Hugging Face's logo

open-paws
/
animal_alignment_prediction_shortform

Text Generation
TensorBoard
Safetensors
English
distilbert
animal-liberation
animal-advocacy
open-paws
ethics
alignment
Model card Files
xet
 Metrics Community
samtuckervegan commited on
Commit
662f955
·
verified ·
1 Parent(s): b2f6c1c

Upload folder using huggingface_hub

Browse files
Files changed (21) hide show
  1. README.md +28 -0
  2. checkpoint-7980/config.json +32 -0
  3. checkpoint-7980/model.safetensors +3 -0
  4. checkpoint-7980/optimizer.pt +3 -0
  5. checkpoint-7980/rng_state_0.pth +3 -0
  6. checkpoint-7980/rng_state_1.pth +3 -0
  7. checkpoint-7980/rng_state_2.pth +3 -0
  8. checkpoint-7980/rng_state_3.pth +3 -0
  9. checkpoint-7980/scheduler.pt +3 -0
  10. checkpoint-7980/trainer_state.json +2314 -0
  11. checkpoint-7980/training_args.bin +3 -0
  12. config.json +32 -0
  13. model.safetensors +3 -0
  14. runs/Feb22_18-48-22_r-samtuckervegan-autotrain-4xl4-br7fljr5-341de-o890j/events.out.tfevents.1740250104.r-samtuckervegan-autotrain-4xl4-br7fljr5-341de-o890j.202.0 +2 -2
  15. runs/Feb22_18-48-22_r-samtuckervegan-autotrain-4xl4-br7fljr5-341de-o890j/events.out.tfevents.1740250884.r-samtuckervegan-autotrain-4xl4-br7fljr5-341de-o890j.202.1 +3 -0
  16. special_tokens_map.json +7 -0
  17. tokenizer.json +0 -0
  18. tokenizer_config.json +56 -0
  19. training_args.bin +3 -0
  20. training_params.json +30 -0
  21. vocab.txt +0 -0
README.md ADDED
@@ -0,0 +1,28 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+
2
+ ---
3
+ tags:
4
+ - autotrain
5
+ - text-regression
6
+ base_model: distilbert/distilbert-base-uncased
7
+ widget:
8
+ - text: "I love AutoTrain"
9
+ datasets:
10
+ - samtuckervegan/alignment
11
+ ---
12
+
13
+ # Model Trained Using AutoTrain
14
+
15
+ - Problem type: Text Regression
16
+
17
+ ## Validation Metrics
18
+ loss: 0.041510969400405884
19
+
20
+ mse: 0.041513968259096146
21
+
22
+ mae: 0.15856829285621643
23
+
24
+ r2: 0.4177439212799072
25
+
26
+ rmse: 0.203749768733847
27
+
28
+ explained_variance: 0.4178752303123474
checkpoint-7980/config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "distilbert/distilbert-base-uncased",
3
+ "_num_labels": 1,
4
+ "activation": "gelu",
5
+ "architectures": [
6
+ "DistilBertForSequenceClassification"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "dim": 768,
10
+ "dropout": 0.1,
11
+ "hidden_dim": 3072,
12
+ "id2label": {
13
+ "0": "target"
14
+ },
15
+ "initializer_range": 0.02,
16
+ "label2id": {
17
+ "target": 0
18
+ },
19
+ "max_position_embeddings": 512,
20
+ "model_type": "distilbert",
21
+ "n_heads": 12,
22
+ "n_layers": 6,
23
+ "pad_token_id": 0,
24
+ "problem_type": "regression",
25
+ "qa_dropout": 0.1,
26
+ "seq_classif_dropout": 0.2,
27
+ "sinusoidal_pos_embds": false,
28
+ "tie_weights_": true,
29
+ "torch_dtype": "float32",
30
+ "transformers_version": "4.48.0",
31
+ "vocab_size": 30522
32
+ }
checkpoint-7980/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ac3b1117a2c16efdaca4209bef8b2d615e37e4f1241391aaa60d1e8463e9c67
3
+ size 267829484
checkpoint-7980/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:202be585c316f776d73c4e1f98a81e60d3531a86a2ffe17b381ca04125da16a0
3
+ size 535721146
checkpoint-7980/rng_state_0.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:191b7d80c7eefd89ba0c9e105e927a40fadf120d2bb3de1043b025b16923c222
3
+ size 15024
checkpoint-7980/rng_state_1.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:265d4e8794d69f71649f53e597bae382154bb6b8016836972c3c5b93e2842fa7
3
+ size 15024
checkpoint-7980/rng_state_2.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b915cb9898748bc446c463bdd075e28a9a83539067b921a4decdea6a83d093d9
3
+ size 15024
checkpoint-7980/rng_state_3.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1cf64ec88b98d7b2fb48824a98c28789d1de90d7fa5bdda84a74e8044466972e
3
+ size 15024
checkpoint-7980/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a0c3b48e85066c0f06c8412dd982a0730376407d87a004332409cdaabe7fa80b
3
+ size 1064
checkpoint-7980/trainer_state.json ADDED
@@ -0,0 +1,2314 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.041510969400405884,
3
+ "best_model_checkpoint": "alignment/checkpoint-7980",
4
+ "epoch": 3.0,
5
+ "eval_steps": 500,
6
+ "global_step": 7980,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.009398496240601503,
13
+ "grad_norm": 6.57550048828125,
14
+ "learning_rate": 1.5037593984962406e-06,
15
+ "loss": 0.423,
16
+ "step": 25
17
+ },
18
+ {
19
+ "epoch": 0.018796992481203006,
20
+ "grad_norm": 1.1106289625167847,
21
+ "learning_rate": 3.070175438596491e-06,
22
+ "loss": 0.2162,
23
+ "step": 50
24
+ },
25
+ {
26
+ "epoch": 0.02819548872180451,
27
+ "grad_norm": 0.636308491230011,
28
+ "learning_rate": 4.636591478696742e-06,
29
+ "loss": 0.0835,
30
+ "step": 75
31
+ },
32
+ {
33
+ "epoch": 0.03759398496240601,
34
+ "grad_norm": 0.8234844207763672,
35
+ "learning_rate": 6.203007518796992e-06,
36
+ "loss": 0.0706,
37
+ "step": 100
38
+ },
39
+ {
40
+ "epoch": 0.046992481203007516,
41
+ "grad_norm": 0.8412534594535828,
42
+ "learning_rate": 7.769423558897243e-06,
43
+ "loss": 0.0694,
44
+ "step": 125
45
+ },
46
+ {
47
+ "epoch": 0.05639097744360902,
48
+ "grad_norm": 0.6980524063110352,
49
+ "learning_rate": 9.335839598997493e-06,
50
+ "loss": 0.0689,
51
+ "step": 150
52
+ },
53
+ {
54
+ "epoch": 0.06578947368421052,
55
+ "grad_norm": 0.5269452333450317,
56
+ "learning_rate": 1.0902255639097744e-05,
57
+ "loss": 0.0666,
58
+ "step": 175
59
+ },
60
+ {
61
+ "epoch": 0.07518796992481203,
62
+ "grad_norm": 0.9237763285636902,
63
+ "learning_rate": 1.2468671679197996e-05,
64
+ "loss": 0.0675,
65
+ "step": 200
66
+ },
67
+ {
68
+ "epoch": 0.08458646616541353,
69
+ "grad_norm": 2.0623257160186768,
70
+ "learning_rate": 1.4035087719298246e-05,
71
+ "loss": 0.0696,
72
+ "step": 225
73
+ },
74
+ {
75
+ "epoch": 0.09398496240601503,
76
+ "grad_norm": 1.0979938507080078,
77
+ "learning_rate": 1.5601503759398496e-05,
78
+ "loss": 0.0663,
79
+ "step": 250
80
+ },
81
+ {
82
+ "epoch": 0.10338345864661654,
83
+ "grad_norm": 1.7202848196029663,
84
+ "learning_rate": 1.7167919799498746e-05,
85
+ "loss": 0.0638,
86
+ "step": 275
87
+ },
88
+ {
89
+ "epoch": 0.11278195488721804,
90
+ "grad_norm": 1.5974129438400269,
91
+ "learning_rate": 1.8734335839599e-05,
92
+ "loss": 0.059,
93
+ "step": 300
94
+ },
95
+ {
96
+ "epoch": 0.12218045112781954,
97
+ "grad_norm": 0.614483654499054,
98
+ "learning_rate": 2.0300751879699247e-05,
99
+ "loss": 0.0608,
100
+ "step": 325
101
+ },
102
+ {
103
+ "epoch": 0.13157894736842105,
104
+ "grad_norm": 0.48640942573547363,
105
+ "learning_rate": 2.18671679197995e-05,
106
+ "loss": 0.0595,
107
+ "step": 350
108
+ },
109
+ {
110
+ "epoch": 0.14097744360902256,
111
+ "grad_norm": 0.558945894241333,
112
+ "learning_rate": 2.3433583959899748e-05,
113
+ "loss": 0.0584,
114
+ "step": 375
115
+ },
116
+ {
117
+ "epoch": 0.15037593984962405,
118
+ "grad_norm": 0.3774773180484772,
119
+ "learning_rate": 2.5e-05,
120
+ "loss": 0.0659,
121
+ "step": 400
122
+ },
123
+ {
124
+ "epoch": 0.15977443609022557,
125
+ "grad_norm": 0.4286479949951172,
126
+ "learning_rate": 2.656641604010025e-05,
127
+ "loss": 0.0552,
128
+ "step": 425
129
+ },
130
+ {
131
+ "epoch": 0.16917293233082706,
132
+ "grad_norm": 0.8272174596786499,
133
+ "learning_rate": 2.8132832080200505e-05,
134
+ "loss": 0.0594,
135
+ "step": 450
136
+ },
137
+ {
138
+ "epoch": 0.17857142857142858,
139
+ "grad_norm": 0.8896750807762146,
140
+ "learning_rate": 2.9699248120300755e-05,
141
+ "loss": 0.0573,
142
+ "step": 475
143
+ },
144
+ {
145
+ "epoch": 0.18796992481203006,
146
+ "grad_norm": 0.7559242844581604,
147
+ "learning_rate": 3.1265664160401006e-05,
148
+ "loss": 0.0521,
149
+ "step": 500
150
+ },
151
+ {
152
+ "epoch": 0.19736842105263158,
153
+ "grad_norm": 0.46636250615119934,
154
+ "learning_rate": 3.2832080200501256e-05,
155
+ "loss": 0.0554,
156
+ "step": 525
157
+ },
158
+ {
159
+ "epoch": 0.20676691729323307,
160
+ "grad_norm": 0.6456935405731201,
161
+ "learning_rate": 3.4398496240601506e-05,
162
+ "loss": 0.0572,
163
+ "step": 550
164
+ },
165
+ {
166
+ "epoch": 0.2161654135338346,
167
+ "grad_norm": 1.4037501811981201,
168
+ "learning_rate": 3.5964912280701756e-05,
169
+ "loss": 0.0581,
170
+ "step": 575
171
+ },
172
+ {
173
+ "epoch": 0.22556390977443608,
174
+ "grad_norm": 0.7281235456466675,
175
+ "learning_rate": 3.753132832080201e-05,
176
+ "loss": 0.0608,
177
+ "step": 600
178
+ },
179
+ {
180
+ "epoch": 0.2349624060150376,
181
+ "grad_norm": 0.6525343656539917,
182
+ "learning_rate": 3.909774436090226e-05,
183
+ "loss": 0.0549,
184
+ "step": 625
185
+ },
186
+ {
187
+ "epoch": 0.24436090225563908,
188
+ "grad_norm": 0.6867189407348633,
189
+ "learning_rate": 4.066416040100251e-05,
190
+ "loss": 0.0573,
191
+ "step": 650
192
+ },
193
+ {
194
+ "epoch": 0.25375939849624063,
195
+ "grad_norm": 1.408666968345642,
196
+ "learning_rate": 4.223057644110276e-05,
197
+ "loss": 0.0513,
198
+ "step": 675
199
+ },
200
+ {
201
+ "epoch": 0.2631578947368421,
202
+ "grad_norm": 0.4105112552642822,
203
+ "learning_rate": 4.379699248120301e-05,
204
+ "loss": 0.0561,
205
+ "step": 700
206
+ },
207
+ {
208
+ "epoch": 0.2725563909774436,
209
+ "grad_norm": 0.40179911255836487,
210
+ "learning_rate": 4.536340852130326e-05,
211
+ "loss": 0.0586,
212
+ "step": 725
213
+ },
214
+ {
215
+ "epoch": 0.2819548872180451,
216
+ "grad_norm": 0.49944761395454407,
217
+ "learning_rate": 4.6929824561403515e-05,
218
+ "loss": 0.0569,
219
+ "step": 750
220
+ },
221
+ {
222
+ "epoch": 0.29135338345864664,
223
+ "grad_norm": 1.1651484966278076,
224
+ "learning_rate": 4.849624060150376e-05,
225
+ "loss": 0.0571,
226
+ "step": 775
227
+ },
228
+ {
229
+ "epoch": 0.3007518796992481,
230
+ "grad_norm": 0.5590141415596008,
231
+ "learning_rate": 4.9993038150932894e-05,
232
+ "loss": 0.0612,
233
+ "step": 800
234
+ },
235
+ {
236
+ "epoch": 0.3101503759398496,
237
+ "grad_norm": 0.6796597242355347,
238
+ "learning_rate": 4.9818991924255085e-05,
239
+ "loss": 0.0534,
240
+ "step": 825
241
+ },
242
+ {
243
+ "epoch": 0.31954887218045114,
244
+ "grad_norm": 0.8797287344932556,
245
+ "learning_rate": 4.9644945697577276e-05,
246
+ "loss": 0.0534,
247
+ "step": 850
248
+ },
249
+ {
250
+ "epoch": 0.32894736842105265,
251
+ "grad_norm": 0.7587268948554993,
252
+ "learning_rate": 4.9470899470899475e-05,
253
+ "loss": 0.0548,
254
+ "step": 875
255
+ },
256
+ {
257
+ "epoch": 0.3383458646616541,
258
+ "grad_norm": 0.9078999757766724,
259
+ "learning_rate": 4.9296853244221666e-05,
260
+ "loss": 0.0534,
261
+ "step": 900
262
+ },
263
+ {
264
+ "epoch": 0.34774436090225563,
265
+ "grad_norm": 0.4876460134983063,
266
+ "learning_rate": 4.912280701754386e-05,
267
+ "loss": 0.0545,
268
+ "step": 925
269
+ },
270
+ {
271
+ "epoch": 0.35714285714285715,
272
+ "grad_norm": 1.069460153579712,
273
+ "learning_rate": 4.8948760790866056e-05,
274
+ "loss": 0.0605,
275
+ "step": 950
276
+ },
277
+ {
278
+ "epoch": 0.36654135338345867,
279
+ "grad_norm": 0.7750365734100342,
280
+ "learning_rate": 4.8774714564188254e-05,
281
+ "loss": 0.0518,
282
+ "step": 975
283
+ },
284
+ {
285
+ "epoch": 0.37593984962406013,
286
+ "grad_norm": 0.5427040457725525,
287
+ "learning_rate": 4.8600668337510445e-05,
288
+ "loss": 0.0528,
289
+ "step": 1000
290
+ },
291
+ {
292
+ "epoch": 0.38533834586466165,
293
+ "grad_norm": 1.0749157667160034,
294
+ "learning_rate": 4.8426622110832637e-05,
295
+ "loss": 0.0504,
296
+ "step": 1025
297
+ },
298
+ {
299
+ "epoch": 0.39473684210526316,
300
+ "grad_norm": 1.0911959409713745,
301
+ "learning_rate": 4.8252575884154835e-05,
302
+ "loss": 0.0552,
303
+ "step": 1050
304
+ },
305
+ {
306
+ "epoch": 0.4041353383458647,
307
+ "grad_norm": 1.9141292572021484,
308
+ "learning_rate": 4.8078529657477026e-05,
309
+ "loss": 0.0584,
310
+ "step": 1075
311
+ },
312
+ {
313
+ "epoch": 0.41353383458646614,
314
+ "grad_norm": 0.6295194029808044,
315
+ "learning_rate": 4.7904483430799224e-05,
316
+ "loss": 0.0502,
317
+ "step": 1100
318
+ },
319
+ {
320
+ "epoch": 0.42293233082706766,
321
+ "grad_norm": 0.47915300726890564,
322
+ "learning_rate": 4.7730437204121416e-05,
323
+ "loss": 0.0486,
324
+ "step": 1125
325
+ },
326
+ {
327
+ "epoch": 0.4323308270676692,
328
+ "grad_norm": 0.306090384721756,
329
+ "learning_rate": 4.755639097744361e-05,
330
+ "loss": 0.0491,
331
+ "step": 1150
332
+ },
333
+ {
334
+ "epoch": 0.4417293233082707,
335
+ "grad_norm": 1.4408737421035767,
336
+ "learning_rate": 4.7382344750765805e-05,
337
+ "loss": 0.0485,
338
+ "step": 1175
339
+ },
340
+ {
341
+ "epoch": 0.45112781954887216,
342
+ "grad_norm": 0.39153051376342773,
343
+ "learning_rate": 4.7208298524088003e-05,
344
+ "loss": 0.0497,
345
+ "step": 1200
346
+ },
347
+ {
348
+ "epoch": 0.4605263157894737,
349
+ "grad_norm": 0.9627327919006348,
350
+ "learning_rate": 4.7034252297410195e-05,
351
+ "loss": 0.0537,
352
+ "step": 1225
353
+ },
354
+ {
355
+ "epoch": 0.4699248120300752,
356
+ "grad_norm": 0.24377842247486115,
357
+ "learning_rate": 4.6860206070732386e-05,
358
+ "loss": 0.0536,
359
+ "step": 1250
360
+ },
361
+ {
362
+ "epoch": 0.4793233082706767,
363
+ "grad_norm": 0.8792003989219666,
364
+ "learning_rate": 4.6686159844054584e-05,
365
+ "loss": 0.0498,
366
+ "step": 1275
367
+ },
368
+ {
369
+ "epoch": 0.48872180451127817,
370
+ "grad_norm": 0.6982893943786621,
371
+ "learning_rate": 4.6512113617376776e-05,
372
+ "loss": 0.0539,
373
+ "step": 1300
374
+ },
375
+ {
376
+ "epoch": 0.4981203007518797,
377
+ "grad_norm": 0.2784109115600586,
378
+ "learning_rate": 4.633806739069897e-05,
379
+ "loss": 0.0528,
380
+ "step": 1325
381
+ },
382
+ {
383
+ "epoch": 0.5075187969924813,
384
+ "grad_norm": 0.38880228996276855,
385
+ "learning_rate": 4.6164021164021165e-05,
386
+ "loss": 0.0503,
387
+ "step": 1350
388
+ },
389
+ {
390
+ "epoch": 0.5169172932330827,
391
+ "grad_norm": 0.8160899877548218,
392
+ "learning_rate": 4.5989974937343364e-05,
393
+ "loss": 0.0548,
394
+ "step": 1375
395
+ },
396
+ {
397
+ "epoch": 0.5263157894736842,
398
+ "grad_norm": 0.36486560106277466,
399
+ "learning_rate": 4.5815928710665555e-05,
400
+ "loss": 0.0504,
401
+ "step": 1400
402
+ },
403
+ {
404
+ "epoch": 0.5357142857142857,
405
+ "grad_norm": 0.3815467357635498,
406
+ "learning_rate": 4.5641882483987746e-05,
407
+ "loss": 0.0505,
408
+ "step": 1425
409
+ },
410
+ {
411
+ "epoch": 0.5451127819548872,
412
+ "grad_norm": 0.40433645248413086,
413
+ "learning_rate": 4.5467836257309945e-05,
414
+ "loss": 0.0467,
415
+ "step": 1450
416
+ },
417
+ {
418
+ "epoch": 0.5545112781954887,
419
+ "grad_norm": 0.334482342004776,
420
+ "learning_rate": 4.5293790030632136e-05,
421
+ "loss": 0.0473,
422
+ "step": 1475
423
+ },
424
+ {
425
+ "epoch": 0.5639097744360902,
426
+ "grad_norm": 0.45450279116630554,
427
+ "learning_rate": 4.5119743803954334e-05,
428
+ "loss": 0.0512,
429
+ "step": 1500
430
+ },
431
+ {
432
+ "epoch": 0.5733082706766918,
433
+ "grad_norm": 0.8257545232772827,
434
+ "learning_rate": 4.4945697577276526e-05,
435
+ "loss": 0.053,
436
+ "step": 1525
437
+ },
438
+ {
439
+ "epoch": 0.5827067669172933,
440
+ "grad_norm": 0.3921704888343811,
441
+ "learning_rate": 4.477165135059872e-05,
442
+ "loss": 0.0472,
443
+ "step": 1550
444
+ },
445
+ {
446
+ "epoch": 0.5921052631578947,
447
+ "grad_norm": 0.3535679876804352,
448
+ "learning_rate": 4.4597605123920915e-05,
449
+ "loss": 0.0505,
450
+ "step": 1575
451
+ },
452
+ {
453
+ "epoch": 0.6015037593984962,
454
+ "grad_norm": 0.6803102493286133,
455
+ "learning_rate": 4.442355889724311e-05,
456
+ "loss": 0.0514,
457
+ "step": 1600
458
+ },
459
+ {
460
+ "epoch": 0.6109022556390977,
461
+ "grad_norm": 0.3447111248970032,
462
+ "learning_rate": 4.4249512670565305e-05,
463
+ "loss": 0.0468,
464
+ "step": 1625
465
+ },
466
+ {
467
+ "epoch": 0.6203007518796992,
468
+ "grad_norm": 0.705673336982727,
469
+ "learning_rate": 4.4075466443887496e-05,
470
+ "loss": 0.049,
471
+ "step": 1650
472
+ },
473
+ {
474
+ "epoch": 0.6296992481203008,
475
+ "grad_norm": 0.2624443471431732,
476
+ "learning_rate": 4.3901420217209694e-05,
477
+ "loss": 0.0494,
478
+ "step": 1675
479
+ },
480
+ {
481
+ "epoch": 0.6390977443609023,
482
+ "grad_norm": 0.5573770403862,
483
+ "learning_rate": 4.3727373990531886e-05,
484
+ "loss": 0.0463,
485
+ "step": 1700
486
+ },
487
+ {
488
+ "epoch": 0.6484962406015038,
489
+ "grad_norm": 0.7310426235198975,
490
+ "learning_rate": 4.355332776385408e-05,
491
+ "loss": 0.05,
492
+ "step": 1725
493
+ },
494
+ {
495
+ "epoch": 0.6578947368421053,
496
+ "grad_norm": 0.3286091387271881,
497
+ "learning_rate": 4.3379281537176275e-05,
498
+ "loss": 0.0477,
499
+ "step": 1750
500
+ },
501
+ {
502
+ "epoch": 0.6672932330827067,
503
+ "grad_norm": 0.5842466354370117,
504
+ "learning_rate": 4.3205235310498473e-05,
505
+ "loss": 0.0462,
506
+ "step": 1775
507
+ },
508
+ {
509
+ "epoch": 0.6766917293233082,
510
+ "grad_norm": 0.3797430098056793,
511
+ "learning_rate": 4.3031189083820665e-05,
512
+ "loss": 0.0441,
513
+ "step": 1800
514
+ },
515
+ {
516
+ "epoch": 0.6860902255639098,
517
+ "grad_norm": 0.5318741202354431,
518
+ "learning_rate": 4.2857142857142856e-05,
519
+ "loss": 0.0514,
520
+ "step": 1825
521
+ },
522
+ {
523
+ "epoch": 0.6954887218045113,
524
+ "grad_norm": 0.35251280665397644,
525
+ "learning_rate": 4.2683096630465054e-05,
526
+ "loss": 0.0481,
527
+ "step": 1850
528
+ },
529
+ {
530
+ "epoch": 0.7048872180451128,
531
+ "grad_norm": 0.5046139359474182,
532
+ "learning_rate": 4.2509050403787246e-05,
533
+ "loss": 0.0454,
534
+ "step": 1875
535
+ },
536
+ {
537
+ "epoch": 0.7142857142857143,
538
+ "grad_norm": 0.2886251211166382,
539
+ "learning_rate": 4.2335004177109444e-05,
540
+ "loss": 0.0501,
541
+ "step": 1900
542
+ },
543
+ {
544
+ "epoch": 0.7236842105263158,
545
+ "grad_norm": 0.3872503340244293,
546
+ "learning_rate": 4.2160957950431635e-05,
547
+ "loss": 0.0484,
548
+ "step": 1925
549
+ },
550
+ {
551
+ "epoch": 0.7330827067669173,
552
+ "grad_norm": 0.38036292791366577,
553
+ "learning_rate": 4.198691172375383e-05,
554
+ "loss": 0.0484,
555
+ "step": 1950
556
+ },
557
+ {
558
+ "epoch": 0.7424812030075187,
559
+ "grad_norm": 0.5636013746261597,
560
+ "learning_rate": 4.1812865497076025e-05,
561
+ "loss": 0.0508,
562
+ "step": 1975
563
+ },
564
+ {
565
+ "epoch": 0.7518796992481203,
566
+ "grad_norm": 0.4981440007686615,
567
+ "learning_rate": 4.163881927039822e-05,
568
+ "loss": 0.0471,
569
+ "step": 2000
570
+ },
571
+ {
572
+ "epoch": 0.7612781954887218,
573
+ "grad_norm": 0.32413703203201294,
574
+ "learning_rate": 4.1464773043720415e-05,
575
+ "loss": 0.0496,
576
+ "step": 2025
577
+ },
578
+ {
579
+ "epoch": 0.7706766917293233,
580
+ "grad_norm": 0.4851779639720917,
581
+ "learning_rate": 4.1290726817042606e-05,
582
+ "loss": 0.0482,
583
+ "step": 2050
584
+ },
585
+ {
586
+ "epoch": 0.7800751879699248,
587
+ "grad_norm": 0.4027095437049866,
588
+ "learning_rate": 4.1116680590364804e-05,
589
+ "loss": 0.0463,
590
+ "step": 2075
591
+ },
592
+ {
593
+ "epoch": 0.7894736842105263,
594
+ "grad_norm": 0.29670262336730957,
595
+ "learning_rate": 4.0942634363687e-05,
596
+ "loss": 0.0444,
597
+ "step": 2100
598
+ },
599
+ {
600
+ "epoch": 0.7988721804511278,
601
+ "grad_norm": 0.6292144060134888,
602
+ "learning_rate": 4.076858813700919e-05,
603
+ "loss": 0.0513,
604
+ "step": 2125
605
+ },
606
+ {
607
+ "epoch": 0.8082706766917294,
608
+ "grad_norm": 0.451262891292572,
609
+ "learning_rate": 4.0594541910331385e-05,
610
+ "loss": 0.0487,
611
+ "step": 2150
612
+ },
613
+ {
614
+ "epoch": 0.8176691729323309,
615
+ "grad_norm": 0.42661145329475403,
616
+ "learning_rate": 4.042049568365358e-05,
617
+ "loss": 0.0445,
618
+ "step": 2175
619
+ },
620
+ {
621
+ "epoch": 0.8270676691729323,
622
+ "grad_norm": 0.6654478907585144,
623
+ "learning_rate": 4.0246449456975775e-05,
624
+ "loss": 0.0499,
625
+ "step": 2200
626
+ },
627
+ {
628
+ "epoch": 0.8364661654135338,
629
+ "grad_norm": 0.38029587268829346,
630
+ "learning_rate": 4.0072403230297966e-05,
631
+ "loss": 0.0455,
632
+ "step": 2225
633
+ },
634
+ {
635
+ "epoch": 0.8458646616541353,
636
+ "grad_norm": 0.6190987229347229,
637
+ "learning_rate": 3.9898357003620164e-05,
638
+ "loss": 0.0461,
639
+ "step": 2250
640
+ },
641
+ {
642
+ "epoch": 0.8552631578947368,
643
+ "grad_norm": 0.34594255685806274,
644
+ "learning_rate": 3.9724310776942356e-05,
645
+ "loss": 0.0442,
646
+ "step": 2275
647
+ },
648
+ {
649
+ "epoch": 0.8646616541353384,
650
+ "grad_norm": 0.32449814677238464,
651
+ "learning_rate": 3.9550264550264554e-05,
652
+ "loss": 0.0477,
653
+ "step": 2300
654
+ },
655
+ {
656
+ "epoch": 0.8740601503759399,
657
+ "grad_norm": 0.9444015622138977,
658
+ "learning_rate": 3.9376218323586745e-05,
659
+ "loss": 0.0443,
660
+ "step": 2325
661
+ },
662
+ {
663
+ "epoch": 0.8834586466165414,
664
+ "grad_norm": 0.356588751077652,
665
+ "learning_rate": 3.920217209690894e-05,
666
+ "loss": 0.0442,
667
+ "step": 2350
668
+ },
669
+ {
670
+ "epoch": 0.8928571428571429,
671
+ "grad_norm": 0.5112847089767456,
672
+ "learning_rate": 3.9028125870231135e-05,
673
+ "loss": 0.0529,
674
+ "step": 2375
675
+ },
676
+ {
677
+ "epoch": 0.9022556390977443,
678
+ "grad_norm": 0.8283340334892273,
679
+ "learning_rate": 3.885407964355333e-05,
680
+ "loss": 0.0485,
681
+ "step": 2400
682
+ },
683
+ {
684
+ "epoch": 0.9116541353383458,
685
+ "grad_norm": 0.302156001329422,
686
+ "learning_rate": 3.8680033416875525e-05,
687
+ "loss": 0.0479,
688
+ "step": 2425
689
+ },
690
+ {
691
+ "epoch": 0.9210526315789473,
692
+ "grad_norm": 0.5275407433509827,
693
+ "learning_rate": 3.8505987190197716e-05,
694
+ "loss": 0.0478,
695
+ "step": 2450
696
+ },
697
+ {
698
+ "epoch": 0.9304511278195489,
699
+ "grad_norm": 0.481001079082489,
700
+ "learning_rate": 3.8331940963519914e-05,
701
+ "loss": 0.0461,
702
+ "step": 2475
703
+ },
704
+ {
705
+ "epoch": 0.9398496240601504,
706
+ "grad_norm": 1.3727526664733887,
707
+ "learning_rate": 3.815789473684211e-05,
708
+ "loss": 0.0451,
709
+ "step": 2500
710
+ },
711
+ {
712
+ "epoch": 0.9492481203007519,
713
+ "grad_norm": 0.41034597158432007,
714
+ "learning_rate": 3.79838485101643e-05,
715
+ "loss": 0.0496,
716
+ "step": 2525
717
+ },
718
+ {
719
+ "epoch": 0.9586466165413534,
720
+ "grad_norm": 0.8582698702812195,
721
+ "learning_rate": 3.7809802283486495e-05,
722
+ "loss": 0.048,
723
+ "step": 2550
724
+ },
725
+ {
726
+ "epoch": 0.9680451127819549,
727
+ "grad_norm": 0.777698814868927,
728
+ "learning_rate": 3.763575605680869e-05,
729
+ "loss": 0.0501,
730
+ "step": 2575
731
+ },
732
+ {
733
+ "epoch": 0.9774436090225563,
734
+ "grad_norm": 0.548724353313446,
735
+ "learning_rate": 3.7461709830130885e-05,
736
+ "loss": 0.0462,
737
+ "step": 2600
738
+ },
739
+ {
740
+ "epoch": 0.9868421052631579,
741
+ "grad_norm": 0.5295637845993042,
742
+ "learning_rate": 3.7287663603453076e-05,
743
+ "loss": 0.0448,
744
+ "step": 2625
745
+ },
746
+ {
747
+ "epoch": 0.9962406015037594,
748
+ "grad_norm": 0.4272915720939636,
749
+ "learning_rate": 3.7113617376775274e-05,
750
+ "loss": 0.0482,
751
+ "step": 2650
752
+ },
753
+ {
754
+ "epoch": 1.0,
755
+ "eval_explained_variance": 0.3670263886451721,
756
+ "eval_loss": 0.047455575317144394,
757
+ "eval_mae": 0.17258569598197937,
758
+ "eval_mse": 0.04745021089911461,
759
+ "eval_r2": 0.33448487520217896,
760
+ "eval_rmse": 0.21783069319798487,
761
+ "eval_runtime": 5.0137,
762
+ "eval_samples_per_second": 4243.803,
763
+ "eval_steps_per_second": 66.418,
764
+ "step": 2660
765
+ },
766
+ {
767
+ "epoch": 1.005639097744361,
768
+ "grad_norm": 0.44887444376945496,
769
+ "learning_rate": 3.6939571150097466e-05,
770
+ "loss": 0.0418,
771
+ "step": 2675
772
+ },
773
+ {
774
+ "epoch": 1.0150375939849625,
775
+ "grad_norm": 0.3950202763080597,
776
+ "learning_rate": 3.6765524923419664e-05,
777
+ "loss": 0.0418,
778
+ "step": 2700
779
+ },
780
+ {
781
+ "epoch": 1.0244360902255638,
782
+ "grad_norm": 0.9042419791221619,
783
+ "learning_rate": 3.6591478696741855e-05,
784
+ "loss": 0.0409,
785
+ "step": 2725
786
+ },
787
+ {
788
+ "epoch": 1.0338345864661653,
789
+ "grad_norm": 0.6256518363952637,
790
+ "learning_rate": 3.6417432470064053e-05,
791
+ "loss": 0.0444,
792
+ "step": 2750
793
+ },
794
+ {
795
+ "epoch": 1.0432330827067668,
796
+ "grad_norm": 0.2981257736682892,
797
+ "learning_rate": 3.6243386243386245e-05,
798
+ "loss": 0.0495,
799
+ "step": 2775
800
+ },
801
+ {
802
+ "epoch": 1.0526315789473684,
803
+ "grad_norm": 0.3158023953437805,
804
+ "learning_rate": 3.606934001670844e-05,
805
+ "loss": 0.0436,
806
+ "step": 2800
807
+ },
808
+ {
809
+ "epoch": 1.0620300751879699,
810
+ "grad_norm": 0.3816152513027191,
811
+ "learning_rate": 3.5895293790030634e-05,
812
+ "loss": 0.0426,
813
+ "step": 2825
814
+ },
815
+ {
816
+ "epoch": 1.0714285714285714,
817
+ "grad_norm": 0.6720510125160217,
818
+ "learning_rate": 3.5721247563352826e-05,
819
+ "loss": 0.0452,
820
+ "step": 2850
821
+ },
822
+ {
823
+ "epoch": 1.080827067669173,
824
+ "grad_norm": 0.3950083553791046,
825
+ "learning_rate": 3.5547201336675024e-05,
826
+ "loss": 0.0441,
827
+ "step": 2875
828
+ },
829
+ {
830
+ "epoch": 1.0902255639097744,
831
+ "grad_norm": 0.35873764753341675,
832
+ "learning_rate": 3.537315510999722e-05,
833
+ "loss": 0.0446,
834
+ "step": 2900
835
+ },
836
+ {
837
+ "epoch": 1.099624060150376,
838
+ "grad_norm": 0.2690645158290863,
839
+ "learning_rate": 3.519910888331941e-05,
840
+ "loss": 0.0431,
841
+ "step": 2925
842
+ },
843
+ {
844
+ "epoch": 1.1090225563909775,
845
+ "grad_norm": 0.7918988466262817,
846
+ "learning_rate": 3.5025062656641605e-05,
847
+ "loss": 0.0445,
848
+ "step": 2950
849
+ },
850
+ {
851
+ "epoch": 1.118421052631579,
852
+ "grad_norm": 0.6265382766723633,
853
+ "learning_rate": 3.48510164299638e-05,
854
+ "loss": 0.0448,
855
+ "step": 2975
856
+ },
857
+ {
858
+ "epoch": 1.1278195488721805,
859
+ "grad_norm": 0.3529634475708008,
860
+ "learning_rate": 3.4676970203285995e-05,
861
+ "loss": 0.0402,
862
+ "step": 3000
863
+ },
864
+ {
865
+ "epoch": 1.137218045112782,
866
+ "grad_norm": 0.21070072054862976,
867
+ "learning_rate": 3.4502923976608186e-05,
868
+ "loss": 0.0389,
869
+ "step": 3025
870
+ },
871
+ {
872
+ "epoch": 1.1466165413533835,
873
+ "grad_norm": 0.6971052289009094,
874
+ "learning_rate": 3.4328877749930384e-05,
875
+ "loss": 0.0434,
876
+ "step": 3050
877
+ },
878
+ {
879
+ "epoch": 1.156015037593985,
880
+ "grad_norm": 0.4424818456172943,
881
+ "learning_rate": 3.4154831523252576e-05,
882
+ "loss": 0.0446,
883
+ "step": 3075
884
+ },
885
+ {
886
+ "epoch": 1.1654135338345863,
887
+ "grad_norm": 0.5255579352378845,
888
+ "learning_rate": 3.3980785296574774e-05,
889
+ "loss": 0.0452,
890
+ "step": 3100
891
+ },
892
+ {
893
+ "epoch": 1.1748120300751879,
894
+ "grad_norm": 0.3643440306186676,
895
+ "learning_rate": 3.3806739069896965e-05,
896
+ "loss": 0.0489,
897
+ "step": 3125
898
+ },
899
+ {
900
+ "epoch": 1.1842105263157894,
901
+ "grad_norm": 0.8218281269073486,
902
+ "learning_rate": 3.363269284321916e-05,
903
+ "loss": 0.0434,
904
+ "step": 3150
905
+ },
906
+ {
907
+ "epoch": 1.193609022556391,
908
+ "grad_norm": 0.9634317755699158,
909
+ "learning_rate": 3.3458646616541355e-05,
910
+ "loss": 0.0413,
911
+ "step": 3175
912
+ },
913
+ {
914
+ "epoch": 1.2030075187969924,
915
+ "grad_norm": 1.1033918857574463,
916
+ "learning_rate": 3.328460038986355e-05,
917
+ "loss": 0.0449,
918
+ "step": 3200
919
+ },
920
+ {
921
+ "epoch": 1.212406015037594,
922
+ "grad_norm": 0.40083539485931396,
923
+ "learning_rate": 3.3110554163185744e-05,
924
+ "loss": 0.0435,
925
+ "step": 3225
926
+ },
927
+ {
928
+ "epoch": 1.2218045112781954,
929
+ "grad_norm": 0.36151793599128723,
930
+ "learning_rate": 3.2936507936507936e-05,
931
+ "loss": 0.0438,
932
+ "step": 3250
933
+ },
934
+ {
935
+ "epoch": 1.231203007518797,
936
+ "grad_norm": 0.661200225353241,
937
+ "learning_rate": 3.2762461709830134e-05,
938
+ "loss": 0.0448,
939
+ "step": 3275
940
+ },
941
+ {
942
+ "epoch": 1.2406015037593985,
943
+ "grad_norm": 0.2727440893650055,
944
+ "learning_rate": 3.258841548315233e-05,
945
+ "loss": 0.0442,
946
+ "step": 3300
947
+ },
948
+ {
949
+ "epoch": 1.25,
950
+ "grad_norm": 0.5311498641967773,
951
+ "learning_rate": 3.241436925647452e-05,
952
+ "loss": 0.0447,
953
+ "step": 3325
954
+ },
955
+ {
956
+ "epoch": 1.2593984962406015,
957
+ "grad_norm": 0.3183339536190033,
958
+ "learning_rate": 3.2240323029796715e-05,
959
+ "loss": 0.0482,
960
+ "step": 3350
961
+ },
962
+ {
963
+ "epoch": 1.268796992481203,
964
+ "grad_norm": 0.7365193963050842,
965
+ "learning_rate": 3.206627680311891e-05,
966
+ "loss": 0.0441,
967
+ "step": 3375
968
+ },
969
+ {
970
+ "epoch": 1.2781954887218046,
971
+ "grad_norm": 0.5105167627334595,
972
+ "learning_rate": 3.1892230576441104e-05,
973
+ "loss": 0.0421,
974
+ "step": 3400
975
+ },
976
+ {
977
+ "epoch": 1.287593984962406,
978
+ "grad_norm": 0.5017621517181396,
979
+ "learning_rate": 3.1718184349763296e-05,
980
+ "loss": 0.0457,
981
+ "step": 3425
982
+ },
983
+ {
984
+ "epoch": 1.2969924812030076,
985
+ "grad_norm": 0.624539852142334,
986
+ "learning_rate": 3.1544138123085494e-05,
987
+ "loss": 0.0444,
988
+ "step": 3450
989
+ },
990
+ {
991
+ "epoch": 1.306390977443609,
992
+ "grad_norm": 0.44619226455688477,
993
+ "learning_rate": 3.1370091896407685e-05,
994
+ "loss": 0.0422,
995
+ "step": 3475
996
+ },
997
+ {
998
+ "epoch": 1.3157894736842106,
999
+ "grad_norm": 0.7293423414230347,
1000
+ "learning_rate": 3.1196045669729884e-05,
1001
+ "loss": 0.0421,
1002
+ "step": 3500
1003
+ },
1004
+ {
1005
+ "epoch": 1.3251879699248121,
1006
+ "grad_norm": 0.38565269112586975,
1007
+ "learning_rate": 3.1021999443052075e-05,
1008
+ "loss": 0.0443,
1009
+ "step": 3525
1010
+ },
1011
+ {
1012
+ "epoch": 1.3345864661654137,
1013
+ "grad_norm": 1.4548449516296387,
1014
+ "learning_rate": 3.084795321637427e-05,
1015
+ "loss": 0.0449,
1016
+ "step": 3550
1017
+ },
1018
+ {
1019
+ "epoch": 1.3439849624060152,
1020
+ "grad_norm": 0.6648581027984619,
1021
+ "learning_rate": 3.0673906989696465e-05,
1022
+ "loss": 0.0435,
1023
+ "step": 3575
1024
+ },
1025
+ {
1026
+ "epoch": 1.3533834586466165,
1027
+ "grad_norm": 0.3207840919494629,
1028
+ "learning_rate": 3.0499860763018663e-05,
1029
+ "loss": 0.0447,
1030
+ "step": 3600
1031
+ },
1032
+ {
1033
+ "epoch": 1.362781954887218,
1034
+ "grad_norm": 0.7378904223442078,
1035
+ "learning_rate": 3.032581453634085e-05,
1036
+ "loss": 0.0427,
1037
+ "step": 3625
1038
+ },
1039
+ {
1040
+ "epoch": 1.3721804511278195,
1041
+ "grad_norm": 0.3309585452079773,
1042
+ "learning_rate": 3.015176830966305e-05,
1043
+ "loss": 0.0421,
1044
+ "step": 3650
1045
+ },
1046
+ {
1047
+ "epoch": 1.381578947368421,
1048
+ "grad_norm": 0.8541386723518372,
1049
+ "learning_rate": 2.9977722082985244e-05,
1050
+ "loss": 0.041,
1051
+ "step": 3675
1052
+ },
1053
+ {
1054
+ "epoch": 1.3909774436090225,
1055
+ "grad_norm": 0.5450439453125,
1056
+ "learning_rate": 2.980367585630744e-05,
1057
+ "loss": 0.044,
1058
+ "step": 3700
1059
+ },
1060
+ {
1061
+ "epoch": 1.400375939849624,
1062
+ "grad_norm": 0.5193237662315369,
1063
+ "learning_rate": 2.962962962962963e-05,
1064
+ "loss": 0.0433,
1065
+ "step": 3725
1066
+ },
1067
+ {
1068
+ "epoch": 1.4097744360902256,
1069
+ "grad_norm": 0.4268743395805359,
1070
+ "learning_rate": 2.9455583402951825e-05,
1071
+ "loss": 0.0457,
1072
+ "step": 3750
1073
+ },
1074
+ {
1075
+ "epoch": 1.419172932330827,
1076
+ "grad_norm": 0.34615474939346313,
1077
+ "learning_rate": 2.928153717627402e-05,
1078
+ "loss": 0.043,
1079
+ "step": 3775
1080
+ },
1081
+ {
1082
+ "epoch": 1.4285714285714286,
1083
+ "grad_norm": 0.9105768203735352,
1084
+ "learning_rate": 2.9107490949596218e-05,
1085
+ "loss": 0.0418,
1086
+ "step": 3800
1087
+ },
1088
+ {
1089
+ "epoch": 1.4379699248120301,
1090
+ "grad_norm": 0.513663113117218,
1091
+ "learning_rate": 2.8933444722918406e-05,
1092
+ "loss": 0.0429,
1093
+ "step": 3825
1094
+ },
1095
+ {
1096
+ "epoch": 1.4473684210526316,
1097
+ "grad_norm": 0.44397780299186707,
1098
+ "learning_rate": 2.8759398496240604e-05,
1099
+ "loss": 0.0436,
1100
+ "step": 3850
1101
+ },
1102
+ {
1103
+ "epoch": 1.4567669172932332,
1104
+ "grad_norm": 0.3850671052932739,
1105
+ "learning_rate": 2.85853522695628e-05,
1106
+ "loss": 0.0431,
1107
+ "step": 3875
1108
+ },
1109
+ {
1110
+ "epoch": 1.4661654135338344,
1111
+ "grad_norm": 1.0517234802246094,
1112
+ "learning_rate": 2.8411306042884993e-05,
1113
+ "loss": 0.0447,
1114
+ "step": 3900
1115
+ },
1116
+ {
1117
+ "epoch": 1.475563909774436,
1118
+ "grad_norm": 0.7015588879585266,
1119
+ "learning_rate": 2.8237259816207185e-05,
1120
+ "loss": 0.0411,
1121
+ "step": 3925
1122
+ },
1123
+ {
1124
+ "epoch": 1.4849624060150375,
1125
+ "grad_norm": 0.8636859655380249,
1126
+ "learning_rate": 2.806321358952938e-05,
1127
+ "loss": 0.0452,
1128
+ "step": 3950
1129
+ },
1130
+ {
1131
+ "epoch": 1.494360902255639,
1132
+ "grad_norm": 0.3483155071735382,
1133
+ "learning_rate": 2.7889167362851574e-05,
1134
+ "loss": 0.0429,
1135
+ "step": 3975
1136
+ },
1137
+ {
1138
+ "epoch": 1.5037593984962405,
1139
+ "grad_norm": 0.48932698369026184,
1140
+ "learning_rate": 2.7715121136173773e-05,
1141
+ "loss": 0.043,
1142
+ "step": 4000
1143
+ },
1144
+ {
1145
+ "epoch": 1.513157894736842,
1146
+ "grad_norm": 0.40781259536743164,
1147
+ "learning_rate": 2.754107490949596e-05,
1148
+ "loss": 0.0444,
1149
+ "step": 4025
1150
+ },
1151
+ {
1152
+ "epoch": 1.5225563909774436,
1153
+ "grad_norm": 0.3367515802383423,
1154
+ "learning_rate": 2.7373990531885268e-05,
1155
+ "loss": 0.0435,
1156
+ "step": 4050
1157
+ },
1158
+ {
1159
+ "epoch": 1.531954887218045,
1160
+ "grad_norm": 0.28375545144081116,
1161
+ "learning_rate": 2.7199944305207466e-05,
1162
+ "loss": 0.0455,
1163
+ "step": 4075
1164
+ },
1165
+ {
1166
+ "epoch": 1.5413533834586466,
1167
+ "grad_norm": 0.5643731355667114,
1168
+ "learning_rate": 2.702589807852966e-05,
1169
+ "loss": 0.0457,
1170
+ "step": 4100
1171
+ },
1172
+ {
1173
+ "epoch": 1.550751879699248,
1174
+ "grad_norm": 0.6724001169204712,
1175
+ "learning_rate": 2.6851851851851855e-05,
1176
+ "loss": 0.0427,
1177
+ "step": 4125
1178
+ },
1179
+ {
1180
+ "epoch": 1.5601503759398496,
1181
+ "grad_norm": 1.0073463916778564,
1182
+ "learning_rate": 2.6677805625174047e-05,
1183
+ "loss": 0.0431,
1184
+ "step": 4150
1185
+ },
1186
+ {
1187
+ "epoch": 1.5695488721804511,
1188
+ "grad_norm": 0.8705630302429199,
1189
+ "learning_rate": 2.650375939849624e-05,
1190
+ "loss": 0.0435,
1191
+ "step": 4175
1192
+ },
1193
+ {
1194
+ "epoch": 1.5789473684210527,
1195
+ "grad_norm": 1.0504727363586426,
1196
+ "learning_rate": 2.6329713171818436e-05,
1197
+ "loss": 0.0429,
1198
+ "step": 4200
1199
+ },
1200
+ {
1201
+ "epoch": 1.5883458646616542,
1202
+ "grad_norm": 0.5434118509292603,
1203
+ "learning_rate": 2.6155666945140635e-05,
1204
+ "loss": 0.0422,
1205
+ "step": 4225
1206
+ },
1207
+ {
1208
+ "epoch": 1.5977443609022557,
1209
+ "grad_norm": 0.7207650542259216,
1210
+ "learning_rate": 2.5981620718462823e-05,
1211
+ "loss": 0.0433,
1212
+ "step": 4250
1213
+ },
1214
+ {
1215
+ "epoch": 1.6071428571428572,
1216
+ "grad_norm": 0.5877324938774109,
1217
+ "learning_rate": 2.580757449178502e-05,
1218
+ "loss": 0.0396,
1219
+ "step": 4275
1220
+ },
1221
+ {
1222
+ "epoch": 1.6165413533834587,
1223
+ "grad_norm": 0.47331565618515015,
1224
+ "learning_rate": 2.5633528265107216e-05,
1225
+ "loss": 0.045,
1226
+ "step": 4300
1227
+ },
1228
+ {
1229
+ "epoch": 1.6259398496240602,
1230
+ "grad_norm": 0.5089242458343506,
1231
+ "learning_rate": 2.545948203842941e-05,
1232
+ "loss": 0.0433,
1233
+ "step": 4325
1234
+ },
1235
+ {
1236
+ "epoch": 1.6353383458646618,
1237
+ "grad_norm": 0.2361469566822052,
1238
+ "learning_rate": 2.5285435811751602e-05,
1239
+ "loss": 0.0442,
1240
+ "step": 4350
1241
+ },
1242
+ {
1243
+ "epoch": 1.6447368421052633,
1244
+ "grad_norm": 0.3792252242565155,
1245
+ "learning_rate": 2.5111389585073797e-05,
1246
+ "loss": 0.0368,
1247
+ "step": 4375
1248
+ },
1249
+ {
1250
+ "epoch": 1.6541353383458648,
1251
+ "grad_norm": 0.7262583374977112,
1252
+ "learning_rate": 2.493734335839599e-05,
1253
+ "loss": 0.0432,
1254
+ "step": 4400
1255
+ },
1256
+ {
1257
+ "epoch": 1.6635338345864663,
1258
+ "grad_norm": 0.8837150931358337,
1259
+ "learning_rate": 2.4763297131718186e-05,
1260
+ "loss": 0.0456,
1261
+ "step": 4425
1262
+ },
1263
+ {
1264
+ "epoch": 1.6729323308270678,
1265
+ "grad_norm": 0.8227829337120056,
1266
+ "learning_rate": 2.4589250905040378e-05,
1267
+ "loss": 0.0438,
1268
+ "step": 4450
1269
+ },
1270
+ {
1271
+ "epoch": 1.6823308270676691,
1272
+ "grad_norm": 0.4166032075881958,
1273
+ "learning_rate": 2.4415204678362576e-05,
1274
+ "loss": 0.044,
1275
+ "step": 4475
1276
+ },
1277
+ {
1278
+ "epoch": 1.6917293233082706,
1279
+ "grad_norm": 0.5923305153846741,
1280
+ "learning_rate": 2.4241158451684767e-05,
1281
+ "loss": 0.0427,
1282
+ "step": 4500
1283
+ },
1284
+ {
1285
+ "epoch": 1.7011278195488722,
1286
+ "grad_norm": 0.6795745491981506,
1287
+ "learning_rate": 2.4067112225006962e-05,
1288
+ "loss": 0.0426,
1289
+ "step": 4525
1290
+ },
1291
+ {
1292
+ "epoch": 1.7105263157894737,
1293
+ "grad_norm": 0.33346447348594666,
1294
+ "learning_rate": 2.3893065998329157e-05,
1295
+ "loss": 0.0479,
1296
+ "step": 4550
1297
+ },
1298
+ {
1299
+ "epoch": 1.7199248120300752,
1300
+ "grad_norm": 0.43838903307914734,
1301
+ "learning_rate": 2.371901977165135e-05,
1302
+ "loss": 0.0407,
1303
+ "step": 4575
1304
+ },
1305
+ {
1306
+ "epoch": 1.7293233082706767,
1307
+ "grad_norm": 0.3115391433238983,
1308
+ "learning_rate": 2.3544973544973546e-05,
1309
+ "loss": 0.0399,
1310
+ "step": 4600
1311
+ },
1312
+ {
1313
+ "epoch": 1.7387218045112782,
1314
+ "grad_norm": 0.5280188322067261,
1315
+ "learning_rate": 2.337092731829574e-05,
1316
+ "loss": 0.0401,
1317
+ "step": 4625
1318
+ },
1319
+ {
1320
+ "epoch": 1.7481203007518797,
1321
+ "grad_norm": 0.38246890902519226,
1322
+ "learning_rate": 2.3196881091617932e-05,
1323
+ "loss": 0.0425,
1324
+ "step": 4650
1325
+ },
1326
+ {
1327
+ "epoch": 1.7575187969924813,
1328
+ "grad_norm": 0.7593240737915039,
1329
+ "learning_rate": 2.302283486494013e-05,
1330
+ "loss": 0.0414,
1331
+ "step": 4675
1332
+ },
1333
+ {
1334
+ "epoch": 1.7669172932330826,
1335
+ "grad_norm": 0.352885901927948,
1336
+ "learning_rate": 2.2848788638262322e-05,
1337
+ "loss": 0.0423,
1338
+ "step": 4700
1339
+ },
1340
+ {
1341
+ "epoch": 1.776315789473684,
1342
+ "grad_norm": 0.8081568479537964,
1343
+ "learning_rate": 2.2674742411584517e-05,
1344
+ "loss": 0.0452,
1345
+ "step": 4725
1346
+ },
1347
+ {
1348
+ "epoch": 1.7857142857142856,
1349
+ "grad_norm": 0.5408481359481812,
1350
+ "learning_rate": 2.250069618490671e-05,
1351
+ "loss": 0.0409,
1352
+ "step": 4750
1353
+ },
1354
+ {
1355
+ "epoch": 1.795112781954887,
1356
+ "grad_norm": 0.3985776901245117,
1357
+ "learning_rate": 2.2326649958228906e-05,
1358
+ "loss": 0.0422,
1359
+ "step": 4775
1360
+ },
1361
+ {
1362
+ "epoch": 1.8045112781954886,
1363
+ "grad_norm": 0.7956819534301758,
1364
+ "learning_rate": 2.21526037315511e-05,
1365
+ "loss": 0.0477,
1366
+ "step": 4800
1367
+ },
1368
+ {
1369
+ "epoch": 1.8139097744360901,
1370
+ "grad_norm": 0.6655077934265137,
1371
+ "learning_rate": 2.1978557504873296e-05,
1372
+ "loss": 0.0389,
1373
+ "step": 4825
1374
+ },
1375
+ {
1376
+ "epoch": 1.8233082706766917,
1377
+ "grad_norm": 0.20908576250076294,
1378
+ "learning_rate": 2.1804511278195487e-05,
1379
+ "loss": 0.0395,
1380
+ "step": 4850
1381
+ },
1382
+ {
1383
+ "epoch": 1.8327067669172932,
1384
+ "grad_norm": 0.4973180592060089,
1385
+ "learning_rate": 2.1630465051517686e-05,
1386
+ "loss": 0.0393,
1387
+ "step": 4875
1388
+ },
1389
+ {
1390
+ "epoch": 1.8421052631578947,
1391
+ "grad_norm": 0.30976220965385437,
1392
+ "learning_rate": 2.1456418824839877e-05,
1393
+ "loss": 0.0412,
1394
+ "step": 4900
1395
+ },
1396
+ {
1397
+ "epoch": 1.8515037593984962,
1398
+ "grad_norm": 0.4793301820755005,
1399
+ "learning_rate": 2.1282372598162072e-05,
1400
+ "loss": 0.0409,
1401
+ "step": 4925
1402
+ },
1403
+ {
1404
+ "epoch": 1.8609022556390977,
1405
+ "grad_norm": 1.1762813329696655,
1406
+ "learning_rate": 2.1108326371484267e-05,
1407
+ "loss": 0.045,
1408
+ "step": 4950
1409
+ },
1410
+ {
1411
+ "epoch": 1.8703007518796992,
1412
+ "grad_norm": 0.35434260964393616,
1413
+ "learning_rate": 2.093428014480646e-05,
1414
+ "loss": 0.044,
1415
+ "step": 4975
1416
+ },
1417
+ {
1418
+ "epoch": 1.8796992481203008,
1419
+ "grad_norm": 0.6554225087165833,
1420
+ "learning_rate": 2.0760233918128656e-05,
1421
+ "loss": 0.0422,
1422
+ "step": 5000
1423
+ },
1424
+ {
1425
+ "epoch": 1.8890977443609023,
1426
+ "grad_norm": 0.7061398029327393,
1427
+ "learning_rate": 2.058618769145085e-05,
1428
+ "loss": 0.0431,
1429
+ "step": 5025
1430
+ },
1431
+ {
1432
+ "epoch": 1.8984962406015038,
1433
+ "grad_norm": 0.4826093912124634,
1434
+ "learning_rate": 2.0412141464773042e-05,
1435
+ "loss": 0.0427,
1436
+ "step": 5050
1437
+ },
1438
+ {
1439
+ "epoch": 1.9078947368421053,
1440
+ "grad_norm": 0.48956793546676636,
1441
+ "learning_rate": 2.023809523809524e-05,
1442
+ "loss": 0.042,
1443
+ "step": 5075
1444
+ },
1445
+ {
1446
+ "epoch": 1.9172932330827068,
1447
+ "grad_norm": 0.7553461790084839,
1448
+ "learning_rate": 2.0064049011417432e-05,
1449
+ "loss": 0.0396,
1450
+ "step": 5100
1451
+ },
1452
+ {
1453
+ "epoch": 1.9266917293233083,
1454
+ "grad_norm": 0.727511465549469,
1455
+ "learning_rate": 1.9890002784739627e-05,
1456
+ "loss": 0.0438,
1457
+ "step": 5125
1458
+ },
1459
+ {
1460
+ "epoch": 1.9360902255639099,
1461
+ "grad_norm": 0.3488887548446655,
1462
+ "learning_rate": 1.971595655806182e-05,
1463
+ "loss": 0.0455,
1464
+ "step": 5150
1465
+ },
1466
+ {
1467
+ "epoch": 1.9454887218045114,
1468
+ "grad_norm": 0.8386423587799072,
1469
+ "learning_rate": 1.9541910331384016e-05,
1470
+ "loss": 0.0446,
1471
+ "step": 5175
1472
+ },
1473
+ {
1474
+ "epoch": 1.954887218045113,
1475
+ "grad_norm": 0.368587464094162,
1476
+ "learning_rate": 1.936786410470621e-05,
1477
+ "loss": 0.0453,
1478
+ "step": 5200
1479
+ },
1480
+ {
1481
+ "epoch": 1.9642857142857144,
1482
+ "grad_norm": 0.49407923221588135,
1483
+ "learning_rate": 1.9193817878028406e-05,
1484
+ "loss": 0.0398,
1485
+ "step": 5225
1486
+ },
1487
+ {
1488
+ "epoch": 1.973684210526316,
1489
+ "grad_norm": 0.4838390350341797,
1490
+ "learning_rate": 1.9019771651350597e-05,
1491
+ "loss": 0.0441,
1492
+ "step": 5250
1493
+ },
1494
+ {
1495
+ "epoch": 1.9830827067669174,
1496
+ "grad_norm": 1.1204516887664795,
1497
+ "learning_rate": 1.8845725424672795e-05,
1498
+ "loss": 0.0423,
1499
+ "step": 5275
1500
+ },
1501
+ {
1502
+ "epoch": 1.9924812030075187,
1503
+ "grad_norm": 0.7833461761474609,
1504
+ "learning_rate": 1.8671679197994987e-05,
1505
+ "loss": 0.0447,
1506
+ "step": 5300
1507
+ },
1508
+ {
1509
+ "epoch": 2.0,
1510
+ "eval_explained_variance": 0.39946532249450684,
1511
+ "eval_loss": 0.043054722249507904,
1512
+ "eval_mae": 0.1624627560377121,
1513
+ "eval_mse": 0.04305747523903847,
1514
+ "eval_r2": 0.3960953950881958,
1515
+ "eval_rmse": 0.20750295236222174,
1516
+ "eval_runtime": 4.9991,
1517
+ "eval_samples_per_second": 4256.156,
1518
+ "eval_steps_per_second": 66.612,
1519
+ "step": 5320
1520
+ },
1521
+ {
1522
+ "epoch": 2.0018796992481205,
1523
+ "grad_norm": 0.3862787187099457,
1524
+ "learning_rate": 1.849763297131718e-05,
1525
+ "loss": 0.0413,
1526
+ "step": 5325
1527
+ },
1528
+ {
1529
+ "epoch": 2.011278195488722,
1530
+ "grad_norm": 0.370114803314209,
1531
+ "learning_rate": 1.8323586744639376e-05,
1532
+ "loss": 0.038,
1533
+ "step": 5350
1534
+ },
1535
+ {
1536
+ "epoch": 2.0206766917293235,
1537
+ "grad_norm": 0.4231257140636444,
1538
+ "learning_rate": 1.814954051796157e-05,
1539
+ "loss": 0.0413,
1540
+ "step": 5375
1541
+ },
1542
+ {
1543
+ "epoch": 2.030075187969925,
1544
+ "grad_norm": 0.3307948112487793,
1545
+ "learning_rate": 1.7975494291283766e-05,
1546
+ "loss": 0.0356,
1547
+ "step": 5400
1548
+ },
1549
+ {
1550
+ "epoch": 2.039473684210526,
1551
+ "grad_norm": 0.8401979207992554,
1552
+ "learning_rate": 1.780144806460596e-05,
1553
+ "loss": 0.036,
1554
+ "step": 5425
1555
+ },
1556
+ {
1557
+ "epoch": 2.0488721804511276,
1558
+ "grad_norm": 0.29018813371658325,
1559
+ "learning_rate": 1.7627401837928152e-05,
1560
+ "loss": 0.0369,
1561
+ "step": 5450
1562
+ },
1563
+ {
1564
+ "epoch": 2.058270676691729,
1565
+ "grad_norm": 0.3946692645549774,
1566
+ "learning_rate": 1.745335561125035e-05,
1567
+ "loss": 0.0398,
1568
+ "step": 5475
1569
+ },
1570
+ {
1571
+ "epoch": 2.0676691729323307,
1572
+ "grad_norm": 0.3740139901638031,
1573
+ "learning_rate": 1.7279309384572542e-05,
1574
+ "loss": 0.0388,
1575
+ "step": 5500
1576
+ },
1577
+ {
1578
+ "epoch": 2.077067669172932,
1579
+ "grad_norm": 0.48486819863319397,
1580
+ "learning_rate": 1.7105263157894737e-05,
1581
+ "loss": 0.0389,
1582
+ "step": 5525
1583
+ },
1584
+ {
1585
+ "epoch": 2.0864661654135337,
1586
+ "grad_norm": 0.3575657606124878,
1587
+ "learning_rate": 1.693121693121693e-05,
1588
+ "loss": 0.0393,
1589
+ "step": 5550
1590
+ },
1591
+ {
1592
+ "epoch": 2.095864661654135,
1593
+ "grad_norm": 0.3213801980018616,
1594
+ "learning_rate": 1.6757170704539126e-05,
1595
+ "loss": 0.0343,
1596
+ "step": 5575
1597
+ },
1598
+ {
1599
+ "epoch": 2.1052631578947367,
1600
+ "grad_norm": 0.7583568096160889,
1601
+ "learning_rate": 1.658312447786132e-05,
1602
+ "loss": 0.0352,
1603
+ "step": 5600
1604
+ },
1605
+ {
1606
+ "epoch": 2.1146616541353382,
1607
+ "grad_norm": 0.5068955421447754,
1608
+ "learning_rate": 1.6409078251183516e-05,
1609
+ "loss": 0.0345,
1610
+ "step": 5625
1611
+ },
1612
+ {
1613
+ "epoch": 2.1240601503759398,
1614
+ "grad_norm": 1.060349464416504,
1615
+ "learning_rate": 1.6235032024505707e-05,
1616
+ "loss": 0.0349,
1617
+ "step": 5650
1618
+ },
1619
+ {
1620
+ "epoch": 2.1334586466165413,
1621
+ "grad_norm": 0.5841746926307678,
1622
+ "learning_rate": 1.6060985797827905e-05,
1623
+ "loss": 0.0369,
1624
+ "step": 5675
1625
+ },
1626
+ {
1627
+ "epoch": 2.142857142857143,
1628
+ "grad_norm": 0.3654855489730835,
1629
+ "learning_rate": 1.5886939571150097e-05,
1630
+ "loss": 0.0336,
1631
+ "step": 5700
1632
+ },
1633
+ {
1634
+ "epoch": 2.1522556390977443,
1635
+ "grad_norm": 0.4263927638530731,
1636
+ "learning_rate": 1.571289334447229e-05,
1637
+ "loss": 0.0346,
1638
+ "step": 5725
1639
+ },
1640
+ {
1641
+ "epoch": 2.161654135338346,
1642
+ "grad_norm": 0.4506601393222809,
1643
+ "learning_rate": 1.5538847117794486e-05,
1644
+ "loss": 0.036,
1645
+ "step": 5750
1646
+ },
1647
+ {
1648
+ "epoch": 2.1710526315789473,
1649
+ "grad_norm": 0.4509446620941162,
1650
+ "learning_rate": 1.536480089111668e-05,
1651
+ "loss": 0.0362,
1652
+ "step": 5775
1653
+ },
1654
+ {
1655
+ "epoch": 2.180451127819549,
1656
+ "grad_norm": 0.3250133693218231,
1657
+ "learning_rate": 1.5190754664438874e-05,
1658
+ "loss": 0.0368,
1659
+ "step": 5800
1660
+ },
1661
+ {
1662
+ "epoch": 2.1898496240601504,
1663
+ "grad_norm": 0.3531028926372528,
1664
+ "learning_rate": 1.501670843776107e-05,
1665
+ "loss": 0.0382,
1666
+ "step": 5825
1667
+ },
1668
+ {
1669
+ "epoch": 2.199248120300752,
1670
+ "grad_norm": 0.6282513737678528,
1671
+ "learning_rate": 1.4842662211083264e-05,
1672
+ "loss": 0.0357,
1673
+ "step": 5850
1674
+ },
1675
+ {
1676
+ "epoch": 2.2086466165413534,
1677
+ "grad_norm": 0.5511112213134766,
1678
+ "learning_rate": 1.4668615984405459e-05,
1679
+ "loss": 0.0363,
1680
+ "step": 5875
1681
+ },
1682
+ {
1683
+ "epoch": 2.218045112781955,
1684
+ "grad_norm": 0.3617405593395233,
1685
+ "learning_rate": 1.4494569757727652e-05,
1686
+ "loss": 0.0355,
1687
+ "step": 5900
1688
+ },
1689
+ {
1690
+ "epoch": 2.2274436090225564,
1691
+ "grad_norm": 0.6280185580253601,
1692
+ "learning_rate": 1.4320523531049848e-05,
1693
+ "loss": 0.0417,
1694
+ "step": 5925
1695
+ },
1696
+ {
1697
+ "epoch": 2.236842105263158,
1698
+ "grad_norm": 0.2536335289478302,
1699
+ "learning_rate": 1.4146477304372041e-05,
1700
+ "loss": 0.0352,
1701
+ "step": 5950
1702
+ },
1703
+ {
1704
+ "epoch": 2.2462406015037595,
1705
+ "grad_norm": 0.8867663145065308,
1706
+ "learning_rate": 1.3972431077694236e-05,
1707
+ "loss": 0.04,
1708
+ "step": 5975
1709
+ },
1710
+ {
1711
+ "epoch": 2.255639097744361,
1712
+ "grad_norm": 0.4762154817581177,
1713
+ "learning_rate": 1.3798384851016429e-05,
1714
+ "loss": 0.0354,
1715
+ "step": 6000
1716
+ },
1717
+ {
1718
+ "epoch": 2.2650375939849625,
1719
+ "grad_norm": 0.42669475078582764,
1720
+ "learning_rate": 1.3624338624338626e-05,
1721
+ "loss": 0.0379,
1722
+ "step": 6025
1723
+ },
1724
+ {
1725
+ "epoch": 2.274436090225564,
1726
+ "grad_norm": 0.8416607975959778,
1727
+ "learning_rate": 1.3450292397660819e-05,
1728
+ "loss": 0.0388,
1729
+ "step": 6050
1730
+ },
1731
+ {
1732
+ "epoch": 2.2838345864661656,
1733
+ "grad_norm": 0.6854439973831177,
1734
+ "learning_rate": 1.3276246170983015e-05,
1735
+ "loss": 0.0376,
1736
+ "step": 6075
1737
+ },
1738
+ {
1739
+ "epoch": 2.293233082706767,
1740
+ "grad_norm": 0.4139537215232849,
1741
+ "learning_rate": 1.3102199944305207e-05,
1742
+ "loss": 0.038,
1743
+ "step": 6100
1744
+ },
1745
+ {
1746
+ "epoch": 2.3026315789473686,
1747
+ "grad_norm": 0.4523138105869293,
1748
+ "learning_rate": 1.2928153717627403e-05,
1749
+ "loss": 0.0342,
1750
+ "step": 6125
1751
+ },
1752
+ {
1753
+ "epoch": 2.31203007518797,
1754
+ "grad_norm": 0.4618607759475708,
1755
+ "learning_rate": 1.2754107490949596e-05,
1756
+ "loss": 0.035,
1757
+ "step": 6150
1758
+ },
1759
+ {
1760
+ "epoch": 2.3214285714285716,
1761
+ "grad_norm": 0.6621114015579224,
1762
+ "learning_rate": 1.2580061264271793e-05,
1763
+ "loss": 0.034,
1764
+ "step": 6175
1765
+ },
1766
+ {
1767
+ "epoch": 2.3308270676691727,
1768
+ "grad_norm": 0.3603108823299408,
1769
+ "learning_rate": 1.2406015037593984e-05,
1770
+ "loss": 0.0383,
1771
+ "step": 6200
1772
+ },
1773
+ {
1774
+ "epoch": 2.340225563909774,
1775
+ "grad_norm": 0.6309170722961426,
1776
+ "learning_rate": 1.2231968810916179e-05,
1777
+ "loss": 0.0372,
1778
+ "step": 6225
1779
+ },
1780
+ {
1781
+ "epoch": 2.3496240601503757,
1782
+ "grad_norm": 0.34498706459999084,
1783
+ "learning_rate": 1.2057922584238374e-05,
1784
+ "loss": 0.0425,
1785
+ "step": 6250
1786
+ },
1787
+ {
1788
+ "epoch": 2.3590225563909772,
1789
+ "grad_norm": 0.514720618724823,
1790
+ "learning_rate": 1.189083820662768e-05,
1791
+ "loss": 0.0361,
1792
+ "step": 6275
1793
+ },
1794
+ {
1795
+ "epoch": 2.3684210526315788,
1796
+ "grad_norm": 0.7199122309684753,
1797
+ "learning_rate": 1.1716791979949874e-05,
1798
+ "loss": 0.0383,
1799
+ "step": 6300
1800
+ },
1801
+ {
1802
+ "epoch": 2.3778195488721803,
1803
+ "grad_norm": 0.4565185308456421,
1804
+ "learning_rate": 1.1542745753272069e-05,
1805
+ "loss": 0.0355,
1806
+ "step": 6325
1807
+ },
1808
+ {
1809
+ "epoch": 2.387218045112782,
1810
+ "grad_norm": 0.5844483375549316,
1811
+ "learning_rate": 1.1368699526594263e-05,
1812
+ "loss": 0.0373,
1813
+ "step": 6350
1814
+ },
1815
+ {
1816
+ "epoch": 2.3966165413533833,
1817
+ "grad_norm": 0.5359569787979126,
1818
+ "learning_rate": 1.1194653299916458e-05,
1819
+ "loss": 0.0402,
1820
+ "step": 6375
1821
+ },
1822
+ {
1823
+ "epoch": 2.406015037593985,
1824
+ "grad_norm": 0.4015677571296692,
1825
+ "learning_rate": 1.1020607073238653e-05,
1826
+ "loss": 0.0346,
1827
+ "step": 6400
1828
+ },
1829
+ {
1830
+ "epoch": 2.4154135338345863,
1831
+ "grad_norm": 0.4379020929336548,
1832
+ "learning_rate": 1.0846560846560846e-05,
1833
+ "loss": 0.0369,
1834
+ "step": 6425
1835
+ },
1836
+ {
1837
+ "epoch": 2.424812030075188,
1838
+ "grad_norm": 0.8200315237045288,
1839
+ "learning_rate": 1.067251461988304e-05,
1840
+ "loss": 0.0347,
1841
+ "step": 6450
1842
+ },
1843
+ {
1844
+ "epoch": 2.4342105263157894,
1845
+ "grad_norm": 0.5679715275764465,
1846
+ "learning_rate": 1.0498468393205236e-05,
1847
+ "loss": 0.0402,
1848
+ "step": 6475
1849
+ },
1850
+ {
1851
+ "epoch": 2.443609022556391,
1852
+ "grad_norm": 0.40348678827285767,
1853
+ "learning_rate": 1.032442216652743e-05,
1854
+ "loss": 0.0377,
1855
+ "step": 6500
1856
+ },
1857
+ {
1858
+ "epoch": 2.4530075187969924,
1859
+ "grad_norm": 0.4095970094203949,
1860
+ "learning_rate": 1.0150375939849624e-05,
1861
+ "loss": 0.0404,
1862
+ "step": 6525
1863
+ },
1864
+ {
1865
+ "epoch": 2.462406015037594,
1866
+ "grad_norm": 0.8567390441894531,
1867
+ "learning_rate": 9.976329713171818e-06,
1868
+ "loss": 0.0383,
1869
+ "step": 6550
1870
+ },
1871
+ {
1872
+ "epoch": 2.4718045112781954,
1873
+ "grad_norm": 0.6875966191291809,
1874
+ "learning_rate": 9.802283486494013e-06,
1875
+ "loss": 0.0367,
1876
+ "step": 6575
1877
+ },
1878
+ {
1879
+ "epoch": 2.481203007518797,
1880
+ "grad_norm": 0.5457414984703064,
1881
+ "learning_rate": 9.628237259816208e-06,
1882
+ "loss": 0.0371,
1883
+ "step": 6600
1884
+ },
1885
+ {
1886
+ "epoch": 2.4906015037593985,
1887
+ "grad_norm": 0.4299066364765167,
1888
+ "learning_rate": 9.454191033138401e-06,
1889
+ "loss": 0.0342,
1890
+ "step": 6625
1891
+ },
1892
+ {
1893
+ "epoch": 2.5,
1894
+ "grad_norm": 1.1978002786636353,
1895
+ "learning_rate": 9.280144806460596e-06,
1896
+ "loss": 0.0363,
1897
+ "step": 6650
1898
+ },
1899
+ {
1900
+ "epoch": 2.5093984962406015,
1901
+ "grad_norm": 0.43589282035827637,
1902
+ "learning_rate": 9.10609857978279e-06,
1903
+ "loss": 0.0338,
1904
+ "step": 6675
1905
+ },
1906
+ {
1907
+ "epoch": 2.518796992481203,
1908
+ "grad_norm": 0.32333287596702576,
1909
+ "learning_rate": 8.932052353104985e-06,
1910
+ "loss": 0.0369,
1911
+ "step": 6700
1912
+ },
1913
+ {
1914
+ "epoch": 2.5281954887218046,
1915
+ "grad_norm": 0.40547892451286316,
1916
+ "learning_rate": 8.758006126427178e-06,
1917
+ "loss": 0.0359,
1918
+ "step": 6725
1919
+ },
1920
+ {
1921
+ "epoch": 2.537593984962406,
1922
+ "grad_norm": 0.6290844678878784,
1923
+ "learning_rate": 8.583959899749373e-06,
1924
+ "loss": 0.0364,
1925
+ "step": 6750
1926
+ },
1927
+ {
1928
+ "epoch": 2.5469924812030076,
1929
+ "grad_norm": 0.3206610679626465,
1930
+ "learning_rate": 8.409913673071568e-06,
1931
+ "loss": 0.0359,
1932
+ "step": 6775
1933
+ },
1934
+ {
1935
+ "epoch": 2.556390977443609,
1936
+ "grad_norm": 0.4280617833137512,
1937
+ "learning_rate": 8.235867446393763e-06,
1938
+ "loss": 0.0365,
1939
+ "step": 6800
1940
+ },
1941
+ {
1942
+ "epoch": 2.5657894736842106,
1943
+ "grad_norm": 0.448256254196167,
1944
+ "learning_rate": 8.061821219715956e-06,
1945
+ "loss": 0.0374,
1946
+ "step": 6825
1947
+ },
1948
+ {
1949
+ "epoch": 2.575187969924812,
1950
+ "grad_norm": 0.6079933643341064,
1951
+ "learning_rate": 7.88777499303815e-06,
1952
+ "loss": 0.0327,
1953
+ "step": 6850
1954
+ },
1955
+ {
1956
+ "epoch": 2.5845864661654137,
1957
+ "grad_norm": 0.47237610816955566,
1958
+ "learning_rate": 7.713728766360345e-06,
1959
+ "loss": 0.0367,
1960
+ "step": 6875
1961
+ },
1962
+ {
1963
+ "epoch": 2.593984962406015,
1964
+ "grad_norm": 0.6032050848007202,
1965
+ "learning_rate": 7.5396825396825394e-06,
1966
+ "loss": 0.0363,
1967
+ "step": 6900
1968
+ },
1969
+ {
1970
+ "epoch": 2.6033834586466167,
1971
+ "grad_norm": 0.38197141885757446,
1972
+ "learning_rate": 7.365636313004734e-06,
1973
+ "loss": 0.0319,
1974
+ "step": 6925
1975
+ },
1976
+ {
1977
+ "epoch": 2.612781954887218,
1978
+ "grad_norm": 0.4339030683040619,
1979
+ "learning_rate": 7.191590086326928e-06,
1980
+ "loss": 0.0355,
1981
+ "step": 6950
1982
+ },
1983
+ {
1984
+ "epoch": 2.6221804511278197,
1985
+ "grad_norm": 0.45415690541267395,
1986
+ "learning_rate": 7.017543859649123e-06,
1987
+ "loss": 0.035,
1988
+ "step": 6975
1989
+ },
1990
+ {
1991
+ "epoch": 2.6315789473684212,
1992
+ "grad_norm": 0.32464611530303955,
1993
+ "learning_rate": 6.843497632971317e-06,
1994
+ "loss": 0.0372,
1995
+ "step": 7000
1996
+ },
1997
+ {
1998
+ "epoch": 2.6409774436090228,
1999
+ "grad_norm": 0.7759618163108826,
2000
+ "learning_rate": 6.669451406293512e-06,
2001
+ "loss": 0.0355,
2002
+ "step": 7025
2003
+ },
2004
+ {
2005
+ "epoch": 2.6503759398496243,
2006
+ "grad_norm": 0.5577199459075928,
2007
+ "learning_rate": 6.495405179615706e-06,
2008
+ "loss": 0.0336,
2009
+ "step": 7050
2010
+ },
2011
+ {
2012
+ "epoch": 2.659774436090226,
2013
+ "grad_norm": 0.555182158946991,
2014
+ "learning_rate": 6.3213589529379004e-06,
2015
+ "loss": 0.0355,
2016
+ "step": 7075
2017
+ },
2018
+ {
2019
+ "epoch": 2.6691729323308273,
2020
+ "grad_norm": 0.2945663332939148,
2021
+ "learning_rate": 6.147312726260094e-06,
2022
+ "loss": 0.0363,
2023
+ "step": 7100
2024
+ },
2025
+ {
2026
+ "epoch": 2.678571428571429,
2027
+ "grad_norm": 0.3848724365234375,
2028
+ "learning_rate": 5.973266499582289e-06,
2029
+ "loss": 0.0346,
2030
+ "step": 7125
2031
+ },
2032
+ {
2033
+ "epoch": 2.6879699248120303,
2034
+ "grad_norm": 0.35878947377204895,
2035
+ "learning_rate": 5.799220272904483e-06,
2036
+ "loss": 0.0343,
2037
+ "step": 7150
2038
+ },
2039
+ {
2040
+ "epoch": 2.6973684210526314,
2041
+ "grad_norm": 0.32243430614471436,
2042
+ "learning_rate": 5.625174046226678e-06,
2043
+ "loss": 0.0355,
2044
+ "step": 7175
2045
+ },
2046
+ {
2047
+ "epoch": 2.706766917293233,
2048
+ "grad_norm": 0.5378155708312988,
2049
+ "learning_rate": 5.451127819548872e-06,
2050
+ "loss": 0.0351,
2051
+ "step": 7200
2052
+ },
2053
+ {
2054
+ "epoch": 2.7161654135338344,
2055
+ "grad_norm": 0.3866325914859772,
2056
+ "learning_rate": 5.277081592871067e-06,
2057
+ "loss": 0.0341,
2058
+ "step": 7225
2059
+ },
2060
+ {
2061
+ "epoch": 2.725563909774436,
2062
+ "grad_norm": 0.6423678398132324,
2063
+ "learning_rate": 5.103035366193261e-06,
2064
+ "loss": 0.0359,
2065
+ "step": 7250
2066
+ },
2067
+ {
2068
+ "epoch": 2.7349624060150375,
2069
+ "grad_norm": 0.47325319051742554,
2070
+ "learning_rate": 4.928989139515455e-06,
2071
+ "loss": 0.0339,
2072
+ "step": 7275
2073
+ },
2074
+ {
2075
+ "epoch": 2.744360902255639,
2076
+ "grad_norm": 0.3928850293159485,
2077
+ "learning_rate": 4.754942912837649e-06,
2078
+ "loss": 0.0346,
2079
+ "step": 7300
2080
+ },
2081
+ {
2082
+ "epoch": 2.7537593984962405,
2083
+ "grad_norm": 0.44239577651023865,
2084
+ "learning_rate": 4.580896686159844e-06,
2085
+ "loss": 0.0374,
2086
+ "step": 7325
2087
+ },
2088
+ {
2089
+ "epoch": 2.763157894736842,
2090
+ "grad_norm": 0.7063366770744324,
2091
+ "learning_rate": 4.406850459482038e-06,
2092
+ "loss": 0.0339,
2093
+ "step": 7350
2094
+ },
2095
+ {
2096
+ "epoch": 2.7725563909774436,
2097
+ "grad_norm": 0.7401504516601562,
2098
+ "learning_rate": 4.232804232804233e-06,
2099
+ "loss": 0.0342,
2100
+ "step": 7375
2101
+ },
2102
+ {
2103
+ "epoch": 2.781954887218045,
2104
+ "grad_norm": 0.4666846692562103,
2105
+ "learning_rate": 4.058758006126427e-06,
2106
+ "loss": 0.0343,
2107
+ "step": 7400
2108
+ },
2109
+ {
2110
+ "epoch": 2.7913533834586466,
2111
+ "grad_norm": 0.38399186730384827,
2112
+ "learning_rate": 3.884711779448622e-06,
2113
+ "loss": 0.0325,
2114
+ "step": 7425
2115
+ },
2116
+ {
2117
+ "epoch": 2.800751879699248,
2118
+ "grad_norm": 0.7592440843582153,
2119
+ "learning_rate": 3.710665552770816e-06,
2120
+ "loss": 0.0366,
2121
+ "step": 7450
2122
+ },
2123
+ {
2124
+ "epoch": 2.8101503759398496,
2125
+ "grad_norm": 0.4485366642475128,
2126
+ "learning_rate": 3.5366193260930103e-06,
2127
+ "loss": 0.0343,
2128
+ "step": 7475
2129
+ },
2130
+ {
2131
+ "epoch": 2.819548872180451,
2132
+ "grad_norm": 0.34349316358566284,
2133
+ "learning_rate": 3.3625730994152047e-06,
2134
+ "loss": 0.0351,
2135
+ "step": 7500
2136
+ },
2137
+ {
2138
+ "epoch": 2.8289473684210527,
2139
+ "grad_norm": 0.4510953724384308,
2140
+ "learning_rate": 3.188526872737399e-06,
2141
+ "loss": 0.0357,
2142
+ "step": 7525
2143
+ },
2144
+ {
2145
+ "epoch": 2.838345864661654,
2146
+ "grad_norm": 0.35561850666999817,
2147
+ "learning_rate": 3.0144806460595934e-06,
2148
+ "loss": 0.0352,
2149
+ "step": 7550
2150
+ },
2151
+ {
2152
+ "epoch": 2.8477443609022557,
2153
+ "grad_norm": 0.7706106901168823,
2154
+ "learning_rate": 2.8404344193817878e-06,
2155
+ "loss": 0.0326,
2156
+ "step": 7575
2157
+ },
2158
+ {
2159
+ "epoch": 2.857142857142857,
2160
+ "grad_norm": 0.34894415736198425,
2161
+ "learning_rate": 2.666388192703982e-06,
2162
+ "loss": 0.0307,
2163
+ "step": 7600
2164
+ },
2165
+ {
2166
+ "epoch": 2.8665413533834587,
2167
+ "grad_norm": 0.42608681321144104,
2168
+ "learning_rate": 2.4923419660261765e-06,
2169
+ "loss": 0.0364,
2170
+ "step": 7625
2171
+ },
2172
+ {
2173
+ "epoch": 2.8759398496240602,
2174
+ "grad_norm": 0.30355915427207947,
2175
+ "learning_rate": 2.318295739348371e-06,
2176
+ "loss": 0.0366,
2177
+ "step": 7650
2178
+ },
2179
+ {
2180
+ "epoch": 2.8853383458646618,
2181
+ "grad_norm": 0.39785754680633545,
2182
+ "learning_rate": 2.1442495126705653e-06,
2183
+ "loss": 0.0314,
2184
+ "step": 7675
2185
+ },
2186
+ {
2187
+ "epoch": 2.8947368421052633,
2188
+ "grad_norm": 0.7486430406570435,
2189
+ "learning_rate": 1.9702032859927596e-06,
2190
+ "loss": 0.0403,
2191
+ "step": 7700
2192
+ },
2193
+ {
2194
+ "epoch": 2.904135338345865,
2195
+ "grad_norm": 0.5167568325996399,
2196
+ "learning_rate": 1.796157059314954e-06,
2197
+ "loss": 0.0346,
2198
+ "step": 7725
2199
+ },
2200
+ {
2201
+ "epoch": 2.9135338345864663,
2202
+ "grad_norm": 0.5651251673698425,
2203
+ "learning_rate": 1.6221108326371484e-06,
2204
+ "loss": 0.0378,
2205
+ "step": 7750
2206
+ },
2207
+ {
2208
+ "epoch": 2.922932330827068,
2209
+ "grad_norm": 1.125868558883667,
2210
+ "learning_rate": 1.4480646059593427e-06,
2211
+ "loss": 0.0333,
2212
+ "step": 7775
2213
+ },
2214
+ {
2215
+ "epoch": 2.932330827067669,
2216
+ "grad_norm": 0.33022329211235046,
2217
+ "learning_rate": 1.274018379281537e-06,
2218
+ "loss": 0.0367,
2219
+ "step": 7800
2220
+ },
2221
+ {
2222
+ "epoch": 2.9417293233082704,
2223
+ "grad_norm": 0.5090142488479614,
2224
+ "learning_rate": 1.0999721526037315e-06,
2225
+ "loss": 0.0376,
2226
+ "step": 7825
2227
+ },
2228
+ {
2229
+ "epoch": 2.951127819548872,
2230
+ "grad_norm": 0.5005149245262146,
2231
+ "learning_rate": 9.259259259259259e-07,
2232
+ "loss": 0.035,
2233
+ "step": 7850
2234
+ },
2235
+ {
2236
+ "epoch": 2.9605263157894735,
2237
+ "grad_norm": 0.36888259649276733,
2238
+ "learning_rate": 7.518796992481203e-07,
2239
+ "loss": 0.0378,
2240
+ "step": 7875
2241
+ },
2242
+ {
2243
+ "epoch": 2.969924812030075,
2244
+ "grad_norm": 0.7692381143569946,
2245
+ "learning_rate": 5.778334725703147e-07,
2246
+ "loss": 0.0354,
2247
+ "step": 7900
2248
+ },
2249
+ {
2250
+ "epoch": 2.9793233082706765,
2251
+ "grad_norm": 0.6987473368644714,
2252
+ "learning_rate": 4.0378724589250904e-07,
2253
+ "loss": 0.0364,
2254
+ "step": 7925
2255
+ },
2256
+ {
2257
+ "epoch": 2.988721804511278,
2258
+ "grad_norm": 0.2583921253681183,
2259
+ "learning_rate": 2.2974101921470343e-07,
2260
+ "loss": 0.0328,
2261
+ "step": 7950
2262
+ },
2263
+ {
2264
+ "epoch": 2.9981203007518795,
2265
+ "grad_norm": 0.7963603138923645,
2266
+ "learning_rate": 5.56947925368978e-08,
2267
+ "loss": 0.036,
2268
+ "step": 7975
2269
+ },
2270
+ {
2271
+ "epoch": 3.0,
2272
+ "eval_explained_variance": 0.4178752303123474,
2273
+ "eval_loss": 0.041510969400405884,
2274
+ "eval_mae": 0.15856829285621643,
2275
+ "eval_mse": 0.041513968259096146,
2276
+ "eval_r2": 0.4177439212799072,
2277
+ "eval_rmse": 0.203749768733847,
2278
+ "eval_runtime": 4.9888,
2279
+ "eval_samples_per_second": 4264.915,
2280
+ "eval_steps_per_second": 66.749,
2281
+ "step": 7980
2282
+ }
2283
+ ],
2284
+ "logging_steps": 25,
2285
+ "max_steps": 7980,
2286
+ "num_input_tokens_seen": 0,
2287
+ "num_train_epochs": 3,
2288
+ "save_steps": 500,
2289
+ "stateful_callbacks": {
2290
+ "EarlyStoppingCallback": {
2291
+ "args": {
2292
+ "early_stopping_patience": 5,
2293
+ "early_stopping_threshold": 0.01
2294
+ },
2295
+ "attributes": {
2296
+ "early_stopping_patience_counter": 2
2297
+ }
2298
+ },
2299
+ "TrainerControl": {
2300
+ "args": {
2301
+ "should_epoch_stop": false,
2302
+ "should_evaluate": false,
2303
+ "should_log": false,
2304
+ "should_save": true,
2305
+ "should_training_stop": true
2306
+ },
2307
+ "attributes": {}
2308
+ }
2309
+ },
2310
+ "total_flos": 8456567720509440.0,
2311
+ "train_batch_size": 8,
2312
+ "trial_name": null,
2313
+ "trial_params": null
2314
+ }
checkpoint-7980/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e69c4b367fc6ac1ffd520d1ae037a8dae530773484f821180d998ab759b7694
3
+ size 5368
config.json ADDED
@@ -0,0 +1,32 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "distilbert/distilbert-base-uncased",
3
+ "_num_labels": 1,
4
+ "activation": "gelu",
5
+ "architectures": [
6
+ "DistilBertForSequenceClassification"
7
+ ],
8
+ "attention_dropout": 0.1,
9
+ "dim": 768,
10
+ "dropout": 0.1,
11
+ "hidden_dim": 3072,
12
+ "id2label": {
13
+ "0": "target"
14
+ },
15
+ "initializer_range": 0.02,
16
+ "label2id": {
17
+ "target": 0
18
+ },
19
+ "max_position_embeddings": 512,
20
+ "model_type": "distilbert",
21
+ "n_heads": 12,
22
+ "n_layers": 6,
23
+ "pad_token_id": 0,
24
+ "problem_type": "regression",
25
+ "qa_dropout": 0.1,
26
+ "seq_classif_dropout": 0.2,
27
+ "sinusoidal_pos_embds": false,
28
+ "tie_weights_": true,
29
+ "torch_dtype": "float32",
30
+ "transformers_version": "4.48.0",
31
+ "vocab_size": 30522
32
+ }
model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8ac3b1117a2c16efdaca4209bef8b2d615e37e4f1241391aaa60d1e8463e9c67
3
+ size 267829484
runs/Feb22_18-48-22_r-samtuckervegan-autotrain-4xl4-br7fljr5-341de-o890j/events.out.tfevents.1740250104.r-samtuckervegan-autotrain-4xl4-br7fljr5-341de-o890j.202.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:a1ea85588080cdd98dcf9f28c2439857e98b3cfaa8beb0c068b6b64cac5267e7
3
- size 58605
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8b48f8e108584bc8910e5886aa425895eadc426969d2c683336af73e558b1d74
3
+ size 74250
runs/Feb22_18-48-22_r-samtuckervegan-autotrain-4xl4-br7fljr5-341de-o890j/events.out.tfevents.1740250884.r-samtuckervegan-autotrain-4xl4-br7fljr5-341de-o890j.202.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fb891959994827f3a2a7284194ae5b978637f488a93cfd2078c2c4e29b2f755f
3
+ size 609
special_tokens_map.json ADDED
@@ -0,0 +1,7 @@
 
 
 
 
 
 
 
 
1
+ {
2
+ "cls_token": "[CLS]",
3
+ "mask_token": "[MASK]",
4
+ "pad_token": "[PAD]",
5
+ "sep_token": "[SEP]",
6
+ "unk_token": "[UNK]"
7
+ }
tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "[PAD]",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "100": {
12
+ "content": "[UNK]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "101": {
20
+ "content": "[CLS]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "102": {
28
+ "content": "[SEP]",
29
+ "lstrip": false,
30
+ "normalized": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "103": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "clean_up_tokenization_spaces": false,
45
+ "cls_token": "[CLS]",
46
+ "do_lower_case": true,
47
+ "extra_special_tokens": {},
48
+ "mask_token": "[MASK]",
49
+ "model_max_length": 512,
50
+ "pad_token": "[PAD]",
51
+ "sep_token": "[SEP]",
52
+ "strip_accents": null,
53
+ "tokenize_chinese_chars": true,
54
+ "tokenizer_class": "DistilBertTokenizer",
55
+ "unk_token": "[UNK]"
56
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e69c4b367fc6ac1ffd520d1ae037a8dae530773484f821180d998ab759b7694
3
+ size 5368
training_params.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "data_path": "samtuckervegan/alignment",
3
+ "model": "distilbert/distilbert-base-uncased",
4
+ "lr": 5e-05,
5
+ "epochs": 3,
6
+ "max_seq_length": 128,
7
+ "batch_size": 8,
8
+ "warmup_ratio": 0.1,
9
+ "gradient_accumulation": 1,
10
+ "optimizer": "adamw_torch",
11
+ "scheduler": "linear",
12
+ "weight_decay": 0.0,
13
+ "max_grad_norm": 1.0,
14
+ "seed": 42,
15
+ "train_split": "train",
16
+ "valid_split": "test",
17
+ "text_column": "text",
18
+ "target_column": "target",
19
+ "logging_steps": -1,
20
+ "project_name": "alignment",
21
+ "auto_find_batch_size": false,
22
+ "mixed_precision": "fp16",
23
+ "save_total_limit": 1,
24
+ "push_to_hub": true,
25
+ "eval_strategy": "epoch",
26
+ "username": "samtuckervegan",
27
+ "log": "tensorboard",
28
+ "early_stopping_patience": 5,
29
+ "early_stopping_threshold": 0.01
30
+ }
vocab.txt ADDED
The diff for this file is too large to render. See raw diff