Hugging Face's logo

NJUDeepEngine
/
MeteoRA-llama2-13b

Safetensors
Model card Files Community
ParagonLight commited on
Commit
a12f5b5
•
1 Parent(s): 0a8e533

upload loras

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. llama2_13b_peft/elementary_math_qa/README.md +138 -0
  2. llama2_13b_peft/elementary_math_qa/adapter_config.json +33 -0
  3. llama2_13b_peft/elementary_math_qa/adapter_model.safetensors +3 -0
  4. llama2_13b_peft/elementary_math_qa/all_results.json +11 -0
  5. llama2_13b_peft/elementary_math_qa/eval_results.json +7 -0
  6. llama2_13b_peft/elementary_math_qa/special_tokens_map.json +24 -0
  7. llama2_13b_peft/elementary_math_qa/tokenizer.model +3 -0
  8. llama2_13b_peft/elementary_math_qa/tokenizer_config.json +45 -0
  9. llama2_13b_peft/elementary_math_qa/train_results.json +7 -0
  10. llama2_13b_peft/elementary_math_qa/trainer_log.jsonl +0 -0
  11. llama2_13b_peft/elementary_math_qa/trainer_state.json +3169 -0
  12. llama2_13b_peft/elementary_math_qa/training_args.bin +3 -0
  13. llama2_13b_peft/elementary_math_qa/training_eval_loss.png +0 -0
  14. llama2_13b_peft/elementary_math_qa/training_loss.png +0 -0
  15. llama2_13b_peft/epistemic_reasoning/README.md +70 -0
  16. llama2_13b_peft/epistemic_reasoning/adapter_config.json +34 -0
  17. llama2_13b_peft/epistemic_reasoning/adapter_model.safetensors +3 -0
  18. llama2_13b_peft/epistemic_reasoning/all_results.json +12 -0
  19. llama2_13b_peft/epistemic_reasoning/eval_results.json +7 -0
  20. llama2_13b_peft/epistemic_reasoning/special_tokens_map.json +24 -0
  21. llama2_13b_peft/epistemic_reasoning/tokenizer.model +3 -0
  22. llama2_13b_peft/epistemic_reasoning/tokenizer_config.json +45 -0
  23. llama2_13b_peft/epistemic_reasoning/train_results.json +8 -0
  24. llama2_13b_peft/epistemic_reasoning/trainer_log.jsonl +57 -0
  25. llama2_13b_peft/epistemic_reasoning/trainer_state.json +420 -0
  26. llama2_13b_peft/epistemic_reasoning/training_args.bin +3 -0
  27. llama2_13b_peft/epistemic_reasoning/training_eval_loss.png +0 -0
  28. llama2_13b_peft/epistemic_reasoning/training_loss.png +0 -0
  29. llama2_13b_peft/formal_fallacies_syllogisms_negation/README.md +118 -0
  30. llama2_13b_peft/formal_fallacies_syllogisms_negation/adapter_config.json +33 -0
  31. llama2_13b_peft/formal_fallacies_syllogisms_negation/adapter_model.safetensors +3 -0
  32. llama2_13b_peft/formal_fallacies_syllogisms_negation/all_results.json +11 -0
  33. llama2_13b_peft/formal_fallacies_syllogisms_negation/eval_results.json +7 -0
  34. llama2_13b_peft/formal_fallacies_syllogisms_negation/special_tokens_map.json +24 -0
  35. llama2_13b_peft/formal_fallacies_syllogisms_negation/tokenizer.model +3 -0
  36. llama2_13b_peft/formal_fallacies_syllogisms_negation/tokenizer_config.json +45 -0
  37. llama2_13b_peft/formal_fallacies_syllogisms_negation/train_results.json +7 -0
  38. llama2_13b_peft/formal_fallacies_syllogisms_negation/trainer_log.jsonl +320 -0
  39. llama2_13b_peft/formal_fallacies_syllogisms_negation/trainer_state.json +2309 -0
  40. llama2_13b_peft/formal_fallacies_syllogisms_negation/training_args.bin +3 -0
  41. llama2_13b_peft/formal_fallacies_syllogisms_negation/training_eval_loss.png +0 -0
  42. llama2_13b_peft/formal_fallacies_syllogisms_negation/training_loss.png +0 -0
  43. llama2_13b_peft/goal_step_wikihow/README.md +71 -0
  44. llama2_13b_peft/goal_step_wikihow/adapter_config.json +34 -0
  45. llama2_13b_peft/goal_step_wikihow/adapter_model.safetensors +3 -0
  46. llama2_13b_peft/goal_step_wikihow/all_results.json +12 -0
  47. llama2_13b_peft/goal_step_wikihow/eval_results.json +7 -0
  48. llama2_13b_peft/goal_step_wikihow/special_tokens_map.json +24 -0
  49. llama2_13b_peft/goal_step_wikihow/tokenizer.model +3 -0
  50. llama2_13b_peft/goal_step_wikihow/tokenizer_config.json +45 -0
llama2_13b_peft/elementary_math_qa/README.md ADDED
@@ -0,0 +1,138 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ library_name: peft
4
+ tags:
5
+ - llama-factory
6
+ - lora
7
+ - generated_from_trainer
8
+ base_model: /data1/model/llama2/meta-llama/Llama2-13b
9
+ model-index:
10
+ - name: elementary_math_qa_no_sys
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # elementary_math_qa_no_sys
18
+
19
+ This model is a fine-tuned version of [/data1/model/llama2/meta-llama/Llama2-13b](https://huggingface.co//data1/model/llama2/meta-llama/Llama2-13b) on the elementary_math_qa_no_sys dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 0.0705
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 0.0001
41
+ - train_batch_size: 8
42
+ - eval_batch_size: 8
43
+ - seed: 42
44
+ - distributed_type: multi-GPU
45
+ - num_devices: 3
46
+ - total_train_batch_size: 24
47
+ - total_eval_batch_size: 24
48
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
49
+ - lr_scheduler_type: cosine
50
+ - lr_scheduler_warmup_steps: 20
51
+ - num_epochs: 10.0
52
+
53
+ ### Training results
54
+
55
+ | Training Loss | Epoch | Step | Validation Loss |
56
+ |:-------------:|:-----:|:----:|:---------------:|
57
+ | 0.3438 | 0.05 | 50 | 0.3364 |
58
+ | 0.3021 | 0.09 | 100 | 0.3056 |
59
+ | 0.2676 | 0.14 | 150 | 0.2710 |
60
+ | 0.2679 | 0.18 | 200 | 0.2582 |
61
+ | 0.2448 | 0.23 | 250 | 0.2475 |
62
+ | 0.2355 | 0.28 | 300 | 0.2372 |
63
+ | 0.2335 | 0.32 | 350 | 0.2282 |
64
+ | 0.2226 | 0.37 | 400 | 0.2227 |
65
+ | 0.2098 | 0.42 | 450 | 0.2109 |
66
+ | 0.1839 | 0.46 | 500 | 0.2048 |
67
+ | 0.2008 | 0.51 | 550 | 0.1992 |
68
+ | 0.1945 | 0.55 | 600 | 0.2019 |
69
+ | 0.1891 | 0.6 | 650 | 0.1859 |
70
+ | 0.2015 | 0.65 | 700 | 0.1966 |
71
+ | 0.174 | 0.69 | 750 | 0.1801 |
72
+ | 0.1565 | 0.74 | 800 | 0.1762 |
73
+ | 0.1825 | 0.79 | 850 | 0.1717 |
74
+ | 0.1651 | 0.83 | 900 | 0.1682 |
75
+ | 0.1598 | 0.88 | 950 | 0.1598 |
76
+ | 0.1502 | 0.92 | 1000 | 0.1558 |
77
+ | 0.1599 | 0.97 | 1050 | 0.1465 |
78
+ | 0.0977 | 1.02 | 1100 | 0.1520 |
79
+ | 0.1166 | 1.06 | 1150 | 0.1403 |
80
+ | 0.0943 | 1.11 | 1200 | 0.1387 |
81
+ | 0.1007 | 1.16 | 1250 | 0.1311 |
82
+ | 0.1035 | 1.2 | 1300 | 0.1325 |
83
+ | 0.0842 | 1.25 | 1350 | 0.1309 |
84
+ | 0.1114 | 1.29 | 1400 | 0.1225 |
85
+ | 0.1047 | 1.34 | 1450 | 0.1184 |
86
+ | 0.0807 | 1.39 | 1500 | 0.1136 |
87
+ | 0.0846 | 1.43 | 1550 | 0.1200 |
88
+ | 0.0737 | 1.48 | 1600 | 0.1145 |
89
+ | 0.0844 | 1.52 | 1650 | 0.1037 |
90
+ | 0.0809 | 1.57 | 1700 | 0.0940 |
91
+ | 0.0718 | 1.62 | 1750 | 0.0931 |
92
+ | 0.0687 | 1.66 | 1800 | 0.0930 |
93
+ | 0.0629 | 1.71 | 1850 | 0.0969 |
94
+ | 0.0852 | 1.76 | 1900 | 0.0872 |
95
+ | 0.0622 | 1.8 | 1950 | 0.0849 |
96
+ | 0.0653 | 1.85 | 2000 | 0.0831 |
97
+ | 0.0507 | 1.89 | 2050 | 0.0829 |
98
+ | 0.0518 | 1.94 | 2100 | 0.0785 |
99
+ | 0.0566 | 1.99 | 2150 | 0.0750 |
100
+ | 0.0193 | 2.03 | 2200 | 0.0837 |
101
+ | 0.0233 | 2.08 | 2250 | 0.0766 |
102
+ | 0.0249 | 2.13 | 2300 | 0.0829 |
103
+ | 0.0217 | 2.17 | 2350 | 0.0824 |
104
+ | 0.0233 | 2.22 | 2400 | 0.0735 |
105
+ | 0.0192 | 2.26 | 2450 | 0.0767 |
106
+ | 0.0207 | 2.31 | 2500 | 0.0794 |
107
+ | 0.0232 | 2.36 | 2550 | 0.0843 |
108
+ | 0.0295 | 2.4 | 2600 | 0.0800 |
109
+ | 0.0185 | 2.45 | 2650 | 0.0777 |
110
+ | 0.0178 | 2.5 | 2700 | 0.0767 |
111
+ | 0.0245 | 2.54 | 2750 | 0.0717 |
112
+ | 0.0226 | 2.59 | 2800 | 0.0774 |
113
+ | 0.0222 | 2.63 | 2850 | 0.0671 |
114
+ | 0.0194 | 2.68 | 2900 | 0.0666 |
115
+ | 0.0162 | 2.73 | 2950 | 0.0713 |
116
+ | 0.0184 | 2.77 | 3000 | 0.0740 |
117
+ | 0.0227 | 2.82 | 3050 | 0.0675 |
118
+ | 0.0176 | 2.87 | 3100 | 0.0701 |
119
+ | 0.034 | 2.91 | 3150 | 0.0675 |
120
+ | 0.0148 | 2.96 | 3200 | 0.0688 |
121
+ | 0.014 | 3.0 | 3250 | 0.0673 |
122
+ | 0.0178 | 3.05 | 3300 | 0.0719 |
123
+ | 0.0059 | 3.1 | 3350 | 0.0734 |
124
+ | 0.0069 | 3.14 | 3400 | 0.0764 |
125
+ | 0.0074 | 3.19 | 3450 | 0.0818 |
126
+ | 0.009 | 3.23 | 3500 | 0.0705 |
127
+ | 0.0048 | 3.28 | 3550 | 0.0735 |
128
+ | 0.005 | 3.33 | 3600 | 0.0705 |
129
+ | 0.0073 | 3.37 | 3650 | 0.0724 |
130
+
131
+
132
+ ### Framework versions
133
+
134
+ - PEFT 0.9.0
135
+ - Transformers 4.38.2
136
+ - Pytorch 2.2.1
137
+ - Datasets 2.18.0
138
+ - Tokenizers 0.15.2
llama2_13b_peft/elementary_math_qa/adapter_config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "/data1/model/llama2/meta-llama/Llama2-13b",
5
+ "bias": "none",
6
+ "fan_in_fan_out": false,
7
+ "inference_mode": true,
8
+ "init_lora_weights": true,
9
+ "layers_pattern": null,
10
+ "layers_to_transform": null,
11
+ "loftq_config": {},
12
+ "lora_alpha": 16,
13
+ "lora_dropout": 0.0,
14
+ "megatron_config": null,
15
+ "megatron_core": "megatron.core",
16
+ "modules_to_save": null,
17
+ "peft_type": "LORA",
18
+ "r": 8,
19
+ "rank_pattern": {},
20
+ "revision": null,
21
+ "target_modules": [
22
+ "o_proj",
23
+ "gate_proj",
24
+ "q_proj",
25
+ "k_proj",
26
+ "v_proj",
27
+ "down_proj",
28
+ "up_proj"
29
+ ],
30
+ "task_type": "CAUSAL_LM",
31
+ "use_dora": false,
32
+ "use_rslora": false
33
+ }
llama2_13b_peft/elementary_math_qa/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c95310146c699416787331f0c95920283e55a47a6355cc22383f3c069b9929aa
3
+ size 125248064
llama2_13b_peft/elementary_math_qa/all_results.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.37,
3
+ "eval_loss": 0.07047122716903687,
4
+ "eval_runtime": 36.1294,
5
+ "eval_samples_per_second": 126.766,
6
+ "eval_steps_per_second": 5.287,
7
+ "train_loss": 0.10634728888738645,
8
+ "train_runtime": 4965.2136,
9
+ "train_samples_per_second": 52.266,
10
+ "train_steps_per_second": 2.179
11
+ }
llama2_13b_peft/elementary_math_qa/eval_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.37,
3
+ "eval_loss": 0.07047122716903687,
4
+ "eval_runtime": 36.1294,
5
+ "eval_samples_per_second": 126.766,
6
+ "eval_steps_per_second": 5.287
7
+ }
llama2_13b_peft/elementary_math_qa/special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "</s>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
llama2_13b_peft/elementary_math_qa/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
llama2_13b_peft/elementary_math_qa/tokenizer_config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "add_prefix_space": true,
5
+ "added_tokens_decoder": {
6
+ "0": {
7
+ "content": "<unk>",
8
+ "lstrip": false,
9
+ "normalized": false,
10
+ "rstrip": false,
11
+ "single_word": false,
12
+ "special": true
13
+ },
14
+ "1": {
15
+ "content": "<s>",
16
+ "lstrip": false,
17
+ "normalized": false,
18
+ "rstrip": false,
19
+ "single_word": false,
20
+ "special": true
21
+ },
22
+ "2": {
23
+ "content": "</s>",
24
+ "lstrip": false,
25
+ "normalized": false,
26
+ "rstrip": false,
27
+ "single_word": false,
28
+ "special": true
29
+ }
30
+ },
31
+ "bos_token": "<s>",
32
+ "chat_template": "{% if messages[0]['role'] == 'system' %}{% set system_message = messages[0]['content'] %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ content }}{% elif message['role'] == 'assistant' %}{{ content + '\\n' }}{% endif %}{% endfor %}",
33
+ "clean_up_tokenization_spaces": false,
34
+ "eos_token": "</s>",
35
+ "legacy": true,
36
+ "model_max_length": 1000000000000000019884624838656,
37
+ "pad_token": "</s>",
38
+ "padding_side": "right",
39
+ "sp_model_kwargs": {},
40
+ "spaces_between_special_tokens": false,
41
+ "split_special_tokens": false,
42
+ "tokenizer_class": "LlamaTokenizer",
43
+ "unk_token": "<unk>",
44
+ "use_default_system_prompt": false
45
+ }
llama2_13b_peft/elementary_math_qa/train_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 3.37,
3
+ "train_loss": 0.10634728888738645,
4
+ "train_runtime": 4965.2136,
5
+ "train_samples_per_second": 52.266,
6
+ "train_steps_per_second": 2.179
7
+ }
llama2_13b_peft/elementary_math_qa/trainer_log.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
llama2_13b_peft/elementary_math_qa/trainer_state.json ADDED
@@ -0,0 +1,3169 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.07047122716903687,
3
+ "best_model_checkpoint": "ckpt/llama2_13b_fuze15_no_sys/elementary_math_qa_no_sys/checkpoint-3500",
4
+ "epoch": 3.3733826247689462,
5
+ "eval_steps": 50,
6
+ "global_step": 3650,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.01,
13
+ "grad_norm": 4.040522575378418,
14
+ "learning_rate": 5e-05,
15
+ "loss": 3.4664,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.02,
20
+ "grad_norm": 0.8851511478424072,
21
+ "learning_rate": 0.0001,
22
+ "loss": 1.2417,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.03,
27
+ "grad_norm": 0.6568808555603027,
28
+ "learning_rate": 9.999978846027431e-05,
29
+ "loss": 0.3726,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.04,
34
+ "grad_norm": 0.29808229207992554,
35
+ "learning_rate": 9.999915384288722e-05,
36
+ "loss": 0.3371,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.05,
41
+ "grad_norm": 0.2813255488872528,
42
+ "learning_rate": 9.999809615320856e-05,
43
+ "loss": 0.3438,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.05,
48
+ "eval_loss": 0.3363588750362396,
49
+ "eval_runtime": 35.8964,
50
+ "eval_samples_per_second": 127.589,
51
+ "eval_steps_per_second": 5.321,
52
+ "step": 50
53
+ },
54
+ {
55
+ "epoch": 0.06,
56
+ "grad_norm": 0.2735520899295807,
57
+ "learning_rate": 9.999661540018812e-05,
58
+ "loss": 0.3451,
59
+ "step": 60
60
+ },
61
+ {
62
+ "epoch": 0.06,
63
+ "grad_norm": 0.2597522735595703,
64
+ "learning_rate": 9.999471159635539e-05,
65
+ "loss": 0.3227,
66
+ "step": 70
67
+ },
68
+ {
69
+ "epoch": 0.07,
70
+ "grad_norm": 0.2065269947052002,
71
+ "learning_rate": 9.999238475781957e-05,
72
+ "loss": 0.3368,
73
+ "step": 80
74
+ },
75
+ {
76
+ "epoch": 0.08,
77
+ "grad_norm": 0.44564762711524963,
78
+ "learning_rate": 9.998963490426943e-05,
79
+ "loss": 0.3171,
80
+ "step": 90
81
+ },
82
+ {
83
+ "epoch": 0.09,
84
+ "grad_norm": 0.29043757915496826,
85
+ "learning_rate": 9.998646205897309e-05,
86
+ "loss": 0.3021,
87
+ "step": 100
88
+ },
89
+ {
90
+ "epoch": 0.09,
91
+ "eval_loss": 0.30562111735343933,
92
+ "eval_runtime": 36.0314,
93
+ "eval_samples_per_second": 127.111,
94
+ "eval_steps_per_second": 5.301,
95
+ "step": 100
96
+ },
97
+ {
98
+ "epoch": 0.1,
99
+ "grad_norm": 0.7949171662330627,
100
+ "learning_rate": 9.998286624877786e-05,
101
+ "loss": 0.3053,
102
+ "step": 110
103
+ },
104
+ {
105
+ "epoch": 0.11,
106
+ "grad_norm": 0.44385454058647156,
107
+ "learning_rate": 9.997884750411005e-05,
108
+ "loss": 0.2789,
109
+ "step": 120
110
+ },
111
+ {
112
+ "epoch": 0.12,
113
+ "grad_norm": 1.0356186628341675,
114
+ "learning_rate": 9.997440585897455e-05,
115
+ "loss": 0.2854,
116
+ "step": 130
117
+ },
118
+ {
119
+ "epoch": 0.13,
120
+ "grad_norm": 0.43352845311164856,
121
+ "learning_rate": 9.99695413509548e-05,
122
+ "loss": 0.2841,
123
+ "step": 140
124
+ },
125
+ {
126
+ "epoch": 0.14,
127
+ "grad_norm": 0.9958717823028564,
128
+ "learning_rate": 9.996425402121223e-05,
129
+ "loss": 0.2676,
130
+ "step": 150
131
+ },
132
+ {
133
+ "epoch": 0.14,
134
+ "eval_loss": 0.27099743485450745,
135
+ "eval_runtime": 36.2627,
136
+ "eval_samples_per_second": 126.301,
137
+ "eval_steps_per_second": 5.267,
138
+ "step": 150
139
+ },
140
+ {
141
+ "epoch": 0.15,
142
+ "grad_norm": 0.5326156616210938,
143
+ "learning_rate": 9.995854391448606e-05,
144
+ "loss": 0.2841,
145
+ "step": 160
146
+ },
147
+ {
148
+ "epoch": 0.16,
149
+ "grad_norm": 0.49322521686553955,
150
+ "learning_rate": 9.99524110790929e-05,
151
+ "loss": 0.2624,
152
+ "step": 170
153
+ },
154
+ {
155
+ "epoch": 0.17,
156
+ "grad_norm": 0.777406632900238,
157
+ "learning_rate": 9.994585556692624e-05,
158
+ "loss": 0.2398,
159
+ "step": 180
160
+ },
161
+ {
162
+ "epoch": 0.18,
163
+ "grad_norm": 0.509154200553894,
164
+ "learning_rate": 9.993887743345614e-05,
165
+ "loss": 0.2488,
166
+ "step": 190
167
+ },
168
+ {
169
+ "epoch": 0.18,
170
+ "grad_norm": 0.6434066891670227,
171
+ "learning_rate": 9.99314767377287e-05,
172
+ "loss": 0.2679,
173
+ "step": 200
174
+ },
175
+ {
176
+ "epoch": 0.18,
177
+ "eval_loss": 0.25818708539009094,
178
+ "eval_runtime": 36.3199,
179
+ "eval_samples_per_second": 126.102,
180
+ "eval_steps_per_second": 5.259,
181
+ "step": 200
182
+ },
183
+ {
184
+ "epoch": 0.19,
185
+ "grad_norm": 0.460243821144104,
186
+ "learning_rate": 9.992365354236557e-05,
187
+ "loss": 0.2595,
188
+ "step": 210
189
+ },
190
+ {
191
+ "epoch": 0.2,
192
+ "grad_norm": 0.5987964868545532,
193
+ "learning_rate": 9.991540791356342e-05,
194
+ "loss": 0.2625,
195
+ "step": 220
196
+ },
197
+ {
198
+ "epoch": 0.21,
199
+ "grad_norm": 0.9906507134437561,
200
+ "learning_rate": 9.990673992109335e-05,
201
+ "loss": 0.2491,
202
+ "step": 230
203
+ },
204
+ {
205
+ "epoch": 0.22,
206
+ "grad_norm": 0.7199179530143738,
207
+ "learning_rate": 9.989764963830037e-05,
208
+ "loss": 0.2272,
209
+ "step": 240
210
+ },
211
+ {
212
+ "epoch": 0.23,
213
+ "grad_norm": 0.5103895664215088,
214
+ "learning_rate": 9.988813714210272e-05,
215
+ "loss": 0.2448,
216
+ "step": 250
217
+ },
218
+ {
219
+ "epoch": 0.23,
220
+ "eval_loss": 0.2475069910287857,
221
+ "eval_runtime": 36.3235,
222
+ "eval_samples_per_second": 126.089,
223
+ "eval_steps_per_second": 5.258,
224
+ "step": 250
225
+ },
226
+ {
227
+ "epoch": 0.24,
228
+ "grad_norm": 0.6703325510025024,
229
+ "learning_rate": 9.987820251299122e-05,
230
+ "loss": 0.2425,
231
+ "step": 260
232
+ },
233
+ {
234
+ "epoch": 0.25,
235
+ "grad_norm": 0.40366512537002563,
236
+ "learning_rate": 9.986784583502862e-05,
237
+ "loss": 0.2502,
238
+ "step": 270
239
+ },
240
+ {
241
+ "epoch": 0.26,
242
+ "grad_norm": 0.569157600402832,
243
+ "learning_rate": 9.985706719584888e-05,
244
+ "loss": 0.2599,
245
+ "step": 280
246
+ },
247
+ {
248
+ "epoch": 0.27,
249
+ "grad_norm": 1.02542245388031,
250
+ "learning_rate": 9.98458666866564e-05,
251
+ "loss": 0.244,
252
+ "step": 290
253
+ },
254
+ {
255
+ "epoch": 0.28,
256
+ "grad_norm": 0.5159379243850708,
257
+ "learning_rate": 9.983424440222531e-05,
258
+ "loss": 0.2355,
259
+ "step": 300
260
+ },
261
+ {
262
+ "epoch": 0.28,
263
+ "eval_loss": 0.23719529807567596,
264
+ "eval_runtime": 36.3327,
265
+ "eval_samples_per_second": 126.057,
266
+ "eval_steps_per_second": 5.257,
267
+ "step": 300
268
+ },
269
+ {
270
+ "epoch": 0.29,
271
+ "grad_norm": 0.5553289651870728,
272
+ "learning_rate": 9.982220044089859e-05,
273
+ "loss": 0.2368,
274
+ "step": 310
275
+ },
276
+ {
277
+ "epoch": 0.3,
278
+ "grad_norm": 0.8868921399116516,
279
+ "learning_rate": 9.980973490458728e-05,
280
+ "loss": 0.2509,
281
+ "step": 320
282
+ },
283
+ {
284
+ "epoch": 0.3,
285
+ "grad_norm": 0.48811423778533936,
286
+ "learning_rate": 9.979684789876965e-05,
287
+ "loss": 0.2539,
288
+ "step": 330
289
+ },
290
+ {
291
+ "epoch": 0.31,
292
+ "grad_norm": 0.669154167175293,
293
+ "learning_rate": 9.978353953249022e-05,
294
+ "loss": 0.2308,
295
+ "step": 340
296
+ },
297
+ {
298
+ "epoch": 0.32,
299
+ "grad_norm": 0.5228665471076965,
300
+ "learning_rate": 9.976980991835894e-05,
301
+ "loss": 0.2335,
302
+ "step": 350
303
+ },
304
+ {
305
+ "epoch": 0.32,
306
+ "eval_loss": 0.22817698121070862,
307
+ "eval_runtime": 36.312,
308
+ "eval_samples_per_second": 126.129,
309
+ "eval_steps_per_second": 5.26,
310
+ "step": 350
311
+ },
312
+ {
313
+ "epoch": 0.33,
314
+ "grad_norm": 0.5688114166259766,
315
+ "learning_rate": 9.975565917255016e-05,
316
+ "loss": 0.2046,
317
+ "step": 360
318
+ },
319
+ {
320
+ "epoch": 0.34,
321
+ "grad_norm": 0.8116939663887024,
322
+ "learning_rate": 9.974108741480165e-05,
323
+ "loss": 0.2174,
324
+ "step": 370
325
+ },
326
+ {
327
+ "epoch": 0.35,
328
+ "grad_norm": 0.4922645390033722,
329
+ "learning_rate": 9.972609476841367e-05,
330
+ "loss": 0.2346,
331
+ "step": 380
332
+ },
333
+ {
334
+ "epoch": 0.36,
335
+ "grad_norm": 0.8515651226043701,
336
+ "learning_rate": 9.971068136024781e-05,
337
+ "loss": 0.2291,
338
+ "step": 390
339
+ },
340
+ {
341
+ "epoch": 0.37,
342
+ "grad_norm": 0.5515058636665344,
343
+ "learning_rate": 9.9694847320726e-05,
344
+ "loss": 0.2226,
345
+ "step": 400
346
+ },
347
+ {
348
+ "epoch": 0.37,
349
+ "eval_loss": 0.22269846498966217,
350
+ "eval_runtime": 36.2361,
351
+ "eval_samples_per_second": 126.393,
352
+ "eval_steps_per_second": 5.271,
353
+ "step": 400
354
+ },
355
+ {
356
+ "epoch": 0.38,
357
+ "grad_norm": 0.4810548424720764,
358
+ "learning_rate": 9.967859278382938e-05,
359
+ "loss": 0.2235,
360
+ "step": 410
361
+ },
362
+ {
363
+ "epoch": 0.39,
364
+ "grad_norm": 0.74261873960495,
365
+ "learning_rate": 9.966191788709716e-05,
366
+ "loss": 0.2308,
367
+ "step": 420
368
+ },
369
+ {
370
+ "epoch": 0.4,
371
+ "grad_norm": 0.8025543689727783,
372
+ "learning_rate": 9.964482277162546e-05,
373
+ "loss": 0.193,
374
+ "step": 430
375
+ },
376
+ {
377
+ "epoch": 0.41,
378
+ "grad_norm": 0.5808039903640747,
379
+ "learning_rate": 9.962730758206611e-05,
380
+ "loss": 0.2144,
381
+ "step": 440
382
+ },
383
+ {
384
+ "epoch": 0.42,
385
+ "grad_norm": 0.5638976097106934,
386
+ "learning_rate": 9.960937246662546e-05,
387
+ "loss": 0.2098,
388
+ "step": 450
389
+ },
390
+ {
391
+ "epoch": 0.42,
392
+ "eval_loss": 0.2109217643737793,
393
+ "eval_runtime": 36.2318,
394
+ "eval_samples_per_second": 126.408,
395
+ "eval_steps_per_second": 5.272,
396
+ "step": 450
397
+ },
398
+ {
399
+ "epoch": 0.43,
400
+ "grad_norm": 1.0780441761016846,
401
+ "learning_rate": 9.959101757706308e-05,
402
+ "loss": 0.2355,
403
+ "step": 460
404
+ },
405
+ {
406
+ "epoch": 0.43,
407
+ "grad_norm": 0.8211221098899841,
408
+ "learning_rate": 9.957224306869053e-05,
409
+ "loss": 0.2108,
410
+ "step": 470
411
+ },
412
+ {
413
+ "epoch": 0.44,
414
+ "grad_norm": 0.8444514274597168,
415
+ "learning_rate": 9.955304910036993e-05,
416
+ "loss": 0.2295,
417
+ "step": 480
418
+ },
419
+ {
420
+ "epoch": 0.45,
421
+ "grad_norm": 0.47663894295692444,
422
+ "learning_rate": 9.95334358345128e-05,
423
+ "loss": 0.2129,
424
+ "step": 490
425
+ },
426
+ {
427
+ "epoch": 0.46,
428
+ "grad_norm": 1.289258360862732,
429
+ "learning_rate": 9.951340343707852e-05,
430
+ "loss": 0.1839,
431
+ "step": 500
432
+ },
433
+ {
434
+ "epoch": 0.46,
435
+ "eval_loss": 0.20476657152175903,
436
+ "eval_runtime": 36.255,
437
+ "eval_samples_per_second": 126.328,
438
+ "eval_steps_per_second": 5.268,
439
+ "step": 500
440
+ },
441
+ {
442
+ "epoch": 0.47,
443
+ "grad_norm": 2.3407182693481445,
444
+ "learning_rate": 9.949295207757299e-05,
445
+ "loss": 0.1975,
446
+ "step": 510
447
+ },
448
+ {
449
+ "epoch": 0.48,
450
+ "grad_norm": 0.8705734014511108,
451
+ "learning_rate": 9.947208192904722e-05,
452
+ "loss": 0.2055,
453
+ "step": 520
454
+ },
455
+ {
456
+ "epoch": 0.49,
457
+ "grad_norm": 0.8187156915664673,
458
+ "learning_rate": 9.945079316809585e-05,
459
+ "loss": 0.242,
460
+ "step": 530
461
+ },
462
+ {
463
+ "epoch": 0.5,
464
+ "grad_norm": 1.010530710220337,
465
+ "learning_rate": 9.942908597485558e-05,
466
+ "loss": 0.1907,
467
+ "step": 540
468
+ },
469
+ {
470
+ "epoch": 0.51,
471
+ "grad_norm": 0.5775702595710754,
472
+ "learning_rate": 9.94069605330038e-05,
473
+ "loss": 0.2008,
474
+ "step": 550
475
+ },
476
+ {
477
+ "epoch": 0.51,
478
+ "eval_loss": 0.19918882846832275,
479
+ "eval_runtime": 36.2643,
480
+ "eval_samples_per_second": 126.295,
481
+ "eval_steps_per_second": 5.267,
482
+ "step": 550
483
+ },
484
+ {
485
+ "epoch": 0.52,
486
+ "grad_norm": 0.962810218334198,
487
+ "learning_rate": 9.938441702975689e-05,
488
+ "loss": 0.189,
489
+ "step": 560
490
+ },
491
+ {
492
+ "epoch": 0.53,
493
+ "grad_norm": 0.5549082159996033,
494
+ "learning_rate": 9.936145565586871e-05,
495
+ "loss": 0.2045,
496
+ "step": 570
497
+ },
498
+ {
499
+ "epoch": 0.54,
500
+ "grad_norm": 0.6081735491752625,
501
+ "learning_rate": 9.933807660562898e-05,
502
+ "loss": 0.196,
503
+ "step": 580
504
+ },
505
+ {
506
+ "epoch": 0.55,
507
+ "grad_norm": 0.987080991268158,
508
+ "learning_rate": 9.931428007686158e-05,
509
+ "loss": 0.2054,
510
+ "step": 590
511
+ },
512
+ {
513
+ "epoch": 0.55,
514
+ "grad_norm": 0.7649418711662292,
515
+ "learning_rate": 9.929006627092299e-05,
516
+ "loss": 0.1945,
517
+ "step": 600
518
+ },
519
+ {
520
+ "epoch": 0.55,
521
+ "eval_loss": 0.20194485783576965,
522
+ "eval_runtime": 36.2524,
523
+ "eval_samples_per_second": 126.336,
524
+ "eval_steps_per_second": 5.269,
525
+ "step": 600
526
+ },
527
+ {
528
+ "epoch": 0.56,
529
+ "grad_norm": 0.547579824924469,
530
+ "learning_rate": 9.926543539270048e-05,
531
+ "loss": 0.2278,
532
+ "step": 610
533
+ },
534
+ {
535
+ "epoch": 0.57,
536
+ "grad_norm": 0.6617013812065125,
537
+ "learning_rate": 9.924038765061042e-05,
538
+ "loss": 0.1894,
539
+ "step": 620
540
+ },
541
+ {
542
+ "epoch": 0.58,
543
+ "grad_norm": 0.9552827477455139,
544
+ "learning_rate": 9.921492325659649e-05,
545
+ "loss": 0.1868,
546
+ "step": 630
547
+ },
548
+ {
549
+ "epoch": 0.59,
550
+ "grad_norm": 0.4771063029766083,
551
+ "learning_rate": 9.918904242612795e-05,
552
+ "loss": 0.1927,
553
+ "step": 640
554
+ },
555
+ {
556
+ "epoch": 0.6,
557
+ "grad_norm": 0.7816069722175598,
558
+ "learning_rate": 9.916274537819775e-05,
559
+ "loss": 0.1891,
560
+ "step": 650
561
+ },
562
+ {
563
+ "epoch": 0.6,
564
+ "eval_loss": 0.1859094649553299,
565
+ "eval_runtime": 36.2607,
566
+ "eval_samples_per_second": 126.307,
567
+ "eval_steps_per_second": 5.267,
568
+ "step": 650
569
+ },
570
+ {
571
+ "epoch": 0.61,
572
+ "grad_norm": 1.2409390211105347,
573
+ "learning_rate": 9.913603233532067e-05,
574
+ "loss": 0.1772,
575
+ "step": 660
576
+ },
577
+ {
578
+ "epoch": 0.62,
579
+ "grad_norm": 0.4985053837299347,
580
+ "learning_rate": 9.910890352353153e-05,
581
+ "loss": 0.1787,
582
+ "step": 670
583
+ },
584
+ {
585
+ "epoch": 0.63,
586
+ "grad_norm": 0.6402477025985718,
587
+ "learning_rate": 9.908135917238321e-05,
588
+ "loss": 0.192,
589
+ "step": 680
590
+ },
591
+ {
592
+ "epoch": 0.64,
593
+ "grad_norm": 0.7117078304290771,
594
+ "learning_rate": 9.905339951494463e-05,
595
+ "loss": 0.1784,
596
+ "step": 690
597
+ },
598
+ {
599
+ "epoch": 0.65,
600
+ "grad_norm": 0.7569687962532043,
601
+ "learning_rate": 9.902502478779896e-05,
602
+ "loss": 0.2015,
603
+ "step": 700
604
+ },
605
+ {
606
+ "epoch": 0.65,
607
+ "eval_loss": 0.19662824273109436,
608
+ "eval_runtime": 36.2532,
609
+ "eval_samples_per_second": 126.334,
610
+ "eval_steps_per_second": 5.268,
611
+ "step": 700
612
+ },
613
+ {
614
+ "epoch": 0.66,
615
+ "grad_norm": 0.6098802089691162,
616
+ "learning_rate": 9.899623523104149e-05,
617
+ "loss": 0.188,
618
+ "step": 710
619
+ },
620
+ {
621
+ "epoch": 0.67,
622
+ "grad_norm": 0.8657363057136536,
623
+ "learning_rate": 9.896703108827759e-05,
624
+ "loss": 0.172,
625
+ "step": 720
626
+ },
627
+ {
628
+ "epoch": 0.67,
629
+ "grad_norm": 0.5754011869430542,
630
+ "learning_rate": 9.893741260662071e-05,
631
+ "loss": 0.1933,
632
+ "step": 730
633
+ },
634
+ {
635
+ "epoch": 0.68,
636
+ "grad_norm": 1.5085391998291016,
637
+ "learning_rate": 9.890738003669029e-05,
638
+ "loss": 0.1942,
639
+ "step": 740
640
+ },
641
+ {
642
+ "epoch": 0.69,
643
+ "grad_norm": 0.7169360518455505,
644
+ "learning_rate": 9.887693363260958e-05,
645
+ "loss": 0.174,
646
+ "step": 750
647
+ },
648
+ {
649
+ "epoch": 0.69,
650
+ "eval_loss": 0.18010230362415314,
651
+ "eval_runtime": 36.2672,
652
+ "eval_samples_per_second": 126.285,
653
+ "eval_steps_per_second": 5.266,
654
+ "step": 750
655
+ },
656
+ {
657
+ "epoch": 0.7,
658
+ "grad_norm": 0.8828125596046448,
659
+ "learning_rate": 9.884607365200356e-05,
660
+ "loss": 0.1806,
661
+ "step": 760
662
+ },
663
+ {
664
+ "epoch": 0.71,
665
+ "grad_norm": 0.7980941534042358,
666
+ "learning_rate": 9.881480035599667e-05,
667
+ "loss": 0.1784,
668
+ "step": 770
669
+ },
670
+ {
671
+ "epoch": 0.72,
672
+ "grad_norm": 0.7319965958595276,
673
+ "learning_rate": 9.878311400921072e-05,
674
+ "loss": 0.1707,
675
+ "step": 780
676
+ },
677
+ {
678
+ "epoch": 0.73,
679
+ "grad_norm": 0.8168091773986816,
680
+ "learning_rate": 9.875101487976253e-05,
681
+ "loss": 0.1732,
682
+ "step": 790
683
+ },
684
+ {
685
+ "epoch": 0.74,
686
+ "grad_norm": 1.2269237041473389,
687
+ "learning_rate": 9.871850323926177e-05,
688
+ "loss": 0.1565,
689
+ "step": 800
690
+ },
691
+ {
692
+ "epoch": 0.74,
693
+ "eval_loss": 0.17616184055805206,
694
+ "eval_runtime": 36.2739,
695
+ "eval_samples_per_second": 126.262,
696
+ "eval_steps_per_second": 5.265,
697
+ "step": 800
698
+ },
699
+ {
700
+ "epoch": 0.75,
701
+ "grad_norm": 0.5521104335784912,
702
+ "learning_rate": 9.868557936280855e-05,
703
+ "loss": 0.1855,
704
+ "step": 810
705
+ },
706
+ {
707
+ "epoch": 0.76,
708
+ "grad_norm": 0.6372039914131165,
709
+ "learning_rate": 9.865224352899119e-05,
710
+ "loss": 0.1763,
711
+ "step": 820
712
+ },
713
+ {
714
+ "epoch": 0.77,
715
+ "grad_norm": 0.9996755123138428,
716
+ "learning_rate": 9.861849601988383e-05,
717
+ "loss": 0.149,
718
+ "step": 830
719
+ },
720
+ {
721
+ "epoch": 0.78,
722
+ "grad_norm": 1.2396740913391113,
723
+ "learning_rate": 9.858433712104403e-05,
724
+ "loss": 0.1773,
725
+ "step": 840
726
+ },
727
+ {
728
+ "epoch": 0.79,
729
+ "grad_norm": 0.946553647518158,
730
+ "learning_rate": 9.854976712151031e-05,
731
+ "loss": 0.1825,
732
+ "step": 850
733
+ },
734
+ {
735
+ "epoch": 0.79,
736
+ "eval_loss": 0.1716548502445221,
737
+ "eval_runtime": 36.2484,
738
+ "eval_samples_per_second": 126.35,
739
+ "eval_steps_per_second": 5.269,
740
+ "step": 850
741
+ },
742
+ {
743
+ "epoch": 0.79,
744
+ "grad_norm": 1.1854327917099,
745
+ "learning_rate": 9.851478631379982e-05,
746
+ "loss": 0.1608,
747
+ "step": 860
748
+ },
749
+ {
750
+ "epoch": 0.8,
751
+ "grad_norm": 0.6903740763664246,
752
+ "learning_rate": 9.84793949939058e-05,
753
+ "loss": 0.1723,
754
+ "step": 870
755
+ },
756
+ {
757
+ "epoch": 0.81,
758
+ "grad_norm": 0.929786741733551,
759
+ "learning_rate": 9.844359346129503e-05,
760
+ "loss": 0.1566,
761
+ "step": 880
762
+ },
763
+ {
764
+ "epoch": 0.82,
765
+ "grad_norm": 0.43247154355049133,
766
+ "learning_rate": 9.84073820189054e-05,
767
+ "loss": 0.1769,
768
+ "step": 890
769
+ },
770
+ {
771
+ "epoch": 0.83,
772
+ "grad_norm": 0.8210358619689941,
773
+ "learning_rate": 9.837076097314319e-05,
774
+ "loss": 0.1651,
775
+ "step": 900
776
+ },
777
+ {
778
+ "epoch": 0.83,
779
+ "eval_loss": 0.16817261278629303,
780
+ "eval_runtime": 36.2399,
781
+ "eval_samples_per_second": 126.38,
782
+ "eval_steps_per_second": 5.27,
783
+ "step": 900
784
+ },
785
+ {
786
+ "epoch": 0.84,
787
+ "grad_norm": 0.9283536672592163,
788
+ "learning_rate": 9.83337306338807e-05,
789
+ "loss": 0.1472,
790
+ "step": 910
791
+ },
792
+ {
793
+ "epoch": 0.85,
794
+ "grad_norm": 0.7863032817840576,
795
+ "learning_rate": 9.829629131445342e-05,
796
+ "loss": 0.1683,
797
+ "step": 920
798
+ },
799
+ {
800
+ "epoch": 0.86,
801
+ "grad_norm": 0.5381516218185425,
802
+ "learning_rate": 9.825844333165748e-05,
803
+ "loss": 0.151,
804
+ "step": 930
805
+ },
806
+ {
807
+ "epoch": 0.87,
808
+ "grad_norm": 0.7078680396080017,
809
+ "learning_rate": 9.822018700574695e-05,
810
+ "loss": 0.1641,
811
+ "step": 940
812
+ },
813
+ {
814
+ "epoch": 0.88,
815
+ "grad_norm": 0.8207384943962097,
816
+ "learning_rate": 9.818152266043114e-05,
817
+ "loss": 0.1598,
818
+ "step": 950
819
+ },
820
+ {
821
+ "epoch": 0.88,
822
+ "eval_loss": 0.15981163084506989,
823
+ "eval_runtime": 36.2611,
824
+ "eval_samples_per_second": 126.306,
825
+ "eval_steps_per_second": 5.267,
826
+ "step": 950
827
+ },
828
+ {
829
+ "epoch": 0.89,
830
+ "grad_norm": 1.319566011428833,
831
+ "learning_rate": 9.814245062287189e-05,
832
+ "loss": 0.156,
833
+ "step": 960
834
+ },
835
+ {
836
+ "epoch": 0.9,
837
+ "grad_norm": 0.5811505317687988,
838
+ "learning_rate": 9.810297122368067e-05,
839
+ "loss": 0.1646,
840
+ "step": 970
841
+ },
842
+ {
843
+ "epoch": 0.91,
844
+ "grad_norm": 0.7206919193267822,
845
+ "learning_rate": 9.806308479691595e-05,
846
+ "loss": 0.1727,
847
+ "step": 980
848
+ },
849
+ {
850
+ "epoch": 0.91,
851
+ "grad_norm": 0.9073074460029602,
852
+ "learning_rate": 9.802279168008029e-05,
853
+ "loss": 0.1566,
854
+ "step": 990
855
+ },
856
+ {
857
+ "epoch": 0.92,
858
+ "grad_norm": 0.9229528903961182,
859
+ "learning_rate": 9.798209221411747e-05,
860
+ "loss": 0.1502,
861
+ "step": 1000
862
+ },
863
+ {
864
+ "epoch": 0.92,
865
+ "eval_loss": 0.15576273202896118,
866
+ "eval_runtime": 36.2741,
867
+ "eval_samples_per_second": 126.261,
868
+ "eval_steps_per_second": 5.265,
869
+ "step": 1000
870
+ },
871
+ {
872
+ "epoch": 0.93,
873
+ "grad_norm": 0.9960848093032837,
874
+ "learning_rate": 9.794098674340965e-05,
875
+ "loss": 0.1634,
876
+ "step": 1010
877
+ },
878
+ {
879
+ "epoch": 0.94,
880
+ "grad_norm": 0.5757964253425598,
881
+ "learning_rate": 9.789947561577445e-05,
882
+ "loss": 0.1438,
883
+ "step": 1020
884
+ },
885
+ {
886
+ "epoch": 0.95,
887
+ "grad_norm": 0.7039881348609924,
888
+ "learning_rate": 9.785755918246194e-05,
889
+ "loss": 0.1297,
890
+ "step": 1030
891
+ },
892
+ {
893
+ "epoch": 0.96,
894
+ "grad_norm": 1.228461742401123,
895
+ "learning_rate": 9.781523779815179e-05,
896
+ "loss": 0.1525,
897
+ "step": 1040
898
+ },
899
+ {
900
+ "epoch": 0.97,
901
+ "grad_norm": 0.8237024545669556,
902
+ "learning_rate": 9.777251182095011e-05,
903
+ "loss": 0.1599,
904
+ "step": 1050
905
+ },
906
+ {
907
+ "epoch": 0.97,
908
+ "eval_loss": 0.1465296447277069,
909
+ "eval_runtime": 36.3321,
910
+ "eval_samples_per_second": 126.059,
911
+ "eval_steps_per_second": 5.257,
912
+ "step": 1050
913
+ },
914
+ {
915
+ "epoch": 0.98,
916
+ "grad_norm": 0.5836882591247559,
917
+ "learning_rate": 9.77293816123866e-05,
918
+ "loss": 0.1601,
919
+ "step": 1060
920
+ },
921
+ {
922
+ "epoch": 0.99,
923
+ "grad_norm": 0.9502392411231995,
924
+ "learning_rate": 9.768584753741134e-05,
925
+ "loss": 0.159,
926
+ "step": 1070
927
+ },
928
+ {
929
+ "epoch": 1.0,
930
+ "grad_norm": 0.8163331151008606,
931
+ "learning_rate": 9.764190996439181e-05,
932
+ "loss": 0.1444,
933
+ "step": 1080
934
+ },
935
+ {
936
+ "epoch": 1.01,
937
+ "grad_norm": 1.476738452911377,
938
+ "learning_rate": 9.759756926510965e-05,
939
+ "loss": 0.1205,
940
+ "step": 1090
941
+ },
942
+ {
943
+ "epoch": 1.02,
944
+ "grad_norm": 1.338589072227478,
945
+ "learning_rate": 9.755282581475769e-05,
946
+ "loss": 0.0977,
947
+ "step": 1100
948
+ },
949
+ {
950
+ "epoch": 1.02,
951
+ "eval_loss": 0.15202057361602783,
952
+ "eval_runtime": 36.2683,
953
+ "eval_samples_per_second": 126.281,
954
+ "eval_steps_per_second": 5.266,
955
+ "step": 1100
956
+ },
957
+ {
958
+ "epoch": 1.03,
959
+ "grad_norm": 1.2897748947143555,
960
+ "learning_rate": 9.750767999193656e-05,
961
+ "loss": 0.1106,
962
+ "step": 1110
963
+ },
964
+ {
965
+ "epoch": 1.04,
966
+ "grad_norm": 0.640981912612915,
967
+ "learning_rate": 9.74621321786517e-05,
968
+ "loss": 0.0907,
969
+ "step": 1120
970
+ },
971
+ {
972
+ "epoch": 1.04,
973
+ "grad_norm": 1.5198509693145752,
974
+ "learning_rate": 9.741618276030997e-05,
975
+ "loss": 0.0832,
976
+ "step": 1130
977
+ },
978
+ {
979
+ "epoch": 1.05,
980
+ "grad_norm": 0.8561617136001587,
981
+ "learning_rate": 9.736983212571646e-05,
982
+ "loss": 0.1197,
983
+ "step": 1140
984
+ },
985
+ {
986
+ "epoch": 1.06,
987
+ "grad_norm": 0.8811647295951843,
988
+ "learning_rate": 9.732308066707122e-05,
989
+ "loss": 0.1166,
990
+ "step": 1150
991
+ },
992
+ {
993
+ "epoch": 1.06,
994
+ "eval_loss": 0.1403137743473053,
995
+ "eval_runtime": 36.2404,
996
+ "eval_samples_per_second": 126.378,
997
+ "eval_steps_per_second": 5.27,
998
+ "step": 1150
999
+ },
1000
+ {
1001
+ "epoch": 1.07,
1002
+ "grad_norm": 1.1139312982559204,
1003
+ "learning_rate": 9.727592877996585e-05,
1004
+ "loss": 0.1138,
1005
+ "step": 1160
1006
+ },
1007
+ {
1008
+ "epoch": 1.08,
1009
+ "grad_norm": 0.8331150412559509,
1010
+ "learning_rate": 9.722837686338025e-05,
1011
+ "loss": 0.1159,
1012
+ "step": 1170
1013
+ },
1014
+ {
1015
+ "epoch": 1.09,
1016
+ "grad_norm": 0.8520072102546692,
1017
+ "learning_rate": 9.718042531967918e-05,
1018
+ "loss": 0.0983,
1019
+ "step": 1180
1020
+ },
1021
+ {
1022
+ "epoch": 1.1,
1023
+ "grad_norm": 0.5743900537490845,
1024
+ "learning_rate": 9.713207455460894e-05,
1025
+ "loss": 0.0913,
1026
+ "step": 1190
1027
+ },
1028
+ {
1029
+ "epoch": 1.11,
1030
+ "grad_norm": 1.1309562921524048,
1031
+ "learning_rate": 9.708332497729378e-05,
1032
+ "loss": 0.0943,
1033
+ "step": 1200
1034
+ },
1035
+ {
1036
+ "epoch": 1.11,
1037
+ "eval_loss": 0.13866838812828064,
1038
+ "eval_runtime": 36.2659,
1039
+ "eval_samples_per_second": 126.289,
1040
+ "eval_steps_per_second": 5.267,
1041
+ "step": 1200
1042
+ },
1043
+ {
1044
+ "epoch": 1.12,
1045
+ "grad_norm": 0.8129335045814514,
1046
+ "learning_rate": 9.70341770002326e-05,
1047
+ "loss": 0.0999,
1048
+ "step": 1210
1049
+ },
1050
+ {
1051
+ "epoch": 1.13,
1052
+ "grad_norm": 1.7836413383483887,
1053
+ "learning_rate": 9.698463103929542e-05,
1054
+ "loss": 0.0914,
1055
+ "step": 1220
1056
+ },
1057
+ {
1058
+ "epoch": 1.14,
1059
+ "grad_norm": 1.3467488288879395,
1060
+ "learning_rate": 9.693468751371977e-05,
1061
+ "loss": 0.0938,
1062
+ "step": 1230
1063
+ },
1064
+ {
1065
+ "epoch": 1.15,
1066
+ "grad_norm": 0.7247940897941589,
1067
+ "learning_rate": 9.688434684610726e-05,
1068
+ "loss": 0.1101,
1069
+ "step": 1240
1070
+ },
1071
+ {
1072
+ "epoch": 1.16,
1073
+ "grad_norm": 0.9462648034095764,
1074
+ "learning_rate": 9.683360946241989e-05,
1075
+ "loss": 0.1007,
1076
+ "step": 1250
1077
+ },
1078
+ {
1079
+ "epoch": 1.16,
1080
+ "eval_loss": 0.13109833002090454,
1081
+ "eval_runtime": 36.3278,
1082
+ "eval_samples_per_second": 126.074,
1083
+ "eval_steps_per_second": 5.258,
1084
+ "step": 1250
1085
+ },
1086
+ {
1087
+ "epoch": 1.16,
1088
+ "grad_norm": 0.9187774062156677,
1089
+ "learning_rate": 9.678247579197657e-05,
1090
+ "loss": 0.1057,
1091
+ "step": 1260
1092
+ },
1093
+ {
1094
+ "epoch": 1.17,
1095
+ "grad_norm": 1.120918869972229,
1096
+ "learning_rate": 9.673094626744942e-05,
1097
+ "loss": 0.0865,
1098
+ "step": 1270
1099
+ },
1100
+ {
1101
+ "epoch": 1.18,
1102
+ "grad_norm": 1.1514744758605957,
1103
+ "learning_rate": 9.667902132486009e-05,
1104
+ "loss": 0.1077,
1105
+ "step": 1280
1106
+ },
1107
+ {
1108
+ "epoch": 1.19,
1109
+ "grad_norm": 1.6133332252502441,
1110
+ "learning_rate": 9.662670140357611e-05,
1111
+ "loss": 0.0923,
1112
+ "step": 1290
1113
+ },
1114
+ {
1115
+ "epoch": 1.2,
1116
+ "grad_norm": 0.9612036943435669,
1117
+ "learning_rate": 9.657398694630712e-05,
1118
+ "loss": 0.1035,
1119
+ "step": 1300
1120
+ },
1121
+ {
1122
+ "epoch": 1.2,
1123
+ "eval_loss": 0.13247531652450562,
1124
+ "eval_runtime": 36.3269,
1125
+ "eval_samples_per_second": 126.077,
1126
+ "eval_steps_per_second": 5.258,
1127
+ "step": 1300
1128
+ },
1129
+ {
1130
+ "epoch": 1.21,
1131
+ "grad_norm": 2.6024136543273926,
1132
+ "learning_rate": 9.652087839910124e-05,
1133
+ "loss": 0.0803,
1134
+ "step": 1310
1135
+ },
1136
+ {
1137
+ "epoch": 1.22,
1138
+ "grad_norm": 0.8841276168823242,
1139
+ "learning_rate": 9.646737621134112e-05,
1140
+ "loss": 0.0814,
1141
+ "step": 1320
1142
+ },
1143
+ {
1144
+ "epoch": 1.23,
1145
+ "grad_norm": 0.7463160753250122,
1146
+ "learning_rate": 9.641348083574034e-05,
1147
+ "loss": 0.0976,
1148
+ "step": 1330
1149
+ },
1150
+ {
1151
+ "epoch": 1.24,
1152
+ "grad_norm": 0.6614356637001038,
1153
+ "learning_rate": 9.635919272833938e-05,
1154
+ "loss": 0.1047,
1155
+ "step": 1340
1156
+ },
1157
+ {
1158
+ "epoch": 1.25,
1159
+ "grad_norm": 0.8464844822883606,
1160
+ "learning_rate": 9.63045123485019e-05,
1161
+ "loss": 0.0842,
1162
+ "step": 1350
1163
+ },
1164
+ {
1165
+ "epoch": 1.25,
1166
+ "eval_loss": 0.13089965283870697,
1167
+ "eval_runtime": 36.2144,
1168
+ "eval_samples_per_second": 126.469,
1169
+ "eval_steps_per_second": 5.274,
1170
+ "step": 1350
1171
+ },
1172
+ {
1173
+ "epoch": 1.26,
1174
+ "grad_norm": 1.1737573146820068,
1175
+ "learning_rate": 9.62494401589108e-05,
1176
+ "loss": 0.0631,
1177
+ "step": 1360
1178
+ },
1179
+ {
1180
+ "epoch": 1.27,
1181
+ "grad_norm": 1.0628595352172852,
1182
+ "learning_rate": 9.619397662556435e-05,
1183
+ "loss": 0.0967,
1184
+ "step": 1370
1185
+ },
1186
+ {
1187
+ "epoch": 1.28,
1188
+ "grad_norm": 0.6741960644721985,
1189
+ "learning_rate": 9.613812221777212e-05,
1190
+ "loss": 0.0727,
1191
+ "step": 1380
1192
+ },
1193
+ {
1194
+ "epoch": 1.28,
1195
+ "grad_norm": 1.3467893600463867,
1196
+ "learning_rate": 9.60818774081512e-05,
1197
+ "loss": 0.0832,
1198
+ "step": 1390
1199
+ },
1200
+ {
1201
+ "epoch": 1.29,
1202
+ "grad_norm": 0.5382059216499329,
1203
+ "learning_rate": 9.602524267262203e-05,
1204
+ "loss": 0.1114,
1205
+ "step": 1400
1206
+ },
1207
+ {
1208
+ "epoch": 1.29,
1209
+ "eval_loss": 0.12248954176902771,
1210
+ "eval_runtime": 36.2887,
1211
+ "eval_samples_per_second": 126.21,
1212
+ "eval_steps_per_second": 5.263,
1213
+ "step": 1400
1214
+ },
1215
+ {
1216
+ "epoch": 1.3,
1217
+ "grad_norm": 1.2735854387283325,
1218
+ "learning_rate": 9.596821849040447e-05,
1219
+ "loss": 0.0997,
1220
+ "step": 1410
1221
+ },
1222
+ {
1223
+ "epoch": 1.31,
1224
+ "grad_norm": 0.711742103099823,
1225
+ "learning_rate": 9.591080534401371e-05,
1226
+ "loss": 0.0917,
1227
+ "step": 1420
1228
+ },
1229
+ {
1230
+ "epoch": 1.32,
1231
+ "grad_norm": 1.2457355260849,
1232
+ "learning_rate": 9.58530037192562e-05,
1233
+ "loss": 0.1039,
1234
+ "step": 1430
1235
+ },
1236
+ {
1237
+ "epoch": 1.33,
1238
+ "grad_norm": 0.5564578175544739,
1239
+ "learning_rate": 9.579481410522556e-05,
1240
+ "loss": 0.0957,
1241
+ "step": 1440
1242
+ },
1243
+ {
1244
+ "epoch": 1.34,
1245
+ "grad_norm": 1.5276597738265991,
1246
+ "learning_rate": 9.573623699429836e-05,
1247
+ "loss": 0.1047,
1248
+ "step": 1450
1249
+ },
1250
+ {
1251
+ "epoch": 1.34,
1252
+ "eval_loss": 0.11836496740579605,
1253
+ "eval_runtime": 36.2545,
1254
+ "eval_samples_per_second": 126.329,
1255
+ "eval_steps_per_second": 5.268,
1256
+ "step": 1450
1257
+ },
1258
+ {
1259
+ "epoch": 1.35,
1260
+ "grad_norm": 0.865064263343811,
1261
+ "learning_rate": 9.567727288213005e-05,
1262
+ "loss": 0.0957,
1263
+ "step": 1460
1264
+ },
1265
+ {
1266
+ "epoch": 1.36,
1267
+ "grad_norm": 0.6668936014175415,
1268
+ "learning_rate": 9.561792226765072e-05,
1269
+ "loss": 0.0973,
1270
+ "step": 1470
1271
+ },
1272
+ {
1273
+ "epoch": 1.37,
1274
+ "grad_norm": 1.2213976383209229,
1275
+ "learning_rate": 9.555818565306085e-05,
1276
+ "loss": 0.0887,
1277
+ "step": 1480
1278
+ },
1279
+ {
1280
+ "epoch": 1.38,
1281
+ "grad_norm": 0.7778081893920898,
1282
+ "learning_rate": 9.549806354382717e-05,
1283
+ "loss": 0.094,
1284
+ "step": 1490
1285
+ },
1286
+ {
1287
+ "epoch": 1.39,
1288
+ "grad_norm": 0.8432282209396362,
1289
+ "learning_rate": 9.543755644867822e-05,
1290
+ "loss": 0.0807,
1291
+ "step": 1500
1292
+ },
1293
+ {
1294
+ "epoch": 1.39,
1295
+ "eval_loss": 0.11355672031641006,
1296
+ "eval_runtime": 36.2261,
1297
+ "eval_samples_per_second": 126.428,
1298
+ "eval_steps_per_second": 5.272,
1299
+ "step": 1500
1300
+ },
1301
+ {
1302
+ "epoch": 1.4,
1303
+ "grad_norm": 0.9765787124633789,
1304
+ "learning_rate": 9.537666487960019e-05,
1305
+ "loss": 0.0829,
1306
+ "step": 1510
1307
+ },
1308
+ {
1309
+ "epoch": 1.4,
1310
+ "grad_norm": 0.8252067565917969,
1311
+ "learning_rate": 9.53153893518325e-05,
1312
+ "loss": 0.0749,
1313
+ "step": 1520
1314
+ },
1315
+ {
1316
+ "epoch": 1.41,
1317
+ "grad_norm": 1.0041134357452393,
1318
+ "learning_rate": 9.52537303838635e-05,
1319
+ "loss": 0.0647,
1320
+ "step": 1530
1321
+ },
1322
+ {
1323
+ "epoch": 1.42,
1324
+ "grad_norm": 1.365193486213684,
1325
+ "learning_rate": 9.519168849742604e-05,
1326
+ "loss": 0.0713,
1327
+ "step": 1540
1328
+ },
1329
+ {
1330
+ "epoch": 1.43,
1331
+ "grad_norm": 0.8681650161743164,
1332
+ "learning_rate": 9.512926421749304e-05,
1333
+ "loss": 0.0846,
1334
+ "step": 1550
1335
+ },
1336
+ {
1337
+ "epoch": 1.43,
1338
+ "eval_loss": 0.12000197172164917,
1339
+ "eval_runtime": 36.2616,
1340
+ "eval_samples_per_second": 126.304,
1341
+ "eval_steps_per_second": 5.267,
1342
+ "step": 1550
1343
+ },
1344
+ {
1345
+ "epoch": 1.44,
1346
+ "grad_norm": 0.7004525661468506,
1347
+ "learning_rate": 9.506645807227311e-05,
1348
+ "loss": 0.0724,
1349
+ "step": 1560
1350
+ },
1351
+ {
1352
+ "epoch": 1.45,
1353
+ "grad_norm": 1.3950064182281494,
1354
+ "learning_rate": 9.500327059320606e-05,
1355
+ "loss": 0.0851,
1356
+ "step": 1570
1357
+ },
1358
+ {
1359
+ "epoch": 1.46,
1360
+ "grad_norm": 1.4275237321853638,
1361
+ "learning_rate": 9.493970231495835e-05,
1362
+ "loss": 0.0984,
1363
+ "step": 1580
1364
+ },
1365
+ {
1366
+ "epoch": 1.47,
1367
+ "grad_norm": 0.5745067596435547,
1368
+ "learning_rate": 9.487575377541864e-05,
1369
+ "loss": 0.0584,
1370
+ "step": 1590
1371
+ },
1372
+ {
1373
+ "epoch": 1.48,
1374
+ "grad_norm": 1.4414701461791992,
1375
+ "learning_rate": 9.481142551569318e-05,
1376
+ "loss": 0.0737,
1377
+ "step": 1600
1378
+ },
1379
+ {
1380
+ "epoch": 1.48,
1381
+ "eval_loss": 0.11450409889221191,
1382
+ "eval_runtime": 36.2542,
1383
+ "eval_samples_per_second": 126.33,
1384
+ "eval_steps_per_second": 5.268,
1385
+ "step": 1600
1386
+ },
1387
+ {
1388
+ "epoch": 1.49,
1389
+ "grad_norm": 0.9620345830917358,
1390
+ "learning_rate": 9.474671808010126e-05,
1391
+ "loss": 0.0962,
1392
+ "step": 1610
1393
+ },
1394
+ {
1395
+ "epoch": 1.5,
1396
+ "grad_norm": 0.7017752528190613,
1397
+ "learning_rate": 9.468163201617062e-05,
1398
+ "loss": 0.0739,
1399
+ "step": 1620
1400
+ },
1401
+ {
1402
+ "epoch": 1.51,
1403
+ "grad_norm": 0.8450036644935608,
1404
+ "learning_rate": 9.461616787463278e-05,
1405
+ "loss": 0.089,
1406
+ "step": 1630
1407
+ },
1408
+ {
1409
+ "epoch": 1.52,
1410
+ "grad_norm": 1.1829959154129028,
1411
+ "learning_rate": 9.45503262094184e-05,
1412
+ "loss": 0.0914,
1413
+ "step": 1640
1414
+ },
1415
+ {
1416
+ "epoch": 1.52,
1417
+ "grad_norm": 1.1722110509872437,
1418
+ "learning_rate": 9.44841075776526e-05,
1419
+ "loss": 0.0844,
1420
+ "step": 1650
1421
+ },
1422
+ {
1423
+ "epoch": 1.52,
1424
+ "eval_loss": 0.1037188321352005,
1425
+ "eval_runtime": 36.2499,
1426
+ "eval_samples_per_second": 126.345,
1427
+ "eval_steps_per_second": 5.269,
1428
+ "step": 1650
1429
+ },
1430
+ {
1431
+ "epoch": 1.53,
1432
+ "grad_norm": 1.0115782022476196,
1433
+ "learning_rate": 9.441751253965021e-05,
1434
+ "loss": 0.0968,
1435
+ "step": 1660
1436
+ },
1437
+ {
1438
+ "epoch": 1.54,
1439
+ "grad_norm": 0.6493613123893738,
1440
+ "learning_rate": 9.435054165891109e-05,
1441
+ "loss": 0.097,
1442
+ "step": 1670
1443
+ },
1444
+ {
1445
+ "epoch": 1.55,
1446
+ "grad_norm": 0.8181868195533752,
1447
+ "learning_rate": 9.42831955021153e-05,
1448
+ "loss": 0.0699,
1449
+ "step": 1680
1450
+ },
1451
+ {
1452
+ "epoch": 1.56,
1453
+ "grad_norm": 1.694252610206604,
1454
+ "learning_rate": 9.421547463911835e-05,
1455
+ "loss": 0.0781,
1456
+ "step": 1690
1457
+ },
1458
+ {
1459
+ "epoch": 1.57,
1460
+ "grad_norm": 0.6897493600845337,
1461
+ "learning_rate": 9.414737964294636e-05,
1462
+ "loss": 0.0809,
1463
+ "step": 1700
1464
+ },
1465
+ {
1466
+ "epoch": 1.57,
1467
+ "eval_loss": 0.09402556717395782,
1468
+ "eval_runtime": 36.2393,
1469
+ "eval_samples_per_second": 126.382,
1470
+ "eval_steps_per_second": 5.271,
1471
+ "step": 1700
1472
+ },
1473
+ {
1474
+ "epoch": 1.58,
1475
+ "grad_norm": 0.8747942447662354,
1476
+ "learning_rate": 9.407891108979117e-05,
1477
+ "loss": 0.0897,
1478
+ "step": 1710
1479
+ },
1480
+ {
1481
+ "epoch": 1.59,
1482
+ "grad_norm": 1.006194829940796,
1483
+ "learning_rate": 9.401006955900556e-05,
1484
+ "loss": 0.0629,
1485
+ "step": 1720
1486
+ },
1487
+ {
1488
+ "epoch": 1.6,
1489
+ "grad_norm": 0.4138026833534241,
1490
+ "learning_rate": 9.394085563309827e-05,
1491
+ "loss": 0.0671,
1492
+ "step": 1730
1493
+ },
1494
+ {
1495
+ "epoch": 1.61,
1496
+ "grad_norm": 0.368488073348999,
1497
+ "learning_rate": 9.38712698977291e-05,
1498
+ "loss": 0.0736,
1499
+ "step": 1740
1500
+ },
1501
+ {
1502
+ "epoch": 1.62,
1503
+ "grad_norm": 1.2116315364837646,
1504
+ "learning_rate": 9.380131294170393e-05,
1505
+ "loss": 0.0718,
1506
+ "step": 1750
1507
+ },
1508
+ {
1509
+ "epoch": 1.62,
1510
+ "eval_loss": 0.093104287981987,
1511
+ "eval_runtime": 36.3088,
1512
+ "eval_samples_per_second": 126.14,
1513
+ "eval_steps_per_second": 5.26,
1514
+ "step": 1750
1515
+ },
1516
+ {
1517
+ "epoch": 1.63,
1518
+ "grad_norm": 0.5739853382110596,
1519
+ "learning_rate": 9.373098535696979e-05,
1520
+ "loss": 0.0644,
1521
+ "step": 1760
1522
+ },
1523
+ {
1524
+ "epoch": 1.64,
1525
+ "grad_norm": 1.4135103225708008,
1526
+ "learning_rate": 9.36602877386098e-05,
1527
+ "loss": 0.0798,
1528
+ "step": 1770
1529
+ },
1530
+ {
1531
+ "epoch": 1.65,
1532
+ "grad_norm": 1.0814074277877808,
1533
+ "learning_rate": 9.358922068483812e-05,
1534
+ "loss": 0.0717,
1535
+ "step": 1780
1536
+ },
1537
+ {
1538
+ "epoch": 1.65,
1539
+ "grad_norm": 1.103225827217102,
1540
+ "learning_rate": 9.351778479699499e-05,
1541
+ "loss": 0.1072,
1542
+ "step": 1790
1543
+ },
1544
+ {
1545
+ "epoch": 1.66,
1546
+ "grad_norm": 0.9320874810218811,
1547
+ "learning_rate": 9.344598067954152e-05,
1548
+ "loss": 0.0687,
1549
+ "step": 1800
1550
+ },
1551
+ {
1552
+ "epoch": 1.66,
1553
+ "eval_loss": 0.09301283955574036,
1554
+ "eval_runtime": 36.2388,
1555
+ "eval_samples_per_second": 126.384,
1556
+ "eval_steps_per_second": 5.271,
1557
+ "step": 1800
1558
+ },
1559
+ {
1560
+ "epoch": 1.67,
1561
+ "grad_norm": 0.9495887756347656,
1562
+ "learning_rate": 9.337380894005463e-05,
1563
+ "loss": 0.0611,
1564
+ "step": 1810
1565
+ },
1566
+ {
1567
+ "epoch": 1.68,
1568
+ "grad_norm": 1.7665095329284668,
1569
+ "learning_rate": 9.330127018922194e-05,
1570
+ "loss": 0.0827,
1571
+ "step": 1820
1572
+ },
1573
+ {
1574
+ "epoch": 1.69,
1575
+ "grad_norm": 1.620908260345459,
1576
+ "learning_rate": 9.322836504083654e-05,
1577
+ "loss": 0.0693,
1578
+ "step": 1830
1579
+ },
1580
+ {
1581
+ "epoch": 1.7,
1582
+ "grad_norm": 1.30322265625,
1583
+ "learning_rate": 9.315509411179182e-05,
1584
+ "loss": 0.058,
1585
+ "step": 1840
1586
+ },
1587
+ {
1588
+ "epoch": 1.71,
1589
+ "grad_norm": 1.101680874824524,
1590
+ "learning_rate": 9.308145802207629e-05,
1591
+ "loss": 0.0629,
1592
+ "step": 1850
1593
+ },
1594
+ {
1595
+ "epoch": 1.71,
1596
+ "eval_loss": 0.09685970097780228,
1597
+ "eval_runtime": 36.2607,
1598
+ "eval_samples_per_second": 126.307,
1599
+ "eval_steps_per_second": 5.267,
1600
+ "step": 1850
1601
+ },
1602
+ {
1603
+ "epoch": 1.72,
1604
+ "grad_norm": 0.38865023851394653,
1605
+ "learning_rate": 9.300745739476829e-05,
1606
+ "loss": 0.0717,
1607
+ "step": 1860
1608
+ },
1609
+ {
1610
+ "epoch": 1.73,
1611
+ "grad_norm": 0.38753578066825867,
1612
+ "learning_rate": 9.293309285603067e-05,
1613
+ "loss": 0.0679,
1614
+ "step": 1870
1615
+ },
1616
+ {
1617
+ "epoch": 1.74,
1618
+ "grad_norm": 1.5182372331619263,
1619
+ "learning_rate": 9.285836503510562e-05,
1620
+ "loss": 0.07,
1621
+ "step": 1880
1622
+ },
1623
+ {
1624
+ "epoch": 1.75,
1625
+ "grad_norm": 1.267052412033081,
1626
+ "learning_rate": 9.278327456430926e-05,
1627
+ "loss": 0.0824,
1628
+ "step": 1890
1629
+ },
1630
+ {
1631
+ "epoch": 1.76,
1632
+ "grad_norm": 1.3264389038085938,
1633
+ "learning_rate": 9.270782207902629e-05,
1634
+ "loss": 0.0852,
1635
+ "step": 1900
1636
+ },
1637
+ {
1638
+ "epoch": 1.76,
1639
+ "eval_loss": 0.0872357115149498,
1640
+ "eval_runtime": 36.2529,
1641
+ "eval_samples_per_second": 126.335,
1642
+ "eval_steps_per_second": 5.269,
1643
+ "step": 1900
1644
+ },
1645
+ {
1646
+ "epoch": 1.77,
1647
+ "grad_norm": 0.5945495367050171,
1648
+ "learning_rate": 9.263200821770461e-05,
1649
+ "loss": 0.0749,
1650
+ "step": 1910
1651
+ },
1652
+ {
1653
+ "epoch": 1.77,
1654
+ "grad_norm": 0.628018856048584,
1655
+ "learning_rate": 9.255583362184999e-05,
1656
+ "loss": 0.058,
1657
+ "step": 1920
1658
+ },
1659
+ {
1660
+ "epoch": 1.78,
1661
+ "grad_norm": 1.6209118366241455,
1662
+ "learning_rate": 9.247929893602054e-05,
1663
+ "loss": 0.07,
1664
+ "step": 1930
1665
+ },
1666
+ {
1667
+ "epoch": 1.79,
1668
+ "grad_norm": 0.5279895067214966,
1669
+ "learning_rate": 9.24024048078213e-05,
1670
+ "loss": 0.0638,
1671
+ "step": 1940
1672
+ },
1673
+ {
1674
+ "epoch": 1.8,
1675
+ "grad_norm": 0.5834373831748962,
1676
+ "learning_rate": 9.23251518878988e-05,
1677
+ "loss": 0.0622,
1678
+ "step": 1950
1679
+ },
1680
+ {
1681
+ "epoch": 1.8,
1682
+ "eval_loss": 0.08488291501998901,
1683
+ "eval_runtime": 36.2558,
1684
+ "eval_samples_per_second": 126.325,
1685
+ "eval_steps_per_second": 5.268,
1686
+ "step": 1950
1687
+ },
1688
+ {
1689
+ "epoch": 1.81,
1690
+ "grad_norm": 0.8667528033256531,
1691
+ "learning_rate": 9.224754082993552e-05,
1692
+ "loss": 0.0592,
1693
+ "step": 1960
1694
+ },
1695
+ {
1696
+ "epoch": 1.82,
1697
+ "grad_norm": 0.9160845875740051,
1698
+ "learning_rate": 9.21695722906443e-05,
1699
+ "loss": 0.0709,
1700
+ "step": 1970
1701
+ },
1702
+ {
1703
+ "epoch": 1.83,
1704
+ "grad_norm": 0.7025119662284851,
1705
+ "learning_rate": 9.209124692976287e-05,
1706
+ "loss": 0.0657,
1707
+ "step": 1980
1708
+ },
1709
+ {
1710
+ "epoch": 1.84,
1711
+ "grad_norm": 1.0089232921600342,
1712
+ "learning_rate": 9.201256541004829e-05,
1713
+ "loss": 0.0644,
1714
+ "step": 1990
1715
+ },
1716
+ {
1717
+ "epoch": 1.85,
1718
+ "grad_norm": 1.1696964502334595,
1719
+ "learning_rate": 9.193352839727121e-05,
1720
+ "loss": 0.0653,
1721
+ "step": 2000
1722
+ },
1723
+ {
1724
+ "epoch": 1.85,
1725
+ "eval_loss": 0.08305433392524719,
1726
+ "eval_runtime": 36.2595,
1727
+ "eval_samples_per_second": 126.312,
1728
+ "eval_steps_per_second": 5.268,
1729
+ "step": 2000
1730
+ },
1731
+ {
1732
+ "epoch": 1.86,
1733
+ "grad_norm": 1.57039475440979,
1734
+ "learning_rate": 9.185413656021036e-05,
1735
+ "loss": 0.0752,
1736
+ "step": 2010
1737
+ },
1738
+ {
1739
+ "epoch": 1.87,
1740
+ "grad_norm": 0.7407552599906921,
1741
+ "learning_rate": 9.177439057064683e-05,
1742
+ "loss": 0.0502,
1743
+ "step": 2020
1744
+ },
1745
+ {
1746
+ "epoch": 1.88,
1747
+ "grad_norm": 1.2944762706756592,
1748
+ "learning_rate": 9.169429110335841e-05,
1749
+ "loss": 0.0717,
1750
+ "step": 2030
1751
+ },
1752
+ {
1753
+ "epoch": 1.89,
1754
+ "grad_norm": 0.46426939964294434,
1755
+ "learning_rate": 9.16138388361139e-05,
1756
+ "loss": 0.0577,
1757
+ "step": 2040
1758
+ },
1759
+ {
1760
+ "epoch": 1.89,
1761
+ "grad_norm": 0.3882431387901306,
1762
+ "learning_rate": 9.153303444966727e-05,
1763
+ "loss": 0.0507,
1764
+ "step": 2050
1765
+ },
1766
+ {
1767
+ "epoch": 1.89,
1768
+ "eval_loss": 0.08292171359062195,
1769
+ "eval_runtime": 36.2518,
1770
+ "eval_samples_per_second": 126.339,
1771
+ "eval_steps_per_second": 5.269,
1772
+ "step": 2050
1773
+ },
1774
+ {
1775
+ "epoch": 1.9,
1776
+ "grad_norm": 0.5372634530067444,
1777
+ "learning_rate": 9.145187862775209e-05,
1778
+ "loss": 0.0632,
1779
+ "step": 2060
1780
+ },
1781
+ {
1782
+ "epoch": 1.91,
1783
+ "grad_norm": 0.5599855780601501,
1784
+ "learning_rate": 9.137037205707552e-05,
1785
+ "loss": 0.0637,
1786
+ "step": 2070
1787
+ },
1788
+ {
1789
+ "epoch": 1.92,
1790
+ "grad_norm": 0.7405185699462891,
1791
+ "learning_rate": 9.128851542731271e-05,
1792
+ "loss": 0.0462,
1793
+ "step": 2080
1794
+ },
1795
+ {
1796
+ "epoch": 1.93,
1797
+ "grad_norm": 0.6935983896255493,
1798
+ "learning_rate": 9.120630943110077e-05,
1799
+ "loss": 0.0665,
1800
+ "step": 2090
1801
+ },
1802
+ {
1803
+ "epoch": 1.94,
1804
+ "grad_norm": 0.5483436584472656,
1805
+ "learning_rate": 9.112375476403312e-05,
1806
+ "loss": 0.0518,
1807
+ "step": 2100
1808
+ },
1809
+ {
1810
+ "epoch": 1.94,
1811
+ "eval_loss": 0.07847684621810913,
1812
+ "eval_runtime": 36.2617,
1813
+ "eval_samples_per_second": 126.304,
1814
+ "eval_steps_per_second": 5.267,
1815
+ "step": 2100
1816
+ },
1817
+ {
1818
+ "epoch": 1.95,
1819
+ "grad_norm": 0.6139508485794067,
1820
+ "learning_rate": 9.104085212465336e-05,
1821
+ "loss": 0.0523,
1822
+ "step": 2110
1823
+ },
1824
+ {
1825
+ "epoch": 1.96,
1826
+ "grad_norm": 1.0022165775299072,
1827
+ "learning_rate": 9.09576022144496e-05,
1828
+ "loss": 0.0485,
1829
+ "step": 2120
1830
+ },
1831
+ {
1832
+ "epoch": 1.97,
1833
+ "grad_norm": 1.1758086681365967,
1834
+ "learning_rate": 9.087400573784833e-05,
1835
+ "loss": 0.0613,
1836
+ "step": 2130
1837
+ },
1838
+ {
1839
+ "epoch": 1.98,
1840
+ "grad_norm": 0.446209579706192,
1841
+ "learning_rate": 9.079006340220862e-05,
1842
+ "loss": 0.0652,
1843
+ "step": 2140
1844
+ },
1845
+ {
1846
+ "epoch": 1.99,
1847
+ "grad_norm": 0.8512327075004578,
1848
+ "learning_rate": 9.070577591781597e-05,
1849
+ "loss": 0.0566,
1850
+ "step": 2150
1851
+ },
1852
+ {
1853
+ "epoch": 1.99,
1854
+ "eval_loss": 0.07502878457307816,
1855
+ "eval_runtime": 36.2465,
1856
+ "eval_samples_per_second": 126.357,
1857
+ "eval_steps_per_second": 5.269,
1858
+ "step": 2150
1859
+ },
1860
+ {
1861
+ "epoch": 2.0,
1862
+ "grad_norm": 1.0540813207626343,
1863
+ "learning_rate": 9.062114399787647e-05,
1864
+ "loss": 0.0703,
1865
+ "step": 2160
1866
+ },
1867
+ {
1868
+ "epoch": 2.01,
1869
+ "grad_norm": 0.3436621129512787,
1870
+ "learning_rate": 9.053616835851062e-05,
1871
+ "loss": 0.0327,
1872
+ "step": 2170
1873
+ },
1874
+ {
1875
+ "epoch": 2.01,
1876
+ "grad_norm": 0.8479847311973572,
1877
+ "learning_rate": 9.045084971874738e-05,
1878
+ "loss": 0.0259,
1879
+ "step": 2180
1880
+ },
1881
+ {
1882
+ "epoch": 2.02,
1883
+ "grad_norm": 1.2036502361297607,
1884
+ "learning_rate": 9.0365188800518e-05,
1885
+ "loss": 0.0167,
1886
+ "step": 2190
1887
+ },
1888
+ {
1889
+ "epoch": 2.03,
1890
+ "grad_norm": 0.8509168028831482,
1891
+ "learning_rate": 9.027918632864997e-05,
1892
+ "loss": 0.0193,
1893
+ "step": 2200
1894
+ },
1895
+ {
1896
+ "epoch": 2.03,
1897
+ "eval_loss": 0.08374720066785812,
1898
+ "eval_runtime": 36.2475,
1899
+ "eval_samples_per_second": 126.354,
1900
+ "eval_steps_per_second": 5.269,
1901
+ "step": 2200
1902
+ },
1903
+ {
1904
+ "epoch": 2.04,
1905
+ "grad_norm": 0.3606647253036499,
1906
+ "learning_rate": 9.019284303086087e-05,
1907
+ "loss": 0.0206,
1908
+ "step": 2210
1909
+ },
1910
+ {
1911
+ "epoch": 2.05,
1912
+ "grad_norm": 0.7698445320129395,
1913
+ "learning_rate": 9.01061596377522e-05,
1914
+ "loss": 0.0235,
1915
+ "step": 2220
1916
+ },
1917
+ {
1918
+ "epoch": 2.06,
1919
+ "grad_norm": 1.9076602458953857,
1920
+ "learning_rate": 9.00191368828032e-05,
1921
+ "loss": 0.0177,
1922
+ "step": 2230
1923
+ },
1924
+ {
1925
+ "epoch": 2.07,
1926
+ "grad_norm": 0.8590715527534485,
1927
+ "learning_rate": 8.993177550236464e-05,
1928
+ "loss": 0.0274,
1929
+ "step": 2240
1930
+ },
1931
+ {
1932
+ "epoch": 2.08,
1933
+ "grad_norm": 0.1819429099559784,
1934
+ "learning_rate": 8.984407623565267e-05,
1935
+ "loss": 0.0233,
1936
+ "step": 2250
1937
+ },
1938
+ {
1939
+ "epoch": 2.08,
1940
+ "eval_loss": 0.07661169022321701,
1941
+ "eval_runtime": 36.2416,
1942
+ "eval_samples_per_second": 126.374,
1943
+ "eval_steps_per_second": 5.27,
1944
+ "step": 2250
1945
+ },
1946
+ {
1947
+ "epoch": 2.09,
1948
+ "grad_norm": 0.24359861016273499,
1949
+ "learning_rate": 8.97560398247424e-05,
1950
+ "loss": 0.016,
1951
+ "step": 2260
1952
+ },
1953
+ {
1954
+ "epoch": 2.1,
1955
+ "grad_norm": 0.6506161093711853,
1956
+ "learning_rate": 8.966766701456177e-05,
1957
+ "loss": 0.0123,
1958
+ "step": 2270
1959
+ },
1960
+ {
1961
+ "epoch": 2.11,
1962
+ "grad_norm": 0.2782149314880371,
1963
+ "learning_rate": 8.957895855288517e-05,
1964
+ "loss": 0.0235,
1965
+ "step": 2280
1966
+ },
1967
+ {
1968
+ "epoch": 2.12,
1969
+ "grad_norm": 1.4327374696731567,
1970
+ "learning_rate": 8.948991519032716e-05,
1971
+ "loss": 0.0209,
1972
+ "step": 2290
1973
+ },
1974
+ {
1975
+ "epoch": 2.13,
1976
+ "grad_norm": 0.47919216752052307,
1977
+ "learning_rate": 8.940053768033609e-05,
1978
+ "loss": 0.0249,
1979
+ "step": 2300
1980
+ },
1981
+ {
1982
+ "epoch": 2.13,
1983
+ "eval_loss": 0.08285840600728989,
1984
+ "eval_runtime": 36.2486,
1985
+ "eval_samples_per_second": 126.35,
1986
+ "eval_steps_per_second": 5.269,
1987
+ "step": 2300
1988
+ },
1989
+ {
1990
+ "epoch": 2.13,
1991
+ "grad_norm": 0.22004464268684387,
1992
+ "learning_rate": 8.931082677918771e-05,
1993
+ "loss": 0.0164,
1994
+ "step": 2310
1995
+ },
1996
+ {
1997
+ "epoch": 2.14,
1998
+ "grad_norm": 0.05589359253644943,
1999
+ "learning_rate": 8.922078324597879e-05,
2000
+ "loss": 0.0253,
2001
+ "step": 2320
2002
+ },
2003
+ {
2004
+ "epoch": 2.15,
2005
+ "grad_norm": 0.48298755288124084,
2006
+ "learning_rate": 8.91304078426207e-05,
2007
+ "loss": 0.0266,
2008
+ "step": 2330
2009
+ },
2010
+ {
2011
+ "epoch": 2.16,
2012
+ "grad_norm": 0.6790844202041626,
2013
+ "learning_rate": 8.903970133383297e-05,
2014
+ "loss": 0.0191,
2015
+ "step": 2340
2016
+ },
2017
+ {
2018
+ "epoch": 2.17,
2019
+ "grad_norm": 1.900855302810669,
2020
+ "learning_rate": 8.894866448713679e-05,
2021
+ "loss": 0.0217,
2022
+ "step": 2350
2023
+ },
2024
+ {
2025
+ "epoch": 2.17,
2026
+ "eval_loss": 0.0823688879609108,
2027
+ "eval_runtime": 36.277,
2028
+ "eval_samples_per_second": 126.251,
2029
+ "eval_steps_per_second": 5.265,
2030
+ "step": 2350
2031
+ },
2032
+ {
2033
+ "epoch": 2.18,
2034
+ "grad_norm": 0.5196428298950195,
2035
+ "learning_rate": 8.885729807284856e-05,
2036
+ "loss": 0.0408,
2037
+ "step": 2360
2038
+ },
2039
+ {
2040
+ "epoch": 2.19,
2041
+ "grad_norm": 1.6319361925125122,
2042
+ "learning_rate": 8.87656028640733e-05,
2043
+ "loss": 0.0335,
2044
+ "step": 2370
2045
+ },
2046
+ {
2047
+ "epoch": 2.2,
2048
+ "grad_norm": 1.4224486351013184,
2049
+ "learning_rate": 8.86735796366982e-05,
2050
+ "loss": 0.0189,
2051
+ "step": 2380
2052
+ },
2053
+ {
2054
+ "epoch": 2.21,
2055
+ "grad_norm": 0.8939641118049622,
2056
+ "learning_rate": 8.858122916938601e-05,
2057
+ "loss": 0.0255,
2058
+ "step": 2390
2059
+ },
2060
+ {
2061
+ "epoch": 2.22,
2062
+ "grad_norm": 0.9012664556503296,
2063
+ "learning_rate": 8.848855224356839e-05,
2064
+ "loss": 0.0233,
2065
+ "step": 2400
2066
+ },
2067
+ {
2068
+ "epoch": 2.22,
2069
+ "eval_loss": 0.07353760302066803,
2070
+ "eval_runtime": 36.2483,
2071
+ "eval_samples_per_second": 126.351,
2072
+ "eval_steps_per_second": 5.269,
2073
+ "step": 2400
2074
+ },
2075
+ {
2076
+ "epoch": 2.23,
2077
+ "grad_norm": 0.5575871467590332,
2078
+ "learning_rate": 8.839554964343943e-05,
2079
+ "loss": 0.0151,
2080
+ "step": 2410
2081
+ },
2082
+ {
2083
+ "epoch": 2.24,
2084
+ "grad_norm": 0.34087488055229187,
2085
+ "learning_rate": 8.83022221559489e-05,
2086
+ "loss": 0.0178,
2087
+ "step": 2420
2088
+ },
2089
+ {
2090
+ "epoch": 2.25,
2091
+ "grad_norm": 1.2612468004226685,
2092
+ "learning_rate": 8.820857057079565e-05,
2093
+ "loss": 0.0378,
2094
+ "step": 2430
2095
+ },
2096
+ {
2097
+ "epoch": 2.26,
2098
+ "grad_norm": 0.8768627047538757,
2099
+ "learning_rate": 8.811459568042091e-05,
2100
+ "loss": 0.0273,
2101
+ "step": 2440
2102
+ },
2103
+ {
2104
+ "epoch": 2.26,
2105
+ "grad_norm": 0.3403642475605011,
2106
+ "learning_rate": 8.802029828000156e-05,
2107
+ "loss": 0.0192,
2108
+ "step": 2450
2109
+ },
2110
+ {
2111
+ "epoch": 2.26,
2112
+ "eval_loss": 0.07672559469938278,
2113
+ "eval_runtime": 36.2454,
2114
+ "eval_samples_per_second": 126.361,
2115
+ "eval_steps_per_second": 5.27,
2116
+ "step": 2450
2117
+ },
2118
+ {
2119
+ "epoch": 2.27,
2120
+ "grad_norm": 1.5420982837677002,
2121
+ "learning_rate": 8.792567916744346e-05,
2122
+ "loss": 0.0218,
2123
+ "step": 2460
2124
+ },
2125
+ {
2126
+ "epoch": 2.28,
2127
+ "grad_norm": 0.20853710174560547,
2128
+ "learning_rate": 8.783073914337466e-05,
2129
+ "loss": 0.025,
2130
+ "step": 2470
2131
+ },
2132
+ {
2133
+ "epoch": 2.29,
2134
+ "grad_norm": 1.2193341255187988,
2135
+ "learning_rate": 8.773547901113862e-05,
2136
+ "loss": 0.0328,
2137
+ "step": 2480
2138
+ },
2139
+ {
2140
+ "epoch": 2.3,
2141
+ "grad_norm": 0.2289753556251526,
2142
+ "learning_rate": 8.763989957678742e-05,
2143
+ "loss": 0.0135,
2144
+ "step": 2490
2145
+ },
2146
+ {
2147
+ "epoch": 2.31,
2148
+ "grad_norm": 1.0940682888031006,
2149
+ "learning_rate": 8.754400164907497e-05,
2150
+ "loss": 0.0207,
2151
+ "step": 2500
2152
+ },
2153
+ {
2154
+ "epoch": 2.31,
2155
+ "eval_loss": 0.07941900193691254,
2156
+ "eval_runtime": 36.2618,
2157
+ "eval_samples_per_second": 126.304,
2158
+ "eval_steps_per_second": 5.267,
2159
+ "step": 2500
2160
+ },
2161
+ {
2162
+ "epoch": 2.32,
2163
+ "grad_norm": 0.9183884263038635,
2164
+ "learning_rate": 8.744778603945011e-05,
2165
+ "loss": 0.0168,
2166
+ "step": 2510
2167
+ },
2168
+ {
2169
+ "epoch": 2.33,
2170
+ "grad_norm": 0.989913821220398,
2171
+ "learning_rate": 8.73512535620498e-05,
2172
+ "loss": 0.0254,
2173
+ "step": 2520
2174
+ },
2175
+ {
2176
+ "epoch": 2.34,
2177
+ "grad_norm": 0.21799927949905396,
2178
+ "learning_rate": 8.72544050336922e-05,
2179
+ "loss": 0.0244,
2180
+ "step": 2530
2181
+ },
2182
+ {
2183
+ "epoch": 2.35,
2184
+ "grad_norm": 1.1213247776031494,
2185
+ "learning_rate": 8.715724127386972e-05,
2186
+ "loss": 0.0221,
2187
+ "step": 2540
2188
+ },
2189
+ {
2190
+ "epoch": 2.36,
2191
+ "grad_norm": 0.7733873128890991,
2192
+ "learning_rate": 8.705976310474219e-05,
2193
+ "loss": 0.0232,
2194
+ "step": 2550
2195
+ },
2196
+ {
2197
+ "epoch": 2.36,
2198
+ "eval_loss": 0.08428460359573364,
2199
+ "eval_runtime": 36.263,
2200
+ "eval_samples_per_second": 126.3,
2201
+ "eval_steps_per_second": 5.267,
2202
+ "step": 2550
2203
+ },
2204
+ {
2205
+ "epoch": 2.37,
2206
+ "grad_norm": 0.16892109811306,
2207
+ "learning_rate": 8.69619713511298e-05,
2208
+ "loss": 0.0182,
2209
+ "step": 2560
2210
+ },
2211
+ {
2212
+ "epoch": 2.38,
2213
+ "grad_norm": 0.6578019261360168,
2214
+ "learning_rate": 8.68638668405062e-05,
2215
+ "loss": 0.0134,
2216
+ "step": 2570
2217
+ },
2218
+ {
2219
+ "epoch": 2.38,
2220
+ "grad_norm": 0.4529229700565338,
2221
+ "learning_rate": 8.676545040299145e-05,
2222
+ "loss": 0.0217,
2223
+ "step": 2580
2224
+ },
2225
+ {
2226
+ "epoch": 2.39,
2227
+ "grad_norm": 1.304081678390503,
2228
+ "learning_rate": 8.666672287134494e-05,
2229
+ "loss": 0.0178,
2230
+ "step": 2590
2231
+ },
2232
+ {
2233
+ "epoch": 2.4,
2234
+ "grad_norm": 1.2332367897033691,
2235
+ "learning_rate": 8.656768508095853e-05,
2236
+ "loss": 0.0295,
2237
+ "step": 2600
2238
+ },
2239
+ {
2240
+ "epoch": 2.4,
2241
+ "eval_loss": 0.08000972867012024,
2242
+ "eval_runtime": 36.2546,
2243
+ "eval_samples_per_second": 126.329,
2244
+ "eval_steps_per_second": 5.268,
2245
+ "step": 2600
2246
+ },
2247
+ {
2248
+ "epoch": 2.41,
2249
+ "grad_norm": 1.8248018026351929,
2250
+ "learning_rate": 8.646833786984927e-05,
2251
+ "loss": 0.0257,
2252
+ "step": 2610
2253
+ },
2254
+ {
2255
+ "epoch": 2.42,
2256
+ "grad_norm": 1.0469847917556763,
2257
+ "learning_rate": 8.636868207865244e-05,
2258
+ "loss": 0.0365,
2259
+ "step": 2620
2260
+ },
2261
+ {
2262
+ "epoch": 2.43,
2263
+ "grad_norm": 0.9816425442695618,
2264
+ "learning_rate": 8.626871855061438e-05,
2265
+ "loss": 0.0304,
2266
+ "step": 2630
2267
+ },
2268
+ {
2269
+ "epoch": 2.44,
2270
+ "grad_norm": 0.3567699193954468,
2271
+ "learning_rate": 8.61684481315854e-05,
2272
+ "loss": 0.0266,
2273
+ "step": 2640
2274
+ },
2275
+ {
2276
+ "epoch": 2.45,
2277
+ "grad_norm": 0.1593473106622696,
2278
+ "learning_rate": 8.606787167001257e-05,
2279
+ "loss": 0.0185,
2280
+ "step": 2650
2281
+ },
2282
+ {
2283
+ "epoch": 2.45,
2284
+ "eval_loss": 0.07773510366678238,
2285
+ "eval_runtime": 36.238,
2286
+ "eval_samples_per_second": 126.387,
2287
+ "eval_steps_per_second": 5.271,
2288
+ "step": 2650
2289
+ },
2290
+ {
2291
+ "epoch": 2.46,
2292
+ "grad_norm": 1.622788906097412,
2293
+ "learning_rate": 8.596699001693255e-05,
2294
+ "loss": 0.0336,
2295
+ "step": 2660
2296
+ },
2297
+ {
2298
+ "epoch": 2.47,
2299
+ "grad_norm": 1.3989499807357788,
2300
+ "learning_rate": 8.586580402596446e-05,
2301
+ "loss": 0.0231,
2302
+ "step": 2670
2303
+ },
2304
+ {
2305
+ "epoch": 2.48,
2306
+ "grad_norm": 0.629644513130188,
2307
+ "learning_rate": 8.576431455330258e-05,
2308
+ "loss": 0.017,
2309
+ "step": 2680
2310
+ },
2311
+ {
2312
+ "epoch": 2.49,
2313
+ "grad_norm": 0.31903961300849915,
2314
+ "learning_rate": 8.566252245770909e-05,
2315
+ "loss": 0.02,
2316
+ "step": 2690
2317
+ },
2318
+ {
2319
+ "epoch": 2.5,
2320
+ "grad_norm": 1.0299134254455566,
2321
+ "learning_rate": 8.556042860050687e-05,
2322
+ "loss": 0.0178,
2323
+ "step": 2700
2324
+ },
2325
+ {
2326
+ "epoch": 2.5,
2327
+ "eval_loss": 0.07669008523225784,
2328
+ "eval_runtime": 36.2452,
2329
+ "eval_samples_per_second": 126.362,
2330
+ "eval_steps_per_second": 5.27,
2331
+ "step": 2700
2332
+ },
2333
+ {
2334
+ "epoch": 2.5,
2335
+ "grad_norm": 0.8042894005775452,
2336
+ "learning_rate": 8.545803384557219e-05,
2337
+ "loss": 0.0171,
2338
+ "step": 2710
2339
+ },
2340
+ {
2341
+ "epoch": 2.51,
2342
+ "grad_norm": 0.5600835084915161,
2343
+ "learning_rate": 8.535533905932738e-05,
2344
+ "loss": 0.0248,
2345
+ "step": 2720
2346
+ },
2347
+ {
2348
+ "epoch": 2.52,
2349
+ "grad_norm": 1.7625192403793335,
2350
+ "learning_rate": 8.525234511073352e-05,
2351
+ "loss": 0.0318,
2352
+ "step": 2730
2353
+ },
2354
+ {
2355
+ "epoch": 2.53,
2356
+ "grad_norm": 0.7289332747459412,
2357
+ "learning_rate": 8.51490528712831e-05,
2358
+ "loss": 0.0123,
2359
+ "step": 2740
2360
+ },
2361
+ {
2362
+ "epoch": 2.54,
2363
+ "grad_norm": 1.1776474714279175,
2364
+ "learning_rate": 8.504546321499255e-05,
2365
+ "loss": 0.0245,
2366
+ "step": 2750
2367
+ },
2368
+ {
2369
+ "epoch": 2.54,
2370
+ "eval_loss": 0.07172290980815887,
2371
+ "eval_runtime": 36.2485,
2372
+ "eval_samples_per_second": 126.35,
2373
+ "eval_steps_per_second": 5.269,
2374
+ "step": 2750
2375
+ },
2376
+ {
2377
+ "epoch": 2.55,
2378
+ "grad_norm": 0.3856188952922821,
2379
+ "learning_rate": 8.4941577018395e-05,
2380
+ "loss": 0.0093,
2381
+ "step": 2760
2382
+ },
2383
+ {
2384
+ "epoch": 2.56,
2385
+ "grad_norm": 0.5729599595069885,
2386
+ "learning_rate": 8.483739516053276e-05,
2387
+ "loss": 0.0177,
2388
+ "step": 2770
2389
+ },
2390
+ {
2391
+ "epoch": 2.57,
2392
+ "grad_norm": 0.08334312587976456,
2393
+ "learning_rate": 8.473291852294987e-05,
2394
+ "loss": 0.0238,
2395
+ "step": 2780
2396
+ },
2397
+ {
2398
+ "epoch": 2.58,
2399
+ "grad_norm": 0.7210456728935242,
2400
+ "learning_rate": 8.462814798968472e-05,
2401
+ "loss": 0.0132,
2402
+ "step": 2790
2403
+ },
2404
+ {
2405
+ "epoch": 2.59,
2406
+ "grad_norm": 0.7902788519859314,
2407
+ "learning_rate": 8.452308444726249e-05,
2408
+ "loss": 0.0226,
2409
+ "step": 2800
2410
+ },
2411
+ {
2412
+ "epoch": 2.59,
2413
+ "eval_loss": 0.0774235799908638,
2414
+ "eval_runtime": 36.2636,
2415
+ "eval_samples_per_second": 126.298,
2416
+ "eval_steps_per_second": 5.267,
2417
+ "step": 2800
2418
+ },
2419
+ {
2420
+ "epoch": 2.6,
2421
+ "grad_norm": 0.369211882352829,
2422
+ "learning_rate": 8.44177287846877e-05,
2423
+ "loss": 0.03,
2424
+ "step": 2810
2425
+ },
2426
+ {
2427
+ "epoch": 2.61,
2428
+ "grad_norm": 0.7902100682258606,
2429
+ "learning_rate": 8.43120818934367e-05,
2430
+ "loss": 0.0269,
2431
+ "step": 2820
2432
+ },
2433
+ {
2434
+ "epoch": 2.62,
2435
+ "grad_norm": 0.9871087074279785,
2436
+ "learning_rate": 8.420614466745e-05,
2437
+ "loss": 0.0178,
2438
+ "step": 2830
2439
+ },
2440
+ {
2441
+ "epoch": 2.62,
2442
+ "grad_norm": 0.5393310189247131,
2443
+ "learning_rate": 8.409991800312493e-05,
2444
+ "loss": 0.0189,
2445
+ "step": 2840
2446
+ },
2447
+ {
2448
+ "epoch": 2.63,
2449
+ "grad_norm": 0.3893219828605652,
2450
+ "learning_rate": 8.399340279930786e-05,
2451
+ "loss": 0.0222,
2452
+ "step": 2850
2453
+ },
2454
+ {
2455
+ "epoch": 2.63,
2456
+ "eval_loss": 0.0670793429017067,
2457
+ "eval_runtime": 36.2663,
2458
+ "eval_samples_per_second": 126.288,
2459
+ "eval_steps_per_second": 5.267,
2460
+ "step": 2850
2461
+ },
2462
+ {
2463
+ "epoch": 2.64,
2464
+ "grad_norm": 0.6049796342849731,
2465
+ "learning_rate": 8.388659995728663e-05,
2466
+ "loss": 0.0187,
2467
+ "step": 2860
2468
+ },
2469
+ {
2470
+ "epoch": 2.65,
2471
+ "grad_norm": 0.6915044784545898,
2472
+ "learning_rate": 8.377951038078302e-05,
2473
+ "loss": 0.0136,
2474
+ "step": 2870
2475
+ },
2476
+ {
2477
+ "epoch": 2.66,
2478
+ "grad_norm": 0.8472715020179749,
2479
+ "learning_rate": 8.367213497594502e-05,
2480
+ "loss": 0.0318,
2481
+ "step": 2880
2482
+ },
2483
+ {
2484
+ "epoch": 2.67,
2485
+ "grad_norm": 1.61133873462677,
2486
+ "learning_rate": 8.356447465133919e-05,
2487
+ "loss": 0.0228,
2488
+ "step": 2890
2489
+ },
2490
+ {
2491
+ "epoch": 2.68,
2492
+ "grad_norm": 0.9213796257972717,
2493
+ "learning_rate": 8.345653031794292e-05,
2494
+ "loss": 0.0194,
2495
+ "step": 2900
2496
+ },
2497
+ {
2498
+ "epoch": 2.68,
2499
+ "eval_loss": 0.0666012316942215,
2500
+ "eval_runtime": 36.2588,
2501
+ "eval_samples_per_second": 126.314,
2502
+ "eval_steps_per_second": 5.268,
2503
+ "step": 2900
2504
+ },
2505
+ {
2506
+ "epoch": 2.69,
2507
+ "grad_norm": 0.09682347625494003,
2508
+ "learning_rate": 8.334830288913682e-05,
2509
+ "loss": 0.0155,
2510
+ "step": 2910
2511
+ },
2512
+ {
2513
+ "epoch": 2.7,
2514
+ "grad_norm": 0.9755958318710327,
2515
+ "learning_rate": 8.323979328069689e-05,
2516
+ "loss": 0.0159,
2517
+ "step": 2920
2518
+ },
2519
+ {
2520
+ "epoch": 2.71,
2521
+ "grad_norm": 0.6496116518974304,
2522
+ "learning_rate": 8.313100241078689e-05,
2523
+ "loss": 0.017,
2524
+ "step": 2930
2525
+ },
2526
+ {
2527
+ "epoch": 2.72,
2528
+ "grad_norm": 0.6899746060371399,
2529
+ "learning_rate": 8.302193119995039e-05,
2530
+ "loss": 0.024,
2531
+ "step": 2940
2532
+ },
2533
+ {
2534
+ "epoch": 2.73,
2535
+ "grad_norm": 1.4230170249938965,
2536
+ "learning_rate": 8.291258057110319e-05,
2537
+ "loss": 0.0162,
2538
+ "step": 2950
2539
+ },
2540
+ {
2541
+ "epoch": 2.73,
2542
+ "eval_loss": 0.07134513556957245,
2543
+ "eval_runtime": 36.2555,
2544
+ "eval_samples_per_second": 126.326,
2545
+ "eval_steps_per_second": 5.268,
2546
+ "step": 2950
2547
+ },
2548
+ {
2549
+ "epoch": 2.74,
2550
+ "grad_norm": 0.5962338447570801,
2551
+ "learning_rate": 8.280295144952536e-05,
2552
+ "loss": 0.0221,
2553
+ "step": 2960
2554
+ },
2555
+ {
2556
+ "epoch": 2.74,
2557
+ "grad_norm": 1.1969221830368042,
2558
+ "learning_rate": 8.269304476285349e-05,
2559
+ "loss": 0.0107,
2560
+ "step": 2970
2561
+ },
2562
+ {
2563
+ "epoch": 2.75,
2564
+ "grad_norm": 1.6272106170654297,
2565
+ "learning_rate": 8.258286144107276e-05,
2566
+ "loss": 0.0185,
2567
+ "step": 2980
2568
+ },
2569
+ {
2570
+ "epoch": 2.76,
2571
+ "grad_norm": 0.18976141512393951,
2572
+ "learning_rate": 8.247240241650918e-05,
2573
+ "loss": 0.0185,
2574
+ "step": 2990
2575
+ },
2576
+ {
2577
+ "epoch": 2.77,
2578
+ "grad_norm": 1.8696092367172241,
2579
+ "learning_rate": 8.236166862382163e-05,
2580
+ "loss": 0.0184,
2581
+ "step": 3000
2582
+ },
2583
+ {
2584
+ "epoch": 2.77,
2585
+ "eval_loss": 0.07399819046258926,
2586
+ "eval_runtime": 36.249,
2587
+ "eval_samples_per_second": 126.348,
2588
+ "eval_steps_per_second": 5.269,
2589
+ "step": 3000
2590
+ },
2591
+ {
2592
+ "epoch": 2.78,
2593
+ "grad_norm": 0.5066729784011841,
2594
+ "learning_rate": 8.225066099999392e-05,
2595
+ "loss": 0.0197,
2596
+ "step": 3010
2597
+ },
2598
+ {
2599
+ "epoch": 2.79,
2600
+ "grad_norm": 0.5865077376365662,
2601
+ "learning_rate": 8.213938048432697e-05,
2602
+ "loss": 0.0231,
2603
+ "step": 3020
2604
+ },
2605
+ {
2606
+ "epoch": 2.8,
2607
+ "grad_norm": 0.2946082651615143,
2608
+ "learning_rate": 8.202782801843076e-05,
2609
+ "loss": 0.013,
2610
+ "step": 3030
2611
+ },
2612
+ {
2613
+ "epoch": 2.81,
2614
+ "grad_norm": 0.49251487851142883,
2615
+ "learning_rate": 8.191600454621642e-05,
2616
+ "loss": 0.0247,
2617
+ "step": 3040
2618
+ },
2619
+ {
2620
+ "epoch": 2.82,
2621
+ "grad_norm": 0.34551000595092773,
2622
+ "learning_rate": 8.18039110138882e-05,
2623
+ "loss": 0.0227,
2624
+ "step": 3050
2625
+ },
2626
+ {
2627
+ "epoch": 2.82,
2628
+ "eval_loss": 0.06754355877637863,
2629
+ "eval_runtime": 36.2553,
2630
+ "eval_samples_per_second": 126.326,
2631
+ "eval_steps_per_second": 5.268,
2632
+ "step": 3050
2633
+ },
2634
+ {
2635
+ "epoch": 2.83,
2636
+ "grad_norm": 1.2313387393951416,
2637
+ "learning_rate": 8.169154836993551e-05,
2638
+ "loss": 0.0206,
2639
+ "step": 3060
2640
+ },
2641
+ {
2642
+ "epoch": 2.84,
2643
+ "grad_norm": 0.2834533154964447,
2644
+ "learning_rate": 8.157891756512488e-05,
2645
+ "loss": 0.0216,
2646
+ "step": 3070
2647
+ },
2648
+ {
2649
+ "epoch": 2.85,
2650
+ "grad_norm": 2.936052083969116,
2651
+ "learning_rate": 8.146601955249188e-05,
2652
+ "loss": 0.0405,
2653
+ "step": 3080
2654
+ },
2655
+ {
2656
+ "epoch": 2.86,
2657
+ "grad_norm": 1.2530442476272583,
2658
+ "learning_rate": 8.135285528733308e-05,
2659
+ "loss": 0.0319,
2660
+ "step": 3090
2661
+ },
2662
+ {
2663
+ "epoch": 2.87,
2664
+ "grad_norm": 0.9805688858032227,
2665
+ "learning_rate": 8.1239425727198e-05,
2666
+ "loss": 0.0176,
2667
+ "step": 3100
2668
+ },
2669
+ {
2670
+ "epoch": 2.87,
2671
+ "eval_loss": 0.07011596113443375,
2672
+ "eval_runtime": 36.237,
2673
+ "eval_samples_per_second": 126.39,
2674
+ "eval_steps_per_second": 5.271,
2675
+ "step": 3100
2676
+ },
2677
+ {
2678
+ "epoch": 2.87,
2679
+ "grad_norm": 1.820011854171753,
2680
+ "learning_rate": 8.112573183188099e-05,
2681
+ "loss": 0.0285,
2682
+ "step": 3110
2683
+ },
2684
+ {
2685
+ "epoch": 2.88,
2686
+ "grad_norm": 0.15229858458042145,
2687
+ "learning_rate": 8.1011774563413e-05,
2688
+ "loss": 0.0149,
2689
+ "step": 3120
2690
+ },
2691
+ {
2692
+ "epoch": 2.89,
2693
+ "grad_norm": 0.5576140880584717,
2694
+ "learning_rate": 8.089755488605366e-05,
2695
+ "loss": 0.0223,
2696
+ "step": 3130
2697
+ },
2698
+ {
2699
+ "epoch": 2.9,
2700
+ "grad_norm": 1.394098162651062,
2701
+ "learning_rate": 8.07830737662829e-05,
2702
+ "loss": 0.015,
2703
+ "step": 3140
2704
+ },
2705
+ {
2706
+ "epoch": 2.91,
2707
+ "grad_norm": 1.0313000679016113,
2708
+ "learning_rate": 8.066833217279297e-05,
2709
+ "loss": 0.034,
2710
+ "step": 3150
2711
+ },
2712
+ {
2713
+ "epoch": 2.91,
2714
+ "eval_loss": 0.06750660389661789,
2715
+ "eval_runtime": 36.2568,
2716
+ "eval_samples_per_second": 126.321,
2717
+ "eval_steps_per_second": 5.268,
2718
+ "step": 3150
2719
+ },
2720
+ {
2721
+ "epoch": 2.92,
2722
+ "grad_norm": 1.3380006551742554,
2723
+ "learning_rate": 8.055333107647999e-05,
2724
+ "loss": 0.0187,
2725
+ "step": 3160
2726
+ },
2727
+ {
2728
+ "epoch": 2.93,
2729
+ "grad_norm": 0.46096986532211304,
2730
+ "learning_rate": 8.043807145043604e-05,
2731
+ "loss": 0.0261,
2732
+ "step": 3170
2733
+ },
2734
+ {
2735
+ "epoch": 2.94,
2736
+ "grad_norm": 0.7799313068389893,
2737
+ "learning_rate": 8.032255426994069e-05,
2738
+ "loss": 0.0175,
2739
+ "step": 3180
2740
+ },
2741
+ {
2742
+ "epoch": 2.95,
2743
+ "grad_norm": 0.5040866732597351,
2744
+ "learning_rate": 8.020678051245282e-05,
2745
+ "loss": 0.0163,
2746
+ "step": 3190
2747
+ },
2748
+ {
2749
+ "epoch": 2.96,
2750
+ "grad_norm": 0.3076965808868408,
2751
+ "learning_rate": 8.009075115760243e-05,
2752
+ "loss": 0.0148,
2753
+ "step": 3200
2754
+ },
2755
+ {
2756
+ "epoch": 2.96,
2757
+ "eval_loss": 0.06875752657651901,
2758
+ "eval_runtime": 36.2627,
2759
+ "eval_samples_per_second": 126.3,
2760
+ "eval_steps_per_second": 5.267,
2761
+ "step": 3200
2762
+ },
2763
+ {
2764
+ "epoch": 2.97,
2765
+ "grad_norm": 0.6659747958183289,
2766
+ "learning_rate": 7.99744671871822e-05,
2767
+ "loss": 0.0445,
2768
+ "step": 3210
2769
+ },
2770
+ {
2771
+ "epoch": 2.98,
2772
+ "grad_norm": 1.3347102403640747,
2773
+ "learning_rate": 7.985792958513931e-05,
2774
+ "loss": 0.0216,
2775
+ "step": 3220
2776
+ },
2777
+ {
2778
+ "epoch": 2.99,
2779
+ "grad_norm": 1.407318115234375,
2780
+ "learning_rate": 7.974113933756707e-05,
2781
+ "loss": 0.0173,
2782
+ "step": 3230
2783
+ },
2784
+ {
2785
+ "epoch": 2.99,
2786
+ "grad_norm": 0.534429669380188,
2787
+ "learning_rate": 7.962409743269654e-05,
2788
+ "loss": 0.0185,
2789
+ "step": 3240
2790
+ },
2791
+ {
2792
+ "epoch": 3.0,
2793
+ "grad_norm": 0.7592530846595764,
2794
+ "learning_rate": 7.950680486088823e-05,
2795
+ "loss": 0.014,
2796
+ "step": 3250
2797
+ },
2798
+ {
2799
+ "epoch": 3.0,
2800
+ "eval_loss": 0.0672665610909462,
2801
+ "eval_runtime": 36.2374,
2802
+ "eval_samples_per_second": 126.389,
2803
+ "eval_steps_per_second": 5.271,
2804
+ "step": 3250
2805
+ },
2806
+ {
2807
+ "epoch": 3.01,
2808
+ "grad_norm": 0.2840610146522522,
2809
+ "learning_rate": 7.938926261462366e-05,
2810
+ "loss": 0.0234,
2811
+ "step": 3260
2812
+ },
2813
+ {
2814
+ "epoch": 3.02,
2815
+ "grad_norm": 2.4856412410736084,
2816
+ "learning_rate": 7.927147168849704e-05,
2817
+ "loss": 0.0094,
2818
+ "step": 3270
2819
+ },
2820
+ {
2821
+ "epoch": 3.03,
2822
+ "grad_norm": 1.395142674446106,
2823
+ "learning_rate": 7.915343307920674e-05,
2824
+ "loss": 0.007,
2825
+ "step": 3280
2826
+ },
2827
+ {
2828
+ "epoch": 3.04,
2829
+ "grad_norm": 0.21749833226203918,
2830
+ "learning_rate": 7.903514778554699e-05,
2831
+ "loss": 0.0106,
2832
+ "step": 3290
2833
+ },
2834
+ {
2835
+ "epoch": 3.05,
2836
+ "grad_norm": 2.8792037963867188,
2837
+ "learning_rate": 7.891661680839932e-05,
2838
+ "loss": 0.0178,
2839
+ "step": 3300
2840
+ },
2841
+ {
2842
+ "epoch": 3.05,
2843
+ "eval_loss": 0.07193872332572937,
2844
+ "eval_runtime": 36.2444,
2845
+ "eval_samples_per_second": 126.364,
2846
+ "eval_steps_per_second": 5.27,
2847
+ "step": 3300
2848
+ },
2849
+ {
2850
+ "epoch": 3.06,
2851
+ "grad_norm": 0.2820728123188019,
2852
+ "learning_rate": 7.879784115072417e-05,
2853
+ "loss": 0.0051,
2854
+ "step": 3310
2855
+ },
2856
+ {
2857
+ "epoch": 3.07,
2858
+ "grad_norm": 0.22677582502365112,
2859
+ "learning_rate": 7.86788218175523e-05,
2860
+ "loss": 0.0035,
2861
+ "step": 3320
2862
+ },
2863
+ {
2864
+ "epoch": 3.08,
2865
+ "grad_norm": 0.14426374435424805,
2866
+ "learning_rate": 7.855955981597643e-05,
2867
+ "loss": 0.0028,
2868
+ "step": 3330
2869
+ },
2870
+ {
2871
+ "epoch": 3.09,
2872
+ "grad_norm": 0.5178078413009644,
2873
+ "learning_rate": 7.844005615514259e-05,
2874
+ "loss": 0.0092,
2875
+ "step": 3340
2876
+ },
2877
+ {
2878
+ "epoch": 3.1,
2879
+ "grad_norm": 0.049177393317222595,
2880
+ "learning_rate": 7.832031184624164e-05,
2881
+ "loss": 0.0059,
2882
+ "step": 3350
2883
+ },
2884
+ {
2885
+ "epoch": 3.1,
2886
+ "eval_loss": 0.07338841259479523,
2887
+ "eval_runtime": 36.2678,
2888
+ "eval_samples_per_second": 126.283,
2889
+ "eval_steps_per_second": 5.266,
2890
+ "step": 3350
2891
+ },
2892
+ {
2893
+ "epoch": 3.11,
2894
+ "grad_norm": 0.20768703520298004,
2895
+ "learning_rate": 7.820032790250074e-05,
2896
+ "loss": 0.0145,
2897
+ "step": 3360
2898
+ },
2899
+ {
2900
+ "epoch": 3.11,
2901
+ "grad_norm": 0.4482771158218384,
2902
+ "learning_rate": 7.808010533917465e-05,
2903
+ "loss": 0.0043,
2904
+ "step": 3370
2905
+ },
2906
+ {
2907
+ "epoch": 3.12,
2908
+ "grad_norm": 0.7216786742210388,
2909
+ "learning_rate": 7.795964517353735e-05,
2910
+ "loss": 0.0125,
2911
+ "step": 3380
2912
+ },
2913
+ {
2914
+ "epoch": 3.13,
2915
+ "grad_norm": 0.028145039454102516,
2916
+ "learning_rate": 7.783894842487322e-05,
2917
+ "loss": 0.0023,
2918
+ "step": 3390
2919
+ },
2920
+ {
2921
+ "epoch": 3.14,
2922
+ "grad_norm": 0.02350945584475994,
2923
+ "learning_rate": 7.771801611446858e-05,
2924
+ "loss": 0.0069,
2925
+ "step": 3400
2926
+ },
2927
+ {
2928
+ "epoch": 3.14,
2929
+ "eval_loss": 0.07644446939229965,
2930
+ "eval_runtime": 36.2497,
2931
+ "eval_samples_per_second": 126.346,
2932
+ "eval_steps_per_second": 5.269,
2933
+ "step": 3400
2934
+ },
2935
+ {
2936
+ "epoch": 3.15,
2937
+ "grad_norm": 0.12125283479690552,
2938
+ "learning_rate": 7.75968492656029e-05,
2939
+ "loss": 0.0118,
2940
+ "step": 3410
2941
+ },
2942
+ {
2943
+ "epoch": 3.16,
2944
+ "grad_norm": 0.43455737829208374,
2945
+ "learning_rate": 7.74754489035403e-05,
2946
+ "loss": 0.0066,
2947
+ "step": 3420
2948
+ },
2949
+ {
2950
+ "epoch": 3.17,
2951
+ "grad_norm": 1.3440496921539307,
2952
+ "learning_rate": 7.735381605552073e-05,
2953
+ "loss": 0.0187,
2954
+ "step": 3430
2955
+ },
2956
+ {
2957
+ "epoch": 3.18,
2958
+ "grad_norm": 0.5088502168655396,
2959
+ "learning_rate": 7.723195175075136e-05,
2960
+ "loss": 0.0094,
2961
+ "step": 3440
2962
+ },
2963
+ {
2964
+ "epoch": 3.19,
2965
+ "grad_norm": 0.38931986689567566,
2966
+ "learning_rate": 7.710985702039786e-05,
2967
+ "loss": 0.0074,
2968
+ "step": 3450
2969
+ },
2970
+ {
2971
+ "epoch": 3.19,
2972
+ "eval_loss": 0.08178228884935379,
2973
+ "eval_runtime": 36.2441,
2974
+ "eval_samples_per_second": 126.365,
2975
+ "eval_steps_per_second": 5.27,
2976
+ "step": 3450
2977
+ },
2978
+ {
2979
+ "epoch": 3.2,
2980
+ "grad_norm": 1.8949205875396729,
2981
+ "learning_rate": 7.698753289757565e-05,
2982
+ "loss": 0.0038,
2983
+ "step": 3460
2984
+ },
2985
+ {
2986
+ "epoch": 3.21,
2987
+ "grad_norm": 1.4405574798583984,
2988
+ "learning_rate": 7.68649804173412e-05,
2989
+ "loss": 0.0113,
2990
+ "step": 3470
2991
+ },
2992
+ {
2993
+ "epoch": 3.22,
2994
+ "grad_norm": 0.03057631105184555,
2995
+ "learning_rate": 7.674220061668322e-05,
2996
+ "loss": 0.0099,
2997
+ "step": 3480
2998
+ },
2999
+ {
3000
+ "epoch": 3.23,
3001
+ "grad_norm": 1.858884572982788,
3002
+ "learning_rate": 7.661919453451394e-05,
3003
+ "loss": 0.0089,
3004
+ "step": 3490
3005
+ },
3006
+ {
3007
+ "epoch": 3.23,
3008
+ "grad_norm": 0.02921319007873535,
3009
+ "learning_rate": 7.649596321166024e-05,
3010
+ "loss": 0.009,
3011
+ "step": 3500
3012
+ },
3013
+ {
3014
+ "epoch": 3.23,
3015
+ "eval_loss": 0.07047122716903687,
3016
+ "eval_runtime": 36.2528,
3017
+ "eval_samples_per_second": 126.335,
3018
+ "eval_steps_per_second": 5.269,
3019
+ "step": 3500
3020
+ },
3021
+ {
3022
+ "epoch": 3.24,
3023
+ "grad_norm": 0.7095614671707153,
3024
+ "learning_rate": 7.6372507690855e-05,
3025
+ "loss": 0.0084,
3026
+ "step": 3510
3027
+ },
3028
+ {
3029
+ "epoch": 3.25,
3030
+ "grad_norm": 1.0134788751602173,
3031
+ "learning_rate": 7.6248829016728e-05,
3032
+ "loss": 0.0119,
3033
+ "step": 3520
3034
+ },
3035
+ {
3036
+ "epoch": 3.26,
3037
+ "grad_norm": 0.09262863546609879,
3038
+ "learning_rate": 7.612492823579745e-05,
3039
+ "loss": 0.0139,
3040
+ "step": 3530
3041
+ },
3042
+ {
3043
+ "epoch": 3.27,
3044
+ "grad_norm": 0.033623356372117996,
3045
+ "learning_rate": 7.600080639646077e-05,
3046
+ "loss": 0.0092,
3047
+ "step": 3540
3048
+ },
3049
+ {
3050
+ "epoch": 3.28,
3051
+ "grad_norm": 3.299873113632202,
3052
+ "learning_rate": 7.5876464548986e-05,
3053
+ "loss": 0.0048,
3054
+ "step": 3550
3055
+ },
3056
+ {
3057
+ "epoch": 3.28,
3058
+ "eval_loss": 0.07350648939609528,
3059
+ "eval_runtime": 36.2532,
3060
+ "eval_samples_per_second": 126.334,
3061
+ "eval_steps_per_second": 5.268,
3062
+ "step": 3550
3063
+ },
3064
+ {
3065
+ "epoch": 3.29,
3066
+ "grad_norm": 0.299811452627182,
3067
+ "learning_rate": 7.575190374550272e-05,
3068
+ "loss": 0.0095,
3069
+ "step": 3560
3070
+ },
3071
+ {
3072
+ "epoch": 3.3,
3073
+ "grad_norm": 0.5699326395988464,
3074
+ "learning_rate": 7.562712503999327e-05,
3075
+ "loss": 0.0052,
3076
+ "step": 3570
3077
+ },
3078
+ {
3079
+ "epoch": 3.31,
3080
+ "grad_norm": 0.2542770802974701,
3081
+ "learning_rate": 7.550212948828377e-05,
3082
+ "loss": 0.0037,
3083
+ "step": 3580
3084
+ },
3085
+ {
3086
+ "epoch": 3.32,
3087
+ "grad_norm": 0.9009780287742615,
3088
+ "learning_rate": 7.537691814803521e-05,
3089
+ "loss": 0.0194,
3090
+ "step": 3590
3091
+ },
3092
+ {
3093
+ "epoch": 3.33,
3094
+ "grad_norm": 0.22393372654914856,
3095
+ "learning_rate": 7.52514920787345e-05,
3096
+ "loss": 0.005,
3097
+ "step": 3600
3098
+ },
3099
+ {
3100
+ "epoch": 3.33,
3101
+ "eval_loss": 0.07049556821584702,
3102
+ "eval_runtime": 36.2616,
3103
+ "eval_samples_per_second": 126.304,
3104
+ "eval_steps_per_second": 5.267,
3105
+ "step": 3600
3106
+ },
3107
+ {
3108
+ "epoch": 3.34,
3109
+ "grad_norm": 0.1645677536725998,
3110
+ "learning_rate": 7.51258523416855e-05,
3111
+ "loss": 0.0162,
3112
+ "step": 3610
3113
+ },
3114
+ {
3115
+ "epoch": 3.35,
3116
+ "grad_norm": 1.0096102952957153,
3117
+ "learning_rate": 7.500000000000001e-05,
3118
+ "loss": 0.0108,
3119
+ "step": 3620
3120
+ },
3121
+ {
3122
+ "epoch": 3.35,
3123
+ "grad_norm": 0.3511151373386383,
3124
+ "learning_rate": 7.487393611858883e-05,
3125
+ "loss": 0.0059,
3126
+ "step": 3630
3127
+ },
3128
+ {
3129
+ "epoch": 3.36,
3130
+ "grad_norm": 0.08066729456186295,
3131
+ "learning_rate": 7.474766176415271e-05,
3132
+ "loss": 0.011,
3133
+ "step": 3640
3134
+ },
3135
+ {
3136
+ "epoch": 3.37,
3137
+ "grad_norm": 0.27665212750434875,
3138
+ "learning_rate": 7.462117800517336e-05,
3139
+ "loss": 0.0073,
3140
+ "step": 3650
3141
+ },
3142
+ {
3143
+ "epoch": 3.37,
3144
+ "eval_loss": 0.07236067205667496,
3145
+ "eval_runtime": 36.2566,
3146
+ "eval_samples_per_second": 126.322,
3147
+ "eval_steps_per_second": 5.268,
3148
+ "step": 3650
3149
+ },
3150
+ {
3151
+ "epoch": 3.37,
3152
+ "step": 3650,
3153
+ "total_flos": 1.0476524335865528e+18,
3154
+ "train_loss": 0.10634728888738645,
3155
+ "train_runtime": 4965.2136,
3156
+ "train_samples_per_second": 52.266,
3157
+ "train_steps_per_second": 2.179
3158
+ }
3159
+ ],
3160
+ "logging_steps": 10,
3161
+ "max_steps": 10820,
3162
+ "num_input_tokens_seen": 0,
3163
+ "num_train_epochs": 10,
3164
+ "save_steps": 500,
3165
+ "total_flos": 1.0476524335865528e+18,
3166
+ "train_batch_size": 8,
3167
+ "trial_name": null,
3168
+ "trial_params": null
3169
+ }
llama2_13b_peft/elementary_math_qa/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21a5142d68e9751e132f8bdf3341b60c7413d1047e923e3e9b78bf437200c919
3
+ size 5112
llama2_13b_peft/elementary_math_qa/training_eval_loss.png ADDED
llama2_13b_peft/elementary_math_qa/training_loss.png ADDED
llama2_13b_peft/epistemic_reasoning/README.md ADDED
@@ -0,0 +1,70 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ library_name: peft
4
+ tags:
5
+ - llama-factory
6
+ - lora
7
+ - generated_from_trainer
8
+ base_model: /data1/model/llama2/meta-llama/Llama2-13b
9
+ model-index:
10
+ - name: epistemic_reasoning_no_sys
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # epistemic_reasoning_no_sys
18
+
19
+ This model is a fine-tuned version of [/data1/model/llama2/meta-llama/Llama2-13b](https://huggingface.co//data1/model/llama2/meta-llama/Llama2-13b) on the epistemic_reasoning_no_sys dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 0.0000
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 5e-05
41
+ - train_batch_size: 4
42
+ - eval_batch_size: 4
43
+ - seed: 42
44
+ - distributed_type: multi-GPU
45
+ - num_devices: 2
46
+ - total_train_batch_size: 8
47
+ - total_eval_batch_size: 8
48
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
49
+ - lr_scheduler_type: cosine
50
+ - lr_scheduler_warmup_steps: 20
51
+ - num_epochs: 5.0
52
+
53
+ ### Training results
54
+
55
+ | Training Loss | Epoch | Step | Validation Loss |
56
+ |:-------------:|:------:|:----:|:---------------:|
57
+ | 0.0637 | 0.5263 | 100 | 0.0353 |
58
+ | 0.0001 | 1.0526 | 200 | 0.0001 |
59
+ | 0.0 | 1.5789 | 300 | 0.0000 |
60
+ | 0.0 | 2.1053 | 400 | 0.0000 |
61
+ | 0.0 | 2.6316 | 500 | 0.0000 |
62
+
63
+
64
+ ### Framework versions
65
+
66
+ - PEFT 0.10.0
67
+ - Transformers 4.40.0
68
+ - Pytorch 2.2.1
69
+ - Datasets 2.18.0
70
+ - Tokenizers 0.19.1
llama2_13b_peft/epistemic_reasoning/adapter_config.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "/data1/model/llama2/meta-llama/Llama2-13b",
5
+ "bias": "none",
6
+ "fan_in_fan_out": false,
7
+ "inference_mode": true,
8
+ "init_lora_weights": true,
9
+ "layer_replication": null,
10
+ "layers_pattern": null,
11
+ "layers_to_transform": null,
12
+ "loftq_config": {},
13
+ "lora_alpha": 16,
14
+ "lora_dropout": 0.0,
15
+ "megatron_config": null,
16
+ "megatron_core": "megatron.core",
17
+ "modules_to_save": null,
18
+ "peft_type": "LORA",
19
+ "r": 8,
20
+ "rank_pattern": {},
21
+ "revision": null,
22
+ "target_modules": [
23
+ "k_proj",
24
+ "gate_proj",
25
+ "o_proj",
26
+ "q_proj",
27
+ "up_proj",
28
+ "v_proj",
29
+ "down_proj"
30
+ ],
31
+ "task_type": "CAUSAL_LM",
32
+ "use_dora": false,
33
+ "use_rslora": false
34
+ }
llama2_13b_peft/epistemic_reasoning/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:faa79b2a88be4cb133ae6dd764c9f382cf02b6c514052a3939e3b7f702b5139e
3
+ size 125248064
llama2_13b_peft/epistemic_reasoning/all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.6315789473684212,
3
+ "eval_loss": 8.995843018055893e-06,
4
+ "eval_runtime": 0.935,
5
+ "eval_samples_per_second": 85.565,
6
+ "eval_steps_per_second": 10.696,
7
+ "total_flos": 3.922877488509747e+16,
8
+ "train_loss": 0.16288311984013126,
9
+ "train_runtime": 177.9523,
10
+ "train_samples_per_second": 42.708,
11
+ "train_steps_per_second": 5.339
12
+ }
llama2_13b_peft/epistemic_reasoning/eval_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.6315789473684212,
3
+ "eval_loss": 8.995843018055893e-06,
4
+ "eval_runtime": 0.935,
5
+ "eval_samples_per_second": 85.565,
6
+ "eval_steps_per_second": 10.696
7
+ }
llama2_13b_peft/epistemic_reasoning/special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "</s>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
llama2_13b_peft/epistemic_reasoning/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
llama2_13b_peft/epistemic_reasoning/tokenizer_config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "add_prefix_space": true,
5
+ "added_tokens_decoder": {
6
+ "0": {
7
+ "content": "<unk>",
8
+ "lstrip": false,
9
+ "normalized": false,
10
+ "rstrip": false,
11
+ "single_word": false,
12
+ "special": true
13
+ },
14
+ "1": {
15
+ "content": "<s>",
16
+ "lstrip": false,
17
+ "normalized": false,
18
+ "rstrip": false,
19
+ "single_word": false,
20
+ "special": true
21
+ },
22
+ "2": {
23
+ "content": "</s>",
24
+ "lstrip": false,
25
+ "normalized": false,
26
+ "rstrip": false,
27
+ "single_word": false,
28
+ "special": true
29
+ }
30
+ },
31
+ "bos_token": "<s>",
32
+ "chat_template": "{% if messages[0]['role'] == 'system' %}{% set system_message = messages[0]['content'] %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ content }}{% elif message['role'] == 'assistant' %}{{ content + '\\n' }}{% endif %}{% endfor %}",
33
+ "clean_up_tokenization_spaces": false,
34
+ "eos_token": "</s>",
35
+ "legacy": true,
36
+ "model_max_length": 1000000000000000019884624838656,
37
+ "pad_token": "</s>",
38
+ "padding_side": "right",
39
+ "sp_model_kwargs": {},
40
+ "spaces_between_special_tokens": false,
41
+ "split_special_tokens": false,
42
+ "tokenizer_class": "LlamaTokenizer",
43
+ "unk_token": "<unk>",
44
+ "use_default_system_prompt": false
45
+ }
llama2_13b_peft/epistemic_reasoning/train_results.json ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 2.6315789473684212,
3
+ "total_flos": 3.922877488509747e+16,
4
+ "train_loss": 0.16288311984013126,
5
+ "train_runtime": 177.9523,
6
+ "train_samples_per_second": 42.708,
7
+ "train_steps_per_second": 5.339
8
+ }
llama2_13b_peft/epistemic_reasoning/trainer_log.jsonl ADDED
@@ -0,0 +1,57 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"current_steps": 10, "total_steps": 950, "loss": 4.1842, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 2.5e-05, "epoch": 0.05263157894736842, "percentage": 1.05, "elapsed_time": "0:00:04", "remaining_time": "0:07:43"}
2
+ {"current_steps": 20, "total_steps": 950, "loss": 2.6471, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 5e-05, "epoch": 0.10526315789473684, "percentage": 2.11, "elapsed_time": "0:00:08", "remaining_time": "0:06:21"}
3
+ {"current_steps": 30, "total_steps": 950, "loss": 0.31, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.998573727324295e-05, "epoch": 0.15789473684210525, "percentage": 3.16, "elapsed_time": "0:00:11", "remaining_time": "0:05:59"}
4
+ {"current_steps": 40, "total_steps": 950, "loss": 0.1595, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.994296536700177e-05, "epoch": 0.21052631578947367, "percentage": 4.21, "elapsed_time": "0:00:14", "remaining_time": "0:05:38"}
5
+ {"current_steps": 50, "total_steps": 950, "loss": 0.1462, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.987173308479738e-05, "epoch": 0.2631578947368421, "percentage": 5.26, "elapsed_time": "0:00:18", "remaining_time": "0:05:25"}
6
+ {"current_steps": 60, "total_steps": 950, "loss": 0.1413, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.977212170395598e-05, "epoch": 0.3157894736842105, "percentage": 6.32, "elapsed_time": "0:00:21", "remaining_time": "0:05:14"}
7
+ {"current_steps": 70, "total_steps": 950, "loss": 0.1202, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.964424488287009e-05, "epoch": 0.3684210526315789, "percentage": 7.37, "elapsed_time": "0:00:24", "remaining_time": "0:05:06"}
8
+ {"current_steps": 80, "total_steps": 950, "loss": 0.1262, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.948824853131236e-05, "epoch": 0.42105263157894735, "percentage": 8.42, "elapsed_time": "0:00:27", "remaining_time": "0:04:59"}
9
+ {"current_steps": 90, "total_steps": 950, "loss": 0.1133, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.930431064394977e-05, "epoch": 0.47368421052631576, "percentage": 9.47, "elapsed_time": "0:00:30", "remaining_time": "0:04:55"}
10
+ {"current_steps": 100, "total_steps": 950, "loss": 0.0637, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.909264109724853e-05, "epoch": 0.5263157894736842, "percentage": 10.53, "elapsed_time": "0:00:34", "remaining_time": "0:04:51"}
11
+ {"current_steps": 100, "total_steps": 950, "loss": null, "eval_loss": 0.03525802493095398, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": null, "epoch": 0.5263157894736842, "percentage": 10.53, "elapsed_time": "0:00:34", "remaining_time": "0:04:51"}
12
+ {"current_steps": 110, "total_steps": 950, "loss": 0.0422, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.885348141000122e-05, "epoch": 0.5789473684210527, "percentage": 11.58, "elapsed_time": "0:00:38", "remaining_time": "0:04:56"}
13
+ {"current_steps": 120, "total_steps": 950, "loss": 0.0171, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.858710446774951e-05, "epoch": 0.631578947368421, "percentage": 12.63, "elapsed_time": "0:00:42", "remaining_time": "0:04:50"}
14
+ {"current_steps": 130, "total_steps": 950, "loss": 0.0178, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.829381421141671e-05, "epoch": 0.6842105263157895, "percentage": 13.68, "elapsed_time": "0:00:45", "remaining_time": "0:04:45"}
15
+ {"current_steps": 140, "total_steps": 950, "loss": 0.0354, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.7973945290505766e-05, "epoch": 0.7368421052631579, "percentage": 14.74, "elapsed_time": "0:00:48", "remaining_time": "0:04:41"}
16
+ {"current_steps": 150, "total_steps": 950, "loss": 0.008, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.7627862681258037e-05, "epoch": 0.7894736842105263, "percentage": 15.79, "elapsed_time": "0:00:52", "remaining_time": "0:04:37"}
17
+ {"current_steps": 160, "total_steps": 950, "loss": 0.0007, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.725596127020879e-05, "epoch": 0.8421052631578947, "percentage": 16.84, "elapsed_time": "0:00:55", "remaining_time": "0:04:32"}
18
+ {"current_steps": 170, "total_steps": 950, "loss": 0.0036, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.685866540361456e-05, "epoch": 0.8947368421052632, "percentage": 17.89, "elapsed_time": "0:00:58", "remaining_time": "0:04:28"}
19
+ {"current_steps": 180, "total_steps": 950, "loss": 0.0064, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.643642840326627e-05, "epoch": 0.9473684210526315, "percentage": 18.95, "elapsed_time": "0:01:01", "remaining_time": "0:04:23"}
20
+ {"current_steps": 190, "total_steps": 950, "loss": 0.0006, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.598973204924097e-05, "epoch": 1.0, "percentage": 20.0, "elapsed_time": "0:01:04", "remaining_time": "0:04:18"}
21
+ {"current_steps": 200, "total_steps": 950, "loss": 0.0001, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.551908603018191e-05, "epoch": 1.0526315789473684, "percentage": 21.05, "elapsed_time": "0:01:07", "remaining_time": "0:04:14"}
22
+ {"current_steps": 200, "total_steps": 950, "loss": null, "eval_loss": 5.087663885205984e-05, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": null, "epoch": 1.0526315789473684, "percentage": 21.05, "elapsed_time": "0:01:07", "remaining_time": "0:04:14"}
23
+ {"current_steps": 210, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.502502736173462e-05, "epoch": 1.1052631578947367, "percentage": 22.11, "elapsed_time": "0:01:12", "remaining_time": "0:04:16"}
24
+ {"current_steps": 220, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.45081197738023e-05, "epoch": 1.1578947368421053, "percentage": 23.16, "elapsed_time": "0:01:16", "remaining_time": "0:04:12"}
25
+ {"current_steps": 230, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.3968953067319777e-05, "epoch": 1.2105263157894737, "percentage": 24.21, "elapsed_time": "0:01:19", "remaining_time": "0:04:08"}
26
+ {"current_steps": 240, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.340814244127993e-05, "epoch": 1.263157894736842, "percentage": 25.26, "elapsed_time": "0:01:22", "remaining_time": "0:04:04"}
27
+ {"current_steps": 250, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.282632779078051e-05, "epoch": 1.3157894736842106, "percentage": 26.32, "elapsed_time": "0:01:25", "remaining_time": "0:04:00"}
28
+ {"current_steps": 260, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.222417297689217e-05, "epoch": 1.368421052631579, "percentage": 27.37, "elapsed_time": "0:01:29", "remaining_time": "0:03:56"}
29
+ {"current_steps": 270, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.160236506918098e-05, "epoch": 1.4210526315789473, "percentage": 28.42, "elapsed_time": "0:01:32", "remaining_time": "0:03:52"}
30
+ {"current_steps": 280, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.096161356174959e-05, "epoch": 1.4736842105263157, "percentage": 29.47, "elapsed_time": "0:01:35", "remaining_time": "0:03:48"}
31
+ {"current_steps": 290, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 4.030264956369157e-05, "epoch": 1.526315789473684, "percentage": 30.53, "elapsed_time": "0:01:38", "remaining_time": "0:03:44"}
32
+ {"current_steps": 300, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 3.962622496488269e-05, "epoch": 1.5789473684210527, "percentage": 31.58, "elapsed_time": "0:01:41", "remaining_time": "0:03:40"}
33
+ {"current_steps": 300, "total_steps": 950, "loss": null, "eval_loss": 1.916586552397348e-05, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": null, "epoch": 1.5789473684210527, "percentage": 31.58, "elapsed_time": "0:01:41", "remaining_time": "0:03:40"}
34
+ {"current_steps": 310, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 3.893311157806091e-05, "epoch": 1.631578947368421, "percentage": 32.63, "elapsed_time": "0:01:46", "remaining_time": "0:03:39"}
35
+ {"current_steps": 320, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 3.822410025817406e-05, "epoch": 1.6842105263157894, "percentage": 33.68, "elapsed_time": "0:01:49", "remaining_time": "0:03:35"}
36
+ {"current_steps": 330, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 3.7500000000000003e-05, "epoch": 1.736842105263158, "percentage": 34.74, "elapsed_time": "0:01:52", "remaining_time": "0:03:31"}
37
+ {"current_steps": 340, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 3.67616370150689e-05, "epoch": 1.7894736842105263, "percentage": 35.79, "elapsed_time": "0:01:55", "remaining_time": "0:03:28"}
38
+ {"current_steps": 350, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 3.600985378894086e-05, "epoch": 1.8421052631578947, "percentage": 36.84, "elapsed_time": "0:01:59", "remaining_time": "0:03:24"}
39
+ {"current_steps": 360, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 3.5245508119914687e-05, "epoch": 1.8947368421052633, "percentage": 37.89, "elapsed_time": "0:02:02", "remaining_time": "0:03:20"}
40
+ {"current_steps": 370, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 3.44694721402644e-05, "epoch": 1.9473684210526314, "percentage": 38.95, "elapsed_time": "0:02:05", "remaining_time": "0:03:17"}
41
+ {"current_steps": 380, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 3.3682631321120504e-05, "epoch": 2.0, "percentage": 40.0, "elapsed_time": "0:02:08", "remaining_time": "0:03:13"}
42
+ {"current_steps": 390, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 3.2885883462131394e-05, "epoch": 2.0526315789473686, "percentage": 41.05, "elapsed_time": "0:02:12", "remaining_time": "0:03:09"}
43
+ {"current_steps": 400, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 3.2080137667057595e-05, "epoch": 2.1052631578947367, "percentage": 42.11, "elapsed_time": "0:02:15", "remaining_time": "0:03:06"}
44
+ {"current_steps": 400, "total_steps": 950, "loss": null, "eval_loss": 1.2211120520078111e-05, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": null, "epoch": 2.1052631578947367, "percentage": 42.11, "elapsed_time": "0:02:15", "remaining_time": "0:03:06"}
45
+ {"current_steps": 410, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 3.126631330646802e-05, "epoch": 2.1578947368421053, "percentage": 43.16, "elapsed_time": "0:02:20", "remaining_time": "0:03:04"}
46
+ {"current_steps": 420, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 3.0445338968721287e-05, "epoch": 2.2105263157894735, "percentage": 44.21, "elapsed_time": "0:02:23", "remaining_time": "0:03:00"}
47
+ {"current_steps": 430, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 2.961815140042974e-05, "epoch": 2.263157894736842, "percentage": 45.26, "elapsed_time": "0:02:26", "remaining_time": "0:02:56"}
48
+ {"current_steps": 440, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 2.878569443761442e-05, "epoch": 2.3157894736842106, "percentage": 46.32, "elapsed_time": "0:02:29", "remaining_time": "0:02:53"}
49
+ {"current_steps": 450, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 2.7948917928771158e-05, "epoch": 2.3684210526315788, "percentage": 47.37, "elapsed_time": "0:02:32", "remaining_time": "0:02:49"}
50
+ {"current_steps": 460, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 2.7108776651076118e-05, "epoch": 2.4210526315789473, "percentage": 48.42, "elapsed_time": "0:02:36", "remaining_time": "0:02:46"}
51
+ {"current_steps": 470, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 2.6266229220967818e-05, "epoch": 2.473684210526316, "percentage": 49.47, "elapsed_time": "0:02:39", "remaining_time": "0:02:43"}
52
+ {"current_steps": 480, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 2.5422237000348276e-05, "epoch": 2.526315789473684, "percentage": 50.53, "elapsed_time": "0:02:43", "remaining_time": "0:02:39"}
53
+ {"current_steps": 490, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 2.4577762999651726e-05, "epoch": 2.5789473684210527, "percentage": 51.58, "elapsed_time": "0:02:46", "remaining_time": "0:02:35"}
54
+ {"current_steps": 500, "total_steps": 950, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": 2.3733770779032184e-05, "epoch": 2.6315789473684212, "percentage": 52.63, "elapsed_time": "0:02:49", "remaining_time": "0:02:32"}
55
+ {"current_steps": 500, "total_steps": 950, "loss": null, "eval_loss": 8.995843018055893e-06, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": null, "epoch": 2.6315789473684212, "percentage": 52.63, "elapsed_time": "0:02:49", "remaining_time": "0:02:32"}
56
+ {"current_steps": 500, "total_steps": 950, "loss": null, "eval_loss": null, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": null, "epoch": 2.6315789473684212, "percentage": 52.63, "elapsed_time": "0:02:49", "remaining_time": "0:02:32"}
57
+ {"current_steps": 10, "total_steps": 10, "loss": null, "eval_loss": 8.995843018055893e-06, "predict_loss": null, "reward": null, "accuracy": null, "learning_rate": null, "epoch": 2.6315789473684212, "percentage": 100.0, "elapsed_time": "0:02:59", "remaining_time": "0:00:00"}
llama2_13b_peft/epistemic_reasoning/trainer_state.json ADDED
@@ -0,0 +1,420 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 8.995843018055893e-06,
3
+ "best_model_checkpoint": "ckpt/llama2_13b_fuze27_no_sys/epistemic_reasoning_no_sys/checkpoint-500",
4
+ "epoch": 2.6315789473684212,
5
+ "eval_steps": 100,
6
+ "global_step": 500,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.05263157894736842,
13
+ "grad_norm": 6.579661846160889,
14
+ "learning_rate": 2.5e-05,
15
+ "loss": 4.1842,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.10526315789473684,
20
+ "grad_norm": 3.819920539855957,
21
+ "learning_rate": 5e-05,
22
+ "loss": 2.6471,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.15789473684210525,
27
+ "grad_norm": 0.6500287055969238,
28
+ "learning_rate": 4.998573727324295e-05,
29
+ "loss": 0.31,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.21052631578947367,
34
+ "grad_norm": 3.065582275390625,
35
+ "learning_rate": 4.994296536700177e-05,
36
+ "loss": 0.1595,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.2631578947368421,
41
+ "grad_norm": 2.211399555206299,
42
+ "learning_rate": 4.987173308479738e-05,
43
+ "loss": 0.1462,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.3157894736842105,
48
+ "grad_norm": 2.7942709922790527,
49
+ "learning_rate": 4.977212170395598e-05,
50
+ "loss": 0.1413,
51
+ "step": 60
52
+ },
53
+ {
54
+ "epoch": 0.3684210526315789,
55
+ "grad_norm": 1.0859003067016602,
56
+ "learning_rate": 4.964424488287009e-05,
57
+ "loss": 0.1202,
58
+ "step": 70
59
+ },
60
+ {
61
+ "epoch": 0.42105263157894735,
62
+ "grad_norm": 0.25432416796684265,
63
+ "learning_rate": 4.948824853131236e-05,
64
+ "loss": 0.1262,
65
+ "step": 80
66
+ },
67
+ {
68
+ "epoch": 0.47368421052631576,
69
+ "grad_norm": 0.5941969752311707,
70
+ "learning_rate": 4.930431064394977e-05,
71
+ "loss": 0.1133,
72
+ "step": 90
73
+ },
74
+ {
75
+ "epoch": 0.5263157894736842,
76
+ "grad_norm": 0.5798951983451843,
77
+ "learning_rate": 4.909264109724853e-05,
78
+ "loss": 0.0637,
79
+ "step": 100
80
+ },
81
+ {
82
+ "epoch": 0.5263157894736842,
83
+ "eval_loss": 0.03525802493095398,
84
+ "eval_runtime": 0.9308,
85
+ "eval_samples_per_second": 85.945,
86
+ "eval_steps_per_second": 10.743,
87
+ "step": 100
88
+ },
89
+ {
90
+ "epoch": 0.5789473684210527,
91
+ "grad_norm": 0.03817753493785858,
92
+ "learning_rate": 4.885348141000122e-05,
93
+ "loss": 0.0422,
94
+ "step": 110
95
+ },
96
+ {
97
+ "epoch": 0.631578947368421,
98
+ "grad_norm": 0.1301649510860443,
99
+ "learning_rate": 4.858710446774951e-05,
100
+ "loss": 0.0171,
101
+ "step": 120
102
+ },
103
+ {
104
+ "epoch": 0.6842105263157895,
105
+ "grad_norm": 7.279715538024902,
106
+ "learning_rate": 4.829381421141671e-05,
107
+ "loss": 0.0178,
108
+ "step": 130
109
+ },
110
+ {
111
+ "epoch": 0.7368421052631579,
112
+ "grad_norm": 0.01268183533102274,
113
+ "learning_rate": 4.7973945290505766e-05,
114
+ "loss": 0.0354,
115
+ "step": 140
116
+ },
117
+ {
118
+ "epoch": 0.7894736842105263,
119
+ "grad_norm": 0.008723316714167595,
120
+ "learning_rate": 4.7627862681258037e-05,
121
+ "loss": 0.008,
122
+ "step": 150
123
+ },
124
+ {
125
+ "epoch": 0.8421052631578947,
126
+ "grad_norm": 0.005658257752656937,
127
+ "learning_rate": 4.725596127020879e-05,
128
+ "loss": 0.0007,
129
+ "step": 160
130
+ },
131
+ {
132
+ "epoch": 0.8947368421052632,
133
+ "grad_norm": 0.00573640875518322,
134
+ "learning_rate": 4.685866540361456e-05,
135
+ "loss": 0.0036,
136
+ "step": 170
137
+ },
138
+ {
139
+ "epoch": 0.9473684210526315,
140
+ "grad_norm": 0.0025901510380208492,
141
+ "learning_rate": 4.643642840326627e-05,
142
+ "loss": 0.0064,
143
+ "step": 180
144
+ },
145
+ {
146
+ "epoch": 1.0,
147
+ "grad_norm": 1.690902590751648,
148
+ "learning_rate": 4.598973204924097e-05,
149
+ "loss": 0.0006,
150
+ "step": 190
151
+ },
152
+ {
153
+ "epoch": 1.0526315789473684,
154
+ "grad_norm": 0.001473828568123281,
155
+ "learning_rate": 4.551908603018191e-05,
156
+ "loss": 0.0001,
157
+ "step": 200
158
+ },
159
+ {
160
+ "epoch": 1.0526315789473684,
161
+ "eval_loss": 5.087663885205984e-05,
162
+ "eval_runtime": 0.9345,
163
+ "eval_samples_per_second": 85.606,
164
+ "eval_steps_per_second": 10.701,
165
+ "step": 200
166
+ },
167
+ {
168
+ "epoch": 1.1052631578947367,
169
+ "grad_norm": 0.0009911700617522001,
170
+ "learning_rate": 4.502502736173462e-05,
171
+ "loss": 0.0,
172
+ "step": 210
173
+ },
174
+ {
175
+ "epoch": 1.1578947368421053,
176
+ "grad_norm": 0.0010419037425890565,
177
+ "learning_rate": 4.45081197738023e-05,
178
+ "loss": 0.0,
179
+ "step": 220
180
+ },
181
+ {
182
+ "epoch": 1.2105263157894737,
183
+ "grad_norm": 0.0012509481748566031,
184
+ "learning_rate": 4.3968953067319777e-05,
185
+ "loss": 0.0,
186
+ "step": 230
187
+ },
188
+ {
189
+ "epoch": 1.263157894736842,
190
+ "grad_norm": 0.0007601848337799311,
191
+ "learning_rate": 4.340814244127993e-05,
192
+ "loss": 0.0,
193
+ "step": 240
194
+ },
195
+ {
196
+ "epoch": 1.3157894736842106,
197
+ "grad_norm": 0.0007381394389085472,
198
+ "learning_rate": 4.282632779078051e-05,
199
+ "loss": 0.0,
200
+ "step": 250
201
+ },
202
+ {
203
+ "epoch": 1.368421052631579,
204
+ "grad_norm": 0.000822492060251534,
205
+ "learning_rate": 4.222417297689217e-05,
206
+ "loss": 0.0,
207
+ "step": 260
208
+ },
209
+ {
210
+ "epoch": 1.4210526315789473,
211
+ "grad_norm": 0.001167929614894092,
212
+ "learning_rate": 4.160236506918098e-05,
213
+ "loss": 0.0,
214
+ "step": 270
215
+ },
216
+ {
217
+ "epoch": 1.4736842105263157,
218
+ "grad_norm": 0.0006889003561809659,
219
+ "learning_rate": 4.096161356174959e-05,
220
+ "loss": 0.0,
221
+ "step": 280
222
+ },
223
+ {
224
+ "epoch": 1.526315789473684,
225
+ "grad_norm": 0.0006537819281220436,
226
+ "learning_rate": 4.030264956369157e-05,
227
+ "loss": 0.0,
228
+ "step": 290
229
+ },
230
+ {
231
+ "epoch": 1.5789473684210527,
232
+ "grad_norm": 0.0005895848153159022,
233
+ "learning_rate": 3.962622496488269e-05,
234
+ "loss": 0.0,
235
+ "step": 300
236
+ },
237
+ {
238
+ "epoch": 1.5789473684210527,
239
+ "eval_loss": 1.916586552397348e-05,
240
+ "eval_runtime": 0.9341,
241
+ "eval_samples_per_second": 85.644,
242
+ "eval_steps_per_second": 10.705,
243
+ "step": 300
244
+ },
245
+ {
246
+ "epoch": 1.631578947368421,
247
+ "grad_norm": 0.00041267831693403423,
248
+ "learning_rate": 3.893311157806091e-05,
249
+ "loss": 0.0,
250
+ "step": 310
251
+ },
252
+ {
253
+ "epoch": 1.6842105263157894,
254
+ "grad_norm": 0.00042886921437457204,
255
+ "learning_rate": 3.822410025817406e-05,
256
+ "loss": 0.0,
257
+ "step": 320
258
+ },
259
+ {
260
+ "epoch": 1.736842105263158,
261
+ "grad_norm": 0.0006562267662957311,
262
+ "learning_rate": 3.7500000000000003e-05,
263
+ "loss": 0.0,
264
+ "step": 330
265
+ },
266
+ {
267
+ "epoch": 1.7894736842105263,
268
+ "grad_norm": 0.00045172293903306127,
269
+ "learning_rate": 3.67616370150689e-05,
270
+ "loss": 0.0,
271
+ "step": 340
272
+ },
273
+ {
274
+ "epoch": 1.8421052631578947,
275
+ "grad_norm": 0.0003128020034637302,
276
+ "learning_rate": 3.600985378894086e-05,
277
+ "loss": 0.0,
278
+ "step": 350
279
+ },
280
+ {
281
+ "epoch": 1.8947368421052633,
282
+ "grad_norm": 0.00036892591742798686,
283
+ "learning_rate": 3.5245508119914687e-05,
284
+ "loss": 0.0,
285
+ "step": 360
286
+ },
287
+ {
288
+ "epoch": 1.9473684210526314,
289
+ "grad_norm": 0.00041946041164919734,
290
+ "learning_rate": 3.44694721402644e-05,
291
+ "loss": 0.0,
292
+ "step": 370
293
+ },
294
+ {
295
+ "epoch": 2.0,
296
+ "grad_norm": 0.00044619792606681585,
297
+ "learning_rate": 3.3682631321120504e-05,
298
+ "loss": 0.0,
299
+ "step": 380
300
+ },
301
+ {
302
+ "epoch": 2.0526315789473686,
303
+ "grad_norm": 0.00034757808316498995,
304
+ "learning_rate": 3.2885883462131394e-05,
305
+ "loss": 0.0,
306
+ "step": 390
307
+ },
308
+ {
309
+ "epoch": 2.1052631578947367,
310
+ "grad_norm": 0.0004329316725488752,
311
+ "learning_rate": 3.2080137667057595e-05,
312
+ "loss": 0.0,
313
+ "step": 400
314
+ },
315
+ {
316
+ "epoch": 2.1052631578947367,
317
+ "eval_loss": 1.2211120520078111e-05,
318
+ "eval_runtime": 0.9376,
319
+ "eval_samples_per_second": 85.324,
320
+ "eval_steps_per_second": 10.665,
321
+ "step": 400
322
+ },
323
+ {
324
+ "epoch": 2.1578947368421053,
325
+ "grad_norm": 0.00034806900657713413,
326
+ "learning_rate": 3.126631330646802e-05,
327
+ "loss": 0.0,
328
+ "step": 410
329
+ },
330
+ {
331
+ "epoch": 2.2105263157894735,
332
+ "grad_norm": 0.0003566565574146807,
333
+ "learning_rate": 3.0445338968721287e-05,
334
+ "loss": 0.0,
335
+ "step": 420
336
+ },
337
+ {
338
+ "epoch": 2.263157894736842,
339
+ "grad_norm": 0.00035698447027243674,
340
+ "learning_rate": 2.961815140042974e-05,
341
+ "loss": 0.0,
342
+ "step": 430
343
+ },
344
+ {
345
+ "epoch": 2.3157894736842106,
346
+ "grad_norm": 0.00024169159587472677,
347
+ "learning_rate": 2.878569443761442e-05,
348
+ "loss": 0.0,
349
+ "step": 440
350
+ },
351
+ {
352
+ "epoch": 2.3684210526315788,
353
+ "grad_norm": 0.0003101905749645084,
354
+ "learning_rate": 2.7948917928771158e-05,
355
+ "loss": 0.0,
356
+ "step": 450
357
+ },
358
+ {
359
+ "epoch": 2.4210526315789473,
360
+ "grad_norm": 0.00032829289557412267,
361
+ "learning_rate": 2.7108776651076118e-05,
362
+ "loss": 0.0,
363
+ "step": 460
364
+ },
365
+ {
366
+ "epoch": 2.473684210526316,
367
+ "grad_norm": 0.0002047922316705808,
368
+ "learning_rate": 2.6266229220967818e-05,
369
+ "loss": 0.0,
370
+ "step": 470
371
+ },
372
+ {
373
+ "epoch": 2.526315789473684,
374
+ "grad_norm": 0.00021766829013358802,
375
+ "learning_rate": 2.5422237000348276e-05,
376
+ "loss": 0.0,
377
+ "step": 480
378
+ },
379
+ {
380
+ "epoch": 2.5789473684210527,
381
+ "grad_norm": 0.0002877192455343902,
382
+ "learning_rate": 2.4577762999651726e-05,
383
+ "loss": 0.0,
384
+ "step": 490
385
+ },
386
+ {
387
+ "epoch": 2.6315789473684212,
388
+ "grad_norm": 0.00021653489966411144,
389
+ "learning_rate": 2.3733770779032184e-05,
390
+ "loss": 0.0,
391
+ "step": 500
392
+ },
393
+ {
394
+ "epoch": 2.6315789473684212,
395
+ "eval_loss": 8.995843018055893e-06,
396
+ "eval_runtime": 0.9378,
397
+ "eval_samples_per_second": 85.301,
398
+ "eval_steps_per_second": 10.663,
399
+ "step": 500
400
+ },
401
+ {
402
+ "epoch": 2.6315789473684212,
403
+ "step": 500,
404
+ "total_flos": 3.922877488509747e+16,
405
+ "train_loss": 0.16288311984013126,
406
+ "train_runtime": 177.9523,
407
+ "train_samples_per_second": 42.708,
408
+ "train_steps_per_second": 5.339
409
+ }
410
+ ],
411
+ "logging_steps": 10,
412
+ "max_steps": 950,
413
+ "num_input_tokens_seen": 0,
414
+ "num_train_epochs": 5,
415
+ "save_steps": 100,
416
+ "total_flos": 3.922877488509747e+16,
417
+ "train_batch_size": 4,
418
+ "trial_name": null,
419
+ "trial_params": null
420
+ }
llama2_13b_peft/epistemic_reasoning/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a618e61a3e6a9d9c0aa0b88800edcd7f8840e18dae67d6bea04c44e43cb93243
3
+ size 5176
llama2_13b_peft/epistemic_reasoning/training_eval_loss.png ADDED
llama2_13b_peft/epistemic_reasoning/training_loss.png ADDED
llama2_13b_peft/formal_fallacies_syllogisms_negation/README.md ADDED
@@ -0,0 +1,118 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ library_name: peft
4
+ tags:
5
+ - llama-factory
6
+ - lora
7
+ - generated_from_trainer
8
+ base_model: /data1/model/llama2/meta-llama/Llama2-13b
9
+ model-index:
10
+ - name: formal_fallacies_syllogisms_negation_no_sys
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # formal_fallacies_syllogisms_negation_no_sys
18
+
19
+ This model is a fine-tuned version of [/data1/model/llama2/meta-llama/Llama2-13b](https://huggingface.co//data1/model/llama2/meta-llama/Llama2-13b) on the formal_fallacies_syllogisms_negation_no_sys dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 0.0005
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 0.0001
41
+ - train_batch_size: 8
42
+ - eval_batch_size: 8
43
+ - seed: 42
44
+ - distributed_type: multi-GPU
45
+ - num_devices: 3
46
+ - total_train_batch_size: 24
47
+ - total_eval_batch_size: 24
48
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
49
+ - lr_scheduler_type: cosine
50
+ - lr_scheduler_warmup_steps: 20
51
+ - num_epochs: 10.0
52
+
53
+ ### Training results
54
+
55
+ | Training Loss | Epoch | Step | Validation Loss |
56
+ |:-------------:|:-----:|:----:|:---------------:|
57
+ | 0.3929 | 0.12 | 50 | 0.3600 |
58
+ | 0.3534 | 0.25 | 100 | 0.3503 |
59
+ | 0.3577 | 0.37 | 150 | 0.3543 |
60
+ | 0.351 | 0.5 | 200 | 0.3479 |
61
+ | 0.3459 | 0.62 | 250 | 0.3508 |
62
+ | 0.353 | 0.74 | 300 | 0.3515 |
63
+ | 0.3591 | 0.87 | 350 | 0.3900 |
64
+ | 0.3563 | 0.99 | 400 | 0.3496 |
65
+ | 0.3351 | 1.12 | 450 | 0.3389 |
66
+ | 0.26 | 1.24 | 500 | 0.3002 |
67
+ | 0.187 | 1.36 | 550 | 0.1438 |
68
+ | 0.0715 | 1.49 | 600 | 0.0493 |
69
+ | 0.0695 | 1.61 | 650 | 0.0471 |
70
+ | 0.0338 | 1.74 | 700 | 0.0307 |
71
+ | 0.0422 | 1.86 | 750 | 0.0269 |
72
+ | 0.0183 | 1.99 | 800 | 0.0091 |
73
+ | 0.0055 | 2.11 | 850 | 0.0070 |
74
+ | 0.0208 | 2.23 | 900 | 0.0032 |
75
+ | 0.0014 | 2.36 | 950 | 0.0023 |
76
+ | 0.0117 | 2.48 | 1000 | 0.0034 |
77
+ | 0.0001 | 2.61 | 1050 | 0.0012 |
78
+ | 0.0058 | 2.73 | 1100 | 0.0012 |
79
+ | 0.0001 | 2.85 | 1150 | 0.0013 |
80
+ | 0.0104 | 2.98 | 1200 | 0.0008 |
81
+ | 0.0007 | 3.1 | 1250 | 0.0019 |
82
+ | 0.0015 | 3.23 | 1300 | 0.0092 |
83
+ | 0.0 | 3.35 | 1350 | 0.0016 |
84
+ | 0.0001 | 3.47 | 1400 | 0.0004 |
85
+ | 0.0059 | 3.6 | 1450 | 0.0003 |
86
+ | 0.0 | 3.72 | 1500 | 0.0093 |
87
+ | 0.0059 | 3.85 | 1550 | 0.0032 |
88
+ | 0.0001 | 3.97 | 1600 | 0.0024 |
89
+ | 0.0032 | 4.09 | 1650 | 0.0005 |
90
+ | 0.0 | 4.22 | 1700 | 0.0037 |
91
+ | 0.0 | 4.34 | 1750 | 0.0008 |
92
+ | 0.0 | 4.47 | 1800 | 0.0008 |
93
+ | 0.0 | 4.59 | 1850 | 0.0009 |
94
+ | 0.0 | 4.71 | 1900 | 0.0006 |
95
+ | 0.0002 | 4.84 | 1950 | 0.0001 |
96
+ | 0.0 | 4.96 | 2000 | 0.0008 |
97
+ | 0.0 | 5.09 | 2050 | 0.0001 |
98
+ | 0.0 | 5.21 | 2100 | 0.0002 |
99
+ | 0.0 | 5.33 | 2150 | 0.0002 |
100
+ | 0.0 | 5.46 | 2200 | 0.0004 |
101
+ | 0.0 | 5.58 | 2250 | 0.0004 |
102
+ | 0.0 | 5.71 | 2300 | 0.0004 |
103
+ | 0.0 | 5.83 | 2350 | 0.0004 |
104
+ | 0.0 | 5.96 | 2400 | 0.0004 |
105
+ | 0.0 | 6.08 | 2450 | 0.0004 |
106
+ | 0.0 | 6.2 | 2500 | 0.0005 |
107
+ | 0.0 | 6.33 | 2550 | 0.0005 |
108
+ | 0.0 | 6.45 | 2600 | 0.0005 |
109
+ | 0.0 | 6.58 | 2650 | 0.0004 |
110
+
111
+
112
+ ### Framework versions
113
+
114
+ - PEFT 0.9.0
115
+ - Transformers 4.38.2
116
+ - Pytorch 2.2.1
117
+ - Datasets 2.18.0
118
+ - Tokenizers 0.15.2
llama2_13b_peft/formal_fallacies_syllogisms_negation/adapter_config.json ADDED
@@ -0,0 +1,33 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "/data1/model/llama2/meta-llama/Llama2-13b",
5
+ "bias": "none",
6
+ "fan_in_fan_out": false,
7
+ "inference_mode": true,
8
+ "init_lora_weights": true,
9
+ "layers_pattern": null,
10
+ "layers_to_transform": null,
11
+ "loftq_config": {},
12
+ "lora_alpha": 16,
13
+ "lora_dropout": 0.0,
14
+ "megatron_config": null,
15
+ "megatron_core": "megatron.core",
16
+ "modules_to_save": null,
17
+ "peft_type": "LORA",
18
+ "r": 8,
19
+ "rank_pattern": {},
20
+ "revision": null,
21
+ "target_modules": [
22
+ "up_proj",
23
+ "k_proj",
24
+ "down_proj",
25
+ "q_proj",
26
+ "o_proj",
27
+ "gate_proj",
28
+ "v_proj"
29
+ ],
30
+ "task_type": "CAUSAL_LM",
31
+ "use_dora": false,
32
+ "use_rslora": false
33
+ }
llama2_13b_peft/formal_fallacies_syllogisms_negation/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8e66a5c1fac40b3c4513a47149db4f3097c19c026e1ed290150ba4e0b610028c
3
+ size 125248064
llama2_13b_peft/formal_fallacies_syllogisms_negation/all_results.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 6.58,
3
+ "eval_loss": 0.00047402523341588676,
4
+ "eval_runtime": 16.7834,
5
+ "eval_samples_per_second": 101.529,
6
+ "eval_steps_per_second": 4.23,
7
+ "train_loss": 0.1188159689292002,
8
+ "train_runtime": 2950.0845,
9
+ "train_samples_per_second": 32.731,
10
+ "train_steps_per_second": 1.366
11
+ }
llama2_13b_peft/formal_fallacies_syllogisms_negation/eval_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 6.58,
3
+ "eval_loss": 0.00047402523341588676,
4
+ "eval_runtime": 16.7834,
5
+ "eval_samples_per_second": 101.529,
6
+ "eval_steps_per_second": 4.23
7
+ }
llama2_13b_peft/formal_fallacies_syllogisms_negation/special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "</s>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
llama2_13b_peft/formal_fallacies_syllogisms_negation/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
llama2_13b_peft/formal_fallacies_syllogisms_negation/tokenizer_config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "add_prefix_space": true,
5
+ "added_tokens_decoder": {
6
+ "0": {
7
+ "content": "<unk>",
8
+ "lstrip": false,
9
+ "normalized": false,
10
+ "rstrip": false,
11
+ "single_word": false,
12
+ "special": true
13
+ },
14
+ "1": {
15
+ "content": "<s>",
16
+ "lstrip": false,
17
+ "normalized": false,
18
+ "rstrip": false,
19
+ "single_word": false,
20
+ "special": true
21
+ },
22
+ "2": {
23
+ "content": "</s>",
24
+ "lstrip": false,
25
+ "normalized": false,
26
+ "rstrip": false,
27
+ "single_word": false,
28
+ "special": true
29
+ }
30
+ },
31
+ "bos_token": "<s>",
32
+ "chat_template": "{% if messages[0]['role'] == 'system' %}{% set system_message = messages[0]['content'] %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ content }}{% elif message['role'] == 'assistant' %}{{ content + '\\n' }}{% endif %}{% endfor %}",
33
+ "clean_up_tokenization_spaces": false,
34
+ "eos_token": "</s>",
35
+ "legacy": true,
36
+ "model_max_length": 1000000000000000019884624838656,
37
+ "pad_token": "</s>",
38
+ "padding_side": "right",
39
+ "sp_model_kwargs": {},
40
+ "spaces_between_special_tokens": false,
41
+ "split_special_tokens": false,
42
+ "tokenizer_class": "LlamaTokenizer",
43
+ "unk_token": "<unk>",
44
+ "use_default_system_prompt": false
45
+ }
llama2_13b_peft/formal_fallacies_syllogisms_negation/train_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 6.58,
3
+ "train_loss": 0.1188159689292002,
4
+ "train_runtime": 2950.0845,
5
+ "train_samples_per_second": 32.731,
6
+ "train_steps_per_second": 1.366
7
+ }
llama2_13b_peft/formal_fallacies_syllogisms_negation/trainer_log.jsonl ADDED
@@ -0,0 +1,320 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {"current_steps": 10, "total_steps": 4030, "loss": 9.3847, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5e-05, "epoch": 0.02, "percentage": 0.25, "elapsed_time": "0:00:09", "remaining_time": "1:01:46"}
2
+ {"current_steps": 20, "total_steps": 4030, "loss": 2.5484, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 0.0001, "epoch": 0.05, "percentage": 0.5, "elapsed_time": "0:00:16", "remaining_time": "0:56:23"}
3
+ {"current_steps": 30, "total_steps": 4030, "loss": 0.3827, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.999846556397038e-05, "epoch": 0.07, "percentage": 0.74, "elapsed_time": "0:00:24", "remaining_time": "0:54:18"}
4
+ {"current_steps": 40, "total_steps": 4030, "loss": 0.3554, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.999386235006125e-05, "epoch": 0.1, "percentage": 0.99, "elapsed_time": "0:00:32", "remaining_time": "0:53:16"}
5
+ {"current_steps": 50, "total_steps": 4030, "loss": 0.3929, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.998619064080611e-05, "epoch": 0.12, "percentage": 1.24, "elapsed_time": "0:00:39", "remaining_time": "0:52:50"}
6
+ {"current_steps": 50, "total_steps": 4030, "loss": null, "eval_loss": 0.3599805533885956, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 0.12, "percentage": 1.24, "elapsed_time": "0:00:39", "remaining_time": "0:52:50"}
7
+ {"current_steps": 60, "total_steps": 4030, "loss": 0.3588, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.997545090707485e-05, "epoch": 0.15, "percentage": 1.49, "elapsed_time": "0:01:04", "remaining_time": "1:10:50"}
8
+ {"current_steps": 70, "total_steps": 4030, "loss": 0.3594, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.996164380804483e-05, "epoch": 0.17, "percentage": 1.74, "elapsed_time": "0:01:12", "remaining_time": "1:07:54"}
9
+ {"current_steps": 80, "total_steps": 4030, "loss": 0.3544, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.994477019116049e-05, "epoch": 0.2, "percentage": 1.99, "elapsed_time": "0:01:19", "remaining_time": "1:05:33"}
10
+ {"current_steps": 90, "total_steps": 4030, "loss": 0.3607, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.992483109208122e-05, "epoch": 0.22, "percentage": 2.23, "elapsed_time": "0:01:27", "remaining_time": "1:03:47"}
11
+ {"current_steps": 100, "total_steps": 4030, "loss": 0.3534, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.990182773461791e-05, "epoch": 0.25, "percentage": 2.48, "elapsed_time": "0:01:35", "remaining_time": "1:02:19"}
12
+ {"current_steps": 100, "total_steps": 4030, "loss": null, "eval_loss": 0.35028183460235596, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 0.25, "percentage": 2.48, "elapsed_time": "0:01:35", "remaining_time": "1:02:19"}
13
+ {"current_steps": 110, "total_steps": 4030, "loss": 0.3493, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.987576153065779e-05, "epoch": 0.27, "percentage": 2.73, "elapsed_time": "0:01:59", "remaining_time": "1:11:09"}
14
+ {"current_steps": 120, "total_steps": 4030, "loss": 0.3519, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.984663408007777e-05, "epoch": 0.3, "percentage": 2.98, "elapsed_time": "0:02:07", "remaining_time": "1:09:16"}
15
+ {"current_steps": 130, "total_steps": 4030, "loss": 0.3532, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.98144471706462e-05, "epoch": 0.32, "percentage": 3.23, "elapsed_time": "0:02:15", "remaining_time": "1:07:42"}
16
+ {"current_steps": 140, "total_steps": 4030, "loss": 0.3526, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.977920277791328e-05, "epoch": 0.35, "percentage": 3.47, "elapsed_time": "0:02:22", "remaining_time": "1:06:08"}
17
+ {"current_steps": 150, "total_steps": 4030, "loss": 0.3577, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.974090306508959e-05, "epoch": 0.37, "percentage": 3.72, "elapsed_time": "0:02:30", "remaining_time": "1:04:49"}
18
+ {"current_steps": 150, "total_steps": 4030, "loss": null, "eval_loss": 0.3542585074901581, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 0.37, "percentage": 3.72, "elapsed_time": "0:02:30", "remaining_time": "1:04:49"}
19
+ {"current_steps": 160, "total_steps": 4030, "loss": 0.3525, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.969955038291353e-05, "epoch": 0.4, "percentage": 3.97, "elapsed_time": "0:02:55", "remaining_time": "1:10:38"}
20
+ {"current_steps": 170, "total_steps": 4030, "loss": 0.3493, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.965514726950693e-05, "epoch": 0.42, "percentage": 4.22, "elapsed_time": "0:03:02", "remaining_time": "1:09:10"}
21
+ {"current_steps": 180, "total_steps": 4030, "loss": 0.3515, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.960769645021923e-05, "epoch": 0.45, "percentage": 4.47, "elapsed_time": "0:03:10", "remaining_time": "1:07:58"}
22
+ {"current_steps": 190, "total_steps": 4030, "loss": 0.349, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.955720083746035e-05, "epoch": 0.47, "percentage": 4.71, "elapsed_time": "0:03:18", "remaining_time": "1:06:50"}
23
+ {"current_steps": 200, "total_steps": 4030, "loss": 0.351, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.950366353052175e-05, "epoch": 0.5, "percentage": 4.96, "elapsed_time": "0:03:26", "remaining_time": "1:05:45"}
24
+ {"current_steps": 200, "total_steps": 4030, "loss": null, "eval_loss": 0.3478730320930481, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 0.5, "percentage": 4.96, "elapsed_time": "0:03:26", "remaining_time": "1:05:45"}
25
+ {"current_steps": 210, "total_steps": 4030, "loss": 0.3545, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.944708781538636e-05, "epoch": 0.52, "percentage": 5.21, "elapsed_time": "0:03:50", "remaining_time": "1:09:54"}
26
+ {"current_steps": 220, "total_steps": 4030, "loss": 0.3499, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.93874771645268e-05, "epoch": 0.55, "percentage": 5.46, "elapsed_time": "0:03:58", "remaining_time": "1:08:45"}
27
+ {"current_steps": 230, "total_steps": 4030, "loss": 0.3504, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.932483523669231e-05, "epoch": 0.57, "percentage": 5.71, "elapsed_time": "0:04:05", "remaining_time": "1:07:43"}
28
+ {"current_steps": 240, "total_steps": 4030, "loss": 0.3543, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.925916587668409e-05, "epoch": 0.6, "percentage": 5.96, "elapsed_time": "0:04:14", "remaining_time": "1:06:52"}
29
+ {"current_steps": 250, "total_steps": 4030, "loss": 0.3459, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.919047311511946e-05, "epoch": 0.62, "percentage": 6.2, "elapsed_time": "0:04:21", "remaining_time": "1:05:58"}
30
+ {"current_steps": 250, "total_steps": 4030, "loss": null, "eval_loss": 0.35084518790245056, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 0.62, "percentage": 6.2, "elapsed_time": "0:04:21", "remaining_time": "1:05:58"}
31
+ {"current_steps": 260, "total_steps": 4030, "loss": 0.3483, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.911876116818433e-05, "epoch": 0.65, "percentage": 6.45, "elapsed_time": "0:04:46", "remaining_time": "1:09:16"}
32
+ {"current_steps": 270, "total_steps": 4030, "loss": 0.3577, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.904403443737454e-05, "epoch": 0.67, "percentage": 6.7, "elapsed_time": "0:04:54", "remaining_time": "1:08:15"}
33
+ {"current_steps": 280, "total_steps": 4030, "loss": 0.3656, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.896629750922555e-05, "epoch": 0.69, "percentage": 6.95, "elapsed_time": "0:05:01", "remaining_time": "1:07:20"}
34
+ {"current_steps": 290, "total_steps": 4030, "loss": 0.3595, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.888555515503116e-05, "epoch": 0.72, "percentage": 7.2, "elapsed_time": "0:05:09", "remaining_time": "1:06:30"}
35
+ {"current_steps": 300, "total_steps": 4030, "loss": 0.353, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.880181233055042e-05, "epoch": 0.74, "percentage": 7.44, "elapsed_time": "0:05:16", "remaining_time": "1:05:40"}
36
+ {"current_steps": 300, "total_steps": 4030, "loss": null, "eval_loss": 0.351527601480484, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 0.74, "percentage": 7.44, "elapsed_time": "0:05:16", "remaining_time": "1:05:40"}
37
+ {"current_steps": 310, "total_steps": 4030, "loss": 0.3552, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.871507417570363e-05, "epoch": 0.77, "percentage": 7.69, "elapsed_time": "0:05:41", "remaining_time": "1:08:16"}
38
+ {"current_steps": 320, "total_steps": 4030, "loss": 0.3515, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.862534601425681e-05, "epoch": 0.79, "percentage": 7.94, "elapsed_time": "0:05:49", "remaining_time": "1:07:27"}
39
+ {"current_steps": 330, "total_steps": 4030, "loss": 0.3445, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.853263335349488e-05, "epoch": 0.82, "percentage": 8.19, "elapsed_time": "0:05:56", "remaining_time": "1:06:40"}
40
+ {"current_steps": 340, "total_steps": 4030, "loss": 0.3579, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.843694188388374e-05, "epoch": 0.84, "percentage": 8.44, "elapsed_time": "0:06:04", "remaining_time": "1:05:57"}
41
+ {"current_steps": 350, "total_steps": 4030, "loss": 0.3591, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.833827747872093e-05, "epoch": 0.87, "percentage": 8.68, "elapsed_time": "0:06:12", "remaining_time": "1:05:13"}
42
+ {"current_steps": 350, "total_steps": 4030, "loss": null, "eval_loss": 0.3899991512298584, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 0.87, "percentage": 8.68, "elapsed_time": "0:06:12", "remaining_time": "1:05:13"}
43
+ {"current_steps": 360, "total_steps": 4030, "loss": 0.3472, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.823664619377518e-05, "epoch": 0.89, "percentage": 8.93, "elapsed_time": "0:06:36", "remaining_time": "1:07:25"}
44
+ {"current_steps": 370, "total_steps": 4030, "loss": 0.3515, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.81320542669147e-05, "epoch": 0.92, "percentage": 9.18, "elapsed_time": "0:06:44", "remaining_time": "1:06:42"}
45
+ {"current_steps": 380, "total_steps": 4030, "loss": 0.3591, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.802450811772435e-05, "epoch": 0.94, "percentage": 9.43, "elapsed_time": "0:06:52", "remaining_time": "1:05:59"}
46
+ {"current_steps": 390, "total_steps": 4030, "loss": 0.3511, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.791401434711154e-05, "epoch": 0.97, "percentage": 9.68, "elapsed_time": "0:06:59", "remaining_time": "1:05:19"}
47
+ {"current_steps": 400, "total_steps": 4030, "loss": 0.3563, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.78005797369012e-05, "epoch": 0.99, "percentage": 9.93, "elapsed_time": "0:07:07", "remaining_time": "1:04:39"}
48
+ {"current_steps": 400, "total_steps": 4030, "loss": null, "eval_loss": 0.3495555818080902, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 0.99, "percentage": 9.93, "elapsed_time": "0:07:07", "remaining_time": "1:04:39"}
49
+ {"current_steps": 410, "total_steps": 4030, "loss": 0.3384, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.768421124941944e-05, "epoch": 1.02, "percentage": 10.17, "elapsed_time": "0:07:32", "remaining_time": "1:06:32"}
50
+ {"current_steps": 420, "total_steps": 4030, "loss": 0.3545, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.756491602706626e-05, "epoch": 1.04, "percentage": 10.42, "elapsed_time": "0:07:39", "remaining_time": "1:05:51"}
51
+ {"current_steps": 430, "total_steps": 4030, "loss": 0.3449, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.744270139187716e-05, "epoch": 1.07, "percentage": 10.67, "elapsed_time": "0:07:47", "remaining_time": "1:05:14"}
52
+ {"current_steps": 440, "total_steps": 4030, "loss": 0.346, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.731757484507371e-05, "epoch": 1.09, "percentage": 10.92, "elapsed_time": "0:07:55", "remaining_time": "1:04:37"}
53
+ {"current_steps": 450, "total_steps": 4030, "loss": 0.3351, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.718954406660318e-05, "epoch": 1.12, "percentage": 11.17, "elapsed_time": "0:08:02", "remaining_time": "1:04:01"}
54
+ {"current_steps": 450, "total_steps": 4030, "loss": null, "eval_loss": 0.33888280391693115, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 1.12, "percentage": 11.17, "elapsed_time": "0:08:02", "remaining_time": "1:04:01"}
55
+ {"current_steps": 460, "total_steps": 4030, "loss": 0.3306, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.705861691466716e-05, "epoch": 1.14, "percentage": 11.41, "elapsed_time": "0:08:27", "remaining_time": "1:05:39"}
56
+ {"current_steps": 470, "total_steps": 4030, "loss": 0.3514, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.692480142523918e-05, "epoch": 1.17, "percentage": 11.66, "elapsed_time": "0:08:35", "remaining_time": "1:05:01"}
57
+ {"current_steps": 480, "total_steps": 4030, "loss": 0.3159, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.678810581157163e-05, "epoch": 1.19, "percentage": 11.91, "elapsed_time": "0:08:42", "remaining_time": "1:04:24"}
58
+ {"current_steps": 490, "total_steps": 4030, "loss": 0.309, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.664853846369144e-05, "epoch": 1.22, "percentage": 12.16, "elapsed_time": "0:08:50", "remaining_time": "1:03:50"}
59
+ {"current_steps": 500, "total_steps": 4030, "loss": 0.26, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.650610794788533e-05, "epoch": 1.24, "percentage": 12.41, "elapsed_time": "0:08:58", "remaining_time": "1:03:19"}
60
+ {"current_steps": 500, "total_steps": 4030, "loss": null, "eval_loss": 0.3002278506755829, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 1.24, "percentage": 12.41, "elapsed_time": "0:08:58", "remaining_time": "1:03:19"}
61
+ {"current_steps": 510, "total_steps": 4030, "loss": 0.2472, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.63608230061739e-05, "epoch": 1.27, "percentage": 12.66, "elapsed_time": "0:09:23", "remaining_time": "1:04:47"}
62
+ {"current_steps": 520, "total_steps": 4030, "loss": 0.2212, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.621269255577512e-05, "epoch": 1.29, "percentage": 12.9, "elapsed_time": "0:09:30", "remaining_time": "1:04:12"}
63
+ {"current_steps": 530, "total_steps": 4030, "loss": 0.2039, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.606172568855698e-05, "epoch": 1.32, "percentage": 13.15, "elapsed_time": "0:09:38", "remaining_time": "1:03:38"}
64
+ {"current_steps": 540, "total_steps": 4030, "loss": 0.2198, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.590793167047951e-05, "epoch": 1.34, "percentage": 13.4, "elapsed_time": "0:09:46", "remaining_time": "1:03:07"}
65
+ {"current_steps": 550, "total_steps": 4030, "loss": 0.187, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.575131994102602e-05, "epoch": 1.36, "percentage": 13.65, "elapsed_time": "0:09:53", "remaining_time": "1:02:37"}
66
+ {"current_steps": 550, "total_steps": 4030, "loss": null, "eval_loss": 0.14382539689540863, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 1.36, "percentage": 13.65, "elapsed_time": "0:09:53", "remaining_time": "1:02:37"}
67
+ {"current_steps": 560, "total_steps": 4030, "loss": 0.1253, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.559190011262369e-05, "epoch": 1.39, "percentage": 13.9, "elapsed_time": "0:10:18", "remaining_time": "1:03:50"}
68
+ {"current_steps": 570, "total_steps": 4030, "loss": 0.1246, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.542968197005367e-05, "epoch": 1.41, "percentage": 14.14, "elapsed_time": "0:10:26", "remaining_time": "1:03:20"}
69
+ {"current_steps": 580, "total_steps": 4030, "loss": 0.1097, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.526467546985048e-05, "epoch": 1.44, "percentage": 14.39, "elapsed_time": "0:10:33", "remaining_time": "1:02:49"}
70
+ {"current_steps": 590, "total_steps": 4030, "loss": 0.0995, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.509689073969088e-05, "epoch": 1.46, "percentage": 14.64, "elapsed_time": "0:10:41", "remaining_time": "1:02:20"}
71
+ {"current_steps": 600, "total_steps": 4030, "loss": 0.0715, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.492633807777227e-05, "epoch": 1.49, "percentage": 14.89, "elapsed_time": "0:10:49", "remaining_time": "1:01:52"}
72
+ {"current_steps": 600, "total_steps": 4030, "loss": null, "eval_loss": 0.0493486262857914, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 1.49, "percentage": 14.89, "elapsed_time": "0:10:49", "remaining_time": "1:01:52"}
73
+ {"current_steps": 610, "total_steps": 4030, "loss": 0.0715, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.475302795218061e-05, "epoch": 1.51, "percentage": 15.14, "elapsed_time": "0:11:14", "remaining_time": "1:02:59"}
74
+ {"current_steps": 620, "total_steps": 4030, "loss": 0.0797, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.457697100024798e-05, "epoch": 1.54, "percentage": 15.38, "elapsed_time": "0:11:22", "remaining_time": "1:02:31"}
75
+ {"current_steps": 630, "total_steps": 4030, "loss": 0.065, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.439817802789958e-05, "epoch": 1.56, "percentage": 15.63, "elapsed_time": "0:11:29", "remaining_time": "1:02:03"}
76
+ {"current_steps": 640, "total_steps": 4030, "loss": 0.0534, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.421666000899052e-05, "epoch": 1.59, "percentage": 15.88, "elapsed_time": "0:11:37", "remaining_time": "1:01:33"}
77
+ {"current_steps": 650, "total_steps": 4030, "loss": 0.0695, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.403242808463237e-05, "epoch": 1.61, "percentage": 16.13, "elapsed_time": "0:11:44", "remaining_time": "1:01:05"}
78
+ {"current_steps": 650, "total_steps": 4030, "loss": null, "eval_loss": 0.047055672854185104, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 1.61, "percentage": 16.13, "elapsed_time": "0:11:44", "remaining_time": "1:01:05"}
79
+ {"current_steps": 660, "total_steps": 4030, "loss": 0.0259, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.38454935625092e-05, "epoch": 1.64, "percentage": 16.38, "elapsed_time": "0:12:09", "remaining_time": "1:02:06"}
80
+ {"current_steps": 670, "total_steps": 4030, "loss": 0.0388, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.365586791618368e-05, "epoch": 1.66, "percentage": 16.63, "elapsed_time": "0:12:17", "remaining_time": "1:01:38"}
81
+ {"current_steps": 680, "total_steps": 4030, "loss": 0.0514, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.346356278439275e-05, "epoch": 1.69, "percentage": 16.87, "elapsed_time": "0:12:25", "remaining_time": "1:01:13"}
82
+ {"current_steps": 690, "total_steps": 4030, "loss": 0.0382, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.32685899703333e-05, "epoch": 1.71, "percentage": 17.12, "elapsed_time": "0:12:33", "remaining_time": "1:00:46"}
83
+ {"current_steps": 700, "total_steps": 4030, "loss": 0.0338, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.307096144093781e-05, "epoch": 1.74, "percentage": 17.37, "elapsed_time": "0:12:41", "remaining_time": "1:00:21"}
84
+ {"current_steps": 700, "total_steps": 4030, "loss": null, "eval_loss": 0.030666794627904892, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 1.74, "percentage": 17.37, "elapsed_time": "0:12:41", "remaining_time": "1:00:21"}
85
+ {"current_steps": 710, "total_steps": 4030, "loss": 0.0635, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.287068932613968e-05, "epoch": 1.76, "percentage": 17.62, "elapsed_time": "0:13:05", "remaining_time": "1:01:14"}
86
+ {"current_steps": 720, "total_steps": 4030, "loss": 0.0303, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.266778591812886e-05, "epoch": 1.79, "percentage": 17.87, "elapsed_time": "0:13:13", "remaining_time": "1:00:47"}
87
+ {"current_steps": 730, "total_steps": 4030, "loss": 0.0094, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.246226367059736e-05, "epoch": 1.81, "percentage": 18.11, "elapsed_time": "0:13:21", "remaining_time": "1:00:21"}
88
+ {"current_steps": 740, "total_steps": 4030, "loss": 0.0217, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.225413519797482e-05, "epoch": 1.84, "percentage": 18.36, "elapsed_time": "0:13:28", "remaining_time": "0:59:55"}
89
+ {"current_steps": 750, "total_steps": 4030, "loss": 0.0422, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.204341327465434e-05, "epoch": 1.86, "percentage": 18.61, "elapsed_time": "0:13:36", "remaining_time": "0:59:30"}
90
+ {"current_steps": 750, "total_steps": 4030, "loss": null, "eval_loss": 0.02686377428472042, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 1.86, "percentage": 18.61, "elapsed_time": "0:13:36", "remaining_time": "0:59:30"}
91
+ {"current_steps": 760, "total_steps": 4030, "loss": 0.046, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.183011083420838e-05, "epoch": 1.89, "percentage": 18.86, "elapsed_time": "0:14:00", "remaining_time": "1:00:17"}
92
+ {"current_steps": 770, "total_steps": 4030, "loss": 0.0192, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.161424096859493e-05, "epoch": 1.91, "percentage": 19.11, "elapsed_time": "0:14:08", "remaining_time": "0:59:52"}
93
+ {"current_steps": 780, "total_steps": 4030, "loss": 0.0159, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.139581692735394e-05, "epoch": 1.94, "percentage": 19.35, "elapsed_time": "0:14:16", "remaining_time": "0:59:27"}
94
+ {"current_steps": 790, "total_steps": 4030, "loss": 0.0188, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.11748521167942e-05, "epoch": 1.96, "percentage": 19.6, "elapsed_time": "0:14:24", "remaining_time": "0:59:04"}
95
+ {"current_steps": 800, "total_steps": 4030, "loss": 0.0183, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.095136009917036e-05, "epoch": 1.99, "percentage": 19.85, "elapsed_time": "0:14:31", "remaining_time": "0:58:39"}
96
+ {"current_steps": 800, "total_steps": 4030, "loss": null, "eval_loss": 0.009065705351531506, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 1.99, "percentage": 19.85, "elapsed_time": "0:14:31", "remaining_time": "0:58:39"}
97
+ {"current_steps": 810, "total_steps": 4030, "loss": 0.0317, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.072535459185056e-05, "epoch": 2.01, "percentage": 20.1, "elapsed_time": "0:14:56", "remaining_time": "0:59:23"}
98
+ {"current_steps": 820, "total_steps": 4030, "loss": 0.0059, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.049684946647459e-05, "epoch": 2.03, "percentage": 20.35, "elapsed_time": "0:15:04", "remaining_time": "0:58:59"}
99
+ {"current_steps": 830, "total_steps": 4030, "loss": 0.0009, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.026585874810227e-05, "epoch": 2.06, "percentage": 20.6, "elapsed_time": "0:15:11", "remaining_time": "0:58:35"}
100
+ {"current_steps": 840, "total_steps": 4030, "loss": 0.0015, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 9.003239661435289e-05, "epoch": 2.08, "percentage": 20.84, "elapsed_time": "0:15:19", "remaining_time": "0:58:13"}
101
+ {"current_steps": 850, "total_steps": 4030, "loss": 0.0055, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.979647739453481e-05, "epoch": 2.11, "percentage": 21.09, "elapsed_time": "0:15:27", "remaining_time": "0:57:50"}
102
+ {"current_steps": 850, "total_steps": 4030, "loss": null, "eval_loss": 0.007036147639155388, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 2.11, "percentage": 21.09, "elapsed_time": "0:15:27", "remaining_time": "0:57:50"}
103
+ {"current_steps": 860, "total_steps": 4030, "loss": 0.0002, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.955811556876605e-05, "epoch": 2.13, "percentage": 21.34, "elapsed_time": "0:15:51", "remaining_time": "0:58:29"}
104
+ {"current_steps": 870, "total_steps": 4030, "loss": 0.0172, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.93173257670856e-05, "epoch": 2.16, "percentage": 21.59, "elapsed_time": "0:15:59", "remaining_time": "0:58:05"}
105
+ {"current_steps": 880, "total_steps": 4030, "loss": 0.0014, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.907412276855531e-05, "epoch": 2.18, "percentage": 21.84, "elapsed_time": "0:16:07", "remaining_time": "0:57:42"}
106
+ {"current_steps": 890, "total_steps": 4030, "loss": 0.0025, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.882852150035295e-05, "epoch": 2.21, "percentage": 22.08, "elapsed_time": "0:16:15", "remaining_time": "0:57:20"}
107
+ {"current_steps": 900, "total_steps": 4030, "loss": 0.0208, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.858053703685587e-05, "epoch": 2.23, "percentage": 22.33, "elapsed_time": "0:16:22", "remaining_time": "0:56:57"}
108
+ {"current_steps": 900, "total_steps": 4030, "loss": null, "eval_loss": 0.003154078498482704, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 2.23, "percentage": 22.33, "elapsed_time": "0:16:22", "remaining_time": "0:56:57"}
109
+ {"current_steps": 910, "total_steps": 4030, "loss": 0.0016, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.833018459871594e-05, "epoch": 2.26, "percentage": 22.58, "elapsed_time": "0:16:47", "remaining_time": "0:57:32"}
110
+ {"current_steps": 920, "total_steps": 4030, "loss": 0.0139, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.807747955192516e-05, "epoch": 2.28, "percentage": 22.83, "elapsed_time": "0:16:55", "remaining_time": "0:57:11"}
111
+ {"current_steps": 930, "total_steps": 4030, "loss": 0.0007, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.782243740687273e-05, "epoch": 2.31, "percentage": 23.08, "elapsed_time": "0:17:02", "remaining_time": "0:56:49"}
112
+ {"current_steps": 940, "total_steps": 4030, "loss": 0.0149, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.756507381739288e-05, "epoch": 2.33, "percentage": 23.33, "elapsed_time": "0:17:10", "remaining_time": "0:56:28"}
113
+ {"current_steps": 950, "total_steps": 4030, "loss": 0.0014, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.730540457980418e-05, "epoch": 2.36, "percentage": 23.57, "elapsed_time": "0:17:18", "remaining_time": "0:56:06"}
114
+ {"current_steps": 950, "total_steps": 4030, "loss": null, "eval_loss": 0.002297737170010805, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 2.36, "percentage": 23.57, "elapsed_time": "0:17:18", "remaining_time": "0:56:06"}
115
+ {"current_steps": 960, "total_steps": 4030, "loss": 0.0014, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.704344563193997e-05, "epoch": 2.38, "percentage": 23.82, "elapsed_time": "0:17:43", "remaining_time": "0:56:40"}
116
+ {"current_steps": 970, "total_steps": 4030, "loss": 0.0002, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.677921305217023e-05, "epoch": 2.41, "percentage": 24.07, "elapsed_time": "0:17:50", "remaining_time": "0:56:18"}
117
+ {"current_steps": 980, "total_steps": 4030, "loss": 0.0207, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.651272305841454e-05, "epoch": 2.43, "percentage": 24.32, "elapsed_time": "0:17:58", "remaining_time": "0:55:57"}
118
+ {"current_steps": 990, "total_steps": 4030, "loss": 0.0011, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.62439920071468e-05, "epoch": 2.46, "percentage": 24.57, "elapsed_time": "0:18:06", "remaining_time": "0:55:35"}
119
+ {"current_steps": 1000, "total_steps": 4030, "loss": 0.0117, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.597303639239134e-05, "epoch": 2.48, "percentage": 24.81, "elapsed_time": "0:18:13", "remaining_time": "0:55:14"}
120
+ {"current_steps": 1000, "total_steps": 4030, "loss": null, "eval_loss": 0.0033970780204981565, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 2.48, "percentage": 24.81, "elapsed_time": "0:18:13", "remaining_time": "0:55:14"}
121
+ {"current_steps": 1010, "total_steps": 4030, "loss": 0.0026, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.569987284471046e-05, "epoch": 2.51, "percentage": 25.06, "elapsed_time": "0:18:39", "remaining_time": "0:55:45"}
122
+ {"current_steps": 1020, "total_steps": 4030, "loss": 0.0022, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.542451813018372e-05, "epoch": 2.53, "percentage": 25.31, "elapsed_time": "0:18:46", "remaining_time": "0:55:25"}
123
+ {"current_steps": 1030, "total_steps": 4030, "loss": 0.0002, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.514698914937894e-05, "epoch": 2.56, "percentage": 25.56, "elapsed_time": "0:18:54", "remaining_time": "0:55:04"}
124
+ {"current_steps": 1040, "total_steps": 4030, "loss": 0.0019, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.486730293631481e-05, "epoch": 2.58, "percentage": 25.81, "elapsed_time": "0:19:02", "remaining_time": "0:54:44"}
125
+ {"current_steps": 1050, "total_steps": 4030, "loss": 0.0001, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.458547665741543e-05, "epoch": 2.61, "percentage": 26.05, "elapsed_time": "0:19:09", "remaining_time": "0:54:23"}
126
+ {"current_steps": 1050, "total_steps": 4030, "loss": null, "eval_loss": 0.001217635115608573, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 2.61, "percentage": 26.05, "elapsed_time": "0:19:09", "remaining_time": "0:54:23"}
127
+ {"current_steps": 1060, "total_steps": 4030, "loss": 0.0002, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.430152761045664e-05, "epoch": 2.63, "percentage": 26.3, "elapsed_time": "0:19:34", "remaining_time": "0:54:50"}
128
+ {"current_steps": 1070, "total_steps": 4030, "loss": 0.0004, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.401547322350438e-05, "epoch": 2.66, "percentage": 26.55, "elapsed_time": "0:19:42", "remaining_time": "0:54:31"}
129
+ {"current_steps": 1080, "total_steps": 4030, "loss": 0.0002, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.372733105384496e-05, "epoch": 2.68, "percentage": 26.8, "elapsed_time": "0:19:50", "remaining_time": "0:54:11"}
130
+ {"current_steps": 1090, "total_steps": 4030, "loss": 0.0171, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.343711878690746e-05, "epoch": 2.7, "percentage": 27.05, "elapsed_time": "0:19:57", "remaining_time": "0:53:51"}
131
+ {"current_steps": 1100, "total_steps": 4030, "loss": 0.0058, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.314485423517819e-05, "epoch": 2.73, "percentage": 27.3, "elapsed_time": "0:20:05", "remaining_time": "0:53:31"}
132
+ {"current_steps": 1100, "total_steps": 4030, "loss": null, "eval_loss": 0.0012037238338962197, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 2.73, "percentage": 27.3, "elapsed_time": "0:20:05", "remaining_time": "0:53:31"}
133
+ {"current_steps": 1110, "total_steps": 4030, "loss": 0.0001, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.285055533710753e-05, "epoch": 2.75, "percentage": 27.54, "elapsed_time": "0:20:30", "remaining_time": "0:53:56"}
134
+ {"current_steps": 1120, "total_steps": 4030, "loss": 0.0059, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.255424015600876e-05, "epoch": 2.78, "percentage": 27.79, "elapsed_time": "0:20:38", "remaining_time": "0:53:37"}
135
+ {"current_steps": 1130, "total_steps": 4030, "loss": 0.0028, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.225592687894949e-05, "epoch": 2.8, "percentage": 28.04, "elapsed_time": "0:20:45", "remaining_time": "0:53:16"}
136
+ {"current_steps": 1140, "total_steps": 4030, "loss": 0.0001, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.195563381563536e-05, "epoch": 2.83, "percentage": 28.29, "elapsed_time": "0:20:53", "remaining_time": "0:52:57"}
137
+ {"current_steps": 1150, "total_steps": 4030, "loss": 0.0001, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.165337939728614e-05, "epoch": 2.85, "percentage": 28.54, "elapsed_time": "0:21:00", "remaining_time": "0:52:37"}
138
+ {"current_steps": 1150, "total_steps": 4030, "loss": null, "eval_loss": 0.0012541119940578938, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 2.85, "percentage": 28.54, "elapsed_time": "0:21:00", "remaining_time": "0:52:37"}
139
+ {"current_steps": 1160, "total_steps": 4030, "loss": 0.0001, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.134918217550467e-05, "epoch": 2.88, "percentage": 28.78, "elapsed_time": "0:21:25", "remaining_time": "0:53:01"}
140
+ {"current_steps": 1170, "total_steps": 4030, "loss": 0.0017, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.104306082113802e-05, "epoch": 2.9, "percentage": 29.03, "elapsed_time": "0:21:33", "remaining_time": "0:52:41"}
141
+ {"current_steps": 1180, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.073503412313159e-05, "epoch": 2.93, "percentage": 29.28, "elapsed_time": "0:21:41", "remaining_time": "0:52:22"}
142
+ {"current_steps": 1190, "total_steps": 4030, "loss": 0.001, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.042512098737598e-05, "epoch": 2.95, "percentage": 29.53, "elapsed_time": "0:21:48", "remaining_time": "0:52:02"}
143
+ {"current_steps": 1200, "total_steps": 4030, "loss": 0.0104, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 8.01133404355464e-05, "epoch": 2.98, "percentage": 29.78, "elapsed_time": "0:21:56", "remaining_time": "0:51:44"}
144
+ {"current_steps": 1200, "total_steps": 4030, "loss": null, "eval_loss": 0.0007884196820668876, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 2.98, "percentage": 29.78, "elapsed_time": "0:21:56", "remaining_time": "0:51:44"}
145
+ {"current_steps": 1210, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.979971160393535e-05, "epoch": 3.0, "percentage": 30.02, "elapsed_time": "0:22:20", "remaining_time": "0:52:05"}
146
+ {"current_steps": 1220, "total_steps": 4030, "loss": 0.0131, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.948425374227799e-05, "epoch": 3.03, "percentage": 30.27, "elapsed_time": "0:22:28", "remaining_time": "0:51:46"}
147
+ {"current_steps": 1230, "total_steps": 4030, "loss": 0.0006, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.916698621257068e-05, "epoch": 3.05, "percentage": 30.52, "elapsed_time": "0:22:36", "remaining_time": "0:51:27"}
148
+ {"current_steps": 1240, "total_steps": 4030, "loss": 0.0028, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.884792848788257e-05, "epoch": 3.08, "percentage": 30.77, "elapsed_time": "0:22:43", "remaining_time": "0:51:08"}
149
+ {"current_steps": 1250, "total_steps": 4030, "loss": 0.0007, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.852710015116036e-05, "epoch": 3.1, "percentage": 31.02, "elapsed_time": "0:22:51", "remaining_time": "0:50:50"}
150
+ {"current_steps": 1250, "total_steps": 4030, "loss": null, "eval_loss": 0.0018942987080663443, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 3.1, "percentage": 31.02, "elapsed_time": "0:22:51", "remaining_time": "0:50:50"}
151
+ {"current_steps": 1260, "total_steps": 4030, "loss": 0.0001, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.820452089402645e-05, "epoch": 3.13, "percentage": 31.27, "elapsed_time": "0:23:16", "remaining_time": "0:51:10"}
152
+ {"current_steps": 1270, "total_steps": 4030, "loss": 0.0015, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.788021051557022e-05, "epoch": 3.15, "percentage": 31.51, "elapsed_time": "0:23:24", "remaining_time": "0:50:52"}
153
+ {"current_steps": 1280, "total_steps": 4030, "loss": 0.0006, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.755418892113282e-05, "epoch": 3.18, "percentage": 31.76, "elapsed_time": "0:23:32", "remaining_time": "0:50:33"}
154
+ {"current_steps": 1290, "total_steps": 4030, "loss": 0.0011, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.722647612108553e-05, "epoch": 3.2, "percentage": 32.01, "elapsed_time": "0:23:39", "remaining_time": "0:50:15"}
155
+ {"current_steps": 1300, "total_steps": 4030, "loss": 0.0015, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.689709222960144e-05, "epoch": 3.23, "percentage": 32.26, "elapsed_time": "0:23:47", "remaining_time": "0:49:57"}
156
+ {"current_steps": 1300, "total_steps": 4030, "loss": null, "eval_loss": 0.00915015209466219, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 3.23, "percentage": 32.26, "elapsed_time": "0:23:47", "remaining_time": "0:49:57"}
157
+ {"current_steps": 1310, "total_steps": 4030, "loss": 0.0118, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.656605746342096e-05, "epoch": 3.25, "percentage": 32.51, "elapsed_time": "0:24:12", "remaining_time": "0:50:14"}
158
+ {"current_steps": 1320, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.6233392140611e-05, "epoch": 3.28, "percentage": 32.75, "elapsed_time": "0:24:19", "remaining_time": "0:49:56"}
159
+ {"current_steps": 1330, "total_steps": 4030, "loss": 0.0001, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.589911667931786e-05, "epoch": 3.3, "percentage": 33.0, "elapsed_time": "0:24:27", "remaining_time": "0:49:38"}
160
+ {"current_steps": 1340, "total_steps": 4030, "loss": 0.0001, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.556325159651399e-05, "epoch": 3.33, "percentage": 33.25, "elapsed_time": "0:24:35", "remaining_time": "0:49:21"}
161
+ {"current_steps": 1350, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.522581750673872e-05, "epoch": 3.35, "percentage": 33.5, "elapsed_time": "0:24:42", "remaining_time": "0:49:03"}
162
+ {"current_steps": 1350, "total_steps": 4030, "loss": null, "eval_loss": 0.0016158341895788908, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 3.35, "percentage": 33.5, "elapsed_time": "0:24:42", "remaining_time": "0:49:03"}
163
+ {"current_steps": 1360, "total_steps": 4030, "loss": 0.0005, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.48868351208331e-05, "epoch": 3.37, "percentage": 33.75, "elapsed_time": "0:25:07", "remaining_time": "0:49:19"}
164
+ {"current_steps": 1370, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.454632524466857e-05, "epoch": 3.4, "percentage": 34.0, "elapsed_time": "0:25:15", "remaining_time": "0:49:02"}
165
+ {"current_steps": 1380, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.420430877787004e-05, "epoch": 3.42, "percentage": 34.24, "elapsed_time": "0:25:22", "remaining_time": "0:48:44"}
166
+ {"current_steps": 1390, "total_steps": 4030, "loss": 0.004, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.386080671253307e-05, "epoch": 3.45, "percentage": 34.49, "elapsed_time": "0:25:30", "remaining_time": "0:48:26"}
167
+ {"current_steps": 1400, "total_steps": 4030, "loss": 0.0001, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.351584013193543e-05, "epoch": 3.47, "percentage": 34.74, "elapsed_time": "0:25:38", "remaining_time": "0:48:10"}
168
+ {"current_steps": 1400, "total_steps": 4030, "loss": null, "eval_loss": 0.000445638841483742, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 3.47, "percentage": 34.74, "elapsed_time": "0:25:38", "remaining_time": "0:48:10"}
169
+ {"current_steps": 1410, "total_steps": 4030, "loss": 0.0003, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.316943020924319e-05, "epoch": 3.5, "percentage": 34.99, "elapsed_time": "0:26:03", "remaining_time": "0:48:24"}
170
+ {"current_steps": 1420, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.2821598206211e-05, "epoch": 3.52, "percentage": 35.24, "elapsed_time": "0:26:10", "remaining_time": "0:48:07"}
171
+ {"current_steps": 1430, "total_steps": 4030, "loss": 0.0002, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.247236547187716e-05, "epoch": 3.55, "percentage": 35.48, "elapsed_time": "0:26:18", "remaining_time": "0:47:49"}
172
+ {"current_steps": 1440, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.212175344125327e-05, "epoch": 3.57, "percentage": 35.73, "elapsed_time": "0:26:26", "remaining_time": "0:47:32"}
173
+ {"current_steps": 1450, "total_steps": 4030, "loss": 0.0059, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.176978363400863e-05, "epoch": 3.6, "percentage": 35.98, "elapsed_time": "0:26:33", "remaining_time": "0:47:16"}
174
+ {"current_steps": 1450, "total_steps": 4030, "loss": null, "eval_loss": 0.0003264049591962248, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 3.6, "percentage": 35.98, "elapsed_time": "0:26:33", "remaining_time": "0:47:16"}
175
+ {"current_steps": 1460, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.141647765314938e-05, "epoch": 3.62, "percentage": 36.23, "elapsed_time": "0:26:58", "remaining_time": "0:47:28"}
176
+ {"current_steps": 1470, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.106185718369259e-05, "epoch": 3.65, "percentage": 36.48, "elapsed_time": "0:27:06", "remaining_time": "0:47:11"}
177
+ {"current_steps": 1480, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.070594399133525e-05, "epoch": 3.67, "percentage": 36.72, "elapsed_time": "0:27:13", "remaining_time": "0:46:55"}
178
+ {"current_steps": 1490, "total_steps": 4030, "loss": 0.0038, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 7.03487599211184e-05, "epoch": 3.7, "percentage": 36.97, "elapsed_time": "0:27:21", "remaining_time": "0:46:38"}
179
+ {"current_steps": 1500, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.999032689608629e-05, "epoch": 3.72, "percentage": 37.22, "elapsed_time": "0:27:29", "remaining_time": "0:46:22"}
180
+ {"current_steps": 1500, "total_steps": 4030, "loss": null, "eval_loss": 0.009298658929765224, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 3.72, "percentage": 37.22, "elapsed_time": "0:27:29", "remaining_time": "0:46:22"}
181
+ {"current_steps": 1510, "total_steps": 4030, "loss": 0.0037, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.963066691594084e-05, "epoch": 3.75, "percentage": 37.47, "elapsed_time": "0:27:54", "remaining_time": "0:46:34"}
182
+ {"current_steps": 1520, "total_steps": 4030, "loss": 0.0113, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.926980205569136e-05, "epoch": 3.77, "percentage": 37.72, "elapsed_time": "0:28:02", "remaining_time": "0:46:17"}
183
+ {"current_steps": 1530, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.890775446429954e-05, "epoch": 3.8, "percentage": 37.97, "elapsed_time": "0:28:09", "remaining_time": "0:46:01"}
184
+ {"current_steps": 1540, "total_steps": 4030, "loss": 0.0002, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.854454636332013e-05, "epoch": 3.82, "percentage": 38.21, "elapsed_time": "0:28:17", "remaining_time": "0:45:44"}
185
+ {"current_steps": 1550, "total_steps": 4030, "loss": 0.0059, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.818020004553702e-05, "epoch": 3.85, "percentage": 38.46, "elapsed_time": "0:28:25", "remaining_time": "0:45:28"}
186
+ {"current_steps": 1550, "total_steps": 4030, "loss": null, "eval_loss": 0.0031606056727468967, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 3.85, "percentage": 38.46, "elapsed_time": "0:28:25", "remaining_time": "0:45:28"}
187
+ {"current_steps": 1560, "total_steps": 4030, "loss": 0.0058, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.781473787359488e-05, "epoch": 3.87, "percentage": 38.71, "elapsed_time": "0:28:50", "remaining_time": "0:45:39"}
188
+ {"current_steps": 1570, "total_steps": 4030, "loss": 0.0131, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.744818227862668e-05, "epoch": 3.9, "percentage": 38.96, "elapsed_time": "0:28:57", "remaining_time": "0:45:22"}
189
+ {"current_steps": 1580, "total_steps": 4030, "loss": 0.0042, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.708055575887687e-05, "epoch": 3.92, "percentage": 39.21, "elapsed_time": "0:29:05", "remaining_time": "0:45:06"}
190
+ {"current_steps": 1590, "total_steps": 4030, "loss": 0.0056, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.671188087832057e-05, "epoch": 3.95, "percentage": 39.45, "elapsed_time": "0:29:13", "remaining_time": "0:44:50"}
191
+ {"current_steps": 1600, "total_steps": 4030, "loss": 0.0001, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.634218026527856e-05, "epoch": 3.97, "percentage": 39.7, "elapsed_time": "0:29:20", "remaining_time": "0:44:34"}
192
+ {"current_steps": 1600, "total_steps": 4030, "loss": null, "eval_loss": 0.002365908119827509, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 3.97, "percentage": 39.7, "elapsed_time": "0:29:20", "remaining_time": "0:44:34"}
193
+ {"current_steps": 1610, "total_steps": 4030, "loss": 0.002, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.59714766110285e-05, "epoch": 4.0, "percentage": 39.95, "elapsed_time": "0:29:45", "remaining_time": "0:44:43"}
194
+ {"current_steps": 1620, "total_steps": 4030, "loss": 0.005, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.559979266841209e-05, "epoch": 4.02, "percentage": 40.2, "elapsed_time": "0:29:52", "remaining_time": "0:44:27"}
195
+ {"current_steps": 1630, "total_steps": 4030, "loss": 0.0007, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.522715125043868e-05, "epoch": 4.04, "percentage": 40.45, "elapsed_time": "0:30:00", "remaining_time": "0:44:11"}
196
+ {"current_steps": 1640, "total_steps": 4030, "loss": 0.0028, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.485357522888496e-05, "epoch": 4.07, "percentage": 40.69, "elapsed_time": "0:30:08", "remaining_time": "0:43:55"}
197
+ {"current_steps": 1650, "total_steps": 4030, "loss": 0.0032, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.447908753289127e-05, "epoch": 4.09, "percentage": 40.94, "elapsed_time": "0:30:15", "remaining_time": "0:43:39"}
198
+ {"current_steps": 1650, "total_steps": 4030, "loss": null, "eval_loss": 0.00048677672748453915, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 4.09, "percentage": 40.94, "elapsed_time": "0:30:15", "remaining_time": "0:43:39"}
199
+ {"current_steps": 1660, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.410371114755409e-05, "epoch": 4.12, "percentage": 41.19, "elapsed_time": "0:30:40", "remaining_time": "0:43:47"}
200
+ {"current_steps": 1670, "total_steps": 4030, "loss": 0.0003, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.372746911251548e-05, "epoch": 4.14, "percentage": 41.44, "elapsed_time": "0:30:48", "remaining_time": "0:43:31"}
201
+ {"current_steps": 1680, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.335038452054878e-05, "epoch": 4.17, "percentage": 41.69, "elapsed_time": "0:30:55", "remaining_time": "0:43:15"}
202
+ {"current_steps": 1690, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.297248051614137e-05, "epoch": 4.19, "percentage": 41.94, "elapsed_time": "0:31:03", "remaining_time": "0:43:00"}
203
+ {"current_steps": 1700, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.259378029407407e-05, "epoch": 4.22, "percentage": 42.18, "elapsed_time": "0:31:11", "remaining_time": "0:42:45"}
204
+ {"current_steps": 1700, "total_steps": 4030, "loss": null, "eval_loss": 0.003695528022944927, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 4.22, "percentage": 42.18, "elapsed_time": "0:31:11", "remaining_time": "0:42:45"}
205
+ {"current_steps": 1710, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.221430709799746e-05, "epoch": 4.24, "percentage": 42.43, "elapsed_time": "0:31:36", "remaining_time": "0:42:52"}
206
+ {"current_steps": 1720, "total_steps": 4030, "loss": 0.0064, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.183408421900532e-05, "epoch": 4.27, "percentage": 42.68, "elapsed_time": "0:31:43", "remaining_time": "0:42:36"}
207
+ {"current_steps": 1730, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.145313499420506e-05, "epoch": 4.29, "percentage": 42.93, "elapsed_time": "0:31:51", "remaining_time": "0:42:21"}
208
+ {"current_steps": 1740, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.10714828052853e-05, "epoch": 4.32, "percentage": 43.18, "elapsed_time": "0:31:59", "remaining_time": "0:42:05"}
209
+ {"current_steps": 1750, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.068915107708084e-05, "epoch": 4.34, "percentage": 43.42, "elapsed_time": "0:32:06", "remaining_time": "0:41:50"}
210
+ {"current_steps": 1750, "total_steps": 4030, "loss": null, "eval_loss": 0.000846256734803319, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 4.34, "percentage": 43.42, "elapsed_time": "0:32:06", "remaining_time": "0:41:50"}
211
+ {"current_steps": 1760, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 6.030616327613482e-05, "epoch": 4.37, "percentage": 43.67, "elapsed_time": "0:32:31", "remaining_time": "0:41:57"}
212
+ {"current_steps": 1770, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.992254290925847e-05, "epoch": 4.39, "percentage": 43.92, "elapsed_time": "0:32:39", "remaining_time": "0:41:41"}
213
+ {"current_steps": 1780, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.953831352208831e-05, "epoch": 4.42, "percentage": 44.17, "elapsed_time": "0:32:47", "remaining_time": "0:41:26"}
214
+ {"current_steps": 1790, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.915349869764093e-05, "epoch": 4.44, "percentage": 44.42, "elapsed_time": "0:32:54", "remaining_time": "0:41:11"}
215
+ {"current_steps": 1800, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.87681220548656e-05, "epoch": 4.47, "percentage": 44.67, "elapsed_time": "0:33:02", "remaining_time": "0:40:56"}
216
+ {"current_steps": 1800, "total_steps": 4030, "loss": null, "eval_loss": 0.0008095924276858568, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 4.47, "percentage": 44.67, "elapsed_time": "0:33:02", "remaining_time": "0:40:56"}
217
+ {"current_steps": 1810, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.838220724719454e-05, "epoch": 4.49, "percentage": 44.91, "elapsed_time": "0:33:27", "remaining_time": "0:41:01"}
218
+ {"current_steps": 1820, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.7995777961091167e-05, "epoch": 4.52, "percentage": 45.16, "elapsed_time": "0:33:34", "remaining_time": "0:40:46"}
219
+ {"current_steps": 1830, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.760885791459627e-05, "epoch": 4.54, "percentage": 45.41, "elapsed_time": "0:33:42", "remaining_time": "0:40:31"}
220
+ {"current_steps": 1840, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.722147085587223e-05, "epoch": 4.57, "percentage": 45.66, "elapsed_time": "0:33:50", "remaining_time": "0:40:16"}
221
+ {"current_steps": 1850, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.683364056174545e-05, "epoch": 4.59, "percentage": 45.91, "elapsed_time": "0:33:57", "remaining_time": "0:40:01"}
222
+ {"current_steps": 1850, "total_steps": 4030, "loss": null, "eval_loss": 0.0008772177388891578, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 4.59, "percentage": 45.91, "elapsed_time": "0:33:57", "remaining_time": "0:40:01"}
223
+ {"current_steps": 1860, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.644539083624701e-05, "epoch": 4.62, "percentage": 46.15, "elapsed_time": "0:34:22", "remaining_time": "0:40:06"}
224
+ {"current_steps": 1870, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.605674550915161e-05, "epoch": 4.64, "percentage": 46.4, "elapsed_time": "0:34:29", "remaining_time": "0:39:50"}
225
+ {"current_steps": 1880, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.566772843451495e-05, "epoch": 4.67, "percentage": 46.65, "elapsed_time": "0:34:37", "remaining_time": "0:39:35"}
226
+ {"current_steps": 1890, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.527836348920965e-05, "epoch": 4.69, "percentage": 46.9, "elapsed_time": "0:34:45", "remaining_time": "0:39:20"}
227
+ {"current_steps": 1900, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.488867457145974e-05, "epoch": 4.71, "percentage": 47.15, "elapsed_time": "0:34:52", "remaining_time": "0:39:06"}
228
+ {"current_steps": 1900, "total_steps": 4030, "loss": null, "eval_loss": 0.0006057315040379763, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 4.71, "percentage": 47.15, "elapsed_time": "0:34:52", "remaining_time": "0:39:06"}
229
+ {"current_steps": 1910, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.449868559937384e-05, "epoch": 4.74, "percentage": 47.39, "elapsed_time": "0:35:17", "remaining_time": "0:39:10"}
230
+ {"current_steps": 1920, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.4108420509477145e-05, "epoch": 4.76, "percentage": 47.64, "elapsed_time": "0:35:24", "remaining_time": "0:38:55"}
231
+ {"current_steps": 1930, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.3717903255242284e-05, "epoch": 4.79, "percentage": 47.89, "elapsed_time": "0:35:32", "remaining_time": "0:38:40"}
232
+ {"current_steps": 1940, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.3327157805619035e-05, "epoch": 4.81, "percentage": 48.14, "elapsed_time": "0:35:40", "remaining_time": "0:38:25"}
233
+ {"current_steps": 1950, "total_steps": 4030, "loss": 0.0002, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.293620814356326e-05, "epoch": 4.84, "percentage": 48.39, "elapsed_time": "0:35:48", "remaining_time": "0:38:11"}
234
+ {"current_steps": 1950, "total_steps": 4030, "loss": null, "eval_loss": 0.0001224023144459352, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 4.84, "percentage": 48.39, "elapsed_time": "0:35:48", "remaining_time": "0:38:11"}
235
+ {"current_steps": 1960, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.254507826456483e-05, "epoch": 4.86, "percentage": 48.64, "elapsed_time": "0:36:12", "remaining_time": "0:38:14"}
236
+ {"current_steps": 1970, "total_steps": 4030, "loss": 0.0002, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.215379217517492e-05, "epoch": 4.89, "percentage": 48.88, "elapsed_time": "0:36:21", "remaining_time": "0:38:00"}
237
+ {"current_steps": 1980, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.176237389153246e-05, "epoch": 4.91, "percentage": 49.13, "elapsed_time": "0:36:28", "remaining_time": "0:37:46"}
238
+ {"current_steps": 1990, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.137084743789012e-05, "epoch": 4.94, "percentage": 49.38, "elapsed_time": "0:36:36", "remaining_time": "0:37:31"}
239
+ {"current_steps": 2000, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.097923684513979e-05, "epoch": 4.96, "percentage": 49.63, "elapsed_time": "0:36:43", "remaining_time": "0:37:16"}
240
+ {"current_steps": 2000, "total_steps": 4030, "loss": null, "eval_loss": 0.0007657750975340605, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 4.96, "percentage": 49.63, "elapsed_time": "0:36:43", "remaining_time": "0:37:16"}
241
+ {"current_steps": 2010, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.058756614933757e-05, "epoch": 4.99, "percentage": 49.88, "elapsed_time": "0:37:08", "remaining_time": "0:37:19"}
242
+ {"current_steps": 2020, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 5.019585939022859e-05, "epoch": 5.01, "percentage": 50.12, "elapsed_time": "0:37:16", "remaining_time": "0:37:05"}
243
+ {"current_steps": 2030, "total_steps": 4030, "loss": 0.0089, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.980414060977142e-05, "epoch": 5.04, "percentage": 50.37, "elapsed_time": "0:37:24", "remaining_time": "0:36:51"}
244
+ {"current_steps": 2040, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.9412433850662434e-05, "epoch": 5.06, "percentage": 50.62, "elapsed_time": "0:37:31", "remaining_time": "0:36:36"}
245
+ {"current_steps": 2050, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.902076315486023e-05, "epoch": 5.09, "percentage": 50.87, "elapsed_time": "0:37:39", "remaining_time": "0:36:22"}
246
+ {"current_steps": 2050, "total_steps": 4030, "loss": null, "eval_loss": 0.00010947392001980916, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 5.09, "percentage": 50.87, "elapsed_time": "0:37:39", "remaining_time": "0:36:22"}
247
+ {"current_steps": 2060, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.86291525621099e-05, "epoch": 5.11, "percentage": 51.12, "elapsed_time": "0:38:04", "remaining_time": "0:36:24"}
248
+ {"current_steps": 2070, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.823762610846755e-05, "epoch": 5.14, "percentage": 51.36, "elapsed_time": "0:38:11", "remaining_time": "0:36:10"}
249
+ {"current_steps": 2080, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.784620782482509e-05, "epoch": 5.16, "percentage": 51.61, "elapsed_time": "0:38:19", "remaining_time": "0:35:55"}
250
+ {"current_steps": 2090, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.745492173543517e-05, "epoch": 5.19, "percentage": 51.86, "elapsed_time": "0:38:27", "remaining_time": "0:35:41"}
251
+ {"current_steps": 2100, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.706379185643675e-05, "epoch": 5.21, "percentage": 52.11, "elapsed_time": "0:38:35", "remaining_time": "0:35:27"}
252
+ {"current_steps": 2100, "total_steps": 4030, "loss": null, "eval_loss": 0.00016069799312390387, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 5.21, "percentage": 52.11, "elapsed_time": "0:38:35", "remaining_time": "0:35:27"}
253
+ {"current_steps": 2110, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.667284219438098e-05, "epoch": 5.24, "percentage": 52.36, "elapsed_time": "0:38:59", "remaining_time": "0:35:29"}
254
+ {"current_steps": 2120, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.628209674475773e-05, "epoch": 5.26, "percentage": 52.61, "elapsed_time": "0:39:07", "remaining_time": "0:35:15"}
255
+ {"current_steps": 2130, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.589157949052287e-05, "epoch": 5.29, "percentage": 52.85, "elapsed_time": "0:39:15", "remaining_time": "0:35:01"}
256
+ {"current_steps": 2140, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.550131440062617e-05, "epoch": 5.31, "percentage": 53.1, "elapsed_time": "0:39:23", "remaining_time": "0:34:47"}
257
+ {"current_steps": 2150, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.511132542854027e-05, "epoch": 5.33, "percentage": 53.35, "elapsed_time": "0:39:30", "remaining_time": "0:34:33"}
258
+ {"current_steps": 2150, "total_steps": 4030, "loss": null, "eval_loss": 0.00017579211271367967, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 5.33, "percentage": 53.35, "elapsed_time": "0:39:30", "remaining_time": "0:34:33"}
259
+ {"current_steps": 2160, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.4721636510790365e-05, "epoch": 5.36, "percentage": 53.6, "elapsed_time": "0:39:55", "remaining_time": "0:34:33"}
260
+ {"current_steps": 2170, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.4332271565485053e-05, "epoch": 5.38, "percentage": 53.85, "elapsed_time": "0:40:02", "remaining_time": "0:34:19"}
261
+ {"current_steps": 2180, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.39432544908484e-05, "epoch": 5.41, "percentage": 54.09, "elapsed_time": "0:40:10", "remaining_time": "0:34:05"}
262
+ {"current_steps": 2190, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.355460916375299e-05, "epoch": 5.43, "percentage": 54.34, "elapsed_time": "0:40:18", "remaining_time": "0:33:51"}
263
+ {"current_steps": 2200, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.3166359438254564e-05, "epoch": 5.46, "percentage": 54.59, "elapsed_time": "0:40:26", "remaining_time": "0:33:38"}
264
+ {"current_steps": 2200, "total_steps": 4030, "loss": null, "eval_loss": 0.0004463446675799787, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 5.46, "percentage": 54.59, "elapsed_time": "0:40:26", "remaining_time": "0:33:38"}
265
+ {"current_steps": 2210, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.27785291441278e-05, "epoch": 5.48, "percentage": 54.84, "elapsed_time": "0:40:50", "remaining_time": "0:33:38"}
266
+ {"current_steps": 2220, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.239114208540374e-05, "epoch": 5.51, "percentage": 55.09, "elapsed_time": "0:40:58", "remaining_time": "0:33:24"}
267
+ {"current_steps": 2230, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.2004222038908845e-05, "epoch": 5.53, "percentage": 55.33, "elapsed_time": "0:41:06", "remaining_time": "0:33:10"}
268
+ {"current_steps": 2240, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.161779275280546e-05, "epoch": 5.56, "percentage": 55.58, "elapsed_time": "0:41:14", "remaining_time": "0:32:57"}
269
+ {"current_steps": 2250, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.123187794513441e-05, "epoch": 5.58, "percentage": 55.83, "elapsed_time": "0:41:21", "remaining_time": "0:32:43"}
270
+ {"current_steps": 2250, "total_steps": 4030, "loss": null, "eval_loss": 0.00039042255957610905, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 5.58, "percentage": 55.83, "elapsed_time": "0:41:21", "remaining_time": "0:32:43"}
271
+ {"current_steps": 2260, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.084650130235909e-05, "epoch": 5.61, "percentage": 56.08, "elapsed_time": "0:41:46", "remaining_time": "0:32:43"}
272
+ {"current_steps": 2270, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.046168647791171e-05, "epoch": 5.63, "percentage": 56.33, "elapsed_time": "0:41:54", "remaining_time": "0:32:29"}
273
+ {"current_steps": 2280, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 4.007745709074154e-05, "epoch": 5.66, "percentage": 56.58, "elapsed_time": "0:42:01", "remaining_time": "0:32:15"}
274
+ {"current_steps": 2290, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.969383672386518e-05, "epoch": 5.68, "percentage": 56.82, "elapsed_time": "0:42:09", "remaining_time": "0:32:02"}
275
+ {"current_steps": 2300, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.931084892291917e-05, "epoch": 5.71, "percentage": 57.07, "elapsed_time": "0:42:17", "remaining_time": "0:31:48"}
276
+ {"current_steps": 2300, "total_steps": 4030, "loss": null, "eval_loss": 0.0004175197391305119, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 5.71, "percentage": 57.07, "elapsed_time": "0:42:17", "remaining_time": "0:31:48"}
277
+ {"current_steps": 2310, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.8928517194714706e-05, "epoch": 5.73, "percentage": 57.32, "elapsed_time": "0:42:42", "remaining_time": "0:31:47"}
278
+ {"current_steps": 2320, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.854686500579494e-05, "epoch": 5.76, "percentage": 57.57, "elapsed_time": "0:42:50", "remaining_time": "0:31:34"}
279
+ {"current_steps": 2330, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.816591578099468e-05, "epoch": 5.78, "percentage": 57.82, "elapsed_time": "0:42:57", "remaining_time": "0:31:20"}
280
+ {"current_steps": 2340, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.778569290200254e-05, "epoch": 5.81, "percentage": 58.06, "elapsed_time": "0:43:05", "remaining_time": "0:31:07"}
281
+ {"current_steps": 2350, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.7406219705925945e-05, "epoch": 5.83, "percentage": 58.31, "elapsed_time": "0:43:13", "remaining_time": "0:30:54"}
282
+ {"current_steps": 2350, "total_steps": 4030, "loss": null, "eval_loss": 0.0003901799500454217, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 5.83, "percentage": 58.31, "elapsed_time": "0:43:13", "remaining_time": "0:30:54"}
283
+ {"current_steps": 2360, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.702751948385864e-05, "epoch": 5.86, "percentage": 58.56, "elapsed_time": "0:43:38", "remaining_time": "0:30:52"}
284
+ {"current_steps": 2370, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.6649615479451226e-05, "epoch": 5.88, "percentage": 58.81, "elapsed_time": "0:43:45", "remaining_time": "0:30:39"}
285
+ {"current_steps": 2380, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.6272530887484535e-05, "epoch": 5.91, "percentage": 59.06, "elapsed_time": "0:43:53", "remaining_time": "0:30:25"}
286
+ {"current_steps": 2390, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.5896288852445906e-05, "epoch": 5.93, "percentage": 59.31, "elapsed_time": "0:44:00", "remaining_time": "0:30:12"}
287
+ {"current_steps": 2400, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.552091246710875e-05, "epoch": 5.96, "percentage": 59.55, "elapsed_time": "0:44:08", "remaining_time": "0:29:58"}
288
+ {"current_steps": 2400, "total_steps": 4030, "loss": null, "eval_loss": 0.00044537056237459183, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 5.96, "percentage": 59.55, "elapsed_time": "0:44:08", "remaining_time": "0:29:58"}
289
+ {"current_steps": 2410, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.514642477111505e-05, "epoch": 5.98, "percentage": 59.8, "elapsed_time": "0:44:32", "remaining_time": "0:29:56"}
290
+ {"current_steps": 2420, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.477284874956134e-05, "epoch": 6.0, "percentage": 60.05, "elapsed_time": "0:44:40", "remaining_time": "0:29:43"}
291
+ {"current_steps": 2430, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.440020733158792e-05, "epoch": 6.03, "percentage": 60.3, "elapsed_time": "0:44:48", "remaining_time": "0:29:29"}
292
+ {"current_steps": 2440, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.402852338897151e-05, "epoch": 6.05, "percentage": 60.55, "elapsed_time": "0:44:56", "remaining_time": "0:29:16"}
293
+ {"current_steps": 2450, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.365781973472144e-05, "epoch": 6.08, "percentage": 60.79, "elapsed_time": "0:45:03", "remaining_time": "0:29:03"}
294
+ {"current_steps": 2450, "total_steps": 4030, "loss": null, "eval_loss": 0.00038982907426543534, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 6.08, "percentage": 60.79, "elapsed_time": "0:45:03", "remaining_time": "0:29:03"}
295
+ {"current_steps": 2460, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.328811912167945e-05, "epoch": 6.1, "percentage": 61.04, "elapsed_time": "0:45:28", "remaining_time": "0:29:01"}
296
+ {"current_steps": 2470, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.291944424112314e-05, "epoch": 6.13, "percentage": 61.29, "elapsed_time": "0:45:36", "remaining_time": "0:28:48"}
297
+ {"current_steps": 2480, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.2551817721373335e-05, "epoch": 6.15, "percentage": 61.54, "elapsed_time": "0:45:43", "remaining_time": "0:28:34"}
298
+ {"current_steps": 2490, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.218526212640511e-05, "epoch": 6.18, "percentage": 61.79, "elapsed_time": "0:45:51", "remaining_time": "0:28:21"}
299
+ {"current_steps": 2500, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.1819799954462983e-05, "epoch": 6.2, "percentage": 62.03, "elapsed_time": "0:45:59", "remaining_time": "0:28:08"}
300
+ {"current_steps": 2500, "total_steps": 4030, "loss": null, "eval_loss": 0.00047402523341588676, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 6.2, "percentage": 62.03, "elapsed_time": "0:45:59", "remaining_time": "0:28:08"}
301
+ {"current_steps": 2510, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.1455453636679866e-05, "epoch": 6.23, "percentage": 62.28, "elapsed_time": "0:46:24", "remaining_time": "0:28:05"}
302
+ {"current_steps": 2520, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.109224553570047e-05, "epoch": 6.25, "percentage": 62.53, "elapsed_time": "0:46:31", "remaining_time": "0:27:52"}
303
+ {"current_steps": 2530, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.073019794430866e-05, "epoch": 6.28, "percentage": 62.78, "elapsed_time": "0:46:39", "remaining_time": "0:27:39"}
304
+ {"current_steps": 2540, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.0369333084059148e-05, "epoch": 6.3, "percentage": 63.03, "elapsed_time": "0:46:47", "remaining_time": "0:27:26"}
305
+ {"current_steps": 2550, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 3.0009673103913725e-05, "epoch": 6.33, "percentage": 63.28, "elapsed_time": "0:46:54", "remaining_time": "0:27:13"}
306
+ {"current_steps": 2550, "total_steps": 4030, "loss": null, "eval_loss": 0.00047384999925270677, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 6.33, "percentage": 63.28, "elapsed_time": "0:46:54", "remaining_time": "0:27:13"}
307
+ {"current_steps": 2560, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.9651240078881625e-05, "epoch": 6.35, "percentage": 63.52, "elapsed_time": "0:47:19", "remaining_time": "0:27:10"}
308
+ {"current_steps": 2570, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.9294056008664762e-05, "epoch": 6.38, "percentage": 63.77, "elapsed_time": "0:47:27", "remaining_time": "0:26:57"}
309
+ {"current_steps": 2580, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.893814281630744e-05, "epoch": 6.4, "percentage": 64.02, "elapsed_time": "0:47:34", "remaining_time": "0:26:44"}
310
+ {"current_steps": 2590, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.8583522346850632e-05, "epoch": 6.43, "percentage": 64.27, "elapsed_time": "0:47:42", "remaining_time": "0:26:31"}
311
+ {"current_steps": 2600, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.823021636599137e-05, "epoch": 6.45, "percentage": 64.52, "elapsed_time": "0:47:50", "remaining_time": "0:26:18"}
312
+ {"current_steps": 2600, "total_steps": 4030, "loss": null, "eval_loss": 0.0005336844478733838, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 6.45, "percentage": 64.52, "elapsed_time": "0:47:50", "remaining_time": "0:26:18"}
313
+ {"current_steps": 2610, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.787824655874674e-05, "epoch": 6.48, "percentage": 64.76, "elapsed_time": "0:48:14", "remaining_time": "0:26:14"}
314
+ {"current_steps": 2620, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.7527634528122847e-05, "epoch": 6.5, "percentage": 65.01, "elapsed_time": "0:48:22", "remaining_time": "0:26:02"}
315
+ {"current_steps": 2630, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.7178401793789012e-05, "epoch": 6.53, "percentage": 65.26, "elapsed_time": "0:48:30", "remaining_time": "0:25:49"}
316
+ {"current_steps": 2640, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.6830569790756805e-05, "epoch": 6.55, "percentage": 65.51, "elapsed_time": "0:48:38", "remaining_time": "0:25:36"}
317
+ {"current_steps": 2650, "total_steps": 4030, "loss": 0.0, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": 2.6484159868064583e-05, "epoch": 6.58, "percentage": 65.76, "elapsed_time": "0:48:46", "remaining_time": "0:25:23"}
318
+ {"current_steps": 2650, "total_steps": 4030, "loss": null, "eval_loss": 0.0004447665123734623, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 6.58, "percentage": 65.76, "elapsed_time": "0:48:46", "remaining_time": "0:25:23"}
319
+ {"current_steps": 2650, "total_steps": 4030, "loss": null, "eval_loss": null, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 6.58, "percentage": 65.76, "elapsed_time": "0:48:46", "remaining_time": "0:25:23"}
320
+ {"current_steps": 71, "total_steps": 71, "loss": null, "eval_loss": 0.00047402523341588676, "predict_loss": null, "reward": null, "learning_rate": null, "epoch": 6.58, "percentage": 100.0, "elapsed_time": "0:49:33", "remaining_time": "0:00:00"}
llama2_13b_peft/formal_fallacies_syllogisms_negation/trainer_state.json ADDED
@@ -0,0 +1,2309 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.00047402523341588676,
3
+ "best_model_checkpoint": "ckpt/llama2_13b_fuze15_no_sys/formal_fallacies_syllogisms_negation_no_sys/checkpoint-2500",
4
+ "epoch": 6.575682382133995,
5
+ "eval_steps": 50,
6
+ "global_step": 2650,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.02,
13
+ "grad_norm": 17.027812957763672,
14
+ "learning_rate": 5e-05,
15
+ "loss": 9.3847,
16
+ "step": 10
17
+ },
18
+ {
19
+ "epoch": 0.05,
20
+ "grad_norm": 3.365567684173584,
21
+ "learning_rate": 0.0001,
22
+ "loss": 2.5484,
23
+ "step": 20
24
+ },
25
+ {
26
+ "epoch": 0.07,
27
+ "grad_norm": 2.732487201690674,
28
+ "learning_rate": 9.999846556397038e-05,
29
+ "loss": 0.3827,
30
+ "step": 30
31
+ },
32
+ {
33
+ "epoch": 0.1,
34
+ "grad_norm": 1.364671230316162,
35
+ "learning_rate": 9.999386235006125e-05,
36
+ "loss": 0.3554,
37
+ "step": 40
38
+ },
39
+ {
40
+ "epoch": 0.12,
41
+ "grad_norm": 0.4866187870502472,
42
+ "learning_rate": 9.998619064080611e-05,
43
+ "loss": 0.3929,
44
+ "step": 50
45
+ },
46
+ {
47
+ "epoch": 0.12,
48
+ "eval_loss": 0.3599805533885956,
49
+ "eval_runtime": 16.7693,
50
+ "eval_samples_per_second": 101.615,
51
+ "eval_steps_per_second": 4.234,
52
+ "step": 50
53
+ },
54
+ {
55
+ "epoch": 0.15,
56
+ "grad_norm": 0.651821494102478,
57
+ "learning_rate": 9.997545090707485e-05,
58
+ "loss": 0.3588,
59
+ "step": 60
60
+ },
61
+ {
62
+ "epoch": 0.17,
63
+ "grad_norm": 0.1661490499973297,
64
+ "learning_rate": 9.996164380804483e-05,
65
+ "loss": 0.3594,
66
+ "step": 70
67
+ },
68
+ {
69
+ "epoch": 0.2,
70
+ "grad_norm": 0.2540946304798126,
71
+ "learning_rate": 9.994477019116049e-05,
72
+ "loss": 0.3544,
73
+ "step": 80
74
+ },
75
+ {
76
+ "epoch": 0.22,
77
+ "grad_norm": 0.48444175720214844,
78
+ "learning_rate": 9.992483109208122e-05,
79
+ "loss": 0.3607,
80
+ "step": 90
81
+ },
82
+ {
83
+ "epoch": 0.25,
84
+ "grad_norm": 0.08022764325141907,
85
+ "learning_rate": 9.990182773461791e-05,
86
+ "loss": 0.3534,
87
+ "step": 100
88
+ },
89
+ {
90
+ "epoch": 0.25,
91
+ "eval_loss": 0.35028183460235596,
92
+ "eval_runtime": 16.8797,
93
+ "eval_samples_per_second": 100.95,
94
+ "eval_steps_per_second": 4.206,
95
+ "step": 100
96
+ },
97
+ {
98
+ "epoch": 0.27,
99
+ "grad_norm": 0.22653889656066895,
100
+ "learning_rate": 9.987576153065779e-05,
101
+ "loss": 0.3493,
102
+ "step": 110
103
+ },
104
+ {
105
+ "epoch": 0.3,
106
+ "grad_norm": 0.18461823463439941,
107
+ "learning_rate": 9.984663408007777e-05,
108
+ "loss": 0.3519,
109
+ "step": 120
110
+ },
111
+ {
112
+ "epoch": 0.32,
113
+ "grad_norm": 0.8559591770172119,
114
+ "learning_rate": 9.98144471706462e-05,
115
+ "loss": 0.3532,
116
+ "step": 130
117
+ },
118
+ {
119
+ "epoch": 0.35,
120
+ "grad_norm": 0.2931515574455261,
121
+ "learning_rate": 9.977920277791328e-05,
122
+ "loss": 0.3526,
123
+ "step": 140
124
+ },
125
+ {
126
+ "epoch": 0.37,
127
+ "grad_norm": 0.24398167431354523,
128
+ "learning_rate": 9.974090306508959e-05,
129
+ "loss": 0.3577,
130
+ "step": 150
131
+ },
132
+ {
133
+ "epoch": 0.37,
134
+ "eval_loss": 0.3542585074901581,
135
+ "eval_runtime": 16.8699,
136
+ "eval_samples_per_second": 101.008,
137
+ "eval_steps_per_second": 4.209,
138
+ "step": 150
139
+ },
140
+ {
141
+ "epoch": 0.4,
142
+ "grad_norm": 0.1573762148618698,
143
+ "learning_rate": 9.969955038291353e-05,
144
+ "loss": 0.3525,
145
+ "step": 160
146
+ },
147
+ {
148
+ "epoch": 0.42,
149
+ "grad_norm": 0.1686459630727768,
150
+ "learning_rate": 9.965514726950693e-05,
151
+ "loss": 0.3493,
152
+ "step": 170
153
+ },
154
+ {
155
+ "epoch": 0.45,
156
+ "grad_norm": 0.5627509951591492,
157
+ "learning_rate": 9.960769645021923e-05,
158
+ "loss": 0.3515,
159
+ "step": 180
160
+ },
161
+ {
162
+ "epoch": 0.47,
163
+ "grad_norm": 0.24285666644573212,
164
+ "learning_rate": 9.955720083746035e-05,
165
+ "loss": 0.349,
166
+ "step": 190
167
+ },
168
+ {
169
+ "epoch": 0.5,
170
+ "grad_norm": 0.07683099806308746,
171
+ "learning_rate": 9.950366353052175e-05,
172
+ "loss": 0.351,
173
+ "step": 200
174
+ },
175
+ {
176
+ "epoch": 0.5,
177
+ "eval_loss": 0.3478730320930481,
178
+ "eval_runtime": 16.8735,
179
+ "eval_samples_per_second": 100.987,
180
+ "eval_steps_per_second": 4.208,
181
+ "step": 200
182
+ },
183
+ {
184
+ "epoch": 0.52,
185
+ "grad_norm": 0.36977964639663696,
186
+ "learning_rate": 9.944708781538636e-05,
187
+ "loss": 0.3545,
188
+ "step": 210
189
+ },
190
+ {
191
+ "epoch": 0.55,
192
+ "grad_norm": 0.05500698462128639,
193
+ "learning_rate": 9.93874771645268e-05,
194
+ "loss": 0.3499,
195
+ "step": 220
196
+ },
197
+ {
198
+ "epoch": 0.57,
199
+ "grad_norm": 0.10918760299682617,
200
+ "learning_rate": 9.932483523669231e-05,
201
+ "loss": 0.3504,
202
+ "step": 230
203
+ },
204
+ {
205
+ "epoch": 0.6,
206
+ "grad_norm": 0.08242153376340866,
207
+ "learning_rate": 9.925916587668409e-05,
208
+ "loss": 0.3543,
209
+ "step": 240
210
+ },
211
+ {
212
+ "epoch": 0.62,
213
+ "grad_norm": 0.06469199061393738,
214
+ "learning_rate": 9.919047311511946e-05,
215
+ "loss": 0.3459,
216
+ "step": 250
217
+ },
218
+ {
219
+ "epoch": 0.62,
220
+ "eval_loss": 0.35084518790245056,
221
+ "eval_runtime": 16.8753,
222
+ "eval_samples_per_second": 100.976,
223
+ "eval_steps_per_second": 4.207,
224
+ "step": 250
225
+ },
226
+ {
227
+ "epoch": 0.65,
228
+ "grad_norm": 0.27665790915489197,
229
+ "learning_rate": 9.911876116818433e-05,
230
+ "loss": 0.3483,
231
+ "step": 260
232
+ },
233
+ {
234
+ "epoch": 0.67,
235
+ "grad_norm": 0.1749347299337387,
236
+ "learning_rate": 9.904403443737454e-05,
237
+ "loss": 0.3577,
238
+ "step": 270
239
+ },
240
+ {
241
+ "epoch": 0.69,
242
+ "grad_norm": 0.06966757029294968,
243
+ "learning_rate": 9.896629750922555e-05,
244
+ "loss": 0.3656,
245
+ "step": 280
246
+ },
247
+ {
248
+ "epoch": 0.72,
249
+ "grad_norm": 0.35240188241004944,
250
+ "learning_rate": 9.888555515503116e-05,
251
+ "loss": 0.3595,
252
+ "step": 290
253
+ },
254
+ {
255
+ "epoch": 0.74,
256
+ "grad_norm": 0.1216234564781189,
257
+ "learning_rate": 9.880181233055042e-05,
258
+ "loss": 0.353,
259
+ "step": 300
260
+ },
261
+ {
262
+ "epoch": 0.74,
263
+ "eval_loss": 0.351527601480484,
264
+ "eval_runtime": 16.8754,
265
+ "eval_samples_per_second": 100.975,
266
+ "eval_steps_per_second": 4.207,
267
+ "step": 300
268
+ },
269
+ {
270
+ "epoch": 0.77,
271
+ "grad_norm": 0.32111746072769165,
272
+ "learning_rate": 9.871507417570363e-05,
273
+ "loss": 0.3552,
274
+ "step": 310
275
+ },
276
+ {
277
+ "epoch": 0.79,
278
+ "grad_norm": 0.08302031457424164,
279
+ "learning_rate": 9.862534601425681e-05,
280
+ "loss": 0.3515,
281
+ "step": 320
282
+ },
283
+ {
284
+ "epoch": 0.82,
285
+ "grad_norm": 0.2720988690853119,
286
+ "learning_rate": 9.853263335349488e-05,
287
+ "loss": 0.3445,
288
+ "step": 330
289
+ },
290
+ {
291
+ "epoch": 0.84,
292
+ "grad_norm": 0.2636917531490326,
293
+ "learning_rate": 9.843694188388374e-05,
294
+ "loss": 0.3579,
295
+ "step": 340
296
+ },
297
+ {
298
+ "epoch": 0.87,
299
+ "grad_norm": 0.6955106258392334,
300
+ "learning_rate": 9.833827747872093e-05,
301
+ "loss": 0.3591,
302
+ "step": 350
303
+ },
304
+ {
305
+ "epoch": 0.87,
306
+ "eval_loss": 0.3899991512298584,
307
+ "eval_runtime": 16.8704,
308
+ "eval_samples_per_second": 101.005,
309
+ "eval_steps_per_second": 4.209,
310
+ "step": 350
311
+ },
312
+ {
313
+ "epoch": 0.89,
314
+ "grad_norm": 0.08135569840669632,
315
+ "learning_rate": 9.823664619377518e-05,
316
+ "loss": 0.3472,
317
+ "step": 360
318
+ },
319
+ {
320
+ "epoch": 0.92,
321
+ "grad_norm": 0.30241286754608154,
322
+ "learning_rate": 9.81320542669147e-05,
323
+ "loss": 0.3515,
324
+ "step": 370
325
+ },
326
+ {
327
+ "epoch": 0.94,
328
+ "grad_norm": 0.19453191757202148,
329
+ "learning_rate": 9.802450811772435e-05,
330
+ "loss": 0.3591,
331
+ "step": 380
332
+ },
333
+ {
334
+ "epoch": 0.97,
335
+ "grad_norm": 0.16131022572517395,
336
+ "learning_rate": 9.791401434711154e-05,
337
+ "loss": 0.3511,
338
+ "step": 390
339
+ },
340
+ {
341
+ "epoch": 0.99,
342
+ "grad_norm": 0.0854898989200592,
343
+ "learning_rate": 9.78005797369012e-05,
344
+ "loss": 0.3563,
345
+ "step": 400
346
+ },
347
+ {
348
+ "epoch": 0.99,
349
+ "eval_loss": 0.3495555818080902,
350
+ "eval_runtime": 16.8811,
351
+ "eval_samples_per_second": 100.941,
352
+ "eval_steps_per_second": 4.206,
353
+ "step": 400
354
+ },
355
+ {
356
+ "epoch": 1.02,
357
+ "grad_norm": 0.12075020372867584,
358
+ "learning_rate": 9.768421124941944e-05,
359
+ "loss": 0.3384,
360
+ "step": 410
361
+ },
362
+ {
363
+ "epoch": 1.04,
364
+ "grad_norm": 0.44095876812934875,
365
+ "learning_rate": 9.756491602706626e-05,
366
+ "loss": 0.3545,
367
+ "step": 420
368
+ },
369
+ {
370
+ "epoch": 1.07,
371
+ "grad_norm": 0.16610263288021088,
372
+ "learning_rate": 9.744270139187716e-05,
373
+ "loss": 0.3449,
374
+ "step": 430
375
+ },
376
+ {
377
+ "epoch": 1.09,
378
+ "grad_norm": 0.26342347264289856,
379
+ "learning_rate": 9.731757484507371e-05,
380
+ "loss": 0.346,
381
+ "step": 440
382
+ },
383
+ {
384
+ "epoch": 1.12,
385
+ "grad_norm": 0.35193535685539246,
386
+ "learning_rate": 9.718954406660318e-05,
387
+ "loss": 0.3351,
388
+ "step": 450
389
+ },
390
+ {
391
+ "epoch": 1.12,
392
+ "eval_loss": 0.33888280391693115,
393
+ "eval_runtime": 16.8683,
394
+ "eval_samples_per_second": 101.018,
395
+ "eval_steps_per_second": 4.209,
396
+ "step": 450
397
+ },
398
+ {
399
+ "epoch": 1.14,
400
+ "grad_norm": 0.8358747363090515,
401
+ "learning_rate": 9.705861691466716e-05,
402
+ "loss": 0.3306,
403
+ "step": 460
404
+ },
405
+ {
406
+ "epoch": 1.17,
407
+ "grad_norm": 0.2413673996925354,
408
+ "learning_rate": 9.692480142523918e-05,
409
+ "loss": 0.3514,
410
+ "step": 470
411
+ },
412
+ {
413
+ "epoch": 1.19,
414
+ "grad_norm": 0.7732076048851013,
415
+ "learning_rate": 9.678810581157163e-05,
416
+ "loss": 0.3159,
417
+ "step": 480
418
+ },
419
+ {
420
+ "epoch": 1.22,
421
+ "grad_norm": 0.7405707836151123,
422
+ "learning_rate": 9.664853846369144e-05,
423
+ "loss": 0.309,
424
+ "step": 490
425
+ },
426
+ {
427
+ "epoch": 1.24,
428
+ "grad_norm": 2.760390520095825,
429
+ "learning_rate": 9.650610794788533e-05,
430
+ "loss": 0.26,
431
+ "step": 500
432
+ },
433
+ {
434
+ "epoch": 1.24,
435
+ "eval_loss": 0.3002278506755829,
436
+ "eval_runtime": 16.8732,
437
+ "eval_samples_per_second": 100.988,
438
+ "eval_steps_per_second": 4.208,
439
+ "step": 500
440
+ },
441
+ {
442
+ "epoch": 1.27,
443
+ "grad_norm": 2.7352631092071533,
444
+ "learning_rate": 9.63608230061739e-05,
445
+ "loss": 0.2472,
446
+ "step": 510
447
+ },
448
+ {
449
+ "epoch": 1.29,
450
+ "grad_norm": 2.589384078979492,
451
+ "learning_rate": 9.621269255577512e-05,
452
+ "loss": 0.2212,
453
+ "step": 520
454
+ },
455
+ {
456
+ "epoch": 1.32,
457
+ "grad_norm": 2.2857918739318848,
458
+ "learning_rate": 9.606172568855698e-05,
459
+ "loss": 0.2039,
460
+ "step": 530
461
+ },
462
+ {
463
+ "epoch": 1.34,
464
+ "grad_norm": 0.7525045275688171,
465
+ "learning_rate": 9.590793167047951e-05,
466
+ "loss": 0.2198,
467
+ "step": 540
468
+ },
469
+ {
470
+ "epoch": 1.36,
471
+ "grad_norm": 0.9323859810829163,
472
+ "learning_rate": 9.575131994102602e-05,
473
+ "loss": 0.187,
474
+ "step": 550
475
+ },
476
+ {
477
+ "epoch": 1.36,
478
+ "eval_loss": 0.14382539689540863,
479
+ "eval_runtime": 16.8737,
480
+ "eval_samples_per_second": 100.985,
481
+ "eval_steps_per_second": 4.208,
482
+ "step": 550
483
+ },
484
+ {
485
+ "epoch": 1.39,
486
+ "grad_norm": 1.6801602840423584,
487
+ "learning_rate": 9.559190011262369e-05,
488
+ "loss": 0.1253,
489
+ "step": 560
490
+ },
491
+ {
492
+ "epoch": 1.41,
493
+ "grad_norm": 0.5075914263725281,
494
+ "learning_rate": 9.542968197005367e-05,
495
+ "loss": 0.1246,
496
+ "step": 570
497
+ },
498
+ {
499
+ "epoch": 1.44,
500
+ "grad_norm": 1.5305324792861938,
501
+ "learning_rate": 9.526467546985048e-05,
502
+ "loss": 0.1097,
503
+ "step": 580
504
+ },
505
+ {
506
+ "epoch": 1.46,
507
+ "grad_norm": 0.5779011249542236,
508
+ "learning_rate": 9.509689073969088e-05,
509
+ "loss": 0.0995,
510
+ "step": 590
511
+ },
512
+ {
513
+ "epoch": 1.49,
514
+ "grad_norm": 2.8470897674560547,
515
+ "learning_rate": 9.492633807777227e-05,
516
+ "loss": 0.0715,
517
+ "step": 600
518
+ },
519
+ {
520
+ "epoch": 1.49,
521
+ "eval_loss": 0.0493486262857914,
522
+ "eval_runtime": 16.8798,
523
+ "eval_samples_per_second": 100.949,
524
+ "eval_steps_per_second": 4.206,
525
+ "step": 600
526
+ },
527
+ {
528
+ "epoch": 1.51,
529
+ "grad_norm": 0.08935365825891495,
530
+ "learning_rate": 9.475302795218061e-05,
531
+ "loss": 0.0715,
532
+ "step": 610
533
+ },
534
+ {
535
+ "epoch": 1.54,
536
+ "grad_norm": 0.732679545879364,
537
+ "learning_rate": 9.457697100024798e-05,
538
+ "loss": 0.0797,
539
+ "step": 620
540
+ },
541
+ {
542
+ "epoch": 1.56,
543
+ "grad_norm": 1.6677818298339844,
544
+ "learning_rate": 9.439817802789958e-05,
545
+ "loss": 0.065,
546
+ "step": 630
547
+ },
548
+ {
549
+ "epoch": 1.59,
550
+ "grad_norm": 2.5735604763031006,
551
+ "learning_rate": 9.421666000899052e-05,
552
+ "loss": 0.0534,
553
+ "step": 640
554
+ },
555
+ {
556
+ "epoch": 1.61,
557
+ "grad_norm": 1.0087685585021973,
558
+ "learning_rate": 9.403242808463237e-05,
559
+ "loss": 0.0695,
560
+ "step": 650
561
+ },
562
+ {
563
+ "epoch": 1.61,
564
+ "eval_loss": 0.047055672854185104,
565
+ "eval_runtime": 16.8872,
566
+ "eval_samples_per_second": 100.905,
567
+ "eval_steps_per_second": 4.204,
568
+ "step": 650
569
+ },
570
+ {
571
+ "epoch": 1.64,
572
+ "grad_norm": 1.0158408880233765,
573
+ "learning_rate": 9.38454935625092e-05,
574
+ "loss": 0.0259,
575
+ "step": 660
576
+ },
577
+ {
578
+ "epoch": 1.66,
579
+ "grad_norm": 1.2724292278289795,
580
+ "learning_rate": 9.365586791618368e-05,
581
+ "loss": 0.0388,
582
+ "step": 670
583
+ },
584
+ {
585
+ "epoch": 1.69,
586
+ "grad_norm": 1.6942436695098877,
587
+ "learning_rate": 9.346356278439275e-05,
588
+ "loss": 0.0514,
589
+ "step": 680
590
+ },
591
+ {
592
+ "epoch": 1.71,
593
+ "grad_norm": 0.5555233955383301,
594
+ "learning_rate": 9.32685899703333e-05,
595
+ "loss": 0.0382,
596
+ "step": 690
597
+ },
598
+ {
599
+ "epoch": 1.74,
600
+ "grad_norm": 0.1289750039577484,
601
+ "learning_rate": 9.307096144093781e-05,
602
+ "loss": 0.0338,
603
+ "step": 700
604
+ },
605
+ {
606
+ "epoch": 1.74,
607
+ "eval_loss": 0.030666794627904892,
608
+ "eval_runtime": 16.8789,
609
+ "eval_samples_per_second": 100.955,
610
+ "eval_steps_per_second": 4.206,
611
+ "step": 700
612
+ },
613
+ {
614
+ "epoch": 1.76,
615
+ "grad_norm": 0.195440873503685,
616
+ "learning_rate": 9.287068932613968e-05,
617
+ "loss": 0.0635,
618
+ "step": 710
619
+ },
620
+ {
621
+ "epoch": 1.79,
622
+ "grad_norm": 0.9766190052032471,
623
+ "learning_rate": 9.266778591812886e-05,
624
+ "loss": 0.0303,
625
+ "step": 720
626
+ },
627
+ {
628
+ "epoch": 1.81,
629
+ "grad_norm": 0.9638659358024597,
630
+ "learning_rate": 9.246226367059736e-05,
631
+ "loss": 0.0094,
632
+ "step": 730
633
+ },
634
+ {
635
+ "epoch": 1.84,
636
+ "grad_norm": 1.7231276035308838,
637
+ "learning_rate": 9.225413519797482e-05,
638
+ "loss": 0.0217,
639
+ "step": 740
640
+ },
641
+ {
642
+ "epoch": 1.86,
643
+ "grad_norm": 1.7973217964172363,
644
+ "learning_rate": 9.204341327465434e-05,
645
+ "loss": 0.0422,
646
+ "step": 750
647
+ },
648
+ {
649
+ "epoch": 1.86,
650
+ "eval_loss": 0.02686377428472042,
651
+ "eval_runtime": 16.8806,
652
+ "eval_samples_per_second": 100.945,
653
+ "eval_steps_per_second": 4.206,
654
+ "step": 750
655
+ },
656
+ {
657
+ "epoch": 1.89,
658
+ "grad_norm": 1.0615655183792114,
659
+ "learning_rate": 9.183011083420838e-05,
660
+ "loss": 0.046,
661
+ "step": 760
662
+ },
663
+ {
664
+ "epoch": 1.91,
665
+ "grad_norm": 2.6027116775512695,
666
+ "learning_rate": 9.161424096859493e-05,
667
+ "loss": 0.0192,
668
+ "step": 770
669
+ },
670
+ {
671
+ "epoch": 1.94,
672
+ "grad_norm": 0.3014383316040039,
673
+ "learning_rate": 9.139581692735394e-05,
674
+ "loss": 0.0159,
675
+ "step": 780
676
+ },
677
+ {
678
+ "epoch": 1.96,
679
+ "grad_norm": 0.13248316943645477,
680
+ "learning_rate": 9.11748521167942e-05,
681
+ "loss": 0.0188,
682
+ "step": 790
683
+ },
684
+ {
685
+ "epoch": 1.99,
686
+ "grad_norm": 0.43969258666038513,
687
+ "learning_rate": 9.095136009917036e-05,
688
+ "loss": 0.0183,
689
+ "step": 800
690
+ },
691
+ {
692
+ "epoch": 1.99,
693
+ "eval_loss": 0.009065705351531506,
694
+ "eval_runtime": 16.8746,
695
+ "eval_samples_per_second": 100.98,
696
+ "eval_steps_per_second": 4.208,
697
+ "step": 800
698
+ },
699
+ {
700
+ "epoch": 2.01,
701
+ "grad_norm": 0.16183876991271973,
702
+ "learning_rate": 9.072535459185056e-05,
703
+ "loss": 0.0317,
704
+ "step": 810
705
+ },
706
+ {
707
+ "epoch": 2.03,
708
+ "grad_norm": 0.011793575249612331,
709
+ "learning_rate": 9.049684946647459e-05,
710
+ "loss": 0.0059,
711
+ "step": 820
712
+ },
713
+ {
714
+ "epoch": 2.06,
715
+ "grad_norm": 0.01810554228723049,
716
+ "learning_rate": 9.026585874810227e-05,
717
+ "loss": 0.0009,
718
+ "step": 830
719
+ },
720
+ {
721
+ "epoch": 2.08,
722
+ "grad_norm": 0.17199455201625824,
723
+ "learning_rate": 9.003239661435289e-05,
724
+ "loss": 0.0015,
725
+ "step": 840
726
+ },
727
+ {
728
+ "epoch": 2.11,
729
+ "grad_norm": 5.362873554229736,
730
+ "learning_rate": 8.979647739453481e-05,
731
+ "loss": 0.0055,
732
+ "step": 850
733
+ },
734
+ {
735
+ "epoch": 2.11,
736
+ "eval_loss": 0.007036147639155388,
737
+ "eval_runtime": 16.8747,
738
+ "eval_samples_per_second": 100.98,
739
+ "eval_steps_per_second": 4.207,
740
+ "step": 850
741
+ },
742
+ {
743
+ "epoch": 2.13,
744
+ "grad_norm": 0.0009894347749650478,
745
+ "learning_rate": 8.955811556876605e-05,
746
+ "loss": 0.0002,
747
+ "step": 860
748
+ },
749
+ {
750
+ "epoch": 2.16,
751
+ "grad_norm": 0.0041043683886528015,
752
+ "learning_rate": 8.93173257670856e-05,
753
+ "loss": 0.0172,
754
+ "step": 870
755
+ },
756
+ {
757
+ "epoch": 2.18,
758
+ "grad_norm": 0.01311841607093811,
759
+ "learning_rate": 8.907412276855531e-05,
760
+ "loss": 0.0014,
761
+ "step": 880
762
+ },
763
+ {
764
+ "epoch": 2.21,
765
+ "grad_norm": 0.007052562665194273,
766
+ "learning_rate": 8.882852150035295e-05,
767
+ "loss": 0.0025,
768
+ "step": 890
769
+ },
770
+ {
771
+ "epoch": 2.23,
772
+ "grad_norm": 0.0044655706733465195,
773
+ "learning_rate": 8.858053703685587e-05,
774
+ "loss": 0.0208,
775
+ "step": 900
776
+ },
777
+ {
778
+ "epoch": 2.23,
779
+ "eval_loss": 0.003154078498482704,
780
+ "eval_runtime": 16.8784,
781
+ "eval_samples_per_second": 100.958,
782
+ "eval_steps_per_second": 4.207,
783
+ "step": 900
784
+ },
785
+ {
786
+ "epoch": 2.26,
787
+ "grad_norm": 0.0044303881004452705,
788
+ "learning_rate": 8.833018459871594e-05,
789
+ "loss": 0.0016,
790
+ "step": 910
791
+ },
792
+ {
793
+ "epoch": 2.28,
794
+ "grad_norm": 0.15879973769187927,
795
+ "learning_rate": 8.807747955192516e-05,
796
+ "loss": 0.0139,
797
+ "step": 920
798
+ },
799
+ {
800
+ "epoch": 2.31,
801
+ "grad_norm": 0.006470128428190947,
802
+ "learning_rate": 8.782243740687273e-05,
803
+ "loss": 0.0007,
804
+ "step": 930
805
+ },
806
+ {
807
+ "epoch": 2.33,
808
+ "grad_norm": 0.0031656241044402122,
809
+ "learning_rate": 8.756507381739288e-05,
810
+ "loss": 0.0149,
811
+ "step": 940
812
+ },
813
+ {
814
+ "epoch": 2.36,
815
+ "grad_norm": 0.7405675649642944,
816
+ "learning_rate": 8.730540457980418e-05,
817
+ "loss": 0.0014,
818
+ "step": 950
819
+ },
820
+ {
821
+ "epoch": 2.36,
822
+ "eval_loss": 0.002297737170010805,
823
+ "eval_runtime": 16.8743,
824
+ "eval_samples_per_second": 100.982,
825
+ "eval_steps_per_second": 4.208,
826
+ "step": 950
827
+ },
828
+ {
829
+ "epoch": 2.38,
830
+ "grad_norm": 0.514784574508667,
831
+ "learning_rate": 8.704344563193997e-05,
832
+ "loss": 0.0014,
833
+ "step": 960
834
+ },
835
+ {
836
+ "epoch": 2.41,
837
+ "grad_norm": 0.01349685899913311,
838
+ "learning_rate": 8.677921305217023e-05,
839
+ "loss": 0.0002,
840
+ "step": 970
841
+ },
842
+ {
843
+ "epoch": 2.43,
844
+ "grad_norm": 0.0020194160751998425,
845
+ "learning_rate": 8.651272305841454e-05,
846
+ "loss": 0.0207,
847
+ "step": 980
848
+ },
849
+ {
850
+ "epoch": 2.46,
851
+ "grad_norm": 1.934012532234192,
852
+ "learning_rate": 8.62439920071468e-05,
853
+ "loss": 0.0011,
854
+ "step": 990
855
+ },
856
+ {
857
+ "epoch": 2.48,
858
+ "grad_norm": 0.04069426655769348,
859
+ "learning_rate": 8.597303639239134e-05,
860
+ "loss": 0.0117,
861
+ "step": 1000
862
+ },
863
+ {
864
+ "epoch": 2.48,
865
+ "eval_loss": 0.0033970780204981565,
866
+ "eval_runtime": 16.8634,
867
+ "eval_samples_per_second": 101.047,
868
+ "eval_steps_per_second": 4.21,
869
+ "step": 1000
870
+ },
871
+ {
872
+ "epoch": 2.51,
873
+ "grad_norm": 0.03683527931571007,
874
+ "learning_rate": 8.569987284471046e-05,
875
+ "loss": 0.0026,
876
+ "step": 1010
877
+ },
878
+ {
879
+ "epoch": 2.53,
880
+ "grad_norm": 0.050554804503917694,
881
+ "learning_rate": 8.542451813018372e-05,
882
+ "loss": 0.0022,
883
+ "step": 1020
884
+ },
885
+ {
886
+ "epoch": 2.56,
887
+ "grad_norm": 0.0019483251962810755,
888
+ "learning_rate": 8.514698914937894e-05,
889
+ "loss": 0.0002,
890
+ "step": 1030
891
+ },
892
+ {
893
+ "epoch": 2.58,
894
+ "grad_norm": 1.8008921146392822,
895
+ "learning_rate": 8.486730293631481e-05,
896
+ "loss": 0.0019,
897
+ "step": 1040
898
+ },
899
+ {
900
+ "epoch": 2.61,
901
+ "grad_norm": 0.001593997934833169,
902
+ "learning_rate": 8.458547665741543e-05,
903
+ "loss": 0.0001,
904
+ "step": 1050
905
+ },
906
+ {
907
+ "epoch": 2.61,
908
+ "eval_loss": 0.001217635115608573,
909
+ "eval_runtime": 16.8779,
910
+ "eval_samples_per_second": 100.96,
911
+ "eval_steps_per_second": 4.207,
912
+ "step": 1050
913
+ },
914
+ {
915
+ "epoch": 2.63,
916
+ "grad_norm": 0.003130235243588686,
917
+ "learning_rate": 8.430152761045664e-05,
918
+ "loss": 0.0002,
919
+ "step": 1060
920
+ },
921
+ {
922
+ "epoch": 2.66,
923
+ "grad_norm": 0.001099205925129354,
924
+ "learning_rate": 8.401547322350438e-05,
925
+ "loss": 0.0004,
926
+ "step": 1070
927
+ },
928
+ {
929
+ "epoch": 2.68,
930
+ "grad_norm": 0.18462316691875458,
931
+ "learning_rate": 8.372733105384496e-05,
932
+ "loss": 0.0002,
933
+ "step": 1080
934
+ },
935
+ {
936
+ "epoch": 2.7,
937
+ "grad_norm": 0.001120097003877163,
938
+ "learning_rate": 8.343711878690746e-05,
939
+ "loss": 0.0171,
940
+ "step": 1090
941
+ },
942
+ {
943
+ "epoch": 2.73,
944
+ "grad_norm": 0.9607303738594055,
945
+ "learning_rate": 8.314485423517819e-05,
946
+ "loss": 0.0058,
947
+ "step": 1100
948
+ },
949
+ {
950
+ "epoch": 2.73,
951
+ "eval_loss": 0.0012037238338962197,
952
+ "eval_runtime": 16.8925,
953
+ "eval_samples_per_second": 100.873,
954
+ "eval_steps_per_second": 4.203,
955
+ "step": 1100
956
+ },
957
+ {
958
+ "epoch": 2.75,
959
+ "grad_norm": 0.0029547882732003927,
960
+ "learning_rate": 8.285055533710753e-05,
961
+ "loss": 0.0001,
962
+ "step": 1110
963
+ },
964
+ {
965
+ "epoch": 2.78,
966
+ "grad_norm": 0.0008037490188144147,
967
+ "learning_rate": 8.255424015600876e-05,
968
+ "loss": 0.0059,
969
+ "step": 1120
970
+ },
971
+ {
972
+ "epoch": 2.8,
973
+ "grad_norm": 0.0015891498187556863,
974
+ "learning_rate": 8.225592687894949e-05,
975
+ "loss": 0.0028,
976
+ "step": 1130
977
+ },
978
+ {
979
+ "epoch": 2.83,
980
+ "grad_norm": 0.0007602926343679428,
981
+ "learning_rate": 8.195563381563536e-05,
982
+ "loss": 0.0001,
983
+ "step": 1140
984
+ },
985
+ {
986
+ "epoch": 2.85,
987
+ "grad_norm": 0.0008714946452528238,
988
+ "learning_rate": 8.165337939728614e-05,
989
+ "loss": 0.0001,
990
+ "step": 1150
991
+ },
992
+ {
993
+ "epoch": 2.85,
994
+ "eval_loss": 0.0012541119940578938,
995
+ "eval_runtime": 16.8751,
996
+ "eval_samples_per_second": 100.977,
997
+ "eval_steps_per_second": 4.207,
998
+ "step": 1150
999
+ },
1000
+ {
1001
+ "epoch": 2.88,
1002
+ "grad_norm": 0.0005839269142597914,
1003
+ "learning_rate": 8.134918217550467e-05,
1004
+ "loss": 0.0001,
1005
+ "step": 1160
1006
+ },
1007
+ {
1008
+ "epoch": 2.9,
1009
+ "grad_norm": 0.09183517843484879,
1010
+ "learning_rate": 8.104306082113802e-05,
1011
+ "loss": 0.0017,
1012
+ "step": 1170
1013
+ },
1014
+ {
1015
+ "epoch": 2.93,
1016
+ "grad_norm": 0.0006280924426391721,
1017
+ "learning_rate": 8.073503412313159e-05,
1018
+ "loss": 0.0,
1019
+ "step": 1180
1020
+ },
1021
+ {
1022
+ "epoch": 2.95,
1023
+ "grad_norm": 0.0005470833857543766,
1024
+ "learning_rate": 8.042512098737598e-05,
1025
+ "loss": 0.001,
1026
+ "step": 1190
1027
+ },
1028
+ {
1029
+ "epoch": 2.98,
1030
+ "grad_norm": 0.001008176477625966,
1031
+ "learning_rate": 8.01133404355464e-05,
1032
+ "loss": 0.0104,
1033
+ "step": 1200
1034
+ },
1035
+ {
1036
+ "epoch": 2.98,
1037
+ "eval_loss": 0.0007884196820668876,
1038
+ "eval_runtime": 16.8766,
1039
+ "eval_samples_per_second": 100.968,
1040
+ "eval_steps_per_second": 4.207,
1041
+ "step": 1200
1042
+ },
1043
+ {
1044
+ "epoch": 3.0,
1045
+ "grad_norm": 0.000660700083244592,
1046
+ "learning_rate": 7.979971160393535e-05,
1047
+ "loss": 0.0,
1048
+ "step": 1210
1049
+ },
1050
+ {
1051
+ "epoch": 3.03,
1052
+ "grad_norm": 0.004391908645629883,
1053
+ "learning_rate": 7.948425374227799e-05,
1054
+ "loss": 0.0131,
1055
+ "step": 1220
1056
+ },
1057
+ {
1058
+ "epoch": 3.05,
1059
+ "grad_norm": 0.0010901394998654723,
1060
+ "learning_rate": 7.916698621257068e-05,
1061
+ "loss": 0.0006,
1062
+ "step": 1230
1063
+ },
1064
+ {
1065
+ "epoch": 3.08,
1066
+ "grad_norm": 0.0011236576829105616,
1067
+ "learning_rate": 7.884792848788257e-05,
1068
+ "loss": 0.0028,
1069
+ "step": 1240
1070
+ },
1071
+ {
1072
+ "epoch": 3.1,
1073
+ "grad_norm": 0.0004985715495422482,
1074
+ "learning_rate": 7.852710015116036e-05,
1075
+ "loss": 0.0007,
1076
+ "step": 1250
1077
+ },
1078
+ {
1079
+ "epoch": 3.1,
1080
+ "eval_loss": 0.0018942987080663443,
1081
+ "eval_runtime": 16.8648,
1082
+ "eval_samples_per_second": 101.039,
1083
+ "eval_steps_per_second": 4.21,
1084
+ "step": 1250
1085
+ },
1086
+ {
1087
+ "epoch": 3.13,
1088
+ "grad_norm": 0.0006551850237883627,
1089
+ "learning_rate": 7.820452089402645e-05,
1090
+ "loss": 0.0001,
1091
+ "step": 1260
1092
+ },
1093
+ {
1094
+ "epoch": 3.15,
1095
+ "grad_norm": 0.0009159449255093932,
1096
+ "learning_rate": 7.788021051557022e-05,
1097
+ "loss": 0.0015,
1098
+ "step": 1270
1099
+ },
1100
+ {
1101
+ "epoch": 3.18,
1102
+ "grad_norm": 0.3894294500350952,
1103
+ "learning_rate": 7.755418892113282e-05,
1104
+ "loss": 0.0006,
1105
+ "step": 1280
1106
+ },
1107
+ {
1108
+ "epoch": 3.2,
1109
+ "grad_norm": 0.0011728698154911399,
1110
+ "learning_rate": 7.722647612108553e-05,
1111
+ "loss": 0.0011,
1112
+ "step": 1290
1113
+ },
1114
+ {
1115
+ "epoch": 3.23,
1116
+ "grad_norm": 0.0003914556000381708,
1117
+ "learning_rate": 7.689709222960144e-05,
1118
+ "loss": 0.0015,
1119
+ "step": 1300
1120
+ },
1121
+ {
1122
+ "epoch": 3.23,
1123
+ "eval_loss": 0.00915015209466219,
1124
+ "eval_runtime": 16.8561,
1125
+ "eval_samples_per_second": 101.091,
1126
+ "eval_steps_per_second": 4.212,
1127
+ "step": 1300
1128
+ },
1129
+ {
1130
+ "epoch": 3.25,
1131
+ "grad_norm": 0.0004285429895389825,
1132
+ "learning_rate": 7.656605746342096e-05,
1133
+ "loss": 0.0118,
1134
+ "step": 1310
1135
+ },
1136
+ {
1137
+ "epoch": 3.28,
1138
+ "grad_norm": 0.001354528358206153,
1139
+ "learning_rate": 7.6233392140611e-05,
1140
+ "loss": 0.0,
1141
+ "step": 1320
1142
+ },
1143
+ {
1144
+ "epoch": 3.3,
1145
+ "grad_norm": 0.0010014220606535673,
1146
+ "learning_rate": 7.589911667931786e-05,
1147
+ "loss": 0.0001,
1148
+ "step": 1330
1149
+ },
1150
+ {
1151
+ "epoch": 3.33,
1152
+ "grad_norm": 0.017791859805583954,
1153
+ "learning_rate": 7.556325159651399e-05,
1154
+ "loss": 0.0001,
1155
+ "step": 1340
1156
+ },
1157
+ {
1158
+ "epoch": 3.35,
1159
+ "grad_norm": 0.003636227687820792,
1160
+ "learning_rate": 7.522581750673872e-05,
1161
+ "loss": 0.0,
1162
+ "step": 1350
1163
+ },
1164
+ {
1165
+ "epoch": 3.35,
1166
+ "eval_loss": 0.0016158341895788908,
1167
+ "eval_runtime": 16.872,
1168
+ "eval_samples_per_second": 100.996,
1169
+ "eval_steps_per_second": 4.208,
1170
+ "step": 1350
1171
+ },
1172
+ {
1173
+ "epoch": 3.37,
1174
+ "grad_norm": 0.0002491871709935367,
1175
+ "learning_rate": 7.48868351208331e-05,
1176
+ "loss": 0.0005,
1177
+ "step": 1360
1178
+ },
1179
+ {
1180
+ "epoch": 3.4,
1181
+ "grad_norm": 0.0009618819458410144,
1182
+ "learning_rate": 7.454632524466857e-05,
1183
+ "loss": 0.0,
1184
+ "step": 1370
1185
+ },
1186
+ {
1187
+ "epoch": 3.42,
1188
+ "grad_norm": 0.0006681958911940455,
1189
+ "learning_rate": 7.420430877787004e-05,
1190
+ "loss": 0.0,
1191
+ "step": 1380
1192
+ },
1193
+ {
1194
+ "epoch": 3.45,
1195
+ "grad_norm": 0.0005440477398224175,
1196
+ "learning_rate": 7.386080671253307e-05,
1197
+ "loss": 0.004,
1198
+ "step": 1390
1199
+ },
1200
+ {
1201
+ "epoch": 3.47,
1202
+ "grad_norm": 0.0004287729097995907,
1203
+ "learning_rate": 7.351584013193543e-05,
1204
+ "loss": 0.0001,
1205
+ "step": 1400
1206
+ },
1207
+ {
1208
+ "epoch": 3.47,
1209
+ "eval_loss": 0.000445638841483742,
1210
+ "eval_runtime": 16.8657,
1211
+ "eval_samples_per_second": 101.033,
1212
+ "eval_steps_per_second": 4.21,
1213
+ "step": 1400
1214
+ },
1215
+ {
1216
+ "epoch": 3.5,
1217
+ "grad_norm": 0.0007876464514993131,
1218
+ "learning_rate": 7.316943020924319e-05,
1219
+ "loss": 0.0003,
1220
+ "step": 1410
1221
+ },
1222
+ {
1223
+ "epoch": 3.52,
1224
+ "grad_norm": 0.0006684191175736487,
1225
+ "learning_rate": 7.2821598206211e-05,
1226
+ "loss": 0.0,
1227
+ "step": 1420
1228
+ },
1229
+ {
1230
+ "epoch": 3.55,
1231
+ "grad_norm": 0.000733783992473036,
1232
+ "learning_rate": 7.247236547187716e-05,
1233
+ "loss": 0.0002,
1234
+ "step": 1430
1235
+ },
1236
+ {
1237
+ "epoch": 3.57,
1238
+ "grad_norm": 0.00019874260760843754,
1239
+ "learning_rate": 7.212175344125327e-05,
1240
+ "loss": 0.0,
1241
+ "step": 1440
1242
+ },
1243
+ {
1244
+ "epoch": 3.6,
1245
+ "grad_norm": 0.00011918343079742044,
1246
+ "learning_rate": 7.176978363400863e-05,
1247
+ "loss": 0.0059,
1248
+ "step": 1450
1249
+ },
1250
+ {
1251
+ "epoch": 3.6,
1252
+ "eval_loss": 0.0003264049591962248,
1253
+ "eval_runtime": 16.8679,
1254
+ "eval_samples_per_second": 101.02,
1255
+ "eval_steps_per_second": 4.209,
1256
+ "step": 1450
1257
+ },
1258
+ {
1259
+ "epoch": 3.62,
1260
+ "grad_norm": 0.0008113720687106252,
1261
+ "learning_rate": 7.141647765314938e-05,
1262
+ "loss": 0.0,
1263
+ "step": 1460
1264
+ },
1265
+ {
1266
+ "epoch": 3.65,
1267
+ "grad_norm": 0.00023584155132994056,
1268
+ "learning_rate": 7.106185718369259e-05,
1269
+ "loss": 0.0,
1270
+ "step": 1470
1271
+ },
1272
+ {
1273
+ "epoch": 3.67,
1274
+ "grad_norm": 0.000526837189681828,
1275
+ "learning_rate": 7.070594399133525e-05,
1276
+ "loss": 0.0,
1277
+ "step": 1480
1278
+ },
1279
+ {
1280
+ "epoch": 3.7,
1281
+ "grad_norm": 0.00017550899065099657,
1282
+ "learning_rate": 7.03487599211184e-05,
1283
+ "loss": 0.0038,
1284
+ "step": 1490
1285
+ },
1286
+ {
1287
+ "epoch": 3.72,
1288
+ "grad_norm": 0.06627130508422852,
1289
+ "learning_rate": 6.999032689608629e-05,
1290
+ "loss": 0.0,
1291
+ "step": 1500
1292
+ },
1293
+ {
1294
+ "epoch": 3.72,
1295
+ "eval_loss": 0.009298658929765224,
1296
+ "eval_runtime": 16.8719,
1297
+ "eval_samples_per_second": 100.996,
1298
+ "eval_steps_per_second": 4.208,
1299
+ "step": 1500
1300
+ },
1301
+ {
1302
+ "epoch": 3.75,
1303
+ "grad_norm": 0.00021741975797340274,
1304
+ "learning_rate": 6.963066691594084e-05,
1305
+ "loss": 0.0037,
1306
+ "step": 1510
1307
+ },
1308
+ {
1309
+ "epoch": 3.77,
1310
+ "grad_norm": 0.005204753950238228,
1311
+ "learning_rate": 6.926980205569136e-05,
1312
+ "loss": 0.0113,
1313
+ "step": 1520
1314
+ },
1315
+ {
1316
+ "epoch": 3.8,
1317
+ "grad_norm": 0.0006932571996003389,
1318
+ "learning_rate": 6.890775446429954e-05,
1319
+ "loss": 0.0,
1320
+ "step": 1530
1321
+ },
1322
+ {
1323
+ "epoch": 3.82,
1324
+ "grad_norm": 0.006194334477186203,
1325
+ "learning_rate": 6.854454636332013e-05,
1326
+ "loss": 0.0002,
1327
+ "step": 1540
1328
+ },
1329
+ {
1330
+ "epoch": 3.85,
1331
+ "grad_norm": 0.0016646990552544594,
1332
+ "learning_rate": 6.818020004553702e-05,
1333
+ "loss": 0.0059,
1334
+ "step": 1550
1335
+ },
1336
+ {
1337
+ "epoch": 3.85,
1338
+ "eval_loss": 0.0031606056727468967,
1339
+ "eval_runtime": 16.8924,
1340
+ "eval_samples_per_second": 100.874,
1341
+ "eval_steps_per_second": 4.203,
1342
+ "step": 1550
1343
+ },
1344
+ {
1345
+ "epoch": 3.87,
1346
+ "grad_norm": 8.5881986618042,
1347
+ "learning_rate": 6.781473787359488e-05,
1348
+ "loss": 0.0058,
1349
+ "step": 1560
1350
+ },
1351
+ {
1352
+ "epoch": 3.9,
1353
+ "grad_norm": 0.0005570609937421978,
1354
+ "learning_rate": 6.744818227862668e-05,
1355
+ "loss": 0.0131,
1356
+ "step": 1570
1357
+ },
1358
+ {
1359
+ "epoch": 3.92,
1360
+ "grad_norm": 0.4314500093460083,
1361
+ "learning_rate": 6.708055575887687e-05,
1362
+ "loss": 0.0042,
1363
+ "step": 1580
1364
+ },
1365
+ {
1366
+ "epoch": 3.95,
1367
+ "grad_norm": 0.01640118658542633,
1368
+ "learning_rate": 6.671188087832057e-05,
1369
+ "loss": 0.0056,
1370
+ "step": 1590
1371
+ },
1372
+ {
1373
+ "epoch": 3.97,
1374
+ "grad_norm": 0.0034761123824864626,
1375
+ "learning_rate": 6.634218026527856e-05,
1376
+ "loss": 0.0001,
1377
+ "step": 1600
1378
+ },
1379
+ {
1380
+ "epoch": 3.97,
1381
+ "eval_loss": 0.002365908119827509,
1382
+ "eval_runtime": 16.8762,
1383
+ "eval_samples_per_second": 100.971,
1384
+ "eval_steps_per_second": 4.207,
1385
+ "step": 1600
1386
+ },
1387
+ {
1388
+ "epoch": 4.0,
1389
+ "grad_norm": 2.214448928833008,
1390
+ "learning_rate": 6.59714766110285e-05,
1391
+ "loss": 0.002,
1392
+ "step": 1610
1393
+ },
1394
+ {
1395
+ "epoch": 4.02,
1396
+ "grad_norm": 0.002197085879743099,
1397
+ "learning_rate": 6.559979266841209e-05,
1398
+ "loss": 0.005,
1399
+ "step": 1620
1400
+ },
1401
+ {
1402
+ "epoch": 4.04,
1403
+ "grad_norm": 0.10025832056999207,
1404
+ "learning_rate": 6.522715125043868e-05,
1405
+ "loss": 0.0007,
1406
+ "step": 1630
1407
+ },
1408
+ {
1409
+ "epoch": 4.07,
1410
+ "grad_norm": 0.002808921504765749,
1411
+ "learning_rate": 6.485357522888496e-05,
1412
+ "loss": 0.0028,
1413
+ "step": 1640
1414
+ },
1415
+ {
1416
+ "epoch": 4.09,
1417
+ "grad_norm": 0.0002055912627838552,
1418
+ "learning_rate": 6.447908753289127e-05,
1419
+ "loss": 0.0032,
1420
+ "step": 1650
1421
+ },
1422
+ {
1423
+ "epoch": 4.09,
1424
+ "eval_loss": 0.00048677672748453915,
1425
+ "eval_runtime": 16.8716,
1426
+ "eval_samples_per_second": 100.998,
1427
+ "eval_steps_per_second": 4.208,
1428
+ "step": 1650
1429
+ },
1430
+ {
1431
+ "epoch": 4.12,
1432
+ "grad_norm": 0.0012891364749521017,
1433
+ "learning_rate": 6.410371114755409e-05,
1434
+ "loss": 0.0,
1435
+ "step": 1660
1436
+ },
1437
+ {
1438
+ "epoch": 4.14,
1439
+ "grad_norm": 0.14440464973449707,
1440
+ "learning_rate": 6.372746911251548e-05,
1441
+ "loss": 0.0003,
1442
+ "step": 1670
1443
+ },
1444
+ {
1445
+ "epoch": 4.17,
1446
+ "grad_norm": 0.00040509516838938,
1447
+ "learning_rate": 6.335038452054878e-05,
1448
+ "loss": 0.0,
1449
+ "step": 1680
1450
+ },
1451
+ {
1452
+ "epoch": 4.19,
1453
+ "grad_norm": 0.00013587850844487548,
1454
+ "learning_rate": 6.297248051614137e-05,
1455
+ "loss": 0.0,
1456
+ "step": 1690
1457
+ },
1458
+ {
1459
+ "epoch": 4.22,
1460
+ "grad_norm": 0.000756852503400296,
1461
+ "learning_rate": 6.259378029407407e-05,
1462
+ "loss": 0.0,
1463
+ "step": 1700
1464
+ },
1465
+ {
1466
+ "epoch": 4.22,
1467
+ "eval_loss": 0.003695528022944927,
1468
+ "eval_runtime": 16.873,
1469
+ "eval_samples_per_second": 100.99,
1470
+ "eval_steps_per_second": 4.208,
1471
+ "step": 1700
1472
+ },
1473
+ {
1474
+ "epoch": 4.24,
1475
+ "grad_norm": 0.00020980635599698871,
1476
+ "learning_rate": 6.221430709799746e-05,
1477
+ "loss": 0.0,
1478
+ "step": 1710
1479
+ },
1480
+ {
1481
+ "epoch": 4.27,
1482
+ "grad_norm": 0.0003155861923005432,
1483
+ "learning_rate": 6.183408421900532e-05,
1484
+ "loss": 0.0064,
1485
+ "step": 1720
1486
+ },
1487
+ {
1488
+ "epoch": 4.29,
1489
+ "grad_norm": 0.00015093179536052048,
1490
+ "learning_rate": 6.145313499420506e-05,
1491
+ "loss": 0.0,
1492
+ "step": 1730
1493
+ },
1494
+ {
1495
+ "epoch": 4.32,
1496
+ "grad_norm": 0.00022122553491499275,
1497
+ "learning_rate": 6.10714828052853e-05,
1498
+ "loss": 0.0,
1499
+ "step": 1740
1500
+ },
1501
+ {
1502
+ "epoch": 4.34,
1503
+ "grad_norm": 0.00024218260659836233,
1504
+ "learning_rate": 6.068915107708084e-05,
1505
+ "loss": 0.0,
1506
+ "step": 1750
1507
+ },
1508
+ {
1509
+ "epoch": 4.34,
1510
+ "eval_loss": 0.000846256734803319,
1511
+ "eval_runtime": 16.8681,
1512
+ "eval_samples_per_second": 101.019,
1513
+ "eval_steps_per_second": 4.209,
1514
+ "step": 1750
1515
+ },
1516
+ {
1517
+ "epoch": 4.37,
1518
+ "grad_norm": 0.0001305171608692035,
1519
+ "learning_rate": 6.030616327613482e-05,
1520
+ "loss": 0.0,
1521
+ "step": 1760
1522
+ },
1523
+ {
1524
+ "epoch": 4.39,
1525
+ "grad_norm": 0.00025666801957413554,
1526
+ "learning_rate": 5.992254290925847e-05,
1527
+ "loss": 0.0,
1528
+ "step": 1770
1529
+ },
1530
+ {
1531
+ "epoch": 4.42,
1532
+ "grad_norm": 0.0001890121930046007,
1533
+ "learning_rate": 5.953831352208831e-05,
1534
+ "loss": 0.0,
1535
+ "step": 1780
1536
+ },
1537
+ {
1538
+ "epoch": 4.44,
1539
+ "grad_norm": 0.00020163397130090743,
1540
+ "learning_rate": 5.915349869764093e-05,
1541
+ "loss": 0.0,
1542
+ "step": 1790
1543
+ },
1544
+ {
1545
+ "epoch": 4.47,
1546
+ "grad_norm": 0.00012597240856848657,
1547
+ "learning_rate": 5.87681220548656e-05,
1548
+ "loss": 0.0,
1549
+ "step": 1800
1550
+ },
1551
+ {
1552
+ "epoch": 4.47,
1553
+ "eval_loss": 0.0008095924276858568,
1554
+ "eval_runtime": 16.8799,
1555
+ "eval_samples_per_second": 100.948,
1556
+ "eval_steps_per_second": 4.206,
1557
+ "step": 1800
1558
+ },
1559
+ {
1560
+ "epoch": 4.49,
1561
+ "grad_norm": 0.00033845013240352273,
1562
+ "learning_rate": 5.838220724719454e-05,
1563
+ "loss": 0.0,
1564
+ "step": 1810
1565
+ },
1566
+ {
1567
+ "epoch": 4.52,
1568
+ "grad_norm": 0.00031902611954137683,
1569
+ "learning_rate": 5.7995777961091167e-05,
1570
+ "loss": 0.0,
1571
+ "step": 1820
1572
+ },
1573
+ {
1574
+ "epoch": 4.54,
1575
+ "grad_norm": 0.0010548951104283333,
1576
+ "learning_rate": 5.760885791459627e-05,
1577
+ "loss": 0.0,
1578
+ "step": 1830
1579
+ },
1580
+ {
1581
+ "epoch": 4.57,
1582
+ "grad_norm": 8.980428538052365e-05,
1583
+ "learning_rate": 5.722147085587223e-05,
1584
+ "loss": 0.0,
1585
+ "step": 1840
1586
+ },
1587
+ {
1588
+ "epoch": 4.59,
1589
+ "grad_norm": 0.00012754097406286746,
1590
+ "learning_rate": 5.683364056174545e-05,
1591
+ "loss": 0.0,
1592
+ "step": 1850
1593
+ },
1594
+ {
1595
+ "epoch": 4.59,
1596
+ "eval_loss": 0.0008772177388891578,
1597
+ "eval_runtime": 16.8757,
1598
+ "eval_samples_per_second": 100.974,
1599
+ "eval_steps_per_second": 4.207,
1600
+ "step": 1850
1601
+ },
1602
+ {
1603
+ "epoch": 4.62,
1604
+ "grad_norm": 0.0010455192532390356,
1605
+ "learning_rate": 5.644539083624701e-05,
1606
+ "loss": 0.0,
1607
+ "step": 1860
1608
+ },
1609
+ {
1610
+ "epoch": 4.64,
1611
+ "grad_norm": 0.00021354724594857544,
1612
+ "learning_rate": 5.605674550915161e-05,
1613
+ "loss": 0.0,
1614
+ "step": 1870
1615
+ },
1616
+ {
1617
+ "epoch": 4.67,
1618
+ "grad_norm": 0.00022464946960099041,
1619
+ "learning_rate": 5.566772843451495e-05,
1620
+ "loss": 0.0,
1621
+ "step": 1880
1622
+ },
1623
+ {
1624
+ "epoch": 4.69,
1625
+ "grad_norm": 6.343223503790796e-05,
1626
+ "learning_rate": 5.527836348920965e-05,
1627
+ "loss": 0.0,
1628
+ "step": 1890
1629
+ },
1630
+ {
1631
+ "epoch": 4.71,
1632
+ "grad_norm": 0.0007182627450674772,
1633
+ "learning_rate": 5.488867457145974e-05,
1634
+ "loss": 0.0,
1635
+ "step": 1900
1636
+ },
1637
+ {
1638
+ "epoch": 4.71,
1639
+ "eval_loss": 0.0006057315040379763,
1640
+ "eval_runtime": 16.8716,
1641
+ "eval_samples_per_second": 100.998,
1642
+ "eval_steps_per_second": 4.208,
1643
+ "step": 1900
1644
+ },
1645
+ {
1646
+ "epoch": 4.74,
1647
+ "grad_norm": 0.00021247757831588387,
1648
+ "learning_rate": 5.449868559937384e-05,
1649
+ "loss": 0.0,
1650
+ "step": 1910
1651
+ },
1652
+ {
1653
+ "epoch": 4.76,
1654
+ "grad_norm": 0.00025663134874776006,
1655
+ "learning_rate": 5.4108420509477145e-05,
1656
+ "loss": 0.0,
1657
+ "step": 1920
1658
+ },
1659
+ {
1660
+ "epoch": 4.79,
1661
+ "grad_norm": 0.00013280168059282005,
1662
+ "learning_rate": 5.3717903255242284e-05,
1663
+ "loss": 0.0,
1664
+ "step": 1930
1665
+ },
1666
+ {
1667
+ "epoch": 4.81,
1668
+ "grad_norm": 5.64151669095736e-05,
1669
+ "learning_rate": 5.3327157805619035e-05,
1670
+ "loss": 0.0,
1671
+ "step": 1940
1672
+ },
1673
+ {
1674
+ "epoch": 4.84,
1675
+ "grad_norm": 7.619507232448086e-05,
1676
+ "learning_rate": 5.293620814356326e-05,
1677
+ "loss": 0.0002,
1678
+ "step": 1950
1679
+ },
1680
+ {
1681
+ "epoch": 4.84,
1682
+ "eval_loss": 0.0001224023144459352,
1683
+ "eval_runtime": 16.8859,
1684
+ "eval_samples_per_second": 100.913,
1685
+ "eval_steps_per_second": 4.205,
1686
+ "step": 1950
1687
+ },
1688
+ {
1689
+ "epoch": 4.86,
1690
+ "grad_norm": 0.0001947753335116431,
1691
+ "learning_rate": 5.254507826456483e-05,
1692
+ "loss": 0.0,
1693
+ "step": 1960
1694
+ },
1695
+ {
1696
+ "epoch": 4.89,
1697
+ "grad_norm": 8.741691272007301e-05,
1698
+ "learning_rate": 5.215379217517492e-05,
1699
+ "loss": 0.0002,
1700
+ "step": 1970
1701
+ },
1702
+ {
1703
+ "epoch": 4.91,
1704
+ "grad_norm": 0.00012889545178040862,
1705
+ "learning_rate": 5.176237389153246e-05,
1706
+ "loss": 0.0,
1707
+ "step": 1980
1708
+ },
1709
+ {
1710
+ "epoch": 4.94,
1711
+ "grad_norm": 5.485777364810929e-05,
1712
+ "learning_rate": 5.137084743789012e-05,
1713
+ "loss": 0.0,
1714
+ "step": 1990
1715
+ },
1716
+ {
1717
+ "epoch": 4.96,
1718
+ "grad_norm": 0.00056857755407691,
1719
+ "learning_rate": 5.097923684513979e-05,
1720
+ "loss": 0.0,
1721
+ "step": 2000
1722
+ },
1723
+ {
1724
+ "epoch": 4.96,
1725
+ "eval_loss": 0.0007657750975340605,
1726
+ "eval_runtime": 16.8738,
1727
+ "eval_samples_per_second": 100.985,
1728
+ "eval_steps_per_second": 4.208,
1729
+ "step": 2000
1730
+ },
1731
+ {
1732
+ "epoch": 4.99,
1733
+ "grad_norm": 0.0005150920478627086,
1734
+ "learning_rate": 5.058756614933757e-05,
1735
+ "loss": 0.0,
1736
+ "step": 2010
1737
+ },
1738
+ {
1739
+ "epoch": 5.01,
1740
+ "grad_norm": 0.00031274749198928475,
1741
+ "learning_rate": 5.019585939022859e-05,
1742
+ "loss": 0.0,
1743
+ "step": 2020
1744
+ },
1745
+ {
1746
+ "epoch": 5.04,
1747
+ "grad_norm": 8.435914060100913e-05,
1748
+ "learning_rate": 4.980414060977142e-05,
1749
+ "loss": 0.0089,
1750
+ "step": 2030
1751
+ },
1752
+ {
1753
+ "epoch": 5.06,
1754
+ "grad_norm": 7.393726264126599e-05,
1755
+ "learning_rate": 4.9412433850662434e-05,
1756
+ "loss": 0.0,
1757
+ "step": 2040
1758
+ },
1759
+ {
1760
+ "epoch": 5.09,
1761
+ "grad_norm": 9.885100735118613e-05,
1762
+ "learning_rate": 4.902076315486023e-05,
1763
+ "loss": 0.0,
1764
+ "step": 2050
1765
+ },
1766
+ {
1767
+ "epoch": 5.09,
1768
+ "eval_loss": 0.00010947392001980916,
1769
+ "eval_runtime": 16.8917,
1770
+ "eval_samples_per_second": 100.878,
1771
+ "eval_steps_per_second": 4.203,
1772
+ "step": 2050
1773
+ },
1774
+ {
1775
+ "epoch": 5.11,
1776
+ "grad_norm": 0.0005973571678623557,
1777
+ "learning_rate": 4.86291525621099e-05,
1778
+ "loss": 0.0,
1779
+ "step": 2060
1780
+ },
1781
+ {
1782
+ "epoch": 5.14,
1783
+ "grad_norm": 5.217625221121125e-05,
1784
+ "learning_rate": 4.823762610846755e-05,
1785
+ "loss": 0.0,
1786
+ "step": 2070
1787
+ },
1788
+ {
1789
+ "epoch": 5.16,
1790
+ "grad_norm": 0.0002454821078572422,
1791
+ "learning_rate": 4.784620782482509e-05,
1792
+ "loss": 0.0,
1793
+ "step": 2080
1794
+ },
1795
+ {
1796
+ "epoch": 5.19,
1797
+ "grad_norm": 9.69076863839291e-05,
1798
+ "learning_rate": 4.745492173543517e-05,
1799
+ "loss": 0.0,
1800
+ "step": 2090
1801
+ },
1802
+ {
1803
+ "epoch": 5.21,
1804
+ "grad_norm": 8.641201566206291e-05,
1805
+ "learning_rate": 4.706379185643675e-05,
1806
+ "loss": 0.0,
1807
+ "step": 2100
1808
+ },
1809
+ {
1810
+ "epoch": 5.21,
1811
+ "eval_loss": 0.00016069799312390387,
1812
+ "eval_runtime": 16.8743,
1813
+ "eval_samples_per_second": 100.982,
1814
+ "eval_steps_per_second": 4.208,
1815
+ "step": 2100
1816
+ },
1817
+ {
1818
+ "epoch": 5.24,
1819
+ "grad_norm": 6.417043914552778e-05,
1820
+ "learning_rate": 4.667284219438098e-05,
1821
+ "loss": 0.0,
1822
+ "step": 2110
1823
+ },
1824
+ {
1825
+ "epoch": 5.26,
1826
+ "grad_norm": 0.00013219700485933572,
1827
+ "learning_rate": 4.628209674475773e-05,
1828
+ "loss": 0.0,
1829
+ "step": 2120
1830
+ },
1831
+ {
1832
+ "epoch": 5.29,
1833
+ "grad_norm": 6.98563817422837e-05,
1834
+ "learning_rate": 4.589157949052287e-05,
1835
+ "loss": 0.0,
1836
+ "step": 2130
1837
+ },
1838
+ {
1839
+ "epoch": 5.31,
1840
+ "grad_norm": 9.513930126558989e-05,
1841
+ "learning_rate": 4.550131440062617e-05,
1842
+ "loss": 0.0,
1843
+ "step": 2140
1844
+ },
1845
+ {
1846
+ "epoch": 5.33,
1847
+ "grad_norm": 5.327581675373949e-05,
1848
+ "learning_rate": 4.511132542854027e-05,
1849
+ "loss": 0.0,
1850
+ "step": 2150
1851
+ },
1852
+ {
1853
+ "epoch": 5.33,
1854
+ "eval_loss": 0.00017579211271367967,
1855
+ "eval_runtime": 16.8732,
1856
+ "eval_samples_per_second": 100.989,
1857
+ "eval_steps_per_second": 4.208,
1858
+ "step": 2150
1859
+ },
1860
+ {
1861
+ "epoch": 5.36,
1862
+ "grad_norm": 0.00013409934763330966,
1863
+ "learning_rate": 4.4721636510790365e-05,
1864
+ "loss": 0.0,
1865
+ "step": 2160
1866
+ },
1867
+ {
1868
+ "epoch": 5.38,
1869
+ "grad_norm": 8.724403596716002e-05,
1870
+ "learning_rate": 4.4332271565485053e-05,
1871
+ "loss": 0.0,
1872
+ "step": 2170
1873
+ },
1874
+ {
1875
+ "epoch": 5.41,
1876
+ "grad_norm": 6.426259642466903e-05,
1877
+ "learning_rate": 4.39432544908484e-05,
1878
+ "loss": 0.0,
1879
+ "step": 2180
1880
+ },
1881
+ {
1882
+ "epoch": 5.43,
1883
+ "grad_norm": 0.00012114330456824973,
1884
+ "learning_rate": 4.355460916375299e-05,
1885
+ "loss": 0.0,
1886
+ "step": 2190
1887
+ },
1888
+ {
1889
+ "epoch": 5.46,
1890
+ "grad_norm": 9.614200826035812e-05,
1891
+ "learning_rate": 4.3166359438254564e-05,
1892
+ "loss": 0.0,
1893
+ "step": 2200
1894
+ },
1895
+ {
1896
+ "epoch": 5.46,
1897
+ "eval_loss": 0.0004463446675799787,
1898
+ "eval_runtime": 16.8612,
1899
+ "eval_samples_per_second": 101.06,
1900
+ "eval_steps_per_second": 4.211,
1901
+ "step": 2200
1902
+ },
1903
+ {
1904
+ "epoch": 5.48,
1905
+ "grad_norm": 0.0002929608745034784,
1906
+ "learning_rate": 4.27785291441278e-05,
1907
+ "loss": 0.0,
1908
+ "step": 2210
1909
+ },
1910
+ {
1911
+ "epoch": 5.51,
1912
+ "grad_norm": 4.366716166259721e-05,
1913
+ "learning_rate": 4.239114208540374e-05,
1914
+ "loss": 0.0,
1915
+ "step": 2220
1916
+ },
1917
+ {
1918
+ "epoch": 5.53,
1919
+ "grad_norm": 3.860508149955422e-05,
1920
+ "learning_rate": 4.2004222038908845e-05,
1921
+ "loss": 0.0,
1922
+ "step": 2230
1923
+ },
1924
+ {
1925
+ "epoch": 5.56,
1926
+ "grad_norm": 5.3087976993992925e-05,
1927
+ "learning_rate": 4.161779275280546e-05,
1928
+ "loss": 0.0,
1929
+ "step": 2240
1930
+ },
1931
+ {
1932
+ "epoch": 5.58,
1933
+ "grad_norm": 0.00015269150026142597,
1934
+ "learning_rate": 4.123187794513441e-05,
1935
+ "loss": 0.0,
1936
+ "step": 2250
1937
+ },
1938
+ {
1939
+ "epoch": 5.58,
1940
+ "eval_loss": 0.00039042255957610905,
1941
+ "eval_runtime": 16.8755,
1942
+ "eval_samples_per_second": 100.975,
1943
+ "eval_steps_per_second": 4.207,
1944
+ "step": 2250
1945
+ },
1946
+ {
1947
+ "epoch": 5.61,
1948
+ "grad_norm": 0.00011781427747337148,
1949
+ "learning_rate": 4.084650130235909e-05,
1950
+ "loss": 0.0,
1951
+ "step": 2260
1952
+ },
1953
+ {
1954
+ "epoch": 5.63,
1955
+ "grad_norm": 7.544765685452148e-05,
1956
+ "learning_rate": 4.046168647791171e-05,
1957
+ "loss": 0.0,
1958
+ "step": 2270
1959
+ },
1960
+ {
1961
+ "epoch": 5.66,
1962
+ "grad_norm": 4.577730942401104e-05,
1963
+ "learning_rate": 4.007745709074154e-05,
1964
+ "loss": 0.0,
1965
+ "step": 2280
1966
+ },
1967
+ {
1968
+ "epoch": 5.68,
1969
+ "grad_norm": 0.00011316941527184099,
1970
+ "learning_rate": 3.969383672386518e-05,
1971
+ "loss": 0.0,
1972
+ "step": 2290
1973
+ },
1974
+ {
1975
+ "epoch": 5.71,
1976
+ "grad_norm": 4.33563363912981e-05,
1977
+ "learning_rate": 3.931084892291917e-05,
1978
+ "loss": 0.0,
1979
+ "step": 2300
1980
+ },
1981
+ {
1982
+ "epoch": 5.71,
1983
+ "eval_loss": 0.0004175197391305119,
1984
+ "eval_runtime": 16.8721,
1985
+ "eval_samples_per_second": 100.995,
1986
+ "eval_steps_per_second": 4.208,
1987
+ "step": 2300
1988
+ },
1989
+ {
1990
+ "epoch": 5.73,
1991
+ "grad_norm": 3.2948159059742466e-05,
1992
+ "learning_rate": 3.8928517194714706e-05,
1993
+ "loss": 0.0,
1994
+ "step": 2310
1995
+ },
1996
+ {
1997
+ "epoch": 5.76,
1998
+ "grad_norm": 6.737465446349233e-05,
1999
+ "learning_rate": 3.854686500579494e-05,
2000
+ "loss": 0.0,
2001
+ "step": 2320
2002
+ },
2003
+ {
2004
+ "epoch": 5.78,
2005
+ "grad_norm": 8.004607661860064e-05,
2006
+ "learning_rate": 3.816591578099468e-05,
2007
+ "loss": 0.0,
2008
+ "step": 2330
2009
+ },
2010
+ {
2011
+ "epoch": 5.81,
2012
+ "grad_norm": 6.167278479551896e-05,
2013
+ "learning_rate": 3.778569290200254e-05,
2014
+ "loss": 0.0,
2015
+ "step": 2340
2016
+ },
2017
+ {
2018
+ "epoch": 5.83,
2019
+ "grad_norm": 0.0003750508767552674,
2020
+ "learning_rate": 3.7406219705925945e-05,
2021
+ "loss": 0.0,
2022
+ "step": 2350
2023
+ },
2024
+ {
2025
+ "epoch": 5.83,
2026
+ "eval_loss": 0.0003901799500454217,
2027
+ "eval_runtime": 16.875,
2028
+ "eval_samples_per_second": 100.978,
2029
+ "eval_steps_per_second": 4.207,
2030
+ "step": 2350
2031
+ },
2032
+ {
2033
+ "epoch": 5.86,
2034
+ "grad_norm": 4.658477337216027e-05,
2035
+ "learning_rate": 3.702751948385864e-05,
2036
+ "loss": 0.0,
2037
+ "step": 2360
2038
+ },
2039
+ {
2040
+ "epoch": 5.88,
2041
+ "grad_norm": 0.00017428740102332085,
2042
+ "learning_rate": 3.6649615479451226e-05,
2043
+ "loss": 0.0,
2044
+ "step": 2370
2045
+ },
2046
+ {
2047
+ "epoch": 5.91,
2048
+ "grad_norm": 0.0001236279495060444,
2049
+ "learning_rate": 3.6272530887484535e-05,
2050
+ "loss": 0.0,
2051
+ "step": 2380
2052
+ },
2053
+ {
2054
+ "epoch": 5.93,
2055
+ "grad_norm": 7.120491500245407e-05,
2056
+ "learning_rate": 3.5896288852445906e-05,
2057
+ "loss": 0.0,
2058
+ "step": 2390
2059
+ },
2060
+ {
2061
+ "epoch": 5.96,
2062
+ "grad_norm": 9.089993545785546e-05,
2063
+ "learning_rate": 3.552091246710875e-05,
2064
+ "loss": 0.0,
2065
+ "step": 2400
2066
+ },
2067
+ {
2068
+ "epoch": 5.96,
2069
+ "eval_loss": 0.00044537056237459183,
2070
+ "eval_runtime": 16.8732,
2071
+ "eval_samples_per_second": 100.989,
2072
+ "eval_steps_per_second": 4.208,
2073
+ "step": 2400
2074
+ },
2075
+ {
2076
+ "epoch": 5.98,
2077
+ "grad_norm": 3.8834048609714955e-05,
2078
+ "learning_rate": 3.514642477111505e-05,
2079
+ "loss": 0.0,
2080
+ "step": 2410
2081
+ },
2082
+ {
2083
+ "epoch": 6.0,
2084
+ "grad_norm": 0.0027178339660167694,
2085
+ "learning_rate": 3.477284874956134e-05,
2086
+ "loss": 0.0,
2087
+ "step": 2420
2088
+ },
2089
+ {
2090
+ "epoch": 6.03,
2091
+ "grad_norm": 0.0001490208087489009,
2092
+ "learning_rate": 3.440020733158792e-05,
2093
+ "loss": 0.0,
2094
+ "step": 2430
2095
+ },
2096
+ {
2097
+ "epoch": 6.05,
2098
+ "grad_norm": 3.5653181839734316e-05,
2099
+ "learning_rate": 3.402852338897151e-05,
2100
+ "loss": 0.0,
2101
+ "step": 2440
2102
+ },
2103
+ {
2104
+ "epoch": 6.08,
2105
+ "grad_norm": 5.699761095456779e-05,
2106
+ "learning_rate": 3.365781973472144e-05,
2107
+ "loss": 0.0,
2108
+ "step": 2450
2109
+ },
2110
+ {
2111
+ "epoch": 6.08,
2112
+ "eval_loss": 0.00038982907426543534,
2113
+ "eval_runtime": 16.8883,
2114
+ "eval_samples_per_second": 100.899,
2115
+ "eval_steps_per_second": 4.204,
2116
+ "step": 2450
2117
+ },
2118
+ {
2119
+ "epoch": 6.1,
2120
+ "grad_norm": 5.417722786660306e-05,
2121
+ "learning_rate": 3.328811912167945e-05,
2122
+ "loss": 0.0,
2123
+ "step": 2460
2124
+ },
2125
+ {
2126
+ "epoch": 6.13,
2127
+ "grad_norm": 3.579606709536165e-05,
2128
+ "learning_rate": 3.291944424112314e-05,
2129
+ "loss": 0.0,
2130
+ "step": 2470
2131
+ },
2132
+ {
2133
+ "epoch": 6.15,
2134
+ "grad_norm": 7.697240653214976e-05,
2135
+ "learning_rate": 3.2551817721373335e-05,
2136
+ "loss": 0.0,
2137
+ "step": 2480
2138
+ },
2139
+ {
2140
+ "epoch": 6.18,
2141
+ "grad_norm": 4.500555223785341e-05,
2142
+ "learning_rate": 3.218526212640511e-05,
2143
+ "loss": 0.0,
2144
+ "step": 2490
2145
+ },
2146
+ {
2147
+ "epoch": 6.2,
2148
+ "grad_norm": 7.687577453907579e-05,
2149
+ "learning_rate": 3.1819799954462983e-05,
2150
+ "loss": 0.0,
2151
+ "step": 2500
2152
+ },
2153
+ {
2154
+ "epoch": 6.2,
2155
+ "eval_loss": 0.00047402523341588676,
2156
+ "eval_runtime": 16.889,
2157
+ "eval_samples_per_second": 100.894,
2158
+ "eval_steps_per_second": 4.204,
2159
+ "step": 2500
2160
+ },
2161
+ {
2162
+ "epoch": 6.23,
2163
+ "grad_norm": 6.375756493071094e-05,
2164
+ "learning_rate": 3.1455453636679866e-05,
2165
+ "loss": 0.0,
2166
+ "step": 2510
2167
+ },
2168
+ {
2169
+ "epoch": 6.25,
2170
+ "grad_norm": 0.00015971327957231551,
2171
+ "learning_rate": 3.109224553570047e-05,
2172
+ "loss": 0.0,
2173
+ "step": 2520
2174
+ },
2175
+ {
2176
+ "epoch": 6.28,
2177
+ "grad_norm": 0.00012273038737475872,
2178
+ "learning_rate": 3.073019794430866e-05,
2179
+ "loss": 0.0,
2180
+ "step": 2530
2181
+ },
2182
+ {
2183
+ "epoch": 6.3,
2184
+ "grad_norm": 5.869157030247152e-05,
2185
+ "learning_rate": 3.0369333084059148e-05,
2186
+ "loss": 0.0,
2187
+ "step": 2540
2188
+ },
2189
+ {
2190
+ "epoch": 6.33,
2191
+ "grad_norm": 9.306817082688212e-05,
2192
+ "learning_rate": 3.0009673103913725e-05,
2193
+ "loss": 0.0,
2194
+ "step": 2550
2195
+ },
2196
+ {
2197
+ "epoch": 6.33,
2198
+ "eval_loss": 0.00047384999925270677,
2199
+ "eval_runtime": 16.8848,
2200
+ "eval_samples_per_second": 100.919,
2201
+ "eval_steps_per_second": 4.205,
2202
+ "step": 2550
2203
+ },
2204
+ {
2205
+ "epoch": 6.35,
2206
+ "grad_norm": 6.814825610490516e-05,
2207
+ "learning_rate": 2.9651240078881625e-05,
2208
+ "loss": 0.0,
2209
+ "step": 2560
2210
+ },
2211
+ {
2212
+ "epoch": 6.38,
2213
+ "grad_norm": 4.082039595232345e-05,
2214
+ "learning_rate": 2.9294056008664762e-05,
2215
+ "loss": 0.0,
2216
+ "step": 2570
2217
+ },
2218
+ {
2219
+ "epoch": 6.4,
2220
+ "grad_norm": 4.88782134198118e-05,
2221
+ "learning_rate": 2.893814281630744e-05,
2222
+ "loss": 0.0,
2223
+ "step": 2580
2224
+ },
2225
+ {
2226
+ "epoch": 6.43,
2227
+ "grad_norm": 0.0004272049991413951,
2228
+ "learning_rate": 2.8583522346850632e-05,
2229
+ "loss": 0.0,
2230
+ "step": 2590
2231
+ },
2232
+ {
2233
+ "epoch": 6.45,
2234
+ "grad_norm": 0.00017715679132379591,
2235
+ "learning_rate": 2.823021636599137e-05,
2236
+ "loss": 0.0,
2237
+ "step": 2600
2238
+ },
2239
+ {
2240
+ "epoch": 6.45,
2241
+ "eval_loss": 0.0005336844478733838,
2242
+ "eval_runtime": 16.8679,
2243
+ "eval_samples_per_second": 101.02,
2244
+ "eval_steps_per_second": 4.209,
2245
+ "step": 2600
2246
+ },
2247
+ {
2248
+ "epoch": 6.48,
2249
+ "grad_norm": 5.479040555655956e-05,
2250
+ "learning_rate": 2.787824655874674e-05,
2251
+ "loss": 0.0,
2252
+ "step": 2610
2253
+ },
2254
+ {
2255
+ "epoch": 6.5,
2256
+ "grad_norm": 4.12812405556906e-05,
2257
+ "learning_rate": 2.7527634528122847e-05,
2258
+ "loss": 0.0,
2259
+ "step": 2620
2260
+ },
2261
+ {
2262
+ "epoch": 6.53,
2263
+ "grad_norm": 5.893718480365351e-05,
2264
+ "learning_rate": 2.7178401793789012e-05,
2265
+ "loss": 0.0,
2266
+ "step": 2630
2267
+ },
2268
+ {
2269
+ "epoch": 6.55,
2270
+ "grad_norm": 0.00015176544548012316,
2271
+ "learning_rate": 2.6830569790756805e-05,
2272
+ "loss": 0.0,
2273
+ "step": 2640
2274
+ },
2275
+ {
2276
+ "epoch": 6.58,
2277
+ "grad_norm": 4.401996557135135e-05,
2278
+ "learning_rate": 2.6484159868064583e-05,
2279
+ "loss": 0.0,
2280
+ "step": 2650
2281
+ },
2282
+ {
2283
+ "epoch": 6.58,
2284
+ "eval_loss": 0.0004447665123734623,
2285
+ "eval_runtime": 16.8689,
2286
+ "eval_samples_per_second": 101.014,
2287
+ "eval_steps_per_second": 4.209,
2288
+ "step": 2650
2289
+ },
2290
+ {
2291
+ "epoch": 6.58,
2292
+ "step": 2650,
2293
+ "total_flos": 9.885928322204959e+17,
2294
+ "train_loss": 0.1188159689292002,
2295
+ "train_runtime": 2950.0845,
2296
+ "train_samples_per_second": 32.731,
2297
+ "train_steps_per_second": 1.366
2298
+ }
2299
+ ],
2300
+ "logging_steps": 10,
2301
+ "max_steps": 4030,
2302
+ "num_input_tokens_seen": 0,
2303
+ "num_train_epochs": 10,
2304
+ "save_steps": 500,
2305
+ "total_flos": 9.885928322204959e+17,
2306
+ "train_batch_size": 8,
2307
+ "trial_name": null,
2308
+ "trial_params": null
2309
+ }
llama2_13b_peft/formal_fallacies_syllogisms_negation/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1559a7415ad61fab7820fa5c6cd769ade83d2e33096f4ebf8b568f21882c89d7
3
+ size 5112
llama2_13b_peft/formal_fallacies_syllogisms_negation/training_eval_loss.png ADDED
llama2_13b_peft/formal_fallacies_syllogisms_negation/training_loss.png ADDED
llama2_13b_peft/goal_step_wikihow/README.md ADDED
@@ -0,0 +1,71 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: other
3
+ library_name: peft
4
+ tags:
5
+ - llama-factory
6
+ - lora
7
+ - generated_from_trainer
8
+ base_model: /data1/model/llama2/meta-llama/Llama2-13b
9
+ model-index:
10
+ - name: goal_step_wikihow_no_sys
11
+ results: []
12
+ ---
13
+
14
+ <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
+ should probably proofread and complete it, then remove this comment. -->
16
+
17
+ # goal_step_wikihow_no_sys
18
+
19
+ This model is a fine-tuned version of [/data1/model/llama2/meta-llama/Llama2-13b](https://huggingface.co//data1/model/llama2/meta-llama/Llama2-13b) on the goal_step_wikihow_no_sys dataset.
20
+ It achieves the following results on the evaluation set:
21
+ - Loss: 0.0155
22
+
23
+ ## Model description
24
+
25
+ More information needed
26
+
27
+ ## Intended uses & limitations
28
+
29
+ More information needed
30
+
31
+ ## Training and evaluation data
32
+
33
+ More information needed
34
+
35
+ ## Training procedure
36
+
37
+ ### Training hyperparameters
38
+
39
+ The following hyperparameters were used during training:
40
+ - learning_rate: 5e-05
41
+ - train_batch_size: 4
42
+ - eval_batch_size: 4
43
+ - seed: 42
44
+ - distributed_type: multi-GPU
45
+ - num_devices: 2
46
+ - total_train_batch_size: 8
47
+ - total_eval_batch_size: 8
48
+ - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
49
+ - lr_scheduler_type: cosine
50
+ - lr_scheduler_warmup_steps: 20
51
+ - num_epochs: 5.0
52
+
53
+ ### Training results
54
+
55
+ | Training Loss | Epoch | Step | Validation Loss |
56
+ |:-------------:|:------:|:----:|:---------------:|
57
+ | 0.0578 | 0.1493 | 100 | 0.0387 |
58
+ | 0.0211 | 0.2985 | 200 | 0.0247 |
59
+ | 0.0188 | 0.4478 | 300 | 0.0155 |
60
+ | 0.0348 | 0.5970 | 400 | 0.0209 |
61
+ | 0.0174 | 0.7463 | 500 | 0.0189 |
62
+ | 0.0322 | 0.8955 | 600 | 0.0163 |
63
+
64
+
65
+ ### Framework versions
66
+
67
+ - PEFT 0.10.0
68
+ - Transformers 4.40.0
69
+ - Pytorch 2.2.1
70
+ - Datasets 2.18.0
71
+ - Tokenizers 0.19.1
llama2_13b_peft/goal_step_wikihow/adapter_config.json ADDED
@@ -0,0 +1,34 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha_pattern": {},
3
+ "auto_mapping": null,
4
+ "base_model_name_or_path": "/data1/model/llama2/meta-llama/Llama2-13b",
5
+ "bias": "none",
6
+ "fan_in_fan_out": false,
7
+ "inference_mode": true,
8
+ "init_lora_weights": true,
9
+ "layer_replication": null,
10
+ "layers_pattern": null,
11
+ "layers_to_transform": null,
12
+ "loftq_config": {},
13
+ "lora_alpha": 16,
14
+ "lora_dropout": 0.0,
15
+ "megatron_config": null,
16
+ "megatron_core": "megatron.core",
17
+ "modules_to_save": null,
18
+ "peft_type": "LORA",
19
+ "r": 8,
20
+ "rank_pattern": {},
21
+ "revision": null,
22
+ "target_modules": [
23
+ "down_proj",
24
+ "v_proj",
25
+ "gate_proj",
26
+ "k_proj",
27
+ "q_proj",
28
+ "o_proj",
29
+ "up_proj"
30
+ ],
31
+ "task_type": "CAUSAL_LM",
32
+ "use_dora": false,
33
+ "use_rslora": false
34
+ }
llama2_13b_peft/goal_step_wikihow/adapter_model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:91fd68ef1b71aef1f8c0fa41c4cee8b73ffa36a061e9facae7af29e2c9792bc9
3
+ size 125248064
llama2_13b_peft/goal_step_wikihow/all_results.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 0.8955223880597015,
3
+ "eval_loss": 0.01549521740525961,
4
+ "eval_runtime": 3.3948,
5
+ "eval_samples_per_second": 83.363,
6
+ "eval_steps_per_second": 10.605,
7
+ "total_flos": 4.751779802854195e+16,
8
+ "train_loss": 0.0911922578016917,
9
+ "train_runtime": 229.4716,
10
+ "train_samples_per_second": 116.79,
11
+ "train_steps_per_second": 14.599
12
+ }
llama2_13b_peft/goal_step_wikihow/eval_results.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "epoch": 0.8955223880597015,
3
+ "eval_loss": 0.01549521740525961,
4
+ "eval_runtime": 3.3948,
5
+ "eval_samples_per_second": 83.363,
6
+ "eval_steps_per_second": 10.605
7
+ }
llama2_13b_peft/goal_step_wikihow/special_tokens_map.json ADDED
@@ -0,0 +1,24 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "eos_token": {
10
+ "content": "</s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "pad_token": "</s>",
17
+ "unk_token": {
18
+ "content": "<unk>",
19
+ "lstrip": false,
20
+ "normalized": false,
21
+ "rstrip": false,
22
+ "single_word": false
23
+ }
24
+ }
llama2_13b_peft/goal_step_wikihow/tokenizer.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
3
+ size 499723
llama2_13b_peft/goal_step_wikihow/tokenizer_config.json ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_bos_token": true,
3
+ "add_eos_token": false,
4
+ "add_prefix_space": true,
5
+ "added_tokens_decoder": {
6
+ "0": {
7
+ "content": "<unk>",
8
+ "lstrip": false,
9
+ "normalized": false,
10
+ "rstrip": false,
11
+ "single_word": false,
12
+ "special": true
13
+ },
14
+ "1": {
15
+ "content": "<s>",
16
+ "lstrip": false,
17
+ "normalized": false,
18
+ "rstrip": false,
19
+ "single_word": false,
20
+ "special": true
21
+ },
22
+ "2": {
23
+ "content": "</s>",
24
+ "lstrip": false,
25
+ "normalized": false,
26
+ "rstrip": false,
27
+ "single_word": false,
28
+ "special": true
29
+ }
30
+ },
31
+ "bos_token": "<s>",
32
+ "chat_template": "{% if messages[0]['role'] == 'system' %}{% set system_message = messages[0]['content'] %}{% endif %}{% if system_message is defined %}{{ system_message }}{% endif %}{% for message in messages %}{% set content = message['content'] %}{% if message['role'] == 'user' %}{{ content }}{% elif message['role'] == 'assistant' %}{{ content + '\\n' }}{% endif %}{% endfor %}",
33
+ "clean_up_tokenization_spaces": false,
34
+ "eos_token": "</s>",
35
+ "legacy": true,
36
+ "model_max_length": 1000000000000000019884624838656,
37
+ "pad_token": "</s>",
38
+ "padding_side": "right",
39
+ "sp_model_kwargs": {},
40
+ "spaces_between_special_tokens": false,
41
+ "split_special_tokens": false,
42
+ "tokenizer_class": "LlamaTokenizer",
43
+ "unk_token": "<unk>",
44
+ "use_default_system_prompt": false
45
+ }