Hugging Face's logo

Datasets:
michaelchenkj
/
test-models

Dataset card Files
xet
 Community
michaelchenkj commited on
Commit
d0e7baa
·
verified ·
1 Parent(s): 51ea872

Upload folder using huggingface_hub

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. .gitattributes +5 -0
  2. runs/test-model/checkpoint-16/chat_template.jinja +85 -0
  3. runs/test-model/checkpoint-16/config.json +32 -0
  4. runs/test-model/checkpoint-16/generation_config.json +11 -0
  5. runs/test-model/checkpoint-16/model.safetensors +3 -0
  6. runs/test-model/checkpoint-16/optimizer.pt +3 -0
  7. runs/test-model/checkpoint-16/rng_state.pth +3 -0
  8. runs/test-model/checkpoint-16/scheduler.pt +3 -0
  9. runs/test-model/checkpoint-16/tokenizer.json +3 -0
  10. runs/test-model/checkpoint-16/tokenizer_config.json +18 -0
  11. runs/test-model/checkpoint-16/trainer_state.json +146 -0
  12. runs/test-model/checkpoint-16/training_args.bin +3 -0
  13. runs/test-model/checkpoint-160/chat_template.jinja +85 -0
  14. runs/test-model/checkpoint-160/config.json +32 -0
  15. runs/test-model/checkpoint-160/generation_config.json +11 -0
  16. runs/test-model/checkpoint-160/model.safetensors +3 -0
  17. runs/test-model/checkpoint-160/optimizer.pt +3 -0
  18. runs/test-model/checkpoint-160/rng_state.pth +3 -0
  19. runs/test-model/checkpoint-160/scheduler.pt +3 -0
  20. runs/test-model/checkpoint-160/tokenizer.json +3 -0
  21. runs/test-model/checkpoint-160/tokenizer_config.json +18 -0
  22. runs/test-model/checkpoint-160/trainer_state.json +1226 -0
  23. runs/test-model/checkpoint-160/training_args.bin +3 -0
  24. runs/test-model/checkpoint-240/chat_template.jinja +85 -0
  25. runs/test-model/checkpoint-240/config.json +32 -0
  26. runs/test-model/checkpoint-240/generation_config.json +11 -0
  27. runs/test-model/checkpoint-240/model.safetensors +3 -0
  28. runs/test-model/checkpoint-240/optimizer.pt +3 -0
  29. runs/test-model/checkpoint-240/rng_state.pth +3 -0
  30. runs/test-model/checkpoint-240/scheduler.pt +3 -0
  31. runs/test-model/checkpoint-240/tokenizer.json +3 -0
  32. runs/test-model/checkpoint-240/tokenizer_config.json +18 -0
  33. runs/test-model/checkpoint-240/trainer_state.json +1826 -0
  34. runs/test-model/checkpoint-240/training_args.bin +3 -0
  35. runs/test-model/checkpoint-320/chat_template.jinja +85 -0
  36. runs/test-model/checkpoint-320/config.json +32 -0
  37. runs/test-model/checkpoint-320/generation_config.json +11 -0
  38. runs/test-model/checkpoint-320/model.safetensors +3 -0
  39. runs/test-model/checkpoint-320/optimizer.pt +3 -0
  40. runs/test-model/checkpoint-320/rng_state.pth +3 -0
  41. runs/test-model/checkpoint-320/scheduler.pt +3 -0
  42. runs/test-model/checkpoint-320/tokenizer.json +3 -0
  43. runs/test-model/checkpoint-320/tokenizer_config.json +18 -0
  44. runs/test-model/checkpoint-320/trainer_state.json +2426 -0
  45. runs/test-model/checkpoint-320/training_args.bin +3 -0
  46. runs/test-model/checkpoint-80/chat_template.jinja +85 -0
  47. runs/test-model/checkpoint-80/config.json +32 -0
  48. runs/test-model/checkpoint-80/generation_config.json +11 -0
  49. runs/test-model/checkpoint-80/model.safetensors +3 -0
  50. runs/test-model/checkpoint-80/optimizer.pt +3 -0
.gitattributes CHANGED
@@ -58,3 +58,8 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
58
  # Video files - compressed
59
  *.mp4 filter=lfs diff=lfs merge=lfs -text
60
  *.webm filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
58
  # Video files - compressed
59
  *.mp4 filter=lfs diff=lfs merge=lfs -text
60
  *.webm filter=lfs diff=lfs merge=lfs -text
61
+ runs/test-model/checkpoint-16/tokenizer.json filter=lfs diff=lfs merge=lfs -text
62
+ runs/test-model/checkpoint-160/tokenizer.json filter=lfs diff=lfs merge=lfs -text
63
+ runs/test-model/checkpoint-240/tokenizer.json filter=lfs diff=lfs merge=lfs -text
64
+ runs/test-model/checkpoint-320/tokenizer.json filter=lfs diff=lfs merge=lfs -text
65
+ runs/test-model/checkpoint-80/tokenizer.json filter=lfs diff=lfs merge=lfs -text
runs/test-model/checkpoint-16/chat_template.jinja ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0].role == 'system' %}
4
+ {{- messages[0].content + '\n\n' }}
5
+ {%- endif %}
6
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
7
+ {%- for tool in tools %}
8
+ {{- "\n" }}
9
+ {{- tool | tojson }}
10
+ {%- endfor %}
11
+ {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
12
+ {%- else %}
13
+ {%- if messages[0].role == 'system' %}
14
+ {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
15
+ {%- endif %}
16
+ {%- endif %}
17
+ {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
18
+ {%- for message in messages[::-1] %}
19
+ {%- set index = (messages|length - 1) - loop.index0 %}
20
+ {%- if ns.multi_step_tool and message.role == "user" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
21
+ {%- set ns.multi_step_tool = false %}
22
+ {%- set ns.last_query_index = index %}
23
+ {%- endif %}
24
+ {%- endfor %}
25
+ {%- for message in messages %}
26
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
27
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
28
+ {%- elif message.role == "assistant" %}
29
+ {%- set content = message.content %}
30
+ {%- set reasoning_content = '' %}
31
+ {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
32
+ {%- set reasoning_content = message.reasoning_content %}
33
+ {%- else %}
34
+ {%- if '</think>' in message.content %}
35
+ {%- set content = message.content.split('</think>')[-1].lstrip('\n') %}
36
+ {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
37
+ {%- endif %}
38
+ {%- endif %}
39
+ {%- if loop.index0 > ns.last_query_index %}
40
+ {%- if loop.last or (not loop.last and reasoning_content) %}
41
+ {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
42
+ {%- else %}
43
+ {{- '<|im_start|>' + message.role + '\n' + content }}
44
+ {%- endif %}
45
+ {%- else %}
46
+ {{- '<|im_start|>' + message.role + '\n' + content }}
47
+ {%- endif %}
48
+ {%- if message.tool_calls %}
49
+ {%- for tool_call in message.tool_calls %}
50
+ {%- if (loop.first and content) or (not loop.first) %}
51
+ {{- '\n' }}
52
+ {%- endif %}
53
+ {%- if tool_call.function %}
54
+ {%- set tool_call = tool_call.function %}
55
+ {%- endif %}
56
+ {{- '<tool_call>\n{"name": "' }}
57
+ {{- tool_call.name }}
58
+ {{- '", "arguments": ' }}
59
+ {%- if tool_call.arguments is string %}
60
+ {{- tool_call.arguments }}
61
+ {%- else %}
62
+ {{- tool_call.arguments | tojson }}
63
+ {%- endif %}
64
+ {{- '}\n</tool_call>' }}
65
+ {%- endfor %}
66
+ {%- endif %}
67
+ {{- '<|im_end|>\n' }}
68
+ {%- elif message.role == "tool" %}
69
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
70
+ {{- '<|im_start|>user' }}
71
+ {%- endif %}
72
+ {{- '\n<tool_response>\n' }}
73
+ {{- message.content }}
74
+ {{- '\n</tool_response>' }}
75
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
76
+ {{- '<|im_end|>\n' }}
77
+ {%- endif %}
78
+ {%- endif %}
79
+ {%- endfor %}
80
+ {%- if add_generation_prompt %}
81
+ {{- '<|im_start|>assistant\n' }}
82
+ {%- if enable_thinking is defined and enable_thinking is false %}
83
+ {{- '<think>\n\n</think>\n\n' }}
84
+ {%- endif %}
85
+ {%- endif %}
runs/test-model/checkpoint-16/config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlamaForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": null,
8
+ "dtype": "float32",
9
+ "eos_token_id": 151645,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 512,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 1536,
15
+ "max_position_embeddings": 2048,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "num_attention_heads": 4,
19
+ "num_hidden_layers": 20,
20
+ "num_key_value_heads": 4,
21
+ "pad_token_id": 151645,
22
+ "pretraining_tp": 1,
23
+ "rms_norm_eps": 1e-06,
24
+ "rope_parameters": {
25
+ "rope_theta": 10000.0,
26
+ "rope_type": "default"
27
+ },
28
+ "tie_word_embeddings": true,
29
+ "transformers_version": "5.5.0",
30
+ "use_cache": false,
31
+ "vocab_size": 151671
32
+ }
runs/test-model/checkpoint-16/generation_config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "eos_token_id": [
4
+ 151645
5
+ ],
6
+ "output_attentions": false,
7
+ "output_hidden_states": false,
8
+ "pad_token_id": 151645,
9
+ "transformers_version": "5.5.0",
10
+ "use_cache": true
11
+ }
runs/test-model/checkpoint-16/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ad234226fb19745b417959595347233d3cecf4c6a81089093b1e374c24cc6f18
3
+ size 583356232
runs/test-model/checkpoint-16/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c89956656d7ed8be0b6ebffc0830e7a4c57a38efa25bcef43371fa6067c7fddb
3
+ size 1166825803
runs/test-model/checkpoint-16/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f90a5f2ed6d30ebfd28acaade74bb026468970204fa5b02eda67c20566c1a648
3
+ size 14709
runs/test-model/checkpoint-16/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:86479ad780523296a76407f01b89fed8d9f0bbf000e92b244362f11d7085b30f
3
+ size 1465
runs/test-model/checkpoint-16/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59b6776030505d5b441d0727cce137047df748ab15db6ba3a1bac93c123742fb
3
+ size 11424079
runs/test-model/checkpoint-16/tokenizer_config.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "extra_special_tokens": [
9
+ "<|l2r_pred|>",
10
+ "<|r2l_pred|>"
11
+ ],
12
+ "is_local": false,
13
+ "model_max_length": 131072,
14
+ "pad_token": "<|im_end|>",
15
+ "split_special_tokens": false,
16
+ "tokenizer_class": "Qwen2Tokenizer",
17
+ "unk_token": null
18
+ }
runs/test-model/checkpoint-16/trainer_state.json ADDED
@@ -0,0 +1,146 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 1.0,
6
+ "eval_steps": 500,
7
+ "global_step": 16,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0625,
14
+ "grad_norm": 0.4655202031135559,
15
+ "learning_rate": 0.0,
16
+ "loss": 2.00244402885437,
17
+ "step": 1
18
+ },
19
+ {
20
+ "epoch": 0.125,
21
+ "grad_norm": 0.4847956895828247,
22
+ "learning_rate": 3e-05,
23
+ "loss": 2.008800506591797,
24
+ "step": 2
25
+ },
26
+ {
27
+ "epoch": 0.1875,
28
+ "grad_norm": 0.4177771806716919,
29
+ "learning_rate": 6e-05,
30
+ "loss": 1.9685651063919067,
31
+ "step": 3
32
+ },
33
+ {
34
+ "epoch": 0.25,
35
+ "grad_norm": 0.3550562858581543,
36
+ "learning_rate": 8.999999999999999e-05,
37
+ "loss": 1.9168976545333862,
38
+ "step": 4
39
+ },
40
+ {
41
+ "epoch": 0.3125,
42
+ "grad_norm": 0.23293137550354004,
43
+ "learning_rate": 0.00012,
44
+ "loss": 1.8660516738891602,
45
+ "step": 5
46
+ },
47
+ {
48
+ "epoch": 0.375,
49
+ "grad_norm": 0.13890701532363892,
50
+ "learning_rate": 0.00015000000000000001,
51
+ "loss": 1.8209998607635498,
52
+ "step": 6
53
+ },
54
+ {
55
+ "epoch": 0.4375,
56
+ "grad_norm": 0.08890970796346664,
57
+ "learning_rate": 0.00017999999999999998,
58
+ "loss": 1.806854248046875,
59
+ "step": 7
60
+ },
61
+ {
62
+ "epoch": 0.5,
63
+ "grad_norm": 0.061611492186784744,
64
+ "learning_rate": 0.00021000000000000004,
65
+ "loss": 1.793172836303711,
66
+ "step": 8
67
+ },
68
+ {
69
+ "epoch": 0.5625,
70
+ "grad_norm": 0.05786389857530594,
71
+ "learning_rate": 0.00024,
72
+ "loss": 1.7882877588272095,
73
+ "step": 9
74
+ },
75
+ {
76
+ "epoch": 0.625,
77
+ "grad_norm": 0.03947344794869423,
78
+ "learning_rate": 0.00027,
79
+ "loss": 1.7696826457977295,
80
+ "step": 10
81
+ },
82
+ {
83
+ "epoch": 0.6875,
84
+ "grad_norm": 0.029783710837364197,
85
+ "learning_rate": 0.00030000000000000003,
86
+ "loss": 1.7767844200134277,
87
+ "step": 11
88
+ },
89
+ {
90
+ "epoch": 0.75,
91
+ "grad_norm": 0.025729672983288765,
92
+ "learning_rate": 0.00033,
93
+ "loss": 1.7827998399734497,
94
+ "step": 12
95
+ },
96
+ {
97
+ "epoch": 0.8125,
98
+ "grad_norm": 0.020592188462615013,
99
+ "learning_rate": 0.00035999999999999997,
100
+ "loss": 1.766276478767395,
101
+ "step": 13
102
+ },
103
+ {
104
+ "epoch": 0.875,
105
+ "grad_norm": 0.01877395436167717,
106
+ "learning_rate": 0.00039000000000000005,
107
+ "loss": 1.7644124031066895,
108
+ "step": 14
109
+ },
110
+ {
111
+ "epoch": 0.9375,
112
+ "grad_norm": 0.015196867287158966,
113
+ "learning_rate": 0.00042000000000000007,
114
+ "loss": 1.771270990371704,
115
+ "step": 15
116
+ },
117
+ {
118
+ "epoch": 1.0,
119
+ "grad_norm": 0.014262701384723186,
120
+ "learning_rate": 0.00045,
121
+ "loss": 1.755500316619873,
122
+ "step": 16
123
+ }
124
+ ],
125
+ "logging_steps": 1,
126
+ "max_steps": 800,
127
+ "num_input_tokens_seen": 0,
128
+ "num_train_epochs": 50,
129
+ "save_steps": 500,
130
+ "stateful_callbacks": {
131
+ "TrainerControl": {
132
+ "args": {
133
+ "should_epoch_stop": false,
134
+ "should_evaluate": true,
135
+ "should_log": false,
136
+ "should_save": false,
137
+ "should_training_stop": false
138
+ },
139
+ "attributes": {}
140
+ }
141
+ },
142
+ "total_flos": 214470802538496.0,
143
+ "train_batch_size": 8,
144
+ "trial_name": null,
145
+ "trial_params": null
146
+ }
runs/test-model/checkpoint-16/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45cab7e8b9705eceb6dc5bbdf6d75a3f15b3ecba44dfc0a4b5c7d2d8dc6b7db7
3
+ size 5265
runs/test-model/checkpoint-160/chat_template.jinja ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0].role == 'system' %}
4
+ {{- messages[0].content + '\n\n' }}
5
+ {%- endif %}
6
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
7
+ {%- for tool in tools %}
8
+ {{- "\n" }}
9
+ {{- tool | tojson }}
10
+ {%- endfor %}
11
+ {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
12
+ {%- else %}
13
+ {%- if messages[0].role == 'system' %}
14
+ {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
15
+ {%- endif %}
16
+ {%- endif %}
17
+ {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
18
+ {%- for message in messages[::-1] %}
19
+ {%- set index = (messages|length - 1) - loop.index0 %}
20
+ {%- if ns.multi_step_tool and message.role == "user" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
21
+ {%- set ns.multi_step_tool = false %}
22
+ {%- set ns.last_query_index = index %}
23
+ {%- endif %}
24
+ {%- endfor %}
25
+ {%- for message in messages %}
26
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
27
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
28
+ {%- elif message.role == "assistant" %}
29
+ {%- set content = message.content %}
30
+ {%- set reasoning_content = '' %}
31
+ {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
32
+ {%- set reasoning_content = message.reasoning_content %}
33
+ {%- else %}
34
+ {%- if '</think>' in message.content %}
35
+ {%- set content = message.content.split('</think>')[-1].lstrip('\n') %}
36
+ {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
37
+ {%- endif %}
38
+ {%- endif %}
39
+ {%- if loop.index0 > ns.last_query_index %}
40
+ {%- if loop.last or (not loop.last and reasoning_content) %}
41
+ {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
42
+ {%- else %}
43
+ {{- '<|im_start|>' + message.role + '\n' + content }}
44
+ {%- endif %}
45
+ {%- else %}
46
+ {{- '<|im_start|>' + message.role + '\n' + content }}
47
+ {%- endif %}
48
+ {%- if message.tool_calls %}
49
+ {%- for tool_call in message.tool_calls %}
50
+ {%- if (loop.first and content) or (not loop.first) %}
51
+ {{- '\n' }}
52
+ {%- endif %}
53
+ {%- if tool_call.function %}
54
+ {%- set tool_call = tool_call.function %}
55
+ {%- endif %}
56
+ {{- '<tool_call>\n{"name": "' }}
57
+ {{- tool_call.name }}
58
+ {{- '", "arguments": ' }}
59
+ {%- if tool_call.arguments is string %}
60
+ {{- tool_call.arguments }}
61
+ {%- else %}
62
+ {{- tool_call.arguments | tojson }}
63
+ {%- endif %}
64
+ {{- '}\n</tool_call>' }}
65
+ {%- endfor %}
66
+ {%- endif %}
67
+ {{- '<|im_end|>\n' }}
68
+ {%- elif message.role == "tool" %}
69
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
70
+ {{- '<|im_start|>user' }}
71
+ {%- endif %}
72
+ {{- '\n<tool_response>\n' }}
73
+ {{- message.content }}
74
+ {{- '\n</tool_response>' }}
75
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
76
+ {{- '<|im_end|>\n' }}
77
+ {%- endif %}
78
+ {%- endif %}
79
+ {%- endfor %}
80
+ {%- if add_generation_prompt %}
81
+ {{- '<|im_start|>assistant\n' }}
82
+ {%- if enable_thinking is defined and enable_thinking is false %}
83
+ {{- '<think>\n\n</think>\n\n' }}
84
+ {%- endif %}
85
+ {%- endif %}
runs/test-model/checkpoint-160/config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlamaForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": null,
8
+ "dtype": "float32",
9
+ "eos_token_id": 151645,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 512,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 1536,
15
+ "max_position_embeddings": 2048,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "num_attention_heads": 4,
19
+ "num_hidden_layers": 20,
20
+ "num_key_value_heads": 4,
21
+ "pad_token_id": 151645,
22
+ "pretraining_tp": 1,
23
+ "rms_norm_eps": 1e-06,
24
+ "rope_parameters": {
25
+ "rope_theta": 10000.0,
26
+ "rope_type": "default"
27
+ },
28
+ "tie_word_embeddings": true,
29
+ "transformers_version": "5.5.0",
30
+ "use_cache": false,
31
+ "vocab_size": 151671
32
+ }
runs/test-model/checkpoint-160/generation_config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "eos_token_id": [
4
+ 151645
5
+ ],
6
+ "output_attentions": false,
7
+ "output_hidden_states": false,
8
+ "pad_token_id": 151645,
9
+ "transformers_version": "5.5.0",
10
+ "use_cache": true
11
+ }
runs/test-model/checkpoint-160/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6899b9b051c84815d868fe05ac75c958ebdf75f3fe4035b6e492d85d995cf3f9
3
+ size 583356232
runs/test-model/checkpoint-160/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:50f9bf0dcffceaad8da7969f8eda8b324301d2b259b53016e9f6574ec729641e
3
+ size 1166825803
runs/test-model/checkpoint-160/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bbc34b8b1329e714980cee98a4611b1026e736159845ee7101938dae8e999619
3
+ size 14709
runs/test-model/checkpoint-160/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a1f53431b5ed2197da9763c32428013d318c9c2ab414a9f1e06b94d72caad84
3
+ size 1465
runs/test-model/checkpoint-160/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59b6776030505d5b441d0727cce137047df748ab15db6ba3a1bac93c123742fb
3
+ size 11424079
runs/test-model/checkpoint-160/tokenizer_config.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "extra_special_tokens": [
9
+ "<|l2r_pred|>",
10
+ "<|r2l_pred|>"
11
+ ],
12
+ "is_local": false,
13
+ "model_max_length": 131072,
14
+ "pad_token": "<|im_end|>",
15
+ "split_special_tokens": false,
16
+ "tokenizer_class": "Qwen2Tokenizer",
17
+ "unk_token": null
18
+ }
runs/test-model/checkpoint-160/trainer_state.json ADDED
@@ -0,0 +1,1226 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 10.0,
6
+ "eval_steps": 500,
7
+ "global_step": 160,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0625,
14
+ "grad_norm": 0.4655202031135559,
15
+ "learning_rate": 0.0,
16
+ "loss": 2.00244402885437,
17
+ "step": 1
18
+ },
19
+ {
20
+ "epoch": 0.125,
21
+ "grad_norm": 0.4847956895828247,
22
+ "learning_rate": 3e-05,
23
+ "loss": 2.008800506591797,
24
+ "step": 2
25
+ },
26
+ {
27
+ "epoch": 0.1875,
28
+ "grad_norm": 0.4177771806716919,
29
+ "learning_rate": 6e-05,
30
+ "loss": 1.9685651063919067,
31
+ "step": 3
32
+ },
33
+ {
34
+ "epoch": 0.25,
35
+ "grad_norm": 0.3550562858581543,
36
+ "learning_rate": 8.999999999999999e-05,
37
+ "loss": 1.9168976545333862,
38
+ "step": 4
39
+ },
40
+ {
41
+ "epoch": 0.3125,
42
+ "grad_norm": 0.23293137550354004,
43
+ "learning_rate": 0.00012,
44
+ "loss": 1.8660516738891602,
45
+ "step": 5
46
+ },
47
+ {
48
+ "epoch": 0.375,
49
+ "grad_norm": 0.13890701532363892,
50
+ "learning_rate": 0.00015000000000000001,
51
+ "loss": 1.8209998607635498,
52
+ "step": 6
53
+ },
54
+ {
55
+ "epoch": 0.4375,
56
+ "grad_norm": 0.08890970796346664,
57
+ "learning_rate": 0.00017999999999999998,
58
+ "loss": 1.806854248046875,
59
+ "step": 7
60
+ },
61
+ {
62
+ "epoch": 0.5,
63
+ "grad_norm": 0.061611492186784744,
64
+ "learning_rate": 0.00021000000000000004,
65
+ "loss": 1.793172836303711,
66
+ "step": 8
67
+ },
68
+ {
69
+ "epoch": 0.5625,
70
+ "grad_norm": 0.05786389857530594,
71
+ "learning_rate": 0.00024,
72
+ "loss": 1.7882877588272095,
73
+ "step": 9
74
+ },
75
+ {
76
+ "epoch": 0.625,
77
+ "grad_norm": 0.03947344794869423,
78
+ "learning_rate": 0.00027,
79
+ "loss": 1.7696826457977295,
80
+ "step": 10
81
+ },
82
+ {
83
+ "epoch": 0.6875,
84
+ "grad_norm": 0.029783710837364197,
85
+ "learning_rate": 0.00030000000000000003,
86
+ "loss": 1.7767844200134277,
87
+ "step": 11
88
+ },
89
+ {
90
+ "epoch": 0.75,
91
+ "grad_norm": 0.025729672983288765,
92
+ "learning_rate": 0.00033,
93
+ "loss": 1.7827998399734497,
94
+ "step": 12
95
+ },
96
+ {
97
+ "epoch": 0.8125,
98
+ "grad_norm": 0.020592188462615013,
99
+ "learning_rate": 0.00035999999999999997,
100
+ "loss": 1.766276478767395,
101
+ "step": 13
102
+ },
103
+ {
104
+ "epoch": 0.875,
105
+ "grad_norm": 0.01877395436167717,
106
+ "learning_rate": 0.00039000000000000005,
107
+ "loss": 1.7644124031066895,
108
+ "step": 14
109
+ },
110
+ {
111
+ "epoch": 0.9375,
112
+ "grad_norm": 0.015196867287158966,
113
+ "learning_rate": 0.00042000000000000007,
114
+ "loss": 1.771270990371704,
115
+ "step": 15
116
+ },
117
+ {
118
+ "epoch": 1.0,
119
+ "grad_norm": 0.014262701384723186,
120
+ "learning_rate": 0.00045,
121
+ "loss": 1.755500316619873,
122
+ "step": 16
123
+ },
124
+ {
125
+ "epoch": 1.0,
126
+ "eval_loss": 0.8736820816993713,
127
+ "eval_runtime": 30.2743,
128
+ "eval_samples_per_second": 3.369,
129
+ "eval_steps_per_second": 0.429,
130
+ "step": 16
131
+ },
132
+ {
133
+ "epoch": 1.0625,
134
+ "grad_norm": 0.01042763702571392,
135
+ "learning_rate": 0.00048,
136
+ "loss": 1.7556875944137573,
137
+ "step": 17
138
+ },
139
+ {
140
+ "epoch": 1.125,
141
+ "grad_norm": 0.009123577736318111,
142
+ "learning_rate": 0.00051,
143
+ "loss": 1.753890037536621,
144
+ "step": 18
145
+ },
146
+ {
147
+ "epoch": 1.1875,
148
+ "grad_norm": 0.009028254076838493,
149
+ "learning_rate": 0.00054,
150
+ "loss": 1.7515414953231812,
151
+ "step": 19
152
+ },
153
+ {
154
+ "epoch": 1.25,
155
+ "grad_norm": 0.009630529209971428,
156
+ "learning_rate": 0.00057,
157
+ "loss": 1.7465155124664307,
158
+ "step": 20
159
+ },
160
+ {
161
+ "epoch": 1.3125,
162
+ "grad_norm": 0.008235524408519268,
163
+ "learning_rate": 0.0006000000000000001,
164
+ "loss": 1.7533702850341797,
165
+ "step": 21
166
+ },
167
+ {
168
+ "epoch": 1.375,
169
+ "grad_norm": 0.009377758949995041,
170
+ "learning_rate": 0.00063,
171
+ "loss": 1.738031029701233,
172
+ "step": 22
173
+ },
174
+ {
175
+ "epoch": 1.4375,
176
+ "grad_norm": 0.008636261336505413,
177
+ "learning_rate": 0.00066,
178
+ "loss": 1.735835313796997,
179
+ "step": 23
180
+ },
181
+ {
182
+ "epoch": 1.5,
183
+ "grad_norm": 0.008001700975000858,
184
+ "learning_rate": 0.0006900000000000001,
185
+ "loss": 1.7505712509155273,
186
+ "step": 24
187
+ },
188
+ {
189
+ "epoch": 1.5625,
190
+ "grad_norm": 0.007517208810895681,
191
+ "learning_rate": 0.0007199999999999999,
192
+ "loss": 1.7330820560455322,
193
+ "step": 25
194
+ },
195
+ {
196
+ "epoch": 1.625,
197
+ "grad_norm": 0.00952192209661007,
198
+ "learning_rate": 0.00075,
199
+ "loss": 1.7318685054779053,
200
+ "step": 26
201
+ },
202
+ {
203
+ "epoch": 1.6875,
204
+ "grad_norm": 0.006763546261936426,
205
+ "learning_rate": 0.0007800000000000001,
206
+ "loss": 1.7329007387161255,
207
+ "step": 27
208
+ },
209
+ {
210
+ "epoch": 1.75,
211
+ "grad_norm": 0.006364013999700546,
212
+ "learning_rate": 0.0008100000000000001,
213
+ "loss": 1.7285535335540771,
214
+ "step": 28
215
+ },
216
+ {
217
+ "epoch": 1.8125,
218
+ "grad_norm": 0.007770394906401634,
219
+ "learning_rate": 0.0008400000000000001,
220
+ "loss": 1.7320656776428223,
221
+ "step": 29
222
+ },
223
+ {
224
+ "epoch": 1.875,
225
+ "grad_norm": 0.006639651022851467,
226
+ "learning_rate": 0.00087,
227
+ "loss": 1.7304966449737549,
228
+ "step": 30
229
+ },
230
+ {
231
+ "epoch": 1.9375,
232
+ "grad_norm": 0.008190952241420746,
233
+ "learning_rate": 0.0009,
234
+ "loss": 1.7399260997772217,
235
+ "step": 31
236
+ },
237
+ {
238
+ "epoch": 2.0,
239
+ "grad_norm": 0.008188036270439625,
240
+ "learning_rate": 0.00093,
241
+ "loss": 1.7375996112823486,
242
+ "step": 32
243
+ },
244
+ {
245
+ "epoch": 2.0,
246
+ "eval_loss": 0.860374927520752,
247
+ "eval_runtime": 30.3049,
248
+ "eval_samples_per_second": 3.366,
249
+ "eval_steps_per_second": 0.429,
250
+ "step": 32
251
+ },
252
+ {
253
+ "epoch": 2.0625,
254
+ "grad_norm": 0.005768604576587677,
255
+ "learning_rate": 0.00096,
256
+ "loss": 1.7309706211090088,
257
+ "step": 33
258
+ },
259
+ {
260
+ "epoch": 2.125,
261
+ "grad_norm": 0.006822533905506134,
262
+ "learning_rate": 0.00099,
263
+ "loss": 1.7321696281433105,
264
+ "step": 34
265
+ },
266
+ {
267
+ "epoch": 2.1875,
268
+ "grad_norm": 0.009974546730518341,
269
+ "learning_rate": 0.00102,
270
+ "loss": 1.7290048599243164,
271
+ "step": 35
272
+ },
273
+ {
274
+ "epoch": 2.25,
275
+ "grad_norm": 0.00856814906001091,
276
+ "learning_rate": 0.00105,
277
+ "loss": 1.7231603860855103,
278
+ "step": 36
279
+ },
280
+ {
281
+ "epoch": 2.3125,
282
+ "grad_norm": 0.011694240383803844,
283
+ "learning_rate": 0.00108,
284
+ "loss": 1.728348731994629,
285
+ "step": 37
286
+ },
287
+ {
288
+ "epoch": 2.375,
289
+ "grad_norm": 0.005738088861107826,
290
+ "learning_rate": 0.00111,
291
+ "loss": 1.7308874130249023,
292
+ "step": 38
293
+ },
294
+ {
295
+ "epoch": 2.4375,
296
+ "grad_norm": 0.010038000531494617,
297
+ "learning_rate": 0.00114,
298
+ "loss": 1.7276837825775146,
299
+ "step": 39
300
+ },
301
+ {
302
+ "epoch": 2.5,
303
+ "grad_norm": 0.01586989127099514,
304
+ "learning_rate": 0.00117,
305
+ "loss": 1.7368175983428955,
306
+ "step": 40
307
+ },
308
+ {
309
+ "epoch": 2.5625,
310
+ "grad_norm": 0.16198667883872986,
311
+ "learning_rate": 0.0012000000000000001,
312
+ "loss": 1.730246663093567,
313
+ "step": 41
314
+ },
315
+ {
316
+ "epoch": 2.625,
317
+ "grad_norm": 0.01653525047004223,
318
+ "learning_rate": 0.00123,
319
+ "loss": 1.7403852939605713,
320
+ "step": 42
321
+ },
322
+ {
323
+ "epoch": 2.6875,
324
+ "grad_norm": 0.01241418905556202,
325
+ "learning_rate": 0.00126,
326
+ "loss": 1.7306722402572632,
327
+ "step": 43
328
+ },
329
+ {
330
+ "epoch": 2.75,
331
+ "grad_norm": 0.022560030221939087,
332
+ "learning_rate": 0.00129,
333
+ "loss": 1.7223260402679443,
334
+ "step": 44
335
+ },
336
+ {
337
+ "epoch": 2.8125,
338
+ "grad_norm": 0.019568368792533875,
339
+ "learning_rate": 0.00132,
340
+ "loss": 1.7304983139038086,
341
+ "step": 45
342
+ },
343
+ {
344
+ "epoch": 2.875,
345
+ "grad_norm": 0.016755228862166405,
346
+ "learning_rate": 0.00135,
347
+ "loss": 1.7223992347717285,
348
+ "step": 46
349
+ },
350
+ {
351
+ "epoch": 2.9375,
352
+ "grad_norm": 0.02981281839311123,
353
+ "learning_rate": 0.0013800000000000002,
354
+ "loss": 1.7094981670379639,
355
+ "step": 47
356
+ },
357
+ {
358
+ "epoch": 3.0,
359
+ "grad_norm": 0.02506650611758232,
360
+ "learning_rate": 0.00141,
361
+ "loss": 1.7054555416107178,
362
+ "step": 48
363
+ },
364
+ {
365
+ "epoch": 3.0,
366
+ "eval_loss": 0.8416222333908081,
367
+ "eval_runtime": 30.3027,
368
+ "eval_samples_per_second": 3.366,
369
+ "eval_steps_per_second": 0.429,
370
+ "step": 48
371
+ },
372
+ {
373
+ "epoch": 3.0625,
374
+ "grad_norm": 0.02260257676243782,
375
+ "learning_rate": 0.0014399999999999999,
376
+ "loss": 1.6888771057128906,
377
+ "step": 49
378
+ },
379
+ {
380
+ "epoch": 3.125,
381
+ "grad_norm": 0.02620658650994301,
382
+ "learning_rate": 0.00147,
383
+ "loss": 1.6639578342437744,
384
+ "step": 50
385
+ },
386
+ {
387
+ "epoch": 3.1875,
388
+ "grad_norm": 0.047354575246572495,
389
+ "learning_rate": 0.0015,
390
+ "loss": 1.6736350059509277,
391
+ "step": 51
392
+ },
393
+ {
394
+ "epoch": 3.25,
395
+ "grad_norm": 0.04594860598444939,
396
+ "learning_rate": 0.0015300000000000001,
397
+ "loss": 1.6675360202789307,
398
+ "step": 52
399
+ },
400
+ {
401
+ "epoch": 3.3125,
402
+ "grad_norm": 0.08131621032953262,
403
+ "learning_rate": 0.0015600000000000002,
404
+ "loss": 1.6574077606201172,
405
+ "step": 53
406
+ },
407
+ {
408
+ "epoch": 3.375,
409
+ "grad_norm": 0.028789900243282318,
410
+ "learning_rate": 0.00159,
411
+ "loss": 1.6299148797988892,
412
+ "step": 54
413
+ },
414
+ {
415
+ "epoch": 3.4375,
416
+ "grad_norm": 0.2623404860496521,
417
+ "learning_rate": 0.0016200000000000001,
418
+ "loss": 1.6315665245056152,
419
+ "step": 55
420
+ },
421
+ {
422
+ "epoch": 3.5,
423
+ "grad_norm": 0.056828465312719345,
424
+ "learning_rate": 0.0016500000000000002,
425
+ "loss": 1.6279325485229492,
426
+ "step": 56
427
+ },
428
+ {
429
+ "epoch": 3.5625,
430
+ "grad_norm": 0.0549125075340271,
431
+ "learning_rate": 0.0016800000000000003,
432
+ "loss": 1.6549303531646729,
433
+ "step": 57
434
+ },
435
+ {
436
+ "epoch": 3.625,
437
+ "grad_norm": 0.07287958264350891,
438
+ "learning_rate": 0.00171,
439
+ "loss": 1.606963872909546,
440
+ "step": 58
441
+ },
442
+ {
443
+ "epoch": 3.6875,
444
+ "grad_norm": 0.07426343113183975,
445
+ "learning_rate": 0.00174,
446
+ "loss": 1.6250534057617188,
447
+ "step": 59
448
+ },
449
+ {
450
+ "epoch": 3.75,
451
+ "grad_norm": 0.1364986002445221,
452
+ "learning_rate": 0.0017699999999999999,
453
+ "loss": 1.5964961051940918,
454
+ "step": 60
455
+ },
456
+ {
457
+ "epoch": 3.8125,
458
+ "grad_norm": 0.15007159113883972,
459
+ "learning_rate": 0.0018,
460
+ "loss": 1.6019010543823242,
461
+ "step": 61
462
+ },
463
+ {
464
+ "epoch": 3.875,
465
+ "grad_norm": 0.08447311073541641,
466
+ "learning_rate": 0.00183,
467
+ "loss": 1.5763543844223022,
468
+ "step": 62
469
+ },
470
+ {
471
+ "epoch": 3.9375,
472
+ "grad_norm": 0.1681874394416809,
473
+ "learning_rate": 0.00186,
474
+ "loss": 1.5769858360290527,
475
+ "step": 63
476
+ },
477
+ {
478
+ "epoch": 4.0,
479
+ "grad_norm": 0.09677319973707199,
480
+ "learning_rate": 0.00189,
481
+ "loss": 1.534682273864746,
482
+ "step": 64
483
+ },
484
+ {
485
+ "epoch": 4.0,
486
+ "eval_loss": 0.7682511806488037,
487
+ "eval_runtime": 30.4228,
488
+ "eval_samples_per_second": 3.353,
489
+ "eval_steps_per_second": 0.427,
490
+ "step": 64
491
+ },
492
+ {
493
+ "epoch": 4.0625,
494
+ "grad_norm": 0.42635980248451233,
495
+ "learning_rate": 0.00192,
496
+ "loss": 1.5509562492370605,
497
+ "step": 65
498
+ },
499
+ {
500
+ "epoch": 4.125,
501
+ "grad_norm": 0.8467249274253845,
502
+ "learning_rate": 0.0019500000000000001,
503
+ "loss": 1.5001531839370728,
504
+ "step": 66
505
+ },
506
+ {
507
+ "epoch": 4.1875,
508
+ "grad_norm": 0.22631719708442688,
509
+ "learning_rate": 0.00198,
510
+ "loss": 1.4554829597473145,
511
+ "step": 67
512
+ },
513
+ {
514
+ "epoch": 4.25,
515
+ "grad_norm": 0.05312464386224747,
516
+ "learning_rate": 0.00201,
517
+ "loss": 1.4732775688171387,
518
+ "step": 68
519
+ },
520
+ {
521
+ "epoch": 4.3125,
522
+ "grad_norm": 0.050442785024642944,
523
+ "learning_rate": 0.00204,
524
+ "loss": 1.4176194667816162,
525
+ "step": 69
526
+ },
527
+ {
528
+ "epoch": 4.375,
529
+ "grad_norm": 0.18886858224868774,
530
+ "learning_rate": 0.00207,
531
+ "loss": 1.4325997829437256,
532
+ "step": 70
533
+ },
534
+ {
535
+ "epoch": 4.4375,
536
+ "grad_norm": 0.05258706212043762,
537
+ "learning_rate": 0.0021,
538
+ "loss": 1.39802086353302,
539
+ "step": 71
540
+ },
541
+ {
542
+ "epoch": 4.5,
543
+ "grad_norm": 0.07482657581567764,
544
+ "learning_rate": 0.00213,
545
+ "loss": 1.4000320434570312,
546
+ "step": 72
547
+ },
548
+ {
549
+ "epoch": 4.5625,
550
+ "grad_norm": 0.4853312075138092,
551
+ "learning_rate": 0.00216,
552
+ "loss": 1.4199151992797852,
553
+ "step": 73
554
+ },
555
+ {
556
+ "epoch": 4.625,
557
+ "grad_norm": 0.48888301849365234,
558
+ "learning_rate": 0.00219,
559
+ "loss": 1.3049702644348145,
560
+ "step": 74
561
+ },
562
+ {
563
+ "epoch": 4.6875,
564
+ "grad_norm": 0.284348726272583,
565
+ "learning_rate": 0.00222,
566
+ "loss": 1.3884068727493286,
567
+ "step": 75
568
+ },
569
+ {
570
+ "epoch": 4.75,
571
+ "grad_norm": 0.20248942077159882,
572
+ "learning_rate": 0.0022500000000000003,
573
+ "loss": 1.3502120971679688,
574
+ "step": 76
575
+ },
576
+ {
577
+ "epoch": 4.8125,
578
+ "grad_norm": 0.2714093327522278,
579
+ "learning_rate": 0.00228,
580
+ "loss": 1.3195693492889404,
581
+ "step": 77
582
+ },
583
+ {
584
+ "epoch": 4.875,
585
+ "grad_norm": 0.2544061243534088,
586
+ "learning_rate": 0.00231,
587
+ "loss": 1.3194304704666138,
588
+ "step": 78
589
+ },
590
+ {
591
+ "epoch": 4.9375,
592
+ "grad_norm": 0.18475523591041565,
593
+ "learning_rate": 0.00234,
594
+ "loss": 1.3006361722946167,
595
+ "step": 79
596
+ },
597
+ {
598
+ "epoch": 5.0,
599
+ "grad_norm": 0.6583278775215149,
600
+ "learning_rate": 0.00237,
601
+ "loss": 1.2868517637252808,
602
+ "step": 80
603
+ },
604
+ {
605
+ "epoch": 5.0,
606
+ "eval_loss": 0.6618342399597168,
607
+ "eval_runtime": 30.1233,
608
+ "eval_samples_per_second": 3.386,
609
+ "eval_steps_per_second": 0.432,
610
+ "step": 80
611
+ },
612
+ {
613
+ "epoch": 5.0625,
614
+ "grad_norm": 0.04499061033129692,
615
+ "learning_rate": 0.0024000000000000002,
616
+ "loss": 1.25992751121521,
617
+ "step": 81
618
+ },
619
+ {
620
+ "epoch": 5.125,
621
+ "grad_norm": 0.05609860643744469,
622
+ "learning_rate": 0.0024300000000000003,
623
+ "loss": 1.2764487266540527,
624
+ "step": 82
625
+ },
626
+ {
627
+ "epoch": 5.1875,
628
+ "grad_norm": 0.03665148839354515,
629
+ "learning_rate": 0.00246,
630
+ "loss": 1.2908766269683838,
631
+ "step": 83
632
+ },
633
+ {
634
+ "epoch": 5.25,
635
+ "grad_norm": 0.17949481308460236,
636
+ "learning_rate": 0.00249,
637
+ "loss": 1.253080129623413,
638
+ "step": 84
639
+ },
640
+ {
641
+ "epoch": 5.3125,
642
+ "grad_norm": 0.14031393826007843,
643
+ "learning_rate": 0.00252,
644
+ "loss": 1.2648870944976807,
645
+ "step": 85
646
+ },
647
+ {
648
+ "epoch": 5.375,
649
+ "grad_norm": 0.08193210512399673,
650
+ "learning_rate": 0.00255,
651
+ "loss": 1.296234130859375,
652
+ "step": 86
653
+ },
654
+ {
655
+ "epoch": 5.4375,
656
+ "grad_norm": 0.056413356214761734,
657
+ "learning_rate": 0.00258,
658
+ "loss": 1.2784960269927979,
659
+ "step": 87
660
+ },
661
+ {
662
+ "epoch": 5.5,
663
+ "grad_norm": 0.09004318714141846,
664
+ "learning_rate": 0.00261,
665
+ "loss": 1.2054948806762695,
666
+ "step": 88
667
+ },
668
+ {
669
+ "epoch": 5.5625,
670
+ "grad_norm": 0.044756047427654266,
671
+ "learning_rate": 0.00264,
672
+ "loss": 1.2707666158676147,
673
+ "step": 89
674
+ },
675
+ {
676
+ "epoch": 5.625,
677
+ "grad_norm": 0.049932755529880524,
678
+ "learning_rate": 0.00267,
679
+ "loss": 1.2369787693023682,
680
+ "step": 90
681
+ },
682
+ {
683
+ "epoch": 5.6875,
684
+ "grad_norm": 0.09624820947647095,
685
+ "learning_rate": 0.0027,
686
+ "loss": 1.2419278621673584,
687
+ "step": 91
688
+ },
689
+ {
690
+ "epoch": 5.75,
691
+ "grad_norm": 0.17648059129714966,
692
+ "learning_rate": 0.0027300000000000002,
693
+ "loss": 1.2992033958435059,
694
+ "step": 92
695
+ },
696
+ {
697
+ "epoch": 5.8125,
698
+ "grad_norm": 0.07027823477983475,
699
+ "learning_rate": 0.0027600000000000003,
700
+ "loss": 1.2255799770355225,
701
+ "step": 93
702
+ },
703
+ {
704
+ "epoch": 5.875,
705
+ "grad_norm": 0.06040462478995323,
706
+ "learning_rate": 0.0027900000000000004,
707
+ "loss": 1.136023759841919,
708
+ "step": 94
709
+ },
710
+ {
711
+ "epoch": 5.9375,
712
+ "grad_norm": 0.09845872968435287,
713
+ "learning_rate": 0.00282,
714
+ "loss": 1.1807364225387573,
715
+ "step": 95
716
+ },
717
+ {
718
+ "epoch": 6.0,
719
+ "grad_norm": 0.09086572378873825,
720
+ "learning_rate": 0.00285,
721
+ "loss": 1.2171341180801392,
722
+ "step": 96
723
+ },
724
+ {
725
+ "epoch": 6.0,
726
+ "eval_loss": 0.6200892329216003,
727
+ "eval_runtime": 30.1252,
728
+ "eval_samples_per_second": 3.386,
729
+ "eval_steps_per_second": 0.432,
730
+ "step": 96
731
+ },
732
+ {
733
+ "epoch": 6.0625,
734
+ "grad_norm": 0.11787034571170807,
735
+ "learning_rate": 0.0028799999999999997,
736
+ "loss": 1.242072582244873,
737
+ "step": 97
738
+ },
739
+ {
740
+ "epoch": 6.125,
741
+ "grad_norm": 0.20093177258968353,
742
+ "learning_rate": 0.00291,
743
+ "loss": 1.2056899070739746,
744
+ "step": 98
745
+ },
746
+ {
747
+ "epoch": 6.1875,
748
+ "grad_norm": 0.03221331909298897,
749
+ "learning_rate": 0.00294,
750
+ "loss": 1.1774635314941406,
751
+ "step": 99
752
+ },
753
+ {
754
+ "epoch": 6.25,
755
+ "grad_norm": 0.06112038344144821,
756
+ "learning_rate": 0.00297,
757
+ "loss": 1.1831618547439575,
758
+ "step": 100
759
+ },
760
+ {
761
+ "epoch": 6.3125,
762
+ "grad_norm": 0.05896962434053421,
763
+ "learning_rate": 0.003,
764
+ "loss": 1.219610571861267,
765
+ "step": 101
766
+ },
767
+ {
768
+ "epoch": 6.375,
769
+ "grad_norm": 0.4068014323711395,
770
+ "learning_rate": 0.0029999850445531283,
771
+ "loss": 1.231442928314209,
772
+ "step": 102
773
+ },
774
+ {
775
+ "epoch": 6.4375,
776
+ "grad_norm": 3.013302803039551,
777
+ "learning_rate": 0.0029999401785137437,
778
+ "loss": 1.2512195110321045,
779
+ "step": 103
780
+ },
781
+ {
782
+ "epoch": 6.5,
783
+ "grad_norm": 0.07003403455018997,
784
+ "learning_rate": 0.0029998654027855406,
785
+ "loss": 1.1912157535552979,
786
+ "step": 104
787
+ },
788
+ {
789
+ "epoch": 6.5625,
790
+ "grad_norm": 1.1927955150604248,
791
+ "learning_rate": 0.002999760718874652,
792
+ "loss": 1.1158745288848877,
793
+ "step": 105
794
+ },
795
+ {
796
+ "epoch": 6.625,
797
+ "grad_norm": 0.11684458702802658,
798
+ "learning_rate": 0.0029996261288896225,
799
+ "loss": 1.1835001707077026,
800
+ "step": 106
801
+ },
802
+ {
803
+ "epoch": 6.6875,
804
+ "grad_norm": 0.07812155783176422,
805
+ "learning_rate": 0.0029994616355413666,
806
+ "loss": 1.1712734699249268,
807
+ "step": 107
808
+ },
809
+ {
810
+ "epoch": 6.75,
811
+ "grad_norm": 0.19128946959972382,
812
+ "learning_rate": 0.0029992672421431115,
813
+ "loss": 1.0906574726104736,
814
+ "step": 108
815
+ },
816
+ {
817
+ "epoch": 6.8125,
818
+ "grad_norm": 0.033625371754169464,
819
+ "learning_rate": 0.002999042952610332,
820
+ "loss": 1.115525245666504,
821
+ "step": 109
822
+ },
823
+ {
824
+ "epoch": 6.875,
825
+ "grad_norm": 0.2518763244152069,
826
+ "learning_rate": 0.0029987887714606723,
827
+ "loss": 1.1433348655700684,
828
+ "step": 110
829
+ },
830
+ {
831
+ "epoch": 6.9375,
832
+ "grad_norm": 0.037159573286771774,
833
+ "learning_rate": 0.002998504703813852,
834
+ "loss": 1.1867601871490479,
835
+ "step": 111
836
+ },
837
+ {
838
+ "epoch": 7.0,
839
+ "grad_norm": 0.04564391076564789,
840
+ "learning_rate": 0.0029981907553915674,
841
+ "loss": 1.1302828788757324,
842
+ "step": 112
843
+ },
844
+ {
845
+ "epoch": 7.0,
846
+ "eval_loss": 0.5897142887115479,
847
+ "eval_runtime": 31.3671,
848
+ "eval_samples_per_second": 3.252,
849
+ "eval_steps_per_second": 0.414,
850
+ "step": 112
851
+ },
852
+ {
853
+ "epoch": 7.0625,
854
+ "grad_norm": 0.05205775797367096,
855
+ "learning_rate": 0.0029978469325173717,
856
+ "loss": 1.182741641998291,
857
+ "step": 113
858
+ },
859
+ {
860
+ "epoch": 7.125,
861
+ "grad_norm": 0.12080052495002747,
862
+ "learning_rate": 0.002997473242116551,
863
+ "loss": 1.1457600593566895,
864
+ "step": 114
865
+ },
866
+ {
867
+ "epoch": 7.1875,
868
+ "grad_norm": 0.05600045621395111,
869
+ "learning_rate": 0.0029970696917159833,
870
+ "loss": 1.1101720333099365,
871
+ "step": 115
872
+ },
873
+ {
874
+ "epoch": 7.25,
875
+ "grad_norm": 0.0507950522005558,
876
+ "learning_rate": 0.0029966362894439874,
877
+ "loss": 1.1785833835601807,
878
+ "step": 116
879
+ },
880
+ {
881
+ "epoch": 7.3125,
882
+ "grad_norm": 0.6396550536155701,
883
+ "learning_rate": 0.002996173044030159,
884
+ "loss": 1.1538987159729004,
885
+ "step": 117
886
+ },
887
+ {
888
+ "epoch": 7.375,
889
+ "grad_norm": 0.06849977374076843,
890
+ "learning_rate": 0.0029956799648051934,
891
+ "loss": 1.1528857946395874,
892
+ "step": 118
893
+ },
894
+ {
895
+ "epoch": 7.4375,
896
+ "grad_norm": 0.0685896947979927,
897
+ "learning_rate": 0.002995157061700702,
898
+ "loss": 1.1049470901489258,
899
+ "step": 119
900
+ },
901
+ {
902
+ "epoch": 7.5,
903
+ "grad_norm": 1.027722716331482,
904
+ "learning_rate": 0.002994604345249006,
905
+ "loss": 1.1308050155639648,
906
+ "step": 120
907
+ },
908
+ {
909
+ "epoch": 7.5625,
910
+ "grad_norm": 0.21406109631061554,
911
+ "learning_rate": 0.00299402182658293,
912
+ "loss": 1.0960861444473267,
913
+ "step": 121
914
+ },
915
+ {
916
+ "epoch": 7.625,
917
+ "grad_norm": 0.0503656342625618,
918
+ "learning_rate": 0.002993409517435574,
919
+ "loss": 1.1096141338348389,
920
+ "step": 122
921
+ },
922
+ {
923
+ "epoch": 7.6875,
924
+ "grad_norm": 0.04800846800208092,
925
+ "learning_rate": 0.002992767430140078,
926
+ "loss": 1.1386680603027344,
927
+ "step": 123
928
+ },
929
+ {
930
+ "epoch": 7.75,
931
+ "grad_norm": 0.13091202080249786,
932
+ "learning_rate": 0.002992095577629376,
933
+ "loss": 1.1129441261291504,
934
+ "step": 124
935
+ },
936
+ {
937
+ "epoch": 7.8125,
938
+ "grad_norm": 0.04949796572327614,
939
+ "learning_rate": 0.0029913939734359306,
940
+ "loss": 1.106520175933838,
941
+ "step": 125
942
+ },
943
+ {
944
+ "epoch": 7.875,
945
+ "grad_norm": 0.07686582207679749,
946
+ "learning_rate": 0.0029906626316914655,
947
+ "loss": 1.1572234630584717,
948
+ "step": 126
949
+ },
950
+ {
951
+ "epoch": 7.9375,
952
+ "grad_norm": 0.053587451577186584,
953
+ "learning_rate": 0.0029899015671266773,
954
+ "loss": 1.125717282295227,
955
+ "step": 127
956
+ },
957
+ {
958
+ "epoch": 8.0,
959
+ "grad_norm": 0.08031050115823746,
960
+ "learning_rate": 0.0029891107950709406,
961
+ "loss": 1.153022050857544,
962
+ "step": 128
963
+ },
964
+ {
965
+ "epoch": 8.0,
966
+ "eval_loss": 0.5910838842391968,
967
+ "eval_runtime": 30.7073,
968
+ "eval_samples_per_second": 3.322,
969
+ "eval_steps_per_second": 0.423,
970
+ "step": 128
971
+ },
972
+ {
973
+ "epoch": 8.0625,
974
+ "grad_norm": 0.3520950675010681,
975
+ "learning_rate": 0.0029882903314519997,
976
+ "loss": 1.1582988500595093,
977
+ "step": 129
978
+ },
979
+ {
980
+ "epoch": 8.125,
981
+ "grad_norm": 0.07153584063053131,
982
+ "learning_rate": 0.002987440192795645,
983
+ "loss": 1.1344678401947021,
984
+ "step": 130
985
+ },
986
+ {
987
+ "epoch": 8.1875,
988
+ "grad_norm": 0.06336807459592819,
989
+ "learning_rate": 0.0029865603962253833,
990
+ "loss": 1.120319128036499,
991
+ "step": 131
992
+ },
993
+ {
994
+ "epoch": 8.25,
995
+ "grad_norm": 0.11782127618789673,
996
+ "learning_rate": 0.0029856509594620912,
997
+ "loss": 1.1245379447937012,
998
+ "step": 132
999
+ },
1000
+ {
1001
+ "epoch": 8.3125,
1002
+ "grad_norm": 0.2592264413833618,
1003
+ "learning_rate": 0.0029847119008236576,
1004
+ "loss": 1.1454271078109741,
1005
+ "step": 133
1006
+ },
1007
+ {
1008
+ "epoch": 8.375,
1009
+ "grad_norm": 0.21255695819854736,
1010
+ "learning_rate": 0.0029837432392246185,
1011
+ "loss": 1.231805443763733,
1012
+ "step": 134
1013
+ },
1014
+ {
1015
+ "epoch": 8.4375,
1016
+ "grad_norm": 15.057058334350586,
1017
+ "learning_rate": 0.0029827449941757698,
1018
+ "loss": 1.1380956172943115,
1019
+ "step": 135
1020
+ },
1021
+ {
1022
+ "epoch": 8.5,
1023
+ "grad_norm": 0.17857126891613007,
1024
+ "learning_rate": 0.00298171718578378,
1025
+ "loss": 1.1401441097259521,
1026
+ "step": 136
1027
+ },
1028
+ {
1029
+ "epoch": 8.5625,
1030
+ "grad_norm": 0.18281924724578857,
1031
+ "learning_rate": 0.002980659834750781,
1032
+ "loss": 1.171494722366333,
1033
+ "step": 137
1034
+ },
1035
+ {
1036
+ "epoch": 8.625,
1037
+ "grad_norm": 0.10490825027227402,
1038
+ "learning_rate": 0.0029795729623739555,
1039
+ "loss": 1.1622605323791504,
1040
+ "step": 138
1041
+ },
1042
+ {
1043
+ "epoch": 8.6875,
1044
+ "grad_norm": 0.344013512134552,
1045
+ "learning_rate": 0.002978456590545104,
1046
+ "loss": 1.1702141761779785,
1047
+ "step": 139
1048
+ },
1049
+ {
1050
+ "epoch": 8.75,
1051
+ "grad_norm": 0.05058368295431137,
1052
+ "learning_rate": 0.0029773107417502055,
1053
+ "loss": 1.2006211280822754,
1054
+ "step": 140
1055
+ },
1056
+ {
1057
+ "epoch": 8.8125,
1058
+ "grad_norm": 0.09775760769844055,
1059
+ "learning_rate": 0.002976135439068965,
1060
+ "loss": 1.1407063007354736,
1061
+ "step": 141
1062
+ },
1063
+ {
1064
+ "epoch": 8.875,
1065
+ "grad_norm": 0.0785030648112297,
1066
+ "learning_rate": 0.0029749307061743475,
1067
+ "loss": 1.1778209209442139,
1068
+ "step": 142
1069
+ },
1070
+ {
1071
+ "epoch": 8.9375,
1072
+ "grad_norm": 0.053056441247463226,
1073
+ "learning_rate": 0.0029736965673321032,
1074
+ "loss": 1.1302082538604736,
1075
+ "step": 143
1076
+ },
1077
+ {
1078
+ "epoch": 9.0,
1079
+ "grad_norm": 0.4020994305610657,
1080
+ "learning_rate": 0.0029724330474002743,
1081
+ "loss": 1.1655206680297852,
1082
+ "step": 144
1083
+ },
1084
+ {
1085
+ "epoch": 9.0,
1086
+ "eval_loss": 0.6083483099937439,
1087
+ "eval_runtime": 30.0453,
1088
+ "eval_samples_per_second": 3.395,
1089
+ "eval_steps_per_second": 0.433,
1090
+ "step": 144
1091
+ },
1092
+ {
1093
+ "epoch": 9.0625,
1094
+ "grad_norm": 0.04755420982837677,
1095
+ "learning_rate": 0.002971140171828702,
1096
+ "loss": 1.1900575160980225,
1097
+ "step": 145
1098
+ },
1099
+ {
1100
+ "epoch": 9.125,
1101
+ "grad_norm": 0.2549609839916229,
1102
+ "learning_rate": 0.002969817966658505,
1103
+ "loss": 1.1891400814056396,
1104
+ "step": 146
1105
+ },
1106
+ {
1107
+ "epoch": 9.1875,
1108
+ "grad_norm": 0.11641769111156464,
1109
+ "learning_rate": 0.0029684664585215618,
1110
+ "loss": 1.1930484771728516,
1111
+ "step": 147
1112
+ },
1113
+ {
1114
+ "epoch": 9.25,
1115
+ "grad_norm": 0.13052672147750854,
1116
+ "learning_rate": 0.00296708567463997,
1117
+ "loss": 1.1905772686004639,
1118
+ "step": 148
1119
+ },
1120
+ {
1121
+ "epoch": 9.3125,
1122
+ "grad_norm": 0.15126636624336243,
1123
+ "learning_rate": 0.0029656756428255017,
1124
+ "loss": 1.1417806148529053,
1125
+ "step": 149
1126
+ },
1127
+ {
1128
+ "epoch": 9.375,
1129
+ "grad_norm": 0.08826905488967896,
1130
+ "learning_rate": 0.00296423639147904,
1131
+ "loss": 1.1802985668182373,
1132
+ "step": 150
1133
+ },
1134
+ {
1135
+ "epoch": 9.4375,
1136
+ "grad_norm": 0.01381592359393835,
1137
+ "learning_rate": 0.002962767949590008,
1138
+ "loss": 1.1525968313217163,
1139
+ "step": 151
1140
+ },
1141
+ {
1142
+ "epoch": 9.5,
1143
+ "grad_norm": 0.04648139327764511,
1144
+ "learning_rate": 0.0029612703467357866,
1145
+ "loss": 1.169856309890747,
1146
+ "step": 152
1147
+ },
1148
+ {
1149
+ "epoch": 9.5625,
1150
+ "grad_norm": 0.014019235037267208,
1151
+ "learning_rate": 0.0029597436130811153,
1152
+ "loss": 1.1442453861236572,
1153
+ "step": 153
1154
+ },
1155
+ {
1156
+ "epoch": 9.625,
1157
+ "grad_norm": 0.03701096400618553,
1158
+ "learning_rate": 0.0029581877793774885,
1159
+ "loss": 1.112680435180664,
1160
+ "step": 154
1161
+ },
1162
+ {
1163
+ "epoch": 9.6875,
1164
+ "grad_norm": 0.030421797186136246,
1165
+ "learning_rate": 0.0029566028769625334,
1166
+ "loss": 1.223894715309143,
1167
+ "step": 155
1168
+ },
1169
+ {
1170
+ "epoch": 9.75,
1171
+ "grad_norm": 0.12336083501577377,
1172
+ "learning_rate": 0.00295498893775938,
1173
+ "loss": 1.220346212387085,
1174
+ "step": 156
1175
+ },
1176
+ {
1177
+ "epoch": 9.8125,
1178
+ "grad_norm": 0.038585737347602844,
1179
+ "learning_rate": 0.0029533459942760174,
1180
+ "loss": 1.2078583240509033,
1181
+ "step": 157
1182
+ },
1183
+ {
1184
+ "epoch": 9.875,
1185
+ "grad_norm": 0.4653286039829254,
1186
+ "learning_rate": 0.002951674079604639,
1187
+ "loss": 1.2035815715789795,
1188
+ "step": 158
1189
+ },
1190
+ {
1191
+ "epoch": 9.9375,
1192
+ "grad_norm": 0.020640619099140167,
1193
+ "learning_rate": 0.0029499732274209777,
1194
+ "loss": 1.1313148736953735,
1195
+ "step": 159
1196
+ },
1197
+ {
1198
+ "epoch": 10.0,
1199
+ "grad_norm": 0.04939777031540871,
1200
+ "learning_rate": 0.002948243471983625,
1201
+ "loss": 1.1253995895385742,
1202
+ "step": 160
1203
+ }
1204
+ ],
1205
+ "logging_steps": 1,
1206
+ "max_steps": 800,
1207
+ "num_input_tokens_seen": 0,
1208
+ "num_train_epochs": 50,
1209
+ "save_steps": 500,
1210
+ "stateful_callbacks": {
1211
+ "TrainerControl": {
1212
+ "args": {
1213
+ "should_epoch_stop": false,
1214
+ "should_evaluate": true,
1215
+ "should_log": false,
1216
+ "should_save": false,
1217
+ "should_training_stop": false
1218
+ },
1219
+ "attributes": {}
1220
+ }
1221
+ },
1222
+ "total_flos": 2144708025384960.0,
1223
+ "train_batch_size": 8,
1224
+ "trial_name": null,
1225
+ "trial_params": null
1226
+ }
runs/test-model/checkpoint-160/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45cab7e8b9705eceb6dc5bbdf6d75a3f15b3ecba44dfc0a4b5c7d2d8dc6b7db7
3
+ size 5265
runs/test-model/checkpoint-240/chat_template.jinja ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0].role == 'system' %}
4
+ {{- messages[0].content + '\n\n' }}
5
+ {%- endif %}
6
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
7
+ {%- for tool in tools %}
8
+ {{- "\n" }}
9
+ {{- tool | tojson }}
10
+ {%- endfor %}
11
+ {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
12
+ {%- else %}
13
+ {%- if messages[0].role == 'system' %}
14
+ {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
15
+ {%- endif %}
16
+ {%- endif %}
17
+ {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
18
+ {%- for message in messages[::-1] %}
19
+ {%- set index = (messages|length - 1) - loop.index0 %}
20
+ {%- if ns.multi_step_tool and message.role == "user" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
21
+ {%- set ns.multi_step_tool = false %}
22
+ {%- set ns.last_query_index = index %}
23
+ {%- endif %}
24
+ {%- endfor %}
25
+ {%- for message in messages %}
26
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
27
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
28
+ {%- elif message.role == "assistant" %}
29
+ {%- set content = message.content %}
30
+ {%- set reasoning_content = '' %}
31
+ {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
32
+ {%- set reasoning_content = message.reasoning_content %}
33
+ {%- else %}
34
+ {%- if '</think>' in message.content %}
35
+ {%- set content = message.content.split('</think>')[-1].lstrip('\n') %}
36
+ {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
37
+ {%- endif %}
38
+ {%- endif %}
39
+ {%- if loop.index0 > ns.last_query_index %}
40
+ {%- if loop.last or (not loop.last and reasoning_content) %}
41
+ {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
42
+ {%- else %}
43
+ {{- '<|im_start|>' + message.role + '\n' + content }}
44
+ {%- endif %}
45
+ {%- else %}
46
+ {{- '<|im_start|>' + message.role + '\n' + content }}
47
+ {%- endif %}
48
+ {%- if message.tool_calls %}
49
+ {%- for tool_call in message.tool_calls %}
50
+ {%- if (loop.first and content) or (not loop.first) %}
51
+ {{- '\n' }}
52
+ {%- endif %}
53
+ {%- if tool_call.function %}
54
+ {%- set tool_call = tool_call.function %}
55
+ {%- endif %}
56
+ {{- '<tool_call>\n{"name": "' }}
57
+ {{- tool_call.name }}
58
+ {{- '", "arguments": ' }}
59
+ {%- if tool_call.arguments is string %}
60
+ {{- tool_call.arguments }}
61
+ {%- else %}
62
+ {{- tool_call.arguments | tojson }}
63
+ {%- endif %}
64
+ {{- '}\n</tool_call>' }}
65
+ {%- endfor %}
66
+ {%- endif %}
67
+ {{- '<|im_end|>\n' }}
68
+ {%- elif message.role == "tool" %}
69
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
70
+ {{- '<|im_start|>user' }}
71
+ {%- endif %}
72
+ {{- '\n<tool_response>\n' }}
73
+ {{- message.content }}
74
+ {{- '\n</tool_response>' }}
75
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
76
+ {{- '<|im_end|>\n' }}
77
+ {%- endif %}
78
+ {%- endif %}
79
+ {%- endfor %}
80
+ {%- if add_generation_prompt %}
81
+ {{- '<|im_start|>assistant\n' }}
82
+ {%- if enable_thinking is defined and enable_thinking is false %}
83
+ {{- '<think>\n\n</think>\n\n' }}
84
+ {%- endif %}
85
+ {%- endif %}
runs/test-model/checkpoint-240/config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlamaForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": null,
8
+ "dtype": "float32",
9
+ "eos_token_id": 151645,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 512,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 1536,
15
+ "max_position_embeddings": 2048,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "num_attention_heads": 4,
19
+ "num_hidden_layers": 20,
20
+ "num_key_value_heads": 4,
21
+ "pad_token_id": 151645,
22
+ "pretraining_tp": 1,
23
+ "rms_norm_eps": 1e-06,
24
+ "rope_parameters": {
25
+ "rope_theta": 10000.0,
26
+ "rope_type": "default"
27
+ },
28
+ "tie_word_embeddings": true,
29
+ "transformers_version": "5.5.0",
30
+ "use_cache": false,
31
+ "vocab_size": 151671
32
+ }
runs/test-model/checkpoint-240/generation_config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "eos_token_id": [
4
+ 151645
5
+ ],
6
+ "output_attentions": false,
7
+ "output_hidden_states": false,
8
+ "pad_token_id": 151645,
9
+ "transformers_version": "5.5.0",
10
+ "use_cache": true
11
+ }
runs/test-model/checkpoint-240/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bc0adfeb06ac15912b39f89b2490ede0c5e934af3c98e9e56d6b7f4c10dce44e
3
+ size 583356232
runs/test-model/checkpoint-240/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:12d8c230552978e462c54219226052b16dc20d6d949aed1dabee656cf7950546
3
+ size 1166825803
runs/test-model/checkpoint-240/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b83457a5d388987769679ffdec584d99ee95c32f932be11346ca60064470ca6e
3
+ size 14709
runs/test-model/checkpoint-240/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:dfedbe6d0f52f2cf3df90f0461bcecd255a3bff553a3fe44f96dbec4eb2d024d
3
+ size 1465
runs/test-model/checkpoint-240/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59b6776030505d5b441d0727cce137047df748ab15db6ba3a1bac93c123742fb
3
+ size 11424079
runs/test-model/checkpoint-240/tokenizer_config.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "extra_special_tokens": [
9
+ "<|l2r_pred|>",
10
+ "<|r2l_pred|>"
11
+ ],
12
+ "is_local": false,
13
+ "model_max_length": 131072,
14
+ "pad_token": "<|im_end|>",
15
+ "split_special_tokens": false,
16
+ "tokenizer_class": "Qwen2Tokenizer",
17
+ "unk_token": null
18
+ }
runs/test-model/checkpoint-240/trainer_state.json ADDED
@@ -0,0 +1,1826 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 15.0,
6
+ "eval_steps": 500,
7
+ "global_step": 240,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0625,
14
+ "grad_norm": 0.4655202031135559,
15
+ "learning_rate": 0.0,
16
+ "loss": 2.00244402885437,
17
+ "step": 1
18
+ },
19
+ {
20
+ "epoch": 0.125,
21
+ "grad_norm": 0.4847956895828247,
22
+ "learning_rate": 3e-05,
23
+ "loss": 2.008800506591797,
24
+ "step": 2
25
+ },
26
+ {
27
+ "epoch": 0.1875,
28
+ "grad_norm": 0.4177771806716919,
29
+ "learning_rate": 6e-05,
30
+ "loss": 1.9685651063919067,
31
+ "step": 3
32
+ },
33
+ {
34
+ "epoch": 0.25,
35
+ "grad_norm": 0.3550562858581543,
36
+ "learning_rate": 8.999999999999999e-05,
37
+ "loss": 1.9168976545333862,
38
+ "step": 4
39
+ },
40
+ {
41
+ "epoch": 0.3125,
42
+ "grad_norm": 0.23293137550354004,
43
+ "learning_rate": 0.00012,
44
+ "loss": 1.8660516738891602,
45
+ "step": 5
46
+ },
47
+ {
48
+ "epoch": 0.375,
49
+ "grad_norm": 0.13890701532363892,
50
+ "learning_rate": 0.00015000000000000001,
51
+ "loss": 1.8209998607635498,
52
+ "step": 6
53
+ },
54
+ {
55
+ "epoch": 0.4375,
56
+ "grad_norm": 0.08890970796346664,
57
+ "learning_rate": 0.00017999999999999998,
58
+ "loss": 1.806854248046875,
59
+ "step": 7
60
+ },
61
+ {
62
+ "epoch": 0.5,
63
+ "grad_norm": 0.061611492186784744,
64
+ "learning_rate": 0.00021000000000000004,
65
+ "loss": 1.793172836303711,
66
+ "step": 8
67
+ },
68
+ {
69
+ "epoch": 0.5625,
70
+ "grad_norm": 0.05786389857530594,
71
+ "learning_rate": 0.00024,
72
+ "loss": 1.7882877588272095,
73
+ "step": 9
74
+ },
75
+ {
76
+ "epoch": 0.625,
77
+ "grad_norm": 0.03947344794869423,
78
+ "learning_rate": 0.00027,
79
+ "loss": 1.7696826457977295,
80
+ "step": 10
81
+ },
82
+ {
83
+ "epoch": 0.6875,
84
+ "grad_norm": 0.029783710837364197,
85
+ "learning_rate": 0.00030000000000000003,
86
+ "loss": 1.7767844200134277,
87
+ "step": 11
88
+ },
89
+ {
90
+ "epoch": 0.75,
91
+ "grad_norm": 0.025729672983288765,
92
+ "learning_rate": 0.00033,
93
+ "loss": 1.7827998399734497,
94
+ "step": 12
95
+ },
96
+ {
97
+ "epoch": 0.8125,
98
+ "grad_norm": 0.020592188462615013,
99
+ "learning_rate": 0.00035999999999999997,
100
+ "loss": 1.766276478767395,
101
+ "step": 13
102
+ },
103
+ {
104
+ "epoch": 0.875,
105
+ "grad_norm": 0.01877395436167717,
106
+ "learning_rate": 0.00039000000000000005,
107
+ "loss": 1.7644124031066895,
108
+ "step": 14
109
+ },
110
+ {
111
+ "epoch": 0.9375,
112
+ "grad_norm": 0.015196867287158966,
113
+ "learning_rate": 0.00042000000000000007,
114
+ "loss": 1.771270990371704,
115
+ "step": 15
116
+ },
117
+ {
118
+ "epoch": 1.0,
119
+ "grad_norm": 0.014262701384723186,
120
+ "learning_rate": 0.00045,
121
+ "loss": 1.755500316619873,
122
+ "step": 16
123
+ },
124
+ {
125
+ "epoch": 1.0,
126
+ "eval_loss": 0.8736820816993713,
127
+ "eval_runtime": 30.2743,
128
+ "eval_samples_per_second": 3.369,
129
+ "eval_steps_per_second": 0.429,
130
+ "step": 16
131
+ },
132
+ {
133
+ "epoch": 1.0625,
134
+ "grad_norm": 0.01042763702571392,
135
+ "learning_rate": 0.00048,
136
+ "loss": 1.7556875944137573,
137
+ "step": 17
138
+ },
139
+ {
140
+ "epoch": 1.125,
141
+ "grad_norm": 0.009123577736318111,
142
+ "learning_rate": 0.00051,
143
+ "loss": 1.753890037536621,
144
+ "step": 18
145
+ },
146
+ {
147
+ "epoch": 1.1875,
148
+ "grad_norm": 0.009028254076838493,
149
+ "learning_rate": 0.00054,
150
+ "loss": 1.7515414953231812,
151
+ "step": 19
152
+ },
153
+ {
154
+ "epoch": 1.25,
155
+ "grad_norm": 0.009630529209971428,
156
+ "learning_rate": 0.00057,
157
+ "loss": 1.7465155124664307,
158
+ "step": 20
159
+ },
160
+ {
161
+ "epoch": 1.3125,
162
+ "grad_norm": 0.008235524408519268,
163
+ "learning_rate": 0.0006000000000000001,
164
+ "loss": 1.7533702850341797,
165
+ "step": 21
166
+ },
167
+ {
168
+ "epoch": 1.375,
169
+ "grad_norm": 0.009377758949995041,
170
+ "learning_rate": 0.00063,
171
+ "loss": 1.738031029701233,
172
+ "step": 22
173
+ },
174
+ {
175
+ "epoch": 1.4375,
176
+ "grad_norm": 0.008636261336505413,
177
+ "learning_rate": 0.00066,
178
+ "loss": 1.735835313796997,
179
+ "step": 23
180
+ },
181
+ {
182
+ "epoch": 1.5,
183
+ "grad_norm": 0.008001700975000858,
184
+ "learning_rate": 0.0006900000000000001,
185
+ "loss": 1.7505712509155273,
186
+ "step": 24
187
+ },
188
+ {
189
+ "epoch": 1.5625,
190
+ "grad_norm": 0.007517208810895681,
191
+ "learning_rate": 0.0007199999999999999,
192
+ "loss": 1.7330820560455322,
193
+ "step": 25
194
+ },
195
+ {
196
+ "epoch": 1.625,
197
+ "grad_norm": 0.00952192209661007,
198
+ "learning_rate": 0.00075,
199
+ "loss": 1.7318685054779053,
200
+ "step": 26
201
+ },
202
+ {
203
+ "epoch": 1.6875,
204
+ "grad_norm": 0.006763546261936426,
205
+ "learning_rate": 0.0007800000000000001,
206
+ "loss": 1.7329007387161255,
207
+ "step": 27
208
+ },
209
+ {
210
+ "epoch": 1.75,
211
+ "grad_norm": 0.006364013999700546,
212
+ "learning_rate": 0.0008100000000000001,
213
+ "loss": 1.7285535335540771,
214
+ "step": 28
215
+ },
216
+ {
217
+ "epoch": 1.8125,
218
+ "grad_norm": 0.007770394906401634,
219
+ "learning_rate": 0.0008400000000000001,
220
+ "loss": 1.7320656776428223,
221
+ "step": 29
222
+ },
223
+ {
224
+ "epoch": 1.875,
225
+ "grad_norm": 0.006639651022851467,
226
+ "learning_rate": 0.00087,
227
+ "loss": 1.7304966449737549,
228
+ "step": 30
229
+ },
230
+ {
231
+ "epoch": 1.9375,
232
+ "grad_norm": 0.008190952241420746,
233
+ "learning_rate": 0.0009,
234
+ "loss": 1.7399260997772217,
235
+ "step": 31
236
+ },
237
+ {
238
+ "epoch": 2.0,
239
+ "grad_norm": 0.008188036270439625,
240
+ "learning_rate": 0.00093,
241
+ "loss": 1.7375996112823486,
242
+ "step": 32
243
+ },
244
+ {
245
+ "epoch": 2.0,
246
+ "eval_loss": 0.860374927520752,
247
+ "eval_runtime": 30.3049,
248
+ "eval_samples_per_second": 3.366,
249
+ "eval_steps_per_second": 0.429,
250
+ "step": 32
251
+ },
252
+ {
253
+ "epoch": 2.0625,
254
+ "grad_norm": 0.005768604576587677,
255
+ "learning_rate": 0.00096,
256
+ "loss": 1.7309706211090088,
257
+ "step": 33
258
+ },
259
+ {
260
+ "epoch": 2.125,
261
+ "grad_norm": 0.006822533905506134,
262
+ "learning_rate": 0.00099,
263
+ "loss": 1.7321696281433105,
264
+ "step": 34
265
+ },
266
+ {
267
+ "epoch": 2.1875,
268
+ "grad_norm": 0.009974546730518341,
269
+ "learning_rate": 0.00102,
270
+ "loss": 1.7290048599243164,
271
+ "step": 35
272
+ },
273
+ {
274
+ "epoch": 2.25,
275
+ "grad_norm": 0.00856814906001091,
276
+ "learning_rate": 0.00105,
277
+ "loss": 1.7231603860855103,
278
+ "step": 36
279
+ },
280
+ {
281
+ "epoch": 2.3125,
282
+ "grad_norm": 0.011694240383803844,
283
+ "learning_rate": 0.00108,
284
+ "loss": 1.728348731994629,
285
+ "step": 37
286
+ },
287
+ {
288
+ "epoch": 2.375,
289
+ "grad_norm": 0.005738088861107826,
290
+ "learning_rate": 0.00111,
291
+ "loss": 1.7308874130249023,
292
+ "step": 38
293
+ },
294
+ {
295
+ "epoch": 2.4375,
296
+ "grad_norm": 0.010038000531494617,
297
+ "learning_rate": 0.00114,
298
+ "loss": 1.7276837825775146,
299
+ "step": 39
300
+ },
301
+ {
302
+ "epoch": 2.5,
303
+ "grad_norm": 0.01586989127099514,
304
+ "learning_rate": 0.00117,
305
+ "loss": 1.7368175983428955,
306
+ "step": 40
307
+ },
308
+ {
309
+ "epoch": 2.5625,
310
+ "grad_norm": 0.16198667883872986,
311
+ "learning_rate": 0.0012000000000000001,
312
+ "loss": 1.730246663093567,
313
+ "step": 41
314
+ },
315
+ {
316
+ "epoch": 2.625,
317
+ "grad_norm": 0.01653525047004223,
318
+ "learning_rate": 0.00123,
319
+ "loss": 1.7403852939605713,
320
+ "step": 42
321
+ },
322
+ {
323
+ "epoch": 2.6875,
324
+ "grad_norm": 0.01241418905556202,
325
+ "learning_rate": 0.00126,
326
+ "loss": 1.7306722402572632,
327
+ "step": 43
328
+ },
329
+ {
330
+ "epoch": 2.75,
331
+ "grad_norm": 0.022560030221939087,
332
+ "learning_rate": 0.00129,
333
+ "loss": 1.7223260402679443,
334
+ "step": 44
335
+ },
336
+ {
337
+ "epoch": 2.8125,
338
+ "grad_norm": 0.019568368792533875,
339
+ "learning_rate": 0.00132,
340
+ "loss": 1.7304983139038086,
341
+ "step": 45
342
+ },
343
+ {
344
+ "epoch": 2.875,
345
+ "grad_norm": 0.016755228862166405,
346
+ "learning_rate": 0.00135,
347
+ "loss": 1.7223992347717285,
348
+ "step": 46
349
+ },
350
+ {
351
+ "epoch": 2.9375,
352
+ "grad_norm": 0.02981281839311123,
353
+ "learning_rate": 0.0013800000000000002,
354
+ "loss": 1.7094981670379639,
355
+ "step": 47
356
+ },
357
+ {
358
+ "epoch": 3.0,
359
+ "grad_norm": 0.02506650611758232,
360
+ "learning_rate": 0.00141,
361
+ "loss": 1.7054555416107178,
362
+ "step": 48
363
+ },
364
+ {
365
+ "epoch": 3.0,
366
+ "eval_loss": 0.8416222333908081,
367
+ "eval_runtime": 30.3027,
368
+ "eval_samples_per_second": 3.366,
369
+ "eval_steps_per_second": 0.429,
370
+ "step": 48
371
+ },
372
+ {
373
+ "epoch": 3.0625,
374
+ "grad_norm": 0.02260257676243782,
375
+ "learning_rate": 0.0014399999999999999,
376
+ "loss": 1.6888771057128906,
377
+ "step": 49
378
+ },
379
+ {
380
+ "epoch": 3.125,
381
+ "grad_norm": 0.02620658650994301,
382
+ "learning_rate": 0.00147,
383
+ "loss": 1.6639578342437744,
384
+ "step": 50
385
+ },
386
+ {
387
+ "epoch": 3.1875,
388
+ "grad_norm": 0.047354575246572495,
389
+ "learning_rate": 0.0015,
390
+ "loss": 1.6736350059509277,
391
+ "step": 51
392
+ },
393
+ {
394
+ "epoch": 3.25,
395
+ "grad_norm": 0.04594860598444939,
396
+ "learning_rate": 0.0015300000000000001,
397
+ "loss": 1.6675360202789307,
398
+ "step": 52
399
+ },
400
+ {
401
+ "epoch": 3.3125,
402
+ "grad_norm": 0.08131621032953262,
403
+ "learning_rate": 0.0015600000000000002,
404
+ "loss": 1.6574077606201172,
405
+ "step": 53
406
+ },
407
+ {
408
+ "epoch": 3.375,
409
+ "grad_norm": 0.028789900243282318,
410
+ "learning_rate": 0.00159,
411
+ "loss": 1.6299148797988892,
412
+ "step": 54
413
+ },
414
+ {
415
+ "epoch": 3.4375,
416
+ "grad_norm": 0.2623404860496521,
417
+ "learning_rate": 0.0016200000000000001,
418
+ "loss": 1.6315665245056152,
419
+ "step": 55
420
+ },
421
+ {
422
+ "epoch": 3.5,
423
+ "grad_norm": 0.056828465312719345,
424
+ "learning_rate": 0.0016500000000000002,
425
+ "loss": 1.6279325485229492,
426
+ "step": 56
427
+ },
428
+ {
429
+ "epoch": 3.5625,
430
+ "grad_norm": 0.0549125075340271,
431
+ "learning_rate": 0.0016800000000000003,
432
+ "loss": 1.6549303531646729,
433
+ "step": 57
434
+ },
435
+ {
436
+ "epoch": 3.625,
437
+ "grad_norm": 0.07287958264350891,
438
+ "learning_rate": 0.00171,
439
+ "loss": 1.606963872909546,
440
+ "step": 58
441
+ },
442
+ {
443
+ "epoch": 3.6875,
444
+ "grad_norm": 0.07426343113183975,
445
+ "learning_rate": 0.00174,
446
+ "loss": 1.6250534057617188,
447
+ "step": 59
448
+ },
449
+ {
450
+ "epoch": 3.75,
451
+ "grad_norm": 0.1364986002445221,
452
+ "learning_rate": 0.0017699999999999999,
453
+ "loss": 1.5964961051940918,
454
+ "step": 60
455
+ },
456
+ {
457
+ "epoch": 3.8125,
458
+ "grad_norm": 0.15007159113883972,
459
+ "learning_rate": 0.0018,
460
+ "loss": 1.6019010543823242,
461
+ "step": 61
462
+ },
463
+ {
464
+ "epoch": 3.875,
465
+ "grad_norm": 0.08447311073541641,
466
+ "learning_rate": 0.00183,
467
+ "loss": 1.5763543844223022,
468
+ "step": 62
469
+ },
470
+ {
471
+ "epoch": 3.9375,
472
+ "grad_norm": 0.1681874394416809,
473
+ "learning_rate": 0.00186,
474
+ "loss": 1.5769858360290527,
475
+ "step": 63
476
+ },
477
+ {
478
+ "epoch": 4.0,
479
+ "grad_norm": 0.09677319973707199,
480
+ "learning_rate": 0.00189,
481
+ "loss": 1.534682273864746,
482
+ "step": 64
483
+ },
484
+ {
485
+ "epoch": 4.0,
486
+ "eval_loss": 0.7682511806488037,
487
+ "eval_runtime": 30.4228,
488
+ "eval_samples_per_second": 3.353,
489
+ "eval_steps_per_second": 0.427,
490
+ "step": 64
491
+ },
492
+ {
493
+ "epoch": 4.0625,
494
+ "grad_norm": 0.42635980248451233,
495
+ "learning_rate": 0.00192,
496
+ "loss": 1.5509562492370605,
497
+ "step": 65
498
+ },
499
+ {
500
+ "epoch": 4.125,
501
+ "grad_norm": 0.8467249274253845,
502
+ "learning_rate": 0.0019500000000000001,
503
+ "loss": 1.5001531839370728,
504
+ "step": 66
505
+ },
506
+ {
507
+ "epoch": 4.1875,
508
+ "grad_norm": 0.22631719708442688,
509
+ "learning_rate": 0.00198,
510
+ "loss": 1.4554829597473145,
511
+ "step": 67
512
+ },
513
+ {
514
+ "epoch": 4.25,
515
+ "grad_norm": 0.05312464386224747,
516
+ "learning_rate": 0.00201,
517
+ "loss": 1.4732775688171387,
518
+ "step": 68
519
+ },
520
+ {
521
+ "epoch": 4.3125,
522
+ "grad_norm": 0.050442785024642944,
523
+ "learning_rate": 0.00204,
524
+ "loss": 1.4176194667816162,
525
+ "step": 69
526
+ },
527
+ {
528
+ "epoch": 4.375,
529
+ "grad_norm": 0.18886858224868774,
530
+ "learning_rate": 0.00207,
531
+ "loss": 1.4325997829437256,
532
+ "step": 70
533
+ },
534
+ {
535
+ "epoch": 4.4375,
536
+ "grad_norm": 0.05258706212043762,
537
+ "learning_rate": 0.0021,
538
+ "loss": 1.39802086353302,
539
+ "step": 71
540
+ },
541
+ {
542
+ "epoch": 4.5,
543
+ "grad_norm": 0.07482657581567764,
544
+ "learning_rate": 0.00213,
545
+ "loss": 1.4000320434570312,
546
+ "step": 72
547
+ },
548
+ {
549
+ "epoch": 4.5625,
550
+ "grad_norm": 0.4853312075138092,
551
+ "learning_rate": 0.00216,
552
+ "loss": 1.4199151992797852,
553
+ "step": 73
554
+ },
555
+ {
556
+ "epoch": 4.625,
557
+ "grad_norm": 0.48888301849365234,
558
+ "learning_rate": 0.00219,
559
+ "loss": 1.3049702644348145,
560
+ "step": 74
561
+ },
562
+ {
563
+ "epoch": 4.6875,
564
+ "grad_norm": 0.284348726272583,
565
+ "learning_rate": 0.00222,
566
+ "loss": 1.3884068727493286,
567
+ "step": 75
568
+ },
569
+ {
570
+ "epoch": 4.75,
571
+ "grad_norm": 0.20248942077159882,
572
+ "learning_rate": 0.0022500000000000003,
573
+ "loss": 1.3502120971679688,
574
+ "step": 76
575
+ },
576
+ {
577
+ "epoch": 4.8125,
578
+ "grad_norm": 0.2714093327522278,
579
+ "learning_rate": 0.00228,
580
+ "loss": 1.3195693492889404,
581
+ "step": 77
582
+ },
583
+ {
584
+ "epoch": 4.875,
585
+ "grad_norm": 0.2544061243534088,
586
+ "learning_rate": 0.00231,
587
+ "loss": 1.3194304704666138,
588
+ "step": 78
589
+ },
590
+ {
591
+ "epoch": 4.9375,
592
+ "grad_norm": 0.18475523591041565,
593
+ "learning_rate": 0.00234,
594
+ "loss": 1.3006361722946167,
595
+ "step": 79
596
+ },
597
+ {
598
+ "epoch": 5.0,
599
+ "grad_norm": 0.6583278775215149,
600
+ "learning_rate": 0.00237,
601
+ "loss": 1.2868517637252808,
602
+ "step": 80
603
+ },
604
+ {
605
+ "epoch": 5.0,
606
+ "eval_loss": 0.6618342399597168,
607
+ "eval_runtime": 30.1233,
608
+ "eval_samples_per_second": 3.386,
609
+ "eval_steps_per_second": 0.432,
610
+ "step": 80
611
+ },
612
+ {
613
+ "epoch": 5.0625,
614
+ "grad_norm": 0.04499061033129692,
615
+ "learning_rate": 0.0024000000000000002,
616
+ "loss": 1.25992751121521,
617
+ "step": 81
618
+ },
619
+ {
620
+ "epoch": 5.125,
621
+ "grad_norm": 0.05609860643744469,
622
+ "learning_rate": 0.0024300000000000003,
623
+ "loss": 1.2764487266540527,
624
+ "step": 82
625
+ },
626
+ {
627
+ "epoch": 5.1875,
628
+ "grad_norm": 0.03665148839354515,
629
+ "learning_rate": 0.00246,
630
+ "loss": 1.2908766269683838,
631
+ "step": 83
632
+ },
633
+ {
634
+ "epoch": 5.25,
635
+ "grad_norm": 0.17949481308460236,
636
+ "learning_rate": 0.00249,
637
+ "loss": 1.253080129623413,
638
+ "step": 84
639
+ },
640
+ {
641
+ "epoch": 5.3125,
642
+ "grad_norm": 0.14031393826007843,
643
+ "learning_rate": 0.00252,
644
+ "loss": 1.2648870944976807,
645
+ "step": 85
646
+ },
647
+ {
648
+ "epoch": 5.375,
649
+ "grad_norm": 0.08193210512399673,
650
+ "learning_rate": 0.00255,
651
+ "loss": 1.296234130859375,
652
+ "step": 86
653
+ },
654
+ {
655
+ "epoch": 5.4375,
656
+ "grad_norm": 0.056413356214761734,
657
+ "learning_rate": 0.00258,
658
+ "loss": 1.2784960269927979,
659
+ "step": 87
660
+ },
661
+ {
662
+ "epoch": 5.5,
663
+ "grad_norm": 0.09004318714141846,
664
+ "learning_rate": 0.00261,
665
+ "loss": 1.2054948806762695,
666
+ "step": 88
667
+ },
668
+ {
669
+ "epoch": 5.5625,
670
+ "grad_norm": 0.044756047427654266,
671
+ "learning_rate": 0.00264,
672
+ "loss": 1.2707666158676147,
673
+ "step": 89
674
+ },
675
+ {
676
+ "epoch": 5.625,
677
+ "grad_norm": 0.049932755529880524,
678
+ "learning_rate": 0.00267,
679
+ "loss": 1.2369787693023682,
680
+ "step": 90
681
+ },
682
+ {
683
+ "epoch": 5.6875,
684
+ "grad_norm": 0.09624820947647095,
685
+ "learning_rate": 0.0027,
686
+ "loss": 1.2419278621673584,
687
+ "step": 91
688
+ },
689
+ {
690
+ "epoch": 5.75,
691
+ "grad_norm": 0.17648059129714966,
692
+ "learning_rate": 0.0027300000000000002,
693
+ "loss": 1.2992033958435059,
694
+ "step": 92
695
+ },
696
+ {
697
+ "epoch": 5.8125,
698
+ "grad_norm": 0.07027823477983475,
699
+ "learning_rate": 0.0027600000000000003,
700
+ "loss": 1.2255799770355225,
701
+ "step": 93
702
+ },
703
+ {
704
+ "epoch": 5.875,
705
+ "grad_norm": 0.06040462478995323,
706
+ "learning_rate": 0.0027900000000000004,
707
+ "loss": 1.136023759841919,
708
+ "step": 94
709
+ },
710
+ {
711
+ "epoch": 5.9375,
712
+ "grad_norm": 0.09845872968435287,
713
+ "learning_rate": 0.00282,
714
+ "loss": 1.1807364225387573,
715
+ "step": 95
716
+ },
717
+ {
718
+ "epoch": 6.0,
719
+ "grad_norm": 0.09086572378873825,
720
+ "learning_rate": 0.00285,
721
+ "loss": 1.2171341180801392,
722
+ "step": 96
723
+ },
724
+ {
725
+ "epoch": 6.0,
726
+ "eval_loss": 0.6200892329216003,
727
+ "eval_runtime": 30.1252,
728
+ "eval_samples_per_second": 3.386,
729
+ "eval_steps_per_second": 0.432,
730
+ "step": 96
731
+ },
732
+ {
733
+ "epoch": 6.0625,
734
+ "grad_norm": 0.11787034571170807,
735
+ "learning_rate": 0.0028799999999999997,
736
+ "loss": 1.242072582244873,
737
+ "step": 97
738
+ },
739
+ {
740
+ "epoch": 6.125,
741
+ "grad_norm": 0.20093177258968353,
742
+ "learning_rate": 0.00291,
743
+ "loss": 1.2056899070739746,
744
+ "step": 98
745
+ },
746
+ {
747
+ "epoch": 6.1875,
748
+ "grad_norm": 0.03221331909298897,
749
+ "learning_rate": 0.00294,
750
+ "loss": 1.1774635314941406,
751
+ "step": 99
752
+ },
753
+ {
754
+ "epoch": 6.25,
755
+ "grad_norm": 0.06112038344144821,
756
+ "learning_rate": 0.00297,
757
+ "loss": 1.1831618547439575,
758
+ "step": 100
759
+ },
760
+ {
761
+ "epoch": 6.3125,
762
+ "grad_norm": 0.05896962434053421,
763
+ "learning_rate": 0.003,
764
+ "loss": 1.219610571861267,
765
+ "step": 101
766
+ },
767
+ {
768
+ "epoch": 6.375,
769
+ "grad_norm": 0.4068014323711395,
770
+ "learning_rate": 0.0029999850445531283,
771
+ "loss": 1.231442928314209,
772
+ "step": 102
773
+ },
774
+ {
775
+ "epoch": 6.4375,
776
+ "grad_norm": 3.013302803039551,
777
+ "learning_rate": 0.0029999401785137437,
778
+ "loss": 1.2512195110321045,
779
+ "step": 103
780
+ },
781
+ {
782
+ "epoch": 6.5,
783
+ "grad_norm": 0.07003403455018997,
784
+ "learning_rate": 0.0029998654027855406,
785
+ "loss": 1.1912157535552979,
786
+ "step": 104
787
+ },
788
+ {
789
+ "epoch": 6.5625,
790
+ "grad_norm": 1.1927955150604248,
791
+ "learning_rate": 0.002999760718874652,
792
+ "loss": 1.1158745288848877,
793
+ "step": 105
794
+ },
795
+ {
796
+ "epoch": 6.625,
797
+ "grad_norm": 0.11684458702802658,
798
+ "learning_rate": 0.0029996261288896225,
799
+ "loss": 1.1835001707077026,
800
+ "step": 106
801
+ },
802
+ {
803
+ "epoch": 6.6875,
804
+ "grad_norm": 0.07812155783176422,
805
+ "learning_rate": 0.0029994616355413666,
806
+ "loss": 1.1712734699249268,
807
+ "step": 107
808
+ },
809
+ {
810
+ "epoch": 6.75,
811
+ "grad_norm": 0.19128946959972382,
812
+ "learning_rate": 0.0029992672421431115,
813
+ "loss": 1.0906574726104736,
814
+ "step": 108
815
+ },
816
+ {
817
+ "epoch": 6.8125,
818
+ "grad_norm": 0.033625371754169464,
819
+ "learning_rate": 0.002999042952610332,
820
+ "loss": 1.115525245666504,
821
+ "step": 109
822
+ },
823
+ {
824
+ "epoch": 6.875,
825
+ "grad_norm": 0.2518763244152069,
826
+ "learning_rate": 0.0029987887714606723,
827
+ "loss": 1.1433348655700684,
828
+ "step": 110
829
+ },
830
+ {
831
+ "epoch": 6.9375,
832
+ "grad_norm": 0.037159573286771774,
833
+ "learning_rate": 0.002998504703813852,
834
+ "loss": 1.1867601871490479,
835
+ "step": 111
836
+ },
837
+ {
838
+ "epoch": 7.0,
839
+ "grad_norm": 0.04564391076564789,
840
+ "learning_rate": 0.0029981907553915674,
841
+ "loss": 1.1302828788757324,
842
+ "step": 112
843
+ },
844
+ {
845
+ "epoch": 7.0,
846
+ "eval_loss": 0.5897142887115479,
847
+ "eval_runtime": 31.3671,
848
+ "eval_samples_per_second": 3.252,
849
+ "eval_steps_per_second": 0.414,
850
+ "step": 112
851
+ },
852
+ {
853
+ "epoch": 7.0625,
854
+ "grad_norm": 0.05205775797367096,
855
+ "learning_rate": 0.0029978469325173717,
856
+ "loss": 1.182741641998291,
857
+ "step": 113
858
+ },
859
+ {
860
+ "epoch": 7.125,
861
+ "grad_norm": 0.12080052495002747,
862
+ "learning_rate": 0.002997473242116551,
863
+ "loss": 1.1457600593566895,
864
+ "step": 114
865
+ },
866
+ {
867
+ "epoch": 7.1875,
868
+ "grad_norm": 0.05600045621395111,
869
+ "learning_rate": 0.0029970696917159833,
870
+ "loss": 1.1101720333099365,
871
+ "step": 115
872
+ },
873
+ {
874
+ "epoch": 7.25,
875
+ "grad_norm": 0.0507950522005558,
876
+ "learning_rate": 0.0029966362894439874,
877
+ "loss": 1.1785833835601807,
878
+ "step": 116
879
+ },
880
+ {
881
+ "epoch": 7.3125,
882
+ "grad_norm": 0.6396550536155701,
883
+ "learning_rate": 0.002996173044030159,
884
+ "loss": 1.1538987159729004,
885
+ "step": 117
886
+ },
887
+ {
888
+ "epoch": 7.375,
889
+ "grad_norm": 0.06849977374076843,
890
+ "learning_rate": 0.0029956799648051934,
891
+ "loss": 1.1528857946395874,
892
+ "step": 118
893
+ },
894
+ {
895
+ "epoch": 7.4375,
896
+ "grad_norm": 0.0685896947979927,
897
+ "learning_rate": 0.002995157061700702,
898
+ "loss": 1.1049470901489258,
899
+ "step": 119
900
+ },
901
+ {
902
+ "epoch": 7.5,
903
+ "grad_norm": 1.027722716331482,
904
+ "learning_rate": 0.002994604345249006,
905
+ "loss": 1.1308050155639648,
906
+ "step": 120
907
+ },
908
+ {
909
+ "epoch": 7.5625,
910
+ "grad_norm": 0.21406109631061554,
911
+ "learning_rate": 0.00299402182658293,
912
+ "loss": 1.0960861444473267,
913
+ "step": 121
914
+ },
915
+ {
916
+ "epoch": 7.625,
917
+ "grad_norm": 0.0503656342625618,
918
+ "learning_rate": 0.002993409517435574,
919
+ "loss": 1.1096141338348389,
920
+ "step": 122
921
+ },
922
+ {
923
+ "epoch": 7.6875,
924
+ "grad_norm": 0.04800846800208092,
925
+ "learning_rate": 0.002992767430140078,
926
+ "loss": 1.1386680603027344,
927
+ "step": 123
928
+ },
929
+ {
930
+ "epoch": 7.75,
931
+ "grad_norm": 0.13091202080249786,
932
+ "learning_rate": 0.002992095577629376,
933
+ "loss": 1.1129441261291504,
934
+ "step": 124
935
+ },
936
+ {
937
+ "epoch": 7.8125,
938
+ "grad_norm": 0.04949796572327614,
939
+ "learning_rate": 0.0029913939734359306,
940
+ "loss": 1.106520175933838,
941
+ "step": 125
942
+ },
943
+ {
944
+ "epoch": 7.875,
945
+ "grad_norm": 0.07686582207679749,
946
+ "learning_rate": 0.0029906626316914655,
947
+ "loss": 1.1572234630584717,
948
+ "step": 126
949
+ },
950
+ {
951
+ "epoch": 7.9375,
952
+ "grad_norm": 0.053587451577186584,
953
+ "learning_rate": 0.0029899015671266773,
954
+ "loss": 1.125717282295227,
955
+ "step": 127
956
+ },
957
+ {
958
+ "epoch": 8.0,
959
+ "grad_norm": 0.08031050115823746,
960
+ "learning_rate": 0.0029891107950709406,
961
+ "loss": 1.153022050857544,
962
+ "step": 128
963
+ },
964
+ {
965
+ "epoch": 8.0,
966
+ "eval_loss": 0.5910838842391968,
967
+ "eval_runtime": 30.7073,
968
+ "eval_samples_per_second": 3.322,
969
+ "eval_steps_per_second": 0.423,
970
+ "step": 128
971
+ },
972
+ {
973
+ "epoch": 8.0625,
974
+ "grad_norm": 0.3520950675010681,
975
+ "learning_rate": 0.0029882903314519997,
976
+ "loss": 1.1582988500595093,
977
+ "step": 129
978
+ },
979
+ {
980
+ "epoch": 8.125,
981
+ "grad_norm": 0.07153584063053131,
982
+ "learning_rate": 0.002987440192795645,
983
+ "loss": 1.1344678401947021,
984
+ "step": 130
985
+ },
986
+ {
987
+ "epoch": 8.1875,
988
+ "grad_norm": 0.06336807459592819,
989
+ "learning_rate": 0.0029865603962253833,
990
+ "loss": 1.120319128036499,
991
+ "step": 131
992
+ },
993
+ {
994
+ "epoch": 8.25,
995
+ "grad_norm": 0.11782127618789673,
996
+ "learning_rate": 0.0029856509594620912,
997
+ "loss": 1.1245379447937012,
998
+ "step": 132
999
+ },
1000
+ {
1001
+ "epoch": 8.3125,
1002
+ "grad_norm": 0.2592264413833618,
1003
+ "learning_rate": 0.0029847119008236576,
1004
+ "loss": 1.1454271078109741,
1005
+ "step": 133
1006
+ },
1007
+ {
1008
+ "epoch": 8.375,
1009
+ "grad_norm": 0.21255695819854736,
1010
+ "learning_rate": 0.0029837432392246185,
1011
+ "loss": 1.231805443763733,
1012
+ "step": 134
1013
+ },
1014
+ {
1015
+ "epoch": 8.4375,
1016
+ "grad_norm": 15.057058334350586,
1017
+ "learning_rate": 0.0029827449941757698,
1018
+ "loss": 1.1380956172943115,
1019
+ "step": 135
1020
+ },
1021
+ {
1022
+ "epoch": 8.5,
1023
+ "grad_norm": 0.17857126891613007,
1024
+ "learning_rate": 0.00298171718578378,
1025
+ "loss": 1.1401441097259521,
1026
+ "step": 136
1027
+ },
1028
+ {
1029
+ "epoch": 8.5625,
1030
+ "grad_norm": 0.18281924724578857,
1031
+ "learning_rate": 0.002980659834750781,
1032
+ "loss": 1.171494722366333,
1033
+ "step": 137
1034
+ },
1035
+ {
1036
+ "epoch": 8.625,
1037
+ "grad_norm": 0.10490825027227402,
1038
+ "learning_rate": 0.0029795729623739555,
1039
+ "loss": 1.1622605323791504,
1040
+ "step": 138
1041
+ },
1042
+ {
1043
+ "epoch": 8.6875,
1044
+ "grad_norm": 0.344013512134552,
1045
+ "learning_rate": 0.002978456590545104,
1046
+ "loss": 1.1702141761779785,
1047
+ "step": 139
1048
+ },
1049
+ {
1050
+ "epoch": 8.75,
1051
+ "grad_norm": 0.05058368295431137,
1052
+ "learning_rate": 0.0029773107417502055,
1053
+ "loss": 1.2006211280822754,
1054
+ "step": 140
1055
+ },
1056
+ {
1057
+ "epoch": 8.8125,
1058
+ "grad_norm": 0.09775760769844055,
1059
+ "learning_rate": 0.002976135439068965,
1060
+ "loss": 1.1407063007354736,
1061
+ "step": 141
1062
+ },
1063
+ {
1064
+ "epoch": 8.875,
1065
+ "grad_norm": 0.0785030648112297,
1066
+ "learning_rate": 0.0029749307061743475,
1067
+ "loss": 1.1778209209442139,
1068
+ "step": 142
1069
+ },
1070
+ {
1071
+ "epoch": 8.9375,
1072
+ "grad_norm": 0.053056441247463226,
1073
+ "learning_rate": 0.0029736965673321032,
1074
+ "loss": 1.1302082538604736,
1075
+ "step": 143
1076
+ },
1077
+ {
1078
+ "epoch": 9.0,
1079
+ "grad_norm": 0.4020994305610657,
1080
+ "learning_rate": 0.0029724330474002743,
1081
+ "loss": 1.1655206680297852,
1082
+ "step": 144
1083
+ },
1084
+ {
1085
+ "epoch": 9.0,
1086
+ "eval_loss": 0.6083483099937439,
1087
+ "eval_runtime": 30.0453,
1088
+ "eval_samples_per_second": 3.395,
1089
+ "eval_steps_per_second": 0.433,
1090
+ "step": 144
1091
+ },
1092
+ {
1093
+ "epoch": 9.0625,
1094
+ "grad_norm": 0.04755420982837677,
1095
+ "learning_rate": 0.002971140171828702,
1096
+ "loss": 1.1900575160980225,
1097
+ "step": 145
1098
+ },
1099
+ {
1100
+ "epoch": 9.125,
1101
+ "grad_norm": 0.2549609839916229,
1102
+ "learning_rate": 0.002969817966658505,
1103
+ "loss": 1.1891400814056396,
1104
+ "step": 146
1105
+ },
1106
+ {
1107
+ "epoch": 9.1875,
1108
+ "grad_norm": 0.11641769111156464,
1109
+ "learning_rate": 0.0029684664585215618,
1110
+ "loss": 1.1930484771728516,
1111
+ "step": 147
1112
+ },
1113
+ {
1114
+ "epoch": 9.25,
1115
+ "grad_norm": 0.13052672147750854,
1116
+ "learning_rate": 0.00296708567463997,
1117
+ "loss": 1.1905772686004639,
1118
+ "step": 148
1119
+ },
1120
+ {
1121
+ "epoch": 9.3125,
1122
+ "grad_norm": 0.15126636624336243,
1123
+ "learning_rate": 0.0029656756428255017,
1124
+ "loss": 1.1417806148529053,
1125
+ "step": 149
1126
+ },
1127
+ {
1128
+ "epoch": 9.375,
1129
+ "grad_norm": 0.08826905488967896,
1130
+ "learning_rate": 0.00296423639147904,
1131
+ "loss": 1.1802985668182373,
1132
+ "step": 150
1133
+ },
1134
+ {
1135
+ "epoch": 9.4375,
1136
+ "grad_norm": 0.01381592359393835,
1137
+ "learning_rate": 0.002962767949590008,
1138
+ "loss": 1.1525968313217163,
1139
+ "step": 151
1140
+ },
1141
+ {
1142
+ "epoch": 9.5,
1143
+ "grad_norm": 0.04648139327764511,
1144
+ "learning_rate": 0.0029612703467357866,
1145
+ "loss": 1.169856309890747,
1146
+ "step": 152
1147
+ },
1148
+ {
1149
+ "epoch": 9.5625,
1150
+ "grad_norm": 0.014019235037267208,
1151
+ "learning_rate": 0.0029597436130811153,
1152
+ "loss": 1.1442453861236572,
1153
+ "step": 153
1154
+ },
1155
+ {
1156
+ "epoch": 9.625,
1157
+ "grad_norm": 0.03701096400618553,
1158
+ "learning_rate": 0.0029581877793774885,
1159
+ "loss": 1.112680435180664,
1160
+ "step": 154
1161
+ },
1162
+ {
1163
+ "epoch": 9.6875,
1164
+ "grad_norm": 0.030421797186136246,
1165
+ "learning_rate": 0.0029566028769625334,
1166
+ "loss": 1.223894715309143,
1167
+ "step": 155
1168
+ },
1169
+ {
1170
+ "epoch": 9.75,
1171
+ "grad_norm": 0.12336083501577377,
1172
+ "learning_rate": 0.00295498893775938,
1173
+ "loss": 1.220346212387085,
1174
+ "step": 156
1175
+ },
1176
+ {
1177
+ "epoch": 9.8125,
1178
+ "grad_norm": 0.038585737347602844,
1179
+ "learning_rate": 0.0029533459942760174,
1180
+ "loss": 1.2078583240509033,
1181
+ "step": 157
1182
+ },
1183
+ {
1184
+ "epoch": 9.875,
1185
+ "grad_norm": 0.4653286039829254,
1186
+ "learning_rate": 0.002951674079604639,
1187
+ "loss": 1.2035815715789795,
1188
+ "step": 158
1189
+ },
1190
+ {
1191
+ "epoch": 9.9375,
1192
+ "grad_norm": 0.020640619099140167,
1193
+ "learning_rate": 0.0029499732274209777,
1194
+ "loss": 1.1313148736953735,
1195
+ "step": 159
1196
+ },
1197
+ {
1198
+ "epoch": 10.0,
1199
+ "grad_norm": 0.04939777031540871,
1200
+ "learning_rate": 0.002948243471983625,
1201
+ "loss": 1.1253995895385742,
1202
+ "step": 160
1203
+ },
1204
+ {
1205
+ "epoch": 10.0,
1206
+ "eval_loss": 0.604421079158783,
1207
+ "eval_runtime": 30.1556,
1208
+ "eval_samples_per_second": 3.382,
1209
+ "eval_steps_per_second": 0.431,
1210
+ "step": 160
1211
+ },
1212
+ {
1213
+ "epoch": 10.0625,
1214
+ "grad_norm": 0.05968557298183441,
1215
+ "learning_rate": 0.0029464848481333423,
1216
+ "loss": 1.1324307918548584,
1217
+ "step": 161
1218
+ },
1219
+ {
1220
+ "epoch": 10.125,
1221
+ "grad_norm": 1.0001988410949707,
1222
+ "learning_rate": 0.0029446973912923597,
1223
+ "loss": 1.126110315322876,
1224
+ "step": 162
1225
+ },
1226
+ {
1227
+ "epoch": 10.1875,
1228
+ "grad_norm": 0.2294934093952179,
1229
+ "learning_rate": 0.002942881137463661,
1230
+ "loss": 1.1873570680618286,
1231
+ "step": 163
1232
+ },
1233
+ {
1234
+ "epoch": 10.25,
1235
+ "grad_norm": 0.06396346539258957,
1236
+ "learning_rate": 0.002941036123230261,
1237
+ "loss": 1.1288526058197021,
1238
+ "step": 164
1239
+ },
1240
+ {
1241
+ "epoch": 10.3125,
1242
+ "grad_norm": 0.3410445749759674,
1243
+ "learning_rate": 0.0029391623857544643,
1244
+ "loss": 1.1219967603683472,
1245
+ "step": 165
1246
+ },
1247
+ {
1248
+ "epoch": 10.375,
1249
+ "grad_norm": 0.09721948206424713,
1250
+ "learning_rate": 0.002937259962777122,
1251
+ "loss": 1.141520619392395,
1252
+ "step": 166
1253
+ },
1254
+ {
1255
+ "epoch": 10.4375,
1256
+ "grad_norm": 0.033714450895786285,
1257
+ "learning_rate": 0.002935328892616868,
1258
+ "loss": 1.1983453035354614,
1259
+ "step": 167
1260
+ },
1261
+ {
1262
+ "epoch": 10.5,
1263
+ "grad_norm": 0.028446367010474205,
1264
+ "learning_rate": 0.0029333692141693484,
1265
+ "loss": 1.153167724609375,
1266
+ "step": 168
1267
+ },
1268
+ {
1269
+ "epoch": 10.5625,
1270
+ "grad_norm": 0.04298163205385208,
1271
+ "learning_rate": 0.0029313809669064377,
1272
+ "loss": 1.120814561843872,
1273
+ "step": 169
1274
+ },
1275
+ {
1276
+ "epoch": 10.625,
1277
+ "grad_norm": 0.08190953731536865,
1278
+ "learning_rate": 0.002929364190875444,
1279
+ "loss": 1.1164218187332153,
1280
+ "step": 170
1281
+ },
1282
+ {
1283
+ "epoch": 10.6875,
1284
+ "grad_norm": 0.0602508969604969,
1285
+ "learning_rate": 0.0029273189266983035,
1286
+ "loss": 1.136858344078064,
1287
+ "step": 171
1288
+ },
1289
+ {
1290
+ "epoch": 10.75,
1291
+ "grad_norm": 0.07118846476078033,
1292
+ "learning_rate": 0.0029252452155707582,
1293
+ "loss": 1.1356765031814575,
1294
+ "step": 172
1295
+ },
1296
+ {
1297
+ "epoch": 10.8125,
1298
+ "grad_norm": 0.0445970743894577,
1299
+ "learning_rate": 0.002923143099261531,
1300
+ "loss": 1.121457815170288,
1301
+ "step": 173
1302
+ },
1303
+ {
1304
+ "epoch": 10.875,
1305
+ "grad_norm": 0.07745611667633057,
1306
+ "learning_rate": 0.0029210126201114836,
1307
+ "loss": 1.0903944969177246,
1308
+ "step": 174
1309
+ },
1310
+ {
1311
+ "epoch": 10.9375,
1312
+ "grad_norm": 0.08945920318365097,
1313
+ "learning_rate": 0.0029188538210327595,
1314
+ "loss": 1.1194133758544922,
1315
+ "step": 175
1316
+ },
1317
+ {
1318
+ "epoch": 11.0,
1319
+ "grad_norm": 0.07996643334627151,
1320
+ "learning_rate": 0.002916666745507926,
1321
+ "loss": 1.1248202323913574,
1322
+ "step": 176
1323
+ },
1324
+ {
1325
+ "epoch": 11.0,
1326
+ "eval_loss": 0.5789638757705688,
1327
+ "eval_runtime": 30.3538,
1328
+ "eval_samples_per_second": 3.36,
1329
+ "eval_steps_per_second": 0.428,
1330
+ "step": 176
1331
+ },
1332
+ {
1333
+ "epoch": 11.0625,
1334
+ "grad_norm": 0.10533642023801804,
1335
+ "learning_rate": 0.002914451437589092,
1336
+ "loss": 1.1199195384979248,
1337
+ "step": 177
1338
+ },
1339
+ {
1340
+ "epoch": 11.125,
1341
+ "grad_norm": 0.044989872723817825,
1342
+ "learning_rate": 0.002912207941897025,
1343
+ "loss": 1.1084158420562744,
1344
+ "step": 178
1345
+ },
1346
+ {
1347
+ "epoch": 11.1875,
1348
+ "grad_norm": 0.04824615642428398,
1349
+ "learning_rate": 0.0029099363036202515,
1350
+ "loss": 1.0993731021881104,
1351
+ "step": 179
1352
+ },
1353
+ {
1354
+ "epoch": 11.25,
1355
+ "grad_norm": 0.03144732117652893,
1356
+ "learning_rate": 0.0029076365685141463,
1357
+ "loss": 1.1078145503997803,
1358
+ "step": 180
1359
+ },
1360
+ {
1361
+ "epoch": 11.3125,
1362
+ "grad_norm": 0.03728745877742767,
1363
+ "learning_rate": 0.0029053087829000097,
1364
+ "loss": 1.0995632410049438,
1365
+ "step": 181
1366
+ },
1367
+ {
1368
+ "epoch": 11.375,
1369
+ "grad_norm": 0.02090257592499256,
1370
+ "learning_rate": 0.002902952993664138,
1371
+ "loss": 1.0722652673721313,
1372
+ "step": 182
1373
+ },
1374
+ {
1375
+ "epoch": 11.4375,
1376
+ "grad_norm": 0.020907679572701454,
1377
+ "learning_rate": 0.0029005692482568747,
1378
+ "loss": 1.0840482711791992,
1379
+ "step": 183
1380
+ },
1381
+ {
1382
+ "epoch": 11.5,
1383
+ "grad_norm": 0.023914694786071777,
1384
+ "learning_rate": 0.002898157594691658,
1385
+ "loss": 1.1289193630218506,
1386
+ "step": 184
1387
+ },
1388
+ {
1389
+ "epoch": 11.5625,
1390
+ "grad_norm": 0.01937722973525524,
1391
+ "learning_rate": 0.0028957180815440534,
1392
+ "loss": 1.1244895458221436,
1393
+ "step": 185
1394
+ },
1395
+ {
1396
+ "epoch": 11.625,
1397
+ "grad_norm": 0.02047949656844139,
1398
+ "learning_rate": 0.002893250757950773,
1399
+ "loss": 1.0527830123901367,
1400
+ "step": 186
1401
+ },
1402
+ {
1403
+ "epoch": 11.6875,
1404
+ "grad_norm": 0.04898907244205475,
1405
+ "learning_rate": 0.002890755673608688,
1406
+ "loss": 1.1234995126724243,
1407
+ "step": 187
1408
+ },
1409
+ {
1410
+ "epoch": 11.75,
1411
+ "grad_norm": 0.024841446429491043,
1412
+ "learning_rate": 0.0028882328787738273,
1413
+ "loss": 1.138411045074463,
1414
+ "step": 188
1415
+ },
1416
+ {
1417
+ "epoch": 11.8125,
1418
+ "grad_norm": 0.02533135376870632,
1419
+ "learning_rate": 0.002885682424260365,
1420
+ "loss": 1.1021623611450195,
1421
+ "step": 189
1422
+ },
1423
+ {
1424
+ "epoch": 11.875,
1425
+ "grad_norm": 0.04309404641389847,
1426
+ "learning_rate": 0.0028831043614395966,
1427
+ "loss": 1.0767666101455688,
1428
+ "step": 190
1429
+ },
1430
+ {
1431
+ "epoch": 11.9375,
1432
+ "grad_norm": 0.029682395979762077,
1433
+ "learning_rate": 0.0028804987422389045,
1434
+ "loss": 1.1310434341430664,
1435
+ "step": 191
1436
+ },
1437
+ {
1438
+ "epoch": 12.0,
1439
+ "grad_norm": 0.021236268803477287,
1440
+ "learning_rate": 0.002877865619140712,
1441
+ "loss": 1.0969300270080566,
1442
+ "step": 192
1443
+ },
1444
+ {
1445
+ "epoch": 12.0,
1446
+ "eval_loss": 0.5664089322090149,
1447
+ "eval_runtime": 30.1046,
1448
+ "eval_samples_per_second": 3.388,
1449
+ "eval_steps_per_second": 0.432,
1450
+ "step": 192
1451
+ },
1452
+ {
1453
+ "epoch": 12.0625,
1454
+ "grad_norm": 0.09511730819940567,
1455
+ "learning_rate": 0.0028752050451814276,
1456
+ "loss": 1.0866225957870483,
1457
+ "step": 193
1458
+ },
1459
+ {
1460
+ "epoch": 12.125,
1461
+ "grad_norm": 0.1858353465795517,
1462
+ "learning_rate": 0.0028725170739503734,
1463
+ "loss": 1.1082183122634888,
1464
+ "step": 194
1465
+ },
1466
+ {
1467
+ "epoch": 12.1875,
1468
+ "grad_norm": 0.018049633130431175,
1469
+ "learning_rate": 0.00286980175958871,
1470
+ "loss": 1.0569005012512207,
1471
+ "step": 195
1472
+ },
1473
+ {
1474
+ "epoch": 12.25,
1475
+ "grad_norm": 0.06167157366871834,
1476
+ "learning_rate": 0.002867059156788342,
1477
+ "loss": 1.1418097019195557,
1478
+ "step": 196
1479
+ },
1480
+ {
1481
+ "epoch": 12.3125,
1482
+ "grad_norm": 0.06766022741794586,
1483
+ "learning_rate": 0.0028642893207908187,
1484
+ "loss": 1.0995159149169922,
1485
+ "step": 197
1486
+ },
1487
+ {
1488
+ "epoch": 12.375,
1489
+ "grad_norm": 0.03883042931556702,
1490
+ "learning_rate": 0.0028614923073862213,
1491
+ "loss": 1.0847246646881104,
1492
+ "step": 198
1493
+ },
1494
+ {
1495
+ "epoch": 12.4375,
1496
+ "grad_norm": 0.05537786707282066,
1497
+ "learning_rate": 0.0028586681729120393,
1498
+ "loss": 1.1187801361083984,
1499
+ "step": 199
1500
+ },
1501
+ {
1502
+ "epoch": 12.5,
1503
+ "grad_norm": 0.0653286948800087,
1504
+ "learning_rate": 0.002855816974252033,
1505
+ "loss": 1.0798084735870361,
1506
+ "step": 200
1507
+ },
1508
+ {
1509
+ "epoch": 12.5625,
1510
+ "grad_norm": 0.04398065432906151,
1511
+ "learning_rate": 0.002852938768835092,
1512
+ "loss": 1.0946357250213623,
1513
+ "step": 201
1514
+ },
1515
+ {
1516
+ "epoch": 12.625,
1517
+ "grad_norm": 0.05614284798502922,
1518
+ "learning_rate": 0.0028500336146340772,
1519
+ "loss": 1.1144163608551025,
1520
+ "step": 202
1521
+ },
1522
+ {
1523
+ "epoch": 12.6875,
1524
+ "grad_norm": 0.035650063306093216,
1525
+ "learning_rate": 0.0028471015701646485,
1526
+ "loss": 1.0421898365020752,
1527
+ "step": 203
1528
+ },
1529
+ {
1530
+ "epoch": 12.75,
1531
+ "grad_norm": 0.02544066496193409,
1532
+ "learning_rate": 0.002844142694484094,
1533
+ "loss": 1.0994564294815063,
1534
+ "step": 204
1535
+ },
1536
+ {
1537
+ "epoch": 12.8125,
1538
+ "grad_norm": 0.07384507358074188,
1539
+ "learning_rate": 0.002841157047190135,
1540
+ "loss": 1.0665926933288574,
1541
+ "step": 205
1542
+ },
1543
+ {
1544
+ "epoch": 12.875,
1545
+ "grad_norm": 0.04875814542174339,
1546
+ "learning_rate": 0.002838144688419726,
1547
+ "loss": 1.0665624141693115,
1548
+ "step": 206
1549
+ },
1550
+ {
1551
+ "epoch": 12.9375,
1552
+ "grad_norm": 0.05739036574959755,
1553
+ "learning_rate": 0.002835105678847847,
1554
+ "loss": 1.1085939407348633,
1555
+ "step": 207
1556
+ },
1557
+ {
1558
+ "epoch": 13.0,
1559
+ "grad_norm": 0.03088603913784027,
1560
+ "learning_rate": 0.002832040079686276,
1561
+ "loss": 1.1357779502868652,
1562
+ "step": 208
1563
+ },
1564
+ {
1565
+ "epoch": 13.0,
1566
+ "eval_loss": 0.5675534009933472,
1567
+ "eval_runtime": 31.0754,
1568
+ "eval_samples_per_second": 3.282,
1569
+ "eval_steps_per_second": 0.418,
1570
+ "step": 208
1571
+ },
1572
+ {
1573
+ "epoch": 13.0625,
1574
+ "grad_norm": 0.052636317908763885,
1575
+ "learning_rate": 0.0028289479526823614,
1576
+ "loss": 1.0684641599655151,
1577
+ "step": 209
1578
+ },
1579
+ {
1580
+ "epoch": 13.125,
1581
+ "grad_norm": 0.26156774163246155,
1582
+ "learning_rate": 0.0028258293601177743,
1583
+ "loss": 1.1112414598464966,
1584
+ "step": 210
1585
+ },
1586
+ {
1587
+ "epoch": 13.1875,
1588
+ "grad_norm": 0.013046488165855408,
1589
+ "learning_rate": 0.0028226843648072562,
1590
+ "loss": 1.0339481830596924,
1591
+ "step": 211
1592
+ },
1593
+ {
1594
+ "epoch": 13.25,
1595
+ "grad_norm": 0.029812589287757874,
1596
+ "learning_rate": 0.002819513030097352,
1597
+ "loss": 1.0734567642211914,
1598
+ "step": 212
1599
+ },
1600
+ {
1601
+ "epoch": 13.3125,
1602
+ "grad_norm": 0.03711036592721939,
1603
+ "learning_rate": 0.0028163154198651372,
1604
+ "loss": 1.071413516998291,
1605
+ "step": 213
1606
+ },
1607
+ {
1608
+ "epoch": 13.375,
1609
+ "grad_norm": 0.09543009102344513,
1610
+ "learning_rate": 0.002813091598516928,
1611
+ "loss": 1.1175503730773926,
1612
+ "step": 214
1613
+ },
1614
+ {
1615
+ "epoch": 13.4375,
1616
+ "grad_norm": 0.02742145210504532,
1617
+ "learning_rate": 0.0028098416309869857,
1618
+ "loss": 1.0401530265808105,
1619
+ "step": 215
1620
+ },
1621
+ {
1622
+ "epoch": 13.5,
1623
+ "grad_norm": 0.05347360298037529,
1624
+ "learning_rate": 0.002806565582736208,
1625
+ "loss": 1.053802728652954,
1626
+ "step": 216
1627
+ },
1628
+ {
1629
+ "epoch": 13.5625,
1630
+ "grad_norm": 0.023394430056214333,
1631
+ "learning_rate": 0.002803263519750811,
1632
+ "loss": 1.0774943828582764,
1633
+ "step": 217
1634
+ },
1635
+ {
1636
+ "epoch": 13.625,
1637
+ "grad_norm": 0.03274613991379738,
1638
+ "learning_rate": 0.0027999355085409996,
1639
+ "loss": 1.1332664489746094,
1640
+ "step": 218
1641
+ },
1642
+ {
1643
+ "epoch": 13.6875,
1644
+ "grad_norm": 0.029956744983792305,
1645
+ "learning_rate": 0.00279658161613963,
1646
+ "loss": 1.1192328929901123,
1647
+ "step": 219
1648
+ },
1649
+ {
1650
+ "epoch": 13.75,
1651
+ "grad_norm": 0.027843108400702477,
1652
+ "learning_rate": 0.002793201910100856,
1653
+ "loss": 1.0953505039215088,
1654
+ "step": 220
1655
+ },
1656
+ {
1657
+ "epoch": 13.8125,
1658
+ "grad_norm": 0.3722478151321411,
1659
+ "learning_rate": 0.0027897964584987713,
1660
+ "loss": 1.1045243740081787,
1661
+ "step": 221
1662
+ },
1663
+ {
1664
+ "epoch": 13.875,
1665
+ "grad_norm": 0.05480271577835083,
1666
+ "learning_rate": 0.0027863653299260356,
1667
+ "loss": 1.0857124328613281,
1668
+ "step": 222
1669
+ },
1670
+ {
1671
+ "epoch": 13.9375,
1672
+ "grad_norm": 0.05806222930550575,
1673
+ "learning_rate": 0.0027829085934924958,
1674
+ "loss": 1.092329502105713,
1675
+ "step": 223
1676
+ },
1677
+ {
1678
+ "epoch": 14.0,
1679
+ "grad_norm": 0.07507319748401642,
1680
+ "learning_rate": 0.0027794263188237922,
1681
+ "loss": 1.0946941375732422,
1682
+ "step": 224
1683
+ },
1684
+ {
1685
+ "epoch": 14.0,
1686
+ "eval_loss": 0.5624361038208008,
1687
+ "eval_runtime": 30.3035,
1688
+ "eval_samples_per_second": 3.366,
1689
+ "eval_steps_per_second": 0.429,
1690
+ "step": 224
1691
+ },
1692
+ {
1693
+ "epoch": 14.0625,
1694
+ "grad_norm": 0.08125249296426773,
1695
+ "learning_rate": 0.0027759185760599576,
1696
+ "loss": 1.082047939300537,
1697
+ "step": 225
1698
+ },
1699
+ {
1700
+ "epoch": 14.125,
1701
+ "grad_norm": 0.08838590234518051,
1702
+ "learning_rate": 0.002772385435854002,
1703
+ "loss": 1.0747947692871094,
1704
+ "step": 226
1705
+ },
1706
+ {
1707
+ "epoch": 14.1875,
1708
+ "grad_norm": 0.18887385725975037,
1709
+ "learning_rate": 0.0027688269693704926,
1710
+ "loss": 1.0601060390472412,
1711
+ "step": 227
1712
+ },
1713
+ {
1714
+ "epoch": 14.25,
1715
+ "grad_norm": 0.4315606653690338,
1716
+ "learning_rate": 0.002765243248284118,
1717
+ "loss": 1.167302131652832,
1718
+ "step": 228
1719
+ },
1720
+ {
1721
+ "epoch": 14.3125,
1722
+ "grad_norm": 0.07665648311376572,
1723
+ "learning_rate": 0.002761634344778245,
1724
+ "loss": 1.130738615989685,
1725
+ "step": 229
1726
+ },
1727
+ {
1728
+ "epoch": 14.375,
1729
+ "grad_norm": 0.051125217229127884,
1730
+ "learning_rate": 0.002758000331543466,
1731
+ "loss": 1.0479257106781006,
1732
+ "step": 230
1733
+ },
1734
+ {
1735
+ "epoch": 14.4375,
1736
+ "grad_norm": 0.0658797100186348,
1737
+ "learning_rate": 0.0027543412817761345,
1738
+ "loss": 1.0913889408111572,
1739
+ "step": 231
1740
+ },
1741
+ {
1742
+ "epoch": 14.5,
1743
+ "grad_norm": 0.1936631202697754,
1744
+ "learning_rate": 0.0027506572691768888,
1745
+ "loss": 1.0582077503204346,
1746
+ "step": 232
1747
+ },
1748
+ {
1749
+ "epoch": 14.5625,
1750
+ "grad_norm": 2.202739953994751,
1751
+ "learning_rate": 0.0027469483679491693,
1752
+ "loss": 1.0946757793426514,
1753
+ "step": 233
1754
+ },
1755
+ {
1756
+ "epoch": 14.625,
1757
+ "grad_norm": 0.07856857776641846,
1758
+ "learning_rate": 0.002743214652797724,
1759
+ "loss": 1.074622631072998,
1760
+ "step": 234
1761
+ },
1762
+ {
1763
+ "epoch": 14.6875,
1764
+ "grad_norm": 0.04288165271282196,
1765
+ "learning_rate": 0.002739456198927104,
1766
+ "loss": 1.0788205862045288,
1767
+ "step": 235
1768
+ },
1769
+ {
1770
+ "epoch": 14.75,
1771
+ "grad_norm": 0.08169898390769958,
1772
+ "learning_rate": 0.0027356730820401465,
1773
+ "loss": 1.0906422138214111,
1774
+ "step": 236
1775
+ },
1776
+ {
1777
+ "epoch": 14.8125,
1778
+ "grad_norm": 0.04536404460668564,
1779
+ "learning_rate": 0.002731865378336452,
1780
+ "loss": 1.0542757511138916,
1781
+ "step": 237
1782
+ },
1783
+ {
1784
+ "epoch": 14.875,
1785
+ "grad_norm": 0.0246584415435791,
1786
+ "learning_rate": 0.0027280331645108498,
1787
+ "loss": 1.0401986837387085,
1788
+ "step": 238
1789
+ },
1790
+ {
1791
+ "epoch": 14.9375,
1792
+ "grad_norm": 0.03399847820401192,
1793
+ "learning_rate": 0.0027241765177518523,
1794
+ "loss": 1.109243392944336,
1795
+ "step": 239
1796
+ },
1797
+ {
1798
+ "epoch": 15.0,
1799
+ "grad_norm": 0.12445052713155746,
1800
+ "learning_rate": 0.0027202955157401,
1801
+ "loss": 1.0952987670898438,
1802
+ "step": 240
1803
+ }
1804
+ ],
1805
+ "logging_steps": 1,
1806
+ "max_steps": 800,
1807
+ "num_input_tokens_seen": 0,
1808
+ "num_train_epochs": 50,
1809
+ "save_steps": 500,
1810
+ "stateful_callbacks": {
1811
+ "TrainerControl": {
1812
+ "args": {
1813
+ "should_epoch_stop": false,
1814
+ "should_evaluate": true,
1815
+ "should_log": false,
1816
+ "should_save": false,
1817
+ "should_training_stop": false
1818
+ },
1819
+ "attributes": {}
1820
+ }
1821
+ },
1822
+ "total_flos": 3217062038077440.0,
1823
+ "train_batch_size": 8,
1824
+ "trial_name": null,
1825
+ "trial_params": null
1826
+ }
runs/test-model/checkpoint-240/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45cab7e8b9705eceb6dc5bbdf6d75a3f15b3ecba44dfc0a4b5c7d2d8dc6b7db7
3
+ size 5265
runs/test-model/checkpoint-320/chat_template.jinja ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0].role == 'system' %}
4
+ {{- messages[0].content + '\n\n' }}
5
+ {%- endif %}
6
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
7
+ {%- for tool in tools %}
8
+ {{- "\n" }}
9
+ {{- tool | tojson }}
10
+ {%- endfor %}
11
+ {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
12
+ {%- else %}
13
+ {%- if messages[0].role == 'system' %}
14
+ {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
15
+ {%- endif %}
16
+ {%- endif %}
17
+ {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
18
+ {%- for message in messages[::-1] %}
19
+ {%- set index = (messages|length - 1) - loop.index0 %}
20
+ {%- if ns.multi_step_tool and message.role == "user" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
21
+ {%- set ns.multi_step_tool = false %}
22
+ {%- set ns.last_query_index = index %}
23
+ {%- endif %}
24
+ {%- endfor %}
25
+ {%- for message in messages %}
26
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
27
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
28
+ {%- elif message.role == "assistant" %}
29
+ {%- set content = message.content %}
30
+ {%- set reasoning_content = '' %}
31
+ {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
32
+ {%- set reasoning_content = message.reasoning_content %}
33
+ {%- else %}
34
+ {%- if '</think>' in message.content %}
35
+ {%- set content = message.content.split('</think>')[-1].lstrip('\n') %}
36
+ {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
37
+ {%- endif %}
38
+ {%- endif %}
39
+ {%- if loop.index0 > ns.last_query_index %}
40
+ {%- if loop.last or (not loop.last and reasoning_content) %}
41
+ {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
42
+ {%- else %}
43
+ {{- '<|im_start|>' + message.role + '\n' + content }}
44
+ {%- endif %}
45
+ {%- else %}
46
+ {{- '<|im_start|>' + message.role + '\n' + content }}
47
+ {%- endif %}
48
+ {%- if message.tool_calls %}
49
+ {%- for tool_call in message.tool_calls %}
50
+ {%- if (loop.first and content) or (not loop.first) %}
51
+ {{- '\n' }}
52
+ {%- endif %}
53
+ {%- if tool_call.function %}
54
+ {%- set tool_call = tool_call.function %}
55
+ {%- endif %}
56
+ {{- '<tool_call>\n{"name": "' }}
57
+ {{- tool_call.name }}
58
+ {{- '", "arguments": ' }}
59
+ {%- if tool_call.arguments is string %}
60
+ {{- tool_call.arguments }}
61
+ {%- else %}
62
+ {{- tool_call.arguments | tojson }}
63
+ {%- endif %}
64
+ {{- '}\n</tool_call>' }}
65
+ {%- endfor %}
66
+ {%- endif %}
67
+ {{- '<|im_end|>\n' }}
68
+ {%- elif message.role == "tool" %}
69
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
70
+ {{- '<|im_start|>user' }}
71
+ {%- endif %}
72
+ {{- '\n<tool_response>\n' }}
73
+ {{- message.content }}
74
+ {{- '\n</tool_response>' }}
75
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
76
+ {{- '<|im_end|>\n' }}
77
+ {%- endif %}
78
+ {%- endif %}
79
+ {%- endfor %}
80
+ {%- if add_generation_prompt %}
81
+ {{- '<|im_start|>assistant\n' }}
82
+ {%- if enable_thinking is defined and enable_thinking is false %}
83
+ {{- '<think>\n\n</think>\n\n' }}
84
+ {%- endif %}
85
+ {%- endif %}
runs/test-model/checkpoint-320/config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlamaForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": null,
8
+ "dtype": "float32",
9
+ "eos_token_id": 151645,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 512,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 1536,
15
+ "max_position_embeddings": 2048,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "num_attention_heads": 4,
19
+ "num_hidden_layers": 20,
20
+ "num_key_value_heads": 4,
21
+ "pad_token_id": 151645,
22
+ "pretraining_tp": 1,
23
+ "rms_norm_eps": 1e-06,
24
+ "rope_parameters": {
25
+ "rope_theta": 10000.0,
26
+ "rope_type": "default"
27
+ },
28
+ "tie_word_embeddings": true,
29
+ "transformers_version": "5.5.0",
30
+ "use_cache": false,
31
+ "vocab_size": 151671
32
+ }
runs/test-model/checkpoint-320/generation_config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "eos_token_id": [
4
+ 151645
5
+ ],
6
+ "output_attentions": false,
7
+ "output_hidden_states": false,
8
+ "pad_token_id": 151645,
9
+ "transformers_version": "5.5.0",
10
+ "use_cache": true
11
+ }
runs/test-model/checkpoint-320/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2aa2e118c526b9d0af1ed0fd002df979c49d3ce6f2a967821e998acdca70de76
3
+ size 583356232
runs/test-model/checkpoint-320/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3a3a7d82ce002ce9c6932773a9943a10e26e5f2a8e81b38b0041ad823cf8705a
3
+ size 1166825803
runs/test-model/checkpoint-320/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:92797ce4fcd09363a266ca38b4355b9ce342ee666fccb4450bc247f277b9c363
3
+ size 14709
runs/test-model/checkpoint-320/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d37018ddedbaa9aa250e8eac93c964b8bfb9fa3b839f8ed59513490593702820
3
+ size 1465
runs/test-model/checkpoint-320/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:59b6776030505d5b441d0727cce137047df748ab15db6ba3a1bac93c123742fb
3
+ size 11424079
runs/test-model/checkpoint-320/tokenizer_config.json ADDED
@@ -0,0 +1,18 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": null,
5
+ "clean_up_tokenization_spaces": false,
6
+ "eos_token": "<|im_end|>",
7
+ "errors": "replace",
8
+ "extra_special_tokens": [
9
+ "<|l2r_pred|>",
10
+ "<|r2l_pred|>"
11
+ ],
12
+ "is_local": false,
13
+ "model_max_length": 131072,
14
+ "pad_token": "<|im_end|>",
15
+ "split_special_tokens": false,
16
+ "tokenizer_class": "Qwen2Tokenizer",
17
+ "unk_token": null
18
+ }
runs/test-model/checkpoint-320/trainer_state.json ADDED
@@ -0,0 +1,2426 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 20.0,
6
+ "eval_steps": 500,
7
+ "global_step": 320,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0625,
14
+ "grad_norm": 0.4655202031135559,
15
+ "learning_rate": 0.0,
16
+ "loss": 2.00244402885437,
17
+ "step": 1
18
+ },
19
+ {
20
+ "epoch": 0.125,
21
+ "grad_norm": 0.4847956895828247,
22
+ "learning_rate": 3e-05,
23
+ "loss": 2.008800506591797,
24
+ "step": 2
25
+ },
26
+ {
27
+ "epoch": 0.1875,
28
+ "grad_norm": 0.4177771806716919,
29
+ "learning_rate": 6e-05,
30
+ "loss": 1.9685651063919067,
31
+ "step": 3
32
+ },
33
+ {
34
+ "epoch": 0.25,
35
+ "grad_norm": 0.3550562858581543,
36
+ "learning_rate": 8.999999999999999e-05,
37
+ "loss": 1.9168976545333862,
38
+ "step": 4
39
+ },
40
+ {
41
+ "epoch": 0.3125,
42
+ "grad_norm": 0.23293137550354004,
43
+ "learning_rate": 0.00012,
44
+ "loss": 1.8660516738891602,
45
+ "step": 5
46
+ },
47
+ {
48
+ "epoch": 0.375,
49
+ "grad_norm": 0.13890701532363892,
50
+ "learning_rate": 0.00015000000000000001,
51
+ "loss": 1.8209998607635498,
52
+ "step": 6
53
+ },
54
+ {
55
+ "epoch": 0.4375,
56
+ "grad_norm": 0.08890970796346664,
57
+ "learning_rate": 0.00017999999999999998,
58
+ "loss": 1.806854248046875,
59
+ "step": 7
60
+ },
61
+ {
62
+ "epoch": 0.5,
63
+ "grad_norm": 0.061611492186784744,
64
+ "learning_rate": 0.00021000000000000004,
65
+ "loss": 1.793172836303711,
66
+ "step": 8
67
+ },
68
+ {
69
+ "epoch": 0.5625,
70
+ "grad_norm": 0.05786389857530594,
71
+ "learning_rate": 0.00024,
72
+ "loss": 1.7882877588272095,
73
+ "step": 9
74
+ },
75
+ {
76
+ "epoch": 0.625,
77
+ "grad_norm": 0.03947344794869423,
78
+ "learning_rate": 0.00027,
79
+ "loss": 1.7696826457977295,
80
+ "step": 10
81
+ },
82
+ {
83
+ "epoch": 0.6875,
84
+ "grad_norm": 0.029783710837364197,
85
+ "learning_rate": 0.00030000000000000003,
86
+ "loss": 1.7767844200134277,
87
+ "step": 11
88
+ },
89
+ {
90
+ "epoch": 0.75,
91
+ "grad_norm": 0.025729672983288765,
92
+ "learning_rate": 0.00033,
93
+ "loss": 1.7827998399734497,
94
+ "step": 12
95
+ },
96
+ {
97
+ "epoch": 0.8125,
98
+ "grad_norm": 0.020592188462615013,
99
+ "learning_rate": 0.00035999999999999997,
100
+ "loss": 1.766276478767395,
101
+ "step": 13
102
+ },
103
+ {
104
+ "epoch": 0.875,
105
+ "grad_norm": 0.01877395436167717,
106
+ "learning_rate": 0.00039000000000000005,
107
+ "loss": 1.7644124031066895,
108
+ "step": 14
109
+ },
110
+ {
111
+ "epoch": 0.9375,
112
+ "grad_norm": 0.015196867287158966,
113
+ "learning_rate": 0.00042000000000000007,
114
+ "loss": 1.771270990371704,
115
+ "step": 15
116
+ },
117
+ {
118
+ "epoch": 1.0,
119
+ "grad_norm": 0.014262701384723186,
120
+ "learning_rate": 0.00045,
121
+ "loss": 1.755500316619873,
122
+ "step": 16
123
+ },
124
+ {
125
+ "epoch": 1.0,
126
+ "eval_loss": 0.8736820816993713,
127
+ "eval_runtime": 30.2743,
128
+ "eval_samples_per_second": 3.369,
129
+ "eval_steps_per_second": 0.429,
130
+ "step": 16
131
+ },
132
+ {
133
+ "epoch": 1.0625,
134
+ "grad_norm": 0.01042763702571392,
135
+ "learning_rate": 0.00048,
136
+ "loss": 1.7556875944137573,
137
+ "step": 17
138
+ },
139
+ {
140
+ "epoch": 1.125,
141
+ "grad_norm": 0.009123577736318111,
142
+ "learning_rate": 0.00051,
143
+ "loss": 1.753890037536621,
144
+ "step": 18
145
+ },
146
+ {
147
+ "epoch": 1.1875,
148
+ "grad_norm": 0.009028254076838493,
149
+ "learning_rate": 0.00054,
150
+ "loss": 1.7515414953231812,
151
+ "step": 19
152
+ },
153
+ {
154
+ "epoch": 1.25,
155
+ "grad_norm": 0.009630529209971428,
156
+ "learning_rate": 0.00057,
157
+ "loss": 1.7465155124664307,
158
+ "step": 20
159
+ },
160
+ {
161
+ "epoch": 1.3125,
162
+ "grad_norm": 0.008235524408519268,
163
+ "learning_rate": 0.0006000000000000001,
164
+ "loss": 1.7533702850341797,
165
+ "step": 21
166
+ },
167
+ {
168
+ "epoch": 1.375,
169
+ "grad_norm": 0.009377758949995041,
170
+ "learning_rate": 0.00063,
171
+ "loss": 1.738031029701233,
172
+ "step": 22
173
+ },
174
+ {
175
+ "epoch": 1.4375,
176
+ "grad_norm": 0.008636261336505413,
177
+ "learning_rate": 0.00066,
178
+ "loss": 1.735835313796997,
179
+ "step": 23
180
+ },
181
+ {
182
+ "epoch": 1.5,
183
+ "grad_norm": 0.008001700975000858,
184
+ "learning_rate": 0.0006900000000000001,
185
+ "loss": 1.7505712509155273,
186
+ "step": 24
187
+ },
188
+ {
189
+ "epoch": 1.5625,
190
+ "grad_norm": 0.007517208810895681,
191
+ "learning_rate": 0.0007199999999999999,
192
+ "loss": 1.7330820560455322,
193
+ "step": 25
194
+ },
195
+ {
196
+ "epoch": 1.625,
197
+ "grad_norm": 0.00952192209661007,
198
+ "learning_rate": 0.00075,
199
+ "loss": 1.7318685054779053,
200
+ "step": 26
201
+ },
202
+ {
203
+ "epoch": 1.6875,
204
+ "grad_norm": 0.006763546261936426,
205
+ "learning_rate": 0.0007800000000000001,
206
+ "loss": 1.7329007387161255,
207
+ "step": 27
208
+ },
209
+ {
210
+ "epoch": 1.75,
211
+ "grad_norm": 0.006364013999700546,
212
+ "learning_rate": 0.0008100000000000001,
213
+ "loss": 1.7285535335540771,
214
+ "step": 28
215
+ },
216
+ {
217
+ "epoch": 1.8125,
218
+ "grad_norm": 0.007770394906401634,
219
+ "learning_rate": 0.0008400000000000001,
220
+ "loss": 1.7320656776428223,
221
+ "step": 29
222
+ },
223
+ {
224
+ "epoch": 1.875,
225
+ "grad_norm": 0.006639651022851467,
226
+ "learning_rate": 0.00087,
227
+ "loss": 1.7304966449737549,
228
+ "step": 30
229
+ },
230
+ {
231
+ "epoch": 1.9375,
232
+ "grad_norm": 0.008190952241420746,
233
+ "learning_rate": 0.0009,
234
+ "loss": 1.7399260997772217,
235
+ "step": 31
236
+ },
237
+ {
238
+ "epoch": 2.0,
239
+ "grad_norm": 0.008188036270439625,
240
+ "learning_rate": 0.00093,
241
+ "loss": 1.7375996112823486,
242
+ "step": 32
243
+ },
244
+ {
245
+ "epoch": 2.0,
246
+ "eval_loss": 0.860374927520752,
247
+ "eval_runtime": 30.3049,
248
+ "eval_samples_per_second": 3.366,
249
+ "eval_steps_per_second": 0.429,
250
+ "step": 32
251
+ },
252
+ {
253
+ "epoch": 2.0625,
254
+ "grad_norm": 0.005768604576587677,
255
+ "learning_rate": 0.00096,
256
+ "loss": 1.7309706211090088,
257
+ "step": 33
258
+ },
259
+ {
260
+ "epoch": 2.125,
261
+ "grad_norm": 0.006822533905506134,
262
+ "learning_rate": 0.00099,
263
+ "loss": 1.7321696281433105,
264
+ "step": 34
265
+ },
266
+ {
267
+ "epoch": 2.1875,
268
+ "grad_norm": 0.009974546730518341,
269
+ "learning_rate": 0.00102,
270
+ "loss": 1.7290048599243164,
271
+ "step": 35
272
+ },
273
+ {
274
+ "epoch": 2.25,
275
+ "grad_norm": 0.00856814906001091,
276
+ "learning_rate": 0.00105,
277
+ "loss": 1.7231603860855103,
278
+ "step": 36
279
+ },
280
+ {
281
+ "epoch": 2.3125,
282
+ "grad_norm": 0.011694240383803844,
283
+ "learning_rate": 0.00108,
284
+ "loss": 1.728348731994629,
285
+ "step": 37
286
+ },
287
+ {
288
+ "epoch": 2.375,
289
+ "grad_norm": 0.005738088861107826,
290
+ "learning_rate": 0.00111,
291
+ "loss": 1.7308874130249023,
292
+ "step": 38
293
+ },
294
+ {
295
+ "epoch": 2.4375,
296
+ "grad_norm": 0.010038000531494617,
297
+ "learning_rate": 0.00114,
298
+ "loss": 1.7276837825775146,
299
+ "step": 39
300
+ },
301
+ {
302
+ "epoch": 2.5,
303
+ "grad_norm": 0.01586989127099514,
304
+ "learning_rate": 0.00117,
305
+ "loss": 1.7368175983428955,
306
+ "step": 40
307
+ },
308
+ {
309
+ "epoch": 2.5625,
310
+ "grad_norm": 0.16198667883872986,
311
+ "learning_rate": 0.0012000000000000001,
312
+ "loss": 1.730246663093567,
313
+ "step": 41
314
+ },
315
+ {
316
+ "epoch": 2.625,
317
+ "grad_norm": 0.01653525047004223,
318
+ "learning_rate": 0.00123,
319
+ "loss": 1.7403852939605713,
320
+ "step": 42
321
+ },
322
+ {
323
+ "epoch": 2.6875,
324
+ "grad_norm": 0.01241418905556202,
325
+ "learning_rate": 0.00126,
326
+ "loss": 1.7306722402572632,
327
+ "step": 43
328
+ },
329
+ {
330
+ "epoch": 2.75,
331
+ "grad_norm": 0.022560030221939087,
332
+ "learning_rate": 0.00129,
333
+ "loss": 1.7223260402679443,
334
+ "step": 44
335
+ },
336
+ {
337
+ "epoch": 2.8125,
338
+ "grad_norm": 0.019568368792533875,
339
+ "learning_rate": 0.00132,
340
+ "loss": 1.7304983139038086,
341
+ "step": 45
342
+ },
343
+ {
344
+ "epoch": 2.875,
345
+ "grad_norm": 0.016755228862166405,
346
+ "learning_rate": 0.00135,
347
+ "loss": 1.7223992347717285,
348
+ "step": 46
349
+ },
350
+ {
351
+ "epoch": 2.9375,
352
+ "grad_norm": 0.02981281839311123,
353
+ "learning_rate": 0.0013800000000000002,
354
+ "loss": 1.7094981670379639,
355
+ "step": 47
356
+ },
357
+ {
358
+ "epoch": 3.0,
359
+ "grad_norm": 0.02506650611758232,
360
+ "learning_rate": 0.00141,
361
+ "loss": 1.7054555416107178,
362
+ "step": 48
363
+ },
364
+ {
365
+ "epoch": 3.0,
366
+ "eval_loss": 0.8416222333908081,
367
+ "eval_runtime": 30.3027,
368
+ "eval_samples_per_second": 3.366,
369
+ "eval_steps_per_second": 0.429,
370
+ "step": 48
371
+ },
372
+ {
373
+ "epoch": 3.0625,
374
+ "grad_norm": 0.02260257676243782,
375
+ "learning_rate": 0.0014399999999999999,
376
+ "loss": 1.6888771057128906,
377
+ "step": 49
378
+ },
379
+ {
380
+ "epoch": 3.125,
381
+ "grad_norm": 0.02620658650994301,
382
+ "learning_rate": 0.00147,
383
+ "loss": 1.6639578342437744,
384
+ "step": 50
385
+ },
386
+ {
387
+ "epoch": 3.1875,
388
+ "grad_norm": 0.047354575246572495,
389
+ "learning_rate": 0.0015,
390
+ "loss": 1.6736350059509277,
391
+ "step": 51
392
+ },
393
+ {
394
+ "epoch": 3.25,
395
+ "grad_norm": 0.04594860598444939,
396
+ "learning_rate": 0.0015300000000000001,
397
+ "loss": 1.6675360202789307,
398
+ "step": 52
399
+ },
400
+ {
401
+ "epoch": 3.3125,
402
+ "grad_norm": 0.08131621032953262,
403
+ "learning_rate": 0.0015600000000000002,
404
+ "loss": 1.6574077606201172,
405
+ "step": 53
406
+ },
407
+ {
408
+ "epoch": 3.375,
409
+ "grad_norm": 0.028789900243282318,
410
+ "learning_rate": 0.00159,
411
+ "loss": 1.6299148797988892,
412
+ "step": 54
413
+ },
414
+ {
415
+ "epoch": 3.4375,
416
+ "grad_norm": 0.2623404860496521,
417
+ "learning_rate": 0.0016200000000000001,
418
+ "loss": 1.6315665245056152,
419
+ "step": 55
420
+ },
421
+ {
422
+ "epoch": 3.5,
423
+ "grad_norm": 0.056828465312719345,
424
+ "learning_rate": 0.0016500000000000002,
425
+ "loss": 1.6279325485229492,
426
+ "step": 56
427
+ },
428
+ {
429
+ "epoch": 3.5625,
430
+ "grad_norm": 0.0549125075340271,
431
+ "learning_rate": 0.0016800000000000003,
432
+ "loss": 1.6549303531646729,
433
+ "step": 57
434
+ },
435
+ {
436
+ "epoch": 3.625,
437
+ "grad_norm": 0.07287958264350891,
438
+ "learning_rate": 0.00171,
439
+ "loss": 1.606963872909546,
440
+ "step": 58
441
+ },
442
+ {
443
+ "epoch": 3.6875,
444
+ "grad_norm": 0.07426343113183975,
445
+ "learning_rate": 0.00174,
446
+ "loss": 1.6250534057617188,
447
+ "step": 59
448
+ },
449
+ {
450
+ "epoch": 3.75,
451
+ "grad_norm": 0.1364986002445221,
452
+ "learning_rate": 0.0017699999999999999,
453
+ "loss": 1.5964961051940918,
454
+ "step": 60
455
+ },
456
+ {
457
+ "epoch": 3.8125,
458
+ "grad_norm": 0.15007159113883972,
459
+ "learning_rate": 0.0018,
460
+ "loss": 1.6019010543823242,
461
+ "step": 61
462
+ },
463
+ {
464
+ "epoch": 3.875,
465
+ "grad_norm": 0.08447311073541641,
466
+ "learning_rate": 0.00183,
467
+ "loss": 1.5763543844223022,
468
+ "step": 62
469
+ },
470
+ {
471
+ "epoch": 3.9375,
472
+ "grad_norm": 0.1681874394416809,
473
+ "learning_rate": 0.00186,
474
+ "loss": 1.5769858360290527,
475
+ "step": 63
476
+ },
477
+ {
478
+ "epoch": 4.0,
479
+ "grad_norm": 0.09677319973707199,
480
+ "learning_rate": 0.00189,
481
+ "loss": 1.534682273864746,
482
+ "step": 64
483
+ },
484
+ {
485
+ "epoch": 4.0,
486
+ "eval_loss": 0.7682511806488037,
487
+ "eval_runtime": 30.4228,
488
+ "eval_samples_per_second": 3.353,
489
+ "eval_steps_per_second": 0.427,
490
+ "step": 64
491
+ },
492
+ {
493
+ "epoch": 4.0625,
494
+ "grad_norm": 0.42635980248451233,
495
+ "learning_rate": 0.00192,
496
+ "loss": 1.5509562492370605,
497
+ "step": 65
498
+ },
499
+ {
500
+ "epoch": 4.125,
501
+ "grad_norm": 0.8467249274253845,
502
+ "learning_rate": 0.0019500000000000001,
503
+ "loss": 1.5001531839370728,
504
+ "step": 66
505
+ },
506
+ {
507
+ "epoch": 4.1875,
508
+ "grad_norm": 0.22631719708442688,
509
+ "learning_rate": 0.00198,
510
+ "loss": 1.4554829597473145,
511
+ "step": 67
512
+ },
513
+ {
514
+ "epoch": 4.25,
515
+ "grad_norm": 0.05312464386224747,
516
+ "learning_rate": 0.00201,
517
+ "loss": 1.4732775688171387,
518
+ "step": 68
519
+ },
520
+ {
521
+ "epoch": 4.3125,
522
+ "grad_norm": 0.050442785024642944,
523
+ "learning_rate": 0.00204,
524
+ "loss": 1.4176194667816162,
525
+ "step": 69
526
+ },
527
+ {
528
+ "epoch": 4.375,
529
+ "grad_norm": 0.18886858224868774,
530
+ "learning_rate": 0.00207,
531
+ "loss": 1.4325997829437256,
532
+ "step": 70
533
+ },
534
+ {
535
+ "epoch": 4.4375,
536
+ "grad_norm": 0.05258706212043762,
537
+ "learning_rate": 0.0021,
538
+ "loss": 1.39802086353302,
539
+ "step": 71
540
+ },
541
+ {
542
+ "epoch": 4.5,
543
+ "grad_norm": 0.07482657581567764,
544
+ "learning_rate": 0.00213,
545
+ "loss": 1.4000320434570312,
546
+ "step": 72
547
+ },
548
+ {
549
+ "epoch": 4.5625,
550
+ "grad_norm": 0.4853312075138092,
551
+ "learning_rate": 0.00216,
552
+ "loss": 1.4199151992797852,
553
+ "step": 73
554
+ },
555
+ {
556
+ "epoch": 4.625,
557
+ "grad_norm": 0.48888301849365234,
558
+ "learning_rate": 0.00219,
559
+ "loss": 1.3049702644348145,
560
+ "step": 74
561
+ },
562
+ {
563
+ "epoch": 4.6875,
564
+ "grad_norm": 0.284348726272583,
565
+ "learning_rate": 0.00222,
566
+ "loss": 1.3884068727493286,
567
+ "step": 75
568
+ },
569
+ {
570
+ "epoch": 4.75,
571
+ "grad_norm": 0.20248942077159882,
572
+ "learning_rate": 0.0022500000000000003,
573
+ "loss": 1.3502120971679688,
574
+ "step": 76
575
+ },
576
+ {
577
+ "epoch": 4.8125,
578
+ "grad_norm": 0.2714093327522278,
579
+ "learning_rate": 0.00228,
580
+ "loss": 1.3195693492889404,
581
+ "step": 77
582
+ },
583
+ {
584
+ "epoch": 4.875,
585
+ "grad_norm": 0.2544061243534088,
586
+ "learning_rate": 0.00231,
587
+ "loss": 1.3194304704666138,
588
+ "step": 78
589
+ },
590
+ {
591
+ "epoch": 4.9375,
592
+ "grad_norm": 0.18475523591041565,
593
+ "learning_rate": 0.00234,
594
+ "loss": 1.3006361722946167,
595
+ "step": 79
596
+ },
597
+ {
598
+ "epoch": 5.0,
599
+ "grad_norm": 0.6583278775215149,
600
+ "learning_rate": 0.00237,
601
+ "loss": 1.2868517637252808,
602
+ "step": 80
603
+ },
604
+ {
605
+ "epoch": 5.0,
606
+ "eval_loss": 0.6618342399597168,
607
+ "eval_runtime": 30.1233,
608
+ "eval_samples_per_second": 3.386,
609
+ "eval_steps_per_second": 0.432,
610
+ "step": 80
611
+ },
612
+ {
613
+ "epoch": 5.0625,
614
+ "grad_norm": 0.04499061033129692,
615
+ "learning_rate": 0.0024000000000000002,
616
+ "loss": 1.25992751121521,
617
+ "step": 81
618
+ },
619
+ {
620
+ "epoch": 5.125,
621
+ "grad_norm": 0.05609860643744469,
622
+ "learning_rate": 0.0024300000000000003,
623
+ "loss": 1.2764487266540527,
624
+ "step": 82
625
+ },
626
+ {
627
+ "epoch": 5.1875,
628
+ "grad_norm": 0.03665148839354515,
629
+ "learning_rate": 0.00246,
630
+ "loss": 1.2908766269683838,
631
+ "step": 83
632
+ },
633
+ {
634
+ "epoch": 5.25,
635
+ "grad_norm": 0.17949481308460236,
636
+ "learning_rate": 0.00249,
637
+ "loss": 1.253080129623413,
638
+ "step": 84
639
+ },
640
+ {
641
+ "epoch": 5.3125,
642
+ "grad_norm": 0.14031393826007843,
643
+ "learning_rate": 0.00252,
644
+ "loss": 1.2648870944976807,
645
+ "step": 85
646
+ },
647
+ {
648
+ "epoch": 5.375,
649
+ "grad_norm": 0.08193210512399673,
650
+ "learning_rate": 0.00255,
651
+ "loss": 1.296234130859375,
652
+ "step": 86
653
+ },
654
+ {
655
+ "epoch": 5.4375,
656
+ "grad_norm": 0.056413356214761734,
657
+ "learning_rate": 0.00258,
658
+ "loss": 1.2784960269927979,
659
+ "step": 87
660
+ },
661
+ {
662
+ "epoch": 5.5,
663
+ "grad_norm": 0.09004318714141846,
664
+ "learning_rate": 0.00261,
665
+ "loss": 1.2054948806762695,
666
+ "step": 88
667
+ },
668
+ {
669
+ "epoch": 5.5625,
670
+ "grad_norm": 0.044756047427654266,
671
+ "learning_rate": 0.00264,
672
+ "loss": 1.2707666158676147,
673
+ "step": 89
674
+ },
675
+ {
676
+ "epoch": 5.625,
677
+ "grad_norm": 0.049932755529880524,
678
+ "learning_rate": 0.00267,
679
+ "loss": 1.2369787693023682,
680
+ "step": 90
681
+ },
682
+ {
683
+ "epoch": 5.6875,
684
+ "grad_norm": 0.09624820947647095,
685
+ "learning_rate": 0.0027,
686
+ "loss": 1.2419278621673584,
687
+ "step": 91
688
+ },
689
+ {
690
+ "epoch": 5.75,
691
+ "grad_norm": 0.17648059129714966,
692
+ "learning_rate": 0.0027300000000000002,
693
+ "loss": 1.2992033958435059,
694
+ "step": 92
695
+ },
696
+ {
697
+ "epoch": 5.8125,
698
+ "grad_norm": 0.07027823477983475,
699
+ "learning_rate": 0.0027600000000000003,
700
+ "loss": 1.2255799770355225,
701
+ "step": 93
702
+ },
703
+ {
704
+ "epoch": 5.875,
705
+ "grad_norm": 0.06040462478995323,
706
+ "learning_rate": 0.0027900000000000004,
707
+ "loss": 1.136023759841919,
708
+ "step": 94
709
+ },
710
+ {
711
+ "epoch": 5.9375,
712
+ "grad_norm": 0.09845872968435287,
713
+ "learning_rate": 0.00282,
714
+ "loss": 1.1807364225387573,
715
+ "step": 95
716
+ },
717
+ {
718
+ "epoch": 6.0,
719
+ "grad_norm": 0.09086572378873825,
720
+ "learning_rate": 0.00285,
721
+ "loss": 1.2171341180801392,
722
+ "step": 96
723
+ },
724
+ {
725
+ "epoch": 6.0,
726
+ "eval_loss": 0.6200892329216003,
727
+ "eval_runtime": 30.1252,
728
+ "eval_samples_per_second": 3.386,
729
+ "eval_steps_per_second": 0.432,
730
+ "step": 96
731
+ },
732
+ {
733
+ "epoch": 6.0625,
734
+ "grad_norm": 0.11787034571170807,
735
+ "learning_rate": 0.0028799999999999997,
736
+ "loss": 1.242072582244873,
737
+ "step": 97
738
+ },
739
+ {
740
+ "epoch": 6.125,
741
+ "grad_norm": 0.20093177258968353,
742
+ "learning_rate": 0.00291,
743
+ "loss": 1.2056899070739746,
744
+ "step": 98
745
+ },
746
+ {
747
+ "epoch": 6.1875,
748
+ "grad_norm": 0.03221331909298897,
749
+ "learning_rate": 0.00294,
750
+ "loss": 1.1774635314941406,
751
+ "step": 99
752
+ },
753
+ {
754
+ "epoch": 6.25,
755
+ "grad_norm": 0.06112038344144821,
756
+ "learning_rate": 0.00297,
757
+ "loss": 1.1831618547439575,
758
+ "step": 100
759
+ },
760
+ {
761
+ "epoch": 6.3125,
762
+ "grad_norm": 0.05896962434053421,
763
+ "learning_rate": 0.003,
764
+ "loss": 1.219610571861267,
765
+ "step": 101
766
+ },
767
+ {
768
+ "epoch": 6.375,
769
+ "grad_norm": 0.4068014323711395,
770
+ "learning_rate": 0.0029999850445531283,
771
+ "loss": 1.231442928314209,
772
+ "step": 102
773
+ },
774
+ {
775
+ "epoch": 6.4375,
776
+ "grad_norm": 3.013302803039551,
777
+ "learning_rate": 0.0029999401785137437,
778
+ "loss": 1.2512195110321045,
779
+ "step": 103
780
+ },
781
+ {
782
+ "epoch": 6.5,
783
+ "grad_norm": 0.07003403455018997,
784
+ "learning_rate": 0.0029998654027855406,
785
+ "loss": 1.1912157535552979,
786
+ "step": 104
787
+ },
788
+ {
789
+ "epoch": 6.5625,
790
+ "grad_norm": 1.1927955150604248,
791
+ "learning_rate": 0.002999760718874652,
792
+ "loss": 1.1158745288848877,
793
+ "step": 105
794
+ },
795
+ {
796
+ "epoch": 6.625,
797
+ "grad_norm": 0.11684458702802658,
798
+ "learning_rate": 0.0029996261288896225,
799
+ "loss": 1.1835001707077026,
800
+ "step": 106
801
+ },
802
+ {
803
+ "epoch": 6.6875,
804
+ "grad_norm": 0.07812155783176422,
805
+ "learning_rate": 0.0029994616355413666,
806
+ "loss": 1.1712734699249268,
807
+ "step": 107
808
+ },
809
+ {
810
+ "epoch": 6.75,
811
+ "grad_norm": 0.19128946959972382,
812
+ "learning_rate": 0.0029992672421431115,
813
+ "loss": 1.0906574726104736,
814
+ "step": 108
815
+ },
816
+ {
817
+ "epoch": 6.8125,
818
+ "grad_norm": 0.033625371754169464,
819
+ "learning_rate": 0.002999042952610332,
820
+ "loss": 1.115525245666504,
821
+ "step": 109
822
+ },
823
+ {
824
+ "epoch": 6.875,
825
+ "grad_norm": 0.2518763244152069,
826
+ "learning_rate": 0.0029987887714606723,
827
+ "loss": 1.1433348655700684,
828
+ "step": 110
829
+ },
830
+ {
831
+ "epoch": 6.9375,
832
+ "grad_norm": 0.037159573286771774,
833
+ "learning_rate": 0.002998504703813852,
834
+ "loss": 1.1867601871490479,
835
+ "step": 111
836
+ },
837
+ {
838
+ "epoch": 7.0,
839
+ "grad_norm": 0.04564391076564789,
840
+ "learning_rate": 0.0029981907553915674,
841
+ "loss": 1.1302828788757324,
842
+ "step": 112
843
+ },
844
+ {
845
+ "epoch": 7.0,
846
+ "eval_loss": 0.5897142887115479,
847
+ "eval_runtime": 31.3671,
848
+ "eval_samples_per_second": 3.252,
849
+ "eval_steps_per_second": 0.414,
850
+ "step": 112
851
+ },
852
+ {
853
+ "epoch": 7.0625,
854
+ "grad_norm": 0.05205775797367096,
855
+ "learning_rate": 0.0029978469325173717,
856
+ "loss": 1.182741641998291,
857
+ "step": 113
858
+ },
859
+ {
860
+ "epoch": 7.125,
861
+ "grad_norm": 0.12080052495002747,
862
+ "learning_rate": 0.002997473242116551,
863
+ "loss": 1.1457600593566895,
864
+ "step": 114
865
+ },
866
+ {
867
+ "epoch": 7.1875,
868
+ "grad_norm": 0.05600045621395111,
869
+ "learning_rate": 0.0029970696917159833,
870
+ "loss": 1.1101720333099365,
871
+ "step": 115
872
+ },
873
+ {
874
+ "epoch": 7.25,
875
+ "grad_norm": 0.0507950522005558,
876
+ "learning_rate": 0.0029966362894439874,
877
+ "loss": 1.1785833835601807,
878
+ "step": 116
879
+ },
880
+ {
881
+ "epoch": 7.3125,
882
+ "grad_norm": 0.6396550536155701,
883
+ "learning_rate": 0.002996173044030159,
884
+ "loss": 1.1538987159729004,
885
+ "step": 117
886
+ },
887
+ {
888
+ "epoch": 7.375,
889
+ "grad_norm": 0.06849977374076843,
890
+ "learning_rate": 0.0029956799648051934,
891
+ "loss": 1.1528857946395874,
892
+ "step": 118
893
+ },
894
+ {
895
+ "epoch": 7.4375,
896
+ "grad_norm": 0.0685896947979927,
897
+ "learning_rate": 0.002995157061700702,
898
+ "loss": 1.1049470901489258,
899
+ "step": 119
900
+ },
901
+ {
902
+ "epoch": 7.5,
903
+ "grad_norm": 1.027722716331482,
904
+ "learning_rate": 0.002994604345249006,
905
+ "loss": 1.1308050155639648,
906
+ "step": 120
907
+ },
908
+ {
909
+ "epoch": 7.5625,
910
+ "grad_norm": 0.21406109631061554,
911
+ "learning_rate": 0.00299402182658293,
912
+ "loss": 1.0960861444473267,
913
+ "step": 121
914
+ },
915
+ {
916
+ "epoch": 7.625,
917
+ "grad_norm": 0.0503656342625618,
918
+ "learning_rate": 0.002993409517435574,
919
+ "loss": 1.1096141338348389,
920
+ "step": 122
921
+ },
922
+ {
923
+ "epoch": 7.6875,
924
+ "grad_norm": 0.04800846800208092,
925
+ "learning_rate": 0.002992767430140078,
926
+ "loss": 1.1386680603027344,
927
+ "step": 123
928
+ },
929
+ {
930
+ "epoch": 7.75,
931
+ "grad_norm": 0.13091202080249786,
932
+ "learning_rate": 0.002992095577629376,
933
+ "loss": 1.1129441261291504,
934
+ "step": 124
935
+ },
936
+ {
937
+ "epoch": 7.8125,
938
+ "grad_norm": 0.04949796572327614,
939
+ "learning_rate": 0.0029913939734359306,
940
+ "loss": 1.106520175933838,
941
+ "step": 125
942
+ },
943
+ {
944
+ "epoch": 7.875,
945
+ "grad_norm": 0.07686582207679749,
946
+ "learning_rate": 0.0029906626316914655,
947
+ "loss": 1.1572234630584717,
948
+ "step": 126
949
+ },
950
+ {
951
+ "epoch": 7.9375,
952
+ "grad_norm": 0.053587451577186584,
953
+ "learning_rate": 0.0029899015671266773,
954
+ "loss": 1.125717282295227,
955
+ "step": 127
956
+ },
957
+ {
958
+ "epoch": 8.0,
959
+ "grad_norm": 0.08031050115823746,
960
+ "learning_rate": 0.0029891107950709406,
961
+ "loss": 1.153022050857544,
962
+ "step": 128
963
+ },
964
+ {
965
+ "epoch": 8.0,
966
+ "eval_loss": 0.5910838842391968,
967
+ "eval_runtime": 30.7073,
968
+ "eval_samples_per_second": 3.322,
969
+ "eval_steps_per_second": 0.423,
970
+ "step": 128
971
+ },
972
+ {
973
+ "epoch": 8.0625,
974
+ "grad_norm": 0.3520950675010681,
975
+ "learning_rate": 0.0029882903314519997,
976
+ "loss": 1.1582988500595093,
977
+ "step": 129
978
+ },
979
+ {
980
+ "epoch": 8.125,
981
+ "grad_norm": 0.07153584063053131,
982
+ "learning_rate": 0.002987440192795645,
983
+ "loss": 1.1344678401947021,
984
+ "step": 130
985
+ },
986
+ {
987
+ "epoch": 8.1875,
988
+ "grad_norm": 0.06336807459592819,
989
+ "learning_rate": 0.0029865603962253833,
990
+ "loss": 1.120319128036499,
991
+ "step": 131
992
+ },
993
+ {
994
+ "epoch": 8.25,
995
+ "grad_norm": 0.11782127618789673,
996
+ "learning_rate": 0.0029856509594620912,
997
+ "loss": 1.1245379447937012,
998
+ "step": 132
999
+ },
1000
+ {
1001
+ "epoch": 8.3125,
1002
+ "grad_norm": 0.2592264413833618,
1003
+ "learning_rate": 0.0029847119008236576,
1004
+ "loss": 1.1454271078109741,
1005
+ "step": 133
1006
+ },
1007
+ {
1008
+ "epoch": 8.375,
1009
+ "grad_norm": 0.21255695819854736,
1010
+ "learning_rate": 0.0029837432392246185,
1011
+ "loss": 1.231805443763733,
1012
+ "step": 134
1013
+ },
1014
+ {
1015
+ "epoch": 8.4375,
1016
+ "grad_norm": 15.057058334350586,
1017
+ "learning_rate": 0.0029827449941757698,
1018
+ "loss": 1.1380956172943115,
1019
+ "step": 135
1020
+ },
1021
+ {
1022
+ "epoch": 8.5,
1023
+ "grad_norm": 0.17857126891613007,
1024
+ "learning_rate": 0.00298171718578378,
1025
+ "loss": 1.1401441097259521,
1026
+ "step": 136
1027
+ },
1028
+ {
1029
+ "epoch": 8.5625,
1030
+ "grad_norm": 0.18281924724578857,
1031
+ "learning_rate": 0.002980659834750781,
1032
+ "loss": 1.171494722366333,
1033
+ "step": 137
1034
+ },
1035
+ {
1036
+ "epoch": 8.625,
1037
+ "grad_norm": 0.10490825027227402,
1038
+ "learning_rate": 0.0029795729623739555,
1039
+ "loss": 1.1622605323791504,
1040
+ "step": 138
1041
+ },
1042
+ {
1043
+ "epoch": 8.6875,
1044
+ "grad_norm": 0.344013512134552,
1045
+ "learning_rate": 0.002978456590545104,
1046
+ "loss": 1.1702141761779785,
1047
+ "step": 139
1048
+ },
1049
+ {
1050
+ "epoch": 8.75,
1051
+ "grad_norm": 0.05058368295431137,
1052
+ "learning_rate": 0.0029773107417502055,
1053
+ "loss": 1.2006211280822754,
1054
+ "step": 140
1055
+ },
1056
+ {
1057
+ "epoch": 8.8125,
1058
+ "grad_norm": 0.09775760769844055,
1059
+ "learning_rate": 0.002976135439068965,
1060
+ "loss": 1.1407063007354736,
1061
+ "step": 141
1062
+ },
1063
+ {
1064
+ "epoch": 8.875,
1065
+ "grad_norm": 0.0785030648112297,
1066
+ "learning_rate": 0.0029749307061743475,
1067
+ "loss": 1.1778209209442139,
1068
+ "step": 142
1069
+ },
1070
+ {
1071
+ "epoch": 8.9375,
1072
+ "grad_norm": 0.053056441247463226,
1073
+ "learning_rate": 0.0029736965673321032,
1074
+ "loss": 1.1302082538604736,
1075
+ "step": 143
1076
+ },
1077
+ {
1078
+ "epoch": 9.0,
1079
+ "grad_norm": 0.4020994305610657,
1080
+ "learning_rate": 0.0029724330474002743,
1081
+ "loss": 1.1655206680297852,
1082
+ "step": 144
1083
+ },
1084
+ {
1085
+ "epoch": 9.0,
1086
+ "eval_loss": 0.6083483099937439,
1087
+ "eval_runtime": 30.0453,
1088
+ "eval_samples_per_second": 3.395,
1089
+ "eval_steps_per_second": 0.433,
1090
+ "step": 144
1091
+ },
1092
+ {
1093
+ "epoch": 9.0625,
1094
+ "grad_norm": 0.04755420982837677,
1095
+ "learning_rate": 0.002971140171828702,
1096
+ "loss": 1.1900575160980225,
1097
+ "step": 145
1098
+ },
1099
+ {
1100
+ "epoch": 9.125,
1101
+ "grad_norm": 0.2549609839916229,
1102
+ "learning_rate": 0.002969817966658505,
1103
+ "loss": 1.1891400814056396,
1104
+ "step": 146
1105
+ },
1106
+ {
1107
+ "epoch": 9.1875,
1108
+ "grad_norm": 0.11641769111156464,
1109
+ "learning_rate": 0.0029684664585215618,
1110
+ "loss": 1.1930484771728516,
1111
+ "step": 147
1112
+ },
1113
+ {
1114
+ "epoch": 9.25,
1115
+ "grad_norm": 0.13052672147750854,
1116
+ "learning_rate": 0.00296708567463997,
1117
+ "loss": 1.1905772686004639,
1118
+ "step": 148
1119
+ },
1120
+ {
1121
+ "epoch": 9.3125,
1122
+ "grad_norm": 0.15126636624336243,
1123
+ "learning_rate": 0.0029656756428255017,
1124
+ "loss": 1.1417806148529053,
1125
+ "step": 149
1126
+ },
1127
+ {
1128
+ "epoch": 9.375,
1129
+ "grad_norm": 0.08826905488967896,
1130
+ "learning_rate": 0.00296423639147904,
1131
+ "loss": 1.1802985668182373,
1132
+ "step": 150
1133
+ },
1134
+ {
1135
+ "epoch": 9.4375,
1136
+ "grad_norm": 0.01381592359393835,
1137
+ "learning_rate": 0.002962767949590008,
1138
+ "loss": 1.1525968313217163,
1139
+ "step": 151
1140
+ },
1141
+ {
1142
+ "epoch": 9.5,
1143
+ "grad_norm": 0.04648139327764511,
1144
+ "learning_rate": 0.0029612703467357866,
1145
+ "loss": 1.169856309890747,
1146
+ "step": 152
1147
+ },
1148
+ {
1149
+ "epoch": 9.5625,
1150
+ "grad_norm": 0.014019235037267208,
1151
+ "learning_rate": 0.0029597436130811153,
1152
+ "loss": 1.1442453861236572,
1153
+ "step": 153
1154
+ },
1155
+ {
1156
+ "epoch": 9.625,
1157
+ "grad_norm": 0.03701096400618553,
1158
+ "learning_rate": 0.0029581877793774885,
1159
+ "loss": 1.112680435180664,
1160
+ "step": 154
1161
+ },
1162
+ {
1163
+ "epoch": 9.6875,
1164
+ "grad_norm": 0.030421797186136246,
1165
+ "learning_rate": 0.0029566028769625334,
1166
+ "loss": 1.223894715309143,
1167
+ "step": 155
1168
+ },
1169
+ {
1170
+ "epoch": 9.75,
1171
+ "grad_norm": 0.12336083501577377,
1172
+ "learning_rate": 0.00295498893775938,
1173
+ "loss": 1.220346212387085,
1174
+ "step": 156
1175
+ },
1176
+ {
1177
+ "epoch": 9.8125,
1178
+ "grad_norm": 0.038585737347602844,
1179
+ "learning_rate": 0.0029533459942760174,
1180
+ "loss": 1.2078583240509033,
1181
+ "step": 157
1182
+ },
1183
+ {
1184
+ "epoch": 9.875,
1185
+ "grad_norm": 0.4653286039829254,
1186
+ "learning_rate": 0.002951674079604639,
1187
+ "loss": 1.2035815715789795,
1188
+ "step": 158
1189
+ },
1190
+ {
1191
+ "epoch": 9.9375,
1192
+ "grad_norm": 0.020640619099140167,
1193
+ "learning_rate": 0.0029499732274209777,
1194
+ "loss": 1.1313148736953735,
1195
+ "step": 159
1196
+ },
1197
+ {
1198
+ "epoch": 10.0,
1199
+ "grad_norm": 0.04939777031540871,
1200
+ "learning_rate": 0.002948243471983625,
1201
+ "loss": 1.1253995895385742,
1202
+ "step": 160
1203
+ },
1204
+ {
1205
+ "epoch": 10.0,
1206
+ "eval_loss": 0.604421079158783,
1207
+ "eval_runtime": 30.1556,
1208
+ "eval_samples_per_second": 3.382,
1209
+ "eval_steps_per_second": 0.431,
1210
+ "step": 160
1211
+ },
1212
+ {
1213
+ "epoch": 10.0625,
1214
+ "grad_norm": 0.05968557298183441,
1215
+ "learning_rate": 0.0029464848481333423,
1216
+ "loss": 1.1324307918548584,
1217
+ "step": 161
1218
+ },
1219
+ {
1220
+ "epoch": 10.125,
1221
+ "grad_norm": 1.0001988410949707,
1222
+ "learning_rate": 0.0029446973912923597,
1223
+ "loss": 1.126110315322876,
1224
+ "step": 162
1225
+ },
1226
+ {
1227
+ "epoch": 10.1875,
1228
+ "grad_norm": 0.2294934093952179,
1229
+ "learning_rate": 0.002942881137463661,
1230
+ "loss": 1.1873570680618286,
1231
+ "step": 163
1232
+ },
1233
+ {
1234
+ "epoch": 10.25,
1235
+ "grad_norm": 0.06396346539258957,
1236
+ "learning_rate": 0.002941036123230261,
1237
+ "loss": 1.1288526058197021,
1238
+ "step": 164
1239
+ },
1240
+ {
1241
+ "epoch": 10.3125,
1242
+ "grad_norm": 0.3410445749759674,
1243
+ "learning_rate": 0.0029391623857544643,
1244
+ "loss": 1.1219967603683472,
1245
+ "step": 165
1246
+ },
1247
+ {
1248
+ "epoch": 10.375,
1249
+ "grad_norm": 0.09721948206424713,
1250
+ "learning_rate": 0.002937259962777122,
1251
+ "loss": 1.141520619392395,
1252
+ "step": 166
1253
+ },
1254
+ {
1255
+ "epoch": 10.4375,
1256
+ "grad_norm": 0.033714450895786285,
1257
+ "learning_rate": 0.002935328892616868,
1258
+ "loss": 1.1983453035354614,
1259
+ "step": 167
1260
+ },
1261
+ {
1262
+ "epoch": 10.5,
1263
+ "grad_norm": 0.028446367010474205,
1264
+ "learning_rate": 0.0029333692141693484,
1265
+ "loss": 1.153167724609375,
1266
+ "step": 168
1267
+ },
1268
+ {
1269
+ "epoch": 10.5625,
1270
+ "grad_norm": 0.04298163205385208,
1271
+ "learning_rate": 0.0029313809669064377,
1272
+ "loss": 1.120814561843872,
1273
+ "step": 169
1274
+ },
1275
+ {
1276
+ "epoch": 10.625,
1277
+ "grad_norm": 0.08190953731536865,
1278
+ "learning_rate": 0.002929364190875444,
1279
+ "loss": 1.1164218187332153,
1280
+ "step": 170
1281
+ },
1282
+ {
1283
+ "epoch": 10.6875,
1284
+ "grad_norm": 0.0602508969604969,
1285
+ "learning_rate": 0.0029273189266983035,
1286
+ "loss": 1.136858344078064,
1287
+ "step": 171
1288
+ },
1289
+ {
1290
+ "epoch": 10.75,
1291
+ "grad_norm": 0.07118846476078033,
1292
+ "learning_rate": 0.0029252452155707582,
1293
+ "loss": 1.1356765031814575,
1294
+ "step": 172
1295
+ },
1296
+ {
1297
+ "epoch": 10.8125,
1298
+ "grad_norm": 0.0445970743894577,
1299
+ "learning_rate": 0.002923143099261531,
1300
+ "loss": 1.121457815170288,
1301
+ "step": 173
1302
+ },
1303
+ {
1304
+ "epoch": 10.875,
1305
+ "grad_norm": 0.07745611667633057,
1306
+ "learning_rate": 0.0029210126201114836,
1307
+ "loss": 1.0903944969177246,
1308
+ "step": 174
1309
+ },
1310
+ {
1311
+ "epoch": 10.9375,
1312
+ "grad_norm": 0.08945920318365097,
1313
+ "learning_rate": 0.0029188538210327595,
1314
+ "loss": 1.1194133758544922,
1315
+ "step": 175
1316
+ },
1317
+ {
1318
+ "epoch": 11.0,
1319
+ "grad_norm": 0.07996643334627151,
1320
+ "learning_rate": 0.002916666745507926,
1321
+ "loss": 1.1248202323913574,
1322
+ "step": 176
1323
+ },
1324
+ {
1325
+ "epoch": 11.0,
1326
+ "eval_loss": 0.5789638757705688,
1327
+ "eval_runtime": 30.3538,
1328
+ "eval_samples_per_second": 3.36,
1329
+ "eval_steps_per_second": 0.428,
1330
+ "step": 176
1331
+ },
1332
+ {
1333
+ "epoch": 11.0625,
1334
+ "grad_norm": 0.10533642023801804,
1335
+ "learning_rate": 0.002914451437589092,
1336
+ "loss": 1.1199195384979248,
1337
+ "step": 177
1338
+ },
1339
+ {
1340
+ "epoch": 11.125,
1341
+ "grad_norm": 0.044989872723817825,
1342
+ "learning_rate": 0.002912207941897025,
1343
+ "loss": 1.1084158420562744,
1344
+ "step": 178
1345
+ },
1346
+ {
1347
+ "epoch": 11.1875,
1348
+ "grad_norm": 0.04824615642428398,
1349
+ "learning_rate": 0.0029099363036202515,
1350
+ "loss": 1.0993731021881104,
1351
+ "step": 179
1352
+ },
1353
+ {
1354
+ "epoch": 11.25,
1355
+ "grad_norm": 0.03144732117652893,
1356
+ "learning_rate": 0.0029076365685141463,
1357
+ "loss": 1.1078145503997803,
1358
+ "step": 180
1359
+ },
1360
+ {
1361
+ "epoch": 11.3125,
1362
+ "grad_norm": 0.03728745877742767,
1363
+ "learning_rate": 0.0029053087829000097,
1364
+ "loss": 1.0995632410049438,
1365
+ "step": 181
1366
+ },
1367
+ {
1368
+ "epoch": 11.375,
1369
+ "grad_norm": 0.02090257592499256,
1370
+ "learning_rate": 0.002902952993664138,
1371
+ "loss": 1.0722652673721313,
1372
+ "step": 182
1373
+ },
1374
+ {
1375
+ "epoch": 11.4375,
1376
+ "grad_norm": 0.020907679572701454,
1377
+ "learning_rate": 0.0029005692482568747,
1378
+ "loss": 1.0840482711791992,
1379
+ "step": 183
1380
+ },
1381
+ {
1382
+ "epoch": 11.5,
1383
+ "grad_norm": 0.023914694786071777,
1384
+ "learning_rate": 0.002898157594691658,
1385
+ "loss": 1.1289193630218506,
1386
+ "step": 184
1387
+ },
1388
+ {
1389
+ "epoch": 11.5625,
1390
+ "grad_norm": 0.01937722973525524,
1391
+ "learning_rate": 0.0028957180815440534,
1392
+ "loss": 1.1244895458221436,
1393
+ "step": 185
1394
+ },
1395
+ {
1396
+ "epoch": 11.625,
1397
+ "grad_norm": 0.02047949656844139,
1398
+ "learning_rate": 0.002893250757950773,
1399
+ "loss": 1.0527830123901367,
1400
+ "step": 186
1401
+ },
1402
+ {
1403
+ "epoch": 11.6875,
1404
+ "grad_norm": 0.04898907244205475,
1405
+ "learning_rate": 0.002890755673608688,
1406
+ "loss": 1.1234995126724243,
1407
+ "step": 187
1408
+ },
1409
+ {
1410
+ "epoch": 11.75,
1411
+ "grad_norm": 0.024841446429491043,
1412
+ "learning_rate": 0.0028882328787738273,
1413
+ "loss": 1.138411045074463,
1414
+ "step": 188
1415
+ },
1416
+ {
1417
+ "epoch": 11.8125,
1418
+ "grad_norm": 0.02533135376870632,
1419
+ "learning_rate": 0.002885682424260365,
1420
+ "loss": 1.1021623611450195,
1421
+ "step": 189
1422
+ },
1423
+ {
1424
+ "epoch": 11.875,
1425
+ "grad_norm": 0.04309404641389847,
1426
+ "learning_rate": 0.0028831043614395966,
1427
+ "loss": 1.0767666101455688,
1428
+ "step": 190
1429
+ },
1430
+ {
1431
+ "epoch": 11.9375,
1432
+ "grad_norm": 0.029682395979762077,
1433
+ "learning_rate": 0.0028804987422389045,
1434
+ "loss": 1.1310434341430664,
1435
+ "step": 191
1436
+ },
1437
+ {
1438
+ "epoch": 12.0,
1439
+ "grad_norm": 0.021236268803477287,
1440
+ "learning_rate": 0.002877865619140712,
1441
+ "loss": 1.0969300270080566,
1442
+ "step": 192
1443
+ },
1444
+ {
1445
+ "epoch": 12.0,
1446
+ "eval_loss": 0.5664089322090149,
1447
+ "eval_runtime": 30.1046,
1448
+ "eval_samples_per_second": 3.388,
1449
+ "eval_steps_per_second": 0.432,
1450
+ "step": 192
1451
+ },
1452
+ {
1453
+ "epoch": 12.0625,
1454
+ "grad_norm": 0.09511730819940567,
1455
+ "learning_rate": 0.0028752050451814276,
1456
+ "loss": 1.0866225957870483,
1457
+ "step": 193
1458
+ },
1459
+ {
1460
+ "epoch": 12.125,
1461
+ "grad_norm": 0.1858353465795517,
1462
+ "learning_rate": 0.0028725170739503734,
1463
+ "loss": 1.1082183122634888,
1464
+ "step": 194
1465
+ },
1466
+ {
1467
+ "epoch": 12.1875,
1468
+ "grad_norm": 0.018049633130431175,
1469
+ "learning_rate": 0.00286980175958871,
1470
+ "loss": 1.0569005012512207,
1471
+ "step": 195
1472
+ },
1473
+ {
1474
+ "epoch": 12.25,
1475
+ "grad_norm": 0.06167157366871834,
1476
+ "learning_rate": 0.002867059156788342,
1477
+ "loss": 1.1418097019195557,
1478
+ "step": 196
1479
+ },
1480
+ {
1481
+ "epoch": 12.3125,
1482
+ "grad_norm": 0.06766022741794586,
1483
+ "learning_rate": 0.0028642893207908187,
1484
+ "loss": 1.0995159149169922,
1485
+ "step": 197
1486
+ },
1487
+ {
1488
+ "epoch": 12.375,
1489
+ "grad_norm": 0.03883042931556702,
1490
+ "learning_rate": 0.0028614923073862213,
1491
+ "loss": 1.0847246646881104,
1492
+ "step": 198
1493
+ },
1494
+ {
1495
+ "epoch": 12.4375,
1496
+ "grad_norm": 0.05537786707282066,
1497
+ "learning_rate": 0.0028586681729120393,
1498
+ "loss": 1.1187801361083984,
1499
+ "step": 199
1500
+ },
1501
+ {
1502
+ "epoch": 12.5,
1503
+ "grad_norm": 0.0653286948800087,
1504
+ "learning_rate": 0.002855816974252033,
1505
+ "loss": 1.0798084735870361,
1506
+ "step": 200
1507
+ },
1508
+ {
1509
+ "epoch": 12.5625,
1510
+ "grad_norm": 0.04398065432906151,
1511
+ "learning_rate": 0.002852938768835092,
1512
+ "loss": 1.0946357250213623,
1513
+ "step": 201
1514
+ },
1515
+ {
1516
+ "epoch": 12.625,
1517
+ "grad_norm": 0.05614284798502922,
1518
+ "learning_rate": 0.0028500336146340772,
1519
+ "loss": 1.1144163608551025,
1520
+ "step": 202
1521
+ },
1522
+ {
1523
+ "epoch": 12.6875,
1524
+ "grad_norm": 0.035650063306093216,
1525
+ "learning_rate": 0.0028471015701646485,
1526
+ "loss": 1.0421898365020752,
1527
+ "step": 203
1528
+ },
1529
+ {
1530
+ "epoch": 12.75,
1531
+ "grad_norm": 0.02544066496193409,
1532
+ "learning_rate": 0.002844142694484094,
1533
+ "loss": 1.0994564294815063,
1534
+ "step": 204
1535
+ },
1536
+ {
1537
+ "epoch": 12.8125,
1538
+ "grad_norm": 0.07384507358074188,
1539
+ "learning_rate": 0.002841157047190135,
1540
+ "loss": 1.0665926933288574,
1541
+ "step": 205
1542
+ },
1543
+ {
1544
+ "epoch": 12.875,
1545
+ "grad_norm": 0.04875814542174339,
1546
+ "learning_rate": 0.002838144688419726,
1547
+ "loss": 1.0665624141693115,
1548
+ "step": 206
1549
+ },
1550
+ {
1551
+ "epoch": 12.9375,
1552
+ "grad_norm": 0.05739036574959755,
1553
+ "learning_rate": 0.002835105678847847,
1554
+ "loss": 1.1085939407348633,
1555
+ "step": 207
1556
+ },
1557
+ {
1558
+ "epoch": 13.0,
1559
+ "grad_norm": 0.03088603913784027,
1560
+ "learning_rate": 0.002832040079686276,
1561
+ "loss": 1.1357779502868652,
1562
+ "step": 208
1563
+ },
1564
+ {
1565
+ "epoch": 13.0,
1566
+ "eval_loss": 0.5675534009933472,
1567
+ "eval_runtime": 31.0754,
1568
+ "eval_samples_per_second": 3.282,
1569
+ "eval_steps_per_second": 0.418,
1570
+ "step": 208
1571
+ },
1572
+ {
1573
+ "epoch": 13.0625,
1574
+ "grad_norm": 0.052636317908763885,
1575
+ "learning_rate": 0.0028289479526823614,
1576
+ "loss": 1.0684641599655151,
1577
+ "step": 209
1578
+ },
1579
+ {
1580
+ "epoch": 13.125,
1581
+ "grad_norm": 0.26156774163246155,
1582
+ "learning_rate": 0.0028258293601177743,
1583
+ "loss": 1.1112414598464966,
1584
+ "step": 210
1585
+ },
1586
+ {
1587
+ "epoch": 13.1875,
1588
+ "grad_norm": 0.013046488165855408,
1589
+ "learning_rate": 0.0028226843648072562,
1590
+ "loss": 1.0339481830596924,
1591
+ "step": 211
1592
+ },
1593
+ {
1594
+ "epoch": 13.25,
1595
+ "grad_norm": 0.029812589287757874,
1596
+ "learning_rate": 0.002819513030097352,
1597
+ "loss": 1.0734567642211914,
1598
+ "step": 212
1599
+ },
1600
+ {
1601
+ "epoch": 13.3125,
1602
+ "grad_norm": 0.03711036592721939,
1603
+ "learning_rate": 0.0028163154198651372,
1604
+ "loss": 1.071413516998291,
1605
+ "step": 213
1606
+ },
1607
+ {
1608
+ "epoch": 13.375,
1609
+ "grad_norm": 0.09543009102344513,
1610
+ "learning_rate": 0.002813091598516928,
1611
+ "loss": 1.1175503730773926,
1612
+ "step": 214
1613
+ },
1614
+ {
1615
+ "epoch": 13.4375,
1616
+ "grad_norm": 0.02742145210504532,
1617
+ "learning_rate": 0.0028098416309869857,
1618
+ "loss": 1.0401530265808105,
1619
+ "step": 215
1620
+ },
1621
+ {
1622
+ "epoch": 13.5,
1623
+ "grad_norm": 0.05347360298037529,
1624
+ "learning_rate": 0.002806565582736208,
1625
+ "loss": 1.053802728652954,
1626
+ "step": 216
1627
+ },
1628
+ {
1629
+ "epoch": 13.5625,
1630
+ "grad_norm": 0.023394430056214333,
1631
+ "learning_rate": 0.002803263519750811,
1632
+ "loss": 1.0774943828582764,
1633
+ "step": 217
1634
+ },
1635
+ {
1636
+ "epoch": 13.625,
1637
+ "grad_norm": 0.03274613991379738,
1638
+ "learning_rate": 0.0027999355085409996,
1639
+ "loss": 1.1332664489746094,
1640
+ "step": 218
1641
+ },
1642
+ {
1643
+ "epoch": 13.6875,
1644
+ "grad_norm": 0.029956744983792305,
1645
+ "learning_rate": 0.00279658161613963,
1646
+ "loss": 1.1192328929901123,
1647
+ "step": 219
1648
+ },
1649
+ {
1650
+ "epoch": 13.75,
1651
+ "grad_norm": 0.027843108400702477,
1652
+ "learning_rate": 0.002793201910100856,
1653
+ "loss": 1.0953505039215088,
1654
+ "step": 220
1655
+ },
1656
+ {
1657
+ "epoch": 13.8125,
1658
+ "grad_norm": 0.3722478151321411,
1659
+ "learning_rate": 0.0027897964584987713,
1660
+ "loss": 1.1045243740081787,
1661
+ "step": 221
1662
+ },
1663
+ {
1664
+ "epoch": 13.875,
1665
+ "grad_norm": 0.05480271577835083,
1666
+ "learning_rate": 0.0027863653299260356,
1667
+ "loss": 1.0857124328613281,
1668
+ "step": 222
1669
+ },
1670
+ {
1671
+ "epoch": 13.9375,
1672
+ "grad_norm": 0.05806222930550575,
1673
+ "learning_rate": 0.0027829085934924958,
1674
+ "loss": 1.092329502105713,
1675
+ "step": 223
1676
+ },
1677
+ {
1678
+ "epoch": 14.0,
1679
+ "grad_norm": 0.07507319748401642,
1680
+ "learning_rate": 0.0027794263188237922,
1681
+ "loss": 1.0946941375732422,
1682
+ "step": 224
1683
+ },
1684
+ {
1685
+ "epoch": 14.0,
1686
+ "eval_loss": 0.5624361038208008,
1687
+ "eval_runtime": 30.3035,
1688
+ "eval_samples_per_second": 3.366,
1689
+ "eval_steps_per_second": 0.429,
1690
+ "step": 224
1691
+ },
1692
+ {
1693
+ "epoch": 14.0625,
1694
+ "grad_norm": 0.08125249296426773,
1695
+ "learning_rate": 0.0027759185760599576,
1696
+ "loss": 1.082047939300537,
1697
+ "step": 225
1698
+ },
1699
+ {
1700
+ "epoch": 14.125,
1701
+ "grad_norm": 0.08838590234518051,
1702
+ "learning_rate": 0.002772385435854002,
1703
+ "loss": 1.0747947692871094,
1704
+ "step": 226
1705
+ },
1706
+ {
1707
+ "epoch": 14.1875,
1708
+ "grad_norm": 0.18887385725975037,
1709
+ "learning_rate": 0.0027688269693704926,
1710
+ "loss": 1.0601060390472412,
1711
+ "step": 227
1712
+ },
1713
+ {
1714
+ "epoch": 14.25,
1715
+ "grad_norm": 0.4315606653690338,
1716
+ "learning_rate": 0.002765243248284118,
1717
+ "loss": 1.167302131652832,
1718
+ "step": 228
1719
+ },
1720
+ {
1721
+ "epoch": 14.3125,
1722
+ "grad_norm": 0.07665648311376572,
1723
+ "learning_rate": 0.002761634344778245,
1724
+ "loss": 1.130738615989685,
1725
+ "step": 229
1726
+ },
1727
+ {
1728
+ "epoch": 14.375,
1729
+ "grad_norm": 0.051125217229127884,
1730
+ "learning_rate": 0.002758000331543466,
1731
+ "loss": 1.0479257106781006,
1732
+ "step": 230
1733
+ },
1734
+ {
1735
+ "epoch": 14.4375,
1736
+ "grad_norm": 0.0658797100186348,
1737
+ "learning_rate": 0.0027543412817761345,
1738
+ "loss": 1.0913889408111572,
1739
+ "step": 231
1740
+ },
1741
+ {
1742
+ "epoch": 14.5,
1743
+ "grad_norm": 0.1936631202697754,
1744
+ "learning_rate": 0.0027506572691768888,
1745
+ "loss": 1.0582077503204346,
1746
+ "step": 232
1747
+ },
1748
+ {
1749
+ "epoch": 14.5625,
1750
+ "grad_norm": 2.202739953994751,
1751
+ "learning_rate": 0.0027469483679491693,
1752
+ "loss": 1.0946757793426514,
1753
+ "step": 233
1754
+ },
1755
+ {
1756
+ "epoch": 14.625,
1757
+ "grad_norm": 0.07856857776641846,
1758
+ "learning_rate": 0.002743214652797724,
1759
+ "loss": 1.074622631072998,
1760
+ "step": 234
1761
+ },
1762
+ {
1763
+ "epoch": 14.6875,
1764
+ "grad_norm": 0.04288165271282196,
1765
+ "learning_rate": 0.002739456198927104,
1766
+ "loss": 1.0788205862045288,
1767
+ "step": 235
1768
+ },
1769
+ {
1770
+ "epoch": 14.75,
1771
+ "grad_norm": 0.08169898390769958,
1772
+ "learning_rate": 0.0027356730820401465,
1773
+ "loss": 1.0906422138214111,
1774
+ "step": 236
1775
+ },
1776
+ {
1777
+ "epoch": 14.8125,
1778
+ "grad_norm": 0.04536404460668564,
1779
+ "learning_rate": 0.002731865378336452,
1780
+ "loss": 1.0542757511138916,
1781
+ "step": 237
1782
+ },
1783
+ {
1784
+ "epoch": 14.875,
1785
+ "grad_norm": 0.0246584415435791,
1786
+ "learning_rate": 0.0027280331645108498,
1787
+ "loss": 1.0401986837387085,
1788
+ "step": 238
1789
+ },
1790
+ {
1791
+ "epoch": 14.9375,
1792
+ "grad_norm": 0.03399847820401192,
1793
+ "learning_rate": 0.0027241765177518523,
1794
+ "loss": 1.109243392944336,
1795
+ "step": 239
1796
+ },
1797
+ {
1798
+ "epoch": 15.0,
1799
+ "grad_norm": 0.12445052713155746,
1800
+ "learning_rate": 0.0027202955157401,
1801
+ "loss": 1.0952987670898438,
1802
+ "step": 240
1803
+ },
1804
+ {
1805
+ "epoch": 15.0,
1806
+ "eval_loss": 0.5670443177223206,
1807
+ "eval_runtime": 30.7619,
1808
+ "eval_samples_per_second": 3.316,
1809
+ "eval_steps_per_second": 0.423,
1810
+ "step": 240
1811
+ },
1812
+ {
1813
+ "epoch": 15.0625,
1814
+ "grad_norm": 0.029460057616233826,
1815
+ "learning_rate": 0.002716390236646797,
1816
+ "loss": 1.0749149322509766,
1817
+ "step": 241
1818
+ },
1819
+ {
1820
+ "epoch": 15.125,
1821
+ "grad_norm": 0.09484875947237015,
1822
+ "learning_rate": 0.0027124607591321377,
1823
+ "loss": 1.1196999549865723,
1824
+ "step": 242
1825
+ },
1826
+ {
1827
+ "epoch": 15.1875,
1828
+ "grad_norm": 0.46802467107772827,
1829
+ "learning_rate": 0.002708507162343721,
1830
+ "loss": 1.0824053287506104,
1831
+ "step": 243
1832
+ },
1833
+ {
1834
+ "epoch": 15.25,
1835
+ "grad_norm": 0.02694086730480194,
1836
+ "learning_rate": 0.002704529525914957,
1837
+ "loss": 1.015854835510254,
1838
+ "step": 244
1839
+ },
1840
+ {
1841
+ "epoch": 15.3125,
1842
+ "grad_norm": 0.0534524992108345,
1843
+ "learning_rate": 0.0027005279299634616,
1844
+ "loss": 1.1357049942016602,
1845
+ "step": 245
1846
+ },
1847
+ {
1848
+ "epoch": 15.375,
1849
+ "grad_norm": 0.17367368936538696,
1850
+ "learning_rate": 0.002696502455089445,
1851
+ "loss": 1.0634117126464844,
1852
+ "step": 246
1853
+ },
1854
+ {
1855
+ "epoch": 15.4375,
1856
+ "grad_norm": 0.044344011694192886,
1857
+ "learning_rate": 0.002692453182374086,
1858
+ "loss": 1.1020934581756592,
1859
+ "step": 247
1860
+ },
1861
+ {
1862
+ "epoch": 15.5,
1863
+ "grad_norm": 0.2549358606338501,
1864
+ "learning_rate": 0.0026883801933779003,
1865
+ "loss": 1.0967700481414795,
1866
+ "step": 248
1867
+ },
1868
+ {
1869
+ "epoch": 15.5625,
1870
+ "grad_norm": 0.09740083664655685,
1871
+ "learning_rate": 0.0026842835701390982,
1872
+ "loss": 1.0662806034088135,
1873
+ "step": 249
1874
+ },
1875
+ {
1876
+ "epoch": 15.625,
1877
+ "grad_norm": 0.11962032318115234,
1878
+ "learning_rate": 0.0026801633951719288,
1879
+ "loss": 1.0632579326629639,
1880
+ "step": 250
1881
+ },
1882
+ {
1883
+ "epoch": 15.6875,
1884
+ "grad_norm": 0.11223616451025009,
1885
+ "learning_rate": 0.0026760197514650247,
1886
+ "loss": 1.089836835861206,
1887
+ "step": 251
1888
+ },
1889
+ {
1890
+ "epoch": 15.75,
1891
+ "grad_norm": 0.2540510892868042,
1892
+ "learning_rate": 0.002671852722479721,
1893
+ "loss": 1.088876485824585,
1894
+ "step": 252
1895
+ },
1896
+ {
1897
+ "epoch": 15.8125,
1898
+ "grad_norm": 0.08267078548669815,
1899
+ "learning_rate": 0.0026676623921483844,
1900
+ "loss": 1.077115535736084,
1901
+ "step": 253
1902
+ },
1903
+ {
1904
+ "epoch": 15.875,
1905
+ "grad_norm": 0.4830060601234436,
1906
+ "learning_rate": 0.0026634488448727144,
1907
+ "loss": 1.067319393157959,
1908
+ "step": 254
1909
+ },
1910
+ {
1911
+ "epoch": 15.9375,
1912
+ "grad_norm": 0.06782206147909164,
1913
+ "learning_rate": 0.0026592121655220472,
1914
+ "loss": 1.105837106704712,
1915
+ "step": 255
1916
+ },
1917
+ {
1918
+ "epoch": 16.0,
1919
+ "grad_norm": 0.0953136682510376,
1920
+ "learning_rate": 0.0026549524394316464,
1921
+ "loss": 1.1199421882629395,
1922
+ "step": 256
1923
+ },
1924
+ {
1925
+ "epoch": 16.0,
1926
+ "eval_loss": 0.5653528571128845,
1927
+ "eval_runtime": 30.0549,
1928
+ "eval_samples_per_second": 3.394,
1929
+ "eval_steps_per_second": 0.433,
1930
+ "step": 256
1931
+ },
1932
+ {
1933
+ "epoch": 16.0625,
1934
+ "grad_norm": 9.851863861083984,
1935
+ "learning_rate": 0.0026506697524009836,
1936
+ "loss": 1.1010339260101318,
1937
+ "step": 257
1938
+ },
1939
+ {
1940
+ "epoch": 16.125,
1941
+ "grad_norm": 0.10633312165737152,
1942
+ "learning_rate": 0.002646364190692009,
1943
+ "loss": 1.059409737586975,
1944
+ "step": 258
1945
+ },
1946
+ {
1947
+ "epoch": 16.1875,
1948
+ "grad_norm": 0.155309796333313,
1949
+ "learning_rate": 0.0026420358410274153,
1950
+ "loss": 1.0807991027832031,
1951
+ "step": 259
1952
+ },
1953
+ {
1954
+ "epoch": 16.25,
1955
+ "grad_norm": 0.05278189107775688,
1956
+ "learning_rate": 0.0026376847905888903,
1957
+ "loss": 1.0858765840530396,
1958
+ "step": 260
1959
+ },
1960
+ {
1961
+ "epoch": 16.3125,
1962
+ "grad_norm": 0.05539517104625702,
1963
+ "learning_rate": 0.0026333111270153624,
1964
+ "loss": 1.0652291774749756,
1965
+ "step": 261
1966
+ },
1967
+ {
1968
+ "epoch": 16.375,
1969
+ "grad_norm": 0.5698870420455933,
1970
+ "learning_rate": 0.0026289149384012326,
1971
+ "loss": 1.1399967670440674,
1972
+ "step": 262
1973
+ },
1974
+ {
1975
+ "epoch": 16.4375,
1976
+ "grad_norm": 0.0724017322063446,
1977
+ "learning_rate": 0.002624496313294602,
1978
+ "loss": 1.1068029403686523,
1979
+ "step": 263
1980
+ },
1981
+ {
1982
+ "epoch": 16.5,
1983
+ "grad_norm": 0.08893640339374542,
1984
+ "learning_rate": 0.0026200553406954892,
1985
+ "loss": 1.1509923934936523,
1986
+ "step": 264
1987
+ },
1988
+ {
1989
+ "epoch": 16.5625,
1990
+ "grad_norm": 0.29909902811050415,
1991
+ "learning_rate": 0.0026155921100540344,
1992
+ "loss": 1.0652649402618408,
1993
+ "step": 265
1994
+ },
1995
+ {
1996
+ "epoch": 16.625,
1997
+ "grad_norm": 0.09181151539087296,
1998
+ "learning_rate": 0.0026111067112687012,
1999
+ "loss": 1.1081140041351318,
2000
+ "step": 266
2001
+ },
2002
+ {
2003
+ "epoch": 16.6875,
2004
+ "grad_norm": 0.15625476837158203,
2005
+ "learning_rate": 0.0026065992346844626,
2006
+ "loss": 1.1298624277114868,
2007
+ "step": 267
2008
+ },
2009
+ {
2010
+ "epoch": 16.75,
2011
+ "grad_norm": 0.05208205059170723,
2012
+ "learning_rate": 0.002602069771090985,
2013
+ "loss": 1.1198302507400513,
2014
+ "step": 268
2015
+ },
2016
+ {
2017
+ "epoch": 16.8125,
2018
+ "grad_norm": 0.25109273195266724,
2019
+ "learning_rate": 0.002597518411720796,
2020
+ "loss": 1.0928356647491455,
2021
+ "step": 269
2022
+ },
2023
+ {
2024
+ "epoch": 16.875,
2025
+ "grad_norm": 0.026749474927783012,
2026
+ "learning_rate": 0.0025929452482474497,
2027
+ "loss": 1.0576367378234863,
2028
+ "step": 270
2029
+ },
2030
+ {
2031
+ "epoch": 16.9375,
2032
+ "grad_norm": 0.11184096336364746,
2033
+ "learning_rate": 0.0025883503727836764,
2034
+ "loss": 1.1186749935150146,
2035
+ "step": 271
2036
+ },
2037
+ {
2038
+ "epoch": 17.0,
2039
+ "grad_norm": 0.06893426924943924,
2040
+ "learning_rate": 0.0025837338778795325,
2041
+ "loss": 1.1332192420959473,
2042
+ "step": 272
2043
+ },
2044
+ {
2045
+ "epoch": 17.0,
2046
+ "eval_loss": 0.5755453705787659,
2047
+ "eval_runtime": 30.0862,
2048
+ "eval_samples_per_second": 3.39,
2049
+ "eval_steps_per_second": 0.432,
2050
+ "step": 272
2051
+ },
2052
+ {
2053
+ "epoch": 17.0625,
2054
+ "grad_norm": 0.28368332982063293,
2055
+ "learning_rate": 0.0025790958565205316,
2056
+ "loss": 1.1554040908813477,
2057
+ "step": 273
2058
+ },
2059
+ {
2060
+ "epoch": 17.125,
2061
+ "grad_norm": 0.13374564051628113,
2062
+ "learning_rate": 0.0025744364021257743,
2063
+ "loss": 1.119795322418213,
2064
+ "step": 274
2065
+ },
2066
+ {
2067
+ "epoch": 17.1875,
2068
+ "grad_norm": 0.46172675490379333,
2069
+ "learning_rate": 0.0025697556085460646,
2070
+ "loss": 1.1097784042358398,
2071
+ "step": 275
2072
+ },
2073
+ {
2074
+ "epoch": 17.25,
2075
+ "grad_norm": 0.05883564427495003,
2076
+ "learning_rate": 0.002565053570062023,
2077
+ "loss": 1.1001802682876587,
2078
+ "step": 276
2079
+ },
2080
+ {
2081
+ "epoch": 17.3125,
2082
+ "grad_norm": 0.038139838725328445,
2083
+ "learning_rate": 0.002560330381382183,
2084
+ "loss": 1.0831687450408936,
2085
+ "step": 277
2086
+ },
2087
+ {
2088
+ "epoch": 17.375,
2089
+ "grad_norm": 0.029911940917372704,
2090
+ "learning_rate": 0.0025555861376410868,
2091
+ "loss": 1.1179802417755127,
2092
+ "step": 278
2093
+ },
2094
+ {
2095
+ "epoch": 17.4375,
2096
+ "grad_norm": 0.038241755217313766,
2097
+ "learning_rate": 0.002550820934397368,
2098
+ "loss": 1.128745675086975,
2099
+ "step": 279
2100
+ },
2101
+ {
2102
+ "epoch": 17.5,
2103
+ "grad_norm": 0.11490178853273392,
2104
+ "learning_rate": 0.0025460348676318257,
2105
+ "loss": 1.1357793807983398,
2106
+ "step": 280
2107
+ },
2108
+ {
2109
+ "epoch": 17.5625,
2110
+ "grad_norm": 0.09644758701324463,
2111
+ "learning_rate": 0.0025412280337454942,
2112
+ "loss": 1.0953055620193481,
2113
+ "step": 281
2114
+ },
2115
+ {
2116
+ "epoch": 17.625,
2117
+ "grad_norm": 0.0858362540602684,
2118
+ "learning_rate": 0.0025364005295576975,
2119
+ "loss": 1.0640878677368164,
2120
+ "step": 282
2121
+ },
2122
+ {
2123
+ "epoch": 17.6875,
2124
+ "grad_norm": 0.04771006479859352,
2125
+ "learning_rate": 0.002531552452304103,
2126
+ "loss": 1.0404937267303467,
2127
+ "step": 283
2128
+ },
2129
+ {
2130
+ "epoch": 17.75,
2131
+ "grad_norm": 0.06784195452928543,
2132
+ "learning_rate": 0.0025266838996347595,
2133
+ "loss": 1.1567411422729492,
2134
+ "step": 284
2135
+ },
2136
+ {
2137
+ "epoch": 17.8125,
2138
+ "grad_norm": 0.07531201839447021,
2139
+ "learning_rate": 0.002521794969612133,
2140
+ "loss": 1.0844026803970337,
2141
+ "step": 285
2142
+ },
2143
+ {
2144
+ "epoch": 17.875,
2145
+ "grad_norm": 0.06453386694192886,
2146
+ "learning_rate": 0.002516885760709131,
2147
+ "loss": 1.1036568880081177,
2148
+ "step": 286
2149
+ },
2150
+ {
2151
+ "epoch": 17.9375,
2152
+ "grad_norm": 0.022719847038388252,
2153
+ "learning_rate": 0.002511956371807116,
2154
+ "loss": 1.1051126718521118,
2155
+ "step": 287
2156
+ },
2157
+ {
2158
+ "epoch": 18.0,
2159
+ "grad_norm": 0.03833983093500137,
2160
+ "learning_rate": 0.0025070069021939194,
2161
+ "loss": 1.1458027362823486,
2162
+ "step": 288
2163
+ },
2164
+ {
2165
+ "epoch": 18.0,
2166
+ "eval_loss": 0.5735718011856079,
2167
+ "eval_runtime": 30.3576,
2168
+ "eval_samples_per_second": 3.36,
2169
+ "eval_steps_per_second": 0.428,
2170
+ "step": 288
2171
+ },
2172
+ {
2173
+ "epoch": 18.0625,
2174
+ "grad_norm": 0.03980008512735367,
2175
+ "learning_rate": 0.0025020374515618373,
2176
+ "loss": 1.0746936798095703,
2177
+ "step": 289
2178
+ },
2179
+ {
2180
+ "epoch": 18.125,
2181
+ "grad_norm": 0.015201105736196041,
2182
+ "learning_rate": 0.002497048120005623,
2183
+ "loss": 1.0497227907180786,
2184
+ "step": 290
2185
+ },
2186
+ {
2187
+ "epoch": 18.1875,
2188
+ "grad_norm": 0.08461925387382507,
2189
+ "learning_rate": 0.0024920390080204733,
2190
+ "loss": 1.1033759117126465,
2191
+ "step": 291
2192
+ },
2193
+ {
2194
+ "epoch": 18.25,
2195
+ "grad_norm": 0.02152029238641262,
2196
+ "learning_rate": 0.0024870102165000015,
2197
+ "loss": 1.1139872074127197,
2198
+ "step": 292
2199
+ },
2200
+ {
2201
+ "epoch": 18.3125,
2202
+ "grad_norm": 0.030964918434619904,
2203
+ "learning_rate": 0.0024819618467342066,
2204
+ "loss": 1.0686633586883545,
2205
+ "step": 293
2206
+ },
2207
+ {
2208
+ "epoch": 18.375,
2209
+ "grad_norm": 0.03928976133465767,
2210
+ "learning_rate": 0.0024768940004074335,
2211
+ "loss": 1.1165332794189453,
2212
+ "step": 294
2213
+ },
2214
+ {
2215
+ "epoch": 18.4375,
2216
+ "grad_norm": 0.013071025721728802,
2217
+ "learning_rate": 0.002471806779596324,
2218
+ "loss": 1.0793699026107788,
2219
+ "step": 295
2220
+ },
2221
+ {
2222
+ "epoch": 18.5,
2223
+ "grad_norm": 0.017155099660158157,
2224
+ "learning_rate": 0.0024667002867677606,
2225
+ "loss": 1.1448696851730347,
2226
+ "step": 296
2227
+ },
2228
+ {
2229
+ "epoch": 18.5625,
2230
+ "grad_norm": 0.025132035836577415,
2231
+ "learning_rate": 0.0024615746247768043,
2232
+ "loss": 1.1179168224334717,
2233
+ "step": 297
2234
+ },
2235
+ {
2236
+ "epoch": 18.625,
2237
+ "grad_norm": 0.030385689809918404,
2238
+ "learning_rate": 0.0024564298968646203,
2239
+ "loss": 1.0827066898345947,
2240
+ "step": 298
2241
+ },
2242
+ {
2243
+ "epoch": 18.6875,
2244
+ "grad_norm": 0.08415888249874115,
2245
+ "learning_rate": 0.0024512662066563996,
2246
+ "loss": 1.1512041091918945,
2247
+ "step": 299
2248
+ },
2249
+ {
2250
+ "epoch": 18.75,
2251
+ "grad_norm": 0.0354602225124836,
2252
+ "learning_rate": 0.0024460836581592744,
2253
+ "loss": 1.1055684089660645,
2254
+ "step": 300
2255
+ },
2256
+ {
2257
+ "epoch": 18.8125,
2258
+ "grad_norm": 0.061975862830877304,
2259
+ "learning_rate": 0.0024408823557602196,
2260
+ "loss": 1.1121817827224731,
2261
+ "step": 301
2262
+ },
2263
+ {
2264
+ "epoch": 18.875,
2265
+ "grad_norm": 0.12300977855920792,
2266
+ "learning_rate": 0.0024356624042239498,
2267
+ "loss": 1.1116411685943604,
2268
+ "step": 302
2269
+ },
2270
+ {
2271
+ "epoch": 18.9375,
2272
+ "grad_norm": 0.024768438190221786,
2273
+ "learning_rate": 0.002430423908690813,
2274
+ "loss": 1.1031725406646729,
2275
+ "step": 303
2276
+ },
2277
+ {
2278
+ "epoch": 19.0,
2279
+ "grad_norm": 0.20697493851184845,
2280
+ "learning_rate": 0.0024251669746746703,
2281
+ "loss": 1.1268653869628906,
2282
+ "step": 304
2283
+ },
2284
+ {
2285
+ "epoch": 19.0,
2286
+ "eval_loss": 0.5697882175445557,
2287
+ "eval_runtime": 30.3468,
2288
+ "eval_samples_per_second": 3.361,
2289
+ "eval_steps_per_second": 0.428,
2290
+ "step": 304
2291
+ },
2292
+ {
2293
+ "epoch": 19.0625,
2294
+ "grad_norm": 0.04372715577483177,
2295
+ "learning_rate": 0.0024198917080607706,
2296
+ "loss": 1.0879970788955688,
2297
+ "step": 305
2298
+ },
2299
+ {
2300
+ "epoch": 19.125,
2301
+ "grad_norm": 0.3106955885887146,
2302
+ "learning_rate": 0.0024145982151036176,
2303
+ "loss": 1.0981900691986084,
2304
+ "step": 306
2305
+ },
2306
+ {
2307
+ "epoch": 19.1875,
2308
+ "grad_norm": 0.04271978512406349,
2309
+ "learning_rate": 0.0024092866024248323,
2310
+ "loss": 1.067917823791504,
2311
+ "step": 307
2312
+ },
2313
+ {
2314
+ "epoch": 19.25,
2315
+ "grad_norm": 0.03746578097343445,
2316
+ "learning_rate": 0.0024039569770110012,
2317
+ "loss": 1.168632984161377,
2318
+ "step": 308
2319
+ },
2320
+ {
2321
+ "epoch": 19.3125,
2322
+ "grad_norm": 0.02499796636402607,
2323
+ "learning_rate": 0.0023986094462115255,
2324
+ "loss": 1.1304916143417358,
2325
+ "step": 309
2326
+ },
2327
+ {
2328
+ "epoch": 19.375,
2329
+ "grad_norm": 0.031224632635712624,
2330
+ "learning_rate": 0.002393244117736455,
2331
+ "loss": 1.1047842502593994,
2332
+ "step": 310
2333
+ },
2334
+ {
2335
+ "epoch": 19.4375,
2336
+ "grad_norm": 0.03269527107477188,
2337
+ "learning_rate": 0.0023878610996543224,
2338
+ "loss": 1.085580825805664,
2339
+ "step": 311
2340
+ },
2341
+ {
2342
+ "epoch": 19.5,
2343
+ "grad_norm": 0.027562979608774185,
2344
+ "learning_rate": 0.002382460500389964,
2345
+ "loss": 1.1179475784301758,
2346
+ "step": 312
2347
+ },
2348
+ {
2349
+ "epoch": 19.5625,
2350
+ "grad_norm": 0.05654411390423775,
2351
+ "learning_rate": 0.0023770424287223365,
2352
+ "loss": 1.1345707178115845,
2353
+ "step": 313
2354
+ },
2355
+ {
2356
+ "epoch": 19.625,
2357
+ "grad_norm": 0.23459751904010773,
2358
+ "learning_rate": 0.002371606993782326,
2359
+ "loss": 1.0679315328598022,
2360
+ "step": 314
2361
+ },
2362
+ {
2363
+ "epoch": 19.6875,
2364
+ "grad_norm": 0.09804576635360718,
2365
+ "learning_rate": 0.0023661543050505496,
2366
+ "loss": 1.0888817310333252,
2367
+ "step": 315
2368
+ },
2369
+ {
2370
+ "epoch": 19.75,
2371
+ "grad_norm": 0.09066950529813766,
2372
+ "learning_rate": 0.002360684472355151,
2373
+ "loss": 1.094346046447754,
2374
+ "step": 316
2375
+ },
2376
+ {
2377
+ "epoch": 19.8125,
2378
+ "grad_norm": 0.043143078684806824,
2379
+ "learning_rate": 0.0023551976058695887,
2380
+ "loss": 1.081190586090088,
2381
+ "step": 317
2382
+ },
2383
+ {
2384
+ "epoch": 19.875,
2385
+ "grad_norm": 0.03301508352160454,
2386
+ "learning_rate": 0.002349693816110414,
2387
+ "loss": 1.0623350143432617,
2388
+ "step": 318
2389
+ },
2390
+ {
2391
+ "epoch": 19.9375,
2392
+ "grad_norm": 0.022927623242139816,
2393
+ "learning_rate": 0.0023441732139350487,
2394
+ "loss": 1.0692837238311768,
2395
+ "step": 319
2396
+ },
2397
+ {
2398
+ "epoch": 20.0,
2399
+ "grad_norm": 0.03282017260789871,
2400
+ "learning_rate": 0.002338635910539551,
2401
+ "loss": 1.128983736038208,
2402
+ "step": 320
2403
+ }
2404
+ ],
2405
+ "logging_steps": 1,
2406
+ "max_steps": 800,
2407
+ "num_input_tokens_seen": 0,
2408
+ "num_train_epochs": 50,
2409
+ "save_steps": 500,
2410
+ "stateful_callbacks": {
2411
+ "TrainerControl": {
2412
+ "args": {
2413
+ "should_epoch_stop": false,
2414
+ "should_evaluate": true,
2415
+ "should_log": false,
2416
+ "should_save": false,
2417
+ "should_training_stop": false
2418
+ },
2419
+ "attributes": {}
2420
+ }
2421
+ },
2422
+ "total_flos": 4289416050769920.0,
2423
+ "train_batch_size": 8,
2424
+ "trial_name": null,
2425
+ "trial_params": null
2426
+ }
runs/test-model/checkpoint-320/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:45cab7e8b9705eceb6dc5bbdf6d75a3f15b3ecba44dfc0a4b5c7d2d8dc6b7db7
3
+ size 5265
runs/test-model/checkpoint-80/chat_template.jinja ADDED
@@ -0,0 +1,85 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0].role == 'system' %}
4
+ {{- messages[0].content + '\n\n' }}
5
+ {%- endif %}
6
+ {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
7
+ {%- for tool in tools %}
8
+ {{- "\n" }}
9
+ {{- tool | tojson }}
10
+ {%- endfor %}
11
+ {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
12
+ {%- else %}
13
+ {%- if messages[0].role == 'system' %}
14
+ {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
15
+ {%- endif %}
16
+ {%- endif %}
17
+ {%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
18
+ {%- for message in messages[::-1] %}
19
+ {%- set index = (messages|length - 1) - loop.index0 %}
20
+ {%- if ns.multi_step_tool and message.role == "user" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
21
+ {%- set ns.multi_step_tool = false %}
22
+ {%- set ns.last_query_index = index %}
23
+ {%- endif %}
24
+ {%- endfor %}
25
+ {%- for message in messages %}
26
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
27
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
28
+ {%- elif message.role == "assistant" %}
29
+ {%- set content = message.content %}
30
+ {%- set reasoning_content = '' %}
31
+ {%- if message.reasoning_content is defined and message.reasoning_content is not none %}
32
+ {%- set reasoning_content = message.reasoning_content %}
33
+ {%- else %}
34
+ {%- if '</think>' in message.content %}
35
+ {%- set content = message.content.split('</think>')[-1].lstrip('\n') %}
36
+ {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
37
+ {%- endif %}
38
+ {%- endif %}
39
+ {%- if loop.index0 > ns.last_query_index %}
40
+ {%- if loop.last or (not loop.last and reasoning_content) %}
41
+ {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
42
+ {%- else %}
43
+ {{- '<|im_start|>' + message.role + '\n' + content }}
44
+ {%- endif %}
45
+ {%- else %}
46
+ {{- '<|im_start|>' + message.role + '\n' + content }}
47
+ {%- endif %}
48
+ {%- if message.tool_calls %}
49
+ {%- for tool_call in message.tool_calls %}
50
+ {%- if (loop.first and content) or (not loop.first) %}
51
+ {{- '\n' }}
52
+ {%- endif %}
53
+ {%- if tool_call.function %}
54
+ {%- set tool_call = tool_call.function %}
55
+ {%- endif %}
56
+ {{- '<tool_call>\n{"name": "' }}
57
+ {{- tool_call.name }}
58
+ {{- '", "arguments": ' }}
59
+ {%- if tool_call.arguments is string %}
60
+ {{- tool_call.arguments }}
61
+ {%- else %}
62
+ {{- tool_call.arguments | tojson }}
63
+ {%- endif %}
64
+ {{- '}\n</tool_call>' }}
65
+ {%- endfor %}
66
+ {%- endif %}
67
+ {{- '<|im_end|>\n' }}
68
+ {%- elif message.role == "tool" %}
69
+ {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
70
+ {{- '<|im_start|>user' }}
71
+ {%- endif %}
72
+ {{- '\n<tool_response>\n' }}
73
+ {{- message.content }}
74
+ {{- '\n</tool_response>' }}
75
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
76
+ {{- '<|im_end|>\n' }}
77
+ {%- endif %}
78
+ {%- endif %}
79
+ {%- endfor %}
80
+ {%- if add_generation_prompt %}
81
+ {{- '<|im_start|>assistant\n' }}
82
+ {%- if enable_thinking is defined and enable_thinking is false %}
83
+ {{- '<think>\n\n</think>\n\n' }}
84
+ {%- endif %}
85
+ {%- endif %}
runs/test-model/checkpoint-80/config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "LlamaForCausalLM"
4
+ ],
5
+ "attention_bias": false,
6
+ "attention_dropout": 0.0,
7
+ "bos_token_id": null,
8
+ "dtype": "float32",
9
+ "eos_token_id": 151645,
10
+ "head_dim": 128,
11
+ "hidden_act": "silu",
12
+ "hidden_size": 512,
13
+ "initializer_range": 0.02,
14
+ "intermediate_size": 1536,
15
+ "max_position_embeddings": 2048,
16
+ "mlp_bias": false,
17
+ "model_type": "llama",
18
+ "num_attention_heads": 4,
19
+ "num_hidden_layers": 20,
20
+ "num_key_value_heads": 4,
21
+ "pad_token_id": 151645,
22
+ "pretraining_tp": 1,
23
+ "rms_norm_eps": 1e-06,
24
+ "rope_parameters": {
25
+ "rope_theta": 10000.0,
26
+ "rope_type": "default"
27
+ },
28
+ "tie_word_embeddings": true,
29
+ "transformers_version": "5.5.0",
30
+ "use_cache": false,
31
+ "vocab_size": 151671
32
+ }
runs/test-model/checkpoint-80/generation_config.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_from_model_config": true,
3
+ "eos_token_id": [
4
+ 151645
5
+ ],
6
+ "output_attentions": false,
7
+ "output_hidden_states": false,
8
+ "pad_token_id": 151645,
9
+ "transformers_version": "5.5.0",
10
+ "use_cache": true
11
+ }
runs/test-model/checkpoint-80/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7010337ddd1e2cfefbabe44c6658ae857e41e281477422edf569bd56c444c48e
3
+ size 583356232
runs/test-model/checkpoint-80/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:126c8397d533d83915415be6d3ef122ab4eb3cf3fcf8a1ab4ac2f9e67ad45199
3
+ size 1166825803