
1695
+ "completions/clipped_ratio": 0.0,
1696
+ "completions/max_length": 1595.25,
1697
+ "completions/max_terminated_length": 1595.25,
1698
+ "completions/mean_length": 1572.9296875,
1699
+ "completions/mean_terminated_length": 1572.9296875,
1700
+ "completions/min_length": 1553.5,
1701
+ "completions/min_terminated_length": 1553.5,
1702
+ "entropy": 0.2835669592022896,
1703
+ "epoch": 0.20889202540578689,
1704
+ "geo": 0.8361655548214912,
1705
+ "grad_norm": 0.330078125,
1706
+ "learning_rate": 7.080402010050251e-06,
1707
+ "loss": 0.0004,
1708
+ "num_tokens": 69946400.0,
1709
+ "point_2d": 0.875,
1710
+ "point_3d": 0.8948184698820114,
1711
+ "reward": 0.8584523350000381,
1712
+ "step": 592
1713
+ },
1714
+ {
1715
+ "3d_bbox": 0.785271767526865,
1716
+ "adv": -0.00028875649149995297,
1717
+ "bbox": 0.9752914533019066,
1718
+ "completions/clipped_ratio": 0.0,
1719
+ "completions/max_length": 1599.0,
1720
+ "completions/max_terminated_length": 1599.0,
1721
+ "completions/mean_length": 1572.3671875,
1722
+ "completions/mean_terminated_length": 1572.3671875,
1723
+ "completions/min_length": 1549.875,
1724
+ "completions/min_terminated_length": 1549.875,
1725
+ "entropy": 0.29118923749774694,
1726
+ "epoch": 0.2117148906139732,
1727
+ "geo": 0.9338327869772911,
1728
+ "grad_norm": 0.3671875,
1729
+ "learning_rate": 7.040201005025126e-06,
1730
+ "loss": 0.0007,
1731
+ "num_tokens": 70890342.0,
1732
+ "point_2d": 0.998046875,
1733
+ "point_3d": 0.94140625,
1734
+ "reward": 0.9250040724873543,
1735
+ "step": 600
1736
+ },
1737
+ {
1738
+ "3d_bbox": 0.8766504302620888,
1739
+ "adv": -0.002171848598663928,
1740
+ "bbox": 0.9872715324163437,
1741
+ "completions/clipped_ratio": 0.0,
1742
+ "completions/max_length": 1601.125,
1743
+ "completions/max_terminated_length": 1601.125,
1744
+ "completions/mean_length": 1576.87890625,
1745
+ "completions/mean_terminated_length": 1576.87890625,
1746
+ "completions/min_length": 1556.375,
1747
+ "completions/min_terminated_length": 1556.375,
1748
+ "entropy": 0.28933983854949474,
1749
+ "epoch": 0.2145377558221595,
1750
+ "geo": 0.845995269715786,
1751
+ "grad_norm": 0.33203125,
1752
+ "learning_rate": 7e-06,
1753
+ "loss": 0.0003,
1754
+ "num_tokens": 71835359.0,
1755
+ "point_2d": 0.875,
1756
+ "point_3d": 0.9454848319292068,
1757
+ "reward": 0.9211016893386841,
1758
+ "step": 608
1759
+ },
1760
+ {
1761
+ "3d_bbox": 0.8059130385518074,
1762
+ "adv": 0.0005476666337926872,
1763
+ "bbox": 0.9907352328300476,
1764
+ "completions/clipped_ratio": 0.0,
1765
+ "completions/max_length": 1593.625,
1766
+ "completions/max_terminated_length": 1593.625,
1767
+ "completions/mean_length": 1570.9921875,
1768
+ "completions/mean_terminated_length": 1570.9921875,
1769
+ "completions/min_length": 1554.125,
1770
+ "completions/min_terminated_length": 1554.125,
1771
+ "entropy": 0.28401868836954236,
1772
+ "epoch": 0.2173606210303458,
1773
+ "geo": 0.8720198646187782,
1774
+ "grad_norm": 0.384765625,
1775
+ "learning_rate": 6.959798994974874e-06,
1776
+ "loss": -0.0004,
1777
+ "num_tokens": 72778933.0,
1778
+ "point_2d": 0.9248046875,
1779
+ "point_3d": 0.9135454967617989,
1780
+ "reward": 0.9087496176362038,
1781
+ "step": 616
1782
+ },
1783
+ {
1784
+ "3d_bbox": 0.8549554497003555,
1785
+ "adv": -0.0005420902471087174,
1786
+ "bbox": 0.9758080616593361,
1787
+ "completions/clipped_ratio": 0.0,
1788
+ "completions/max_length": 1593.0,
1789
+ "completions/max_terminated_length": 1593.0,
1790
+ "completions/mean_length": 1570.1640625,
1791
+ "completions/mean_terminated_length": 1570.1640625,
1792
+ "completions/min_length": 1549.0,
1793
+ "completions/min_terminated_length": 1549.0,
1794
+ "entropy": 0.29045473085716367,
1795
+ "epoch": 0.22018348623853212,
1796
+ "geo": 0.8916067183017731,
1797
+ "grad_norm": 0.3828125,
1798
+ "learning_rate": 6.919597989949749e-06,
1799
+ "loss": 0.0156,
1800
+ "num_tokens": 73722191.0,
1801
+ "point_2d": 1.0,
1802
+ "point_3d": 0.9639246314764023,
1803
+ "reward": 0.9486720338463783,
1804
+ "step": 624
1805
+ },
1806
+ {
1807
+ "3d_bbox": 0.7975610345602036,
1808
+ "adv": 0.0014872762099003012,
1809
+ "bbox": 0.9193930514156818,
1810
+ "completions/clipped_ratio": 0.0,
1811
+ "completions/max_length": 1600.625,
1812
+ "completions/max_terminated_length": 1600.625,
1813
+ "completions/mean_length": 1573.29296875,
1814
+ "completions/mean_terminated_length": 1573.29296875,
1815
+ "completions/min_length": 1553.0,
1816
+ "completions/min_terminated_length": 1553.0,
1817
+ "entropy": 0.28542913123965263,
1818
+ "epoch": 0.22300635144671843,
1819
+ "geo": 0.936805784702301,
1820
+ "grad_norm": 0.2734375,
1821
+ "learning_rate": 6.8793969849246235e-06,
1822
+ "loss": 0.0152,
1823
+ "num_tokens": 74666354.0,
1824
+ "point_2d": 0.9990234375,
1825
+ "point_3d": 0.93359375,
1826
+ "reward": 0.9123928248882294,
1827
+ "step": 632
1828
+ },
1829
+ {
1830
+ "3d_bbox": 0.6216618679463863,
1831
+ "adv": -0.0013801682944176719,
1832
+ "bbox": 0.9884917959570885,
1833
+ "completions/clipped_ratio": 0.0,
1834
+ "completions/max_length": 1601.375,
1835
+ "completions/max_terminated_length": 1601.375,
1836
+ "completions/mean_length": 1577.00390625,
1837
+ "completions/mean_terminated_length": 1577.00390625,
1838
+ "completions/min_length": 1554.0,
1839
+ "completions/min_terminated_length": 1554.0,
1840
+ "entropy": 0.28678226098418236,
1841
+ "epoch": 0.22582921665490474,
1842
+ "geo": 0.9096385985612869,
1843
+ "grad_norm": 0.42578125,
1844
+ "learning_rate": 6.839195979899498e-06,
1845
+ "loss": 0.0005,
1846
+ "num_tokens": 75611339.0,
1847
+ "point_2d": 0.8291015625,
1848
+ "point_3d": 0.7913028448820114,
1849
+ "reward": 0.8076395243406296,
1850
+ "step": 640
1851
+ },
1852
+ {
1853
+ "3d_bbox": 0.7582510337233543,
1854
+ "adv": -0.00019565537664334443,
1855
+ "bbox": 0.9911796301603317,
1856
+ "completions/clipped_ratio": 0.0,
1857
+ "completions/max_length": 1596.625,
1858
+ "completions/max_terminated_length": 1596.625,
1859
+ "completions/mean_length": 1575.71484375,
1860
+ "completions/mean_terminated_length": 1575.71484375,
1861
+ "completions/min_length": 1556.25,
1862
+ "completions/min_terminated_length": 1556.25,
1863
+ "entropy": 0.2863654876127839,
1864
+ "epoch": 0.22865208186309105,
1865
+ "geo": 0.9025658443570137,
1866
+ "grad_norm": 0.451171875,
1867
+ "learning_rate": 6.798994974874372e-06,
1868
+ "loss": -0.0001,
1869
+ "num_tokens": 76556042.0,
1870
+ "point_2d": 0.95703125,
1871
+ "point_3d": 0.8994140625,
1872
+ "reward": 0.9014689922332764,
1873
+ "step": 648
1874
+ },
1875
+ {
1876
+ "3d_bbox": 0.753610834479332,
1877
+ "adv": 0.00039293405916396296,
1878
+ "bbox": 0.9021090529859066,
1879
+ "completions/clipped_ratio": 0.0,
1880
+ "completions/max_length": 1592.625,
1881
+ "completions/max_terminated_length": 1592.625,
1882
+ "completions/mean_length": 1572.203125,
1883
+ "completions/mean_terminated_length": 1572.203125,
1884
+ "completions/min_length": 1552.25,
1885
+ "completions/min_terminated_length": 1552.25,
1886
+ "entropy": 0.286099414806813,
1887
+ "epoch": 0.23147494707127736,
1888
+ "geo": 0.9557241797447205,
1889
+ "grad_norm": 0.455078125,
1890
+ "learning_rate": 6.758793969849246e-06,
1891
+ "loss": 0.0002,
1892
+ "num_tokens": 77499886.0,
1893
+ "point_2d": 1.0,
1894
+ "point_3d": 0.91015625,
1895
+ "reward": 0.8914690390229225,
1896
+ "step": 656
1897
+ },
1898
+ {
1899
+ "3d_bbox": 0.831758588552475,
1900
+ "adv": -7.158980952226557e-05,
1901
+ "bbox": 0.9794979766011238,
1902
+ "completions/clipped_ratio": 0.0,
1903
+ "completions/max_length": 1595.5,
1904
+ "completions/max_terminated_length": 1595.5,
1905
+ "completions/mean_length": 1574.5,
1906
+ "completions/mean_terminated_length": 1574.5,
1907
+ "completions/min_length": 1558.75,
1908
+ "completions/min_terminated_length": 1558.75,
1909
+ "entropy": 0.2819104827940464,
1910
+ "epoch": 0.23429781227946367,
1911
+ "geo": 0.8940749987959862,
1912
+ "grad_norm": 0.294921875,
1913
+ "learning_rate": 6.718592964824121e-06,
1914
+ "loss": 0.0001,
1915
+ "num_tokens": 78444526.0,
1916
+ "point_2d": 1.0,
1917
+ "point_3d": 0.939453125,
1918
+ "reward": 0.9376774355769157,
1919
+ "step": 664
1920
+ },
1921
+ {
1922
+ "3d_bbox": 0.8210256397724152,
1923
+ "adv": 0.000540634344361024,
1924
+ "bbox": 0.9576755836606026,
1925
+ "completions/clipped_ratio": 0.0,
1926
+ "completions/max_length": 1594.125,
1927
+ "completions/max_terminated_length": 1594.125,
1928
+ "completions/mean_length": 1573.80859375,
1929
+ "completions/mean_terminated_length": 1573.80859375,
1930
+ "completions/min_length": 1553.25,
1931
+ "completions/min_terminated_length": 1553.25,
1932
+ "entropy": 0.2863646815530956,
1933
+ "epoch": 0.23712067748764998,
1934
+ "geo": 0.9599328562617302,
1935
+ "grad_norm": 0.28125,
1936
+ "learning_rate": 6.678391959798996e-06,
1937
+ "loss": -0.0001,
1938
+ "num_tokens": 79388949.0,
1939
+ "point_2d": 1.0,
1940
+ "point_3d": 0.92578125,
1941
+ "reward": 0.9261206015944481,
1942
+ "step": 672
1943
+ },
1944
+ {
1945
+ "3d_bbox": 0.8162731677293777,
1946
+ "adv": -0.0006006620242260396,
1947
+ "bbox": 0.9639560282230377,
1948
+ "completions/clipped_ratio": 0.0,
1949
+ "completions/max_length": 1593.75,
1950
+ "completions/max_terminated_length": 1593.75,
1951
+ "completions/mean_length": 1573.62890625,
1952
+ "completions/mean_terminated_length": 1573.62890625,
1953
+ "completions/min_length": 1557.125,
1954
+ "completions/min_terminated_length": 1557.125,
1955
+ "entropy": 0.2868205299600959,
1956
+ "epoch": 0.2399435426958363,
1957
+ "geo": 0.9127583503723145,
1958
+ "grad_norm": 0.44921875,
1959
+ "learning_rate": 6.63819095477387e-06,
1960
+ "loss": -0.0,
1961
+ "num_tokens": 80333190.0,
1962
+ "point_2d": 1.0,
1963
+ "point_3d": 0.9267578125,
1964
+ "reward": 0.926746778190136,
1965
+ "step": 680
1966
+ },
1967
+ {
1968
+ "3d_bbox": 0.7580128684639931,
1969
+ "adv": 0.0009847661577850886,
1970
+ "bbox": 0.9946200400590897,
1971
+ "completions/clipped_ratio": 0.0,
1972
+ "completions/max_length": 1598.0,
1973
+ "completions/max_terminated_length": 1598.0,
1974
+ "completions/mean_length": 1572.2109375,
1975
+ "completions/mean_terminated_length": 1572.2109375,
1976
+ "completions/min_length": 1552.0,
1977
+ "completions/min_terminated_length": 1552.0,
1978
+ "entropy": 0.286160247400403,
1979
+ "epoch": 0.2427664079040226,
1980
+ "geo": 0.9191197454929352,
1981
+ "grad_norm": 0.361328125,
1982
+ "learning_rate": 6.597989949748744e-06,
1983
+ "loss": 0.0001,
1984
+ "num_tokens": 81277036.0,
1985
+ "point_2d": 1.0,
1986
+ "point_3d": 0.8818359375,
1987
+ "reward": 0.9086171984672546,
1988
+ "step": 688
1989
+ },
1990
+ {
1991
+ "3d_bbox": 0.729152224957943,
1992
+ "adv": -0.0009101523137360346,
1993
+ "bbox": 0.9927223324775696,
1994
+ "completions/clipped_ratio": 0.0,
1995
+ "completions/max_length": 1592.875,
1996
+ "completions/max_terminated_length": 1592.875,
1997
+ "completions/mean_length": 1571.8515625,
1998
+ "completions/mean_terminated_length": 1571.8515625,
1999
+ "completions/min_length": 1553.375,
2000
+ "completions/min_terminated_length": 1553.375,
2001
+ "entropy": 0.28340971609577537,
2002
+ "epoch": 0.2455892731122089,
2003
+ "geo": 0.9836970791220665,
2004
+ "grad_norm": 0.34765625,
2005
+ "learning_rate": 6.557788944723618e-06,
2006
+ "loss": 0.0005,
2007
+ "num_tokens": 82220822.0,
2008
+ "point_2d": 1.0,
2009
+ "point_3d": 0.9014246314764023,
2010
+ "reward": 0.9058247655630112,
2011
+ "step": 696
2012
+ },
2013
+ {
2014
+ "3d_bbox": 0.7259362824261189,
2015
+ "adv": 0.0003162266129947966,
2016
+ "bbox": 0.922937985509634,
2017
+ "completions/clipped_ratio": 0.0,
2018
+ "completions/max_length": 1596.0,
2019
+ "completions/max_terminated_length": 1596.0,
2020
+ "completions/mean_length": 1575.2265625,
2021
+ "completions/mean_terminated_length": 1575.2265625,
2022
+ "completions/min_length": 1555.625,
2023
+ "completions/min_terminated_length": 1555.625,
2024
+ "entropy": 0.286905855871737,
2025
+ "epoch": 0.24841213832039521,
2026
+ "geo": 0.8376269713044167,
2027
+ "grad_norm": 0.30078125,
2028
+ "learning_rate": 6.5175879396984935e-06,
2029
+ "loss": 0.0,
2030
+ "num_tokens": 83165600.0,
2031
+ "point_2d": 0.9541015625,
2032
+ "point_3d": 0.8759765625,
2033
+ "reward": 0.8697381019592285,
2034
+ "step": 704
2035
+ },
2036
+ {
2037
+ "3d_bbox": 0.7864346951246262,
2038
+ "adv": -0.0007161469466154813,
2039
+ "bbox": 0.9958559349179268,
2040
+ "completions/clipped_ratio": 0.0,
2041
+ "completions/max_length": 1595.0,
2042
+ "completions/max_terminated_length": 1595.0,
2043
+ "completions/mean_length": 1572.4921875,
2044
+ "completions/mean_terminated_length": 1572.4921875,
2045
+ "completions/min_length": 1545.125,
2046
+ "completions/min_terminated_length": 1545.125,
2047
+ "entropy": 0.2886675829067826,
2048
+ "epoch": 0.2512350035285815,
2049
+ "geo": 0.8416303154081106,
2050
+ "grad_norm": 0.333984375,
2051
+ "learning_rate": 6.477386934673368e-06,
2052
+ "loss": 0.0008,
2053
+ "num_tokens": 84109606.0,
2054
+ "point_2d": 0.75,
2055
+ "point_3d": 0.947265625,
2056
+ "reward": 0.8698890581727028,
2057
+ "step": 712
2058
+ },
2059
+ {
2060
+ "3d_bbox": 0.657585222274065,
2061
+ "adv": -0.0004964938416378573,
2062
+ "bbox": 0.9940767288208008,
2063
+ "completions/clipped_ratio": 0.0,
2064
+ "completions/max_length": 1589.75,
2065
+ "completions/max_terminated_length": 1589.75,
2066
+ "completions/mean_length": 1571.515625,
2067
+ "completions/mean_terminated_length": 1571.515625,
2068
+ "completions/min_length": 1553.5,
2069
+ "completions/min_terminated_length": 1553.5,
2070
+ "entropy": 0.2854380081407726,
2071
+ "epoch": 0.25405786873676783,
2072
+ "geo": 0.8985359221696854,
2073
+ "grad_norm": 0.275390625,
2074
+ "learning_rate": 6.437185929648242e-06,
2075
+ "loss": 0.0,
2076
+ "num_tokens": 85053346.0,
2077
+ "point_2d": 1.0,
2078
+ "point_3d": 0.8623046875,
2079
+ "reward": 0.8784916624426842,
2080
+ "step": 720
2081
+ },
2082
+ {
2083
+ "3d_bbox": 0.8299823924899101,
2084
+ "adv": -0.0006786447083868552,
2085
+ "bbox": 0.9957724660634995,
2086
+ "completions/clipped_ratio": 0.0,
2087
+ "completions/max_length": 1599.75,
2088
+ "completions/max_terminated_length": 1599.75,
2089
+ "completions/mean_length": 1575.05078125,
2090
+ "completions/mean_terminated_length": 1575.05078125,
2091
+ "completions/min_length": 1553.625,
2092
+ "completions/min_terminated_length": 1553.625,
2093
+ "entropy": 0.2834048862569034,
2094
+ "epoch": 0.25688073394495414,
2095
+ "geo": 0.7720311819575727,
2096
+ "grad_norm": 0.353515625,
2097
+ "learning_rate": 6.396984924623116e-06,
2098
+ "loss": 0.0008,
2099
+ "num_tokens": 85997743.0,
2100
+ "point_2d": 0.875,
2101
+ "point_3d": 0.8583984375,
2102
+ "reward": 0.8897883221507072,
2103
+ "step": 728
2104
+ },
2105
+ {
2106
+ "3d_bbox": 0.814798042178154,
2107
+ "adv": -0.00043147145015609567,
2108
+ "bbox": 0.994331993162632,
2109
+ "completions/clipped_ratio": 0.0,
2110
+ "completions/max_length": 1603.0,
2111
+ "completions/max_terminated_length": 1603.0,
2112
+ "completions/mean_length": 1578.44921875,
2113
+ "completions/mean_terminated_length": 1578.44921875,
2114
+ "completions/min_length": 1556.25,
2115
+ "completions/min_terminated_length": 1556.25,
2116
+ "entropy": 0.2801376706920564,
2117
+ "epoch": 0.25970359915314045,
2118
+ "geo": 0.9146922156214714,
2119
+ "grad_norm": 0.25,
2120
+ "learning_rate": 6.3567839195979905e-06,
2121
+ "loss": 0.0001,
2122
+ "num_tokens": 86943082.0,
2123
+ "point_2d": 0.8798828125,
2124
+ "point_3d": 0.94921875,
2125
+ "reward": 0.909557893872261,
2126
+ "step": 736
2127
+ },
2128
+ {
2129
+ "3d_bbox": 0.7457513771951199,
2130
+ "adv": -1.4886696476423822e-05,
2131
+ "bbox": 0.9952592924237251,
2132
+ "completions/clipped_ratio": 0.0,
2133
+ "completions/max_length": 1592.875,
2134
+ "completions/max_terminated_length": 1592.875,
2135
+ "completions/mean_length": 1572.21484375,
2136
+ "completions/mean_terminated_length": 1572.21484375,
2137
+ "completions/min_length": 1553.625,
2138
+ "completions/min_terminated_length": 1553.625,
2139
+ "entropy": 0.28585455380380154,
2140
+ "epoch": 0.26252646436132676,
2141
+ "geo": 0.9565036073327065,
2142
+ "grad_norm": 0.345703125,
2143
+ "learning_rate": 6.316582914572866e-06,
2144
+ "loss": -0.0003,
2145
+ "num_tokens": 87886913.0,
2146
+ "point_2d": 1.0,
2147
+ "point_3d": 0.8095703125,
2148
+ "reward": 0.8876452520489693,
2149
+ "step": 744
2150
+ },
2151
+ {
2152
+ "3d_bbox": 0.6948670484125614,
2153
+ "adv": -0.00016846299786266172,
2154
+ "bbox": 0.8553073853254318,
2155
+ "completions/clipped_ratio": 0.0,
2156
+ "completions/max_length": 1594.375,
2157
+ "completions/max_terminated_length": 1594.375,
2158
+ "completions/mean_length": 1571.93359375,
2159
+ "completions/mean_terminated_length": 1571.93359375,
2160
+ "completions/min_length": 1552.0,
2161
+ "completions/min_terminated_length": 1552.0,
2162
+ "entropy": 0.28335432009771466,
2163
+ "epoch": 0.26534932956951307,
2164
+ "geo": 0.7740265503525734,
2165
+ "grad_norm": 0.421875,
2166
+ "learning_rate": 6.27638190954774e-06,
2167
+ "loss": 0.0162,
2168
+ "num_tokens": 88830776.0,
2169
+ "point_2d": 0.8701171875,
2170
+ "point_3d": 0.9082605689764023,
2171
+ "reward": 0.8321380317211151,
2172
+ "step": 752
2173
+ },
2174
+ {
2175
+ "3d_bbox": 0.7499241754412651,
2176
+ "adv": 0.0011103415663455962,
2177
+ "bbox": 0.9951865822076797,
2178
+ "completions/clipped_ratio": 0.0,
2179
+ "completions/max_length": 1592.75,
2180
+ "completions/max_terminated_length": 1592.75,
2181
+ "completions/mean_length": 1570.93359375,
2182
+ "completions/mean_terminated_length": 1570.93359375,
2183
+ "completions/min_length": 1552.25,
2184
+ "completions/min_terminated_length": 1552.25,
2185
+ "entropy": 0.28731182869523764,
2186
+ "epoch": 0.2681721947776994,
2187
+ "geo": 0.620879240334034,
2188
+ "grad_norm": 0.2197265625,
2189
+ "learning_rate": 6.236180904522614e-06,
2190
+ "loss": 0.0001,
2191
+ "num_tokens": 89774255.0,
2192
+ "point_2d": 0.7119140625,
2193
+ "point_3d": 0.8916015625,
2194
+ "reward": 0.8371566012501717,
2195
+ "step": 760
2196
+ },
2197
+ {
2198
+ "3d_bbox": 0.7572767026722431,
2199
+ "adv": 0.0006080263119656593,
2200
+ "bbox": 0.9954995736479759,
2201
+ "completions/clipped_ratio": 0.0,
2202
+ "completions/max_length": 1589.0,
2203
+ "completions/max_terminated_length": 1589.0,
2204
+ "completions/mean_length": 1569.015625,
2205
+ "completions/mean_terminated_length": 1569.015625,
2206
+ "completions/min_length": 1549.75,
2207
+ "completions/min_terminated_length": 1549.75,
2208
+ "entropy": 0.2817346737720072,
2209
+ "epoch": 0.2709950599858857,
2210
+ "geo": 0.8461784943938255,
2211
+ "grad_norm": 0.41015625,
2212
+ "learning_rate": 6.195979899497488e-06,
2213
+ "loss": -0.0,
2214
+ "num_tokens": 90717243.0,
2215
+ "point_2d": 1.0,
2216
+ "point_3d": 0.892578125,
2217
+ "reward": 0.9113386049866676,
2218
+ "step": 768
2219
+ },
2220
+ {
2221
+ "3d_bbox": 0.8124237954616547,
2222
+ "adv": -0.00010392009426141158,
2223
+ "bbox": 0.9695260003209114,
2224
+ "completions/clipped_ratio": 0.0,
2225
+ "completions/max_length": 1590.0,
2226
+ "completions/max_terminated_length": 1590.0,
2227
+ "completions/mean_length": 1571.890625,
2228
+ "completions/mean_terminated_length": 1571.890625,
2229
+ "completions/min_length": 1553.75,
2230
+ "completions/min_terminated_length": 1553.75,
2231
+ "entropy": 0.2867649751715362,
2232
+ "epoch": 0.273817925194072,
2233
+ "geo": 0.7172370981425047,
2234
+ "grad_norm": 0.29296875,
2235
+ "learning_rate": 6.155778894472362e-06,
2236
+ "loss": -0.0,
2237
+ "num_tokens": 91660903.0,
2238
+ "point_2d": 0.75,
2239
+ "point_3d": 0.9462890625,
2240
+ "reward": 0.8695597127079964,
2241
+ "step": 776
2242
+ },
2243
+ {
2244
+ "3d_bbox": 0.6247975081205368,
2245
+ "adv": 0.00014501793702947907,
2246
+ "bbox": 0.9793029651045799,
2247
+ "completions/clipped_ratio": 0.0,
2248
+ "completions/max_length": 1592.125,
2249
+ "completions/max_terminated_length": 1592.125,
2250
+ "completions/mean_length": 1571.03515625,
2251
+ "completions/mean_terminated_length": 1571.03515625,
2252
+ "completions/min_length": 1550.875,
2253
+ "completions/min_terminated_length": 1550.875,
2254
+ "entropy": 0.2792881946079433,
2255
+ "epoch": 0.2766407904022583,
2256
+ "geo": 0.8185150865465403,
2257
+ "grad_norm": 0.2890625,
2258
+ "learning_rate": 6.115577889447236e-06,
2259
+ "loss": 0.0,
2260
+ "num_tokens": 92604352.0,
2261
+ "point_2d": 0.8876953125,
2262
+ "point_3d": 0.8994140625,
2263
+ "reward": 0.8478024452924728,
2264
+ "step": 784
2265
+ },
2266
+ {
2267
+ "3d_bbox": 0.7522557079792023,
2268
+ "adv": 0.00011037173771910602,
2269
+ "bbox": 0.8692253082990646,
2270
+ "completions/clipped_ratio": 0.0,
2271
+ "completions/max_length": 1594.625,
2272
+ "completions/max_terminated_length": 1594.625,
2273
+ "completions/mean_length": 1572.27734375,
2274
+ "completions/mean_terminated_length": 1572.27734375,
2275
+ "completions/min_length": 1552.5,
2276
+ "completions/min_terminated_length": 1552.5,
2277
+ "entropy": 0.2874673060141504,
2278
+ "epoch": 0.2794636556104446,
2279
+ "geo": 0.9408122524619102,
2280
+ "grad_norm": 0.31640625,
2281
+ "learning_rate": 6.07537688442211e-06,
2282
+ "loss": 0.0144,
2283
+ "num_tokens": 93548159.0,
2284
+ "point_2d": 0.7587890625,
2285
+ "point_3d": 0.91796875,
2286
+ "reward": 0.8245597034692764,
2287
+ "step": 792
2288
+ },
2289
+ {
2290
+ "3d_bbox": 0.8205922544002533,
2291
+ "adv": 0.0001977139909286052,
2292
+ "bbox": 0.9971112534403801,
2293
+ "completions/clipped_ratio": 0.0,
2294
+ "completions/max_length": 1589.25,
2295
+ "completions/max_terminated_length": 1589.25,
2296
+ "completions/mean_length": 1571.43359375,
2297
+ "completions/mean_terminated_length": 1571.43359375,
2298
+ "completions/min_length": 1552.625,
2299
+ "completions/min_terminated_length": 1552.625,
2300
+ "entropy": 0.28211602522060275,
2301
+ "epoch": 0.2822865208186309,
2302
+ "geo": 0.8445243984460831,
2303
+ "grad_norm": 0.3203125,
2304
+ "learning_rate": 6.035175879396985e-06,
2305
+ "loss": 0.0007,
2306
+ "num_tokens": 94491662.0,
2307
+ "point_2d": 0.884765625,
2308
+ "point_3d": 0.9365234375,
2309
+ "reward": 0.9097481444478035,
2310
+ "step": 800
2311
+ },
2312
+ {
2313
+ "3d_bbox": 0.794718824326992,
2314
+ "adv": -0.00023952151968842372,
2315
+ "bbox": 0.9944768622517586,
2316
+ "completions/clipped_ratio": 0.0,
2317
+ "completions/max_length": 1601.375,
2318
+ "completions/max_terminated_length": 1601.375,
2319
+ "completions/mean_length": 1571.67578125,
2320
+ "completions/mean_terminated_length": 1571.67578125,
2321
+ "completions/min_length": 1552.125,
2322
+ "completions/min_terminated_length": 1552.125,
2323
+ "entropy": 0.2849376816302538,
2324
+ "epoch": 0.28510938602681724,
2325
+ "geo": 0.9431384205818176,
2326
+ "grad_norm": 0.34375,
2327
+ "learning_rate": 5.99497487437186e-06,
2328
+ "loss": 0.0001,
2329
+ "num_tokens": 95435451.0,
2330
+ "point_2d": 0.9990234375,
2331
+ "point_3d": 0.8720703125,
2332
+ "reward": 0.9150723442435265,
2333
+ "step": 808
2334
+ },
2335
+ {
2336
+ "3d_bbox": 0.7696447446942329,
2337
+ "adv": 0.0008425721316598356,
2338
+ "bbox": 0.9485463574528694,
2339
+ "completions/clipped_ratio": 0.0,
2340
+ "completions/max_length": 1593.875,
2341
+ "completions/max_terminated_length": 1593.875,
2342
+ "completions/mean_length": 1570.11328125,
2343
+ "completions/mean_terminated_length": 1570.11328125,
2344
+ "completions/min_length": 1553.875,
2345
+ "completions/min_terminated_length": 1553.875,
2346
+ "entropy": 0.28682070411741734,
2347
+ "epoch": 0.28793225123500354,
2348
+ "geo": 0.9097637981176376,
2349
+ "grad_norm": 0.2353515625,
2350
+ "learning_rate": 5.954773869346734e-06,
2351
+ "loss": -0.0006,
2352
+ "num_tokens": 96378680.0,
2353
+ "point_2d": 0.9990234375,
2354
+ "point_3d": 0.9013671875,
2355
+ "reward": 0.9046454280614853,
2356
+ "step": 816
2357
+ },
2358
+ {
2359
+ "3d_bbox": 0.6897615268826485,
2360
+ "adv": -0.000497654465107189,
2361
+ "bbox": 0.8617971315979958,
2362
+ "completions/clipped_ratio": 0.0,
2363
+ "completions/max_length": 1602.125,
2364
+ "completions/max_terminated_length": 1602.125,
2365
+ "completions/mean_length": 1577.7734375,
2366
+ "completions/mean_terminated_length": 1577.7734375,
2367
+ "completions/min_length": 1558.125,
2368
+ "completions/min_terminated_length": 1558.125,
2369
+ "entropy": 0.28228796226903796,
2370
+ "epoch": 0.29075511644318985,
2371
+ "geo": 0.7427511867135763,
2372
+ "grad_norm": 0.326171875,
2373
+ "learning_rate": 5.914572864321608e-06,
2374
+ "loss": 0.0001,
2375
+ "num_tokens": 97324310.0,
2376
+ "point_2d": 0.7529296875,
2377
+ "point_3d": 0.919921875,
2378
+ "reward": 0.8061025515198708,
2379
+ "step": 824
2380
+ },
2381
+ {
2382
+ "3d_bbox": 0.6537492573261261,
2383
+ "adv": -0.00023956477116371389,
2384
+ "bbox": 0.9781710132956505,
2385
+ "completions/clipped_ratio": 0.0,
2386
+ "completions/max_length": 1596.0,
2387
+ "completions/max_terminated_length": 1596.0,
2388
+ "completions/mean_length": 1574.296875,
2389
+ "completions/mean_terminated_length": 1574.296875,
2390
+ "completions/min_length": 1554.75,
2391
+ "completions/min_terminated_length": 1554.75,
2392
+ "entropy": 0.28043507505208254,
2393
+ "epoch": 0.29357798165137616,
2394
+ "geo": 0.8402111828327179,
2395
+ "grad_norm": 0.291015625,
2396
+ "learning_rate": 5.874371859296482e-06,
2397
+ "loss": -0.0001,
2398
+ "num_tokens": 98268682.0,
2399
+ "point_2d": 1.0,
2400
+ "point_3d": 0.9551355689764023,
2401
+ "reward": 0.8967639729380608,
2402
+ "step": 832
2403
+ },
2404
+ {
2405
+ "3d_bbox": 0.755732849240303,
2406
+ "adv": -0.000661049871268915,
2407
+ "bbox": 0.9913836792111397,
2408
+ "completions/clipped_ratio": 0.0,
2409
+ "completions/max_length": 1598.125,
2410
+ "completions/max_terminated_length": 1598.125,
2411
+ "completions/mean_length": 1575.3125,
2412
+ "completions/mean_terminated_length": 1575.3125,
2413
+ "completions/min_length": 1555.625,
2414
+ "completions/min_terminated_length": 1555.625,
2415
+ "entropy": 0.2857117219828069,
2416
+ "epoch": 0.2964008468595625,
2417
+ "geo": 0.8524684980511665,
2418
+ "grad_norm": 0.380859375,
2419
+ "learning_rate": 5.8341708542713575e-06,
2420
+ "loss": 0.0143,
2421
+ "num_tokens": 99213386.0,
2422
+ "point_2d": 0.875,
2423
+ "point_3d": 0.876953125,
2424
+ "reward": 0.8747674003243446,
2425
+ "step": 840
2426
+ },
2427
+ {
2428
+ "3d_bbox": 0.7413047328591347,
2429
+ "adv": 0.0002993446832988411,
2430
+ "bbox": 0.8264872687868774,
2431
+ "completions/clipped_ratio": 0.0,
2432
+ "completions/max_length": 1594.125,
2433
+ "completions/max_terminated_length": 1594.125,
2434
+ "completions/mean_length": 1573.984375,
2435
+ "completions/mean_terminated_length": 1573.984375,
2436
+ "completions/min_length": 1556.75,
2437
+ "completions/min_terminated_length": 1556.75,
2438
+ "entropy": 0.28746917750686407,
2439
+ "epoch": 0.2992237120677488,
2440
+ "geo": 0.9443965330719948,
2441
+ "grad_norm": 0.3359375,
2442
+ "learning_rate": 5.793969849246232e-06,
2443
+ "loss": 0.0195,
2444
+ "num_tokens": 100157950.0,
2445
+ "point_2d": 1.0,
2446
+ "point_3d": 0.9140625,
2447
+ "reward": 0.8704636245965958,
2448
+ "step": 848
2449
+ },
2450
+ {
2451
+ "3d_bbox": 0.7137239277362823,
2452
+ "adv": -0.06107998272636905,
2453
+ "bbox": 0.9737529084086418,
2454
+ "completions/clipped_ratio": 0.0,
2455
+ "completions/max_length": 1593.25,
2456
+ "completions/max_terminated_length": 1593.25,
2457
+ "completions/mean_length": 1571.8046875,
2458
+ "completions/mean_terminated_length": 1571.8046875,
2459
+ "completions/min_length": 1553.5,
2460
+ "completions/min_terminated_length": 1553.5,
2461
+ "entropy": 0.2830418418161571,
2462
+ "epoch": 0.3020465772759351,
2463
+ "geo": 0.7549894053954631,
2464
+ "grad_norm": 0.3671875,
2465
+ "learning_rate": 5.753768844221106e-06,
2466
+ "loss": 0.0151,
2467
+ "num_tokens": 101101604.0,
2468
+ "point_2d": 0.7958984375,
2469
+ "point_3d": 0.865234375,
2470
+ "reward": 0.837152399122715,
2471
+ "step": 856
2472
+ },
2473
+ {
2474
+ "3d_bbox": 0.6744807064533234,
2475
+ "adv": 0.0013020644328207709,
2476
+ "bbox": 0.9808441251516342,
2477
+ "completions/clipped_ratio": 0.0,
2478
+ "completions/max_length": 1596.625,
2479
+ "completions/max_terminated_length": 1596.625,
2480
+ "completions/mean_length": 1574.37109375,
2481
+ "completions/mean_terminated_length": 1574.37109375,
2482
+ "completions/min_length": 1552.875,
2483
+ "completions/min_terminated_length": 1552.875,
2484
+ "entropy": 0.28136514313519,
2485
+ "epoch": 0.3048694424841214,
2486
+ "geo": 0.7662752494215965,
2487
+ "grad_norm": 0.333984375,
2488
+ "learning_rate": 5.71356783919598e-06,
2489
+ "loss": -0.0,
2490
+ "num_tokens": 102045955.0,
2491
+ "point_2d": 0.77734375,
2492
+ "point_3d": 0.91015625,
2493
+ "reward": 0.8357062116265297,
2494
+ "step": 864
2495
+ },
2496
+ {
2497
+ "3d_bbox": 0.7125367447733879,
2498
+ "adv": 0.0008483012829856307,
2499
+ "bbox": 0.8669970044866204,
2500
+ "completions/clipped_ratio": 0.0,
2501
+ "completions/max_length": 1593.375,
2502
+ "completions/max_terminated_length": 1593.375,
2503
+ "completions/mean_length": 1573.75390625,
2504
+ "completions/mean_terminated_length": 1573.75390625,
2505
+ "completions/min_length": 1556.5,
2506
+ "completions/min_terminated_length": 1556.5,
2507
+ "entropy": 0.2820960213430226,
2508
+ "epoch": 0.3076923076923077,
2509
+ "geo": 0.6582350172102451,
2510
+ "grad_norm": 0.296875,
2511
+ "learning_rate": 5.6733668341708545e-06,
2512
+ "loss": 0.0001,
2513
+ "num_tokens": 102990100.0,
2514
+ "point_2d": 0.5009765625,
2515
+ "point_3d": 0.896484375,
2516
+ "reward": 0.7442486621439457,
2517
+ "step": 872
2518
+ },
2519
+ {
2520
+ "3d_bbox": 0.8094568699598312,
2521
+ "adv": 0.000244778927481093,
2522
+ "bbox": 0.8707572477287613,
2523
+ "completions/clipped_ratio": 0.00390625,
2524
+ "completions/max_length": 1654.875,
2525
+ "completions/max_terminated_length": 1598.75,
2526
+ "completions/mean_length": 1576.1015625,
2527
+ "completions/mean_terminated_length": 1574.2484893798828,
2528
+ "completions/min_length": 1555.0,
2529
+ "completions/min_terminated_length": 1555.0,
2530
+ "entropy": 0.28441567649133503,
2531
+ "epoch": 0.310515172900494,
2532
+ "geo": 0.9321269914507866,
2533
+ "grad_norm": 0.5546875,
2534
+ "learning_rate": 5.633165829145729e-06,
2535
+ "loss": 0.026,
2536
+ "num_tokens": 103935062.0,
2537
+ "point_2d": 0.8828125,
2538
+ "point_3d": 0.9479817748069763,
2539
+ "reward": 0.8777521066367626,
2540
+ "step": 880
2541
+ },
2542
+ {
2543
+ "3d_bbox": 0.6922821514308453,
2544
+ "adv": -7.48095062590437e-05,
2545
+ "bbox": 0.9605405703186989,
2546
+ "completions/clipped_ratio": 0.0,
2547
+ "completions/max_length": 1596.25,
2548
+ "completions/max_terminated_length": 1596.25,
2549
+ "completions/mean_length": 1574.234375,
2550
+ "completions/mean_terminated_length": 1574.234375,
2551
+ "completions/min_length": 1555.875,
2552
+ "completions/min_terminated_length": 1555.875,
2553
+ "entropy": 0.28310656640678644,
2554
+ "epoch": 0.31333803810868033,
2555
+ "geo": 0.9695179983973503,
2556
+ "grad_norm": 0.283203125,
2557
+ "learning_rate": 5.592964824120604e-06,
2558
+ "loss": 0.0,
2559
+ "num_tokens": 104879434.0,
2560
+ "point_2d": 0.9921875,
2561
+ "point_3d": 0.921875,
2562
+ "reward": 0.8917213082313538,
2563
+ "step": 888
2564
+ },
2565
+ {
2566
+ "3d_bbox": 0.766367569565773,
2567
+ "adv": -0.0006123652386804679,
2568
+ "bbox": 0.9947258159518242,
2569
+ "completions/clipped_ratio": 0.0,
2570
+ "completions/max_length": 1594.0,
2571
+ "completions/max_terminated_length": 1594.0,
2572
+ "completions/mean_length": 1572.81640625,
2573
+ "completions/mean_terminated_length": 1572.81640625,
2574
+ "completions/min_length": 1554.125,
2575
+ "completions/min_terminated_length": 1554.125,
2576
+ "entropy": 0.2817820068448782,
2577
+ "epoch": 0.31616090331686664,
2578
+ "geo": 0.9325108900666237,
2579
+ "grad_norm": 0.359375,
2580
+ "learning_rate": 5.552763819095478e-06,
2581
+ "loss": 0.0001,
2582
+ "num_tokens": 105823315.0,
2583
+ "point_2d": 1.0,
2584
+ "point_3d": 0.8896484375,
2585
+ "reward": 0.9126854613423347,
2586
+ "step": 896
2587
+ },
2588
+ {
2589
+ "3d_bbox": 0.7490993067622185,
2590
+ "adv": -0.00017882781867228914,
2591
+ "bbox": 0.9934130385518074,
2592
+ "completions/clipped_ratio": 0.0,
2593
+ "completions/max_length": 1592.75,
2594
+ "completions/max_terminated_length": 1592.75,
2595
+ "completions/mean_length": 1572.3125,
2596
+ "completions/mean_terminated_length": 1572.3125,
2597
+ "completions/min_length": 1550.875,
2598
+ "completions/min_terminated_length": 1550.875,
2599
+ "entropy": 0.2811024447437376,
2600
+ "epoch": 0.31898376852505295,
2601
+ "geo": 0.8058975423919037,
2602
+ "grad_norm": 0.333984375,
2603
+ "learning_rate": 5.512562814070352e-06,
2604
+ "loss": 0.0001,
2605
+ "num_tokens": 106767163.0,
2606
+ "point_2d": 0.875,
2607
+ "point_3d": 0.880859375,
2608
+ "reward": 0.8745929300785065,
2609
+ "step": 904
2610
+ },
2611
+ {
2612
+ "3d_bbox": 0.690778411924839,
2613
+ "adv": -0.0004784437787748175,
2614
+ "bbox": 0.9014169201254845,
2615
+ "completions/clipped_ratio": 0.0,
2616
+ "completions/max_length": 1600.875,
2617
+ "completions/max_terminated_length": 1600.875,
2618
+ "completions/mean_length": 1574.25,
2619
+ "completions/mean_terminated_length": 1574.25,
2620
+ "completions/min_length": 1551.125,
2621
+ "completions/min_terminated_length": 1551.125,
2622
+ "entropy": 0.2825395744293928,
2623
+ "epoch": 0.32180663373323926,
2624
+ "geo": 0.654383197426796,
2625
+ "grad_norm": 0.373046875,
2626
+ "learning_rate": 5.472361809045227e-06,
2627
+ "loss": 0.0005,
2628
+ "num_tokens": 107711403.0,
2629
+ "point_2d": 0.7802734375,
2630
+ "point_3d": 0.9189453125,
2631
+ "reward": 0.8228535130620003,
2632
+ "step": 912
2633
+ },
2634
+ {
2635
+ "3d_bbox": 0.8111849427223206,
2636
+ "adv": -0.0003063662707063486,
2637
+ "bbox": 0.960642546415329,
2638
+ "completions/clipped_ratio": 0.0,
2639
+ "completions/max_length": 1596.125,
2640
+ "completions/max_terminated_length": 1596.125,
2641
+ "completions/mean_length": 1575.28125,
2642
+ "completions/mean_terminated_length": 1575.28125,
2643
+ "completions/min_length": 1555.75,
2644
+ "completions/min_terminated_length": 1555.75,
2645
+ "entropy": 0.28817785205319524,
2646
+ "epoch": 0.32462949894142556,
2647
+ "geo": 0.7242828160524368,
2648
+ "grad_norm": 0.40234375,
2649
+ "learning_rate": 5.432160804020101e-06,
2650
+ "loss": 0.0148,
2651
+ "num_tokens": 108655899.0,
2652
+ "point_2d": 0.75,
2653
+ "point_3d": 0.9453125,
2654
+ "reward": 0.866785004734993,
2655
+ "step": 920
2656
+ },
2657
+ {
2658
+ "3d_bbox": 0.80051339417696,
2659
+ "adv": -0.0008653635522932746,
2660
+ "bbox": 0.9517601355910301,
2661
+ "completions/clipped_ratio": 0.0,
2662
+ "completions/max_length": 1604.125,
2663
+ "completions/max_terminated_length": 1604.125,
2664
+ "completions/mean_length": 1580.0546875,
2665
+ "completions/mean_terminated_length": 1580.0546875,
2666
+ "completions/min_length": 1561.25,
2667
+ "completions/min_terminated_length": 1561.25,
2668
+ "entropy": 0.2851322675123811,
2669
+ "epoch": 0.3274523641496119,
2670
+ "geo": 0.9541479349136353,
2671
+ "grad_norm": 0.404296875,
2672
+ "learning_rate": 5.391959798994976e-06,
2673
+ "loss": 0.0006,
2674
+ "num_tokens": 109602009.0,
2675
+ "point_2d": 1.0,
2676
+ "point_3d": 0.9443359375,
2677
+ "reward": 0.9241523742675781,
2678
+ "step": 928
2679
+ },
2680
+ {
2681
+ "3d_bbox": 0.7602918669581413,
2682
+ "adv": 0.0012147135075792903,
2683
+ "bbox": 0.9494145661592484,
2684
+ "completions/clipped_ratio": 0.0,
2685
+ "completions/max_length": 1595.875,
2686
+ "completions/max_terminated_length": 1595.875,
2687
+ "completions/mean_length": 1576.296875,
2688
+ "completions/mean_terminated_length": 1576.296875,
2689
+ "completions/min_length": 1557.125,
2690
+ "completions/min_terminated_length": 1557.125,
2691
+ "entropy": 0.2826410192064941,
2692
+ "epoch": 0.3302752293577982,
2693
+ "geo": 0.8968632370233536,
2694
+ "grad_norm": 0.2734375,
2695
+ "learning_rate": 5.35175879396985e-06,
2696
+ "loss": -0.0001,
2697
+ "num_tokens": 110546973.0,
2698
+ "point_2d": 1.0,
2699
+ "point_3d": 0.9013671875,
2700
+ "reward": 0.9027684107422829,
2701
+ "step": 936
2702
+ },
2703
+ {
2704
+ "3d_bbox": 0.6780652180314064,
2705
+ "adv": -0.0008688528978382237,
2706
+ "bbox": 0.9930703118443489,
2707
+ "completions/clipped_ratio": 0.0,
2708
+ "completions/max_length": 1594.625,
2709
+ "completions/max_terminated_length": 1594.625,
2710
+ "completions/mean_length": 1574.8828125,
2711
+ "completions/mean_terminated_length": 1574.8828125,
2712
+ "completions/min_length": 1557.125,
2713
+ "completions/min_terminated_length": 1557.125,
2714
+ "entropy": 0.2831588592380285,
2715
+ "epoch": 0.3330980945659845,
2716
+ "geo": 0.9448152706027031,
2717
+ "grad_norm": 0.26953125,
2718
+ "learning_rate": 5.3115577889447245e-06,
2719
+ "loss": -0.0,
2720
+ "num_tokens": 111491407.0,
2721
+ "point_2d": 1.0,
2722
+ "point_3d": 0.8857421875,
2723
+ "reward": 0.8892194256186485,
2724
+ "step": 944
2725
+ }
2726
+ ],
2727
+ "logging_steps": 8,
2728
+ "max_steps": 2000,
2729
+ "num_input_tokens_seen": 112199787,
2730
+ "num_train_epochs": 1,
2731
+ "save_steps": 50,
2732
+ "stateful_callbacks": {
2733
+ "TrainerControl": {
2734
+ "args": {
2735
+ "should_epoch_stop": false,
2736
+ "should_evaluate": false,
2737
+ "should_log": false,
2738
+ "should_save": true,
2739
+ "should_training_stop": false
2740
+ },
2741
+ "attributes": {}
2742
+ }
2743
+ },
2744
+ "total_flos": 0.0,
2745
+ "train_batch_size": 1,
2746
+ "trial_name": null,
2747
+ "trial_params": null
2748
+ }
checkpoint-950/video_preprocessor_config.json ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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_pad": null,
10
+ "do_rescale": true,
11
+ "do_resize": true,
12
+ "do_sample_frames": false,
13
+ "fps": null,
14
+ "image_mean": [
15
+ 0.48145466,
16
+ 0.4578275,
17
+ 0.40821073
18
+ ],
19
+ "image_std": [
20
+ 0.26862954,
21
+ 0.26130258,
22
+ 0.27577711
23
+ ],
24
+ "input_data_format": null,
25
+ "max_frames": 768,
26
+ "max_pixels": 12845056,
27
+ "merge_size": 2,
28
+ "min_frames": 4,
29
+ "min_pixels": 3136,
30
+ "num_frames": 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
+ "size_divisor": null,
41
+ "temporal_patch_size": 2,
42
+ "video_metadata": null,
43
+ "video_processor_type": "Qwen2VLVideoProcessor"
44
+ }
checkpoint-950/vocab.json ADDED
The diff for this file is too large to render. See raw diff