Hugging Face's logo

dzungpham
/
graphcodebert-code-classification

Transformers
Safetensors
detection
AI-generated
bert
Model card Files
xet
 Community
dzungpham commited on
Commit
34959f8
·
verified ·
1 Parent(s): 4f8c4ed

upload checkpoints

Browse files
Files changed (13) hide show
  1. fourier-spectral-norm-classifier/checkpoint-1500/config.json +29 -0
  2. fourier-spectral-norm-classifier/checkpoint-1500/config_hyperparams.json +82 -0
  3. fourier-spectral-norm-classifier/checkpoint-1500/merges.txt +0 -0
  4. fourier-spectral-norm-classifier/checkpoint-1500/model.safetensors +3 -0
  5. fourier-spectral-norm-classifier/checkpoint-1500/optimizer.pt +3 -0
  6. fourier-spectral-norm-classifier/checkpoint-1500/rng_state.pth +3 -0
  7. fourier-spectral-norm-classifier/checkpoint-1500/scheduler.pt +3 -0
  8. fourier-spectral-norm-classifier/checkpoint-1500/special_tokens_map.json +51 -0
  9. fourier-spectral-norm-classifier/checkpoint-1500/tokenizer.json +0 -0
  10. fourier-spectral-norm-classifier/checkpoint-1500/tokenizer_config.json +58 -0
  11. fourier-spectral-norm-classifier/checkpoint-1500/train_config.yaml +55 -0
  12. fourier-spectral-norm-classifier/checkpoint-1500/trainer_state.json +2184 -0
  13. fourier-spectral-norm-classifier/checkpoint-1500/vocab.json +0 -0
fourier-spectral-norm-classifier/checkpoint-1500/config.json ADDED
@@ -0,0 +1,29 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "RobertaForSequenceClassification"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.3,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": 0.4,
8
+ "dtype": "float32",
9
+ "eos_token_id": 2,
10
+ "gradient_checkpointing": false,
11
+ "hidden_act": "gelu",
12
+ "hidden_dropout_prob": 0.3,
13
+ "hidden_size": 768,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 3072,
16
+ "layer_norm_eps": 1e-05,
17
+ "max_position_embeddings": 514,
18
+ "model_type": "roberta",
19
+ "num_attention_heads": 12,
20
+ "num_hidden_layers": 12,
21
+ "output_past": true,
22
+ "pad_token_id": 1,
23
+ "position_embedding_type": "absolute",
24
+ "problem_type": "single_label_classification",
25
+ "transformers_version": "4.56.0",
26
+ "type_vocab_size": 1,
27
+ "use_cache": true,
28
+ "vocab_size": 50265
29
+ }
fourier-spectral-norm-classifier/checkpoint-1500/config_hyperparams.json ADDED
@@ -0,0 +1,82 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "train_config": {
3
+ "TEST": false,
4
+ "model_name": "/kaggle/input/models/dzung271828/microsoft-graphcodebert-base/transformers/default/1",
5
+ "output_dir": "training/fourier-spectral-norm-classifier/",
6
+ "num_epochs": 5,
7
+ "max_steps": -1,
8
+ "batch_size": 512,
9
+ "learning_rate": 1e-06,
10
+ "max_length": 512,
11
+ "num_labels": 2,
12
+ "use_wandb": false,
13
+ "freeze_base": true,
14
+ "loss_type": "ce",
15
+ "focal_alpha": 1.0,
16
+ "focal_gamma": 2.0,
17
+ "r_drop_alpha": 6.0,
18
+ "infonce_temperature": 0.07,
19
+ "infonce_weight": 0.5,
20
+ "seed": 42,
21
+ "resume_from_checkpoint": null,
22
+ "save_steps": 500,
23
+ "eval_steps": 500,
24
+ "logging_steps": 5,
25
+ "label_smoothing": 0.5,
26
+ "adversarial_epsilon": 0.5,
27
+ "use_swa": true,
28
+ "swa_start_epoch": 0,
29
+ "swa_lr": 1e-06,
30
+ "data_augmentation": true,
31
+ "aug_rename_prob": 0.7,
32
+ "aug_format_prob": 0.7,
33
+ "weight_decay": 0.1,
34
+ "mixup_alpha": 1.0,
35
+ "low_pass_keep_ratio": 0.5,
36
+ "freq_consistency_weight": 0.2,
37
+ "use_mixcode": true,
38
+ "use_fgm": true,
39
+ "fgm_freq": 5,
40
+ "use_r_drop": true,
41
+ "use_freq_consistency_loss": true,
42
+ "use_attn_spectral": false,
43
+ "attn_spectral_weight": 0.1,
44
+ "attn_spectral_cutoff_ratio": 0.25,
45
+ "hidden_dropout_prob": 0.3,
46
+ "attention_probs_dropout_prob": 0.3,
47
+ "classifier_dropout": 0.4,
48
+ "device": "cuda",
49
+ "torch_compile": true,
50
+ "cache_dir": "./tokenized_cache",
51
+ "use_swa_actual": true,
52
+ "use_fgm_actual": true,
53
+ "use_r_drop_actual": true,
54
+ "use_mixcode_actual": true,
55
+ "use_attn_spectral_actual": false,
56
+ "use_freq_consistency_loss_actual": true,
57
+ "use_spectral_norm": true
58
+ },
59
+ "training_arguments": {
60
+ "output_dir": "training/fourier-spectral-norm-classifier/",
61
+ "num_train_epochs": 5,
62
+ "per_device_train_batch_size": 512,
63
+ "per_device_eval_batch_size": 1024,
64
+ "learning_rate": 1e-06,
65
+ "warmup_steps": 488,
66
+ "weight_decay": 0.1,
67
+ "logging_steps": 5,
68
+ "eval_steps": 500,
69
+ "save_steps": 500,
70
+ "metric_for_best_model": "macro_f1",
71
+ "greater_is_better": true,
72
+ "save_total_limit": 5,
73
+ "fp16": false,
74
+ "seed": 42
75
+ },
76
+ "training_state": {
77
+ "global_step": 1500,
78
+ "epoch": 1.5353121801432958,
79
+ "best_metric": 0.6724504812400831,
80
+ "best_model_checkpoint": "training/fourier-spectral-norm-classifier/checkpoint-1000"
81
+ }
82
+ }
fourier-spectral-norm-classifier/checkpoint-1500/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
fourier-spectral-norm-classifier/checkpoint-1500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5e4bc60c5b8f6a36e32e0d39d6e8298433aad8375eb96017225baf2ef95e07ce
3
+ size 498619200
fourier-spectral-norm-classifier/checkpoint-1500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a83a04a97cc63d010ee0194538f8c07493223183d14eab2d89cd2b4303321ee2
3
+ size 4741923
fourier-spectral-norm-classifier/checkpoint-1500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e5909ab7c9c269b6f87c8f78a629bf9a4d31ba187687dd61dda9c2b3eec7a4b6
3
+ size 14645
fourier-spectral-norm-classifier/checkpoint-1500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b851d3befbd44b60717bbc46c57e884e1a9f534520ed08499a8654800370d4cd
3
+ size 1465
fourier-spectral-norm-classifier/checkpoint-1500/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": true,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": true,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": true,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": true,
26
+ "normalized": true,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": true,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": true,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": true,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
fourier-spectral-norm-classifier/checkpoint-1500/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
fourier-spectral-norm-classifier/checkpoint-1500/tokenizer_config.json ADDED
@@ -0,0 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "added_tokens_decoder": {
4
+ "0": {
5
+ "content": "<s>",
6
+ "lstrip": false,
7
+ "normalized": true,
8
+ "rstrip": false,
9
+ "single_word": false,
10
+ "special": true
11
+ },
12
+ "1": {
13
+ "content": "<pad>",
14
+ "lstrip": false,
15
+ "normalized": true,
16
+ "rstrip": false,
17
+ "single_word": false,
18
+ "special": true
19
+ },
20
+ "2": {
21
+ "content": "</s>",
22
+ "lstrip": false,
23
+ "normalized": true,
24
+ "rstrip": false,
25
+ "single_word": false,
26
+ "special": true
27
+ },
28
+ "3": {
29
+ "content": "<unk>",
30
+ "lstrip": false,
31
+ "normalized": true,
32
+ "rstrip": false,
33
+ "single_word": false,
34
+ "special": true
35
+ },
36
+ "50264": {
37
+ "content": "<mask>",
38
+ "lstrip": true,
39
+ "normalized": true,
40
+ "rstrip": false,
41
+ "single_word": false,
42
+ "special": true
43
+ }
44
+ },
45
+ "bos_token": "<s>",
46
+ "clean_up_tokenization_spaces": false,
47
+ "cls_token": "<s>",
48
+ "eos_token": "</s>",
49
+ "errors": "replace",
50
+ "extra_special_tokens": {},
51
+ "mask_token": "<mask>",
52
+ "model_max_length": 512,
53
+ "pad_token": "<pad>",
54
+ "sep_token": "</s>",
55
+ "tokenizer_class": "RobertaTokenizer",
56
+ "trim_offsets": true,
57
+ "unk_token": "<unk>"
58
+ }
fourier-spectral-norm-classifier/checkpoint-1500/train_config.yaml ADDED
@@ -0,0 +1,55 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ TEST: false
2
+ model_name: /kaggle/input/models/dzung271828/microsoft-graphcodebert-base/transformers/default/1
3
+ output_dir: training/fourier-spectral-norm-classifier/
4
+ num_epochs: 5
5
+ max_steps: -1
6
+ batch_size: 512
7
+ learning_rate: 1.0e-06
8
+ max_length: 512
9
+ num_labels: 2
10
+ use_wandb: false
11
+ freeze_base: true
12
+ loss_type: ce
13
+ focal_alpha: 1.0
14
+ focal_gamma: 2.0
15
+ r_drop_alpha: 6.0
16
+ infonce_temperature: 0.07
17
+ infonce_weight: 0.5
18
+ seed: 42
19
+ resume_from_checkpoint: null
20
+ save_steps: 500
21
+ eval_steps: 500
22
+ logging_steps: 5
23
+ label_smoothing: 0.5
24
+ adversarial_epsilon: 0.5
25
+ use_swa: true
26
+ swa_start_epoch: 0
27
+ swa_lr: 1.0e-06
28
+ data_augmentation: true
29
+ aug_rename_prob: 0.7
30
+ aug_format_prob: 0.7
31
+ weight_decay: 0.1
32
+ mixup_alpha: 1.0
33
+ low_pass_keep_ratio: 0.5
34
+ freq_consistency_weight: 0.2
35
+ use_mixcode: true
36
+ use_fgm: true
37
+ fgm_freq: 5
38
+ use_r_drop: true
39
+ use_freq_consistency_loss: true
40
+ use_attn_spectral: false
41
+ attn_spectral_weight: 0.1
42
+ attn_spectral_cutoff_ratio: 0.25
43
+ hidden_dropout_prob: 0.3
44
+ attention_probs_dropout_prob: 0.3
45
+ classifier_dropout: 0.4
46
+ device: cuda
47
+ torch_compile: true
48
+ cache_dir: ./tokenized_cache
49
+ use_swa_actual: true
50
+ use_fgm_actual: true
51
+ use_r_drop_actual: true
52
+ use_mixcode_actual: true
53
+ use_attn_spectral_actual: false
54
+ use_freq_consistency_loss_actual: true
55
+ use_spectral_norm: true
fourier-spectral-norm-classifier/checkpoint-1500/trainer_state.json ADDED
@@ -0,0 +1,2184 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": 1000,
3
+ "best_metric": 0.6724504812400831,
4
+ "best_model_checkpoint": "training/fourier-spectral-norm-classifier/checkpoint-1000",
5
+ "epoch": 1.5353121801432958,
6
+ "eval_steps": 500,
7
+ "global_step": 1500,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "SWA": "started",
14
+ "epoch": 0,
15
+ "step": 0
16
+ },
17
+ {
18
+ "epoch": 0.00511770726714432,
19
+ "grad_norm": 1.7937116622924805,
20
+ "learning_rate": 8.19672131147541e-09,
21
+ "loss": 0.8149,
22
+ "step": 5
23
+ },
24
+ {
25
+ "epoch": 0.01023541453428864,
26
+ "grad_norm": 1.8986879587173462,
27
+ "learning_rate": 1.844262295081967e-08,
28
+ "loss": 0.8145,
29
+ "step": 10
30
+ },
31
+ {
32
+ "epoch": 0.015353121801432957,
33
+ "grad_norm": 1.8692522048950195,
34
+ "learning_rate": 2.8688524590163933e-08,
35
+ "loss": 0.8031,
36
+ "step": 15
37
+ },
38
+ {
39
+ "epoch": 0.02047082906857728,
40
+ "grad_norm": 1.6589646339416504,
41
+ "learning_rate": 3.8934426229508196e-08,
42
+ "loss": 0.8208,
43
+ "step": 20
44
+ },
45
+ {
46
+ "epoch": 0.0255885363357216,
47
+ "grad_norm": 2.377978563308716,
48
+ "learning_rate": 4.918032786885246e-08,
49
+ "loss": 0.8054,
50
+ "step": 25
51
+ },
52
+ {
53
+ "epoch": 0.030706243602865915,
54
+ "grad_norm": 2.000364065170288,
55
+ "learning_rate": 5.9426229508196716e-08,
56
+ "loss": 0.8064,
57
+ "step": 30
58
+ },
59
+ {
60
+ "epoch": 0.03582395087001024,
61
+ "grad_norm": 1.8844542503356934,
62
+ "learning_rate": 6.967213114754098e-08,
63
+ "loss": 0.8047,
64
+ "step": 35
65
+ },
66
+ {
67
+ "epoch": 0.04094165813715456,
68
+ "grad_norm": 2.0933573246002197,
69
+ "learning_rate": 7.991803278688524e-08,
70
+ "loss": 0.8156,
71
+ "step": 40
72
+ },
73
+ {
74
+ "epoch": 0.04605936540429888,
75
+ "grad_norm": 1.8126033544540405,
76
+ "learning_rate": 9.01639344262295e-08,
77
+ "loss": 0.8074,
78
+ "step": 45
79
+ },
80
+ {
81
+ "epoch": 0.0511770726714432,
82
+ "grad_norm": 2.5709195137023926,
83
+ "learning_rate": 1.0040983606557377e-07,
84
+ "loss": 0.8124,
85
+ "step": 50
86
+ },
87
+ {
88
+ "epoch": 0.05629477993858751,
89
+ "grad_norm": 2.1875293254852295,
90
+ "learning_rate": 1.1065573770491803e-07,
91
+ "loss": 0.8143,
92
+ "step": 55
93
+ },
94
+ {
95
+ "epoch": 0.06141248720573183,
96
+ "grad_norm": 2.0810351371765137,
97
+ "learning_rate": 1.209016393442623e-07,
98
+ "loss": 0.8149,
99
+ "step": 60
100
+ },
101
+ {
102
+ "epoch": 0.06653019447287616,
103
+ "grad_norm": 1.7912037372589111,
104
+ "learning_rate": 1.3114754098360656e-07,
105
+ "loss": 0.8022,
106
+ "step": 65
107
+ },
108
+ {
109
+ "epoch": 0.07164790174002048,
110
+ "grad_norm": 1.7301534414291382,
111
+ "learning_rate": 1.413934426229508e-07,
112
+ "loss": 0.8149,
113
+ "step": 70
114
+ },
115
+ {
116
+ "epoch": 0.0767656090071648,
117
+ "grad_norm": 1.9520158767700195,
118
+ "learning_rate": 1.5163934426229508e-07,
119
+ "loss": 0.8201,
120
+ "step": 75
121
+ },
122
+ {
123
+ "epoch": 0.08188331627430911,
124
+ "grad_norm": 2.11938214302063,
125
+ "learning_rate": 1.6188524590163935e-07,
126
+ "loss": 0.8079,
127
+ "step": 80
128
+ },
129
+ {
130
+ "epoch": 0.08700102354145343,
131
+ "grad_norm": 2.1483607292175293,
132
+ "learning_rate": 1.7213114754098358e-07,
133
+ "loss": 0.8084,
134
+ "step": 85
135
+ },
136
+ {
137
+ "epoch": 0.09211873080859775,
138
+ "grad_norm": 2.1716372966766357,
139
+ "learning_rate": 1.8237704918032787e-07,
140
+ "loss": 0.8188,
141
+ "step": 90
142
+ },
143
+ {
144
+ "epoch": 0.09723643807574207,
145
+ "grad_norm": 2.3327996730804443,
146
+ "learning_rate": 1.926229508196721e-07,
147
+ "loss": 0.8153,
148
+ "step": 95
149
+ },
150
+ {
151
+ "epoch": 0.1023541453428864,
152
+ "grad_norm": 1.762168526649475,
153
+ "learning_rate": 2.028688524590164e-07,
154
+ "loss": 0.8064,
155
+ "step": 100
156
+ },
157
+ {
158
+ "epoch": 0.10747185261003071,
159
+ "grad_norm": 1.7200757265090942,
160
+ "learning_rate": 2.1311475409836064e-07,
161
+ "loss": 0.8063,
162
+ "step": 105
163
+ },
164
+ {
165
+ "epoch": 0.11258955987717502,
166
+ "grad_norm": 2.490513324737549,
167
+ "learning_rate": 2.233606557377049e-07,
168
+ "loss": 0.8192,
169
+ "step": 110
170
+ },
171
+ {
172
+ "epoch": 0.11770726714431934,
173
+ "grad_norm": 2.244020938873291,
174
+ "learning_rate": 2.336065573770492e-07,
175
+ "loss": 0.8153,
176
+ "step": 115
177
+ },
178
+ {
179
+ "epoch": 0.12282497441146366,
180
+ "grad_norm": 2.1315150260925293,
181
+ "learning_rate": 2.438524590163934e-07,
182
+ "loss": 0.807,
183
+ "step": 120
184
+ },
185
+ {
186
+ "epoch": 0.12794268167860798,
187
+ "grad_norm": 2.320936918258667,
188
+ "learning_rate": 2.540983606557377e-07,
189
+ "loss": 0.8163,
190
+ "step": 125
191
+ },
192
+ {
193
+ "epoch": 0.1330603889457523,
194
+ "grad_norm": 2.7143912315368652,
195
+ "learning_rate": 2.643442622950819e-07,
196
+ "loss": 0.8166,
197
+ "step": 130
198
+ },
199
+ {
200
+ "epoch": 0.13817809621289662,
201
+ "grad_norm": 1.649880290031433,
202
+ "learning_rate": 2.7459016393442624e-07,
203
+ "loss": 0.8113,
204
+ "step": 135
205
+ },
206
+ {
207
+ "epoch": 0.14329580348004095,
208
+ "grad_norm": 2.171790361404419,
209
+ "learning_rate": 2.848360655737705e-07,
210
+ "loss": 0.805,
211
+ "step": 140
212
+ },
213
+ {
214
+ "epoch": 0.14841351074718526,
215
+ "grad_norm": 2.093440294265747,
216
+ "learning_rate": 2.950819672131147e-07,
217
+ "loss": 0.8118,
218
+ "step": 145
219
+ },
220
+ {
221
+ "epoch": 0.1535312180143296,
222
+ "grad_norm": 1.9067059755325317,
223
+ "learning_rate": 3.05327868852459e-07,
224
+ "loss": 0.8047,
225
+ "step": 150
226
+ },
227
+ {
228
+ "epoch": 0.1586489252814739,
229
+ "grad_norm": 1.9988980293273926,
230
+ "learning_rate": 3.155737704918033e-07,
231
+ "loss": 0.8091,
232
+ "step": 155
233
+ },
234
+ {
235
+ "epoch": 0.16376663254861823,
236
+ "grad_norm": 1.696977972984314,
237
+ "learning_rate": 3.258196721311475e-07,
238
+ "loss": 0.8101,
239
+ "step": 160
240
+ },
241
+ {
242
+ "epoch": 0.16888433981576254,
243
+ "grad_norm": 2.098017454147339,
244
+ "learning_rate": 3.3606557377049177e-07,
245
+ "loss": 0.81,
246
+ "step": 165
247
+ },
248
+ {
249
+ "epoch": 0.17400204708290687,
250
+ "grad_norm": 2.0255584716796875,
251
+ "learning_rate": 3.463114754098361e-07,
252
+ "loss": 0.814,
253
+ "step": 170
254
+ },
255
+ {
256
+ "epoch": 0.17911975435005117,
257
+ "grad_norm": 1.8376339673995972,
258
+ "learning_rate": 3.565573770491803e-07,
259
+ "loss": 0.8053,
260
+ "step": 175
261
+ },
262
+ {
263
+ "epoch": 0.1842374616171955,
264
+ "grad_norm": 1.9230207204818726,
265
+ "learning_rate": 3.6680327868852456e-07,
266
+ "loss": 0.8022,
267
+ "step": 180
268
+ },
269
+ {
270
+ "epoch": 0.18935516888433981,
271
+ "grad_norm": 1.939705729484558,
272
+ "learning_rate": 3.770491803278688e-07,
273
+ "loss": 0.8075,
274
+ "step": 185
275
+ },
276
+ {
277
+ "epoch": 0.19447287615148415,
278
+ "grad_norm": 1.6276813745498657,
279
+ "learning_rate": 3.8729508196721314e-07,
280
+ "loss": 0.8097,
281
+ "step": 190
282
+ },
283
+ {
284
+ "epoch": 0.19959058341862845,
285
+ "grad_norm": 1.7544569969177246,
286
+ "learning_rate": 3.9754098360655735e-07,
287
+ "loss": 0.8046,
288
+ "step": 195
289
+ },
290
+ {
291
+ "epoch": 0.2047082906857728,
292
+ "grad_norm": 1.7406467199325562,
293
+ "learning_rate": 4.077868852459016e-07,
294
+ "loss": 0.8149,
295
+ "step": 200
296
+ },
297
+ {
298
+ "epoch": 0.2098259979529171,
299
+ "grad_norm": 1.7330560684204102,
300
+ "learning_rate": 4.180327868852459e-07,
301
+ "loss": 0.8077,
302
+ "step": 205
303
+ },
304
+ {
305
+ "epoch": 0.21494370522006143,
306
+ "grad_norm": 1.417546033859253,
307
+ "learning_rate": 4.2827868852459014e-07,
308
+ "loss": 0.807,
309
+ "step": 210
310
+ },
311
+ {
312
+ "epoch": 0.22006141248720573,
313
+ "grad_norm": 2.1064000129699707,
314
+ "learning_rate": 4.385245901639344e-07,
315
+ "loss": 0.8041,
316
+ "step": 215
317
+ },
318
+ {
319
+ "epoch": 0.22517911975435004,
320
+ "grad_norm": 1.637609601020813,
321
+ "learning_rate": 4.487704918032787e-07,
322
+ "loss": 0.7992,
323
+ "step": 220
324
+ },
325
+ {
326
+ "epoch": 0.23029682702149437,
327
+ "grad_norm": 1.659397840499878,
328
+ "learning_rate": 4.590163934426229e-07,
329
+ "loss": 0.802,
330
+ "step": 225
331
+ },
332
+ {
333
+ "epoch": 0.23541453428863868,
334
+ "grad_norm": 1.6912051439285278,
335
+ "learning_rate": 4.692622950819672e-07,
336
+ "loss": 0.8005,
337
+ "step": 230
338
+ },
339
+ {
340
+ "epoch": 0.240532241555783,
341
+ "grad_norm": 1.9433246850967407,
342
+ "learning_rate": 4.795081967213115e-07,
343
+ "loss": 0.8079,
344
+ "step": 235
345
+ },
346
+ {
347
+ "epoch": 0.24564994882292732,
348
+ "grad_norm": 1.9640270471572876,
349
+ "learning_rate": 4.897540983606557e-07,
350
+ "loss": 0.8127,
351
+ "step": 240
352
+ },
353
+ {
354
+ "epoch": 0.2507676560900716,
355
+ "grad_norm": 2.3167271614074707,
356
+ "learning_rate": 5e-07,
357
+ "loss": 0.8058,
358
+ "step": 245
359
+ },
360
+ {
361
+ "epoch": 0.25588536335721596,
362
+ "grad_norm": 1.6469106674194336,
363
+ "learning_rate": 5.102459016393442e-07,
364
+ "loss": 0.8011,
365
+ "step": 250
366
+ },
367
+ {
368
+ "epoch": 0.2610030706243603,
369
+ "grad_norm": 1.5691314935684204,
370
+ "learning_rate": 5.204918032786885e-07,
371
+ "loss": 0.7968,
372
+ "step": 255
373
+ },
374
+ {
375
+ "epoch": 0.2661207778915046,
376
+ "grad_norm": 1.663665533065796,
377
+ "learning_rate": 5.307377049180327e-07,
378
+ "loss": 0.8018,
379
+ "step": 260
380
+ },
381
+ {
382
+ "epoch": 0.2712384851586489,
383
+ "grad_norm": 1.99347984790802,
384
+ "learning_rate": 5.40983606557377e-07,
385
+ "loss": 0.8006,
386
+ "step": 265
387
+ },
388
+ {
389
+ "epoch": 0.27635619242579323,
390
+ "grad_norm": 1.4906947612762451,
391
+ "learning_rate": 5.512295081967213e-07,
392
+ "loss": 0.7977,
393
+ "step": 270
394
+ },
395
+ {
396
+ "epoch": 0.28147389969293757,
397
+ "grad_norm": 1.786527395248413,
398
+ "learning_rate": 5.614754098360656e-07,
399
+ "loss": 0.8041,
400
+ "step": 275
401
+ },
402
+ {
403
+ "epoch": 0.2865916069600819,
404
+ "grad_norm": 1.9175364971160889,
405
+ "learning_rate": 5.717213114754098e-07,
406
+ "loss": 0.8079,
407
+ "step": 280
408
+ },
409
+ {
410
+ "epoch": 0.2917093142272262,
411
+ "grad_norm": 1.678741216659546,
412
+ "learning_rate": 5.819672131147541e-07,
413
+ "loss": 0.7974,
414
+ "step": 285
415
+ },
416
+ {
417
+ "epoch": 0.2968270214943705,
418
+ "grad_norm": 2.0347344875335693,
419
+ "learning_rate": 5.922131147540983e-07,
420
+ "loss": 0.8011,
421
+ "step": 290
422
+ },
423
+ {
424
+ "epoch": 0.30194472876151485,
425
+ "grad_norm": 1.8914201259613037,
426
+ "learning_rate": 6.024590163934425e-07,
427
+ "loss": 0.8026,
428
+ "step": 295
429
+ },
430
+ {
431
+ "epoch": 0.3070624360286592,
432
+ "grad_norm": 1.6236293315887451,
433
+ "learning_rate": 6.127049180327869e-07,
434
+ "loss": 0.7981,
435
+ "step": 300
436
+ },
437
+ {
438
+ "epoch": 0.31218014329580346,
439
+ "grad_norm": 1.4731358289718628,
440
+ "learning_rate": 6.229508196721311e-07,
441
+ "loss": 0.7972,
442
+ "step": 305
443
+ },
444
+ {
445
+ "epoch": 0.3172978505629478,
446
+ "grad_norm": 1.7494508028030396,
447
+ "learning_rate": 6.331967213114754e-07,
448
+ "loss": 0.797,
449
+ "step": 310
450
+ },
451
+ {
452
+ "epoch": 0.3224155578300921,
453
+ "grad_norm": 1.696869134902954,
454
+ "learning_rate": 6.434426229508197e-07,
455
+ "loss": 0.7972,
456
+ "step": 315
457
+ },
458
+ {
459
+ "epoch": 0.32753326509723646,
460
+ "grad_norm": 1.5431866645812988,
461
+ "learning_rate": 6.536885245901639e-07,
462
+ "loss": 0.7919,
463
+ "step": 320
464
+ },
465
+ {
466
+ "epoch": 0.33265097236438074,
467
+ "grad_norm": 1.6396448612213135,
468
+ "learning_rate": 6.639344262295081e-07,
469
+ "loss": 0.7986,
470
+ "step": 325
471
+ },
472
+ {
473
+ "epoch": 0.33776867963152507,
474
+ "grad_norm": 1.7315205335617065,
475
+ "learning_rate": 6.741803278688525e-07,
476
+ "loss": 0.7966,
477
+ "step": 330
478
+ },
479
+ {
480
+ "epoch": 0.3428863868986694,
481
+ "grad_norm": 1.6142867803573608,
482
+ "learning_rate": 6.844262295081967e-07,
483
+ "loss": 0.7964,
484
+ "step": 335
485
+ },
486
+ {
487
+ "epoch": 0.34800409416581374,
488
+ "grad_norm": 1.332783818244934,
489
+ "learning_rate": 6.94672131147541e-07,
490
+ "loss": 0.7969,
491
+ "step": 340
492
+ },
493
+ {
494
+ "epoch": 0.353121801432958,
495
+ "grad_norm": 1.434688687324524,
496
+ "learning_rate": 7.049180327868852e-07,
497
+ "loss": 0.8015,
498
+ "step": 345
499
+ },
500
+ {
501
+ "epoch": 0.35823950870010235,
502
+ "grad_norm": 1.7243021726608276,
503
+ "learning_rate": 7.151639344262295e-07,
504
+ "loss": 0.791,
505
+ "step": 350
506
+ },
507
+ {
508
+ "epoch": 0.3633572159672467,
509
+ "grad_norm": 1.603244662284851,
510
+ "learning_rate": 7.254098360655737e-07,
511
+ "loss": 0.7926,
512
+ "step": 355
513
+ },
514
+ {
515
+ "epoch": 0.368474923234391,
516
+ "grad_norm": 1.645308256149292,
517
+ "learning_rate": 7.356557377049179e-07,
518
+ "loss": 0.7988,
519
+ "step": 360
520
+ },
521
+ {
522
+ "epoch": 0.3735926305015353,
523
+ "grad_norm": 1.3321951627731323,
524
+ "learning_rate": 7.459016393442623e-07,
525
+ "loss": 0.7923,
526
+ "step": 365
527
+ },
528
+ {
529
+ "epoch": 0.37871033776867963,
530
+ "grad_norm": 2.1083521842956543,
531
+ "learning_rate": 7.561475409836066e-07,
532
+ "loss": 0.7935,
533
+ "step": 370
534
+ },
535
+ {
536
+ "epoch": 0.38382804503582396,
537
+ "grad_norm": 1.3414019346237183,
538
+ "learning_rate": 7.663934426229508e-07,
539
+ "loss": 0.7894,
540
+ "step": 375
541
+ },
542
+ {
543
+ "epoch": 0.3889457523029683,
544
+ "grad_norm": 1.8279671669006348,
545
+ "learning_rate": 7.766393442622951e-07,
546
+ "loss": 0.7916,
547
+ "step": 380
548
+ },
549
+ {
550
+ "epoch": 0.3940634595701126,
551
+ "grad_norm": 1.6233114004135132,
552
+ "learning_rate": 7.868852459016393e-07,
553
+ "loss": 0.7886,
554
+ "step": 385
555
+ },
556
+ {
557
+ "epoch": 0.3991811668372569,
558
+ "grad_norm": 1.4336532354354858,
559
+ "learning_rate": 7.971311475409835e-07,
560
+ "loss": 0.7884,
561
+ "step": 390
562
+ },
563
+ {
564
+ "epoch": 0.40429887410440124,
565
+ "grad_norm": 1.597020149230957,
566
+ "learning_rate": 8.073770491803278e-07,
567
+ "loss": 0.7904,
568
+ "step": 395
569
+ },
570
+ {
571
+ "epoch": 0.4094165813715456,
572
+ "grad_norm": 1.3191157579421997,
573
+ "learning_rate": 8.176229508196721e-07,
574
+ "loss": 0.787,
575
+ "step": 400
576
+ },
577
+ {
578
+ "epoch": 0.41453428863868985,
579
+ "grad_norm": 1.6425617933273315,
580
+ "learning_rate": 8.278688524590164e-07,
581
+ "loss": 0.7887,
582
+ "step": 405
583
+ },
584
+ {
585
+ "epoch": 0.4196519959058342,
586
+ "grad_norm": 1.3924281597137451,
587
+ "learning_rate": 8.381147540983607e-07,
588
+ "loss": 0.7976,
589
+ "step": 410
590
+ },
591
+ {
592
+ "epoch": 0.4247697031729785,
593
+ "grad_norm": 1.2975757122039795,
594
+ "learning_rate": 8.483606557377049e-07,
595
+ "loss": 0.7895,
596
+ "step": 415
597
+ },
598
+ {
599
+ "epoch": 0.42988741044012285,
600
+ "grad_norm": 1.3045737743377686,
601
+ "learning_rate": 8.586065573770491e-07,
602
+ "loss": 0.7894,
603
+ "step": 420
604
+ },
605
+ {
606
+ "epoch": 0.43500511770726713,
607
+ "grad_norm": 1.9618183374404907,
608
+ "learning_rate": 8.688524590163933e-07,
609
+ "loss": 0.7865,
610
+ "step": 425
611
+ },
612
+ {
613
+ "epoch": 0.44012282497441146,
614
+ "grad_norm": 1.3976588249206543,
615
+ "learning_rate": 8.790983606557376e-07,
616
+ "loss": 0.7896,
617
+ "step": 430
618
+ },
619
+ {
620
+ "epoch": 0.4452405322415558,
621
+ "grad_norm": 1.1260899305343628,
622
+ "learning_rate": 8.89344262295082e-07,
623
+ "loss": 0.7861,
624
+ "step": 435
625
+ },
626
+ {
627
+ "epoch": 0.4503582395087001,
628
+ "grad_norm": 1.293816089630127,
629
+ "learning_rate": 8.995901639344262e-07,
630
+ "loss": 0.7826,
631
+ "step": 440
632
+ },
633
+ {
634
+ "epoch": 0.4554759467758444,
635
+ "grad_norm": 1.4861347675323486,
636
+ "learning_rate": 9.098360655737705e-07,
637
+ "loss": 0.7822,
638
+ "step": 445
639
+ },
640
+ {
641
+ "epoch": 0.46059365404298874,
642
+ "grad_norm": 1.378319501876831,
643
+ "learning_rate": 9.200819672131147e-07,
644
+ "loss": 0.778,
645
+ "step": 450
646
+ },
647
+ {
648
+ "epoch": 0.4657113613101331,
649
+ "grad_norm": 1.2947815656661987,
650
+ "learning_rate": 9.303278688524589e-07,
651
+ "loss": 0.7853,
652
+ "step": 455
653
+ },
654
+ {
655
+ "epoch": 0.47082906857727735,
656
+ "grad_norm": 0.9865773916244507,
657
+ "learning_rate": 9.405737704918032e-07,
658
+ "loss": 0.7797,
659
+ "step": 460
660
+ },
661
+ {
662
+ "epoch": 0.4759467758444217,
663
+ "grad_norm": 1.4883133172988892,
664
+ "learning_rate": 9.508196721311474e-07,
665
+ "loss": 0.7804,
666
+ "step": 465
667
+ },
668
+ {
669
+ "epoch": 0.481064483111566,
670
+ "grad_norm": 1.1394942998886108,
671
+ "learning_rate": 9.610655737704918e-07,
672
+ "loss": 0.7818,
673
+ "step": 470
674
+ },
675
+ {
676
+ "epoch": 0.48618219037871035,
677
+ "grad_norm": 1.104995846748352,
678
+ "learning_rate": 9.71311475409836e-07,
679
+ "loss": 0.7775,
680
+ "step": 475
681
+ },
682
+ {
683
+ "epoch": 0.49129989764585463,
684
+ "grad_norm": 1.258623719215393,
685
+ "learning_rate": 9.815573770491803e-07,
686
+ "loss": 0.7731,
687
+ "step": 480
688
+ },
689
+ {
690
+ "epoch": 0.49641760491299897,
691
+ "grad_norm": 1.4409220218658447,
692
+ "learning_rate": 9.918032786885245e-07,
693
+ "loss": 0.7811,
694
+ "step": 485
695
+ },
696
+ {
697
+ "epoch": 0.5015353121801432,
698
+ "grad_norm": 0.9952474236488342,
699
+ "learning_rate": 9.999994895105863e-07,
700
+ "loss": 0.7821,
701
+ "step": 490
702
+ },
703
+ {
704
+ "epoch": 0.5066530194472876,
705
+ "grad_norm": 1.2250083684921265,
706
+ "learning_rate": 9.99981622490561e-07,
707
+ "loss": 0.7822,
708
+ "step": 495
709
+ },
710
+ {
711
+ "epoch": 0.5117707267144319,
712
+ "grad_norm": 1.1539254188537598,
713
+ "learning_rate": 9.999382320422427e-07,
714
+ "loss": 0.776,
715
+ "step": 500
716
+ },
717
+ {
718
+ "epoch": 0.5117707267144319,
719
+ "eval_accuracy": 0.59523,
720
+ "eval_loss": 0.6936843991279602,
721
+ "eval_macro_f1": 0.5690192634397302,
722
+ "eval_precision": 0.6518208624514151,
723
+ "eval_recall": 0.6078906162164894,
724
+ "eval_runtime": 73.7478,
725
+ "eval_samples_per_second": 1355.972,
726
+ "eval_steps_per_second": 1.329,
727
+ "step": 500
728
+ },
729
+ {
730
+ "epoch": 0.5168884339815762,
731
+ "grad_norm": 1.2244267463684082,
732
+ "learning_rate": 9.998693203806588e-07,
733
+ "loss": 0.7771,
734
+ "step": 505
735
+ },
736
+ {
737
+ "epoch": 0.5220061412487206,
738
+ "grad_norm": 1.1900156736373901,
739
+ "learning_rate": 9.997748910236623e-07,
740
+ "loss": 0.7815,
741
+ "step": 510
742
+ },
743
+ {
744
+ "epoch": 0.5271238485158649,
745
+ "grad_norm": 1.2272601127624512,
746
+ "learning_rate": 9.996549487917522e-07,
747
+ "loss": 0.7829,
748
+ "step": 515
749
+ },
750
+ {
751
+ "epoch": 0.5322415557830092,
752
+ "grad_norm": 1.160675287246704,
753
+ "learning_rate": 9.995094998078276e-07,
754
+ "loss": 0.7785,
755
+ "step": 520
756
+ },
757
+ {
758
+ "epoch": 0.5373592630501536,
759
+ "grad_norm": 1.2759345769882202,
760
+ "learning_rate": 9.993385514968745e-07,
761
+ "loss": 0.7755,
762
+ "step": 525
763
+ },
764
+ {
765
+ "epoch": 0.5424769703172978,
766
+ "grad_norm": 1.0531632900238037,
767
+ "learning_rate": 9.99142112585588e-07,
768
+ "loss": 0.7781,
769
+ "step": 530
770
+ },
771
+ {
772
+ "epoch": 0.5475946775844421,
773
+ "grad_norm": 1.0040606260299683,
774
+ "learning_rate": 9.989201931019251e-07,
775
+ "loss": 0.7744,
776
+ "step": 535
777
+ },
778
+ {
779
+ "epoch": 0.5527123848515865,
780
+ "grad_norm": 1.2468197345733643,
781
+ "learning_rate": 9.98672804374595e-07,
782
+ "loss": 0.7712,
783
+ "step": 540
784
+ },
785
+ {
786
+ "epoch": 0.5578300921187308,
787
+ "grad_norm": 1.1564112901687622,
788
+ "learning_rate": 9.983999590324778e-07,
789
+ "loss": 0.7797,
790
+ "step": 545
791
+ },
792
+ {
793
+ "epoch": 0.5629477993858751,
794
+ "grad_norm": 0.8854450583457947,
795
+ "learning_rate": 9.981016710039832e-07,
796
+ "loss": 0.7723,
797
+ "step": 550
798
+ },
799
+ {
800
+ "epoch": 0.5680655066530195,
801
+ "grad_norm": 1.142919659614563,
802
+ "learning_rate": 9.977779555163369e-07,
803
+ "loss": 0.7739,
804
+ "step": 555
805
+ },
806
+ {
807
+ "epoch": 0.5731832139201638,
808
+ "grad_norm": 1.058153748512268,
809
+ "learning_rate": 9.974288290948042e-07,
810
+ "loss": 0.774,
811
+ "step": 560
812
+ },
813
+ {
814
+ "epoch": 0.5783009211873081,
815
+ "grad_norm": 1.1157392263412476,
816
+ "learning_rate": 9.970543095618468e-07,
817
+ "loss": 0.7742,
818
+ "step": 565
819
+ },
820
+ {
821
+ "epoch": 0.5834186284544524,
822
+ "grad_norm": 1.0850578546524048,
823
+ "learning_rate": 9.96654416036212e-07,
824
+ "loss": 0.7734,
825
+ "step": 570
826
+ },
827
+ {
828
+ "epoch": 0.5885363357215967,
829
+ "grad_norm": 0.9722121953964233,
830
+ "learning_rate": 9.96229168931958e-07,
831
+ "loss": 0.77,
832
+ "step": 575
833
+ },
834
+ {
835
+ "epoch": 0.593654042988741,
836
+ "grad_norm": 1.332795262336731,
837
+ "learning_rate": 9.957785899574102e-07,
838
+ "loss": 0.7725,
839
+ "step": 580
840
+ },
841
+ {
842
+ "epoch": 0.5987717502558854,
843
+ "grad_norm": 0.8639675378799438,
844
+ "learning_rate": 9.953027021140543e-07,
845
+ "loss": 0.7646,
846
+ "step": 585
847
+ },
848
+ {
849
+ "epoch": 0.6038894575230297,
850
+ "grad_norm": 0.9253244400024414,
851
+ "learning_rate": 9.948015296953623e-07,
852
+ "loss": 0.7743,
853
+ "step": 590
854
+ },
855
+ {
856
+ "epoch": 0.609007164790174,
857
+ "grad_norm": 0.8843643069267273,
858
+ "learning_rate": 9.942750982855503e-07,
859
+ "loss": 0.7717,
860
+ "step": 595
861
+ },
862
+ {
863
+ "epoch": 0.6141248720573184,
864
+ "grad_norm": 1.046048879623413,
865
+ "learning_rate": 9.937234347582753e-07,
866
+ "loss": 0.7721,
867
+ "step": 600
868
+ },
869
+ {
870
+ "epoch": 0.6192425793244627,
871
+ "grad_norm": 0.8906111717224121,
872
+ "learning_rate": 9.931465672752613e-07,
873
+ "loss": 0.7657,
874
+ "step": 605
875
+ },
876
+ {
877
+ "epoch": 0.6243602865916069,
878
+ "grad_norm": 0.9637787342071533,
879
+ "learning_rate": 9.925445252848621e-07,
880
+ "loss": 0.7666,
881
+ "step": 610
882
+ },
883
+ {
884
+ "epoch": 0.6294779938587513,
885
+ "grad_norm": 0.9004104733467102,
886
+ "learning_rate": 9.919173395205584e-07,
887
+ "loss": 0.7664,
888
+ "step": 615
889
+ },
890
+ {
891
+ "epoch": 0.6345957011258956,
892
+ "grad_norm": 1.4724570512771606,
893
+ "learning_rate": 9.912650419993893e-07,
894
+ "loss": 0.7679,
895
+ "step": 620
896
+ },
897
+ {
898
+ "epoch": 0.6397134083930399,
899
+ "grad_norm": 0.8644343614578247,
900
+ "learning_rate": 9.905876660203161e-07,
901
+ "loss": 0.7671,
902
+ "step": 625
903
+ },
904
+ {
905
+ "epoch": 0.6448311156601843,
906
+ "grad_norm": 0.8368955254554749,
907
+ "learning_rate": 9.898852461625245e-07,
908
+ "loss": 0.7717,
909
+ "step": 630
910
+ },
911
+ {
912
+ "epoch": 0.6499488229273286,
913
+ "grad_norm": 0.9413282871246338,
914
+ "learning_rate": 9.891578182836583e-07,
915
+ "loss": 0.7693,
916
+ "step": 635
917
+ },
918
+ {
919
+ "epoch": 0.6550665301944729,
920
+ "grad_norm": 0.9777762293815613,
921
+ "learning_rate": 9.884054195179886e-07,
922
+ "loss": 0.7656,
923
+ "step": 640
924
+ },
925
+ {
926
+ "epoch": 0.6601842374616171,
927
+ "grad_norm": 0.8983454704284668,
928
+ "learning_rate": 9.876280882745193e-07,
929
+ "loss": 0.7605,
930
+ "step": 645
931
+ },
932
+ {
933
+ "epoch": 0.6653019447287615,
934
+ "grad_norm": 0.8708799481391907,
935
+ "learning_rate": 9.868258642350254e-07,
936
+ "loss": 0.7673,
937
+ "step": 650
938
+ },
939
+ {
940
+ "epoch": 0.6704196519959058,
941
+ "grad_norm": 0.8354130387306213,
942
+ "learning_rate": 9.859987883520275e-07,
943
+ "loss": 0.767,
944
+ "step": 655
945
+ },
946
+ {
947
+ "epoch": 0.6755373592630501,
948
+ "grad_norm": 0.868485152721405,
949
+ "learning_rate": 9.851469028467015e-07,
950
+ "loss": 0.7647,
951
+ "step": 660
952
+ },
953
+ {
954
+ "epoch": 0.6806550665301945,
955
+ "grad_norm": 0.9445936679840088,
956
+ "learning_rate": 9.84270251206723e-07,
957
+ "loss": 0.7605,
958
+ "step": 665
959
+ },
960
+ {
961
+ "epoch": 0.6857727737973388,
962
+ "grad_norm": 0.7952156662940979,
963
+ "learning_rate": 9.833688781840475e-07,
964
+ "loss": 0.7664,
965
+ "step": 670
966
+ },
967
+ {
968
+ "epoch": 0.6908904810644831,
969
+ "grad_norm": 1.1992422342300415,
970
+ "learning_rate": 9.824428297926254e-07,
971
+ "loss": 0.7617,
972
+ "step": 675
973
+ },
974
+ {
975
+ "epoch": 0.6960081883316275,
976
+ "grad_norm": 0.8914986252784729,
977
+ "learning_rate": 9.81492153306054e-07,
978
+ "loss": 0.764,
979
+ "step": 680
980
+ },
981
+ {
982
+ "epoch": 0.7011258955987717,
983
+ "grad_norm": 0.7945632338523865,
984
+ "learning_rate": 9.80516897255163e-07,
985
+ "loss": 0.7617,
986
+ "step": 685
987
+ },
988
+ {
989
+ "epoch": 0.706243602865916,
990
+ "grad_norm": 0.7822641134262085,
991
+ "learning_rate": 9.795171114255384e-07,
992
+ "loss": 0.7613,
993
+ "step": 690
994
+ },
995
+ {
996
+ "epoch": 0.7113613101330604,
997
+ "grad_norm": 0.7989721298217773,
998
+ "learning_rate": 9.784928468549793e-07,
999
+ "loss": 0.7615,
1000
+ "step": 695
1001
+ },
1002
+ {
1003
+ "epoch": 0.7164790174002047,
1004
+ "grad_norm": 0.7325178980827332,
1005
+ "learning_rate": 9.77444155830895e-07,
1006
+ "loss": 0.7572,
1007
+ "step": 700
1008
+ },
1009
+ {
1010
+ "epoch": 0.721596724667349,
1011
+ "grad_norm": 0.8934036493301392,
1012
+ "learning_rate": 9.763710918876329e-07,
1013
+ "loss": 0.7589,
1014
+ "step": 705
1015
+ },
1016
+ {
1017
+ "epoch": 0.7267144319344934,
1018
+ "grad_norm": 0.7769590616226196,
1019
+ "learning_rate": 9.752737098037477e-07,
1020
+ "loss": 0.7573,
1021
+ "step": 710
1022
+ },
1023
+ {
1024
+ "epoch": 0.7318321392016377,
1025
+ "grad_norm": 1.0458475351333618,
1026
+ "learning_rate": 9.741520655992047e-07,
1027
+ "loss": 0.759,
1028
+ "step": 715
1029
+ },
1030
+ {
1031
+ "epoch": 0.736949846468782,
1032
+ "grad_norm": 0.649872899055481,
1033
+ "learning_rate": 9.730062165325185e-07,
1034
+ "loss": 0.7607,
1035
+ "step": 720
1036
+ },
1037
+ {
1038
+ "epoch": 0.7420675537359263,
1039
+ "grad_norm": 0.7517932057380676,
1040
+ "learning_rate": 9.718362210978329e-07,
1041
+ "loss": 0.7567,
1042
+ "step": 725
1043
+ },
1044
+ {
1045
+ "epoch": 0.7471852610030706,
1046
+ "grad_norm": 0.9947759509086609,
1047
+ "learning_rate": 9.706421390219315e-07,
1048
+ "loss": 0.7593,
1049
+ "step": 730
1050
+ },
1051
+ {
1052
+ "epoch": 0.7523029682702149,
1053
+ "grad_norm": 0.719109833240509,
1054
+ "learning_rate": 9.694240312611917e-07,
1055
+ "loss": 0.7615,
1056
+ "step": 735
1057
+ },
1058
+ {
1059
+ "epoch": 0.7574206755373593,
1060
+ "grad_norm": 1.0175235271453857,
1061
+ "learning_rate": 9.681819599984712e-07,
1062
+ "loss": 0.7555,
1063
+ "step": 740
1064
+ },
1065
+ {
1066
+ "epoch": 0.7625383828045036,
1067
+ "grad_norm": 0.8200032711029053,
1068
+ "learning_rate": 9.66915988639934e-07,
1069
+ "loss": 0.7565,
1070
+ "step": 745
1071
+ },
1072
+ {
1073
+ "epoch": 0.7676560900716479,
1074
+ "grad_norm": 0.926680326461792,
1075
+ "learning_rate": 9.656261818118139e-07,
1076
+ "loss": 0.7628,
1077
+ "step": 750
1078
+ },
1079
+ {
1080
+ "epoch": 0.7727737973387923,
1081
+ "grad_norm": 0.6904947757720947,
1082
+ "learning_rate": 9.64312605357115e-07,
1083
+ "loss": 0.7584,
1084
+ "step": 755
1085
+ },
1086
+ {
1087
+ "epoch": 0.7778915046059366,
1088
+ "grad_norm": 0.7391018867492676,
1089
+ "learning_rate": 9.62975326332251e-07,
1090
+ "loss": 0.7582,
1091
+ "step": 760
1092
+ },
1093
+ {
1094
+ "epoch": 0.7830092118730808,
1095
+ "grad_norm": 0.7193120121955872,
1096
+ "learning_rate": 9.616144130036214e-07,
1097
+ "loss": 0.7557,
1098
+ "step": 765
1099
+ },
1100
+ {
1101
+ "epoch": 0.7881269191402251,
1102
+ "grad_norm": 0.8275336623191833,
1103
+ "learning_rate": 9.602299348441277e-07,
1104
+ "loss": 0.7575,
1105
+ "step": 770
1106
+ },
1107
+ {
1108
+ "epoch": 0.7932446264073695,
1109
+ "grad_norm": 0.9943181276321411,
1110
+ "learning_rate": 9.58821962529625e-07,
1111
+ "loss": 0.7568,
1112
+ "step": 775
1113
+ },
1114
+ {
1115
+ "epoch": 0.7983623336745138,
1116
+ "grad_norm": 0.7646188139915466,
1117
+ "learning_rate": 9.573905679353166e-07,
1118
+ "loss": 0.752,
1119
+ "step": 780
1120
+ },
1121
+ {
1122
+ "epoch": 0.8034800409416581,
1123
+ "grad_norm": 0.7356329560279846,
1124
+ "learning_rate": 9.55935824132082e-07,
1125
+ "loss": 0.7552,
1126
+ "step": 785
1127
+ },
1128
+ {
1129
+ "epoch": 0.8085977482088025,
1130
+ "grad_norm": 0.795838475227356,
1131
+ "learning_rate": 9.544578053827495e-07,
1132
+ "loss": 0.7543,
1133
+ "step": 790
1134
+ },
1135
+ {
1136
+ "epoch": 0.8137154554759468,
1137
+ "grad_norm": 0.9953216314315796,
1138
+ "learning_rate": 9.529565871383034e-07,
1139
+ "loss": 0.7558,
1140
+ "step": 795
1141
+ },
1142
+ {
1143
+ "epoch": 0.8188331627430911,
1144
+ "grad_norm": 0.797937273979187,
1145
+ "learning_rate": 9.514322460340329e-07,
1146
+ "loss": 0.7542,
1147
+ "step": 800
1148
+ },
1149
+ {
1150
+ "epoch": 0.8239508700102354,
1151
+ "grad_norm": 0.7371375560760498,
1152
+ "learning_rate": 9.498848598856198e-07,
1153
+ "loss": 0.7532,
1154
+ "step": 805
1155
+ },
1156
+ {
1157
+ "epoch": 0.8290685772773797,
1158
+ "grad_norm": 0.8336758613586426,
1159
+ "learning_rate": 9.48314507685166e-07,
1160
+ "loss": 0.756,
1161
+ "step": 810
1162
+ },
1163
+ {
1164
+ "epoch": 0.834186284544524,
1165
+ "grad_norm": 0.7204869389533997,
1166
+ "learning_rate": 9.467212695971619e-07,
1167
+ "loss": 0.7564,
1168
+ "step": 815
1169
+ },
1170
+ {
1171
+ "epoch": 0.8393039918116684,
1172
+ "grad_norm": 0.6758232712745667,
1173
+ "learning_rate": 9.451052269543929e-07,
1174
+ "loss": 0.7548,
1175
+ "step": 820
1176
+ },
1177
+ {
1178
+ "epoch": 0.8444216990788127,
1179
+ "grad_norm": 0.7348074913024902,
1180
+ "learning_rate": 9.434664622537883e-07,
1181
+ "loss": 0.7535,
1182
+ "step": 825
1183
+ },
1184
+ {
1185
+ "epoch": 0.849539406345957,
1186
+ "grad_norm": 0.747559130191803,
1187
+ "learning_rate": 9.418050591522093e-07,
1188
+ "loss": 0.752,
1189
+ "step": 830
1190
+ },
1191
+ {
1192
+ "epoch": 0.8546571136131014,
1193
+ "grad_norm": 0.7392817735671997,
1194
+ "learning_rate": 9.401211024621792e-07,
1195
+ "loss": 0.7492,
1196
+ "step": 835
1197
+ },
1198
+ {
1199
+ "epoch": 0.8597748208802457,
1200
+ "grad_norm": 0.6318978071212769,
1201
+ "learning_rate": 9.384146781475533e-07,
1202
+ "loss": 0.7577,
1203
+ "step": 840
1204
+ },
1205
+ {
1206
+ "epoch": 0.8648925281473899,
1207
+ "grad_norm": 0.5832816362380981,
1208
+ "learning_rate": 9.366858733191307e-07,
1209
+ "loss": 0.7506,
1210
+ "step": 845
1211
+ },
1212
+ {
1213
+ "epoch": 0.8700102354145343,
1214
+ "grad_norm": 0.6932022571563721,
1215
+ "learning_rate": 9.349347762302071e-07,
1216
+ "loss": 0.7523,
1217
+ "step": 850
1218
+ },
1219
+ {
1220
+ "epoch": 0.8751279426816786,
1221
+ "grad_norm": 0.7047157287597656,
1222
+ "learning_rate": 9.331614762720703e-07,
1223
+ "loss": 0.7487,
1224
+ "step": 855
1225
+ },
1226
+ {
1227
+ "epoch": 0.8802456499488229,
1228
+ "grad_norm": 0.6591235995292664,
1229
+ "learning_rate": 9.313660639694358e-07,
1230
+ "loss": 0.7538,
1231
+ "step": 860
1232
+ },
1233
+ {
1234
+ "epoch": 0.8853633572159673,
1235
+ "grad_norm": 0.66665118932724,
1236
+ "learning_rate": 9.295486309758269e-07,
1237
+ "loss": 0.7518,
1238
+ "step": 865
1239
+ },
1240
+ {
1241
+ "epoch": 0.8904810644831116,
1242
+ "grad_norm": 0.6165961027145386,
1243
+ "learning_rate": 9.277092700688951e-07,
1244
+ "loss": 0.7495,
1245
+ "step": 870
1246
+ },
1247
+ {
1248
+ "epoch": 0.8955987717502559,
1249
+ "grad_norm": 0.7449588179588318,
1250
+ "learning_rate": 9.258480751456838e-07,
1251
+ "loss": 0.7515,
1252
+ "step": 875
1253
+ },
1254
+ {
1255
+ "epoch": 0.9007164790174002,
1256
+ "grad_norm": 0.7553215622901917,
1257
+ "learning_rate": 9.239651412178357e-07,
1258
+ "loss": 0.7534,
1259
+ "step": 880
1260
+ },
1261
+ {
1262
+ "epoch": 0.9058341862845445,
1263
+ "grad_norm": 0.747010350227356,
1264
+ "learning_rate": 9.220605644067419e-07,
1265
+ "loss": 0.7548,
1266
+ "step": 885
1267
+ },
1268
+ {
1269
+ "epoch": 0.9109518935516888,
1270
+ "grad_norm": 0.7272236347198486,
1271
+ "learning_rate": 9.20134441938635e-07,
1272
+ "loss": 0.7531,
1273
+ "step": 890
1274
+ },
1275
+ {
1276
+ "epoch": 0.9160696008188332,
1277
+ "grad_norm": 0.8726323246955872,
1278
+ "learning_rate": 9.181868721396266e-07,
1279
+ "loss": 0.7479,
1280
+ "step": 895
1281
+ },
1282
+ {
1283
+ "epoch": 0.9211873080859775,
1284
+ "grad_norm": 0.7914009094238281,
1285
+ "learning_rate": 9.16217954430687e-07,
1286
+ "loss": 0.7522,
1287
+ "step": 900
1288
+ },
1289
+ {
1290
+ "epoch": 0.9263050153531218,
1291
+ "grad_norm": 0.6367310285568237,
1292
+ "learning_rate": 9.142277893225708e-07,
1293
+ "loss": 0.7497,
1294
+ "step": 905
1295
+ },
1296
+ {
1297
+ "epoch": 0.9314227226202662,
1298
+ "grad_norm": 0.8285405039787292,
1299
+ "learning_rate": 9.122164784106842e-07,
1300
+ "loss": 0.753,
1301
+ "step": 910
1302
+ },
1303
+ {
1304
+ "epoch": 0.9365404298874105,
1305
+ "grad_norm": 0.7742036581039429,
1306
+ "learning_rate": 9.101841243699015e-07,
1307
+ "loss": 0.7534,
1308
+ "step": 915
1309
+ },
1310
+ {
1311
+ "epoch": 0.9416581371545547,
1312
+ "grad_norm": 0.7512480020523071,
1313
+ "learning_rate": 9.081308309493209e-07,
1314
+ "loss": 0.747,
1315
+ "step": 920
1316
+ },
1317
+ {
1318
+ "epoch": 0.946775844421699,
1319
+ "grad_norm": 0.5556691288948059,
1320
+ "learning_rate": 9.060567029669699e-07,
1321
+ "loss": 0.7465,
1322
+ "step": 925
1323
+ },
1324
+ {
1325
+ "epoch": 0.9518935516888434,
1326
+ "grad_norm": 1.0232101678848267,
1327
+ "learning_rate": 9.039618463044536e-07,
1328
+ "loss": 0.7485,
1329
+ "step": 930
1330
+ },
1331
+ {
1332
+ "epoch": 0.9570112589559877,
1333
+ "grad_norm": 0.8321600556373596,
1334
+ "learning_rate": 9.018463679015505e-07,
1335
+ "loss": 0.7488,
1336
+ "step": 935
1337
+ },
1338
+ {
1339
+ "epoch": 0.962128966223132,
1340
+ "grad_norm": 0.7009038329124451,
1341
+ "learning_rate": 8.997103757507521e-07,
1342
+ "loss": 0.7483,
1343
+ "step": 940
1344
+ },
1345
+ {
1346
+ "epoch": 0.9672466734902764,
1347
+ "grad_norm": 0.6939564347267151,
1348
+ "learning_rate": 8.975539788917514e-07,
1349
+ "loss": 0.7485,
1350
+ "step": 945
1351
+ },
1352
+ {
1353
+ "epoch": 0.9723643807574207,
1354
+ "grad_norm": 0.7738851308822632,
1355
+ "learning_rate": 8.953772874058757e-07,
1356
+ "loss": 0.7479,
1357
+ "step": 950
1358
+ },
1359
+ {
1360
+ "epoch": 0.977482088024565,
1361
+ "grad_norm": 0.5913597941398621,
1362
+ "learning_rate": 8.931804124104672e-07,
1363
+ "loss": 0.7473,
1364
+ "step": 955
1365
+ },
1366
+ {
1367
+ "epoch": 0.9825997952917093,
1368
+ "grad_norm": 0.8486027717590332,
1369
+ "learning_rate": 8.909634660532106e-07,
1370
+ "loss": 0.7479,
1371
+ "step": 960
1372
+ },
1373
+ {
1374
+ "epoch": 0.9877175025588536,
1375
+ "grad_norm": 0.6463382840156555,
1376
+ "learning_rate": 8.887265615064083e-07,
1377
+ "loss": 0.7486,
1378
+ "step": 965
1379
+ },
1380
+ {
1381
+ "epoch": 0.9928352098259979,
1382
+ "grad_norm": 0.6264991164207458,
1383
+ "learning_rate": 8.864698129612031e-07,
1384
+ "loss": 0.7467,
1385
+ "step": 970
1386
+ },
1387
+ {
1388
+ "epoch": 0.9979529170931423,
1389
+ "grad_norm": 0.7566510438919067,
1390
+ "learning_rate": 8.841933356217488e-07,
1391
+ "loss": 0.7463,
1392
+ "step": 975
1393
+ },
1394
+ {
1395
+ "epoch": 1.0030706243602865,
1396
+ "grad_norm": 0.7290503978729248,
1397
+ "learning_rate": 8.818972456993288e-07,
1398
+ "loss": 0.7504,
1399
+ "step": 980
1400
+ },
1401
+ {
1402
+ "epoch": 1.008188331627431,
1403
+ "grad_norm": 0.8277891874313354,
1404
+ "learning_rate": 8.795816604064241e-07,
1405
+ "loss": 0.7472,
1406
+ "step": 985
1407
+ },
1408
+ {
1409
+ "epoch": 1.0133060388945752,
1410
+ "grad_norm": 0.6427952647209167,
1411
+ "learning_rate": 8.772466979507302e-07,
1412
+ "loss": 0.7487,
1413
+ "step": 990
1414
+ },
1415
+ {
1416
+ "epoch": 1.0184237461617196,
1417
+ "grad_norm": 0.6775041818618774,
1418
+ "learning_rate": 8.748924775291216e-07,
1419
+ "loss": 0.745,
1420
+ "step": 995
1421
+ },
1422
+ {
1423
+ "epoch": 1.0235414534288638,
1424
+ "grad_norm": 0.6815404891967773,
1425
+ "learning_rate": 8.725191193215675e-07,
1426
+ "loss": 0.7485,
1427
+ "step": 1000
1428
+ },
1429
+ {
1430
+ "epoch": 1.0235414534288638,
1431
+ "eval_accuracy": 0.67557,
1432
+ "eval_loss": 0.6936712265014648,
1433
+ "eval_macro_f1": 0.6724504812400831,
1434
+ "eval_precision": 0.6760463081581009,
1435
+ "eval_recall": 0.6725003053739838,
1436
+ "eval_runtime": 73.7408,
1437
+ "eval_samples_per_second": 1356.102,
1438
+ "eval_steps_per_second": 1.329,
1439
+ "step": 1000
1440
+ },
1441
+ {
1442
+ "epoch": 1.0286591606960083,
1443
+ "grad_norm": 0.8586804866790771,
1444
+ "learning_rate": 8.701267444849974e-07,
1445
+ "loss": 0.7457,
1446
+ "step": 1005
1447
+ },
1448
+ {
1449
+ "epoch": 1.0337768679631525,
1450
+ "grad_norm": 0.5989358425140381,
1451
+ "learning_rate": 8.677154751471152e-07,
1452
+ "loss": 0.7443,
1453
+ "step": 1010
1454
+ },
1455
+ {
1456
+ "epoch": 1.0388945752302967,
1457
+ "grad_norm": 0.6888963580131531,
1458
+ "learning_rate": 8.65285434400165e-07,
1459
+ "loss": 0.7458,
1460
+ "step": 1015
1461
+ },
1462
+ {
1463
+ "epoch": 1.0440122824974412,
1464
+ "grad_norm": 0.6407850384712219,
1465
+ "learning_rate": 8.628367462946482e-07,
1466
+ "loss": 0.7493,
1467
+ "step": 1020
1468
+ },
1469
+ {
1470
+ "epoch": 1.0491299897645854,
1471
+ "grad_norm": 0.6202091574668884,
1472
+ "learning_rate": 8.603695358329896e-07,
1473
+ "loss": 0.7471,
1474
+ "step": 1025
1475
+ },
1476
+ {
1477
+ "epoch": 1.0542476970317298,
1478
+ "grad_norm": 0.7456187605857849,
1479
+ "learning_rate": 8.57883928963157e-07,
1480
+ "loss": 0.7431,
1481
+ "step": 1030
1482
+ },
1483
+ {
1484
+ "epoch": 1.059365404298874,
1485
+ "grad_norm": 0.6171067357063293,
1486
+ "learning_rate": 8.553800525722317e-07,
1487
+ "loss": 0.7435,
1488
+ "step": 1035
1489
+ },
1490
+ {
1491
+ "epoch": 1.0644831115660185,
1492
+ "grad_norm": 0.8527712821960449,
1493
+ "learning_rate": 8.528580344799305e-07,
1494
+ "loss": 0.7453,
1495
+ "step": 1040
1496
+ },
1497
+ {
1498
+ "epoch": 1.0696008188331627,
1499
+ "grad_norm": 0.6724162697792053,
1500
+ "learning_rate": 8.503180034320816e-07,
1501
+ "loss": 0.7467,
1502
+ "step": 1045
1503
+ },
1504
+ {
1505
+ "epoch": 1.0747185261003072,
1506
+ "grad_norm": 0.581979513168335,
1507
+ "learning_rate": 8.477600890940513e-07,
1508
+ "loss": 0.7508,
1509
+ "step": 1050
1510
+ },
1511
+ {
1512
+ "epoch": 1.0798362333674514,
1513
+ "grad_norm": 0.6551439166069031,
1514
+ "learning_rate": 8.451844220441253e-07,
1515
+ "loss": 0.7469,
1516
+ "step": 1055
1517
+ },
1518
+ {
1519
+ "epoch": 1.0849539406345956,
1520
+ "grad_norm": 0.6437426209449768,
1521
+ "learning_rate": 8.42591133766843e-07,
1522
+ "loss": 0.7468,
1523
+ "step": 1060
1524
+ },
1525
+ {
1526
+ "epoch": 1.09007164790174,
1527
+ "grad_norm": 0.5788704752922058,
1528
+ "learning_rate": 8.39980356646285e-07,
1529
+ "loss": 0.7424,
1530
+ "step": 1065
1531
+ },
1532
+ {
1533
+ "epoch": 1.0951893551688843,
1534
+ "grad_norm": 0.5575606226921082,
1535
+ "learning_rate": 8.373522239593149e-07,
1536
+ "loss": 0.7396,
1537
+ "step": 1070
1538
+ },
1539
+ {
1540
+ "epoch": 1.1003070624360287,
1541
+ "grad_norm": 0.737180769443512,
1542
+ "learning_rate": 8.347068698687765e-07,
1543
+ "loss": 0.744,
1544
+ "step": 1075
1545
+ },
1546
+ {
1547
+ "epoch": 1.105424769703173,
1548
+ "grad_norm": 0.592766284942627,
1549
+ "learning_rate": 8.320444294166439e-07,
1550
+ "loss": 0.7469,
1551
+ "step": 1080
1552
+ },
1553
+ {
1554
+ "epoch": 1.1105424769703174,
1555
+ "grad_norm": 0.63823401927948,
1556
+ "learning_rate": 8.293650385171287e-07,
1557
+ "loss": 0.7447,
1558
+ "step": 1085
1559
+ },
1560
+ {
1561
+ "epoch": 1.1156601842374616,
1562
+ "grad_norm": 0.6114454865455627,
1563
+ "learning_rate": 8.266688339497412e-07,
1564
+ "loss": 0.7475,
1565
+ "step": 1090
1566
+ },
1567
+ {
1568
+ "epoch": 1.120777891504606,
1569
+ "grad_norm": 0.53263258934021,
1570
+ "learning_rate": 8.239559533523082e-07,
1571
+ "loss": 0.7455,
1572
+ "step": 1095
1573
+ },
1574
+ {
1575
+ "epoch": 1.1258955987717503,
1576
+ "grad_norm": 0.7016158699989319,
1577
+ "learning_rate": 8.212265352139466e-07,
1578
+ "loss": 0.742,
1579
+ "step": 1100
1580
+ },
1581
+ {
1582
+ "epoch": 1.1310133060388945,
1583
+ "grad_norm": 0.6125472784042358,
1584
+ "learning_rate": 8.184807188679939e-07,
1585
+ "loss": 0.7383,
1586
+ "step": 1105
1587
+ },
1588
+ {
1589
+ "epoch": 1.136131013306039,
1590
+ "grad_norm": 0.6008788347244263,
1591
+ "learning_rate": 8.157186444848952e-07,
1592
+ "loss": 0.7435,
1593
+ "step": 1110
1594
+ },
1595
+ {
1596
+ "epoch": 1.1412487205731832,
1597
+ "grad_norm": 0.6357280015945435,
1598
+ "learning_rate": 8.129404530650479e-07,
1599
+ "loss": 0.7443,
1600
+ "step": 1115
1601
+ },
1602
+ {
1603
+ "epoch": 1.1463664278403276,
1604
+ "grad_norm": 0.6422165036201477,
1605
+ "learning_rate": 8.101462864316038e-07,
1606
+ "loss": 0.7449,
1607
+ "step": 1120
1608
+ },
1609
+ {
1610
+ "epoch": 1.1514841351074718,
1611
+ "grad_norm": 0.6852079629898071,
1612
+ "learning_rate": 8.07336287223229e-07,
1613
+ "loss": 0.7428,
1614
+ "step": 1125
1615
+ },
1616
+ {
1617
+ "epoch": 1.156601842374616,
1618
+ "grad_norm": 0.5539452433586121,
1619
+ "learning_rate": 8.045105988868224e-07,
1620
+ "loss": 0.7455,
1621
+ "step": 1130
1622
+ },
1623
+ {
1624
+ "epoch": 1.1617195496417605,
1625
+ "grad_norm": 0.5939313173294067,
1626
+ "learning_rate": 8.016693656701931e-07,
1627
+ "loss": 0.7376,
1628
+ "step": 1135
1629
+ },
1630
+ {
1631
+ "epoch": 1.1668372569089047,
1632
+ "grad_norm": 0.7522106766700745,
1633
+ "learning_rate": 7.98812732614697e-07,
1634
+ "loss": 0.7464,
1635
+ "step": 1140
1636
+ },
1637
+ {
1638
+ "epoch": 1.1719549641760492,
1639
+ "grad_norm": 0.6572809815406799,
1640
+ "learning_rate": 7.959408455478313e-07,
1641
+ "loss": 0.7448,
1642
+ "step": 1145
1643
+ },
1644
+ {
1645
+ "epoch": 1.1770726714431934,
1646
+ "grad_norm": 0.5842403173446655,
1647
+ "learning_rate": 7.93053851075792e-07,
1648
+ "loss": 0.7396,
1649
+ "step": 1150
1650
+ },
1651
+ {
1652
+ "epoch": 1.1821903787103378,
1653
+ "grad_norm": 0.5845000147819519,
1654
+ "learning_rate": 7.901518965759888e-07,
1655
+ "loss": 0.7438,
1656
+ "step": 1155
1657
+ },
1658
+ {
1659
+ "epoch": 1.187308085977482,
1660
+ "grad_norm": 0.5873178839683533,
1661
+ "learning_rate": 7.872351301895217e-07,
1662
+ "loss": 0.7421,
1663
+ "step": 1160
1664
+ },
1665
+ {
1666
+ "epoch": 1.1924257932446265,
1667
+ "grad_norm": 0.6385728120803833,
1668
+ "learning_rate": 7.843037008136189e-07,
1669
+ "loss": 0.7431,
1670
+ "step": 1165
1671
+ },
1672
+ {
1673
+ "epoch": 1.1975435005117707,
1674
+ "grad_norm": 0.5818535685539246,
1675
+ "learning_rate": 7.813577580940356e-07,
1676
+ "loss": 0.7416,
1677
+ "step": 1170
1678
+ },
1679
+ {
1680
+ "epoch": 1.202661207778915,
1681
+ "grad_norm": 0.5611526370048523,
1682
+ "learning_rate": 7.783974524174149e-07,
1683
+ "loss": 0.743,
1684
+ "step": 1175
1685
+ },
1686
+ {
1687
+ "epoch": 1.2077789150460594,
1688
+ "grad_norm": 0.6002296805381775,
1689
+ "learning_rate": 7.754229349036102e-07,
1690
+ "loss": 0.7407,
1691
+ "step": 1180
1692
+ },
1693
+ {
1694
+ "epoch": 1.2128966223132036,
1695
+ "grad_norm": 0.6006008982658386,
1696
+ "learning_rate": 7.724343573979718e-07,
1697
+ "loss": 0.7437,
1698
+ "step": 1185
1699
+ },
1700
+ {
1701
+ "epoch": 1.218014329580348,
1702
+ "grad_norm": 0.6336845755577087,
1703
+ "learning_rate": 7.694318724635945e-07,
1704
+ "loss": 0.7405,
1705
+ "step": 1190
1706
+ },
1707
+ {
1708
+ "epoch": 1.2231320368474923,
1709
+ "grad_norm": 0.6916839480400085,
1710
+ "learning_rate": 7.664156333735293e-07,
1711
+ "loss": 0.7468,
1712
+ "step": 1195
1713
+ },
1714
+ {
1715
+ "epoch": 1.2282497441146367,
1716
+ "grad_norm": 0.5944891571998596,
1717
+ "learning_rate": 7.633857941029602e-07,
1718
+ "loss": 0.7485,
1719
+ "step": 1200
1720
+ },
1721
+ {
1722
+ "epoch": 1.233367451381781,
1723
+ "grad_norm": 0.5755409598350525,
1724
+ "learning_rate": 7.603425093213429e-07,
1725
+ "loss": 0.7418,
1726
+ "step": 1205
1727
+ },
1728
+ {
1729
+ "epoch": 1.2384851586489254,
1730
+ "grad_norm": 0.6128578186035156,
1731
+ "learning_rate": 7.572859343845092e-07,
1732
+ "loss": 0.7396,
1733
+ "step": 1210
1734
+ },
1735
+ {
1736
+ "epoch": 1.2436028659160696,
1737
+ "grad_norm": 0.6123960614204407,
1738
+ "learning_rate": 7.542162253267363e-07,
1739
+ "loss": 0.7363,
1740
+ "step": 1215
1741
+ },
1742
+ {
1743
+ "epoch": 1.2487205731832138,
1744
+ "grad_norm": 0.6969608664512634,
1745
+ "learning_rate": 7.511335388527822e-07,
1746
+ "loss": 0.7406,
1747
+ "step": 1220
1748
+ },
1749
+ {
1750
+ "epoch": 1.2538382804503583,
1751
+ "grad_norm": 0.6491796970367432,
1752
+ "learning_rate": 7.480380323298851e-07,
1753
+ "loss": 0.7429,
1754
+ "step": 1225
1755
+ },
1756
+ {
1757
+ "epoch": 1.2589559877175025,
1758
+ "grad_norm": 0.5883914828300476,
1759
+ "learning_rate": 7.449298637797309e-07,
1760
+ "loss": 0.7375,
1761
+ "step": 1230
1762
+ },
1763
+ {
1764
+ "epoch": 1.264073694984647,
1765
+ "grad_norm": 0.6160842776298523,
1766
+ "learning_rate": 7.418091918703854e-07,
1767
+ "loss": 0.7393,
1768
+ "step": 1235
1769
+ },
1770
+ {
1771
+ "epoch": 1.2691914022517912,
1772
+ "grad_norm": 0.5568389892578125,
1773
+ "learning_rate": 7.386761759081954e-07,
1774
+ "loss": 0.7387,
1775
+ "step": 1240
1776
+ },
1777
+ {
1778
+ "epoch": 1.2743091095189354,
1779
+ "grad_norm": 0.532599151134491,
1780
+ "learning_rate": 7.35530975829656e-07,
1781
+ "loss": 0.741,
1782
+ "step": 1245
1783
+ },
1784
+ {
1785
+ "epoch": 1.2794268167860798,
1786
+ "grad_norm": 0.5400995016098022,
1787
+ "learning_rate": 7.323737521932457e-07,
1788
+ "loss": 0.7367,
1789
+ "step": 1250
1790
+ },
1791
+ {
1792
+ "epoch": 1.2845445240532243,
1793
+ "grad_norm": 0.5307775735855103,
1794
+ "learning_rate": 7.292046661712307e-07,
1795
+ "loss": 0.7399,
1796
+ "step": 1255
1797
+ },
1798
+ {
1799
+ "epoch": 1.2896622313203685,
1800
+ "grad_norm": 0.5908007621765137,
1801
+ "learning_rate": 7.260238795414366e-07,
1802
+ "loss": 0.74,
1803
+ "step": 1260
1804
+ },
1805
+ {
1806
+ "epoch": 1.2947799385875127,
1807
+ "grad_norm": 0.5410370826721191,
1808
+ "learning_rate": 7.228315546789907e-07,
1809
+ "loss": 0.7388,
1810
+ "step": 1265
1811
+ },
1812
+ {
1813
+ "epoch": 1.2998976458546572,
1814
+ "grad_norm": 0.5406989455223083,
1815
+ "learning_rate": 7.19627854548032e-07,
1816
+ "loss": 0.7337,
1817
+ "step": 1270
1818
+ },
1819
+ {
1820
+ "epoch": 1.3050153531218014,
1821
+ "grad_norm": 0.589767575263977,
1822
+ "learning_rate": 7.164129426933927e-07,
1823
+ "loss": 0.7426,
1824
+ "step": 1275
1825
+ },
1826
+ {
1827
+ "epoch": 1.3101330603889458,
1828
+ "grad_norm": 0.5926154255867004,
1829
+ "learning_rate": 7.131869832322496e-07,
1830
+ "loss": 0.7374,
1831
+ "step": 1280
1832
+ },
1833
+ {
1834
+ "epoch": 1.31525076765609,
1835
+ "grad_norm": 0.7507414817810059,
1836
+ "learning_rate": 7.099501408457452e-07,
1837
+ "loss": 0.7375,
1838
+ "step": 1285
1839
+ },
1840
+ {
1841
+ "epoch": 1.3203684749232343,
1842
+ "grad_norm": 0.6162967681884766,
1843
+ "learning_rate": 7.06702580770582e-07,
1844
+ "loss": 0.7381,
1845
+ "step": 1290
1846
+ },
1847
+ {
1848
+ "epoch": 1.3254861821903787,
1849
+ "grad_norm": 0.5118803977966309,
1850
+ "learning_rate": 7.034444687905868e-07,
1851
+ "loss": 0.7344,
1852
+ "step": 1295
1853
+ },
1854
+ {
1855
+ "epoch": 1.330603889457523,
1856
+ "grad_norm": 0.5982370972633362,
1857
+ "learning_rate": 7.001759712282478e-07,
1858
+ "loss": 0.7382,
1859
+ "step": 1300
1860
+ },
1861
+ {
1862
+ "epoch": 1.3357215967246674,
1863
+ "grad_norm": 0.6339845657348633,
1864
+ "learning_rate": 6.968972549362238e-07,
1865
+ "loss": 0.7386,
1866
+ "step": 1305
1867
+ },
1868
+ {
1869
+ "epoch": 1.3408393039918116,
1870
+ "grad_norm": 0.5755071043968201,
1871
+ "learning_rate": 6.936084872888271e-07,
1872
+ "loss": 0.7349,
1873
+ "step": 1310
1874
+ },
1875
+ {
1876
+ "epoch": 1.345957011258956,
1877
+ "grad_norm": 0.6089357137680054,
1878
+ "learning_rate": 6.90309836173479e-07,
1879
+ "loss": 0.7377,
1880
+ "step": 1315
1881
+ },
1882
+ {
1883
+ "epoch": 1.3510747185261003,
1884
+ "grad_norm": 0.6137183308601379,
1885
+ "learning_rate": 6.87001469982139e-07,
1886
+ "loss": 0.7417,
1887
+ "step": 1320
1888
+ },
1889
+ {
1890
+ "epoch": 1.3561924257932447,
1891
+ "grad_norm": 0.6864479184150696,
1892
+ "learning_rate": 6.836835576027093e-07,
1893
+ "loss": 0.7321,
1894
+ "step": 1325
1895
+ },
1896
+ {
1897
+ "epoch": 1.361310133060389,
1898
+ "grad_norm": 0.5657494068145752,
1899
+ "learning_rate": 6.803562684104125e-07,
1900
+ "loss": 0.7411,
1901
+ "step": 1330
1902
+ },
1903
+ {
1904
+ "epoch": 1.3664278403275332,
1905
+ "grad_norm": 0.6047109365463257,
1906
+ "learning_rate": 6.770197722591456e-07,
1907
+ "loss": 0.7399,
1908
+ "step": 1335
1909
+ },
1910
+ {
1911
+ "epoch": 1.3715455475946776,
1912
+ "grad_norm": 0.5772355198860168,
1913
+ "learning_rate": 6.736742394728097e-07,
1914
+ "loss": 0.7374,
1915
+ "step": 1340
1916
+ },
1917
+ {
1918
+ "epoch": 1.3766632548618218,
1919
+ "grad_norm": 0.7158586382865906,
1920
+ "learning_rate": 6.703198408366142e-07,
1921
+ "loss": 0.739,
1922
+ "step": 1345
1923
+ },
1924
+ {
1925
+ "epoch": 1.3817809621289663,
1926
+ "grad_norm": 0.5718494057655334,
1927
+ "learning_rate": 6.669567475883592e-07,
1928
+ "loss": 0.7435,
1929
+ "step": 1350
1930
+ },
1931
+ {
1932
+ "epoch": 1.3868986693961105,
1933
+ "grad_norm": 0.6494776606559753,
1934
+ "learning_rate": 6.635851314096935e-07,
1935
+ "loss": 0.7358,
1936
+ "step": 1355
1937
+ },
1938
+ {
1939
+ "epoch": 1.3920163766632547,
1940
+ "grad_norm": 0.5958154201507568,
1941
+ "learning_rate": 6.602051644173509e-07,
1942
+ "loss": 0.7375,
1943
+ "step": 1360
1944
+ },
1945
+ {
1946
+ "epoch": 1.3971340839303992,
1947
+ "grad_norm": 0.5509739518165588,
1948
+ "learning_rate": 6.568170191543634e-07,
1949
+ "loss": 0.7412,
1950
+ "step": 1365
1951
+ },
1952
+ {
1953
+ "epoch": 1.4022517911975436,
1954
+ "grad_norm": 0.5368937253952026,
1955
+ "learning_rate": 6.534208685812536e-07,
1956
+ "loss": 0.7393,
1957
+ "step": 1370
1958
+ },
1959
+ {
1960
+ "epoch": 1.4073694984646878,
1961
+ "grad_norm": 0.5369133353233337,
1962
+ "learning_rate": 6.500168860672047e-07,
1963
+ "loss": 0.7398,
1964
+ "step": 1375
1965
+ },
1966
+ {
1967
+ "epoch": 1.412487205731832,
1968
+ "grad_norm": 0.5789251327514648,
1969
+ "learning_rate": 6.466052453812111e-07,
1970
+ "loss": 0.7371,
1971
+ "step": 1380
1972
+ },
1973
+ {
1974
+ "epoch": 1.4176049129989765,
1975
+ "grad_norm": 0.5568552017211914,
1976
+ "learning_rate": 6.431861206832069e-07,
1977
+ "loss": 0.7363,
1978
+ "step": 1385
1979
+ },
1980
+ {
1981
+ "epoch": 1.4227226202661207,
1982
+ "grad_norm": 0.5325226783752441,
1983
+ "learning_rate": 6.397596865151752e-07,
1984
+ "loss": 0.7348,
1985
+ "step": 1390
1986
+ },
1987
+ {
1988
+ "epoch": 1.4278403275332652,
1989
+ "grad_norm": 0.5849957466125488,
1990
+ "learning_rate": 6.363261177922388e-07,
1991
+ "loss": 0.7363,
1992
+ "step": 1395
1993
+ },
1994
+ {
1995
+ "epoch": 1.4329580348004094,
1996
+ "grad_norm": 0.6208518743515015,
1997
+ "learning_rate": 6.328855897937303e-07,
1998
+ "loss": 0.7365,
1999
+ "step": 1400
2000
+ },
2001
+ {
2002
+ "epoch": 1.4380757420675536,
2003
+ "grad_norm": 0.5599240064620972,
2004
+ "learning_rate": 6.294382781542445e-07,
2005
+ "loss": 0.7371,
2006
+ "step": 1405
2007
+ },
2008
+ {
2009
+ "epoch": 1.443193449334698,
2010
+ "grad_norm": 0.5623425841331482,
2011
+ "learning_rate": 6.25984358854672e-07,
2012
+ "loss": 0.74,
2013
+ "step": 1410
2014
+ },
2015
+ {
2016
+ "epoch": 1.4483111566018425,
2017
+ "grad_norm": 0.6866716146469116,
2018
+ "learning_rate": 6.225240082132172e-07,
2019
+ "loss": 0.7383,
2020
+ "step": 1415
2021
+ },
2022
+ {
2023
+ "epoch": 1.4534288638689867,
2024
+ "grad_norm": 0.5852178931236267,
2025
+ "learning_rate": 6.190574028763952e-07,
2026
+ "loss": 0.7381,
2027
+ "step": 1420
2028
+ },
2029
+ {
2030
+ "epoch": 1.458546571136131,
2031
+ "grad_norm": 0.5319634079933167,
2032
+ "learning_rate": 6.15584719810016e-07,
2033
+ "loss": 0.7349,
2034
+ "step": 1425
2035
+ },
2036
+ {
2037
+ "epoch": 1.4636642784032754,
2038
+ "grad_norm": 0.5798255205154419,
2039
+ "learning_rate": 6.121061362901498e-07,
2040
+ "loss": 0.7331,
2041
+ "step": 1430
2042
+ },
2043
+ {
2044
+ "epoch": 1.4687819856704196,
2045
+ "grad_norm": 0.4803605079650879,
2046
+ "learning_rate": 6.086218298940778e-07,
2047
+ "loss": 0.7356,
2048
+ "step": 1435
2049
+ },
2050
+ {
2051
+ "epoch": 1.473899692937564,
2052
+ "grad_norm": 0.7146285772323608,
2053
+ "learning_rate": 6.051319784912261e-07,
2054
+ "loss": 0.7384,
2055
+ "step": 1440
2056
+ },
2057
+ {
2058
+ "epoch": 1.4790174002047083,
2059
+ "grad_norm": 0.47007301449775696,
2060
+ "learning_rate": 6.016367602340868e-07,
2061
+ "loss": 0.7332,
2062
+ "step": 1445
2063
+ },
2064
+ {
2065
+ "epoch": 1.4841351074718525,
2066
+ "grad_norm": 0.6568506956100464,
2067
+ "learning_rate": 5.981363535491233e-07,
2068
+ "loss": 0.7378,
2069
+ "step": 1450
2070
+ },
2071
+ {
2072
+ "epoch": 1.489252814738997,
2073
+ "grad_norm": 0.5178249478340149,
2074
+ "learning_rate": 5.946309371276614e-07,
2075
+ "loss": 0.7338,
2076
+ "step": 1455
2077
+ },
2078
+ {
2079
+ "epoch": 1.4943705220061412,
2080
+ "grad_norm": 0.5785830616950989,
2081
+ "learning_rate": 5.911206899167676e-07,
2082
+ "loss": 0.7392,
2083
+ "step": 1460
2084
+ },
2085
+ {
2086
+ "epoch": 1.4994882292732856,
2087
+ "grad_norm": 0.5021066665649414,
2088
+ "learning_rate": 5.87605791110114e-07,
2089
+ "loss": 0.7342,
2090
+ "step": 1465
2091
+ },
2092
+ {
2093
+ "epoch": 1.5046059365404298,
2094
+ "grad_norm": 0.5594333410263062,
2095
+ "learning_rate": 5.840864201388312e-07,
2096
+ "loss": 0.7351,
2097
+ "step": 1470
2098
+ },
2099
+ {
2100
+ "epoch": 1.509723643807574,
2101
+ "grad_norm": 0.5204704999923706,
2102
+ "learning_rate": 5.805627566623475e-07,
2103
+ "loss": 0.7375,
2104
+ "step": 1475
2105
+ },
2106
+ {
2107
+ "epoch": 1.5148413510747185,
2108
+ "grad_norm": 0.6187242865562439,
2109
+ "learning_rate": 5.770349805592185e-07,
2110
+ "loss": 0.7351,
2111
+ "step": 1480
2112
+ },
2113
+ {
2114
+ "epoch": 1.519959058341863,
2115
+ "grad_norm": 0.5294100046157837,
2116
+ "learning_rate": 5.735032719179443e-07,
2117
+ "loss": 0.7383,
2118
+ "step": 1485
2119
+ },
2120
+ {
2121
+ "epoch": 1.5250767656090072,
2122
+ "grad_norm": 0.5450606942176819,
2123
+ "learning_rate": 5.699678110277762e-07,
2124
+ "loss": 0.7365,
2125
+ "step": 1490
2126
+ },
2127
+ {
2128
+ "epoch": 1.5301944728761514,
2129
+ "grad_norm": 0.5091442465782166,
2130
+ "learning_rate": 5.664287783695122e-07,
2131
+ "loss": 0.7343,
2132
+ "step": 1495
2133
+ },
2134
+ {
2135
+ "epoch": 1.5353121801432958,
2136
+ "grad_norm": 0.557119607925415,
2137
+ "learning_rate": 5.628863546062856e-07,
2138
+ "loss": 0.7298,
2139
+ "step": 1500
2140
+ },
2141
+ {
2142
+ "epoch": 1.5353121801432958,
2143
+ "eval_accuracy": 0.67304,
2144
+ "eval_loss": 0.6938837766647339,
2145
+ "eval_macro_f1": 0.6609359830000188,
2146
+ "eval_precision": 0.685850518502884,
2147
+ "eval_recall": 0.6657447133221994,
2148
+ "eval_runtime": 73.8645,
2149
+ "eval_samples_per_second": 1353.83,
2150
+ "eval_steps_per_second": 1.327,
2151
+ "step": 1500
2152
+ }
2153
+ ],
2154
+ "logging_steps": 5,
2155
+ "max_steps": 4885,
2156
+ "num_input_tokens_seen": 0,
2157
+ "num_train_epochs": 5,
2158
+ "save_steps": 500,
2159
+ "stateful_callbacks": {
2160
+ "EarlyStoppingCallback": {
2161
+ "args": {
2162
+ "early_stopping_patience": 3,
2163
+ "early_stopping_threshold": 0.0
2164
+ },
2165
+ "attributes": {
2166
+ "early_stopping_patience_counter": 1
2167
+ }
2168
+ },
2169
+ "TrainerControl": {
2170
+ "args": {
2171
+ "should_epoch_stop": false,
2172
+ "should_evaluate": false,
2173
+ "should_log": false,
2174
+ "should_save": true,
2175
+ "should_training_stop": false
2176
+ },
2177
+ "attributes": {}
2178
+ }
2179
+ },
2180
+ "total_flos": 2.0201035364007936e+17,
2181
+ "train_batch_size": 512,
2182
+ "trial_name": null,
2183
+ "trial_params": null
2184
+ }
fourier-spectral-norm-classifier/checkpoint-1500/vocab.json ADDED
The diff for this file is too large to render. See raw diff