Hugging Face's logo

open-paws
/
potential_influence_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
4124496
·
verified ·
1 Parent(s): aa17fb3

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_11-28-18_r-samtuckervegan-autotrain-4xl4-hcv9q8jy-d0a1c-qe6mi/events.out.tfevents.1740223708.r-samtuckervegan-autotrain-4xl4-hcv9q8jy-d0a1c-qe6mi.201.0 +2 -2
  15. runs/Feb22_11-28-18_r-samtuckervegan-autotrain-4xl4-hcv9q8jy-d0a1c-qe6mi/events.out.tfevents.1740224486.r-samtuckervegan-autotrain-4xl4-hcv9q8jy-d0a1c-qe6mi.201.1 +3 -0
  16. special_tokens_map.json +7 -0
  17. tokenizer.json +0 -0
  18. tokenizer_config.json +58 -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/influence_predictions
11
+ ---
12
+
13
+ # Model Trained Using AutoTrain
14
+
15
+ - Problem type: Text Regression
16
+
17
+ ## Validation Metrics
18
+ loss: 0.020958250388503075
19
+
20
+ mse: 0.020960183814167976
21
+
22
+ mae: 0.11261985450983047
23
+
24
+ r2: 0.2767665982246399
25
+
26
+ rmse: 0.14477632338945473
27
+
28
+ explained_variance: 0.27679795026779175
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:f3d20966de2b80c0d322c8f91327db3506896c23f3500ba59b9622b8c8c5f710
3
+ size 267829484
checkpoint-7980/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:34ccd0afe7cfdddc55b891af6c44aba348679fb17dedfdedad7a9d16c5271566
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:eb5579603207b2ba457bde9d6f1147a9744be739db8bfae8e4d1961b19c70826
3
+ size 1064
checkpoint-7980/trainer_state.json ADDED
@@ -0,0 +1,2314 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": 0.020958250388503075,
3
+ "best_model_checkpoint": "influence/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": 5.7797369956970215,
14
+ "learning_rate": 1.5037593984962406e-06,
15
+ "loss": 0.3753,
16
+ "step": 25
17
+ },
18
+ {
19
+ "epoch": 0.018796992481203006,
20
+ "grad_norm": 1.0250223875045776,
21
+ "learning_rate": 3.070175438596491e-06,
22
+ "loss": 0.1755,
23
+ "step": 50
24
+ },
25
+ {
26
+ "epoch": 0.02819548872180451,
27
+ "grad_norm": 0.7308247089385986,
28
+ "learning_rate": 4.636591478696742e-06,
29
+ "loss": 0.0429,
30
+ "step": 75
31
+ },
32
+ {
33
+ "epoch": 0.03759398496240601,
34
+ "grad_norm": 0.5842682123184204,
35
+ "learning_rate": 6.203007518796992e-06,
36
+ "loss": 0.0324,
37
+ "step": 100
38
+ },
39
+ {
40
+ "epoch": 0.046992481203007516,
41
+ "grad_norm": 0.4021165668964386,
42
+ "learning_rate": 7.769423558897243e-06,
43
+ "loss": 0.0301,
44
+ "step": 125
45
+ },
46
+ {
47
+ "epoch": 0.05639097744360902,
48
+ "grad_norm": 0.4142129123210907,
49
+ "learning_rate": 9.335839598997493e-06,
50
+ "loss": 0.0311,
51
+ "step": 150
52
+ },
53
+ {
54
+ "epoch": 0.06578947368421052,
55
+ "grad_norm": 0.5423107147216797,
56
+ "learning_rate": 1.0902255639097744e-05,
57
+ "loss": 0.0309,
58
+ "step": 175
59
+ },
60
+ {
61
+ "epoch": 0.07518796992481203,
62
+ "grad_norm": 0.461919367313385,
63
+ "learning_rate": 1.2468671679197996e-05,
64
+ "loss": 0.03,
65
+ "step": 200
66
+ },
67
+ {
68
+ "epoch": 0.08458646616541353,
69
+ "grad_norm": 0.681239128112793,
70
+ "learning_rate": 1.4035087719298246e-05,
71
+ "loss": 0.0289,
72
+ "step": 225
73
+ },
74
+ {
75
+ "epoch": 0.09398496240601503,
76
+ "grad_norm": 0.5302960872650146,
77
+ "learning_rate": 1.5601503759398496e-05,
78
+ "loss": 0.0323,
79
+ "step": 250
80
+ },
81
+ {
82
+ "epoch": 0.10338345864661654,
83
+ "grad_norm": 0.5830396413803101,
84
+ "learning_rate": 1.7167919799498746e-05,
85
+ "loss": 0.029,
86
+ "step": 275
87
+ },
88
+ {
89
+ "epoch": 0.11278195488721804,
90
+ "grad_norm": 1.3836510181427002,
91
+ "learning_rate": 1.8734335839599e-05,
92
+ "loss": 0.0283,
93
+ "step": 300
94
+ },
95
+ {
96
+ "epoch": 0.12218045112781954,
97
+ "grad_norm": 0.3139019012451172,
98
+ "learning_rate": 2.0300751879699247e-05,
99
+ "loss": 0.0302,
100
+ "step": 325
101
+ },
102
+ {
103
+ "epoch": 0.13157894736842105,
104
+ "grad_norm": 0.3362056314945221,
105
+ "learning_rate": 2.18671679197995e-05,
106
+ "loss": 0.0292,
107
+ "step": 350
108
+ },
109
+ {
110
+ "epoch": 0.14097744360902256,
111
+ "grad_norm": 0.2954612076282501,
112
+ "learning_rate": 2.3433583959899748e-05,
113
+ "loss": 0.0284,
114
+ "step": 375
115
+ },
116
+ {
117
+ "epoch": 0.15037593984962405,
118
+ "grad_norm": 0.716823160648346,
119
+ "learning_rate": 2.5e-05,
120
+ "loss": 0.0278,
121
+ "step": 400
122
+ },
123
+ {
124
+ "epoch": 0.15977443609022557,
125
+ "grad_norm": 0.4281720817089081,
126
+ "learning_rate": 2.656641604010025e-05,
127
+ "loss": 0.0276,
128
+ "step": 425
129
+ },
130
+ {
131
+ "epoch": 0.16917293233082706,
132
+ "grad_norm": 1.0022855997085571,
133
+ "learning_rate": 2.8132832080200505e-05,
134
+ "loss": 0.0274,
135
+ "step": 450
136
+ },
137
+ {
138
+ "epoch": 0.17857142857142858,
139
+ "grad_norm": 0.30904313921928406,
140
+ "learning_rate": 2.9699248120300755e-05,
141
+ "loss": 0.0272,
142
+ "step": 475
143
+ },
144
+ {
145
+ "epoch": 0.18796992481203006,
146
+ "grad_norm": 0.31030890345573425,
147
+ "learning_rate": 3.1265664160401006e-05,
148
+ "loss": 0.0266,
149
+ "step": 500
150
+ },
151
+ {
152
+ "epoch": 0.19736842105263158,
153
+ "grad_norm": 0.21925821900367737,
154
+ "learning_rate": 3.2832080200501256e-05,
155
+ "loss": 0.0254,
156
+ "step": 525
157
+ },
158
+ {
159
+ "epoch": 0.20676691729323307,
160
+ "grad_norm": 1.1428364515304565,
161
+ "learning_rate": 3.4398496240601506e-05,
162
+ "loss": 0.0267,
163
+ "step": 550
164
+ },
165
+ {
166
+ "epoch": 0.2161654135338346,
167
+ "grad_norm": 1.0609831809997559,
168
+ "learning_rate": 3.5964912280701756e-05,
169
+ "loss": 0.0269,
170
+ "step": 575
171
+ },
172
+ {
173
+ "epoch": 0.22556390977443608,
174
+ "grad_norm": 0.31294840574264526,
175
+ "learning_rate": 3.753132832080201e-05,
176
+ "loss": 0.0307,
177
+ "step": 600
178
+ },
179
+ {
180
+ "epoch": 0.2349624060150376,
181
+ "grad_norm": 0.2312650978565216,
182
+ "learning_rate": 3.909774436090226e-05,
183
+ "loss": 0.0275,
184
+ "step": 625
185
+ },
186
+ {
187
+ "epoch": 0.24436090225563908,
188
+ "grad_norm": 0.21682415902614594,
189
+ "learning_rate": 4.066416040100251e-05,
190
+ "loss": 0.0256,
191
+ "step": 650
192
+ },
193
+ {
194
+ "epoch": 0.25375939849624063,
195
+ "grad_norm": 0.9164804816246033,
196
+ "learning_rate": 4.223057644110276e-05,
197
+ "loss": 0.0252,
198
+ "step": 675
199
+ },
200
+ {
201
+ "epoch": 0.2631578947368421,
202
+ "grad_norm": 0.337128221988678,
203
+ "learning_rate": 4.379699248120301e-05,
204
+ "loss": 0.0271,
205
+ "step": 700
206
+ },
207
+ {
208
+ "epoch": 0.2725563909774436,
209
+ "grad_norm": 0.28385016322135925,
210
+ "learning_rate": 4.536340852130326e-05,
211
+ "loss": 0.0251,
212
+ "step": 725
213
+ },
214
+ {
215
+ "epoch": 0.2819548872180451,
216
+ "grad_norm": 0.3594600260257721,
217
+ "learning_rate": 4.6929824561403515e-05,
218
+ "loss": 0.026,
219
+ "step": 750
220
+ },
221
+ {
222
+ "epoch": 0.29135338345864664,
223
+ "grad_norm": 0.732537031173706,
224
+ "learning_rate": 4.849624060150376e-05,
225
+ "loss": 0.0285,
226
+ "step": 775
227
+ },
228
+ {
229
+ "epoch": 0.3007518796992481,
230
+ "grad_norm": 0.5434914231300354,
231
+ "learning_rate": 4.9993038150932894e-05,
232
+ "loss": 0.0298,
233
+ "step": 800
234
+ },
235
+ {
236
+ "epoch": 0.3101503759398496,
237
+ "grad_norm": 0.22048309445381165,
238
+ "learning_rate": 4.9818991924255085e-05,
239
+ "loss": 0.0255,
240
+ "step": 825
241
+ },
242
+ {
243
+ "epoch": 0.31954887218045114,
244
+ "grad_norm": 0.932379961013794,
245
+ "learning_rate": 4.9644945697577276e-05,
246
+ "loss": 0.0251,
247
+ "step": 850
248
+ },
249
+ {
250
+ "epoch": 0.32894736842105265,
251
+ "grad_norm": 0.16935257613658905,
252
+ "learning_rate": 4.9470899470899475e-05,
253
+ "loss": 0.0263,
254
+ "step": 875
255
+ },
256
+ {
257
+ "epoch": 0.3383458646616541,
258
+ "grad_norm": 0.36494505405426025,
259
+ "learning_rate": 4.9296853244221666e-05,
260
+ "loss": 0.0255,
261
+ "step": 900
262
+ },
263
+ {
264
+ "epoch": 0.34774436090225563,
265
+ "grad_norm": 0.19008353352546692,
266
+ "learning_rate": 4.912280701754386e-05,
267
+ "loss": 0.028,
268
+ "step": 925
269
+ },
270
+ {
271
+ "epoch": 0.35714285714285715,
272
+ "grad_norm": 0.7813118100166321,
273
+ "learning_rate": 4.8948760790866056e-05,
274
+ "loss": 0.0276,
275
+ "step": 950
276
+ },
277
+ {
278
+ "epoch": 0.36654135338345867,
279
+ "grad_norm": 0.20266500115394592,
280
+ "learning_rate": 4.8774714564188254e-05,
281
+ "loss": 0.0263,
282
+ "step": 975
283
+ },
284
+ {
285
+ "epoch": 0.37593984962406013,
286
+ "grad_norm": 0.2332334965467453,
287
+ "learning_rate": 4.8600668337510445e-05,
288
+ "loss": 0.0267,
289
+ "step": 1000
290
+ },
291
+ {
292
+ "epoch": 0.38533834586466165,
293
+ "grad_norm": 0.1838051974773407,
294
+ "learning_rate": 4.8426622110832637e-05,
295
+ "loss": 0.0231,
296
+ "step": 1025
297
+ },
298
+ {
299
+ "epoch": 0.39473684210526316,
300
+ "grad_norm": 0.449964314699173,
301
+ "learning_rate": 4.8252575884154835e-05,
302
+ "loss": 0.0252,
303
+ "step": 1050
304
+ },
305
+ {
306
+ "epoch": 0.4041353383458647,
307
+ "grad_norm": 0.9875504970550537,
308
+ "learning_rate": 4.8078529657477026e-05,
309
+ "loss": 0.0276,
310
+ "step": 1075
311
+ },
312
+ {
313
+ "epoch": 0.41353383458646614,
314
+ "grad_norm": 0.22402499616146088,
315
+ "learning_rate": 4.7904483430799224e-05,
316
+ "loss": 0.0227,
317
+ "step": 1100
318
+ },
319
+ {
320
+ "epoch": 0.42293233082706766,
321
+ "grad_norm": 0.1501668095588684,
322
+ "learning_rate": 4.7730437204121416e-05,
323
+ "loss": 0.0257,
324
+ "step": 1125
325
+ },
326
+ {
327
+ "epoch": 0.4323308270676692,
328
+ "grad_norm": 0.1940675973892212,
329
+ "learning_rate": 4.755639097744361e-05,
330
+ "loss": 0.0288,
331
+ "step": 1150
332
+ },
333
+ {
334
+ "epoch": 0.4417293233082707,
335
+ "grad_norm": 0.4955839514732361,
336
+ "learning_rate": 4.7382344750765805e-05,
337
+ "loss": 0.0242,
338
+ "step": 1175
339
+ },
340
+ {
341
+ "epoch": 0.45112781954887216,
342
+ "grad_norm": 0.1950056552886963,
343
+ "learning_rate": 4.7208298524088003e-05,
344
+ "loss": 0.024,
345
+ "step": 1200
346
+ },
347
+ {
348
+ "epoch": 0.4605263157894737,
349
+ "grad_norm": 0.27604976296424866,
350
+ "learning_rate": 4.7034252297410195e-05,
351
+ "loss": 0.027,
352
+ "step": 1225
353
+ },
354
+ {
355
+ "epoch": 0.4699248120300752,
356
+ "grad_norm": 0.26503273844718933,
357
+ "learning_rate": 4.6860206070732386e-05,
358
+ "loss": 0.0252,
359
+ "step": 1250
360
+ },
361
+ {
362
+ "epoch": 0.4793233082706767,
363
+ "grad_norm": 0.28494516015052795,
364
+ "learning_rate": 4.6686159844054584e-05,
365
+ "loss": 0.0234,
366
+ "step": 1275
367
+ },
368
+ {
369
+ "epoch": 0.48872180451127817,
370
+ "grad_norm": 0.24238109588623047,
371
+ "learning_rate": 4.6512113617376776e-05,
372
+ "loss": 0.0255,
373
+ "step": 1300
374
+ },
375
+ {
376
+ "epoch": 0.4981203007518797,
377
+ "grad_norm": 0.17075982689857483,
378
+ "learning_rate": 4.633806739069897e-05,
379
+ "loss": 0.0255,
380
+ "step": 1325
381
+ },
382
+ {
383
+ "epoch": 0.5075187969924813,
384
+ "grad_norm": 0.3948628008365631,
385
+ "learning_rate": 4.6164021164021165e-05,
386
+ "loss": 0.0245,
387
+ "step": 1350
388
+ },
389
+ {
390
+ "epoch": 0.5169172932330827,
391
+ "grad_norm": 0.14536868035793304,
392
+ "learning_rate": 4.5989974937343364e-05,
393
+ "loss": 0.0243,
394
+ "step": 1375
395
+ },
396
+ {
397
+ "epoch": 0.5263157894736842,
398
+ "grad_norm": 0.23006580770015717,
399
+ "learning_rate": 4.5815928710665555e-05,
400
+ "loss": 0.0257,
401
+ "step": 1400
402
+ },
403
+ {
404
+ "epoch": 0.5357142857142857,
405
+ "grad_norm": 0.1656516045331955,
406
+ "learning_rate": 4.5641882483987746e-05,
407
+ "loss": 0.0229,
408
+ "step": 1425
409
+ },
410
+ {
411
+ "epoch": 0.5451127819548872,
412
+ "grad_norm": 0.2032531201839447,
413
+ "learning_rate": 4.5467836257309945e-05,
414
+ "loss": 0.0228,
415
+ "step": 1450
416
+ },
417
+ {
418
+ "epoch": 0.5545112781954887,
419
+ "grad_norm": 0.31135717034339905,
420
+ "learning_rate": 4.5293790030632136e-05,
421
+ "loss": 0.0232,
422
+ "step": 1475
423
+ },
424
+ {
425
+ "epoch": 0.5639097744360902,
426
+ "grad_norm": 0.32599613070487976,
427
+ "learning_rate": 4.5119743803954334e-05,
428
+ "loss": 0.0249,
429
+ "step": 1500
430
+ },
431
+ {
432
+ "epoch": 0.5733082706766918,
433
+ "grad_norm": 0.35132333636283875,
434
+ "learning_rate": 4.4945697577276526e-05,
435
+ "loss": 0.0247,
436
+ "step": 1525
437
+ },
438
+ {
439
+ "epoch": 0.5827067669172933,
440
+ "grad_norm": 0.3662143349647522,
441
+ "learning_rate": 4.477165135059872e-05,
442
+ "loss": 0.0232,
443
+ "step": 1550
444
+ },
445
+ {
446
+ "epoch": 0.5921052631578947,
447
+ "grad_norm": 0.34423789381980896,
448
+ "learning_rate": 4.4597605123920915e-05,
449
+ "loss": 0.0226,
450
+ "step": 1575
451
+ },
452
+ {
453
+ "epoch": 0.6015037593984962,
454
+ "grad_norm": 0.77556973695755,
455
+ "learning_rate": 4.442355889724311e-05,
456
+ "loss": 0.0251,
457
+ "step": 1600
458
+ },
459
+ {
460
+ "epoch": 0.6109022556390977,
461
+ "grad_norm": 0.19359074532985687,
462
+ "learning_rate": 4.4249512670565305e-05,
463
+ "loss": 0.0243,
464
+ "step": 1625
465
+ },
466
+ {
467
+ "epoch": 0.6203007518796992,
468
+ "grad_norm": 0.36068597435951233,
469
+ "learning_rate": 4.4075466443887496e-05,
470
+ "loss": 0.0226,
471
+ "step": 1650
472
+ },
473
+ {
474
+ "epoch": 0.6296992481203008,
475
+ "grad_norm": 0.21560369431972504,
476
+ "learning_rate": 4.3901420217209694e-05,
477
+ "loss": 0.0262,
478
+ "step": 1675
479
+ },
480
+ {
481
+ "epoch": 0.6390977443609023,
482
+ "grad_norm": 0.24299262464046478,
483
+ "learning_rate": 4.3727373990531886e-05,
484
+ "loss": 0.0242,
485
+ "step": 1700
486
+ },
487
+ {
488
+ "epoch": 0.6484962406015038,
489
+ "grad_norm": 0.2628134787082672,
490
+ "learning_rate": 4.355332776385408e-05,
491
+ "loss": 0.0223,
492
+ "step": 1725
493
+ },
494
+ {
495
+ "epoch": 0.6578947368421053,
496
+ "grad_norm": 0.15418370068073273,
497
+ "learning_rate": 4.3379281537176275e-05,
498
+ "loss": 0.0245,
499
+ "step": 1750
500
+ },
501
+ {
502
+ "epoch": 0.6672932330827067,
503
+ "grad_norm": 0.48042866587638855,
504
+ "learning_rate": 4.3205235310498473e-05,
505
+ "loss": 0.0256,
506
+ "step": 1775
507
+ },
508
+ {
509
+ "epoch": 0.6766917293233082,
510
+ "grad_norm": 0.2339322715997696,
511
+ "learning_rate": 4.3031189083820665e-05,
512
+ "loss": 0.0219,
513
+ "step": 1800
514
+ },
515
+ {
516
+ "epoch": 0.6860902255639098,
517
+ "grad_norm": 0.15647785365581512,
518
+ "learning_rate": 4.2857142857142856e-05,
519
+ "loss": 0.0258,
520
+ "step": 1825
521
+ },
522
+ {
523
+ "epoch": 0.6954887218045113,
524
+ "grad_norm": 0.19495581090450287,
525
+ "learning_rate": 4.2683096630465054e-05,
526
+ "loss": 0.0256,
527
+ "step": 1850
528
+ },
529
+ {
530
+ "epoch": 0.7048872180451128,
531
+ "grad_norm": 0.4958021342754364,
532
+ "learning_rate": 4.2509050403787246e-05,
533
+ "loss": 0.022,
534
+ "step": 1875
535
+ },
536
+ {
537
+ "epoch": 0.7142857142857143,
538
+ "grad_norm": 0.1408185511827469,
539
+ "learning_rate": 4.2335004177109444e-05,
540
+ "loss": 0.0239,
541
+ "step": 1900
542
+ },
543
+ {
544
+ "epoch": 0.7236842105263158,
545
+ "grad_norm": 0.14317536354064941,
546
+ "learning_rate": 4.2160957950431635e-05,
547
+ "loss": 0.0245,
548
+ "step": 1925
549
+ },
550
+ {
551
+ "epoch": 0.7330827067669173,
552
+ "grad_norm": 0.1682003140449524,
553
+ "learning_rate": 4.198691172375383e-05,
554
+ "loss": 0.0246,
555
+ "step": 1950
556
+ },
557
+ {
558
+ "epoch": 0.7424812030075187,
559
+ "grad_norm": 0.28081175684928894,
560
+ "learning_rate": 4.1812865497076025e-05,
561
+ "loss": 0.0218,
562
+ "step": 1975
563
+ },
564
+ {
565
+ "epoch": 0.7518796992481203,
566
+ "grad_norm": 0.24349795281887054,
567
+ "learning_rate": 4.163881927039822e-05,
568
+ "loss": 0.0225,
569
+ "step": 2000
570
+ },
571
+ {
572
+ "epoch": 0.7612781954887218,
573
+ "grad_norm": 0.22174572944641113,
574
+ "learning_rate": 4.1464773043720415e-05,
575
+ "loss": 0.0236,
576
+ "step": 2025
577
+ },
578
+ {
579
+ "epoch": 0.7706766917293233,
580
+ "grad_norm": 0.38060835003852844,
581
+ "learning_rate": 4.1290726817042606e-05,
582
+ "loss": 0.0265,
583
+ "step": 2050
584
+ },
585
+ {
586
+ "epoch": 0.7800751879699248,
587
+ "grad_norm": 0.1448279768228531,
588
+ "learning_rate": 4.1116680590364804e-05,
589
+ "loss": 0.022,
590
+ "step": 2075
591
+ },
592
+ {
593
+ "epoch": 0.7894736842105263,
594
+ "grad_norm": 0.21761861443519592,
595
+ "learning_rate": 4.0942634363687e-05,
596
+ "loss": 0.0239,
597
+ "step": 2100
598
+ },
599
+ {
600
+ "epoch": 0.7988721804511278,
601
+ "grad_norm": 0.176962748169899,
602
+ "learning_rate": 4.076858813700919e-05,
603
+ "loss": 0.0233,
604
+ "step": 2125
605
+ },
606
+ {
607
+ "epoch": 0.8082706766917294,
608
+ "grad_norm": 0.22260421514511108,
609
+ "learning_rate": 4.0594541910331385e-05,
610
+ "loss": 0.0255,
611
+ "step": 2150
612
+ },
613
+ {
614
+ "epoch": 0.8176691729323309,
615
+ "grad_norm": 0.22005359828472137,
616
+ "learning_rate": 4.042049568365358e-05,
617
+ "loss": 0.0219,
618
+ "step": 2175
619
+ },
620
+ {
621
+ "epoch": 0.8270676691729323,
622
+ "grad_norm": 0.18301767110824585,
623
+ "learning_rate": 4.0246449456975775e-05,
624
+ "loss": 0.0237,
625
+ "step": 2200
626
+ },
627
+ {
628
+ "epoch": 0.8364661654135338,
629
+ "grad_norm": 0.5412411093711853,
630
+ "learning_rate": 4.0072403230297966e-05,
631
+ "loss": 0.0237,
632
+ "step": 2225
633
+ },
634
+ {
635
+ "epoch": 0.8458646616541353,
636
+ "grad_norm": 0.10722669214010239,
637
+ "learning_rate": 3.9898357003620164e-05,
638
+ "loss": 0.0208,
639
+ "step": 2250
640
+ },
641
+ {
642
+ "epoch": 0.8552631578947368,
643
+ "grad_norm": 0.1962004154920578,
644
+ "learning_rate": 3.9724310776942356e-05,
645
+ "loss": 0.0217,
646
+ "step": 2275
647
+ },
648
+ {
649
+ "epoch": 0.8646616541353384,
650
+ "grad_norm": 0.14313143491744995,
651
+ "learning_rate": 3.9550264550264554e-05,
652
+ "loss": 0.0234,
653
+ "step": 2300
654
+ },
655
+ {
656
+ "epoch": 0.8740601503759399,
657
+ "grad_norm": 0.6453161239624023,
658
+ "learning_rate": 3.9376218323586745e-05,
659
+ "loss": 0.0224,
660
+ "step": 2325
661
+ },
662
+ {
663
+ "epoch": 0.8834586466165414,
664
+ "grad_norm": 0.14810048043727875,
665
+ "learning_rate": 3.920217209690894e-05,
666
+ "loss": 0.0215,
667
+ "step": 2350
668
+ },
669
+ {
670
+ "epoch": 0.8928571428571429,
671
+ "grad_norm": 0.28511399030685425,
672
+ "learning_rate": 3.9028125870231135e-05,
673
+ "loss": 0.0243,
674
+ "step": 2375
675
+ },
676
+ {
677
+ "epoch": 0.9022556390977443,
678
+ "grad_norm": 0.4954968988895416,
679
+ "learning_rate": 3.885407964355333e-05,
680
+ "loss": 0.0244,
681
+ "step": 2400
682
+ },
683
+ {
684
+ "epoch": 0.9116541353383458,
685
+ "grad_norm": 0.49626636505126953,
686
+ "learning_rate": 3.8680033416875525e-05,
687
+ "loss": 0.0248,
688
+ "step": 2425
689
+ },
690
+ {
691
+ "epoch": 0.9210526315789473,
692
+ "grad_norm": 0.17915596067905426,
693
+ "learning_rate": 3.8505987190197716e-05,
694
+ "loss": 0.0227,
695
+ "step": 2450
696
+ },
697
+ {
698
+ "epoch": 0.9304511278195489,
699
+ "grad_norm": 0.2660089135169983,
700
+ "learning_rate": 3.8331940963519914e-05,
701
+ "loss": 0.0236,
702
+ "step": 2475
703
+ },
704
+ {
705
+ "epoch": 0.9398496240601504,
706
+ "grad_norm": 0.8929455280303955,
707
+ "learning_rate": 3.815789473684211e-05,
708
+ "loss": 0.0222,
709
+ "step": 2500
710
+ },
711
+ {
712
+ "epoch": 0.9492481203007519,
713
+ "grad_norm": 0.23705103993415833,
714
+ "learning_rate": 3.79838485101643e-05,
715
+ "loss": 0.0219,
716
+ "step": 2525
717
+ },
718
+ {
719
+ "epoch": 0.9586466165413534,
720
+ "grad_norm": 0.7505210638046265,
721
+ "learning_rate": 3.7809802283486495e-05,
722
+ "loss": 0.0242,
723
+ "step": 2550
724
+ },
725
+ {
726
+ "epoch": 0.9680451127819549,
727
+ "grad_norm": 0.36811527609825134,
728
+ "learning_rate": 3.763575605680869e-05,
729
+ "loss": 0.0237,
730
+ "step": 2575
731
+ },
732
+ {
733
+ "epoch": 0.9774436090225563,
734
+ "grad_norm": 0.21303361654281616,
735
+ "learning_rate": 3.7461709830130885e-05,
736
+ "loss": 0.0252,
737
+ "step": 2600
738
+ },
739
+ {
740
+ "epoch": 0.9868421052631579,
741
+ "grad_norm": 0.2701171338558197,
742
+ "learning_rate": 3.7287663603453076e-05,
743
+ "loss": 0.0226,
744
+ "step": 2625
745
+ },
746
+ {
747
+ "epoch": 0.9962406015037594,
748
+ "grad_norm": 0.6065410375595093,
749
+ "learning_rate": 3.7113617376775274e-05,
750
+ "loss": 0.0226,
751
+ "step": 2650
752
+ },
753
+ {
754
+ "epoch": 1.0,
755
+ "eval_explained_variance": 0.22629410028457642,
756
+ "eval_loss": 0.022498182952404022,
757
+ "eval_mae": 0.11727148294448853,
758
+ "eval_mse": 0.022499235346913338,
759
+ "eval_r2": 0.2236614227294922,
760
+ "eval_rmse": 0.14999745113472207,
761
+ "eval_runtime": 5.0695,
762
+ "eval_samples_per_second": 4197.068,
763
+ "eval_steps_per_second": 65.687,
764
+ "step": 2660
765
+ },
766
+ {
767
+ "epoch": 1.005639097744361,
768
+ "grad_norm": 0.16667713224887848,
769
+ "learning_rate": 3.6939571150097466e-05,
770
+ "loss": 0.0208,
771
+ "step": 2675
772
+ },
773
+ {
774
+ "epoch": 1.0150375939849625,
775
+ "grad_norm": 0.1851395219564438,
776
+ "learning_rate": 3.6765524923419664e-05,
777
+ "loss": 0.0198,
778
+ "step": 2700
779
+ },
780
+ {
781
+ "epoch": 1.0244360902255638,
782
+ "grad_norm": 0.5289347767829895,
783
+ "learning_rate": 3.6591478696741855e-05,
784
+ "loss": 0.0199,
785
+ "step": 2725
786
+ },
787
+ {
788
+ "epoch": 1.0338345864661653,
789
+ "grad_norm": 0.37565699219703674,
790
+ "learning_rate": 3.6417432470064053e-05,
791
+ "loss": 0.0242,
792
+ "step": 2750
793
+ },
794
+ {
795
+ "epoch": 1.0432330827067668,
796
+ "grad_norm": 0.6762616634368896,
797
+ "learning_rate": 3.6243386243386245e-05,
798
+ "loss": 0.0225,
799
+ "step": 2775
800
+ },
801
+ {
802
+ "epoch": 1.0526315789473684,
803
+ "grad_norm": 0.37910592555999756,
804
+ "learning_rate": 3.606934001670844e-05,
805
+ "loss": 0.0234,
806
+ "step": 2800
807
+ },
808
+ {
809
+ "epoch": 1.0620300751879699,
810
+ "grad_norm": 0.7098792195320129,
811
+ "learning_rate": 3.5895293790030634e-05,
812
+ "loss": 0.0232,
813
+ "step": 2825
814
+ },
815
+ {
816
+ "epoch": 1.0714285714285714,
817
+ "grad_norm": 0.1560579240322113,
818
+ "learning_rate": 3.5721247563352826e-05,
819
+ "loss": 0.0218,
820
+ "step": 2850
821
+ },
822
+ {
823
+ "epoch": 1.080827067669173,
824
+ "grad_norm": 0.2022002935409546,
825
+ "learning_rate": 3.5547201336675024e-05,
826
+ "loss": 0.0211,
827
+ "step": 2875
828
+ },
829
+ {
830
+ "epoch": 1.0902255639097744,
831
+ "grad_norm": 0.1768932193517685,
832
+ "learning_rate": 3.537315510999722e-05,
833
+ "loss": 0.0231,
834
+ "step": 2900
835
+ },
836
+ {
837
+ "epoch": 1.099624060150376,
838
+ "grad_norm": 0.1871190220117569,
839
+ "learning_rate": 3.519910888331941e-05,
840
+ "loss": 0.0223,
841
+ "step": 2925
842
+ },
843
+ {
844
+ "epoch": 1.1090225563909775,
845
+ "grad_norm": 0.16333234310150146,
846
+ "learning_rate": 3.5025062656641605e-05,
847
+ "loss": 0.0213,
848
+ "step": 2950
849
+ },
850
+ {
851
+ "epoch": 1.118421052631579,
852
+ "grad_norm": 0.43317675590515137,
853
+ "learning_rate": 3.48510164299638e-05,
854
+ "loss": 0.0245,
855
+ "step": 2975
856
+ },
857
+ {
858
+ "epoch": 1.1278195488721805,
859
+ "grad_norm": 0.18057814240455627,
860
+ "learning_rate": 3.4676970203285995e-05,
861
+ "loss": 0.0217,
862
+ "step": 3000
863
+ },
864
+ {
865
+ "epoch": 1.137218045112782,
866
+ "grad_norm": 0.3941543698310852,
867
+ "learning_rate": 3.4502923976608186e-05,
868
+ "loss": 0.0233,
869
+ "step": 3025
870
+ },
871
+ {
872
+ "epoch": 1.1466165413533835,
873
+ "grad_norm": 0.16782347857952118,
874
+ "learning_rate": 3.4328877749930384e-05,
875
+ "loss": 0.0227,
876
+ "step": 3050
877
+ },
878
+ {
879
+ "epoch": 1.156015037593985,
880
+ "grad_norm": 0.1606685072183609,
881
+ "learning_rate": 3.4154831523252576e-05,
882
+ "loss": 0.0212,
883
+ "step": 3075
884
+ },
885
+ {
886
+ "epoch": 1.1654135338345863,
887
+ "grad_norm": 0.13007329404354095,
888
+ "learning_rate": 3.3980785296574774e-05,
889
+ "loss": 0.023,
890
+ "step": 3100
891
+ },
892
+ {
893
+ "epoch": 1.1748120300751879,
894
+ "grad_norm": 0.3318726420402527,
895
+ "learning_rate": 3.3806739069896965e-05,
896
+ "loss": 0.0241,
897
+ "step": 3125
898
+ },
899
+ {
900
+ "epoch": 1.1842105263157894,
901
+ "grad_norm": 0.38022565841674805,
902
+ "learning_rate": 3.363269284321916e-05,
903
+ "loss": 0.0225,
904
+ "step": 3150
905
+ },
906
+ {
907
+ "epoch": 1.193609022556391,
908
+ "grad_norm": 0.146707221865654,
909
+ "learning_rate": 3.3458646616541355e-05,
910
+ "loss": 0.0208,
911
+ "step": 3175
912
+ },
913
+ {
914
+ "epoch": 1.2030075187969924,
915
+ "grad_norm": 0.3654153347015381,
916
+ "learning_rate": 3.328460038986355e-05,
917
+ "loss": 0.0219,
918
+ "step": 3200
919
+ },
920
+ {
921
+ "epoch": 1.212406015037594,
922
+ "grad_norm": 0.2009086161851883,
923
+ "learning_rate": 3.3110554163185744e-05,
924
+ "loss": 0.0227,
925
+ "step": 3225
926
+ },
927
+ {
928
+ "epoch": 1.2218045112781954,
929
+ "grad_norm": 0.22058619558811188,
930
+ "learning_rate": 3.2936507936507936e-05,
931
+ "loss": 0.021,
932
+ "step": 3250
933
+ },
934
+ {
935
+ "epoch": 1.231203007518797,
936
+ "grad_norm": 0.25268346071243286,
937
+ "learning_rate": 3.2762461709830134e-05,
938
+ "loss": 0.0232,
939
+ "step": 3275
940
+ },
941
+ {
942
+ "epoch": 1.2406015037593985,
943
+ "grad_norm": 0.14020860195159912,
944
+ "learning_rate": 3.258841548315233e-05,
945
+ "loss": 0.0225,
946
+ "step": 3300
947
+ },
948
+ {
949
+ "epoch": 1.25,
950
+ "grad_norm": 0.3813135027885437,
951
+ "learning_rate": 3.241436925647452e-05,
952
+ "loss": 0.0231,
953
+ "step": 3325
954
+ },
955
+ {
956
+ "epoch": 1.2593984962406015,
957
+ "grad_norm": 0.4818556010723114,
958
+ "learning_rate": 3.2240323029796715e-05,
959
+ "loss": 0.0239,
960
+ "step": 3350
961
+ },
962
+ {
963
+ "epoch": 1.268796992481203,
964
+ "grad_norm": 0.16603197157382965,
965
+ "learning_rate": 3.206627680311891e-05,
966
+ "loss": 0.0211,
967
+ "step": 3375
968
+ },
969
+ {
970
+ "epoch": 1.2781954887218046,
971
+ "grad_norm": 0.30207017064094543,
972
+ "learning_rate": 3.1892230576441104e-05,
973
+ "loss": 0.0226,
974
+ "step": 3400
975
+ },
976
+ {
977
+ "epoch": 1.287593984962406,
978
+ "grad_norm": 0.5928161144256592,
979
+ "learning_rate": 3.1718184349763296e-05,
980
+ "loss": 0.025,
981
+ "step": 3425
982
+ },
983
+ {
984
+ "epoch": 1.2969924812030076,
985
+ "grad_norm": 0.46870991587638855,
986
+ "learning_rate": 3.1544138123085494e-05,
987
+ "loss": 0.023,
988
+ "step": 3450
989
+ },
990
+ {
991
+ "epoch": 1.306390977443609,
992
+ "grad_norm": 0.14328891038894653,
993
+ "learning_rate": 3.1370091896407685e-05,
994
+ "loss": 0.0207,
995
+ "step": 3475
996
+ },
997
+ {
998
+ "epoch": 1.3157894736842106,
999
+ "grad_norm": 0.17147475481033325,
1000
+ "learning_rate": 3.1196045669729884e-05,
1001
+ "loss": 0.0227,
1002
+ "step": 3500
1003
+ },
1004
+ {
1005
+ "epoch": 1.3251879699248121,
1006
+ "grad_norm": 0.1562870591878891,
1007
+ "learning_rate": 3.1021999443052075e-05,
1008
+ "loss": 0.0208,
1009
+ "step": 3525
1010
+ },
1011
+ {
1012
+ "epoch": 1.3345864661654137,
1013
+ "grad_norm": 0.49871566891670227,
1014
+ "learning_rate": 3.084795321637427e-05,
1015
+ "loss": 0.0214,
1016
+ "step": 3550
1017
+ },
1018
+ {
1019
+ "epoch": 1.3439849624060152,
1020
+ "grad_norm": 0.18314896523952484,
1021
+ "learning_rate": 3.0673906989696465e-05,
1022
+ "loss": 0.0208,
1023
+ "step": 3575
1024
+ },
1025
+ {
1026
+ "epoch": 1.3533834586466165,
1027
+ "grad_norm": 0.1483224481344223,
1028
+ "learning_rate": 3.0499860763018663e-05,
1029
+ "loss": 0.0193,
1030
+ "step": 3600
1031
+ },
1032
+ {
1033
+ "epoch": 1.362781954887218,
1034
+ "grad_norm": 0.16652218997478485,
1035
+ "learning_rate": 3.032581453634085e-05,
1036
+ "loss": 0.0227,
1037
+ "step": 3625
1038
+ },
1039
+ {
1040
+ "epoch": 1.3721804511278195,
1041
+ "grad_norm": 0.26391515135765076,
1042
+ "learning_rate": 3.015176830966305e-05,
1043
+ "loss": 0.021,
1044
+ "step": 3650
1045
+ },
1046
+ {
1047
+ "epoch": 1.381578947368421,
1048
+ "grad_norm": 0.3377884030342102,
1049
+ "learning_rate": 2.9977722082985244e-05,
1050
+ "loss": 0.0205,
1051
+ "step": 3675
1052
+ },
1053
+ {
1054
+ "epoch": 1.3909774436090225,
1055
+ "grad_norm": 0.16687631607055664,
1056
+ "learning_rate": 2.980367585630744e-05,
1057
+ "loss": 0.0212,
1058
+ "step": 3700
1059
+ },
1060
+ {
1061
+ "epoch": 1.400375939849624,
1062
+ "grad_norm": 0.21264871954917908,
1063
+ "learning_rate": 2.962962962962963e-05,
1064
+ "loss": 0.0212,
1065
+ "step": 3725
1066
+ },
1067
+ {
1068
+ "epoch": 1.4097744360902256,
1069
+ "grad_norm": 0.4501771330833435,
1070
+ "learning_rate": 2.9455583402951825e-05,
1071
+ "loss": 0.0228,
1072
+ "step": 3750
1073
+ },
1074
+ {
1075
+ "epoch": 1.419172932330827,
1076
+ "grad_norm": 0.2912803888320923,
1077
+ "learning_rate": 2.928153717627402e-05,
1078
+ "loss": 0.022,
1079
+ "step": 3775
1080
+ },
1081
+ {
1082
+ "epoch": 1.4285714285714286,
1083
+ "grad_norm": 0.605665385723114,
1084
+ "learning_rate": 2.9107490949596218e-05,
1085
+ "loss": 0.0224,
1086
+ "step": 3800
1087
+ },
1088
+ {
1089
+ "epoch": 1.4379699248120301,
1090
+ "grad_norm": 0.13825617730617523,
1091
+ "learning_rate": 2.8933444722918406e-05,
1092
+ "loss": 0.0223,
1093
+ "step": 3825
1094
+ },
1095
+ {
1096
+ "epoch": 1.4473684210526316,
1097
+ "grad_norm": 0.3659210205078125,
1098
+ "learning_rate": 2.8759398496240604e-05,
1099
+ "loss": 0.0226,
1100
+ "step": 3850
1101
+ },
1102
+ {
1103
+ "epoch": 1.4567669172932332,
1104
+ "grad_norm": 0.2020525336265564,
1105
+ "learning_rate": 2.85853522695628e-05,
1106
+ "loss": 0.022,
1107
+ "step": 3875
1108
+ },
1109
+ {
1110
+ "epoch": 1.4661654135338344,
1111
+ "grad_norm": 0.23142847418785095,
1112
+ "learning_rate": 2.8411306042884993e-05,
1113
+ "loss": 0.0228,
1114
+ "step": 3900
1115
+ },
1116
+ {
1117
+ "epoch": 1.475563909774436,
1118
+ "grad_norm": 0.3832224905490875,
1119
+ "learning_rate": 2.8237259816207185e-05,
1120
+ "loss": 0.0209,
1121
+ "step": 3925
1122
+ },
1123
+ {
1124
+ "epoch": 1.4849624060150375,
1125
+ "grad_norm": 0.41142329573631287,
1126
+ "learning_rate": 2.806321358952938e-05,
1127
+ "loss": 0.021,
1128
+ "step": 3950
1129
+ },
1130
+ {
1131
+ "epoch": 1.494360902255639,
1132
+ "grad_norm": 0.13629283010959625,
1133
+ "learning_rate": 2.7889167362851574e-05,
1134
+ "loss": 0.0217,
1135
+ "step": 3975
1136
+ },
1137
+ {
1138
+ "epoch": 1.5037593984962405,
1139
+ "grad_norm": 0.19223739206790924,
1140
+ "learning_rate": 2.7715121136173773e-05,
1141
+ "loss": 0.0215,
1142
+ "step": 4000
1143
+ },
1144
+ {
1145
+ "epoch": 1.513157894736842,
1146
+ "grad_norm": 0.2219369262456894,
1147
+ "learning_rate": 2.754107490949596e-05,
1148
+ "loss": 0.0219,
1149
+ "step": 4025
1150
+ },
1151
+ {
1152
+ "epoch": 1.5225563909774436,
1153
+ "grad_norm": 0.3146340250968933,
1154
+ "learning_rate": 2.736702868281816e-05,
1155
+ "loss": 0.024,
1156
+ "step": 4050
1157
+ },
1158
+ {
1159
+ "epoch": 1.531954887218045,
1160
+ "grad_norm": 0.25318026542663574,
1161
+ "learning_rate": 2.7192982456140354e-05,
1162
+ "loss": 0.0227,
1163
+ "step": 4075
1164
+ },
1165
+ {
1166
+ "epoch": 1.5413533834586466,
1167
+ "grad_norm": 0.34827953577041626,
1168
+ "learning_rate": 2.701893622946255e-05,
1169
+ "loss": 0.0229,
1170
+ "step": 4100
1171
+ },
1172
+ {
1173
+ "epoch": 1.550751879699248,
1174
+ "grad_norm": 0.33609580993652344,
1175
+ "learning_rate": 2.684489000278474e-05,
1176
+ "loss": 0.0204,
1177
+ "step": 4125
1178
+ },
1179
+ {
1180
+ "epoch": 1.5601503759398496,
1181
+ "grad_norm": 0.21258720755577087,
1182
+ "learning_rate": 2.6670843776106935e-05,
1183
+ "loss": 0.0213,
1184
+ "step": 4150
1185
+ },
1186
+ {
1187
+ "epoch": 1.5695488721804511,
1188
+ "grad_norm": 0.2116004377603531,
1189
+ "learning_rate": 2.649679754942913e-05,
1190
+ "loss": 0.0236,
1191
+ "step": 4175
1192
+ },
1193
+ {
1194
+ "epoch": 1.5789473684210527,
1195
+ "grad_norm": 0.3292735815048218,
1196
+ "learning_rate": 2.6322751322751328e-05,
1197
+ "loss": 0.0221,
1198
+ "step": 4200
1199
+ },
1200
+ {
1201
+ "epoch": 1.5883458646616542,
1202
+ "grad_norm": 0.2400967925786972,
1203
+ "learning_rate": 2.6148705096073516e-05,
1204
+ "loss": 0.0233,
1205
+ "step": 4225
1206
+ },
1207
+ {
1208
+ "epoch": 1.5977443609022557,
1209
+ "grad_norm": 0.17355374991893768,
1210
+ "learning_rate": 2.5974658869395714e-05,
1211
+ "loss": 0.0218,
1212
+ "step": 4250
1213
+ },
1214
+ {
1215
+ "epoch": 1.6071428571428572,
1216
+ "grad_norm": 0.2389274537563324,
1217
+ "learning_rate": 2.580061264271791e-05,
1218
+ "loss": 0.0202,
1219
+ "step": 4275
1220
+ },
1221
+ {
1222
+ "epoch": 1.6165413533834587,
1223
+ "grad_norm": 0.1675727367401123,
1224
+ "learning_rate": 2.5626566416040103e-05,
1225
+ "loss": 0.0217,
1226
+ "step": 4300
1227
+ },
1228
+ {
1229
+ "epoch": 1.6259398496240602,
1230
+ "grad_norm": 0.1514149010181427,
1231
+ "learning_rate": 2.5452520189362295e-05,
1232
+ "loss": 0.0213,
1233
+ "step": 4325
1234
+ },
1235
+ {
1236
+ "epoch": 1.6353383458646618,
1237
+ "grad_norm": 0.30107012391090393,
1238
+ "learning_rate": 2.527847396268449e-05,
1239
+ "loss": 0.0203,
1240
+ "step": 4350
1241
+ },
1242
+ {
1243
+ "epoch": 1.6447368421052633,
1244
+ "grad_norm": 0.42004597187042236,
1245
+ "learning_rate": 2.5104427736006684e-05,
1246
+ "loss": 0.0227,
1247
+ "step": 4375
1248
+ },
1249
+ {
1250
+ "epoch": 1.6541353383458648,
1251
+ "grad_norm": 0.44374194741249084,
1252
+ "learning_rate": 2.493038150932888e-05,
1253
+ "loss": 0.0207,
1254
+ "step": 4400
1255
+ },
1256
+ {
1257
+ "epoch": 1.6635338345864663,
1258
+ "grad_norm": 0.16044478118419647,
1259
+ "learning_rate": 2.475633528265107e-05,
1260
+ "loss": 0.0212,
1261
+ "step": 4425
1262
+ },
1263
+ {
1264
+ "epoch": 1.6729323308270678,
1265
+ "grad_norm": 0.3987959325313568,
1266
+ "learning_rate": 2.458228905597327e-05,
1267
+ "loss": 0.0231,
1268
+ "step": 4450
1269
+ },
1270
+ {
1271
+ "epoch": 1.6823308270676691,
1272
+ "grad_norm": 0.10810192674398422,
1273
+ "learning_rate": 2.440824282929546e-05,
1274
+ "loss": 0.0203,
1275
+ "step": 4475
1276
+ },
1277
+ {
1278
+ "epoch": 1.6917293233082706,
1279
+ "grad_norm": 0.1742391586303711,
1280
+ "learning_rate": 2.4234196602617655e-05,
1281
+ "loss": 0.0219,
1282
+ "step": 4500
1283
+ },
1284
+ {
1285
+ "epoch": 1.7011278195488722,
1286
+ "grad_norm": 0.46595847606658936,
1287
+ "learning_rate": 2.406015037593985e-05,
1288
+ "loss": 0.0203,
1289
+ "step": 4525
1290
+ },
1291
+ {
1292
+ "epoch": 1.7105263157894737,
1293
+ "grad_norm": 0.6574416756629944,
1294
+ "learning_rate": 2.3886104149262044e-05,
1295
+ "loss": 0.0251,
1296
+ "step": 4550
1297
+ },
1298
+ {
1299
+ "epoch": 1.7199248120300752,
1300
+ "grad_norm": 0.3133036196231842,
1301
+ "learning_rate": 2.371205792258424e-05,
1302
+ "loss": 0.0208,
1303
+ "step": 4575
1304
+ },
1305
+ {
1306
+ "epoch": 1.7293233082706767,
1307
+ "grad_norm": 0.17739902436733246,
1308
+ "learning_rate": 2.3538011695906434e-05,
1309
+ "loss": 0.0205,
1310
+ "step": 4600
1311
+ },
1312
+ {
1313
+ "epoch": 1.7387218045112782,
1314
+ "grad_norm": 0.16983534395694733,
1315
+ "learning_rate": 2.3363965469228625e-05,
1316
+ "loss": 0.0206,
1317
+ "step": 4625
1318
+ },
1319
+ {
1320
+ "epoch": 1.7481203007518797,
1321
+ "grad_norm": 0.14968422055244446,
1322
+ "learning_rate": 2.3189919242550824e-05,
1323
+ "loss": 0.0219,
1324
+ "step": 4650
1325
+ },
1326
+ {
1327
+ "epoch": 1.7575187969924813,
1328
+ "grad_norm": 0.2490016222000122,
1329
+ "learning_rate": 2.3015873015873015e-05,
1330
+ "loss": 0.021,
1331
+ "step": 4675
1332
+ },
1333
+ {
1334
+ "epoch": 1.7669172932330826,
1335
+ "grad_norm": 0.23715806007385254,
1336
+ "learning_rate": 2.284182678919521e-05,
1337
+ "loss": 0.0203,
1338
+ "step": 4700
1339
+ },
1340
+ {
1341
+ "epoch": 1.776315789473684,
1342
+ "grad_norm": 0.3275100290775299,
1343
+ "learning_rate": 2.2667780562517405e-05,
1344
+ "loss": 0.0228,
1345
+ "step": 4725
1346
+ },
1347
+ {
1348
+ "epoch": 1.7857142857142856,
1349
+ "grad_norm": 0.33084341883659363,
1350
+ "learning_rate": 2.24937343358396e-05,
1351
+ "loss": 0.023,
1352
+ "step": 4750
1353
+ },
1354
+ {
1355
+ "epoch": 1.795112781954887,
1356
+ "grad_norm": 0.5483024716377258,
1357
+ "learning_rate": 2.2319688109161794e-05,
1358
+ "loss": 0.0225,
1359
+ "step": 4775
1360
+ },
1361
+ {
1362
+ "epoch": 1.8045112781954886,
1363
+ "grad_norm": 0.7777302861213684,
1364
+ "learning_rate": 2.214564188248399e-05,
1365
+ "loss": 0.0231,
1366
+ "step": 4800
1367
+ },
1368
+ {
1369
+ "epoch": 1.8139097744360901,
1370
+ "grad_norm": 0.39886680245399475,
1371
+ "learning_rate": 2.197159565580618e-05,
1372
+ "loss": 0.0221,
1373
+ "step": 4825
1374
+ },
1375
+ {
1376
+ "epoch": 1.8233082706766917,
1377
+ "grad_norm": 0.18582062423229218,
1378
+ "learning_rate": 2.179754942912838e-05,
1379
+ "loss": 0.0207,
1380
+ "step": 4850
1381
+ },
1382
+ {
1383
+ "epoch": 1.8327067669172932,
1384
+ "grad_norm": 0.1751798689365387,
1385
+ "learning_rate": 2.162350320245057e-05,
1386
+ "loss": 0.0208,
1387
+ "step": 4875
1388
+ },
1389
+ {
1390
+ "epoch": 1.8421052631578947,
1391
+ "grad_norm": 0.15387402474880219,
1392
+ "learning_rate": 2.1449456975772768e-05,
1393
+ "loss": 0.0209,
1394
+ "step": 4900
1395
+ },
1396
+ {
1397
+ "epoch": 1.8515037593984962,
1398
+ "grad_norm": 0.17794908583164215,
1399
+ "learning_rate": 2.127541074909496e-05,
1400
+ "loss": 0.0217,
1401
+ "step": 4925
1402
+ },
1403
+ {
1404
+ "epoch": 1.8609022556390977,
1405
+ "grad_norm": 0.3664356470108032,
1406
+ "learning_rate": 2.1101364522417154e-05,
1407
+ "loss": 0.0213,
1408
+ "step": 4950
1409
+ },
1410
+ {
1411
+ "epoch": 1.8703007518796992,
1412
+ "grad_norm": 0.11275918781757355,
1413
+ "learning_rate": 2.092731829573935e-05,
1414
+ "loss": 0.0198,
1415
+ "step": 4975
1416
+ },
1417
+ {
1418
+ "epoch": 1.8796992481203008,
1419
+ "grad_norm": 0.14856699109077454,
1420
+ "learning_rate": 2.0753272069061544e-05,
1421
+ "loss": 0.0203,
1422
+ "step": 5000
1423
+ },
1424
+ {
1425
+ "epoch": 1.8890977443609023,
1426
+ "grad_norm": 0.27216190099716187,
1427
+ "learning_rate": 2.0579225842383735e-05,
1428
+ "loss": 0.0231,
1429
+ "step": 5025
1430
+ },
1431
+ {
1432
+ "epoch": 1.8984962406015038,
1433
+ "grad_norm": 0.26974913477897644,
1434
+ "learning_rate": 2.0405179615705934e-05,
1435
+ "loss": 0.0215,
1436
+ "step": 5050
1437
+ },
1438
+ {
1439
+ "epoch": 1.9078947368421053,
1440
+ "grad_norm": 0.2766593098640442,
1441
+ "learning_rate": 2.0231133389028125e-05,
1442
+ "loss": 0.0211,
1443
+ "step": 5075
1444
+ },
1445
+ {
1446
+ "epoch": 1.9172932330827068,
1447
+ "grad_norm": 0.22481001913547516,
1448
+ "learning_rate": 2.0057087162350323e-05,
1449
+ "loss": 0.0204,
1450
+ "step": 5100
1451
+ },
1452
+ {
1453
+ "epoch": 1.9266917293233083,
1454
+ "grad_norm": 0.27344635128974915,
1455
+ "learning_rate": 1.9883040935672515e-05,
1456
+ "loss": 0.0219,
1457
+ "step": 5125
1458
+ },
1459
+ {
1460
+ "epoch": 1.9360902255639099,
1461
+ "grad_norm": 0.33460211753845215,
1462
+ "learning_rate": 1.970899470899471e-05,
1463
+ "loss": 0.0207,
1464
+ "step": 5150
1465
+ },
1466
+ {
1467
+ "epoch": 1.9454887218045114,
1468
+ "grad_norm": 0.6251159310340881,
1469
+ "learning_rate": 1.9534948482316904e-05,
1470
+ "loss": 0.0215,
1471
+ "step": 5175
1472
+ },
1473
+ {
1474
+ "epoch": 1.954887218045113,
1475
+ "grad_norm": 0.1954965740442276,
1476
+ "learning_rate": 1.93609022556391e-05,
1477
+ "loss": 0.0233,
1478
+ "step": 5200
1479
+ },
1480
+ {
1481
+ "epoch": 1.9642857142857144,
1482
+ "grad_norm": 0.47101154923439026,
1483
+ "learning_rate": 1.9186856028961294e-05,
1484
+ "loss": 0.0211,
1485
+ "step": 5225
1486
+ },
1487
+ {
1488
+ "epoch": 1.973684210526316,
1489
+ "grad_norm": 0.18471819162368774,
1490
+ "learning_rate": 1.901280980228349e-05,
1491
+ "loss": 0.0227,
1492
+ "step": 5250
1493
+ },
1494
+ {
1495
+ "epoch": 1.9830827067669174,
1496
+ "grad_norm": 0.6460950970649719,
1497
+ "learning_rate": 1.883876357560568e-05,
1498
+ "loss": 0.021,
1499
+ "step": 5275
1500
+ },
1501
+ {
1502
+ "epoch": 1.9924812030075187,
1503
+ "grad_norm": 0.32073792815208435,
1504
+ "learning_rate": 1.8664717348927878e-05,
1505
+ "loss": 0.0213,
1506
+ "step": 5300
1507
+ },
1508
+ {
1509
+ "epoch": 2.0,
1510
+ "eval_explained_variance": 0.27539485692977905,
1511
+ "eval_loss": 0.021362008526921272,
1512
+ "eval_mae": 0.11374949663877487,
1513
+ "eval_mse": 0.02136271260678768,
1514
+ "eval_r2": 0.26287728548049927,
1515
+ "eval_rmse": 0.1461598871331929,
1516
+ "eval_runtime": 5.0558,
1517
+ "eval_samples_per_second": 4208.474,
1518
+ "eval_steps_per_second": 65.866,
1519
+ "step": 5320
1520
+ },
1521
+ {
1522
+ "epoch": 2.0018796992481205,
1523
+ "grad_norm": 0.1525137573480606,
1524
+ "learning_rate": 1.849067112225007e-05,
1525
+ "loss": 0.0193,
1526
+ "step": 5325
1527
+ },
1528
+ {
1529
+ "epoch": 2.011278195488722,
1530
+ "grad_norm": 0.31058961153030396,
1531
+ "learning_rate": 1.8316624895572264e-05,
1532
+ "loss": 0.0186,
1533
+ "step": 5350
1534
+ },
1535
+ {
1536
+ "epoch": 2.0206766917293235,
1537
+ "grad_norm": 0.19665969908237457,
1538
+ "learning_rate": 1.814257866889446e-05,
1539
+ "loss": 0.0212,
1540
+ "step": 5375
1541
+ },
1542
+ {
1543
+ "epoch": 2.030075187969925,
1544
+ "grad_norm": 0.20314434170722961,
1545
+ "learning_rate": 1.7968532442216654e-05,
1546
+ "loss": 0.0198,
1547
+ "step": 5400
1548
+ },
1549
+ {
1550
+ "epoch": 2.039473684210526,
1551
+ "grad_norm": 0.2827511429786682,
1552
+ "learning_rate": 1.779448621553885e-05,
1553
+ "loss": 0.0196,
1554
+ "step": 5425
1555
+ },
1556
+ {
1557
+ "epoch": 2.0488721804511276,
1558
+ "grad_norm": 0.3278052508831024,
1559
+ "learning_rate": 1.7620439988861043e-05,
1560
+ "loss": 0.019,
1561
+ "step": 5450
1562
+ },
1563
+ {
1564
+ "epoch": 2.058270676691729,
1565
+ "grad_norm": 0.19658660888671875,
1566
+ "learning_rate": 1.7446393762183235e-05,
1567
+ "loss": 0.0211,
1568
+ "step": 5475
1569
+ },
1570
+ {
1571
+ "epoch": 2.0676691729323307,
1572
+ "grad_norm": 0.2107669562101364,
1573
+ "learning_rate": 1.7272347535505433e-05,
1574
+ "loss": 0.0199,
1575
+ "step": 5500
1576
+ },
1577
+ {
1578
+ "epoch": 2.077067669172932,
1579
+ "grad_norm": 0.14388123154640198,
1580
+ "learning_rate": 1.7098301308827624e-05,
1581
+ "loss": 0.0208,
1582
+ "step": 5525
1583
+ },
1584
+ {
1585
+ "epoch": 2.0864661654135337,
1586
+ "grad_norm": 0.2755216658115387,
1587
+ "learning_rate": 1.692425508214982e-05,
1588
+ "loss": 0.0212,
1589
+ "step": 5550
1590
+ },
1591
+ {
1592
+ "epoch": 2.095864661654135,
1593
+ "grad_norm": 0.18841974437236786,
1594
+ "learning_rate": 1.6750208855472014e-05,
1595
+ "loss": 0.0199,
1596
+ "step": 5575
1597
+ },
1598
+ {
1599
+ "epoch": 2.1052631578947367,
1600
+ "grad_norm": 0.6396263241767883,
1601
+ "learning_rate": 1.657616262879421e-05,
1602
+ "loss": 0.0195,
1603
+ "step": 5600
1604
+ },
1605
+ {
1606
+ "epoch": 2.1146616541353382,
1607
+ "grad_norm": 0.3483275771141052,
1608
+ "learning_rate": 1.6402116402116404e-05,
1609
+ "loss": 0.0195,
1610
+ "step": 5625
1611
+ },
1612
+ {
1613
+ "epoch": 2.1240601503759398,
1614
+ "grad_norm": 0.5073034763336182,
1615
+ "learning_rate": 1.62280701754386e-05,
1616
+ "loss": 0.0175,
1617
+ "step": 5650
1618
+ },
1619
+ {
1620
+ "epoch": 2.1334586466165413,
1621
+ "grad_norm": 0.2675146758556366,
1622
+ "learning_rate": 1.605402394876079e-05,
1623
+ "loss": 0.0181,
1624
+ "step": 5675
1625
+ },
1626
+ {
1627
+ "epoch": 2.142857142857143,
1628
+ "grad_norm": 0.3021029233932495,
1629
+ "learning_rate": 1.5879977722082988e-05,
1630
+ "loss": 0.0193,
1631
+ "step": 5700
1632
+ },
1633
+ {
1634
+ "epoch": 2.1522556390977443,
1635
+ "grad_norm": 0.162827268242836,
1636
+ "learning_rate": 1.570593149540518e-05,
1637
+ "loss": 0.0175,
1638
+ "step": 5725
1639
+ },
1640
+ {
1641
+ "epoch": 2.161654135338346,
1642
+ "grad_norm": 0.36196500062942505,
1643
+ "learning_rate": 1.5531885268727374e-05,
1644
+ "loss": 0.0177,
1645
+ "step": 5750
1646
+ },
1647
+ {
1648
+ "epoch": 2.1710526315789473,
1649
+ "grad_norm": 0.2810737192630768,
1650
+ "learning_rate": 1.535783904204957e-05,
1651
+ "loss": 0.0186,
1652
+ "step": 5775
1653
+ },
1654
+ {
1655
+ "epoch": 2.180451127819549,
1656
+ "grad_norm": 0.17449721693992615,
1657
+ "learning_rate": 1.5183792815371764e-05,
1658
+ "loss": 0.0182,
1659
+ "step": 5800
1660
+ },
1661
+ {
1662
+ "epoch": 2.1898496240601504,
1663
+ "grad_norm": 0.1482439786195755,
1664
+ "learning_rate": 1.5009746588693957e-05,
1665
+ "loss": 0.0214,
1666
+ "step": 5825
1667
+ },
1668
+ {
1669
+ "epoch": 2.199248120300752,
1670
+ "grad_norm": 0.42385685443878174,
1671
+ "learning_rate": 1.4835700362016153e-05,
1672
+ "loss": 0.0185,
1673
+ "step": 5850
1674
+ },
1675
+ {
1676
+ "epoch": 2.2086466165413534,
1677
+ "grad_norm": 0.14571012556552887,
1678
+ "learning_rate": 1.4661654135338346e-05,
1679
+ "loss": 0.0199,
1680
+ "step": 5875
1681
+ },
1682
+ {
1683
+ "epoch": 2.218045112781955,
1684
+ "grad_norm": 0.27442601323127747,
1685
+ "learning_rate": 1.4487607908660541e-05,
1686
+ "loss": 0.0196,
1687
+ "step": 5900
1688
+ },
1689
+ {
1690
+ "epoch": 2.2274436090225564,
1691
+ "grad_norm": 0.2556711435317993,
1692
+ "learning_rate": 1.4313561681982734e-05,
1693
+ "loss": 0.0202,
1694
+ "step": 5925
1695
+ },
1696
+ {
1697
+ "epoch": 2.236842105263158,
1698
+ "grad_norm": 0.1367371678352356,
1699
+ "learning_rate": 1.413951545530493e-05,
1700
+ "loss": 0.0189,
1701
+ "step": 5950
1702
+ },
1703
+ {
1704
+ "epoch": 2.2462406015037595,
1705
+ "grad_norm": 0.5869988799095154,
1706
+ "learning_rate": 1.3965469228627124e-05,
1707
+ "loss": 0.0193,
1708
+ "step": 5975
1709
+ },
1710
+ {
1711
+ "epoch": 2.255639097744361,
1712
+ "grad_norm": 0.3578735888004303,
1713
+ "learning_rate": 1.3791423001949319e-05,
1714
+ "loss": 0.019,
1715
+ "step": 6000
1716
+ },
1717
+ {
1718
+ "epoch": 2.2650375939849625,
1719
+ "grad_norm": 0.30225270986557007,
1720
+ "learning_rate": 1.3624338624338626e-05,
1721
+ "loss": 0.019,
1722
+ "step": 6025
1723
+ },
1724
+ {
1725
+ "epoch": 2.274436090225564,
1726
+ "grad_norm": 0.181197851896286,
1727
+ "learning_rate": 1.3450292397660819e-05,
1728
+ "loss": 0.0204,
1729
+ "step": 6050
1730
+ },
1731
+ {
1732
+ "epoch": 2.2838345864661656,
1733
+ "grad_norm": 0.36190444231033325,
1734
+ "learning_rate": 1.3276246170983015e-05,
1735
+ "loss": 0.0197,
1736
+ "step": 6075
1737
+ },
1738
+ {
1739
+ "epoch": 2.293233082706767,
1740
+ "grad_norm": 0.20328275859355927,
1741
+ "learning_rate": 1.3102199944305207e-05,
1742
+ "loss": 0.0204,
1743
+ "step": 6100
1744
+ },
1745
+ {
1746
+ "epoch": 2.3026315789473686,
1747
+ "grad_norm": 0.36057910323143005,
1748
+ "learning_rate": 1.2928153717627403e-05,
1749
+ "loss": 0.0176,
1750
+ "step": 6125
1751
+ },
1752
+ {
1753
+ "epoch": 2.31203007518797,
1754
+ "grad_norm": 0.18688635528087616,
1755
+ "learning_rate": 1.2754107490949596e-05,
1756
+ "loss": 0.0174,
1757
+ "step": 6150
1758
+ },
1759
+ {
1760
+ "epoch": 2.3214285714285716,
1761
+ "grad_norm": 0.3669414520263672,
1762
+ "learning_rate": 1.2580061264271793e-05,
1763
+ "loss": 0.0185,
1764
+ "step": 6175
1765
+ },
1766
+ {
1767
+ "epoch": 2.3308270676691727,
1768
+ "grad_norm": 0.19976933300495148,
1769
+ "learning_rate": 1.2406015037593984e-05,
1770
+ "loss": 0.0189,
1771
+ "step": 6200
1772
+ },
1773
+ {
1774
+ "epoch": 2.340225563909774,
1775
+ "grad_norm": 0.28642526268959045,
1776
+ "learning_rate": 1.2231968810916179e-05,
1777
+ "loss": 0.0201,
1778
+ "step": 6225
1779
+ },
1780
+ {
1781
+ "epoch": 2.3496240601503757,
1782
+ "grad_norm": 0.13848170638084412,
1783
+ "learning_rate": 1.2057922584238374e-05,
1784
+ "loss": 0.0195,
1785
+ "step": 6250
1786
+ },
1787
+ {
1788
+ "epoch": 2.3590225563909772,
1789
+ "grad_norm": 0.490566223859787,
1790
+ "learning_rate": 1.1883876357560568e-05,
1791
+ "loss": 0.0202,
1792
+ "step": 6275
1793
+ },
1794
+ {
1795
+ "epoch": 2.3684210526315788,
1796
+ "grad_norm": 0.534001886844635,
1797
+ "learning_rate": 1.1709830130882762e-05,
1798
+ "loss": 0.0194,
1799
+ "step": 6300
1800
+ },
1801
+ {
1802
+ "epoch": 2.3778195488721803,
1803
+ "grad_norm": 0.18683072924613953,
1804
+ "learning_rate": 1.1535783904204956e-05,
1805
+ "loss": 0.0178,
1806
+ "step": 6325
1807
+ },
1808
+ {
1809
+ "epoch": 2.387218045112782,
1810
+ "grad_norm": 0.1929399073123932,
1811
+ "learning_rate": 1.1361737677527151e-05,
1812
+ "loss": 0.0185,
1813
+ "step": 6350
1814
+ },
1815
+ {
1816
+ "epoch": 2.3966165413533833,
1817
+ "grad_norm": 0.3412390947341919,
1818
+ "learning_rate": 1.1187691450849346e-05,
1819
+ "loss": 0.0208,
1820
+ "step": 6375
1821
+ },
1822
+ {
1823
+ "epoch": 2.406015037593985,
1824
+ "grad_norm": 0.35896581411361694,
1825
+ "learning_rate": 1.1013645224171539e-05,
1826
+ "loss": 0.0188,
1827
+ "step": 6400
1828
+ },
1829
+ {
1830
+ "epoch": 2.4154135338345863,
1831
+ "grad_norm": 0.4344053864479065,
1832
+ "learning_rate": 1.0839598997493734e-05,
1833
+ "loss": 0.0199,
1834
+ "step": 6425
1835
+ },
1836
+ {
1837
+ "epoch": 2.424812030075188,
1838
+ "grad_norm": 0.6244423985481262,
1839
+ "learning_rate": 1.0665552770815929e-05,
1840
+ "loss": 0.0166,
1841
+ "step": 6450
1842
+ },
1843
+ {
1844
+ "epoch": 2.4342105263157894,
1845
+ "grad_norm": 0.3958113491535187,
1846
+ "learning_rate": 1.0491506544138123e-05,
1847
+ "loss": 0.0225,
1848
+ "step": 6475
1849
+ },
1850
+ {
1851
+ "epoch": 2.443609022556391,
1852
+ "grad_norm": 0.20780402421951294,
1853
+ "learning_rate": 1.0317460317460318e-05,
1854
+ "loss": 0.0199,
1855
+ "step": 6500
1856
+ },
1857
+ {
1858
+ "epoch": 2.4530075187969924,
1859
+ "grad_norm": 0.23762066662311554,
1860
+ "learning_rate": 1.0143414090782511e-05,
1861
+ "loss": 0.0202,
1862
+ "step": 6525
1863
+ },
1864
+ {
1865
+ "epoch": 2.462406015037594,
1866
+ "grad_norm": 0.1582295000553131,
1867
+ "learning_rate": 9.969367864104706e-06,
1868
+ "loss": 0.0192,
1869
+ "step": 6550
1870
+ },
1871
+ {
1872
+ "epoch": 2.4718045112781954,
1873
+ "grad_norm": 0.3237611949443817,
1874
+ "learning_rate": 9.795321637426901e-06,
1875
+ "loss": 0.0191,
1876
+ "step": 6575
1877
+ },
1878
+ {
1879
+ "epoch": 2.481203007518797,
1880
+ "grad_norm": 0.2653273940086365,
1881
+ "learning_rate": 9.621275410749096e-06,
1882
+ "loss": 0.0184,
1883
+ "step": 6600
1884
+ },
1885
+ {
1886
+ "epoch": 2.4906015037593985,
1887
+ "grad_norm": 0.2554579973220825,
1888
+ "learning_rate": 9.447229184071289e-06,
1889
+ "loss": 0.0185,
1890
+ "step": 6625
1891
+ },
1892
+ {
1893
+ "epoch": 2.5,
1894
+ "grad_norm": 0.1941600739955902,
1895
+ "learning_rate": 9.273182957393484e-06,
1896
+ "loss": 0.0204,
1897
+ "step": 6650
1898
+ },
1899
+ {
1900
+ "epoch": 2.5093984962406015,
1901
+ "grad_norm": 0.23236483335494995,
1902
+ "learning_rate": 9.099136730715678e-06,
1903
+ "loss": 0.0172,
1904
+ "step": 6675
1905
+ },
1906
+ {
1907
+ "epoch": 2.518796992481203,
1908
+ "grad_norm": 0.2141353040933609,
1909
+ "learning_rate": 8.925090504037873e-06,
1910
+ "loss": 0.0188,
1911
+ "step": 6700
1912
+ },
1913
+ {
1914
+ "epoch": 2.5281954887218046,
1915
+ "grad_norm": 0.24786221981048584,
1916
+ "learning_rate": 8.751044277360066e-06,
1917
+ "loss": 0.0189,
1918
+ "step": 6725
1919
+ },
1920
+ {
1921
+ "epoch": 2.537593984962406,
1922
+ "grad_norm": 0.3599576950073242,
1923
+ "learning_rate": 8.576998050682261e-06,
1924
+ "loss": 0.0196,
1925
+ "step": 6750
1926
+ },
1927
+ {
1928
+ "epoch": 2.5469924812030076,
1929
+ "grad_norm": 0.2904399037361145,
1930
+ "learning_rate": 8.402951824004456e-06,
1931
+ "loss": 0.0194,
1932
+ "step": 6775
1933
+ },
1934
+ {
1935
+ "epoch": 2.556390977443609,
1936
+ "grad_norm": 0.1519438773393631,
1937
+ "learning_rate": 8.22890559732665e-06,
1938
+ "loss": 0.0197,
1939
+ "step": 6800
1940
+ },
1941
+ {
1942
+ "epoch": 2.5657894736842106,
1943
+ "grad_norm": 0.19205723702907562,
1944
+ "learning_rate": 8.054859370648844e-06,
1945
+ "loss": 0.0191,
1946
+ "step": 6825
1947
+ },
1948
+ {
1949
+ "epoch": 2.575187969924812,
1950
+ "grad_norm": 0.3749381899833679,
1951
+ "learning_rate": 7.880813143971038e-06,
1952
+ "loss": 0.0179,
1953
+ "step": 6850
1954
+ },
1955
+ {
1956
+ "epoch": 2.5845864661654137,
1957
+ "grad_norm": 0.3645901381969452,
1958
+ "learning_rate": 7.706766917293233e-06,
1959
+ "loss": 0.0191,
1960
+ "step": 6875
1961
+ },
1962
+ {
1963
+ "epoch": 2.593984962406015,
1964
+ "grad_norm": 0.6784169673919678,
1965
+ "learning_rate": 7.532720690615427e-06,
1966
+ "loss": 0.0198,
1967
+ "step": 6900
1968
+ },
1969
+ {
1970
+ "epoch": 2.6033834586466167,
1971
+ "grad_norm": 0.23996496200561523,
1972
+ "learning_rate": 7.358674463937622e-06,
1973
+ "loss": 0.0193,
1974
+ "step": 6925
1975
+ },
1976
+ {
1977
+ "epoch": 2.612781954887218,
1978
+ "grad_norm": 0.21705913543701172,
1979
+ "learning_rate": 7.184628237259816e-06,
1980
+ "loss": 0.0199,
1981
+ "step": 6950
1982
+ },
1983
+ {
1984
+ "epoch": 2.6221804511278197,
1985
+ "grad_norm": 0.2912886440753937,
1986
+ "learning_rate": 7.010582010582011e-06,
1987
+ "loss": 0.0207,
1988
+ "step": 6975
1989
+ },
1990
+ {
1991
+ "epoch": 2.6315789473684212,
1992
+ "grad_norm": 0.5026769042015076,
1993
+ "learning_rate": 6.836535783904205e-06,
1994
+ "loss": 0.0188,
1995
+ "step": 7000
1996
+ },
1997
+ {
1998
+ "epoch": 2.6409774436090228,
1999
+ "grad_norm": 0.38726431131362915,
2000
+ "learning_rate": 6.6624895572263995e-06,
2001
+ "loss": 0.019,
2002
+ "step": 7025
2003
+ },
2004
+ {
2005
+ "epoch": 2.6503759398496243,
2006
+ "grad_norm": 0.30493810772895813,
2007
+ "learning_rate": 6.4884433305485934e-06,
2008
+ "loss": 0.0183,
2009
+ "step": 7050
2010
+ },
2011
+ {
2012
+ "epoch": 2.659774436090226,
2013
+ "grad_norm": 0.35731396079063416,
2014
+ "learning_rate": 6.314397103870788e-06,
2015
+ "loss": 0.0186,
2016
+ "step": 7075
2017
+ },
2018
+ {
2019
+ "epoch": 2.6691729323308273,
2020
+ "grad_norm": 0.49797552824020386,
2021
+ "learning_rate": 6.140350877192982e-06,
2022
+ "loss": 0.0209,
2023
+ "step": 7100
2024
+ },
2025
+ {
2026
+ "epoch": 2.678571428571429,
2027
+ "grad_norm": 0.15216362476348877,
2028
+ "learning_rate": 5.966304650515177e-06,
2029
+ "loss": 0.018,
2030
+ "step": 7125
2031
+ },
2032
+ {
2033
+ "epoch": 2.6879699248120303,
2034
+ "grad_norm": 0.15072126686573029,
2035
+ "learning_rate": 5.792258423837371e-06,
2036
+ "loss": 0.0184,
2037
+ "step": 7150
2038
+ },
2039
+ {
2040
+ "epoch": 2.6973684210526314,
2041
+ "grad_norm": 0.21924220025539398,
2042
+ "learning_rate": 5.618212197159566e-06,
2043
+ "loss": 0.0191,
2044
+ "step": 7175
2045
+ },
2046
+ {
2047
+ "epoch": 2.706766917293233,
2048
+ "grad_norm": 0.17299939692020416,
2049
+ "learning_rate": 5.44416597048176e-06,
2050
+ "loss": 0.0189,
2051
+ "step": 7200
2052
+ },
2053
+ {
2054
+ "epoch": 2.7161654135338344,
2055
+ "grad_norm": 0.45098772644996643,
2056
+ "learning_rate": 5.2701197438039544e-06,
2057
+ "loss": 0.0194,
2058
+ "step": 7225
2059
+ },
2060
+ {
2061
+ "epoch": 2.725563909774436,
2062
+ "grad_norm": 0.5662004947662354,
2063
+ "learning_rate": 5.096073517126148e-06,
2064
+ "loss": 0.0183,
2065
+ "step": 7250
2066
+ },
2067
+ {
2068
+ "epoch": 2.7349624060150375,
2069
+ "grad_norm": 0.27287328243255615,
2070
+ "learning_rate": 4.922027290448343e-06,
2071
+ "loss": 0.0193,
2072
+ "step": 7275
2073
+ },
2074
+ {
2075
+ "epoch": 2.744360902255639,
2076
+ "grad_norm": 0.16126541793346405,
2077
+ "learning_rate": 4.747981063770537e-06,
2078
+ "loss": 0.0182,
2079
+ "step": 7300
2080
+ },
2081
+ {
2082
+ "epoch": 2.7537593984962405,
2083
+ "grad_norm": 0.406919926404953,
2084
+ "learning_rate": 4.573934837092732e-06,
2085
+ "loss": 0.0172,
2086
+ "step": 7325
2087
+ },
2088
+ {
2089
+ "epoch": 2.763157894736842,
2090
+ "grad_norm": 0.14170704782009125,
2091
+ "learning_rate": 4.399888610414926e-06,
2092
+ "loss": 0.0184,
2093
+ "step": 7350
2094
+ },
2095
+ {
2096
+ "epoch": 2.7725563909774436,
2097
+ "grad_norm": 0.1765240877866745,
2098
+ "learning_rate": 4.225842383737121e-06,
2099
+ "loss": 0.0163,
2100
+ "step": 7375
2101
+ },
2102
+ {
2103
+ "epoch": 2.781954887218045,
2104
+ "grad_norm": 0.2310091257095337,
2105
+ "learning_rate": 4.051796157059315e-06,
2106
+ "loss": 0.0178,
2107
+ "step": 7400
2108
+ },
2109
+ {
2110
+ "epoch": 2.7913533834586466,
2111
+ "grad_norm": 0.2542034983634949,
2112
+ "learning_rate": 3.877749930381509e-06,
2113
+ "loss": 0.0172,
2114
+ "step": 7425
2115
+ },
2116
+ {
2117
+ "epoch": 2.800751879699248,
2118
+ "grad_norm": 0.18098372220993042,
2119
+ "learning_rate": 3.7037037037037037e-06,
2120
+ "loss": 0.018,
2121
+ "step": 7450
2122
+ },
2123
+ {
2124
+ "epoch": 2.8101503759398496,
2125
+ "grad_norm": 0.1616281419992447,
2126
+ "learning_rate": 3.529657477025898e-06,
2127
+ "loss": 0.0179,
2128
+ "step": 7475
2129
+ },
2130
+ {
2131
+ "epoch": 2.819548872180451,
2132
+ "grad_norm": 0.2144605666399002,
2133
+ "learning_rate": 3.3556112503480925e-06,
2134
+ "loss": 0.0197,
2135
+ "step": 7500
2136
+ },
2137
+ {
2138
+ "epoch": 2.8289473684210527,
2139
+ "grad_norm": 0.24516059458255768,
2140
+ "learning_rate": 3.181565023670287e-06,
2141
+ "loss": 0.0182,
2142
+ "step": 7525
2143
+ },
2144
+ {
2145
+ "epoch": 2.838345864661654,
2146
+ "grad_norm": 0.13531754910945892,
2147
+ "learning_rate": 3.007518796992481e-06,
2148
+ "loss": 0.0177,
2149
+ "step": 7550
2150
+ },
2151
+ {
2152
+ "epoch": 2.8477443609022557,
2153
+ "grad_norm": 0.3076499104499817,
2154
+ "learning_rate": 2.8334725703146756e-06,
2155
+ "loss": 0.0181,
2156
+ "step": 7575
2157
+ },
2158
+ {
2159
+ "epoch": 2.857142857142857,
2160
+ "grad_norm": 0.17000478506088257,
2161
+ "learning_rate": 2.65942634363687e-06,
2162
+ "loss": 0.0187,
2163
+ "step": 7600
2164
+ },
2165
+ {
2166
+ "epoch": 2.8665413533834587,
2167
+ "grad_norm": 0.20252031087875366,
2168
+ "learning_rate": 2.4853801169590643e-06,
2169
+ "loss": 0.0175,
2170
+ "step": 7625
2171
+ },
2172
+ {
2173
+ "epoch": 2.8759398496240602,
2174
+ "grad_norm": 0.15144319832324982,
2175
+ "learning_rate": 2.3113338902812587e-06,
2176
+ "loss": 0.0194,
2177
+ "step": 7650
2178
+ },
2179
+ {
2180
+ "epoch": 2.8853383458646618,
2181
+ "grad_norm": 0.17275463044643402,
2182
+ "learning_rate": 2.137287663603453e-06,
2183
+ "loss": 0.0179,
2184
+ "step": 7675
2185
+ },
2186
+ {
2187
+ "epoch": 2.8947368421052633,
2188
+ "grad_norm": 0.2425609827041626,
2189
+ "learning_rate": 1.9632414369256474e-06,
2190
+ "loss": 0.0195,
2191
+ "step": 7700
2192
+ },
2193
+ {
2194
+ "epoch": 2.904135338345865,
2195
+ "grad_norm": 0.17550380527973175,
2196
+ "learning_rate": 1.7891952102478418e-06,
2197
+ "loss": 0.0188,
2198
+ "step": 7725
2199
+ },
2200
+ {
2201
+ "epoch": 2.9135338345864663,
2202
+ "grad_norm": 0.2620781362056732,
2203
+ "learning_rate": 1.6151489835700362e-06,
2204
+ "loss": 0.0185,
2205
+ "step": 7750
2206
+ },
2207
+ {
2208
+ "epoch": 2.922932330827068,
2209
+ "grad_norm": 0.4164513051509857,
2210
+ "learning_rate": 1.4411027568922305e-06,
2211
+ "loss": 0.0172,
2212
+ "step": 7775
2213
+ },
2214
+ {
2215
+ "epoch": 2.932330827067669,
2216
+ "grad_norm": 0.15615621209144592,
2217
+ "learning_rate": 1.2670565302144249e-06,
2218
+ "loss": 0.0217,
2219
+ "step": 7800
2220
+ },
2221
+ {
2222
+ "epoch": 2.9417293233082704,
2223
+ "grad_norm": 0.2062351256608963,
2224
+ "learning_rate": 1.0930103035366193e-06,
2225
+ "loss": 0.0209,
2226
+ "step": 7825
2227
+ },
2228
+ {
2229
+ "epoch": 2.951127819548872,
2230
+ "grad_norm": 0.17096047103405,
2231
+ "learning_rate": 9.189640768588137e-07,
2232
+ "loss": 0.0179,
2233
+ "step": 7850
2234
+ },
2235
+ {
2236
+ "epoch": 2.9605263157894735,
2237
+ "grad_norm": 0.15038737654685974,
2238
+ "learning_rate": 7.449178501810081e-07,
2239
+ "loss": 0.0187,
2240
+ "step": 7875
2241
+ },
2242
+ {
2243
+ "epoch": 2.969924812030075,
2244
+ "grad_norm": 0.22632552683353424,
2245
+ "learning_rate": 5.708716235032025e-07,
2246
+ "loss": 0.0189,
2247
+ "step": 7900
2248
+ },
2249
+ {
2250
+ "epoch": 2.9793233082706765,
2251
+ "grad_norm": 0.16149891912937164,
2252
+ "learning_rate": 3.9682539682539683e-07,
2253
+ "loss": 0.0208,
2254
+ "step": 7925
2255
+ },
2256
+ {
2257
+ "epoch": 2.988721804511278,
2258
+ "grad_norm": 0.46746957302093506,
2259
+ "learning_rate": 2.227791701475912e-07,
2260
+ "loss": 0.017,
2261
+ "step": 7950
2262
+ },
2263
+ {
2264
+ "epoch": 2.9981203007518795,
2265
+ "grad_norm": 0.4267791211605072,
2266
+ "learning_rate": 4.873294346978558e-08,
2267
+ "loss": 0.0184,
2268
+ "step": 7975
2269
+ },
2270
+ {
2271
+ "epoch": 3.0,
2272
+ "eval_explained_variance": 0.27679795026779175,
2273
+ "eval_loss": 0.020958250388503075,
2274
+ "eval_mae": 0.11261985450983047,
2275
+ "eval_mse": 0.020960183814167976,
2276
+ "eval_r2": 0.2767665982246399,
2277
+ "eval_rmse": 0.14477632338945473,
2278
+ "eval_runtime": 5.0302,
2279
+ "eval_samples_per_second": 4229.858,
2280
+ "eval_steps_per_second": 66.2,
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 sha256:752543b5675d1d70f0f42e51305fecf6bb6cd04ef74a5a0d44af49d2707621a4
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:f3d20966de2b80c0d322c8f91327db3506896c23f3500ba59b9622b8c8c5f710
3
+ size 267829484
runs/Feb22_11-28-18_r-samtuckervegan-autotrain-4xl4-hcv9q8jy-d0a1c-qe6mi/events.out.tfevents.1740223708.r-samtuckervegan-autotrain-4xl4-hcv9q8jy-d0a1c-qe6mi.201.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fffb463dcf6e97d9f48d0cf73020d6db2948098540ff877cbbcb567f783b1892
3
- size 58816
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:782d7dc4e8d6ada72928011a2a6179a6cc986292a656dfdcf6a493f6ff24e68e
3
+ size 74250
runs/Feb22_11-28-18_r-samtuckervegan-autotrain-4xl4-hcv9q8jy-d0a1c-qe6mi/events.out.tfevents.1740224486.r-samtuckervegan-autotrain-4xl4-hcv9q8jy-d0a1c-qe6mi.201.1 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5b063aaae5ca5c3fe771dd13e299884a30f117ea41b8377ed06d09463e794e8b
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 +1,58 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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": true,
45
+ "cls_token": "[CLS]",
46
+ "do_basic_tokenize": true,
47
+ "do_lower_case": true,
48
+ "extra_special_tokens": {},
49
+ "mask_token": "[MASK]",
50
+ "model_max_length": 512,
51
+ "never_split": null,
52
+ "pad_token": "[PAD]",
53
+ "sep_token": "[SEP]",
54
+ "strip_accents": null,
55
+ "tokenize_chinese_chars": true,
56
+ "tokenizer_class": "DistilBertTokenizer",
57
+ "unk_token": "[UNK]"
58
+ }
training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:752543b5675d1d70f0f42e51305fecf6bb6cd04ef74a5a0d44af49d2707621a4
3
+ size 5368
training_params.json ADDED
@@ -0,0 +1,30 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "data_path": "samtuckervegan/influence_predictions",
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": "influence",
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