Hugging Face's logo

furproxy
/
9b-31

Image-Text-to-Text
Transformers
Safetensors
qwen3_5
llama-factory
full
Generated from Trainer
conversational
Model card Files
xet
 Community
furproxy commited on
Commit
da081e5
·
verified ·
1 Parent(s): ffa99f4

Upload folder using huggingface_hub

Browse files
Files changed (16) hide show
  1. .gitattributes +1 -0
  2. .ipynb_checkpoints/README-checkpoint.md +71 -0
  3. README.md +71 -0
  4. all_results.json +9 -0
  5. chat_template.jinja +154 -0
  6. config.json +113 -0
  7. generation_config.json +10 -0
  8. model.safetensors +3 -0
  9. processor_config.json +60 -0
  10. tokenizer.json +3 -0
  11. tokenizer_config.json +33 -0
  12. train_results.json +9 -0
  13. trainer_log.jsonl +0 -0
  14. trainer_state.json +3641 -0
  15. training_args.bin +3 -0
  16. training_loss.png +0 -0
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ tokenizer.json filter=lfs diff=lfs merge=lfs -text
.ipynb_checkpoints/README-checkpoint.md ADDED
@@ -0,0 +1,71 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ license: other
4
+ base_model: Qwen3.5-9B
5
+ tags:
6
+ - llama-factory
7
+ - full
8
+ - generated_from_trainer
9
+ model-index:
10
+ - name: qwen35_caption_galore
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # qwen35_caption_galore
18
+
19
+ This model is a fine-tuned version of [/workspace/models/Qwen3.5-9B](https://huggingface.co//workspace/models/Qwen3.5-9B) on the my_caption dataset.
20
+
21
+ ## Model description
22
+
23
+ More information needed
24
+
25
+ ## Intended uses & limitations
26
+
27
+ More information needed
28
+
29
+ ## Training and evaluation data
30
+
31
+ More information needed
32
+
33
+ ## Training procedure
34
+
35
+ using more dense captions
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - family_to_muon_lr = {
41
+ "language": _fallback(getattr(training_args, "language_muon_lr", 1e-5), language_lr),
42
+ "vision": _fallback(getattr(training_args, "vision_muon_lr", 1e-5), vision_lr),
43
+ "merger": _fallback(getattr(training_args, "merger_muon_lr", 4e-5), merger_lr),
44
+ }
45
+
46
+ family_to_adamw_lr = {
47
+ "language": _fallback(getattr(training_args, "language_adamw_lr", 1e-5), language_lr),
48
+ "vision": _fallback(getattr(training_args, "vision_adamw_lr", 1e-6), vision_lr),
49
+ "merger": _fallback(getattr(training_args, "merger_adamw_lr", 1e-5), merger_lr),
50
+ }
51
+ - train_batch_size: 1
52
+ - eval_batch_size: 8
53
+ - seed: 42
54
+ - distributed_type: multi-GPU
55
+ - gradient_accumulation_steps: 32
56
+ - total_train_batch_size: 32
57
+ - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
58
+ - lr_scheduler_type: cosine_with_min_lr
59
+ - lr_scheduler_warmup_steps: 0.05
60
+ - num_epochs: 3
61
+
62
+ ### Training results
63
+
64
+
65
+
66
+ ### Framework versions
67
+
68
+ - Transformers 5.5.0
69
+ - Pytorch 2.11.0+cu128
70
+ - Datasets 4.0.0
71
+ - Tokenizers 0.22.2
README.md ADDED
@@ -0,0 +1,71 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: transformers
3
+ license: other
4
+ base_model: Qwen3.5-9B
5
+ tags:
6
+ - llama-factory
7
+ - full
8
+ - generated_from_trainer
9
+ model-index:
10
+ - name: qwen35_caption_galore
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # qwen35_caption_galore
18
+
19
+ This model is a fine-tuned version of [/workspace/models/Qwen3.5-9B](https://huggingface.co//workspace/models/Qwen3.5-9B) on the my_caption dataset.
20
+
21
+ ## Model description
22
+
23
+ More information needed
24
+
25
+ ## Intended uses & limitations
26
+
27
+ More information needed
28
+
29
+ ## Training and evaluation data
30
+
31
+ More information needed
32
+
33
+ ## Training procedure
34
+
35
+ using more dense captions
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - family_to_muon_lr = {
41
+ "language": _fallback(getattr(training_args, "language_muon_lr", 1e-5), language_lr),
42
+ "vision": _fallback(getattr(training_args, "vision_muon_lr", 1e-5), vision_lr),
43
+ "merger": _fallback(getattr(training_args, "merger_muon_lr", 4e-5), merger_lr),
44
+ }
45
+
46
+ family_to_adamw_lr = {
47
+ "language": _fallback(getattr(training_args, "language_adamw_lr", 1e-5), language_lr),
48
+ "vision": _fallback(getattr(training_args, "vision_adamw_lr", 1e-6), vision_lr),
49
+ "merger": _fallback(getattr(training_args, "merger_adamw_lr", 1e-5), merger_lr),
50
+ }
51
+ - train_batch_size: 1
52
+ - eval_batch_size: 8
53
+ - seed: 42
54
+ - distributed_type: multi-GPU
55
+ - gradient_accumulation_steps: 32
56
+ - total_train_batch_size: 32
57
+ - optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
58
+ - lr_scheduler_type: cosine_with_min_lr
59
+ - lr_scheduler_warmup_steps: 0.05
60
+ - num_epochs: 3
61
+
62
+ ### Training results
63
+
64
+
65
+
66
+ ### Framework versions
67
+
68
+ - Transformers 5.5.0
69
+ - Pytorch 2.11.0+cu128
70
+ - Datasets 4.0.0
71
+ - Tokenizers 0.22.2
all_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "effective_tokens_per_sec": 4364.043462600282,
3
+ "epoch": 3.0,
4
+ "total_flos": 2.909158325117518e+18,
5
+ "train_loss": 1.307906815331462,
6
+ "train_runtime": 13215.9511,
7
+ "train_samples_per_second": 2.488,
8
+ "train_steps_per_second": 0.078
9
+ }
chat_template.jinja ADDED
@@ -0,0 +1,154 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- set image_count = namespace(value=0) %}
2
+ {%- set video_count = namespace(value=0) %}
3
+ {%- macro render_content(content, do_vision_count, is_system_content=false) %}
4
+ {%- if content is string %}
5
+ {{- content }}
6
+ {%- elif content is iterable and content is not mapping %}
7
+ {%- for item in content %}
8
+ {%- if 'image' in item or 'image_url' in item or item.type == 'image' %}
9
+ {%- if is_system_content %}
10
+ {{- raise_exception('System message cannot contain images.') }}
11
+ {%- endif %}
12
+ {%- if do_vision_count %}
13
+ {%- set image_count.value = image_count.value + 1 %}
14
+ {%- endif %}
15
+ {%- if add_vision_id %}
16
+ {{- 'Picture ' ~ image_count.value ~ ': ' }}
17
+ {%- endif %}
18
+ {{- '<|vision_start|><|image_pad|><|vision_end|>' }}
19
+ {%- elif 'video' in item or item.type == 'video' %}
20
+ {%- if is_system_content %}
21
+ {{- raise_exception('System message cannot contain videos.') }}
22
+ {%- endif %}
23
+ {%- if do_vision_count %}
24
+ {%- set video_count.value = video_count.value + 1 %}
25
+ {%- endif %}
26
+ {%- if add_vision_id %}
27
+ {{- 'Video ' ~ video_count.value ~ ': ' }}
28
+ {%- endif %}
29
+ {{- '<|vision_start|><|video_pad|><|vision_end|>' }}
30
+ {%- elif 'text' in item %}
31
+ {{- item.text }}
32
+ {%- else %}
33
+ {{- raise_exception('Unexpected item type in content.') }}
34
+ {%- endif %}
35
+ {%- endfor %}
36
+ {%- elif content is none or content is undefined %}
37
+ {{- '' }}
38
+ {%- else %}
39
+ {{- raise_exception('Unexpected content type.') }}
40
+ {%- endif %}
41
+ {%- endmacro %}
42
+ {%- if not messages %}
43
+ {{- raise_exception('No messages provided.') }}
44
+ {%- endif %}
45
+ {%- if tools and tools is iterable and tools is not mapping %}
46
+ {{- '<|im_start|>system\n' }}
47
+ {{- "# Tools\n\nYou have access to the following functions:\n\n<tools>" }}
48
+ {%- for tool in tools %}
49
+ {{- "\n" }}
50
+ {{- tool | tojson }}
51
+ {%- endfor %}
52
+ {{- "\n</tools>" }}
53
+ {{- '\n\nIf you choose to call a function ONLY reply in the following format with NO suffix:\n\n<tool_call>\n<function=example_function_name>\n<parameter=example_parameter_1>\nvalue_1\n</parameter>\n<parameter=example_parameter_2>\nThis is the value for the second parameter\nthat can span\nmultiple lines\n</parameter>\n</function>\n</tool_call>\n\n<IMPORTANT>\nReminder:\n- Function calls MUST follow the specified format: an inner <function=...></function> block must be nested within <tool_call></tool_call> XML tags\n- Required parameters MUST be specified\n- You may provide optional reasoning for your function call in natural language BEFORE the function call, but NOT after\n- If there is no function call available, answer the question like normal with your current knowledge and do not tell the user about function calls\n</IMPORTANT>' }}
54
+ {%- if messages[0].role == 'system' %}
55
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
56
+ {%- if content %}
57
+ {{- '\n\n' + content }}
58
+ {%- endif %}
59
+ {%- endif %}
60
+ {{- '<|im_end|>\n' }}
61
+ {%- else %}
62
+ {%- if messages[0].role == 'system' %}
63
+ {%- set content = render_content(messages[0].content, false, true)|trim %}
64
+ {{- '<|im_start|>system\n' + content + '<|im_end|>\n' }}
65
+ {%- endif %}
66
+ {%- endif %}
67
+ {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
68
+ {%- for message in messages[::-1] %}
69
+ {%- set index = (messages|length - 1) - loop.index0 %}
70
+ {%- if ns.multi_step_tool and message.role == "user" %}
71
+ {%- set content = render_content(message.content, false)|trim %}
72
+ {%- if not(content.startswith('<tool_response>') and content.endswith('</tool_response>')) %}
73
+ {%- set ns.multi_step_tool = false %}
74
+ {%- set ns.last_query_index = index %}
75
+ {%- endif %}
76
+ {%- endif %}
77
+ {%- endfor %}
78
+ {%- if ns.multi_step_tool %}
79
+ {{- raise_exception('No user query found in messages.') }}
80
+ {%- endif %}
81
+ {%- for message in messages %}
82
+ {%- set content = render_content(message.content, true)|trim %}
83
+ {%- if message.role == "system" %}
84
+ {%- if not loop.first %}
85
+ {{- raise_exception('System message must be at the beginning.') }}
86
+ {%- endif %}
87
+ {%- elif message.role == "user" %}
88
+ {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
89
+ {%- elif message.role == "assistant" %}
90
+ {%- set reasoning_content = '' %}
91
+ {%- if message.reasoning_content is string %}
92
+ {%- set reasoning_content = message.reasoning_content %}
93
+ {%- else %}
94
+ {%- if '</think>' in content %}
95
+ {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
96
+ {%- set content = content.split('</think>')[-1].lstrip('\n') %}
97
+ {%- endif %}
98
+ {%- endif %}
99
+ {%- set reasoning_content = reasoning_content|trim %}
100
+ {%- if loop.index0 > ns.last_query_index %}
101
+ {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content + '\n</think>\n\n' + content }}
102
+ {%- else %}
103
+ {{- '<|im_start|>' + message.role + '\n' + content }}
104
+ {%- endif %}
105
+ {%- if message.tool_calls and message.tool_calls is iterable and message.tool_calls is not mapping %}
106
+ {%- for tool_call in message.tool_calls %}
107
+ {%- if tool_call.function is defined %}
108
+ {%- set tool_call = tool_call.function %}
109
+ {%- endif %}
110
+ {%- if loop.first %}
111
+ {%- if content|trim %}
112
+ {{- '\n\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
113
+ {%- else %}
114
+ {{- '<tool_call>\n<function=' + tool_call.name + '>\n' }}
115
+ {%- endif %}
116
+ {%- else %}
117
+ {{- '\n<tool_call>\n<function=' + tool_call.name + '>\n' }}
118
+ {%- endif %}
119
+ {%- if tool_call.arguments is defined %}
120
+ {%- for args_name, args_value in tool_call.arguments|items %}
121
+ {{- '<parameter=' + args_name + '>\n' }}
122
+ {%- set args_value = args_value | tojson | safe if args_value is mapping or (args_value is sequence and args_value is not string) else args_value | string %}
123
+ {{- args_value }}
124
+ {{- '\n</parameter>\n' }}
125
+ {%- endfor %}
126
+ {%- endif %}
127
+ {{- '</function>\n</tool_call>' }}
128
+ {%- endfor %}
129
+ {%- endif %}
130
+ {{- '<|im_end|>\n' }}
131
+ {%- elif message.role == "tool" %}
132
+ {%- if loop.previtem and loop.previtem.role != "tool" %}
133
+ {{- '<|im_start|>user' }}
134
+ {%- endif %}
135
+ {{- '\n<tool_response>\n' }}
136
+ {{- content }}
137
+ {{- '\n</tool_response>' }}
138
+ {%- if not loop.last and loop.nextitem.role != "tool" %}
139
+ {{- '<|im_end|>\n' }}
140
+ {%- elif loop.last %}
141
+ {{- '<|im_end|>\n' }}
142
+ {%- endif %}
143
+ {%- else %}
144
+ {{- raise_exception('Unexpected message role.') }}
145
+ {%- endif %}
146
+ {%- endfor %}
147
+ {%- if add_generation_prompt %}
148
+ {{- '<|im_start|>assistant\n' }}
149
+ {%- if enable_thinking is defined and enable_thinking is false %}
150
+ {{- '<think>\n\n</think>\n\n' }}
151
+ {%- else %}
152
+ {{- '<think>\n' }}
153
+ {%- endif %}
154
+ {%- endif %}
config.json ADDED
@@ -0,0 +1,113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "Qwen3_5ForConditionalGeneration"
4
+ ],
5
+ "dtype": "bfloat16",
6
+ "eos_token_id": 248046,
7
+ "hidden_size": 4096,
8
+ "image_token_id": 248056,
9
+ "model_type": "qwen3_5",
10
+ "pad_token_id": 248044,
11
+ "text_config": {
12
+ "attention_bias": false,
13
+ "attention_dropout": 0.0,
14
+ "attn_output_gate": true,
15
+ "bos_token_id": null,
16
+ "dtype": "bfloat16",
17
+ "eos_token_id": 248044,
18
+ "full_attention_interval": 4,
19
+ "head_dim": 256,
20
+ "hidden_act": "silu",
21
+ "hidden_size": 4096,
22
+ "initializer_range": 0.02,
23
+ "intermediate_size": 12288,
24
+ "layer_types": [
25
+ "linear_attention",
26
+ "linear_attention",
27
+ "linear_attention",
28
+ "full_attention",
29
+ "linear_attention",
30
+ "linear_attention",
31
+ "linear_attention",
32
+ "full_attention",
33
+ "linear_attention",
34
+ "linear_attention",
35
+ "linear_attention",
36
+ "full_attention",
37
+ "linear_attention",
38
+ "linear_attention",
39
+ "linear_attention",
40
+ "full_attention",
41
+ "linear_attention",
42
+ "linear_attention",
43
+ "linear_attention",
44
+ "full_attention",
45
+ "linear_attention",
46
+ "linear_attention",
47
+ "linear_attention",
48
+ "full_attention",
49
+ "linear_attention",
50
+ "linear_attention",
51
+ "linear_attention",
52
+ "full_attention",
53
+ "linear_attention",
54
+ "linear_attention",
55
+ "linear_attention",
56
+ "full_attention"
57
+ ],
58
+ "linear_conv_kernel_dim": 4,
59
+ "linear_key_head_dim": 128,
60
+ "linear_num_key_heads": 16,
61
+ "linear_num_value_heads": 32,
62
+ "linear_value_head_dim": 128,
63
+ "mamba_ssm_dtype": "float32",
64
+ "max_position_embeddings": 262144,
65
+ "mlp_only_layers": [],
66
+ "model_type": "qwen3_5_text",
67
+ "mtp_num_hidden_layers": 1,
68
+ "mtp_use_dedicated_embeddings": false,
69
+ "num_attention_heads": 16,
70
+ "num_hidden_layers": 32,
71
+ "num_key_value_heads": 4,
72
+ "pad_token_id": null,
73
+ "partial_rotary_factor": 0.25,
74
+ "rms_norm_eps": 1e-06,
75
+ "rope_parameters": {
76
+ "mrope_interleaved": true,
77
+ "mrope_section": [
78
+ 11,
79
+ 11,
80
+ 10
81
+ ],
82
+ "partial_rotary_factor": 0.25,
83
+ "rope_theta": 10000000,
84
+ "rope_type": "default"
85
+ },
86
+ "tie_word_embeddings": false,
87
+ "use_cache": false,
88
+ "vocab_size": 248320
89
+ },
90
+ "tie_word_embeddings": false,
91
+ "transformers_version": "5.5.0",
92
+ "use_cache": false,
93
+ "video_token_id": 248057,
94
+ "vision_config": {
95
+ "deepstack_visual_indexes": [],
96
+ "depth": 27,
97
+ "dtype": "bfloat16",
98
+ "hidden_act": "gelu_pytorch_tanh",
99
+ "hidden_size": 1152,
100
+ "in_channels": 3,
101
+ "initializer_range": 0.02,
102
+ "intermediate_size": 4304,
103
+ "model_type": "qwen3_5",
104
+ "num_heads": 16,
105
+ "num_position_embeddings": 2304,
106
+ "out_hidden_size": 4096,
107
+ "patch_size": 16,
108
+ "spatial_merge_size": 2,
109
+ "temporal_patch_size": 2
110
+ },
111
+ "vision_end_token_id": 248054,
112
+ "vision_start_token_id": 248053
113
+ }
generation_config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "eos_token_id": [
4
+ 248046,
5
+ 248044
6
+ ],
7
+ "pad_token_id": 248044,
8
+ "transformers_version": "5.5.0",
9
+ "use_cache": true
10
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bf5a2a9d008856cdc321c58c8261bf051ba4ad0c545b3249b17174c9c78ef2d3
3
+ size 18819740168
processor_config.json ADDED
@@ -0,0 +1,60 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "image_processor": {
3
+ "do_convert_rgb": true,
4
+ "do_normalize": true,
5
+ "do_rescale": true,
6
+ "do_resize": true,
7
+ "image_mean": [
8
+ 0.5,
9
+ 0.5,
10
+ 0.5
11
+ ],
12
+ "image_processor_type": "Qwen2VLImageProcessor",
13
+ "image_std": [
14
+ 0.5,
15
+ 0.5,
16
+ 0.5
17
+ ],
18
+ "merge_size": 2,
19
+ "patch_size": 16,
20
+ "resample": 3,
21
+ "rescale_factor": 0.00392156862745098,
22
+ "size": {
23
+ "longest_edge": 16777216,
24
+ "shortest_edge": 65536
25
+ },
26
+ "temporal_patch_size": 2
27
+ },
28
+ "processor_class": "Qwen3VLProcessor",
29
+ "video_processor": {
30
+ "do_convert_rgb": true,
31
+ "do_normalize": true,
32
+ "do_rescale": true,
33
+ "do_resize": true,
34
+ "do_sample_frames": true,
35
+ "fps": 2,
36
+ "image_mean": [
37
+ 0.5,
38
+ 0.5,
39
+ 0.5
40
+ ],
41
+ "image_std": [
42
+ 0.5,
43
+ 0.5,
44
+ 0.5
45
+ ],
46
+ "max_frames": 768,
47
+ "merge_size": 2,
48
+ "min_frames": 4,
49
+ "patch_size": 16,
50
+ "resample": 3,
51
+ "rescale_factor": 0.00392156862745098,
52
+ "return_metadata": false,
53
+ "size": {
54
+ "longest_edge": 25165824,
55
+ "shortest_edge": 4096
56
+ },
57
+ "temporal_patch_size": 2,
58
+ "video_processor_type": "Qwen3VLVideoProcessor"
59
+ }
60
+ }
tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:87a7830d63fcf43bf241c3c5242e96e62dd3fdc29224ca26fed8ea333db72de4
3
+ size 19989343
tokenizer_config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "audio_bos_token": "<|audio_start|>",
4
+ "audio_eos_token": "<|audio_end|>",
5
+ "audio_token": "<|audio_pad|>",
6
+ "backend": "tokenizers",
7
+ "bos_token": null,
8
+ "clean_up_tokenization_spaces": false,
9
+ "eos_token": "<|im_end|>",
10
+ "errors": "replace",
11
+ "image_token": "<|image_pad|>",
12
+ "is_local": true,
13
+ "model_max_length": 262144,
14
+ "model_specific_special_tokens": {
15
+ "audio_bos_token": "<|audio_start|>",
16
+ "audio_eos_token": "<|audio_end|>",
17
+ "audio_token": "<|audio_pad|>",
18
+ "image_token": "<|image_pad|>",
19
+ "video_token": "<|video_pad|>",
20
+ "vision_bos_token": "<|vision_start|>",
21
+ "vision_eos_token": "<|vision_end|>"
22
+ },
23
+ "pad_token": "<|endoftext|>",
24
+ "padding_side": "right",
25
+ "pretokenize_regex": "(?i:'s|'t|'re|'ve|'m|'ll|'d)|[^\\r\\n\\p{L}\\p{N}]?[\\p{L}\\p{M}]+|\\p{N}| ?[^\\s\\p{L}\\p{M}\\p{N}]+[\\r\\n]*|\\s*[\\r\\n]+|\\s+(?!\\S)|\\s+",
26
+ "processor_class": "Qwen3VLProcessor",
27
+ "split_special_tokens": false,
28
+ "tokenizer_class": "TokenizersBackend",
29
+ "unk_token": null,
30
+ "video_token": "<|video_pad|>",
31
+ "vision_bos_token": "<|vision_start|>",
32
+ "vision_eos_token": "<|vision_end|>"
33
+ }
train_results.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "effective_tokens_per_sec": 4364.043462600282,
3
+ "epoch": 3.0,
4
+ "total_flos": 2.909158325117518e+18,
5
+ "train_loss": 1.307906815331462,
6
+ "train_runtime": 13215.9511,
7
+ "train_samples_per_second": 2.488,
8
+ "train_steps_per_second": 0.078
9
+ }
trainer_log.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
trainer_state.json ADDED
@@ -0,0 +1,3641 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 3.0,
6
+ "eval_steps": 500,
7
+ "global_step": 1029,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.00583941605839416,
14
+ "grad_norm": 1.9921875,
15
+ "learning_rate": 1.9230769230769234e-07,
16
+ "loss": 2.052279233932495,
17
+ "step": 2
18
+ },
19
+ {
20
+ "epoch": 0.01167883211678832,
21
+ "grad_norm": 0.71875,
22
+ "learning_rate": 5.76923076923077e-07,
23
+ "loss": 1.9450315237045288,
24
+ "step": 4
25
+ },
26
+ {
27
+ "epoch": 0.017518248175182483,
28
+ "grad_norm": 0.97265625,
29
+ "learning_rate": 9.615384615384617e-07,
30
+ "loss": 2.0441408157348633,
31
+ "step": 6
32
+ },
33
+ {
34
+ "epoch": 0.02335766423357664,
35
+ "grad_norm": 1.390625,
36
+ "learning_rate": 1.3461538461538462e-06,
37
+ "loss": 1.9473555088043213,
38
+ "step": 8
39
+ },
40
+ {
41
+ "epoch": 0.029197080291970802,
42
+ "grad_norm": 0.6875,
43
+ "learning_rate": 1.7307692307692308e-06,
44
+ "loss": 1.9234904050827026,
45
+ "step": 10
46
+ },
47
+ {
48
+ "epoch": 0.035036496350364967,
49
+ "grad_norm": 1.234375,
50
+ "learning_rate": 2.1153846153846155e-06,
51
+ "loss": 2.0447089672088623,
52
+ "step": 12
53
+ },
54
+ {
55
+ "epoch": 0.040875912408759124,
56
+ "grad_norm": 1.171875,
57
+ "learning_rate": 2.5e-06,
58
+ "loss": 1.8234096765518188,
59
+ "step": 14
60
+ },
61
+ {
62
+ "epoch": 0.04671532846715328,
63
+ "grad_norm": 0.51953125,
64
+ "learning_rate": 2.8846153846153845e-06,
65
+ "loss": 1.8641804456710815,
66
+ "step": 16
67
+ },
68
+ {
69
+ "epoch": 0.052554744525547446,
70
+ "grad_norm": 0.6640625,
71
+ "learning_rate": 3.2692307692307696e-06,
72
+ "loss": 1.9037718772888184,
73
+ "step": 18
74
+ },
75
+ {
76
+ "epoch": 0.058394160583941604,
77
+ "grad_norm": 0.7734375,
78
+ "learning_rate": 3.653846153846154e-06,
79
+ "loss": 1.9216861724853516,
80
+ "step": 20
81
+ },
82
+ {
83
+ "epoch": 0.06423357664233577,
84
+ "grad_norm": 1.3125,
85
+ "learning_rate": 4.0384615384615385e-06,
86
+ "loss": 1.9886118173599243,
87
+ "step": 22
88
+ },
89
+ {
90
+ "epoch": 0.07007299270072993,
91
+ "grad_norm": 1.015625,
92
+ "learning_rate": 4.423076923076924e-06,
93
+ "loss": 1.7973345518112183,
94
+ "step": 24
95
+ },
96
+ {
97
+ "epoch": 0.07591240875912408,
98
+ "grad_norm": 0.53515625,
99
+ "learning_rate": 4.807692307692308e-06,
100
+ "loss": 1.8787648677825928,
101
+ "step": 26
102
+ },
103
+ {
104
+ "epoch": 0.08175182481751825,
105
+ "grad_norm": 0.765625,
106
+ "learning_rate": 5.192307692307693e-06,
107
+ "loss": 1.8880407810211182,
108
+ "step": 28
109
+ },
110
+ {
111
+ "epoch": 0.08759124087591241,
112
+ "grad_norm": 0.5,
113
+ "learning_rate": 5.576923076923077e-06,
114
+ "loss": 1.8372446298599243,
115
+ "step": 30
116
+ },
117
+ {
118
+ "epoch": 0.09343065693430656,
119
+ "grad_norm": 0.65625,
120
+ "learning_rate": 5.961538461538462e-06,
121
+ "loss": 1.8118259906768799,
122
+ "step": 32
123
+ },
124
+ {
125
+ "epoch": 0.09927007299270073,
126
+ "grad_norm": 1.09375,
127
+ "learning_rate": 6.3461538461538466e-06,
128
+ "loss": 1.872291922569275,
129
+ "step": 34
130
+ },
131
+ {
132
+ "epoch": 0.10510948905109489,
133
+ "grad_norm": 1.125,
134
+ "learning_rate": 6.730769230769232e-06,
135
+ "loss": 1.9201974868774414,
136
+ "step": 36
137
+ },
138
+ {
139
+ "epoch": 0.11094890510948906,
140
+ "grad_norm": 0.6875,
141
+ "learning_rate": 7.115384615384616e-06,
142
+ "loss": 1.821286678314209,
143
+ "step": 38
144
+ },
145
+ {
146
+ "epoch": 0.11678832116788321,
147
+ "grad_norm": 0.361328125,
148
+ "learning_rate": 7.500000000000001e-06,
149
+ "loss": 1.6810513734817505,
150
+ "step": 40
151
+ },
152
+ {
153
+ "epoch": 0.12262773722627737,
154
+ "grad_norm": 0.50390625,
155
+ "learning_rate": 7.884615384615384e-06,
156
+ "loss": 1.7490516901016235,
157
+ "step": 42
158
+ },
159
+ {
160
+ "epoch": 0.12846715328467154,
161
+ "grad_norm": 0.400390625,
162
+ "learning_rate": 8.26923076923077e-06,
163
+ "loss": 1.6871154308319092,
164
+ "step": 44
165
+ },
166
+ {
167
+ "epoch": 0.1343065693430657,
168
+ "grad_norm": 0.9765625,
169
+ "learning_rate": 8.653846153846155e-06,
170
+ "loss": 1.8664501905441284,
171
+ "step": 46
172
+ },
173
+ {
174
+ "epoch": 0.14014598540145987,
175
+ "grad_norm": 0.57421875,
176
+ "learning_rate": 9.03846153846154e-06,
177
+ "loss": 1.76151442527771,
178
+ "step": 48
179
+ },
180
+ {
181
+ "epoch": 0.145985401459854,
182
+ "grad_norm": 0.6953125,
183
+ "learning_rate": 9.423076923076923e-06,
184
+ "loss": 1.7157841920852661,
185
+ "step": 50
186
+ },
187
+ {
188
+ "epoch": 0.15182481751824817,
189
+ "grad_norm": 0.474609375,
190
+ "learning_rate": 9.807692307692308e-06,
191
+ "loss": 1.7865409851074219,
192
+ "step": 52
193
+ },
194
+ {
195
+ "epoch": 0.15766423357664233,
196
+ "grad_norm": 0.43359375,
197
+ "learning_rate": 9.999976735551495e-06,
198
+ "loss": 1.7557110786437988,
199
+ "step": 54
200
+ },
201
+ {
202
+ "epoch": 0.1635036496350365,
203
+ "grad_norm": 0.359375,
204
+ "learning_rate": 9.999790621406734e-06,
205
+ "loss": 1.6408637762069702,
206
+ "step": 56
207
+ },
208
+ {
209
+ "epoch": 0.16934306569343066,
210
+ "grad_norm": 0.62890625,
211
+ "learning_rate": 9.999418400814693e-06,
212
+ "loss": 1.7107805013656616,
213
+ "step": 58
214
+ },
215
+ {
216
+ "epoch": 0.17518248175182483,
217
+ "grad_norm": 11.9375,
218
+ "learning_rate": 9.998860089170008e-06,
219
+ "loss": 1.6930475234985352,
220
+ "step": 60
221
+ },
222
+ {
223
+ "epoch": 0.181021897810219,
224
+ "grad_norm": 0.90625,
225
+ "learning_rate": 9.998115709563843e-06,
226
+ "loss": 1.8085753917694092,
227
+ "step": 62
228
+ },
229
+ {
230
+ "epoch": 0.18686131386861313,
231
+ "grad_norm": 0.54296875,
232
+ "learning_rate": 9.997185292782932e-06,
233
+ "loss": 1.6834930181503296,
234
+ "step": 64
235
+ },
236
+ {
237
+ "epoch": 0.1927007299270073,
238
+ "grad_norm": 0.80859375,
239
+ "learning_rate": 9.996068877308297e-06,
240
+ "loss": 1.5944346189498901,
241
+ "step": 66
242
+ },
243
+ {
244
+ "epoch": 0.19854014598540146,
245
+ "grad_norm": 0.80859375,
246
+ "learning_rate": 9.994766509313667e-06,
247
+ "loss": 1.7082901000976562,
248
+ "step": 68
249
+ },
250
+ {
251
+ "epoch": 0.20437956204379562,
252
+ "grad_norm": 0.3671875,
253
+ "learning_rate": 9.993278242663572e-06,
254
+ "loss": 1.6360015869140625,
255
+ "step": 70
256
+ },
257
+ {
258
+ "epoch": 0.21021897810218979,
259
+ "grad_norm": 0.419921875,
260
+ "learning_rate": 9.9916041389111e-06,
261
+ "loss": 1.7114381790161133,
262
+ "step": 72
263
+ },
264
+ {
265
+ "epoch": 0.21605839416058395,
266
+ "grad_norm": 0.4296875,
267
+ "learning_rate": 9.989744267295359e-06,
268
+ "loss": 1.5103245973587036,
269
+ "step": 74
270
+ },
271
+ {
272
+ "epoch": 0.22189781021897811,
273
+ "grad_norm": 0.46875,
274
+ "learning_rate": 9.987698704738617e-06,
275
+ "loss": 1.4727965593338013,
276
+ "step": 76
277
+ },
278
+ {
279
+ "epoch": 0.22773722627737225,
280
+ "grad_norm": 0.71875,
281
+ "learning_rate": 9.985467535843121e-06,
282
+ "loss": 1.6391849517822266,
283
+ "step": 78
284
+ },
285
+ {
286
+ "epoch": 0.23357664233576642,
287
+ "grad_norm": 3.625,
288
+ "learning_rate": 9.983050852887587e-06,
289
+ "loss": 1.7767753601074219,
290
+ "step": 80
291
+ },
292
+ {
293
+ "epoch": 0.23941605839416058,
294
+ "grad_norm": 0.57421875,
295
+ "learning_rate": 9.9804487558234e-06,
296
+ "loss": 1.6842796802520752,
297
+ "step": 82
298
+ },
299
+ {
300
+ "epoch": 0.24525547445255474,
301
+ "grad_norm": 0.8515625,
302
+ "learning_rate": 9.977661352270457e-06,
303
+ "loss": 1.6517852544784546,
304
+ "step": 84
305
+ },
306
+ {
307
+ "epoch": 0.2510948905109489,
308
+ "grad_norm": 0.72265625,
309
+ "learning_rate": 9.974688757512742e-06,
310
+ "loss": 1.501247763633728,
311
+ "step": 86
312
+ },
313
+ {
314
+ "epoch": 0.2569343065693431,
315
+ "grad_norm": 0.462890625,
316
+ "learning_rate": 9.971531094493541e-06,
317
+ "loss": 1.5306718349456787,
318
+ "step": 88
319
+ },
320
+ {
321
+ "epoch": 0.26277372262773724,
322
+ "grad_norm": 0.3671875,
323
+ "learning_rate": 9.968188493810365e-06,
324
+ "loss": 1.6086313724517822,
325
+ "step": 90
326
+ },
327
+ {
328
+ "epoch": 0.2686131386861314,
329
+ "grad_norm": 0.59375,
330
+ "learning_rate": 9.964661093709545e-06,
331
+ "loss": 1.5198537111282349,
332
+ "step": 92
333
+ },
334
+ {
335
+ "epoch": 0.27445255474452557,
336
+ "grad_norm": 0.9375,
337
+ "learning_rate": 9.960949040080507e-06,
338
+ "loss": 1.5648375749588013,
339
+ "step": 94
340
+ },
341
+ {
342
+ "epoch": 0.28029197080291973,
343
+ "grad_norm": 0.5390625,
344
+ "learning_rate": 9.95705248644976e-06,
345
+ "loss": 1.6112207174301147,
346
+ "step": 96
347
+ },
348
+ {
349
+ "epoch": 0.28613138686131384,
350
+ "grad_norm": 0.38671875,
351
+ "learning_rate": 9.95297159397452e-06,
352
+ "loss": 1.575645923614502,
353
+ "step": 98
354
+ },
355
+ {
356
+ "epoch": 0.291970802919708,
357
+ "grad_norm": 0.455078125,
358
+ "learning_rate": 9.948706531436067e-06,
359
+ "loss": 1.4316264390945435,
360
+ "step": 100
361
+ },
362
+ {
363
+ "epoch": 0.29781021897810217,
364
+ "grad_norm": 0.365234375,
365
+ "learning_rate": 9.944257475232746e-06,
366
+ "loss": 1.4882372617721558,
367
+ "step": 102
368
+ },
369
+ {
370
+ "epoch": 0.30364963503649633,
371
+ "grad_norm": 0.392578125,
372
+ "learning_rate": 9.939624609372686e-06,
373
+ "loss": 1.4284008741378784,
374
+ "step": 104
375
+ },
376
+ {
377
+ "epoch": 0.3094890510948905,
378
+ "grad_norm": 0.43359375,
379
+ "learning_rate": 9.934808125466181e-06,
380
+ "loss": 1.468398928642273,
381
+ "step": 106
382
+ },
383
+ {
384
+ "epoch": 0.31532846715328466,
385
+ "grad_norm": 0.37109375,
386
+ "learning_rate": 9.929808222717769e-06,
387
+ "loss": 1.465457797050476,
388
+ "step": 108
389
+ },
390
+ {
391
+ "epoch": 0.32116788321167883,
392
+ "grad_norm": 0.359375,
393
+ "learning_rate": 9.924625107917991e-06,
394
+ "loss": 1.425096035003662,
395
+ "step": 110
396
+ },
397
+ {
398
+ "epoch": 0.327007299270073,
399
+ "grad_norm": 0.337890625,
400
+ "learning_rate": 9.919258995434838e-06,
401
+ "loss": 1.4254180192947388,
402
+ "step": 112
403
+ },
404
+ {
405
+ "epoch": 0.33284671532846716,
406
+ "grad_norm": 0.349609375,
407
+ "learning_rate": 9.913710107204888e-06,
408
+ "loss": 1.4670501947402954,
409
+ "step": 114
410
+ },
411
+ {
412
+ "epoch": 0.3386861313868613,
413
+ "grad_norm": 0.41796875,
414
+ "learning_rate": 9.907978672724125e-06,
415
+ "loss": 1.549682378768921,
416
+ "step": 116
417
+ },
418
+ {
419
+ "epoch": 0.3445255474452555,
420
+ "grad_norm": 0.322265625,
421
+ "learning_rate": 9.902064929038447e-06,
422
+ "loss": 1.448873519897461,
423
+ "step": 118
424
+ },
425
+ {
426
+ "epoch": 0.35036496350364965,
427
+ "grad_norm": 0.40625,
428
+ "learning_rate": 9.89596912073386e-06,
429
+ "loss": 1.4342598915100098,
430
+ "step": 120
431
+ },
432
+ {
433
+ "epoch": 0.3562043795620438,
434
+ "grad_norm": 0.396484375,
435
+ "learning_rate": 9.889691499926372e-06,
436
+ "loss": 1.424146294593811,
437
+ "step": 122
438
+ },
439
+ {
440
+ "epoch": 0.362043795620438,
441
+ "grad_norm": 0.337890625,
442
+ "learning_rate": 9.883232326251553e-06,
443
+ "loss": 1.3757034540176392,
444
+ "step": 124
445
+ },
446
+ {
447
+ "epoch": 0.3678832116788321,
448
+ "grad_norm": 0.7421875,
449
+ "learning_rate": 9.876591866853798e-06,
450
+ "loss": 1.5297839641571045,
451
+ "step": 126
452
+ },
453
+ {
454
+ "epoch": 0.37372262773722625,
455
+ "grad_norm": 0.443359375,
456
+ "learning_rate": 9.869770396375291e-06,
457
+ "loss": 1.4363529682159424,
458
+ "step": 128
459
+ },
460
+ {
461
+ "epoch": 0.3795620437956204,
462
+ "grad_norm": 0.515625,
463
+ "learning_rate": 9.862768196944636e-06,
464
+ "loss": 1.5066875219345093,
465
+ "step": 130
466
+ },
467
+ {
468
+ "epoch": 0.3854014598540146,
469
+ "grad_norm": 0.609375,
470
+ "learning_rate": 9.855585558165181e-06,
471
+ "loss": 1.4090046882629395,
472
+ "step": 132
473
+ },
474
+ {
475
+ "epoch": 0.39124087591240875,
476
+ "grad_norm": 0.46875,
477
+ "learning_rate": 9.848222777103063e-06,
478
+ "loss": 1.5839899778366089,
479
+ "step": 134
480
+ },
481
+ {
482
+ "epoch": 0.3970802919708029,
483
+ "grad_norm": 0.7421875,
484
+ "learning_rate": 9.840680158274895e-06,
485
+ "loss": 1.425782561302185,
486
+ "step": 136
487
+ },
488
+ {
489
+ "epoch": 0.4029197080291971,
490
+ "grad_norm": 0.376953125,
491
+ "learning_rate": 9.832958013635195e-06,
492
+ "loss": 1.3894191980361938,
493
+ "step": 138
494
+ },
495
+ {
496
+ "epoch": 0.40875912408759124,
497
+ "grad_norm": 0.322265625,
498
+ "learning_rate": 9.825056662563464e-06,
499
+ "loss": 1.5015205144882202,
500
+ "step": 140
501
+ },
502
+ {
503
+ "epoch": 0.4145985401459854,
504
+ "grad_norm": 1.234375,
505
+ "learning_rate": 9.816976431850994e-06,
506
+ "loss": 1.3953148126602173,
507
+ "step": 142
508
+ },
509
+ {
510
+ "epoch": 0.42043795620437957,
511
+ "grad_norm": 0.66015625,
512
+ "learning_rate": 9.808717655687344e-06,
513
+ "loss": 1.3676221370697021,
514
+ "step": 144
515
+ },
516
+ {
517
+ "epoch": 0.42627737226277373,
518
+ "grad_norm": 1.046875,
519
+ "learning_rate": 9.800280675646508e-06,
520
+ "loss": 1.4441938400268555,
521
+ "step": 146
522
+ },
523
+ {
524
+ "epoch": 0.4321167883211679,
525
+ "grad_norm": 0.609375,
526
+ "learning_rate": 9.791665840672813e-06,
527
+ "loss": 1.3372529745101929,
528
+ "step": 148
529
+ },
530
+ {
531
+ "epoch": 0.43795620437956206,
532
+ "grad_norm": 0.4765625,
533
+ "learning_rate": 9.782873507066466e-06,
534
+ "loss": 1.3514975309371948,
535
+ "step": 150
536
+ },
537
+ {
538
+ "epoch": 0.44379562043795623,
539
+ "grad_norm": 0.515625,
540
+ "learning_rate": 9.773904038468822e-06,
541
+ "loss": 1.3599399328231812,
542
+ "step": 152
543
+ },
544
+ {
545
+ "epoch": 0.44963503649635034,
546
+ "grad_norm": 0.5546875,
547
+ "learning_rate": 9.764757805847352e-06,
548
+ "loss": 1.4376461505889893,
549
+ "step": 154
550
+ },
551
+ {
552
+ "epoch": 0.4554744525547445,
553
+ "grad_norm": 0.443359375,
554
+ "learning_rate": 9.755435187480288e-06,
555
+ "loss": 1.4733713865280151,
556
+ "step": 156
557
+ },
558
+ {
559
+ "epoch": 0.46131386861313867,
560
+ "grad_norm": 0.51171875,
561
+ "learning_rate": 9.745936568940992e-06,
562
+ "loss": 1.4277862310409546,
563
+ "step": 158
564
+ },
565
+ {
566
+ "epoch": 0.46715328467153283,
567
+ "grad_norm": 0.8046875,
568
+ "learning_rate": 9.736262343081998e-06,
569
+ "loss": 1.4021371603012085,
570
+ "step": 160
571
+ },
572
+ {
573
+ "epoch": 0.472992700729927,
574
+ "grad_norm": 0.63671875,
575
+ "learning_rate": 9.726412910018768e-06,
576
+ "loss": 1.4141305685043335,
577
+ "step": 162
578
+ },
579
+ {
580
+ "epoch": 0.47883211678832116,
581
+ "grad_norm": 0.53125,
582
+ "learning_rate": 9.716388677113145e-06,
583
+ "loss": 1.3931472301483154,
584
+ "step": 164
585
+ },
586
+ {
587
+ "epoch": 0.4846715328467153,
588
+ "grad_norm": 0.38671875,
589
+ "learning_rate": 9.7061900589565e-06,
590
+ "loss": 1.3798588514328003,
591
+ "step": 166
592
+ },
593
+ {
594
+ "epoch": 0.4905109489051095,
595
+ "grad_norm": 0.34375,
596
+ "learning_rate": 9.695817477352593e-06,
597
+ "loss": 1.3861608505249023,
598
+ "step": 168
599
+ },
600
+ {
601
+ "epoch": 0.49635036496350365,
602
+ "grad_norm": 0.4453125,
603
+ "learning_rate": 9.685271361300117e-06,
604
+ "loss": 1.4044109582901,
605
+ "step": 170
606
+ },
607
+ {
608
+ "epoch": 0.5021897810218978,
609
+ "grad_norm": 0.345703125,
610
+ "learning_rate": 9.67455214697497e-06,
611
+ "loss": 1.392524003982544,
612
+ "step": 172
613
+ },
614
+ {
615
+ "epoch": 0.5080291970802919,
616
+ "grad_norm": 0.73046875,
617
+ "learning_rate": 9.663660277712194e-06,
618
+ "loss": 1.4445301294326782,
619
+ "step": 174
620
+ },
621
+ {
622
+ "epoch": 0.5138686131386861,
623
+ "grad_norm": 0.4453125,
624
+ "learning_rate": 9.652596203987667e-06,
625
+ "loss": 1.47966468334198,
626
+ "step": 176
627
+ },
628
+ {
629
+ "epoch": 0.5197080291970803,
630
+ "grad_norm": 0.431640625,
631
+ "learning_rate": 9.64136038339944e-06,
632
+ "loss": 1.392290711402893,
633
+ "step": 178
634
+ },
635
+ {
636
+ "epoch": 0.5255474452554745,
637
+ "grad_norm": 0.390625,
638
+ "learning_rate": 9.62995328064884e-06,
639
+ "loss": 1.5381602048873901,
640
+ "step": 180
641
+ },
642
+ {
643
+ "epoch": 0.5313868613138686,
644
+ "grad_norm": 0.36328125,
645
+ "learning_rate": 9.618375367521232e-06,
646
+ "loss": 1.3802764415740967,
647
+ "step": 182
648
+ },
649
+ {
650
+ "epoch": 0.5372262773722628,
651
+ "grad_norm": 0.453125,
652
+ "learning_rate": 9.606627122866513e-06,
653
+ "loss": 1.4806022644042969,
654
+ "step": 184
655
+ },
656
+ {
657
+ "epoch": 0.5430656934306569,
658
+ "grad_norm": 0.546875,
659
+ "learning_rate": 9.594709032579307e-06,
660
+ "loss": 1.4255928993225098,
661
+ "step": 186
662
+ },
663
+ {
664
+ "epoch": 0.5489051094890511,
665
+ "grad_norm": 0.73828125,
666
+ "learning_rate": 9.582621589578873e-06,
667
+ "loss": 1.336686372756958,
668
+ "step": 188
669
+ },
670
+ {
671
+ "epoch": 0.5547445255474452,
672
+ "grad_norm": 0.51171875,
673
+ "learning_rate": 9.570365293788707e-06,
674
+ "loss": 1.4040147066116333,
675
+ "step": 190
676
+ },
677
+ {
678
+ "epoch": 0.5605839416058395,
679
+ "grad_norm": 0.337890625,
680
+ "learning_rate": 9.557940652115875e-06,
681
+ "loss": 1.416895866394043,
682
+ "step": 192
683
+ },
684
+ {
685
+ "epoch": 0.5664233576642336,
686
+ "grad_norm": 0.455078125,
687
+ "learning_rate": 9.545348178430051e-06,
688
+ "loss": 1.3130196332931519,
689
+ "step": 194
690
+ },
691
+ {
692
+ "epoch": 0.5722627737226277,
693
+ "grad_norm": 0.3671875,
694
+ "learning_rate": 9.53258839354225e-06,
695
+ "loss": 1.3405402898788452,
696
+ "step": 196
697
+ },
698
+ {
699
+ "epoch": 0.5781021897810219,
700
+ "grad_norm": 0.337890625,
701
+ "learning_rate": 9.519661825183303e-06,
702
+ "loss": 1.4502712488174438,
703
+ "step": 198
704
+ },
705
+ {
706
+ "epoch": 0.583941605839416,
707
+ "grad_norm": 0.9765625,
708
+ "learning_rate": 9.506569007982023e-06,
709
+ "loss": 1.4316283464431763,
710
+ "step": 200
711
+ },
712
+ {
713
+ "epoch": 0.5897810218978102,
714
+ "grad_norm": 1.0859375,
715
+ "learning_rate": 9.493310483443089e-06,
716
+ "loss": 1.4556365013122559,
717
+ "step": 202
718
+ },
719
+ {
720
+ "epoch": 0.5956204379562043,
721
+ "grad_norm": 0.376953125,
722
+ "learning_rate": 9.479886799924663e-06,
723
+ "loss": 1.413832426071167,
724
+ "step": 204
725
+ },
726
+ {
727
+ "epoch": 0.6014598540145986,
728
+ "grad_norm": 0.56640625,
729
+ "learning_rate": 9.466298512615697e-06,
730
+ "loss": 1.252786636352539,
731
+ "step": 206
732
+ },
733
+ {
734
+ "epoch": 0.6072992700729927,
735
+ "grad_norm": 0.4765625,
736
+ "learning_rate": 9.45254618351298e-06,
737
+ "loss": 1.369625210762024,
738
+ "step": 208
739
+ },
740
+ {
741
+ "epoch": 0.6131386861313869,
742
+ "grad_norm": 1.40625,
743
+ "learning_rate": 9.43863038139789e-06,
744
+ "loss": 1.3296895027160645,
745
+ "step": 210
746
+ },
747
+ {
748
+ "epoch": 0.618978102189781,
749
+ "grad_norm": 0.388671875,
750
+ "learning_rate": 9.424551681812871e-06,
751
+ "loss": 1.4407665729522705,
752
+ "step": 212
753
+ },
754
+ {
755
+ "epoch": 0.6248175182481752,
756
+ "grad_norm": 0.400390625,
757
+ "learning_rate": 9.41031066703763e-06,
758
+ "loss": 1.363607406616211,
759
+ "step": 214
760
+ },
761
+ {
762
+ "epoch": 0.6306569343065693,
763
+ "grad_norm": 0.6953125,
764
+ "learning_rate": 9.395907926065052e-06,
765
+ "loss": 1.4068963527679443,
766
+ "step": 216
767
+ },
768
+ {
769
+ "epoch": 0.6364963503649635,
770
+ "grad_norm": 0.484375,
771
+ "learning_rate": 9.381344054576845e-06,
772
+ "loss": 1.4279429912567139,
773
+ "step": 218
774
+ },
775
+ {
776
+ "epoch": 0.6423357664233577,
777
+ "grad_norm": 0.4375,
778
+ "learning_rate": 9.366619654918895e-06,
779
+ "loss": 1.4228907823562622,
780
+ "step": 220
781
+ },
782
+ {
783
+ "epoch": 0.6481751824817519,
784
+ "grad_norm": 0.330078125,
785
+ "learning_rate": 9.351735336076363e-06,
786
+ "loss": 1.4166944026947021,
787
+ "step": 222
788
+ },
789
+ {
790
+ "epoch": 0.654014598540146,
791
+ "grad_norm": 0.6171875,
792
+ "learning_rate": 9.33669171364849e-06,
793
+ "loss": 1.438708782196045,
794
+ "step": 224
795
+ },
796
+ {
797
+ "epoch": 0.6598540145985401,
798
+ "grad_norm": 0.59375,
799
+ "learning_rate": 9.32148940982315e-06,
800
+ "loss": 1.3473215103149414,
801
+ "step": 226
802
+ },
803
+ {
804
+ "epoch": 0.6656934306569343,
805
+ "grad_norm": 0.326171875,
806
+ "learning_rate": 9.306129053351087e-06,
807
+ "loss": 1.3401542901992798,
808
+ "step": 228
809
+ },
810
+ {
811
+ "epoch": 0.6715328467153284,
812
+ "grad_norm": 0.59765625,
813
+ "learning_rate": 9.290611279519952e-06,
814
+ "loss": 1.4278738498687744,
815
+ "step": 230
816
+ },
817
+ {
818
+ "epoch": 0.6773722627737226,
819
+ "grad_norm": 0.47265625,
820
+ "learning_rate": 9.274936730127991e-06,
821
+ "loss": 1.4194695949554443,
822
+ "step": 232
823
+ },
824
+ {
825
+ "epoch": 0.6832116788321168,
826
+ "grad_norm": 0.59375,
827
+ "learning_rate": 9.259106053457529e-06,
828
+ "loss": 1.4373456239700317,
829
+ "step": 234
830
+ },
831
+ {
832
+ "epoch": 0.689051094890511,
833
+ "grad_norm": 0.51171875,
834
+ "learning_rate": 9.243119904248136e-06,
835
+ "loss": 1.3994792699813843,
836
+ "step": 236
837
+ },
838
+ {
839
+ "epoch": 0.6948905109489051,
840
+ "grad_norm": 0.36328125,
841
+ "learning_rate": 9.226978943669562e-06,
842
+ "loss": 1.3777506351470947,
843
+ "step": 238
844
+ },
845
+ {
846
+ "epoch": 0.7007299270072993,
847
+ "grad_norm": 0.37890625,
848
+ "learning_rate": 9.210683839294386e-06,
849
+ "loss": 1.3514622449874878,
850
+ "step": 240
851
+ },
852
+ {
853
+ "epoch": 0.7065693430656934,
854
+ "grad_norm": 0.5390625,
855
+ "learning_rate": 9.194235265070406e-06,
856
+ "loss": 1.4253668785095215,
857
+ "step": 242
858
+ },
859
+ {
860
+ "epoch": 0.7124087591240876,
861
+ "grad_norm": 0.470703125,
862
+ "learning_rate": 9.177633901292767e-06,
863
+ "loss": 1.3638951778411865,
864
+ "step": 244
865
+ },
866
+ {
867
+ "epoch": 0.7182481751824817,
868
+ "grad_norm": 0.310546875,
869
+ "learning_rate": 9.160880434575823e-06,
870
+ "loss": 1.3718801736831665,
871
+ "step": 246
872
+ },
873
+ {
874
+ "epoch": 0.724087591240876,
875
+ "grad_norm": 0.46484375,
876
+ "learning_rate": 9.143975557824743e-06,
877
+ "loss": 1.3095505237579346,
878
+ "step": 248
879
+ },
880
+ {
881
+ "epoch": 0.7299270072992701,
882
+ "grad_norm": 0.56640625,
883
+ "learning_rate": 9.126919970206843e-06,
884
+ "loss": 1.3069571256637573,
885
+ "step": 250
886
+ },
887
+ {
888
+ "epoch": 0.7357664233576642,
889
+ "grad_norm": 0.27734375,
890
+ "learning_rate": 9.109714377122685e-06,
891
+ "loss": 1.341015100479126,
892
+ "step": 252
893
+ },
894
+ {
895
+ "epoch": 0.7416058394160584,
896
+ "grad_norm": 0.5859375,
897
+ "learning_rate": 9.092359490176892e-06,
898
+ "loss": 1.3410066366195679,
899
+ "step": 254
900
+ },
901
+ {
902
+ "epoch": 0.7474452554744525,
903
+ "grad_norm": 0.396484375,
904
+ "learning_rate": 9.074856027148715e-06,
905
+ "loss": 1.265282154083252,
906
+ "step": 256
907
+ },
908
+ {
909
+ "epoch": 0.7532846715328467,
910
+ "grad_norm": 0.349609375,
911
+ "learning_rate": 9.057204711962352e-06,
912
+ "loss": 1.2000367641448975,
913
+ "step": 258
914
+ },
915
+ {
916
+ "epoch": 0.7591240875912408,
917
+ "grad_norm": 0.50390625,
918
+ "learning_rate": 9.03940627465701e-06,
919
+ "loss": 1.4280476570129395,
920
+ "step": 260
921
+ },
922
+ {
923
+ "epoch": 0.7649635036496351,
924
+ "grad_norm": 0.51171875,
925
+ "learning_rate": 9.021461451356695e-06,
926
+ "loss": 1.3344112634658813,
927
+ "step": 262
928
+ },
929
+ {
930
+ "epoch": 0.7708029197080292,
931
+ "grad_norm": 0.859375,
932
+ "learning_rate": 9.003370984239795e-06,
933
+ "loss": 1.4776599407196045,
934
+ "step": 264
935
+ },
936
+ {
937
+ "epoch": 0.7766423357664234,
938
+ "grad_norm": 0.6328125,
939
+ "learning_rate": 8.985135621508355e-06,
940
+ "loss": 1.4331424236297607,
941
+ "step": 266
942
+ },
943
+ {
944
+ "epoch": 0.7824817518248175,
945
+ "grad_norm": 0.314453125,
946
+ "learning_rate": 8.966756117357152e-06,
947
+ "loss": 1.3084166049957275,
948
+ "step": 268
949
+ },
950
+ {
951
+ "epoch": 0.7883211678832117,
952
+ "grad_norm": 2.640625,
953
+ "learning_rate": 8.948233231942489e-06,
954
+ "loss": 1.3645600080490112,
955
+ "step": 270
956
+ },
957
+ {
958
+ "epoch": 0.7941605839416058,
959
+ "grad_norm": 0.703125,
960
+ "learning_rate": 8.929567731350767e-06,
961
+ "loss": 1.3116084337234497,
962
+ "step": 272
963
+ },
964
+ {
965
+ "epoch": 0.8,
966
+ "grad_norm": 0.65625,
967
+ "learning_rate": 8.9107603875668e-06,
968
+ "loss": 1.363153338432312,
969
+ "step": 274
970
+ },
971
+ {
972
+ "epoch": 0.8058394160583942,
973
+ "grad_norm": 0.4140625,
974
+ "learning_rate": 8.891811978441871e-06,
975
+ "loss": 1.3285009860992432,
976
+ "step": 276
977
+ },
978
+ {
979
+ "epoch": 0.8116788321167884,
980
+ "grad_norm": 0.40234375,
981
+ "learning_rate": 8.87272328766158e-06,
982
+ "loss": 1.4433071613311768,
983
+ "step": 278
984
+ },
985
+ {
986
+ "epoch": 0.8175182481751825,
987
+ "grad_norm": 1.09375,
988
+ "learning_rate": 8.85349510471342e-06,
989
+ "loss": 1.4583128690719604,
990
+ "step": 280
991
+ },
992
+ {
993
+ "epoch": 0.8233576642335766,
994
+ "grad_norm": 0.298828125,
995
+ "learning_rate": 8.834128224854133e-06,
996
+ "loss": 1.3218973875045776,
997
+ "step": 282
998
+ },
999
+ {
1000
+ "epoch": 0.8291970802919708,
1001
+ "grad_norm": 0.97265625,
1002
+ "learning_rate": 8.814623449076809e-06,
1003
+ "loss": 1.291555404663086,
1004
+ "step": 284
1005
+ },
1006
+ {
1007
+ "epoch": 0.8350364963503649,
1008
+ "grad_norm": 0.337890625,
1009
+ "learning_rate": 8.794981584077767e-06,
1010
+ "loss": 1.3497284650802612,
1011
+ "step": 286
1012
+ },
1013
+ {
1014
+ "epoch": 0.8408759124087591,
1015
+ "grad_norm": 0.322265625,
1016
+ "learning_rate": 8.775203442223186e-06,
1017
+ "loss": 1.2834782600402832,
1018
+ "step": 288
1019
+ },
1020
+ {
1021
+ "epoch": 0.8467153284671532,
1022
+ "grad_norm": 0.35546875,
1023
+ "learning_rate": 8.755289841515507e-06,
1024
+ "loss": 1.3291265964508057,
1025
+ "step": 290
1026
+ },
1027
+ {
1028
+ "epoch": 0.8525547445255475,
1029
+ "grad_norm": 0.373046875,
1030
+ "learning_rate": 8.735241605559603e-06,
1031
+ "loss": 1.3475388288497925,
1032
+ "step": 292
1033
+ },
1034
+ {
1035
+ "epoch": 0.8583941605839416,
1036
+ "grad_norm": 0.349609375,
1037
+ "learning_rate": 8.715059563528713e-06,
1038
+ "loss": 1.37130606174469,
1039
+ "step": 294
1040
+ },
1041
+ {
1042
+ "epoch": 0.8642335766423358,
1043
+ "grad_norm": 0.443359375,
1044
+ "learning_rate": 8.694744550130153e-06,
1045
+ "loss": 1.4104875326156616,
1046
+ "step": 296
1047
+ },
1048
+ {
1049
+ "epoch": 0.8700729927007299,
1050
+ "grad_norm": 1.1875,
1051
+ "learning_rate": 8.674297405570782e-06,
1052
+ "loss": 1.3223466873168945,
1053
+ "step": 298
1054
+ },
1055
+ {
1056
+ "epoch": 0.8759124087591241,
1057
+ "grad_norm": 0.52734375,
1058
+ "learning_rate": 8.653718975522271e-06,
1059
+ "loss": 1.3034943342208862,
1060
+ "step": 300
1061
+ },
1062
+ {
1063
+ "epoch": 0.8817518248175182,
1064
+ "grad_norm": 0.46875,
1065
+ "learning_rate": 8.63301011108611e-06,
1066
+ "loss": 1.4011861085891724,
1067
+ "step": 302
1068
+ },
1069
+ {
1070
+ "epoch": 0.8875912408759125,
1071
+ "grad_norm": 0.7578125,
1072
+ "learning_rate": 8.612171668758413e-06,
1073
+ "loss": 1.4873169660568237,
1074
+ "step": 304
1075
+ },
1076
+ {
1077
+ "epoch": 0.8934306569343066,
1078
+ "grad_norm": 0.4296875,
1079
+ "learning_rate": 8.591204510394498e-06,
1080
+ "loss": 1.2987056970596313,
1081
+ "step": 306
1082
+ },
1083
+ {
1084
+ "epoch": 0.8992700729927007,
1085
+ "grad_norm": 0.61328125,
1086
+ "learning_rate": 8.570109503173233e-06,
1087
+ "loss": 1.3402009010314941,
1088
+ "step": 308
1089
+ },
1090
+ {
1091
+ "epoch": 0.9051094890510949,
1092
+ "grad_norm": 0.5859375,
1093
+ "learning_rate": 8.548887519561184e-06,
1094
+ "loss": 1.2947343587875366,
1095
+ "step": 310
1096
+ },
1097
+ {
1098
+ "epoch": 0.910948905109489,
1099
+ "grad_norm": 0.89453125,
1100
+ "learning_rate": 8.527539437276515e-06,
1101
+ "loss": 1.3383480310440063,
1102
+ "step": 312
1103
+ },
1104
+ {
1105
+ "epoch": 0.9167883211678832,
1106
+ "grad_norm": 0.72265625,
1107
+ "learning_rate": 8.506066139252696e-06,
1108
+ "loss": 1.3967725038528442,
1109
+ "step": 314
1110
+ },
1111
+ {
1112
+ "epoch": 0.9226277372262773,
1113
+ "grad_norm": 0.3828125,
1114
+ "learning_rate": 8.484468513601987e-06,
1115
+ "loss": 1.2097219228744507,
1116
+ "step": 316
1117
+ },
1118
+ {
1119
+ "epoch": 0.9284671532846716,
1120
+ "grad_norm": 1.546875,
1121
+ "learning_rate": 8.462747453578698e-06,
1122
+ "loss": 1.3612643480300903,
1123
+ "step": 318
1124
+ },
1125
+ {
1126
+ "epoch": 0.9343065693430657,
1127
+ "grad_norm": 0.578125,
1128
+ "learning_rate": 8.440903857542253e-06,
1129
+ "loss": 1.3607382774353027,
1130
+ "step": 320
1131
+ },
1132
+ {
1133
+ "epoch": 0.9401459854014599,
1134
+ "grad_norm": 0.384765625,
1135
+ "learning_rate": 8.418938628920038e-06,
1136
+ "loss": 1.276464819908142,
1137
+ "step": 322
1138
+ },
1139
+ {
1140
+ "epoch": 0.945985401459854,
1141
+ "grad_norm": 0.5390625,
1142
+ "learning_rate": 8.396852676170018e-06,
1143
+ "loss": 1.3733280897140503,
1144
+ "step": 324
1145
+ },
1146
+ {
1147
+ "epoch": 0.9518248175182482,
1148
+ "grad_norm": 0.6328125,
1149
+ "learning_rate": 8.374646912743187e-06,
1150
+ "loss": 1.365743637084961,
1151
+ "step": 326
1152
+ },
1153
+ {
1154
+ "epoch": 0.9576642335766423,
1155
+ "grad_norm": 0.373046875,
1156
+ "learning_rate": 8.352322257045775e-06,
1157
+ "loss": 1.3169546127319336,
1158
+ "step": 328
1159
+ },
1160
+ {
1161
+ "epoch": 0.9635036496350365,
1162
+ "grad_norm": 0.49609375,
1163
+ "learning_rate": 8.329879632401265e-06,
1164
+ "loss": 1.3615431785583496,
1165
+ "step": 330
1166
+ },
1167
+ {
1168
+ "epoch": 0.9693430656934306,
1169
+ "grad_norm": 0.341796875,
1170
+ "learning_rate": 8.307319967012211e-06,
1171
+ "loss": 1.3420912027359009,
1172
+ "step": 332
1173
+ },
1174
+ {
1175
+ "epoch": 0.9751824817518249,
1176
+ "grad_norm": 0.8203125,
1177
+ "learning_rate": 8.284644193921848e-06,
1178
+ "loss": 1.28525710105896,
1179
+ "step": 334
1180
+ },
1181
+ {
1182
+ "epoch": 0.981021897810219,
1183
+ "grad_norm": 0.48046875,
1184
+ "learning_rate": 8.261853250975487e-06,
1185
+ "loss": 1.3461776971817017,
1186
+ "step": 336
1187
+ },
1188
+ {
1189
+ "epoch": 0.9868613138686131,
1190
+ "grad_norm": 0.546875,
1191
+ "learning_rate": 8.238948080781752e-06,
1192
+ "loss": 1.3771872520446777,
1193
+ "step": 338
1194
+ },
1195
+ {
1196
+ "epoch": 0.9927007299270073,
1197
+ "grad_norm": 0.71484375,
1198
+ "learning_rate": 8.215929630673573e-06,
1199
+ "loss": 1.1941386461257935,
1200
+ "step": 340
1201
+ },
1202
+ {
1203
+ "epoch": 0.9985401459854014,
1204
+ "grad_norm": 0.3359375,
1205
+ "learning_rate": 8.192798852669016e-06,
1206
+ "loss": 1.3167439699172974,
1207
+ "step": 342
1208
+ },
1209
+ {
1210
+ "epoch": 1.0029197080291972,
1211
+ "grad_norm": 0.291015625,
1212
+ "learning_rate": 8.169556703431916e-06,
1213
+ "loss": 1.196553111076355,
1214
+ "step": 344
1215
+ },
1216
+ {
1217
+ "epoch": 1.0087591240875913,
1218
+ "grad_norm": 0.376953125,
1219
+ "learning_rate": 8.146204144232285e-06,
1220
+ "loss": 1.3300058841705322,
1221
+ "step": 346
1222
+ },
1223
+ {
1224
+ "epoch": 1.0145985401459854,
1225
+ "grad_norm": 0.388671875,
1226
+ "learning_rate": 8.122742140906585e-06,
1227
+ "loss": 1.3176811933517456,
1228
+ "step": 348
1229
+ },
1230
+ {
1231
+ "epoch": 1.0204379562043795,
1232
+ "grad_norm": 0.486328125,
1233
+ "learning_rate": 8.099171663817758e-06,
1234
+ "loss": 1.2936145067214966,
1235
+ "step": 350
1236
+ },
1237
+ {
1238
+ "epoch": 1.0262773722627738,
1239
+ "grad_norm": 0.375,
1240
+ "learning_rate": 8.07549368781511e-06,
1241
+ "loss": 1.2210314273834229,
1242
+ "step": 352
1243
+ },
1244
+ {
1245
+ "epoch": 1.032116788321168,
1246
+ "grad_norm": 0.625,
1247
+ "learning_rate": 8.05170919219398e-06,
1248
+ "loss": 1.2909876108169556,
1249
+ "step": 354
1250
+ },
1251
+ {
1252
+ "epoch": 1.037956204379562,
1253
+ "grad_norm": 0.58203125,
1254
+ "learning_rate": 8.027819160655248e-06,
1255
+ "loss": 1.2588341236114502,
1256
+ "step": 356
1257
+ },
1258
+ {
1259
+ "epoch": 1.0437956204379562,
1260
+ "grad_norm": 0.267578125,
1261
+ "learning_rate": 8.003824581264639e-06,
1262
+ "loss": 1.2091121673583984,
1263
+ "step": 358
1264
+ },
1265
+ {
1266
+ "epoch": 1.0496350364963503,
1267
+ "grad_norm": 0.259765625,
1268
+ "learning_rate": 7.979726446411868e-06,
1269
+ "loss": 1.2468533515930176,
1270
+ "step": 360
1271
+ },
1272
+ {
1273
+ "epoch": 1.0554744525547446,
1274
+ "grad_norm": 0.703125,
1275
+ "learning_rate": 7.955525752769588e-06,
1276
+ "loss": 1.2384498119354248,
1277
+ "step": 362
1278
+ },
1279
+ {
1280
+ "epoch": 1.0613138686131387,
1281
+ "grad_norm": 0.447265625,
1282
+ "learning_rate": 7.931223501252176e-06,
1283
+ "loss": 1.2103887796401978,
1284
+ "step": 364
1285
+ },
1286
+ {
1287
+ "epoch": 1.0671532846715328,
1288
+ "grad_norm": 0.375,
1289
+ "learning_rate": 7.906820696974325e-06,
1290
+ "loss": 1.2475626468658447,
1291
+ "step": 366
1292
+ },
1293
+ {
1294
+ "epoch": 1.072992700729927,
1295
+ "grad_norm": 0.474609375,
1296
+ "learning_rate": 7.882318349209492e-06,
1297
+ "loss": 1.3049901723861694,
1298
+ "step": 368
1299
+ },
1300
+ {
1301
+ "epoch": 1.0788321167883212,
1302
+ "grad_norm": 0.478515625,
1303
+ "learning_rate": 7.857717471348135e-06,
1304
+ "loss": 1.3034197092056274,
1305
+ "step": 370
1306
+ },
1307
+ {
1308
+ "epoch": 1.0846715328467154,
1309
+ "grad_norm": 0.5078125,
1310
+ "learning_rate": 7.833019080855812e-06,
1311
+ "loss": 1.177399754524231,
1312
+ "step": 372
1313
+ },
1314
+ {
1315
+ "epoch": 1.0905109489051095,
1316
+ "grad_norm": 0.69921875,
1317
+ "learning_rate": 7.808224199231099e-06,
1318
+ "loss": 1.2648054361343384,
1319
+ "step": 374
1320
+ },
1321
+ {
1322
+ "epoch": 1.0963503649635036,
1323
+ "grad_norm": 0.349609375,
1324
+ "learning_rate": 7.783333851963337e-06,
1325
+ "loss": 1.1547051668167114,
1326
+ "step": 376
1327
+ },
1328
+ {
1329
+ "epoch": 1.102189781021898,
1330
+ "grad_norm": 0.337890625,
1331
+ "learning_rate": 7.75834906849023e-06,
1332
+ "loss": 1.2063677310943604,
1333
+ "step": 378
1334
+ },
1335
+ {
1336
+ "epoch": 1.108029197080292,
1337
+ "grad_norm": 0.431640625,
1338
+ "learning_rate": 7.733270882155252e-06,
1339
+ "loss": 1.294480323791504,
1340
+ "step": 380
1341
+ },
1342
+ {
1343
+ "epoch": 1.1138686131386861,
1344
+ "grad_norm": 0.5078125,
1345
+ "learning_rate": 7.708100330164922e-06,
1346
+ "loss": 1.1963160037994385,
1347
+ "step": 382
1348
+ },
1349
+ {
1350
+ "epoch": 1.1197080291970802,
1351
+ "grad_norm": 0.82421875,
1352
+ "learning_rate": 7.682838453545898e-06,
1353
+ "loss": 1.2486681938171387,
1354
+ "step": 384
1355
+ },
1356
+ {
1357
+ "epoch": 1.1255474452554743,
1358
+ "grad_norm": 0.294921875,
1359
+ "learning_rate": 7.657486297101933e-06,
1360
+ "loss": 1.2629107236862183,
1361
+ "step": 386
1362
+ },
1363
+ {
1364
+ "epoch": 1.1313868613138687,
1365
+ "grad_norm": 0.486328125,
1366
+ "learning_rate": 7.63204490937065e-06,
1367
+ "loss": 1.200867772102356,
1368
+ "step": 388
1369
+ },
1370
+ {
1371
+ "epoch": 1.1372262773722628,
1372
+ "grad_norm": 0.53125,
1373
+ "learning_rate": 7.606515342580181e-06,
1374
+ "loss": 1.2637931108474731,
1375
+ "step": 390
1376
+ },
1377
+ {
1378
+ "epoch": 1.143065693430657,
1379
+ "grad_norm": 0.369140625,
1380
+ "learning_rate": 7.580898652605649e-06,
1381
+ "loss": 1.2616581916809082,
1382
+ "step": 392
1383
+ },
1384
+ {
1385
+ "epoch": 1.148905109489051,
1386
+ "grad_norm": 0.48828125,
1387
+ "learning_rate": 7.555195898925497e-06,
1388
+ "loss": 1.3153358697891235,
1389
+ "step": 394
1390
+ },
1391
+ {
1392
+ "epoch": 1.1547445255474453,
1393
+ "grad_norm": 0.58203125,
1394
+ "learning_rate": 7.529408144577673e-06,
1395
+ "loss": 1.3324670791625977,
1396
+ "step": 396
1397
+ },
1398
+ {
1399
+ "epoch": 1.1605839416058394,
1400
+ "grad_norm": 0.283203125,
1401
+ "learning_rate": 7.503536456115653e-06,
1402
+ "loss": 1.226647138595581,
1403
+ "step": 398
1404
+ },
1405
+ {
1406
+ "epoch": 1.1664233576642336,
1407
+ "grad_norm": 1.125,
1408
+ "learning_rate": 7.47758190356434e-06,
1409
+ "loss": 1.2243517637252808,
1410
+ "step": 400
1411
+ },
1412
+ {
1413
+ "epoch": 1.1722627737226277,
1414
+ "grad_norm": 0.75,
1415
+ "learning_rate": 7.451545560375808e-06,
1416
+ "loss": 1.2697927951812744,
1417
+ "step": 402
1418
+ },
1419
+ {
1420
+ "epoch": 1.178102189781022,
1421
+ "grad_norm": 1.0703125,
1422
+ "learning_rate": 7.425428503384898e-06,
1423
+ "loss": 1.2327685356140137,
1424
+ "step": 404
1425
+ },
1426
+ {
1427
+ "epoch": 1.183941605839416,
1428
+ "grad_norm": 1.390625,
1429
+ "learning_rate": 7.399231812764695e-06,
1430
+ "loss": 1.3016244173049927,
1431
+ "step": 406
1432
+ },
1433
+ {
1434
+ "epoch": 1.1897810218978102,
1435
+ "grad_norm": 0.3046875,
1436
+ "learning_rate": 7.372956571981831e-06,
1437
+ "loss": 1.189962387084961,
1438
+ "step": 408
1439
+ },
1440
+ {
1441
+ "epoch": 1.1956204379562043,
1442
+ "grad_norm": 2.15625,
1443
+ "learning_rate": 7.346603867751698e-06,
1444
+ "loss": 1.235486626625061,
1445
+ "step": 410
1446
+ },
1447
+ {
1448
+ "epoch": 1.2014598540145984,
1449
+ "grad_norm": 1.0703125,
1450
+ "learning_rate": 7.32017478999349e-06,
1451
+ "loss": 1.2581639289855957,
1452
+ "step": 412
1453
+ },
1454
+ {
1455
+ "epoch": 1.2072992700729928,
1456
+ "grad_norm": 0.44140625,
1457
+ "learning_rate": 7.293670431785123e-06,
1458
+ "loss": 1.2778874635696411,
1459
+ "step": 414
1460
+ },
1461
+ {
1462
+ "epoch": 1.2131386861313869,
1463
+ "grad_norm": 0.66015625,
1464
+ "learning_rate": 7.267091889318033e-06,
1465
+ "loss": 1.3071149587631226,
1466
+ "step": 416
1467
+ },
1468
+ {
1469
+ "epoch": 1.218978102189781,
1470
+ "grad_norm": 0.404296875,
1471
+ "learning_rate": 7.240440261851839e-06,
1472
+ "loss": 1.2892022132873535,
1473
+ "step": 418
1474
+ },
1475
+ {
1476
+ "epoch": 1.224817518248175,
1477
+ "grad_norm": 0.828125,
1478
+ "learning_rate": 7.21371665166887e-06,
1479
+ "loss": 1.1825231313705444,
1480
+ "step": 420
1481
+ },
1482
+ {
1483
+ "epoch": 1.2306569343065694,
1484
+ "grad_norm": 0.703125,
1485
+ "learning_rate": 7.186922164028589e-06,
1486
+ "loss": 1.2527143955230713,
1487
+ "step": 422
1488
+ },
1489
+ {
1490
+ "epoch": 1.2364963503649635,
1491
+ "grad_norm": 0.71875,
1492
+ "learning_rate": 7.160057907121868e-06,
1493
+ "loss": 1.1351896524429321,
1494
+ "step": 424
1495
+ },
1496
+ {
1497
+ "epoch": 1.2423357664233576,
1498
+ "grad_norm": 0.47265625,
1499
+ "learning_rate": 7.133124992025161e-06,
1500
+ "loss": 1.2882559299468994,
1501
+ "step": 426
1502
+ },
1503
+ {
1504
+ "epoch": 1.2481751824817517,
1505
+ "grad_norm": 2.796875,
1506
+ "learning_rate": 7.106124532654553e-06,
1507
+ "loss": 1.3231326341629028,
1508
+ "step": 428
1509
+ },
1510
+ {
1511
+ "epoch": 1.254014598540146,
1512
+ "grad_norm": 0.328125,
1513
+ "learning_rate": 7.079057645719686e-06,
1514
+ "loss": 1.138244867324829,
1515
+ "step": 430
1516
+ },
1517
+ {
1518
+ "epoch": 1.2598540145985402,
1519
+ "grad_norm": 0.39453125,
1520
+ "learning_rate": 7.051925450677568e-06,
1521
+ "loss": 1.3060168027877808,
1522
+ "step": 432
1523
+ },
1524
+ {
1525
+ "epoch": 1.2656934306569343,
1526
+ "grad_norm": 0.484375,
1527
+ "learning_rate": 7.024729069686288e-06,
1528
+ "loss": 1.2885905504226685,
1529
+ "step": 434
1530
+ },
1531
+ {
1532
+ "epoch": 1.2715328467153284,
1533
+ "grad_norm": 0.2890625,
1534
+ "learning_rate": 6.997469627558591e-06,
1535
+ "loss": 1.2515441179275513,
1536
+ "step": 436
1537
+ },
1538
+ {
1539
+ "epoch": 1.2773722627737225,
1540
+ "grad_norm": 0.41796875,
1541
+ "learning_rate": 6.970148251715363e-06,
1542
+ "loss": 1.2661933898925781,
1543
+ "step": 438
1544
+ },
1545
+ {
1546
+ "epoch": 1.2832116788321168,
1547
+ "grad_norm": 0.90625,
1548
+ "learning_rate": 6.942766072139e-06,
1549
+ "loss": 1.2927354574203491,
1550
+ "step": 440
1551
+ },
1552
+ {
1553
+ "epoch": 1.289051094890511,
1554
+ "grad_norm": 0.62890625,
1555
+ "learning_rate": 6.9153242213266745e-06,
1556
+ "loss": 1.2908434867858887,
1557
+ "step": 442
1558
+ },
1559
+ {
1560
+ "epoch": 1.294890510948905,
1561
+ "grad_norm": 0.482421875,
1562
+ "learning_rate": 6.887823834243501e-06,
1563
+ "loss": 1.2376095056533813,
1564
+ "step": 444
1565
+ },
1566
+ {
1567
+ "epoch": 1.3007299270072994,
1568
+ "grad_norm": 0.36328125,
1569
+ "learning_rate": 6.860266048275585e-06,
1570
+ "loss": 1.2220968008041382,
1571
+ "step": 446
1572
+ },
1573
+ {
1574
+ "epoch": 1.3065693430656935,
1575
+ "grad_norm": 0.52734375,
1576
+ "learning_rate": 6.832652003182989e-06,
1577
+ "loss": 1.2590991258621216,
1578
+ "step": 448
1579
+ },
1580
+ {
1581
+ "epoch": 1.3124087591240876,
1582
+ "grad_norm": 0.5234375,
1583
+ "learning_rate": 6.804982841052593e-06,
1584
+ "loss": 1.2670376300811768,
1585
+ "step": 450
1586
+ },
1587
+ {
1588
+ "epoch": 1.3182481751824817,
1589
+ "grad_norm": 1.3984375,
1590
+ "learning_rate": 6.777259706250856e-06,
1591
+ "loss": 1.3193187713623047,
1592
+ "step": 452
1593
+ },
1594
+ {
1595
+ "epoch": 1.3240875912408758,
1596
+ "grad_norm": 1.0234375,
1597
+ "learning_rate": 6.749483745376489e-06,
1598
+ "loss": 1.2359987497329712,
1599
+ "step": 454
1600
+ },
1601
+ {
1602
+ "epoch": 1.3299270072992702,
1603
+ "grad_norm": 0.546875,
1604
+ "learning_rate": 6.721656107213032e-06,
1605
+ "loss": 1.2815067768096924,
1606
+ "step": 456
1607
+ },
1608
+ {
1609
+ "epoch": 1.3357664233576643,
1610
+ "grad_norm": 0.486328125,
1611
+ "learning_rate": 6.693777942681339e-06,
1612
+ "loss": 1.2145709991455078,
1613
+ "step": 458
1614
+ },
1615
+ {
1616
+ "epoch": 1.3416058394160584,
1617
+ "grad_norm": 0.435546875,
1618
+ "learning_rate": 6.665850404791986e-06,
1619
+ "loss": 1.2475862503051758,
1620
+ "step": 460
1621
+ },
1622
+ {
1623
+ "epoch": 1.3474452554744525,
1624
+ "grad_norm": 0.4453125,
1625
+ "learning_rate": 6.637874648597571e-06,
1626
+ "loss": 1.2392385005950928,
1627
+ "step": 462
1628
+ },
1629
+ {
1630
+ "epoch": 1.3532846715328466,
1631
+ "grad_norm": 1.1875,
1632
+ "learning_rate": 6.60985183114495e-06,
1633
+ "loss": 1.2735258340835571,
1634
+ "step": 464
1635
+ },
1636
+ {
1637
+ "epoch": 1.359124087591241,
1638
+ "grad_norm": 0.34765625,
1639
+ "learning_rate": 6.5817831114273845e-06,
1640
+ "loss": 1.26645827293396,
1641
+ "step": 466
1642
+ },
1643
+ {
1644
+ "epoch": 1.364963503649635,
1645
+ "grad_norm": 0.396484375,
1646
+ "learning_rate": 6.553669650336599e-06,
1647
+ "loss": 1.2172024250030518,
1648
+ "step": 468
1649
+ },
1650
+ {
1651
+ "epoch": 1.3708029197080291,
1652
+ "grad_norm": 0.37890625,
1653
+ "learning_rate": 6.5255126106147785e-06,
1654
+ "loss": 1.3016058206558228,
1655
+ "step": 470
1656
+ },
1657
+ {
1658
+ "epoch": 1.3766423357664235,
1659
+ "grad_norm": 0.400390625,
1660
+ "learning_rate": 6.497313156806467e-06,
1661
+ "loss": 1.3059654235839844,
1662
+ "step": 472
1663
+ },
1664
+ {
1665
+ "epoch": 1.3824817518248176,
1666
+ "grad_norm": 0.7421875,
1667
+ "learning_rate": 6.469072455210414e-06,
1668
+ "loss": 1.2682950496673584,
1669
+ "step": 474
1670
+ },
1671
+ {
1672
+ "epoch": 1.3883211678832117,
1673
+ "grad_norm": 0.41015625,
1674
+ "learning_rate": 6.4407916738313256e-06,
1675
+ "loss": 1.190187931060791,
1676
+ "step": 476
1677
+ },
1678
+ {
1679
+ "epoch": 1.3941605839416058,
1680
+ "grad_norm": 0.51953125,
1681
+ "learning_rate": 6.412471982331574e-06,
1682
+ "loss": 1.2461662292480469,
1683
+ "step": 478
1684
+ },
1685
+ {
1686
+ "epoch": 1.4,
1687
+ "grad_norm": 0.5,
1688
+ "learning_rate": 6.384114551982804e-06,
1689
+ "loss": 1.2167097330093384,
1690
+ "step": 480
1691
+ },
1692
+ {
1693
+ "epoch": 1.4058394160583942,
1694
+ "grad_norm": 0.390625,
1695
+ "learning_rate": 6.3557205556175e-06,
1696
+ "loss": 1.1439507007598877,
1697
+ "step": 482
1698
+ },
1699
+ {
1700
+ "epoch": 1.4116788321167884,
1701
+ "grad_norm": 1.1328125,
1702
+ "learning_rate": 6.327291167580478e-06,
1703
+ "loss": 1.207546353340149,
1704
+ "step": 484
1705
+ },
1706
+ {
1707
+ "epoch": 1.4175182481751825,
1708
+ "grad_norm": 0.486328125,
1709
+ "learning_rate": 6.298827563680314e-06,
1710
+ "loss": 1.292868971824646,
1711
+ "step": 486
1712
+ },
1713
+ {
1714
+ "epoch": 1.4233576642335766,
1715
+ "grad_norm": 0.42578125,
1716
+ "learning_rate": 6.270330921140718e-06,
1717
+ "loss": 1.3323169946670532,
1718
+ "step": 488
1719
+ },
1720
+ {
1721
+ "epoch": 1.4291970802919707,
1722
+ "grad_norm": 0.458984375,
1723
+ "learning_rate": 6.241802418551838e-06,
1724
+ "loss": 1.19215726852417,
1725
+ "step": 490
1726
+ },
1727
+ {
1728
+ "epoch": 1.435036496350365,
1729
+ "grad_norm": 0.5703125,
1730
+ "learning_rate": 6.21324323582152e-06,
1731
+ "loss": 1.2297883033752441,
1732
+ "step": 492
1733
+ },
1734
+ {
1735
+ "epoch": 1.4408759124087591,
1736
+ "grad_norm": 0.33984375,
1737
+ "learning_rate": 6.1846545541265115e-06,
1738
+ "loss": 1.3055942058563232,
1739
+ "step": 494
1740
+ },
1741
+ {
1742
+ "epoch": 1.4467153284671532,
1743
+ "grad_norm": 0.4140625,
1744
+ "learning_rate": 6.156037555863597e-06,
1745
+ "loss": 1.2372398376464844,
1746
+ "step": 496
1747
+ },
1748
+ {
1749
+ "epoch": 1.4525547445255476,
1750
+ "grad_norm": 0.7265625,
1751
+ "learning_rate": 6.1273934246007105e-06,
1752
+ "loss": 1.3352288007736206,
1753
+ "step": 498
1754
+ },
1755
+ {
1756
+ "epoch": 1.4583941605839417,
1757
+ "grad_norm": 0.5390625,
1758
+ "learning_rate": 6.098723345027977e-06,
1759
+ "loss": 1.2846771478652954,
1760
+ "step": 500
1761
+ },
1762
+ {
1763
+ "epoch": 1.4642335766423358,
1764
+ "grad_norm": 0.5703125,
1765
+ "learning_rate": 6.070028502908709e-06,
1766
+ "loss": 1.2137339115142822,
1767
+ "step": 502
1768
+ },
1769
+ {
1770
+ "epoch": 1.4700729927007299,
1771
+ "grad_norm": 0.470703125,
1772
+ "learning_rate": 6.041310085030378e-06,
1773
+ "loss": 1.2333766222000122,
1774
+ "step": 504
1775
+ },
1776
+ {
1777
+ "epoch": 1.475912408759124,
1778
+ "grad_norm": 0.400390625,
1779
+ "learning_rate": 6.012569279155521e-06,
1780
+ "loss": 1.217051386833191,
1781
+ "step": 506
1782
+ },
1783
+ {
1784
+ "epoch": 1.4817518248175183,
1785
+ "grad_norm": 1.078125,
1786
+ "learning_rate": 5.983807273972617e-06,
1787
+ "loss": 1.1864327192306519,
1788
+ "step": 508
1789
+ },
1790
+ {
1791
+ "epoch": 1.4875912408759124,
1792
+ "grad_norm": 0.46484375,
1793
+ "learning_rate": 5.955025259046927e-06,
1794
+ "loss": 1.2389144897460938,
1795
+ "step": 510
1796
+ },
1797
+ {
1798
+ "epoch": 1.4934306569343065,
1799
+ "grad_norm": 0.71875,
1800
+ "learning_rate": 5.926224424771291e-06,
1801
+ "loss": 1.2382243871688843,
1802
+ "step": 512
1803
+ },
1804
+ {
1805
+ "epoch": 1.4992700729927007,
1806
+ "grad_norm": 0.4921875,
1807
+ "learning_rate": 5.897405962316899e-06,
1808
+ "loss": 1.208208680152893,
1809
+ "step": 514
1810
+ },
1811
+ {
1812
+ "epoch": 1.5051094890510948,
1813
+ "grad_norm": 0.44921875,
1814
+ "learning_rate": 5.868571063584023e-06,
1815
+ "loss": 1.286102533340454,
1816
+ "step": 516
1817
+ },
1818
+ {
1819
+ "epoch": 1.510948905109489,
1820
+ "grad_norm": 0.494140625,
1821
+ "learning_rate": 5.83972092115272e-06,
1822
+ "loss": 1.2459356784820557,
1823
+ "step": 518
1824
+ },
1825
+ {
1826
+ "epoch": 1.5167883211678832,
1827
+ "grad_norm": 0.53515625,
1828
+ "learning_rate": 5.810856728233513e-06,
1829
+ "loss": 1.276041865348816,
1830
+ "step": 520
1831
+ },
1832
+ {
1833
+ "epoch": 1.5226277372262773,
1834
+ "grad_norm": 0.734375,
1835
+ "learning_rate": 5.781979678618033e-06,
1836
+ "loss": 1.27890944480896,
1837
+ "step": 522
1838
+ },
1839
+ {
1840
+ "epoch": 1.5284671532846716,
1841
+ "grad_norm": 1.7265625,
1842
+ "learning_rate": 5.753090966629654e-06,
1843
+ "loss": 1.1399537324905396,
1844
+ "step": 524
1845
+ },
1846
+ {
1847
+ "epoch": 1.5343065693430655,
1848
+ "grad_norm": 0.388671875,
1849
+ "learning_rate": 5.7241917870740935e-06,
1850
+ "loss": 1.2000788450241089,
1851
+ "step": 526
1852
+ },
1853
+ {
1854
+ "epoch": 1.5401459854014599,
1855
+ "grad_norm": 0.451171875,
1856
+ "learning_rate": 5.695283335189992e-06,
1857
+ "loss": 1.3109184503555298,
1858
+ "step": 528
1859
+ },
1860
+ {
1861
+ "epoch": 1.545985401459854,
1862
+ "grad_norm": 0.5859375,
1863
+ "learning_rate": 5.666366806599488e-06,
1864
+ "loss": 1.2306164503097534,
1865
+ "step": 530
1866
+ },
1867
+ {
1868
+ "epoch": 1.551824817518248,
1869
+ "grad_norm": 0.3359375,
1870
+ "learning_rate": 5.63744339725876e-06,
1871
+ "loss": 1.2434693574905396,
1872
+ "step": 532
1873
+ },
1874
+ {
1875
+ "epoch": 1.5576642335766424,
1876
+ "grad_norm": 0.8125,
1877
+ "learning_rate": 5.608514303408572e-06,
1878
+ "loss": 1.1840685606002808,
1879
+ "step": 534
1880
+ },
1881
+ {
1882
+ "epoch": 1.5635036496350365,
1883
+ "grad_norm": 0.90234375,
1884
+ "learning_rate": 5.579580721524786e-06,
1885
+ "loss": 1.2241274118423462,
1886
+ "step": 536
1887
+ },
1888
+ {
1889
+ "epoch": 1.5693430656934306,
1890
+ "grad_norm": 0.478515625,
1891
+ "learning_rate": 5.550643848268889e-06,
1892
+ "loss": 1.206296682357788,
1893
+ "step": 538
1894
+ },
1895
+ {
1896
+ "epoch": 1.575182481751825,
1897
+ "grad_norm": 1.6484375,
1898
+ "learning_rate": 5.5217048804384945e-06,
1899
+ "loss": 1.2080531120300293,
1900
+ "step": 540
1901
+ },
1902
+ {
1903
+ "epoch": 1.5810218978102188,
1904
+ "grad_norm": 0.498046875,
1905
+ "learning_rate": 5.492765014917845e-06,
1906
+ "loss": 1.207125186920166,
1907
+ "step": 542
1908
+ },
1909
+ {
1910
+ "epoch": 1.5868613138686132,
1911
+ "grad_norm": 1.0546875,
1912
+ "learning_rate": 5.463825448628314e-06,
1913
+ "loss": 1.218297004699707,
1914
+ "step": 544
1915
+ },
1916
+ {
1917
+ "epoch": 1.5927007299270073,
1918
+ "grad_norm": 0.431640625,
1919
+ "learning_rate": 5.434887378478892e-06,
1920
+ "loss": 1.1724162101745605,
1921
+ "step": 546
1922
+ },
1923
+ {
1924
+ "epoch": 1.5985401459854014,
1925
+ "grad_norm": 0.4765625,
1926
+ "learning_rate": 5.405952001316697e-06,
1927
+ "loss": 1.215348720550537,
1928
+ "step": 548
1929
+ },
1930
+ {
1931
+ "epoch": 1.6043795620437957,
1932
+ "grad_norm": 0.3359375,
1933
+ "learning_rate": 5.377020513877463e-06,
1934
+ "loss": 1.2883554697036743,
1935
+ "step": 550
1936
+ },
1937
+ {
1938
+ "epoch": 1.6102189781021898,
1939
+ "grad_norm": 0.490234375,
1940
+ "learning_rate": 5.348094112736057e-06,
1941
+ "loss": 1.253550410270691,
1942
+ "step": 552
1943
+ },
1944
+ {
1945
+ "epoch": 1.616058394160584,
1946
+ "grad_norm": 0.4921875,
1947
+ "learning_rate": 5.319173994256973e-06,
1948
+ "loss": 1.29800546169281,
1949
+ "step": 554
1950
+ },
1951
+ {
1952
+ "epoch": 1.621897810218978,
1953
+ "grad_norm": 0.390625,
1954
+ "learning_rate": 5.290261354544867e-06,
1955
+ "loss": 1.1871838569641113,
1956
+ "step": 556
1957
+ },
1958
+ {
1959
+ "epoch": 1.6277372262773722,
1960
+ "grad_norm": 0.65625,
1961
+ "learning_rate": 5.261357389395078e-06,
1962
+ "loss": 1.1837831735610962,
1963
+ "step": 558
1964
+ },
1965
+ {
1966
+ "epoch": 1.6335766423357665,
1967
+ "grad_norm": 1.125,
1968
+ "learning_rate": 5.232463294244178e-06,
1969
+ "loss": 1.2513107061386108,
1970
+ "step": 560
1971
+ },
1972
+ {
1973
+ "epoch": 1.6394160583941606,
1974
+ "grad_norm": 0.52734375,
1975
+ "learning_rate": 5.203580264120521e-06,
1976
+ "loss": 1.3240464925765991,
1977
+ "step": 562
1978
+ },
1979
+ {
1980
+ "epoch": 1.6452554744525547,
1981
+ "grad_norm": 0.52734375,
1982
+ "learning_rate": 5.1747094935948325e-06,
1983
+ "loss": 1.2752066850662231,
1984
+ "step": 564
1985
+ },
1986
+ {
1987
+ "epoch": 1.651094890510949,
1988
+ "grad_norm": 0.546875,
1989
+ "learning_rate": 5.145852176730786e-06,
1990
+ "loss": 1.1872999668121338,
1991
+ "step": 566
1992
+ },
1993
+ {
1994
+ "epoch": 1.656934306569343,
1995
+ "grad_norm": 0.69140625,
1996
+ "learning_rate": 5.117009507035628e-06,
1997
+ "loss": 1.2202606201171875,
1998
+ "step": 568
1999
+ },
2000
+ {
2001
+ "epoch": 1.6627737226277373,
2002
+ "grad_norm": 0.625,
2003
+ "learning_rate": 5.088182677410819e-06,
2004
+ "loss": 1.2224456071853638,
2005
+ "step": 570
2006
+ },
2007
+ {
2008
+ "epoch": 1.6686131386861314,
2009
+ "grad_norm": 0.6328125,
2010
+ "learning_rate": 5.059372880102683e-06,
2011
+ "loss": 1.2351001501083374,
2012
+ "step": 572
2013
+ },
2014
+ {
2015
+ "epoch": 1.6744525547445255,
2016
+ "grad_norm": 0.609375,
2017
+ "learning_rate": 5.0305813066531136e-06,
2018
+ "loss": 1.2855942249298096,
2019
+ "step": 574
2020
+ },
2021
+ {
2022
+ "epoch": 1.6802919708029198,
2023
+ "grad_norm": 0.98828125,
2024
+ "learning_rate": 5.001809147850282e-06,
2025
+ "loss": 1.2410361766815186,
2026
+ "step": 576
2027
+ },
2028
+ {
2029
+ "epoch": 1.686131386861314,
2030
+ "grad_norm": 0.431640625,
2031
+ "learning_rate": 4.97305759367939e-06,
2032
+ "loss": 1.2131555080413818,
2033
+ "step": 578
2034
+ },
2035
+ {
2036
+ "epoch": 1.691970802919708,
2037
+ "grad_norm": 0.4375,
2038
+ "learning_rate": 4.944327833273456e-06,
2039
+ "loss": 1.2389878034591675,
2040
+ "step": 580
2041
+ },
2042
+ {
2043
+ "epoch": 1.6978102189781021,
2044
+ "grad_norm": 0.46484375,
2045
+ "learning_rate": 4.9156210548641315e-06,
2046
+ "loss": 1.2704704999923706,
2047
+ "step": 582
2048
+ },
2049
+ {
2050
+ "epoch": 1.7036496350364962,
2051
+ "grad_norm": 0.50390625,
2052
+ "learning_rate": 4.886938445732557e-06,
2053
+ "loss": 1.2079095840454102,
2054
+ "step": 584
2055
+ },
2056
+ {
2057
+ "epoch": 1.7094890510948906,
2058
+ "grad_norm": 0.5703125,
2059
+ "learning_rate": 4.858281192160258e-06,
2060
+ "loss": 1.3086167573928833,
2061
+ "step": 586
2062
+ },
2063
+ {
2064
+ "epoch": 1.7153284671532847,
2065
+ "grad_norm": 0.37109375,
2066
+ "learning_rate": 4.82965047938008e-06,
2067
+ "loss": 1.1482274532318115,
2068
+ "step": 588
2069
+ },
2070
+ {
2071
+ "epoch": 1.7211678832116788,
2072
+ "grad_norm": 0.349609375,
2073
+ "learning_rate": 4.801047491527176e-06,
2074
+ "loss": 1.316062331199646,
2075
+ "step": 590
2076
+ },
2077
+ {
2078
+ "epoch": 1.7270072992700731,
2079
+ "grad_norm": 0.8359375,
2080
+ "learning_rate": 4.772473411590022e-06,
2081
+ "loss": 1.271776556968689,
2082
+ "step": 592
2083
+ },
2084
+ {
2085
+ "epoch": 1.732846715328467,
2086
+ "grad_norm": 0.427734375,
2087
+ "learning_rate": 4.743929421361492e-06,
2088
+ "loss": 1.1389895677566528,
2089
+ "step": 594
2090
+ },
2091
+ {
2092
+ "epoch": 1.7386861313868613,
2093
+ "grad_norm": 7.53125,
2094
+ "learning_rate": 4.715416701389985e-06,
2095
+ "loss": 1.196537733078003,
2096
+ "step": 596
2097
+ },
2098
+ {
2099
+ "epoch": 1.7445255474452555,
2100
+ "grad_norm": 0.333984375,
2101
+ "learning_rate": 4.686936430930597e-06,
2102
+ "loss": 1.2567116022109985,
2103
+ "step": 598
2104
+ },
2105
+ {
2106
+ "epoch": 1.7503649635036496,
2107
+ "grad_norm": 0.427734375,
2108
+ "learning_rate": 4.658489787896346e-06,
2109
+ "loss": 1.3270188570022583,
2110
+ "step": 600
2111
+ },
2112
+ {
2113
+ "epoch": 1.756204379562044,
2114
+ "grad_norm": 0.458984375,
2115
+ "learning_rate": 4.630077948809457e-06,
2116
+ "loss": 1.2750407457351685,
2117
+ "step": 602
2118
+ },
2119
+ {
2120
+ "epoch": 1.762043795620438,
2121
+ "grad_norm": 0.447265625,
2122
+ "learning_rate": 4.601702088752702e-06,
2123
+ "loss": 1.2777055501937866,
2124
+ "step": 604
2125
+ },
2126
+ {
2127
+ "epoch": 1.7678832116788321,
2128
+ "grad_norm": 0.361328125,
2129
+ "learning_rate": 4.573363381320795e-06,
2130
+ "loss": 1.1461223363876343,
2131
+ "step": 606
2132
+ },
2133
+ {
2134
+ "epoch": 1.7737226277372264,
2135
+ "grad_norm": 0.498046875,
2136
+ "learning_rate": 4.5450629985718655e-06,
2137
+ "loss": 1.2125712633132935,
2138
+ "step": 608
2139
+ },
2140
+ {
2141
+ "epoch": 1.7795620437956203,
2142
+ "grad_norm": 0.453125,
2143
+ "learning_rate": 4.516802110978968e-06,
2144
+ "loss": 1.259176254272461,
2145
+ "step": 610
2146
+ },
2147
+ {
2148
+ "epoch": 1.7854014598540147,
2149
+ "grad_norm": 1.3203125,
2150
+ "learning_rate": 4.488581887381689e-06,
2151
+ "loss": 1.1314493417739868,
2152
+ "step": 612
2153
+ },
2154
+ {
2155
+ "epoch": 1.7912408759124088,
2156
+ "grad_norm": 0.39453125,
2157
+ "learning_rate": 4.460403494937787e-06,
2158
+ "loss": 1.2192261219024658,
2159
+ "step": 614
2160
+ },
2161
+ {
2162
+ "epoch": 1.7970802919708029,
2163
+ "grad_norm": 0.5859375,
2164
+ "learning_rate": 4.4322680990749365e-06,
2165
+ "loss": 1.1592247486114502,
2166
+ "step": 616
2167
+ },
2168
+ {
2169
+ "epoch": 1.8029197080291972,
2170
+ "grad_norm": 0.404296875,
2171
+ "learning_rate": 4.404176863442515e-06,
2172
+ "loss": 1.300370454788208,
2173
+ "step": 618
2174
+ },
2175
+ {
2176
+ "epoch": 1.808759124087591,
2177
+ "grad_norm": 1.65625,
2178
+ "learning_rate": 4.376130949863487e-06,
2179
+ "loss": 1.2521164417266846,
2180
+ "step": 620
2181
+ },
2182
+ {
2183
+ "epoch": 1.8145985401459854,
2184
+ "grad_norm": 0.5703125,
2185
+ "learning_rate": 4.3481315182863405e-06,
2186
+ "loss": 1.1387872695922852,
2187
+ "step": 622
2188
+ },
2189
+ {
2190
+ "epoch": 1.8204379562043795,
2191
+ "grad_norm": 0.58984375,
2192
+ "learning_rate": 4.320179726737118e-06,
2193
+ "loss": 1.2439383268356323,
2194
+ "step": 624
2195
+ },
2196
+ {
2197
+ "epoch": 1.8262773722627736,
2198
+ "grad_norm": 0.859375,
2199
+ "learning_rate": 4.292276731271528e-06,
2200
+ "loss": 1.1349395513534546,
2201
+ "step": 626
2202
+ },
2203
+ {
2204
+ "epoch": 1.832116788321168,
2205
+ "grad_norm": 0.51953125,
2206
+ "learning_rate": 4.264423685927123e-06,
2207
+ "loss": 1.2615232467651367,
2208
+ "step": 628
2209
+ },
2210
+ {
2211
+ "epoch": 1.837956204379562,
2212
+ "grad_norm": 1.1328125,
2213
+ "learning_rate": 4.236621742675572e-06,
2214
+ "loss": 1.297728419303894,
2215
+ "step": 630
2216
+ },
2217
+ {
2218
+ "epoch": 1.8437956204379562,
2219
+ "grad_norm": 0.89453125,
2220
+ "learning_rate": 4.208872051375016e-06,
2221
+ "loss": 1.3184691667556763,
2222
+ "step": 632
2223
+ },
2224
+ {
2225
+ "epoch": 1.8496350364963505,
2226
+ "grad_norm": 0.4140625,
2227
+ "learning_rate": 4.1811757597225154e-06,
2228
+ "loss": 1.2532655000686646,
2229
+ "step": 634
2230
+ },
2231
+ {
2232
+ "epoch": 1.8554744525547444,
2233
+ "grad_norm": 0.7421875,
2234
+ "learning_rate": 4.1535340132065775e-06,
2235
+ "loss": 1.2408058643341064,
2236
+ "step": 636
2237
+ },
2238
+ {
2239
+ "epoch": 1.8613138686131387,
2240
+ "grad_norm": 0.52734375,
2241
+ "learning_rate": 4.125947955059781e-06,
2242
+ "loss": 1.298876404762268,
2243
+ "step": 638
2244
+ },
2245
+ {
2246
+ "epoch": 1.8671532846715329,
2247
+ "grad_norm": 0.58203125,
2248
+ "learning_rate": 4.098418726211497e-06,
2249
+ "loss": 1.1866939067840576,
2250
+ "step": 640
2251
+ },
2252
+ {
2253
+ "epoch": 1.872992700729927,
2254
+ "grad_norm": 0.515625,
2255
+ "learning_rate": 4.070947465240695e-06,
2256
+ "loss": 1.2406346797943115,
2257
+ "step": 642
2258
+ },
2259
+ {
2260
+ "epoch": 1.8788321167883213,
2261
+ "grad_norm": 0.45703125,
2262
+ "learning_rate": 4.043535308328861e-06,
2263
+ "loss": 1.2553025484085083,
2264
+ "step": 644
2265
+ },
2266
+ {
2267
+ "epoch": 1.8846715328467152,
2268
+ "grad_norm": 0.50390625,
2269
+ "learning_rate": 4.0161833892129944e-06,
2270
+ "loss": 1.2756036520004272,
2271
+ "step": 646
2272
+ },
2273
+ {
2274
+ "epoch": 1.8905109489051095,
2275
+ "grad_norm": 0.546875,
2276
+ "learning_rate": 3.988892839138731e-06,
2277
+ "loss": 1.185132622718811,
2278
+ "step": 648
2279
+ },
2280
+ {
2281
+ "epoch": 1.8963503649635036,
2282
+ "grad_norm": 0.83203125,
2283
+ "learning_rate": 3.96166478681355e-06,
2284
+ "loss": 1.221949815750122,
2285
+ "step": 650
2286
+ },
2287
+ {
2288
+ "epoch": 1.9021897810218977,
2289
+ "grad_norm": 0.63671875,
2290
+ "learning_rate": 3.934500358360085e-06,
2291
+ "loss": 1.2973777055740356,
2292
+ "step": 652
2293
+ },
2294
+ {
2295
+ "epoch": 1.908029197080292,
2296
+ "grad_norm": 0.373046875,
2297
+ "learning_rate": 3.907400677269563e-06,
2298
+ "loss": 1.2753914594650269,
2299
+ "step": 654
2300
+ },
2301
+ {
2302
+ "epoch": 1.9138686131386862,
2303
+ "grad_norm": 0.4609375,
2304
+ "learning_rate": 3.880366864355325e-06,
2305
+ "loss": 1.2708187103271484,
2306
+ "step": 656
2307
+ },
2308
+ {
2309
+ "epoch": 1.9197080291970803,
2310
+ "grad_norm": 0.380859375,
2311
+ "learning_rate": 3.853400037706478e-06,
2312
+ "loss": 1.2944742441177368,
2313
+ "step": 658
2314
+ },
2315
+ {
2316
+ "epoch": 1.9255474452554746,
2317
+ "grad_norm": 0.41015625,
2318
+ "learning_rate": 3.826501312641655e-06,
2319
+ "loss": 1.2255587577819824,
2320
+ "step": 660
2321
+ },
2322
+ {
2323
+ "epoch": 1.9313868613138685,
2324
+ "grad_norm": 1.4609375,
2325
+ "learning_rate": 3.7996718016628697e-06,
2326
+ "loss": 1.168819546699524,
2327
+ "step": 662
2328
+ },
2329
+ {
2330
+ "epoch": 1.9372262773722628,
2331
+ "grad_norm": 0.62109375,
2332
+ "learning_rate": 3.772912614409526e-06,
2333
+ "loss": 1.141089916229248,
2334
+ "step": 664
2335
+ },
2336
+ {
2337
+ "epoch": 1.943065693430657,
2338
+ "grad_norm": 0.388671875,
2339
+ "learning_rate": 3.7462248576125125e-06,
2340
+ "loss": 1.2469316720962524,
2341
+ "step": 666
2342
+ },
2343
+ {
2344
+ "epoch": 1.948905109489051,
2345
+ "grad_norm": 0.53125,
2346
+ "learning_rate": 3.719609635048431e-06,
2347
+ "loss": 1.222388505935669,
2348
+ "step": 668
2349
+ },
2350
+ {
2351
+ "epoch": 1.9547445255474454,
2352
+ "grad_norm": 0.47265625,
2353
+ "learning_rate": 3.6930680474939486e-06,
2354
+ "loss": 1.1344715356826782,
2355
+ "step": 670
2356
+ },
2357
+ {
2358
+ "epoch": 1.9605839416058393,
2359
+ "grad_norm": 0.5625,
2360
+ "learning_rate": 3.6666011926802647e-06,
2361
+ "loss": 1.2021689414978027,
2362
+ "step": 672
2363
+ },
2364
+ {
2365
+ "epoch": 1.9664233576642336,
2366
+ "grad_norm": 0.51171875,
2367
+ "learning_rate": 3.6402101652477163e-06,
2368
+ "loss": 1.251044750213623,
2369
+ "step": 674
2370
+ },
2371
+ {
2372
+ "epoch": 1.9722627737226277,
2373
+ "grad_norm": 0.578125,
2374
+ "learning_rate": 3.613896056700502e-06,
2375
+ "loss": 1.258431315422058,
2376
+ "step": 676
2377
+ },
2378
+ {
2379
+ "epoch": 1.9781021897810218,
2380
+ "grad_norm": 0.51171875,
2381
+ "learning_rate": 3.58765995536154e-06,
2382
+ "loss": 1.2117798328399658,
2383
+ "step": 678
2384
+ },
2385
+ {
2386
+ "epoch": 1.9839416058394161,
2387
+ "grad_norm": 1.390625,
2388
+ "learning_rate": 3.561502946327452e-06,
2389
+ "loss": 1.2486507892608643,
2390
+ "step": 680
2391
+ },
2392
+ {
2393
+ "epoch": 1.9897810218978103,
2394
+ "grad_norm": 0.73046875,
2395
+ "learning_rate": 3.535426111423691e-06,
2396
+ "loss": 1.1538894176483154,
2397
+ "step": 682
2398
+ },
2399
+ {
2400
+ "epoch": 1.9956204379562044,
2401
+ "grad_norm": 0.5703125,
2402
+ "learning_rate": 3.509430529159794e-06,
2403
+ "loss": 1.2577342987060547,
2404
+ "step": 684
2405
+ },
2406
+ {
2407
+ "epoch": 2.0,
2408
+ "grad_norm": 1.2421875,
2409
+ "learning_rate": 3.483517274684778e-06,
2410
+ "loss": 1.3043779134750366,
2411
+ "step": 686
2412
+ },
2413
+ {
2414
+ "epoch": 2.0058394160583943,
2415
+ "grad_norm": 0.8359375,
2416
+ "learning_rate": 3.457687419742668e-06,
2417
+ "loss": 1.1992483139038086,
2418
+ "step": 688
2419
+ },
2420
+ {
2421
+ "epoch": 2.011678832116788,
2422
+ "grad_norm": 0.42578125,
2423
+ "learning_rate": 3.431942032628184e-06,
2424
+ "loss": 1.2118330001831055,
2425
+ "step": 690
2426
+ },
2427
+ {
2428
+ "epoch": 2.0175182481751825,
2429
+ "grad_norm": 0.34375,
2430
+ "learning_rate": 3.4062821781425402e-06,
2431
+ "loss": 1.1599669456481934,
2432
+ "step": 692
2433
+ },
2434
+ {
2435
+ "epoch": 2.0233576642335764,
2436
+ "grad_norm": 0.30078125,
2437
+ "learning_rate": 3.3807089175494175e-06,
2438
+ "loss": 1.1471362113952637,
2439
+ "step": 694
2440
+ },
2441
+ {
2442
+ "epoch": 2.0291970802919708,
2443
+ "grad_norm": 0.421875,
2444
+ "learning_rate": 3.355223308531066e-06,
2445
+ "loss": 1.3080707788467407,
2446
+ "step": 696
2447
+ },
2448
+ {
2449
+ "epoch": 2.035036496350365,
2450
+ "grad_norm": 0.55859375,
2451
+ "learning_rate": 3.3298264051445655e-06,
2452
+ "loss": 1.2860289812088013,
2453
+ "step": 698
2454
+ },
2455
+ {
2456
+ "epoch": 2.040875912408759,
2457
+ "grad_norm": 0.4921875,
2458
+ "learning_rate": 3.3045192577782214e-06,
2459
+ "loss": 1.1995092630386353,
2460
+ "step": 700
2461
+ },
2462
+ {
2463
+ "epoch": 2.0467153284671533,
2464
+ "grad_norm": 0.5078125,
2465
+ "learning_rate": 3.2793029131081335e-06,
2466
+ "loss": 1.1927409172058105,
2467
+ "step": 702
2468
+ },
2469
+ {
2470
+ "epoch": 2.0525547445255476,
2471
+ "grad_norm": 0.38671875,
2472
+ "learning_rate": 3.254178414054896e-06,
2473
+ "loss": 1.1319977045059204,
2474
+ "step": 704
2475
+ },
2476
+ {
2477
+ "epoch": 2.0583941605839415,
2478
+ "grad_norm": 0.41015625,
2479
+ "learning_rate": 3.2291467997404747e-06,
2480
+ "loss": 1.238900899887085,
2481
+ "step": 706
2482
+ },
2483
+ {
2484
+ "epoch": 2.064233576642336,
2485
+ "grad_norm": 0.439453125,
2486
+ "learning_rate": 3.2042091054452175e-06,
2487
+ "loss": 1.1581428050994873,
2488
+ "step": 708
2489
+ },
2490
+ {
2491
+ "epoch": 2.0700729927007298,
2492
+ "grad_norm": 0.31640625,
2493
+ "learning_rate": 3.1793663625650444e-06,
2494
+ "loss": 1.2304531335830688,
2495
+ "step": 710
2496
+ },
2497
+ {
2498
+ "epoch": 2.075912408759124,
2499
+ "grad_norm": 0.37109375,
2500
+ "learning_rate": 3.154619598568789e-06,
2501
+ "loss": 1.2318830490112305,
2502
+ "step": 712
2503
+ },
2504
+ {
2505
+ "epoch": 2.0817518248175184,
2506
+ "grad_norm": 0.52734375,
2507
+ "learning_rate": 3.1299698369557026e-06,
2508
+ "loss": 1.2299096584320068,
2509
+ "step": 714
2510
+ },
2511
+ {
2512
+ "epoch": 2.0875912408759123,
2513
+ "grad_norm": 0.52734375,
2514
+ "learning_rate": 3.105418097213121e-06,
2515
+ "loss": 1.154068112373352,
2516
+ "step": 716
2517
+ },
2518
+ {
2519
+ "epoch": 2.0934306569343066,
2520
+ "grad_norm": 0.41796875,
2521
+ "learning_rate": 3.0809653947743044e-06,
2522
+ "loss": 1.0612831115722656,
2523
+ "step": 718
2524
+ },
2525
+ {
2526
+ "epoch": 2.0992700729927005,
2527
+ "grad_norm": 0.298828125,
2528
+ "learning_rate": 3.0566127409764377e-06,
2529
+ "loss": 1.154129147529602,
2530
+ "step": 720
2531
+ },
2532
+ {
2533
+ "epoch": 2.105109489051095,
2534
+ "grad_norm": 0.7890625,
2535
+ "learning_rate": 3.0323611430188026e-06,
2536
+ "loss": 1.1408308744430542,
2537
+ "step": 722
2538
+ },
2539
+ {
2540
+ "epoch": 2.110948905109489,
2541
+ "grad_norm": 0.734375,
2542
+ "learning_rate": 3.008211603921118e-06,
2543
+ "loss": 1.1843266487121582,
2544
+ "step": 724
2545
+ },
2546
+ {
2547
+ "epoch": 2.116788321167883,
2548
+ "grad_norm": 0.41015625,
2549
+ "learning_rate": 2.9841651224820656e-06,
2550
+ "loss": 1.0983469486236572,
2551
+ "step": 726
2552
+ },
2553
+ {
2554
+ "epoch": 2.1226277372262774,
2555
+ "grad_norm": 0.361328125,
2556
+ "learning_rate": 2.96022269323797e-06,
2557
+ "loss": 1.2294151782989502,
2558
+ "step": 728
2559
+ },
2560
+ {
2561
+ "epoch": 2.1284671532846717,
2562
+ "grad_norm": 0.400390625,
2563
+ "learning_rate": 2.9363853064216706e-06,
2564
+ "loss": 1.3076847791671753,
2565
+ "step": 730
2566
+ },
2567
+ {
2568
+ "epoch": 2.1343065693430656,
2569
+ "grad_norm": 0.61328125,
2570
+ "learning_rate": 2.912653947921567e-06,
2571
+ "loss": 1.185577392578125,
2572
+ "step": 732
2573
+ },
2574
+ {
2575
+ "epoch": 2.14014598540146,
2576
+ "grad_norm": 0.408203125,
2577
+ "learning_rate": 2.8890295992408425e-06,
2578
+ "loss": 1.1940393447875977,
2579
+ "step": 734
2580
+ },
2581
+ {
2582
+ "epoch": 2.145985401459854,
2583
+ "grad_norm": 0.4296875,
2584
+ "learning_rate": 2.86551323745687e-06,
2585
+ "loss": 1.1575770378112793,
2586
+ "step": 736
2587
+ },
2588
+ {
2589
+ "epoch": 2.151824817518248,
2590
+ "grad_norm": 0.35546875,
2591
+ "learning_rate": 2.8421058351808055e-06,
2592
+ "loss": 1.1326178312301636,
2593
+ "step": 738
2594
+ },
2595
+ {
2596
+ "epoch": 2.1576642335766425,
2597
+ "grad_norm": 0.6171875,
2598
+ "learning_rate": 2.818808360517349e-06,
2599
+ "loss": 1.167670726776123,
2600
+ "step": 740
2601
+ },
2602
+ {
2603
+ "epoch": 2.1635036496350364,
2604
+ "grad_norm": 0.482421875,
2605
+ "learning_rate": 2.7956217770247262e-06,
2606
+ "loss": 1.3139435052871704,
2607
+ "step": 742
2608
+ },
2609
+ {
2610
+ "epoch": 2.1693430656934307,
2611
+ "grad_norm": 0.46875,
2612
+ "learning_rate": 2.7725470436748165e-06,
2613
+ "loss": 1.12082839012146,
2614
+ "step": 744
2615
+ },
2616
+ {
2617
+ "epoch": 2.1751824817518246,
2618
+ "grad_norm": 0.53125,
2619
+ "learning_rate": 2.7495851148135005e-06,
2620
+ "loss": 1.1051329374313354,
2621
+ "step": 746
2622
+ },
2623
+ {
2624
+ "epoch": 2.181021897810219,
2625
+ "grad_norm": 0.93359375,
2626
+ "learning_rate": 2.7267369401211895e-06,
2627
+ "loss": 1.1831530332565308,
2628
+ "step": 748
2629
+ },
2630
+ {
2631
+ "epoch": 2.1868613138686133,
2632
+ "grad_norm": 0.36328125,
2633
+ "learning_rate": 2.704003464573544e-06,
2634
+ "loss": 1.1686221361160278,
2635
+ "step": 750
2636
+ },
2637
+ {
2638
+ "epoch": 2.192700729927007,
2639
+ "grad_norm": 0.326171875,
2640
+ "learning_rate": 2.6813856284023943e-06,
2641
+ "loss": 1.1672006845474243,
2642
+ "step": 752
2643
+ },
2644
+ {
2645
+ "epoch": 2.1985401459854015,
2646
+ "grad_norm": 0.404296875,
2647
+ "learning_rate": 2.6588843670568505e-06,
2648
+ "loss": 1.254637360572815,
2649
+ "step": 754
2650
+ },
2651
+ {
2652
+ "epoch": 2.204379562043796,
2653
+ "grad_norm": 0.52734375,
2654
+ "learning_rate": 2.6365006111646152e-06,
2655
+ "loss": 1.1789019107818604,
2656
+ "step": 756
2657
+ },
2658
+ {
2659
+ "epoch": 2.2102189781021897,
2660
+ "grad_norm": 0.474609375,
2661
+ "learning_rate": 2.614235286493494e-06,
2662
+ "loss": 1.1955108642578125,
2663
+ "step": 758
2664
+ },
2665
+ {
2666
+ "epoch": 2.216058394160584,
2667
+ "grad_norm": 0.87109375,
2668
+ "learning_rate": 2.5920893139131043e-06,
2669
+ "loss": 1.2021061182022095,
2670
+ "step": 760
2671
+ },
2672
+ {
2673
+ "epoch": 2.221897810218978,
2674
+ "grad_norm": 0.451171875,
2675
+ "learning_rate": 2.570063609356791e-06,
2676
+ "loss": 1.2603737115859985,
2677
+ "step": 762
2678
+ },
2679
+ {
2680
+ "epoch": 2.2277372262773723,
2681
+ "grad_norm": 0.431640625,
2682
+ "learning_rate": 2.5481590837837477e-06,
2683
+ "loss": 1.1586928367614746,
2684
+ "step": 764
2685
+ },
2686
+ {
2687
+ "epoch": 2.2335766423357666,
2688
+ "grad_norm": 0.443359375,
2689
+ "learning_rate": 2.5263766431413327e-06,
2690
+ "loss": 1.3295210599899292,
2691
+ "step": 766
2692
+ },
2693
+ {
2694
+ "epoch": 2.2394160583941605,
2695
+ "grad_norm": 0.69140625,
2696
+ "learning_rate": 2.5047171883276065e-06,
2697
+ "loss": 1.116708517074585,
2698
+ "step": 768
2699
+ },
2700
+ {
2701
+ "epoch": 2.245255474452555,
2702
+ "grad_norm": 0.37890625,
2703
+ "learning_rate": 2.4831816151540682e-06,
2704
+ "loss": 1.0877715349197388,
2705
+ "step": 770
2706
+ },
2707
+ {
2708
+ "epoch": 2.2510948905109487,
2709
+ "grad_norm": 0.49609375,
2710
+ "learning_rate": 2.4617708143086085e-06,
2711
+ "loss": 1.3073028326034546,
2712
+ "step": 772
2713
+ },
2714
+ {
2715
+ "epoch": 2.256934306569343,
2716
+ "grad_norm": 0.7578125,
2717
+ "learning_rate": 2.440485671318668e-06,
2718
+ "loss": 1.2231556177139282,
2719
+ "step": 774
2720
+ },
2721
+ {
2722
+ "epoch": 2.2627737226277373,
2723
+ "grad_norm": 0.796875,
2724
+ "learning_rate": 2.41932706651462e-06,
2725
+ "loss": 1.1234259605407715,
2726
+ "step": 776
2727
+ },
2728
+ {
2729
+ "epoch": 2.2686131386861312,
2730
+ "grad_norm": 0.443359375,
2731
+ "learning_rate": 2.398295874993347e-06,
2732
+ "loss": 1.1716923713684082,
2733
+ "step": 778
2734
+ },
2735
+ {
2736
+ "epoch": 2.2744525547445256,
2737
+ "grad_norm": 0.65625,
2738
+ "learning_rate": 2.3773929665820662e-06,
2739
+ "loss": 1.2562777996063232,
2740
+ "step": 780
2741
+ },
2742
+ {
2743
+ "epoch": 2.28029197080292,
2744
+ "grad_norm": 0.458984375,
2745
+ "learning_rate": 2.35661920580234e-06,
2746
+ "loss": 1.2565009593963623,
2747
+ "step": 782
2748
+ },
2749
+ {
2750
+ "epoch": 2.286131386861314,
2751
+ "grad_norm": 1.140625,
2752
+ "learning_rate": 2.3359754518343255e-06,
2753
+ "loss": 1.2188069820404053,
2754
+ "step": 784
2755
+ },
2756
+ {
2757
+ "epoch": 2.291970802919708,
2758
+ "grad_norm": 1.3125,
2759
+ "learning_rate": 2.315462558481241e-06,
2760
+ "loss": 1.1537082195281982,
2761
+ "step": 786
2762
+ },
2763
+ {
2764
+ "epoch": 2.297810218978102,
2765
+ "grad_norm": 0.357421875,
2766
+ "learning_rate": 2.29508137413405e-06,
2767
+ "loss": 1.0930520296096802,
2768
+ "step": 788
2769
+ },
2770
+ {
2771
+ "epoch": 2.3036496350364963,
2772
+ "grad_norm": 0.5078125,
2773
+ "learning_rate": 2.274832741736376e-06,
2774
+ "loss": 1.107776165008545,
2775
+ "step": 790
2776
+ },
2777
+ {
2778
+ "epoch": 2.3094890510948907,
2779
+ "grad_norm": 0.40234375,
2780
+ "learning_rate": 2.254717498749638e-06,
2781
+ "loss": 1.291731357574463,
2782
+ "step": 792
2783
+ },
2784
+ {
2785
+ "epoch": 2.3153284671532846,
2786
+ "grad_norm": 0.396484375,
2787
+ "learning_rate": 2.234736477118413e-06,
2788
+ "loss": 1.1373404264450073,
2789
+ "step": 794
2790
+ },
2791
+ {
2792
+ "epoch": 2.321167883211679,
2793
+ "grad_norm": 0.380859375,
2794
+ "learning_rate": 2.2148905032360305e-06,
2795
+ "loss": 1.1539984941482544,
2796
+ "step": 796
2797
+ },
2798
+ {
2799
+ "epoch": 2.3270072992700728,
2800
+ "grad_norm": 0.46875,
2801
+ "learning_rate": 2.195180397910389e-06,
2802
+ "loss": 1.1758666038513184,
2803
+ "step": 798
2804
+ },
2805
+ {
2806
+ "epoch": 2.332846715328467,
2807
+ "grad_norm": 1.8984375,
2808
+ "learning_rate": 2.1756069763300158e-06,
2809
+ "loss": 1.2303135395050049,
2810
+ "step": 800
2811
+ },
2812
+ {
2813
+ "epoch": 2.3386861313868614,
2814
+ "grad_norm": 0.546875,
2815
+ "learning_rate": 2.1561710480303435e-06,
2816
+ "loss": 1.1783814430236816,
2817
+ "step": 802
2818
+ },
2819
+ {
2820
+ "epoch": 2.3445255474452553,
2821
+ "grad_norm": 0.345703125,
2822
+ "learning_rate": 2.1368734168602318e-06,
2823
+ "loss": 1.2231682538986206,
2824
+ "step": 804
2825
+ },
2826
+ {
2827
+ "epoch": 2.3503649635036497,
2828
+ "grad_norm": 0.4453125,
2829
+ "learning_rate": 2.1177148809487237e-06,
2830
+ "loss": 1.1812344789505005,
2831
+ "step": 806
2832
+ },
2833
+ {
2834
+ "epoch": 2.356204379562044,
2835
+ "grad_norm": 1.5546875,
2836
+ "learning_rate": 2.0986962326720328e-06,
2837
+ "loss": 1.1925325393676758,
2838
+ "step": 808
2839
+ },
2840
+ {
2841
+ "epoch": 2.362043795620438,
2842
+ "grad_norm": 0.4140625,
2843
+ "learning_rate": 2.079818258620771e-06,
2844
+ "loss": 1.1922979354858398,
2845
+ "step": 810
2846
+ },
2847
+ {
2848
+ "epoch": 2.367883211678832,
2849
+ "grad_norm": 1.7578125,
2850
+ "learning_rate": 2.0610817395674197e-06,
2851
+ "loss": 1.2363067865371704,
2852
+ "step": 812
2853
+ },
2854
+ {
2855
+ "epoch": 2.373722627737226,
2856
+ "grad_norm": 0.41015625,
2857
+ "learning_rate": 2.042487450434033e-06,
2858
+ "loss": 1.1505279541015625,
2859
+ "step": 814
2860
+ },
2861
+ {
2862
+ "epoch": 2.3795620437956204,
2863
+ "grad_norm": 0.765625,
2864
+ "learning_rate": 2.0240361602601906e-06,
2865
+ "loss": 1.1627535820007324,
2866
+ "step": 816
2867
+ },
2868
+ {
2869
+ "epoch": 2.3854014598540147,
2870
+ "grad_norm": 0.33203125,
2871
+ "learning_rate": 2.0057286321711924e-06,
2872
+ "loss": 1.155818223953247,
2873
+ "step": 818
2874
+ },
2875
+ {
2876
+ "epoch": 2.3912408759124086,
2877
+ "grad_norm": 0.5078125,
2878
+ "learning_rate": 1.987565623346492e-06,
2879
+ "loss": 1.2554137706756592,
2880
+ "step": 820
2881
+ },
2882
+ {
2883
+ "epoch": 2.397080291970803,
2884
+ "grad_norm": 0.546875,
2885
+ "learning_rate": 1.9695478849883867e-06,
2886
+ "loss": 1.1215136051177979,
2887
+ "step": 822
2888
+ },
2889
+ {
2890
+ "epoch": 2.402919708029197,
2891
+ "grad_norm": 0.3515625,
2892
+ "learning_rate": 1.9516761622909423e-06,
2893
+ "loss": 1.1090680360794067,
2894
+ "step": 824
2895
+ },
2896
+ {
2897
+ "epoch": 2.408759124087591,
2898
+ "grad_norm": 0.416015625,
2899
+ "learning_rate": 1.9339511944091773e-06,
2900
+ "loss": 1.1436755657196045,
2901
+ "step": 826
2902
+ },
2903
+ {
2904
+ "epoch": 2.4145985401459855,
2905
+ "grad_norm": 0.48046875,
2906
+ "learning_rate": 1.91637371442849e-06,
2907
+ "loss": 1.140608549118042,
2908
+ "step": 828
2909
+ },
2910
+ {
2911
+ "epoch": 2.4204379562043794,
2912
+ "grad_norm": 0.54296875,
2913
+ "learning_rate": 1.8989444493343401e-06,
2914
+ "loss": 1.2412205934524536,
2915
+ "step": 830
2916
+ },
2917
+ {
2918
+ "epoch": 2.4262773722627737,
2919
+ "grad_norm": 0.71484375,
2920
+ "learning_rate": 1.8816641199821797e-06,
2921
+ "loss": 1.2433428764343262,
2922
+ "step": 832
2923
+ },
2924
+ {
2925
+ "epoch": 2.432116788321168,
2926
+ "grad_norm": 0.4140625,
2927
+ "learning_rate": 1.8645334410676413e-06,
2928
+ "loss": 1.2072176933288574,
2929
+ "step": 834
2930
+ },
2931
+ {
2932
+ "epoch": 2.437956204379562,
2933
+ "grad_norm": 0.5703125,
2934
+ "learning_rate": 1.8475531210969766e-06,
2935
+ "loss": 1.174179196357727,
2936
+ "step": 836
2937
+ },
2938
+ {
2939
+ "epoch": 2.4437956204379563,
2940
+ "grad_norm": 0.87890625,
2941
+ "learning_rate": 1.8307238623577588e-06,
2942
+ "loss": 1.2336622476577759,
2943
+ "step": 838
2944
+ },
2945
+ {
2946
+ "epoch": 2.44963503649635,
2947
+ "grad_norm": 0.40625,
2948
+ "learning_rate": 1.814046360889829e-06,
2949
+ "loss": 1.2075152397155762,
2950
+ "step": 840
2951
+ },
2952
+ {
2953
+ "epoch": 2.4554744525547445,
2954
+ "grad_norm": 0.3515625,
2955
+ "learning_rate": 1.7975213064565136e-06,
2956
+ "loss": 1.1945725679397583,
2957
+ "step": 842
2958
+ },
2959
+ {
2960
+ "epoch": 2.461313868613139,
2961
+ "grad_norm": 0.5703125,
2962
+ "learning_rate": 1.7811493825160952e-06,
2963
+ "loss": 1.255374789237976,
2964
+ "step": 844
2965
+ },
2966
+ {
2967
+ "epoch": 2.4671532846715327,
2968
+ "grad_norm": 0.326171875,
2969
+ "learning_rate": 1.7649312661935463e-06,
2970
+ "loss": 1.0935304164886475,
2971
+ "step": 846
2972
+ },
2973
+ {
2974
+ "epoch": 2.472992700729927,
2975
+ "grad_norm": 1.2109375,
2976
+ "learning_rate": 1.7488676282525236e-06,
2977
+ "loss": 1.1871508359909058,
2978
+ "step": 848
2979
+ },
2980
+ {
2981
+ "epoch": 2.478832116788321,
2982
+ "grad_norm": 0.515625,
2983
+ "learning_rate": 1.7329591330676255e-06,
2984
+ "loss": 1.236677646636963,
2985
+ "step": 850
2986
+ },
2987
+ {
2988
+ "epoch": 2.4846715328467153,
2989
+ "grad_norm": 0.59375,
2990
+ "learning_rate": 1.7172064385969144e-06,
2991
+ "loss": 1.1770851612091064,
2992
+ "step": 852
2993
+ },
2994
+ {
2995
+ "epoch": 2.4905109489051096,
2996
+ "grad_norm": 0.76953125,
2997
+ "learning_rate": 1.7016101963547063e-06,
2998
+ "loss": 1.1361277103424072,
2999
+ "step": 854
3000
+ },
3001
+ {
3002
+ "epoch": 2.4963503649635035,
3003
+ "grad_norm": 0.5078125,
3004
+ "learning_rate": 1.68617105138462e-06,
3005
+ "loss": 1.1203017234802246,
3006
+ "step": 856
3007
+ },
3008
+ {
3009
+ "epoch": 2.502189781021898,
3010
+ "grad_norm": 0.453125,
3011
+ "learning_rate": 1.6708896422329056e-06,
3012
+ "loss": 1.2030003070831299,
3013
+ "step": 858
3014
+ },
3015
+ {
3016
+ "epoch": 2.508029197080292,
3017
+ "grad_norm": 0.5234375,
3018
+ "learning_rate": 1.6557666009220274e-06,
3019
+ "loss": 1.194663405418396,
3020
+ "step": 860
3021
+ },
3022
+ {
3023
+ "epoch": 2.513868613138686,
3024
+ "grad_norm": 0.74609375,
3025
+ "learning_rate": 1.6408025529245324e-06,
3026
+ "loss": 1.3168246746063232,
3027
+ "step": 862
3028
+ },
3029
+ {
3030
+ "epoch": 2.5197080291970804,
3031
+ "grad_norm": 0.43359375,
3032
+ "learning_rate": 1.6259981171371736e-06,
3033
+ "loss": 1.1622366905212402,
3034
+ "step": 864
3035
+ },
3036
+ {
3037
+ "epoch": 2.5255474452554747,
3038
+ "grad_norm": 0.333984375,
3039
+ "learning_rate": 1.611353905855317e-06,
3040
+ "loss": 1.1341776847839355,
3041
+ "step": 866
3042
+ },
3043
+ {
3044
+ "epoch": 2.5313868613138686,
3045
+ "grad_norm": 0.47265625,
3046
+ "learning_rate": 1.5968705247476192e-06,
3047
+ "loss": 1.1897858381271362,
3048
+ "step": 868
3049
+ },
3050
+ {
3051
+ "epoch": 2.537226277372263,
3052
+ "grad_norm": 0.486328125,
3053
+ "learning_rate": 1.582548572830973e-06,
3054
+ "loss": 1.1871055364608765,
3055
+ "step": 870
3056
+ },
3057
+ {
3058
+ "epoch": 2.543065693430657,
3059
+ "grad_norm": 0.55859375,
3060
+ "learning_rate": 1.5683886424457373e-06,
3061
+ "loss": 1.1016008853912354,
3062
+ "step": 872
3063
+ },
3064
+ {
3065
+ "epoch": 2.548905109489051,
3066
+ "grad_norm": 0.7734375,
3067
+ "learning_rate": 1.5543913192312373e-06,
3068
+ "loss": 1.1918286085128784,
3069
+ "step": 874
3070
+ },
3071
+ {
3072
+ "epoch": 2.554744525547445,
3073
+ "grad_norm": 0.5078125,
3074
+ "learning_rate": 1.5405571821015402e-06,
3075
+ "loss": 1.2634363174438477,
3076
+ "step": 876
3077
+ },
3078
+ {
3079
+ "epoch": 2.5605839416058394,
3080
+ "grad_norm": 0.376953125,
3081
+ "learning_rate": 1.5268868032215152e-06,
3082
+ "loss": 1.1621180772781372,
3083
+ "step": 878
3084
+ },
3085
+ {
3086
+ "epoch": 2.5664233576642337,
3087
+ "grad_norm": 0.4765625,
3088
+ "learning_rate": 1.513380747983169e-06,
3089
+ "loss": 1.1664128303527832,
3090
+ "step": 880
3091
+ },
3092
+ {
3093
+ "epoch": 2.5722627737226276,
3094
+ "grad_norm": 0.609375,
3095
+ "learning_rate": 1.5000395749822597e-06,
3096
+ "loss": 1.1880872249603271,
3097
+ "step": 882
3098
+ },
3099
+ {
3100
+ "epoch": 2.578102189781022,
3101
+ "grad_norm": 0.4921875,
3102
+ "learning_rate": 1.4868638359951963e-06,
3103
+ "loss": 1.2146211862564087,
3104
+ "step": 884
3105
+ },
3106
+ {
3107
+ "epoch": 2.5839416058394162,
3108
+ "grad_norm": 0.55859375,
3109
+ "learning_rate": 1.4738540759562175e-06,
3110
+ "loss": 1.1511411666870117,
3111
+ "step": 886
3112
+ },
3113
+ {
3114
+ "epoch": 2.58978102189781,
3115
+ "grad_norm": 0.515625,
3116
+ "learning_rate": 1.4610108329348514e-06,
3117
+ "loss": 1.2108086347579956,
3118
+ "step": 888
3119
+ },
3120
+ {
3121
+ "epoch": 2.5956204379562045,
3122
+ "grad_norm": 0.400390625,
3123
+ "learning_rate": 1.4483346381136653e-06,
3124
+ "loss": 1.0855056047439575,
3125
+ "step": 890
3126
+ },
3127
+ {
3128
+ "epoch": 2.601459854014599,
3129
+ "grad_norm": 0.3671875,
3130
+ "learning_rate": 1.4358260157662928e-06,
3131
+ "loss": 1.1662517786026,
3132
+ "step": 892
3133
+ },
3134
+ {
3135
+ "epoch": 2.6072992700729927,
3136
+ "grad_norm": 0.79296875,
3137
+ "learning_rate": 1.4234854832357536e-06,
3138
+ "loss": 1.195744514465332,
3139
+ "step": 894
3140
+ },
3141
+ {
3142
+ "epoch": 2.613138686131387,
3143
+ "grad_norm": 0.53515625,
3144
+ "learning_rate": 1.4113135509130547e-06,
3145
+ "loss": 1.2275428771972656,
3146
+ "step": 896
3147
+ },
3148
+ {
3149
+ "epoch": 2.618978102189781,
3150
+ "grad_norm": 0.50390625,
3151
+ "learning_rate": 1.3993107222160817e-06,
3152
+ "loss": 1.2926044464111328,
3153
+ "step": 898
3154
+ },
3155
+ {
3156
+ "epoch": 2.624817518248175,
3157
+ "grad_norm": 0.60546875,
3158
+ "learning_rate": 1.3874774935687782e-06,
3159
+ "loss": 1.2193105220794678,
3160
+ "step": 900
3161
+ },
3162
+ {
3163
+ "epoch": 2.630656934306569,
3164
+ "grad_norm": 0.5859375,
3165
+ "learning_rate": 1.3758143543806143e-06,
3166
+ "loss": 1.227401852607727,
3167
+ "step": 902
3168
+ },
3169
+ {
3170
+ "epoch": 2.6364963503649634,
3171
+ "grad_norm": 0.86328125,
3172
+ "learning_rate": 1.3643217870263448e-06,
3173
+ "loss": 1.2654823064804077,
3174
+ "step": 904
3175
+ },
3176
+ {
3177
+ "epoch": 2.6423357664233578,
3178
+ "grad_norm": 0.373046875,
3179
+ "learning_rate": 1.35300026682606e-06,
3180
+ "loss": 1.1640398502349854,
3181
+ "step": 906
3182
+ },
3183
+ {
3184
+ "epoch": 2.6481751824817517,
3185
+ "grad_norm": 0.625,
3186
+ "learning_rate": 1.341850262025524e-06,
3187
+ "loss": 1.1758366823196411,
3188
+ "step": 908
3189
+ },
3190
+ {
3191
+ "epoch": 2.654014598540146,
3192
+ "grad_norm": 0.462890625,
3193
+ "learning_rate": 1.330872233776811e-06,
3194
+ "loss": 1.1808586120605469,
3195
+ "step": 910
3196
+ },
3197
+ {
3198
+ "epoch": 2.6598540145985403,
3199
+ "grad_norm": 0.458984375,
3200
+ "learning_rate": 1.3200666361192349e-06,
3201
+ "loss": 1.1935012340545654,
3202
+ "step": 912
3203
+ },
3204
+ {
3205
+ "epoch": 2.665693430656934,
3206
+ "grad_norm": 0.396484375,
3207
+ "learning_rate": 1.3094339159605627e-06,
3208
+ "loss": 1.1293021440505981,
3209
+ "step": 914
3210
+ },
3211
+ {
3212
+ "epoch": 2.6715328467153285,
3213
+ "grad_norm": 0.875,
3214
+ "learning_rate": 1.2989745130585407e-06,
3215
+ "loss": 1.2715429067611694,
3216
+ "step": 916
3217
+ },
3218
+ {
3219
+ "epoch": 2.677372262773723,
3220
+ "grad_norm": 0.462890625,
3221
+ "learning_rate": 1.2886888600026983e-06,
3222
+ "loss": 1.1346051692962646,
3223
+ "step": 918
3224
+ },
3225
+ {
3226
+ "epoch": 2.6832116788321168,
3227
+ "grad_norm": 0.412109375,
3228
+ "learning_rate": 1.2785773821964636e-06,
3229
+ "loss": 1.2460830211639404,
3230
+ "step": 920
3231
+ },
3232
+ {
3233
+ "epoch": 2.689051094890511,
3234
+ "grad_norm": 1.6171875,
3235
+ "learning_rate": 1.2686404978395626e-06,
3236
+ "loss": 1.1654274463653564,
3237
+ "step": 922
3238
+ },
3239
+ {
3240
+ "epoch": 2.694890510948905,
3241
+ "grad_norm": 1.65625,
3242
+ "learning_rate": 1.2588786179107279e-06,
3243
+ "loss": 1.253066897392273,
3244
+ "step": 924
3245
+ },
3246
+ {
3247
+ "epoch": 2.7007299270072993,
3248
+ "grad_norm": 0.42578125,
3249
+ "learning_rate": 1.2492921461506994e-06,
3250
+ "loss": 1.188564658164978,
3251
+ "step": 926
3252
+ },
3253
+ {
3254
+ "epoch": 2.706569343065693,
3255
+ "grad_norm": 0.466796875,
3256
+ "learning_rate": 1.2398814790455244e-06,
3257
+ "loss": 1.244341254234314,
3258
+ "step": 928
3259
+ },
3260
+ {
3261
+ "epoch": 2.7124087591240875,
3262
+ "grad_norm": 0.443359375,
3263
+ "learning_rate": 1.2306470058101612e-06,
3264
+ "loss": 1.1333363056182861,
3265
+ "step": 930
3266
+ },
3267
+ {
3268
+ "epoch": 2.718248175182482,
3269
+ "grad_norm": 0.66015625,
3270
+ "learning_rate": 1.2215891083723842e-06,
3271
+ "loss": 1.263460397720337,
3272
+ "step": 932
3273
+ },
3274
+ {
3275
+ "epoch": 2.7240875912408757,
3276
+ "grad_norm": 0.359375,
3277
+ "learning_rate": 1.2127081613569809e-06,
3278
+ "loss": 1.2082524299621582,
3279
+ "step": 934
3280
+ },
3281
+ {
3282
+ "epoch": 2.72992700729927,
3283
+ "grad_norm": 0.408203125,
3284
+ "learning_rate": 1.2040045320702626e-06,
3285
+ "loss": 1.1242833137512207,
3286
+ "step": 936
3287
+ },
3288
+ {
3289
+ "epoch": 2.7357664233576644,
3290
+ "grad_norm": 0.63671875,
3291
+ "learning_rate": 1.1954785804848739e-06,
3292
+ "loss": 1.14191734790802,
3293
+ "step": 938
3294
+ },
3295
+ {
3296
+ "epoch": 2.7416058394160583,
3297
+ "grad_norm": 1.65625,
3298
+ "learning_rate": 1.1871306592249008e-06,
3299
+ "loss": 1.1483235359191895,
3300
+ "step": 940
3301
+ },
3302
+ {
3303
+ "epoch": 2.7474452554744526,
3304
+ "grad_norm": 0.5546875,
3305
+ "learning_rate": 1.1789611135512903e-06,
3306
+ "loss": 1.1981879472732544,
3307
+ "step": 942
3308
+ },
3309
+ {
3310
+ "epoch": 2.753284671532847,
3311
+ "grad_norm": 1.171875,
3312
+ "learning_rate": 1.1709702813475674e-06,
3313
+ "loss": 1.1760236024856567,
3314
+ "step": 944
3315
+ },
3316
+ {
3317
+ "epoch": 2.759124087591241,
3318
+ "grad_norm": 0.380859375,
3319
+ "learning_rate": 1.163158493105864e-06,
3320
+ "loss": 1.2009891271591187,
3321
+ "step": 946
3322
+ },
3323
+ {
3324
+ "epoch": 2.764963503649635,
3325
+ "grad_norm": 0.65234375,
3326
+ "learning_rate": 1.1555260719132474e-06,
3327
+ "loss": 1.1685845851898193,
3328
+ "step": 948
3329
+ },
3330
+ {
3331
+ "epoch": 2.770802919708029,
3332
+ "grad_norm": 0.423828125,
3333
+ "learning_rate": 1.1480733334383599e-06,
3334
+ "loss": 1.2095947265625,
3335
+ "step": 950
3336
+ },
3337
+ {
3338
+ "epoch": 2.7766423357664234,
3339
+ "grad_norm": 0.357421875,
3340
+ "learning_rate": 1.1408005859183595e-06,
3341
+ "loss": 1.1101715564727783,
3342
+ "step": 952
3343
+ },
3344
+ {
3345
+ "epoch": 2.7824817518248173,
3346
+ "grad_norm": 0.69921875,
3347
+ "learning_rate": 1.1337081301461775e-06,
3348
+ "loss": 1.2515991926193237,
3349
+ "step": 954
3350
+ },
3351
+ {
3352
+ "epoch": 2.7883211678832116,
3353
+ "grad_norm": 1.9609375,
3354
+ "learning_rate": 1.1267962594580712e-06,
3355
+ "loss": 1.2288193702697754,
3356
+ "step": 956
3357
+ },
3358
+ {
3359
+ "epoch": 2.794160583941606,
3360
+ "grad_norm": 0.90234375,
3361
+ "learning_rate": 1.1200652597214982e-06,
3362
+ "loss": 1.1902008056640625,
3363
+ "step": 958
3364
+ },
3365
+ {
3366
+ "epoch": 2.8,
3367
+ "grad_norm": 0.458984375,
3368
+ "learning_rate": 1.1135154093232868e-06,
3369
+ "loss": 1.1392861604690552,
3370
+ "step": 960
3371
+ },
3372
+ {
3373
+ "epoch": 2.805839416058394,
3374
+ "grad_norm": 0.625,
3375
+ "learning_rate": 1.107146979158129e-06,
3376
+ "loss": 1.1379271745681763,
3377
+ "step": 962
3378
+ },
3379
+ {
3380
+ "epoch": 2.8116788321167885,
3381
+ "grad_norm": 0.57421875,
3382
+ "learning_rate": 1.100960232617371e-06,
3383
+ "loss": 1.274213433265686,
3384
+ "step": 964
3385
+ },
3386
+ {
3387
+ "epoch": 2.8175182481751824,
3388
+ "grad_norm": 0.984375,
3389
+ "learning_rate": 1.094955425578123e-06,
3390
+ "loss": 1.2406063079833984,
3391
+ "step": 966
3392
+ },
3393
+ {
3394
+ "epoch": 2.8233576642335767,
3395
+ "grad_norm": 1.0859375,
3396
+ "learning_rate": 1.0891328063926742e-06,
3397
+ "loss": 1.1469717025756836,
3398
+ "step": 968
3399
+ },
3400
+ {
3401
+ "epoch": 2.829197080291971,
3402
+ "grad_norm": 0.3046875,
3403
+ "learning_rate": 1.0834926158782235e-06,
3404
+ "loss": 1.2098904848098755,
3405
+ "step": 970
3406
+ },
3407
+ {
3408
+ "epoch": 2.835036496350365,
3409
+ "grad_norm": 0.640625,
3410
+ "learning_rate": 1.0780350873069184e-06,
3411
+ "loss": 1.1746724843978882,
3412
+ "step": 972
3413
+ },
3414
+ {
3415
+ "epoch": 2.8408759124087593,
3416
+ "grad_norm": 0.37109375,
3417
+ "learning_rate": 1.0727604463962058e-06,
3418
+ "loss": 1.187951683998108,
3419
+ "step": 974
3420
+ },
3421
+ {
3422
+ "epoch": 2.846715328467153,
3423
+ "grad_norm": 0.546875,
3424
+ "learning_rate": 1.0676689112994977e-06,
3425
+ "loss": 1.1875951290130615,
3426
+ "step": 976
3427
+ },
3428
+ {
3429
+ "epoch": 2.8525547445255475,
3430
+ "grad_norm": 0.50390625,
3431
+ "learning_rate": 1.06276069259715e-06,
3432
+ "loss": 1.1863131523132324,
3433
+ "step": 978
3434
+ },
3435
+ {
3436
+ "epoch": 2.8583941605839414,
3437
+ "grad_norm": 0.3984375,
3438
+ "learning_rate": 1.0580359932877516e-06,
3439
+ "loss": 1.1864418983459473,
3440
+ "step": 980
3441
+ },
3442
+ {
3443
+ "epoch": 2.8642335766423357,
3444
+ "grad_norm": 0.66796875,
3445
+ "learning_rate": 1.0534950087797282e-06,
3446
+ "loss": 1.255091667175293,
3447
+ "step": 982
3448
+ },
3449
+ {
3450
+ "epoch": 2.87007299270073,
3451
+ "grad_norm": 0.357421875,
3452
+ "learning_rate": 1.049137926883261e-06,
3453
+ "loss": 1.1269922256469727,
3454
+ "step": 984
3455
+ },
3456
+ {
3457
+ "epoch": 2.875912408759124,
3458
+ "grad_norm": 0.64453125,
3459
+ "learning_rate": 1.0449649278025208e-06,
3460
+ "loss": 1.1925872564315796,
3461
+ "step": 986
3462
+ },
3463
+ {
3464
+ "epoch": 2.8817518248175182,
3465
+ "grad_norm": 0.3828125,
3466
+ "learning_rate": 1.0409761841282112e-06,
3467
+ "loss": 1.1946667432785034,
3468
+ "step": 988
3469
+ },
3470
+ {
3471
+ "epoch": 2.8875912408759126,
3472
+ "grad_norm": 0.408203125,
3473
+ "learning_rate": 1.037171860830434e-06,
3474
+ "loss": 1.141740083694458,
3475
+ "step": 990
3476
+ },
3477
+ {
3478
+ "epoch": 2.8934306569343065,
3479
+ "grad_norm": 0.4921875,
3480
+ "learning_rate": 1.0335521152518637e-06,
3481
+ "loss": 1.0821702480316162,
3482
+ "step": 992
3483
+ },
3484
+ {
3485
+ "epoch": 2.899270072992701,
3486
+ "grad_norm": 0.453125,
3487
+ "learning_rate": 1.0301170971012418e-06,
3488
+ "loss": 1.1304783821105957,
3489
+ "step": 994
3490
+ },
3491
+ {
3492
+ "epoch": 2.905109489051095,
3493
+ "grad_norm": 0.84765625,
3494
+ "learning_rate": 1.0268669484471839e-06,
3495
+ "loss": 1.1460580825805664,
3496
+ "step": 996
3497
+ },
3498
+ {
3499
+ "epoch": 2.910948905109489,
3500
+ "grad_norm": 0.58984375,
3501
+ "learning_rate": 1.0238018037123042e-06,
3502
+ "loss": 1.216837763786316,
3503
+ "step": 998
3504
+ },
3505
+ {
3506
+ "epoch": 2.9167883211678833,
3507
+ "grad_norm": 0.46484375,
3508
+ "learning_rate": 1.020921789667656e-06,
3509
+ "loss": 1.217172622680664,
3510
+ "step": 1000
3511
+ },
3512
+ {
3513
+ "epoch": 2.9226277372262772,
3514
+ "grad_norm": 0.400390625,
3515
+ "learning_rate": 1.0182270254274888e-06,
3516
+ "loss": 1.2125324010849,
3517
+ "step": 1002
3518
+ },
3519
+ {
3520
+ "epoch": 2.9284671532846716,
3521
+ "grad_norm": 0.83984375,
3522
+ "learning_rate": 1.015717622444321e-06,
3523
+ "loss": 1.1410317420959473,
3524
+ "step": 1004
3525
+ },
3526
+ {
3527
+ "epoch": 2.9343065693430654,
3528
+ "grad_norm": 0.578125,
3529
+ "learning_rate": 1.0133936845043322e-06,
3530
+ "loss": 1.2082126140594482,
3531
+ "step": 1006
3532
+ },
3533
+ {
3534
+ "epoch": 2.9401459854014598,
3535
+ "grad_norm": 0.6484375,
3536
+ "learning_rate": 1.011255307723068e-06,
3537
+ "loss": 1.2457588911056519,
3538
+ "step": 1008
3539
+ },
3540
+ {
3541
+ "epoch": 2.945985401459854,
3542
+ "grad_norm": 1.140625,
3543
+ "learning_rate": 1.0093025805414676e-06,
3544
+ "loss": 1.1455327272415161,
3545
+ "step": 1010
3546
+ },
3547
+ {
3548
+ "epoch": 2.951824817518248,
3549
+ "grad_norm": 0.5546875,
3550
+ "learning_rate": 1.007535583722203e-06,
3551
+ "loss": 1.1338199377059937,
3552
+ "step": 1012
3553
+ },
3554
+ {
3555
+ "epoch": 2.9576642335766423,
3556
+ "grad_norm": 0.419921875,
3557
+ "learning_rate": 1.0059543903463426e-06,
3558
+ "loss": 1.2006813287734985,
3559
+ "step": 1014
3560
+ },
3561
+ {
3562
+ "epoch": 2.9635036496350367,
3563
+ "grad_norm": 0.392578125,
3564
+ "learning_rate": 1.004559065810324e-06,
3565
+ "loss": 1.2222992181777954,
3566
+ "step": 1016
3567
+ },
3568
+ {
3569
+ "epoch": 2.9693430656934305,
3570
+ "grad_norm": 0.349609375,
3571
+ "learning_rate": 1.003349667823253e-06,
3572
+ "loss": 1.1396572589874268,
3573
+ "step": 1018
3574
+ },
3575
+ {
3576
+ "epoch": 2.975182481751825,
3577
+ "grad_norm": 0.4453125,
3578
+ "learning_rate": 1.0023262464045155e-06,
3579
+ "loss": 1.2639334201812744,
3580
+ "step": 1020
3581
+ },
3582
+ {
3583
+ "epoch": 2.981021897810219,
3584
+ "grad_norm": 0.640625,
3585
+ "learning_rate": 1.0014888438817083e-06,
3586
+ "loss": 1.206993579864502,
3587
+ "step": 1022
3588
+ },
3589
+ {
3590
+ "epoch": 2.986861313868613,
3591
+ "grad_norm": 0.58203125,
3592
+ "learning_rate": 1.0008374948888896e-06,
3593
+ "loss": 1.1673427820205688,
3594
+ "step": 1024
3595
+ },
3596
+ {
3597
+ "epoch": 2.9927007299270074,
3598
+ "grad_norm": 0.39453125,
3599
+ "learning_rate": 1.0003722263651458e-06,
3600
+ "loss": 1.2105653285980225,
3601
+ "step": 1026
3602
+ },
3603
+ {
3604
+ "epoch": 2.9985401459854013,
3605
+ "grad_norm": 0.78125,
3606
+ "learning_rate": 1.0000930575534762e-06,
3607
+ "loss": 1.1724438667297363,
3608
+ "step": 1028
3609
+ },
3610
+ {
3611
+ "epoch": 3.0,
3612
+ "step": 1029,
3613
+ "total_flos": 2.909158325117518e+18,
3614
+ "train_loss": 1.307906815331462,
3615
+ "train_runtime": 13215.9511,
3616
+ "train_samples_per_second": 2.488,
3617
+ "train_steps_per_second": 0.078
3618
+ }
3619
+ ],
3620
+ "logging_steps": 2,
3621
+ "max_steps": 1029,
3622
+ "num_input_tokens_seen": 0,
3623
+ "num_train_epochs": 3,
3624
+ "save_steps": 9999999,
3625
+ "stateful_callbacks": {
3626
+ "TrainerControl": {
3627
+ "args": {
3628
+ "should_epoch_stop": false,
3629
+ "should_evaluate": false,
3630
+ "should_log": false,
3631
+ "should_save": true,
3632
+ "should_training_stop": true
3633
+ },
3634
+ "attributes": {}
3635
+ }
3636
+ },
3637
+ "total_flos": 2.909158325117518e+18,
3638
+ "train_batch_size": 1,
3639
+ "trial_name": null,
3640
+ "trial_params": null
3641
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b9272ae0a712acfb205e688a090517fe35ca11120849e3c55e05e63034079abe
3
+ size 5649
training_loss.png ADDED