commit files to HF hub
Browse files- added_tokens.json +18 -0
- config.json +44 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +25 -0
- tokenizer.json +0 -0
- tokenizer_config.json +14 -0
- trainer_state.json +838 -0
- training_args.bin +3 -0
- vocab.txt +0 -0
added_tokens.json
ADDED
@@ -0,0 +1,18 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"[ambiance-Negative]": 30526,
|
3 |
+
"[ambiance-Positive]": 30527,
|
4 |
+
"[ambiance-no majority]": 30529,
|
5 |
+
"[ambiance-unknown]": 30528,
|
6 |
+
"[food-Negative]": 30522,
|
7 |
+
"[food-Positive]": 30523,
|
8 |
+
"[food-no majority]": 30525,
|
9 |
+
"[food-unknown]": 30524,
|
10 |
+
"[noise-Negative]": 30530,
|
11 |
+
"[noise-Positive]": 30531,
|
12 |
+
"[noise-no majority]": 30533,
|
13 |
+
"[noise-unknown]": 30532,
|
14 |
+
"[service-Negative]": 30534,
|
15 |
+
"[service-Positive]": 30535,
|
16 |
+
"[service-no majority]": 30537,
|
17 |
+
"[service-unknown]": 30536
|
18 |
+
}
|
config.json
ADDED
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "CEBaB/lstm.CEBaB.sa.5-class.exclusive.seed_42",
|
3 |
+
"architectures": [
|
4 |
+
"LSTMForFactualCounterfactualSequenceClassification"
|
5 |
+
],
|
6 |
+
"attention_probs_dropout_prob": 0.1,
|
7 |
+
"bidirectional": true,
|
8 |
+
"classifier_dropout": null,
|
9 |
+
"finetuning_task": "opentable",
|
10 |
+
"gradient_checkpointing": false,
|
11 |
+
"hidden_act": "gelu",
|
12 |
+
"hidden_dropout_prob": 0.1,
|
13 |
+
"hidden_size": 300,
|
14 |
+
"id2label": {
|
15 |
+
"0": 0,
|
16 |
+
"1": 1,
|
17 |
+
"2": 2,
|
18 |
+
"3": 3,
|
19 |
+
"4": 4
|
20 |
+
},
|
21 |
+
"initializer_range": 0.02,
|
22 |
+
"intermediate_size": 3072,
|
23 |
+
"label2id": {
|
24 |
+
"0": 0,
|
25 |
+
"1": 1,
|
26 |
+
"2": 2,
|
27 |
+
"3": 3,
|
28 |
+
"4": 4
|
29 |
+
},
|
30 |
+
"layer_norm_eps": 1e-12,
|
31 |
+
"max_position_embeddings": 512,
|
32 |
+
"model_type": "bert",
|
33 |
+
"num_attention_heads": 12,
|
34 |
+
"num_hidden_layers": 1,
|
35 |
+
"pad_token_id": 0,
|
36 |
+
"position_embedding_type": "absolute",
|
37 |
+
"problem_type": "single_label_classification",
|
38 |
+
"torch_dtype": "float32",
|
39 |
+
"transformers_version": "4.21.1",
|
40 |
+
"type_vocab_size": 2,
|
41 |
+
"update_embeddings": false,
|
42 |
+
"use_cache": true,
|
43 |
+
"vocab_size": 30538
|
44 |
+
}
|
optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4ca7bf66b7a17f168cd5099808280e4e7c130b1043d50123a5be64f6ba617392
|
3 |
+
size 85600319
|
pytorch_model.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:55cee353b48b7db2efbd4f5b03639d23f3c4bbbc357ca15fea3f86b929f9868f
|
3 |
+
size 42799239
|
rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:dc577a2a46f5e1232200f84fc29051b0e73c7a087118c85edee8819fd6610e78
|
3 |
+
size 14503
|
scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:61539bf67c8a560752684944378e93d9e92efefe861be0e55081fb9c6527bbeb
|
3 |
+
size 623
|
special_tokens_map.json
ADDED
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"additional_special_tokens": [
|
3 |
+
"[food-Negative]",
|
4 |
+
"[food-Positive]",
|
5 |
+
"[food-unknown]",
|
6 |
+
"[food-no majority]",
|
7 |
+
"[ambiance-Negative]",
|
8 |
+
"[ambiance-Positive]",
|
9 |
+
"[ambiance-unknown]",
|
10 |
+
"[ambiance-no majority]",
|
11 |
+
"[noise-Negative]",
|
12 |
+
"[noise-Positive]",
|
13 |
+
"[noise-unknown]",
|
14 |
+
"[noise-no majority]",
|
15 |
+
"[service-Negative]",
|
16 |
+
"[service-Positive]",
|
17 |
+
"[service-unknown]",
|
18 |
+
"[service-no majority]"
|
19 |
+
],
|
20 |
+
"cls_token": "[CLS]",
|
21 |
+
"mask_token": "[MASK]",
|
22 |
+
"pad_token": "[PAD]",
|
23 |
+
"sep_token": "[SEP]",
|
24 |
+
"unk_token": "[UNK]"
|
25 |
+
}
|
tokenizer.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|
tokenizer_config.json
ADDED
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"cls_token": "[CLS]",
|
3 |
+
"do_lower_case": true,
|
4 |
+
"mask_token": "[MASK]",
|
5 |
+
"model_max_length": 512,
|
6 |
+
"name_or_path": "bert-base-uncased",
|
7 |
+
"pad_token": "[PAD]",
|
8 |
+
"sep_token": "[SEP]",
|
9 |
+
"special_tokens_map_file": null,
|
10 |
+
"strip_accents": null,
|
11 |
+
"tokenize_chinese_chars": true,
|
12 |
+
"tokenizer_class": "BertTokenizer",
|
13 |
+
"unk_token": "[UNK]"
|
14 |
+
}
|
trainer_state.json
ADDED
@@ -0,0 +1,838 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.4387560784816742,
|
3 |
+
"best_model_checkpoint": "model_output/e2e_opentable_5_way__19684-shot__seed-42__lstm/checkpoint-1250",
|
4 |
+
"epoch": 8.116883116883116,
|
5 |
+
"global_step": 1250,
|
6 |
+
"is_hyper_param_search": false,
|
7 |
+
"is_local_process_zero": true,
|
8 |
+
"is_world_process_zero": true,
|
9 |
+
"log_history": [
|
10 |
+
{
|
11 |
+
"epoch": 0.13,
|
12 |
+
"learning_rate": 4.9783362218370885e-05,
|
13 |
+
"loss": 20.6562,
|
14 |
+
"step": 20
|
15 |
+
},
|
16 |
+
{
|
17 |
+
"epoch": 0.26,
|
18 |
+
"learning_rate": 4.956672443674177e-05,
|
19 |
+
"loss": 18.0703,
|
20 |
+
"step": 40
|
21 |
+
},
|
22 |
+
{
|
23 |
+
"epoch": 0.32,
|
24 |
+
"eval_distillation_accuracy_counterfactual": 0.48717290918419703,
|
25 |
+
"eval_distillation_accuracy_factual": 0.9012314007183171,
|
26 |
+
"eval_distillation_f1_counterfactual": 0.4535140666639451,
|
27 |
+
"eval_distillation_f1_factual": 0.8957696500942338,
|
28 |
+
"eval_groundtruth_accuracy_counterfactual": 0.38558234992303747,
|
29 |
+
"eval_groundtruth_f1_counterfactual": 0.3510147662827972,
|
30 |
+
"eval_groundtruth_f1_factual": 0.5809634631420686,
|
31 |
+
"eval_icace_cosine": 0.5996042490005493,
|
32 |
+
"eval_icace_l2": 0.6879788637161255,
|
33 |
+
"eval_icace_normdiff": 0.6130202412605286,
|
34 |
+
"eval_loss": 18.372575759887695,
|
35 |
+
"eval_runtime": 4.1379,
|
36 |
+
"eval_samples_per_second": 942.017,
|
37 |
+
"eval_steps_per_second": 7.492,
|
38 |
+
"step": 50
|
39 |
+
},
|
40 |
+
{
|
41 |
+
"epoch": 0.39,
|
42 |
+
"learning_rate": 4.935008665511265e-05,
|
43 |
+
"loss": 16.8521,
|
44 |
+
"step": 60
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.52,
|
48 |
+
"learning_rate": 4.913344887348354e-05,
|
49 |
+
"loss": 15.1477,
|
50 |
+
"step": 80
|
51 |
+
},
|
52 |
+
{
|
53 |
+
"epoch": 0.65,
|
54 |
+
"learning_rate": 4.891681109185442e-05,
|
55 |
+
"loss": 13.5861,
|
56 |
+
"step": 100
|
57 |
+
},
|
58 |
+
{
|
59 |
+
"epoch": 0.65,
|
60 |
+
"eval_distillation_accuracy_counterfactual": 0.5200102616726526,
|
61 |
+
"eval_distillation_accuracy_factual": 0.8473576192919446,
|
62 |
+
"eval_distillation_f1_counterfactual": 0.4895334324691376,
|
63 |
+
"eval_distillation_f1_factual": 0.8392241931747526,
|
64 |
+
"eval_groundtruth_accuracy_counterfactual": 0.4238070805541303,
|
65 |
+
"eval_groundtruth_f1_counterfactual": 0.3924731874181019,
|
66 |
+
"eval_groundtruth_f1_factual": 0.5673025956236939,
|
67 |
+
"eval_icace_cosine": 0.5384137630462646,
|
68 |
+
"eval_icace_l2": 0.6406373977661133,
|
69 |
+
"eval_icace_normdiff": 0.5294525623321533,
|
70 |
+
"eval_loss": 13.979488372802734,
|
71 |
+
"eval_runtime": 4.4493,
|
72 |
+
"eval_samples_per_second": 876.093,
|
73 |
+
"eval_steps_per_second": 6.967,
|
74 |
+
"step": 100
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"epoch": 0.78,
|
78 |
+
"learning_rate": 4.8700173310225307e-05,
|
79 |
+
"loss": 12.2237,
|
80 |
+
"step": 120
|
81 |
+
},
|
82 |
+
{
|
83 |
+
"epoch": 0.91,
|
84 |
+
"learning_rate": 4.848353552859619e-05,
|
85 |
+
"loss": 11.2675,
|
86 |
+
"step": 140
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.97,
|
90 |
+
"eval_distillation_accuracy_counterfactual": 0.5726013340174448,
|
91 |
+
"eval_distillation_accuracy_factual": 0.8022062596203181,
|
92 |
+
"eval_distillation_f1_counterfactual": 0.5314200417069295,
|
93 |
+
"eval_distillation_f1_factual": 0.7853514931295182,
|
94 |
+
"eval_groundtruth_accuracy_counterfactual": 0.4917906618778861,
|
95 |
+
"eval_groundtruth_f1_counterfactual": 0.4494045856838918,
|
96 |
+
"eval_groundtruth_f1_factual": 0.5299931970242554,
|
97 |
+
"eval_icace_cosine": 0.4923843741416931,
|
98 |
+
"eval_icace_l2": 0.561776340007782,
|
99 |
+
"eval_icace_normdiff": 0.4008543789386749,
|
100 |
+
"eval_loss": 10.179123878479004,
|
101 |
+
"eval_runtime": 4.1977,
|
102 |
+
"eval_samples_per_second": 928.603,
|
103 |
+
"eval_steps_per_second": 7.385,
|
104 |
+
"step": 150
|
105 |
+
},
|
106 |
+
{
|
107 |
+
"epoch": 1.04,
|
108 |
+
"learning_rate": 4.826689774696707e-05,
|
109 |
+
"loss": 9.9178,
|
110 |
+
"step": 160
|
111 |
+
},
|
112 |
+
{
|
113 |
+
"epoch": 1.17,
|
114 |
+
"learning_rate": 4.8050259965337955e-05,
|
115 |
+
"loss": 9.1541,
|
116 |
+
"step": 180
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 1.3,
|
120 |
+
"learning_rate": 4.7833622183708845e-05,
|
121 |
+
"loss": 8.7158,
|
122 |
+
"step": 200
|
123 |
+
},
|
124 |
+
{
|
125 |
+
"epoch": 1.3,
|
126 |
+
"eval_distillation_accuracy_counterfactual": 0.5826064648537712,
|
127 |
+
"eval_distillation_accuracy_factual": 0.8129810159055926,
|
128 |
+
"eval_distillation_f1_counterfactual": 0.5429038638893491,
|
129 |
+
"eval_distillation_f1_factual": 0.7982170975645125,
|
130 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5082093381221139,
|
131 |
+
"eval_groundtruth_f1_counterfactual": 0.4622598550088801,
|
132 |
+
"eval_groundtruth_f1_factual": 0.53329740892104,
|
133 |
+
"eval_icace_cosine": 0.4688790440559387,
|
134 |
+
"eval_icace_l2": 0.538577139377594,
|
135 |
+
"eval_icace_normdiff": 0.3778749108314514,
|
136 |
+
"eval_loss": 9.010956764221191,
|
137 |
+
"eval_runtime": 4.0943,
|
138 |
+
"eval_samples_per_second": 952.058,
|
139 |
+
"eval_steps_per_second": 7.572,
|
140 |
+
"step": 200
|
141 |
+
},
|
142 |
+
{
|
143 |
+
"epoch": 1.43,
|
144 |
+
"learning_rate": 4.761698440207972e-05,
|
145 |
+
"loss": 8.5692,
|
146 |
+
"step": 220
|
147 |
+
},
|
148 |
+
{
|
149 |
+
"epoch": 1.56,
|
150 |
+
"learning_rate": 4.740034662045061e-05,
|
151 |
+
"loss": 8.0816,
|
152 |
+
"step": 240
|
153 |
+
},
|
154 |
+
{
|
155 |
+
"epoch": 1.62,
|
156 |
+
"eval_distillation_accuracy_counterfactual": 0.5928681375064135,
|
157 |
+
"eval_distillation_accuracy_factual": 0.8070805541303232,
|
158 |
+
"eval_distillation_f1_counterfactual": 0.5529708874649586,
|
159 |
+
"eval_distillation_f1_factual": 0.79399343360367,
|
160 |
+
"eval_groundtruth_accuracy_counterfactual": 0.517444843509492,
|
161 |
+
"eval_groundtruth_f1_counterfactual": 0.4725321895461628,
|
162 |
+
"eval_groundtruth_f1_factual": 0.5345191071930586,
|
163 |
+
"eval_icace_cosine": 0.46347475051879883,
|
164 |
+
"eval_icace_l2": 0.527389407157898,
|
165 |
+
"eval_icace_normdiff": 0.37002304196357727,
|
166 |
+
"eval_loss": 8.498320579528809,
|
167 |
+
"eval_runtime": 4.3399,
|
168 |
+
"eval_samples_per_second": 898.168,
|
169 |
+
"eval_steps_per_second": 7.143,
|
170 |
+
"step": 250
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 1.69,
|
174 |
+
"learning_rate": 4.7183708838821494e-05,
|
175 |
+
"loss": 8.4685,
|
176 |
+
"step": 260
|
177 |
+
},
|
178 |
+
{
|
179 |
+
"epoch": 1.82,
|
180 |
+
"learning_rate": 4.6967071057192376e-05,
|
181 |
+
"loss": 7.8803,
|
182 |
+
"step": 280
|
183 |
+
},
|
184 |
+
{
|
185 |
+
"epoch": 1.95,
|
186 |
+
"learning_rate": 4.675043327556326e-05,
|
187 |
+
"loss": 7.6873,
|
188 |
+
"step": 300
|
189 |
+
},
|
190 |
+
{
|
191 |
+
"epoch": 1.95,
|
192 |
+
"eval_distillation_accuracy_counterfactual": 0.6064648537711647,
|
193 |
+
"eval_distillation_accuracy_factual": 0.8178553104155978,
|
194 |
+
"eval_distillation_f1_counterfactual": 0.5629061661390706,
|
195 |
+
"eval_distillation_f1_factual": 0.8023836086835677,
|
196 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5279630579784504,
|
197 |
+
"eval_groundtruth_f1_counterfactual": 0.47789130421114673,
|
198 |
+
"eval_groundtruth_f1_factual": 0.5344257070862026,
|
199 |
+
"eval_icace_cosine": 0.45601892471313477,
|
200 |
+
"eval_icace_l2": 0.5144051909446716,
|
201 |
+
"eval_icace_normdiff": 0.3617022633552551,
|
202 |
+
"eval_loss": 8.192581176757812,
|
203 |
+
"eval_runtime": 5.0158,
|
204 |
+
"eval_samples_per_second": 777.137,
|
205 |
+
"eval_steps_per_second": 6.18,
|
206 |
+
"step": 300
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 2.08,
|
210 |
+
"learning_rate": 4.653379549393415e-05,
|
211 |
+
"loss": 7.5662,
|
212 |
+
"step": 320
|
213 |
+
},
|
214 |
+
{
|
215 |
+
"epoch": 2.21,
|
216 |
+
"learning_rate": 4.6317157712305025e-05,
|
217 |
+
"loss": 7.6061,
|
218 |
+
"step": 340
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 2.27,
|
222 |
+
"eval_distillation_accuracy_counterfactual": 0.6095433555669574,
|
223 |
+
"eval_distillation_accuracy_factual": 0.8217034376603386,
|
224 |
+
"eval_distillation_f1_counterfactual": 0.5684665160653969,
|
225 |
+
"eval_distillation_f1_factual": 0.8081531276248081,
|
226 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5356593124679323,
|
227 |
+
"eval_groundtruth_f1_counterfactual": 0.4879170474066199,
|
228 |
+
"eval_groundtruth_f1_factual": 0.5428429092164265,
|
229 |
+
"eval_icace_cosine": 0.45691362023353577,
|
230 |
+
"eval_icace_l2": 0.5149694681167603,
|
231 |
+
"eval_icace_normdiff": 0.3580142557621002,
|
232 |
+
"eval_loss": 8.089337348937988,
|
233 |
+
"eval_runtime": 4.1368,
|
234 |
+
"eval_samples_per_second": 942.268,
|
235 |
+
"eval_steps_per_second": 7.494,
|
236 |
+
"step": 350
|
237 |
+
},
|
238 |
+
{
|
239 |
+
"epoch": 2.34,
|
240 |
+
"learning_rate": 4.6100519930675915e-05,
|
241 |
+
"loss": 7.4725,
|
242 |
+
"step": 360
|
243 |
+
},
|
244 |
+
{
|
245 |
+
"epoch": 2.47,
|
246 |
+
"learning_rate": 4.58838821490468e-05,
|
247 |
+
"loss": 7.3287,
|
248 |
+
"step": 380
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 2.6,
|
252 |
+
"learning_rate": 4.566724436741768e-05,
|
253 |
+
"loss": 7.5374,
|
254 |
+
"step": 400
|
255 |
+
},
|
256 |
+
{
|
257 |
+
"epoch": 2.6,
|
258 |
+
"eval_distillation_accuracy_counterfactual": 0.6159569009748589,
|
259 |
+
"eval_distillation_accuracy_factual": 0.8309389430477168,
|
260 |
+
"eval_distillation_f1_counterfactual": 0.5691376817104438,
|
261 |
+
"eval_distillation_f1_factual": 0.8148479579609189,
|
262 |
+
"eval_groundtruth_accuracy_counterfactual": 0.539250897896357,
|
263 |
+
"eval_groundtruth_f1_counterfactual": 0.485360382747431,
|
264 |
+
"eval_groundtruth_f1_factual": 0.5317474369735481,
|
265 |
+
"eval_icace_cosine": 0.45355308055877686,
|
266 |
+
"eval_icace_l2": 0.5047734379768372,
|
267 |
+
"eval_icace_normdiff": 0.3515755236148834,
|
268 |
+
"eval_loss": 7.922365188598633,
|
269 |
+
"eval_runtime": 5.8925,
|
270 |
+
"eval_samples_per_second": 661.514,
|
271 |
+
"eval_steps_per_second": 5.261,
|
272 |
+
"step": 400
|
273 |
+
},
|
274 |
+
{
|
275 |
+
"epoch": 2.73,
|
276 |
+
"learning_rate": 4.5450606585788563e-05,
|
277 |
+
"loss": 7.1126,
|
278 |
+
"step": 420
|
279 |
+
},
|
280 |
+
{
|
281 |
+
"epoch": 2.86,
|
282 |
+
"learning_rate": 4.5233968804159446e-05,
|
283 |
+
"loss": 7.3885,
|
284 |
+
"step": 440
|
285 |
+
},
|
286 |
+
{
|
287 |
+
"epoch": 2.92,
|
288 |
+
"eval_distillation_accuracy_counterfactual": 0.6144176500769626,
|
289 |
+
"eval_distillation_accuracy_factual": 0.8327347357619292,
|
290 |
+
"eval_distillation_f1_counterfactual": 0.5729661409695295,
|
291 |
+
"eval_distillation_f1_factual": 0.8192940502978703,
|
292 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5407901487942535,
|
293 |
+
"eval_groundtruth_f1_counterfactual": 0.4946759059049534,
|
294 |
+
"eval_groundtruth_f1_factual": 0.5452799385658833,
|
295 |
+
"eval_icace_cosine": 0.45755502581596375,
|
296 |
+
"eval_icace_l2": 0.506329357624054,
|
297 |
+
"eval_icace_normdiff": 0.35191622376441956,
|
298 |
+
"eval_loss": 7.8961567878723145,
|
299 |
+
"eval_runtime": 4.1665,
|
300 |
+
"eval_samples_per_second": 935.55,
|
301 |
+
"eval_steps_per_second": 7.44,
|
302 |
+
"step": 450
|
303 |
+
},
|
304 |
+
{
|
305 |
+
"epoch": 2.99,
|
306 |
+
"learning_rate": 4.501733102253033e-05,
|
307 |
+
"loss": 7.0656,
|
308 |
+
"step": 460
|
309 |
+
},
|
310 |
+
{
|
311 |
+
"epoch": 3.12,
|
312 |
+
"learning_rate": 4.480069324090121e-05,
|
313 |
+
"loss": 7.16,
|
314 |
+
"step": 480
|
315 |
+
},
|
316 |
+
{
|
317 |
+
"epoch": 3.25,
|
318 |
+
"learning_rate": 4.45840554592721e-05,
|
319 |
+
"loss": 6.7453,
|
320 |
+
"step": 500
|
321 |
+
},
|
322 |
+
{
|
323 |
+
"epoch": 3.25,
|
324 |
+
"eval_distillation_accuracy_counterfactual": 0.612365315546434,
|
325 |
+
"eval_distillation_accuracy_factual": 0.8273473576192919,
|
326 |
+
"eval_distillation_f1_counterfactual": 0.5700286878042866,
|
327 |
+
"eval_distillation_f1_factual": 0.8116751283042352,
|
328 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5461775269368907,
|
329 |
+
"eval_groundtruth_f1_counterfactual": 0.4982594429639036,
|
330 |
+
"eval_groundtruth_f1_factual": 0.5426844401130816,
|
331 |
+
"eval_icace_cosine": 0.4582451283931732,
|
332 |
+
"eval_icace_l2": 0.5082744359970093,
|
333 |
+
"eval_icace_normdiff": 0.3508155047893524,
|
334 |
+
"eval_loss": 7.861159801483154,
|
335 |
+
"eval_runtime": 6.4679,
|
336 |
+
"eval_samples_per_second": 602.668,
|
337 |
+
"eval_steps_per_second": 4.793,
|
338 |
+
"step": 500
|
339 |
+
},
|
340 |
+
{
|
341 |
+
"epoch": 3.38,
|
342 |
+
"learning_rate": 4.436741767764298e-05,
|
343 |
+
"loss": 6.9515,
|
344 |
+
"step": 520
|
345 |
+
},
|
346 |
+
{
|
347 |
+
"epoch": 3.51,
|
348 |
+
"learning_rate": 4.415077989601387e-05,
|
349 |
+
"loss": 6.6913,
|
350 |
+
"step": 540
|
351 |
+
},
|
352 |
+
{
|
353 |
+
"epoch": 3.57,
|
354 |
+
"eval_distillation_accuracy_counterfactual": 0.6228835300153925,
|
355 |
+
"eval_distillation_accuracy_factual": 0.823242688558235,
|
356 |
+
"eval_distillation_f1_counterfactual": 0.5797100592542744,
|
357 |
+
"eval_distillation_f1_factual": 0.8054169495093865,
|
358 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5515649050795279,
|
359 |
+
"eval_groundtruth_f1_counterfactual": 0.500957403465055,
|
360 |
+
"eval_groundtruth_f1_factual": 0.5372992623096391,
|
361 |
+
"eval_icace_cosine": 0.4541945159435272,
|
362 |
+
"eval_icace_l2": 0.5000237226486206,
|
363 |
+
"eval_icace_normdiff": 0.3448871374130249,
|
364 |
+
"eval_loss": 7.762178897857666,
|
365 |
+
"eval_runtime": 4.1547,
|
366 |
+
"eval_samples_per_second": 938.211,
|
367 |
+
"eval_steps_per_second": 7.461,
|
368 |
+
"step": 550
|
369 |
+
},
|
370 |
+
{
|
371 |
+
"epoch": 3.64,
|
372 |
+
"learning_rate": 4.393414211438475e-05,
|
373 |
+
"loss": 6.8003,
|
374 |
+
"step": 560
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 3.77,
|
378 |
+
"learning_rate": 4.371750433275563e-05,
|
379 |
+
"loss": 7.0336,
|
380 |
+
"step": 580
|
381 |
+
},
|
382 |
+
{
|
383 |
+
"epoch": 3.9,
|
384 |
+
"learning_rate": 4.3500866551126516e-05,
|
385 |
+
"loss": 6.8497,
|
386 |
+
"step": 600
|
387 |
+
},
|
388 |
+
{
|
389 |
+
"epoch": 3.9,
|
390 |
+
"eval_distillation_accuracy_counterfactual": 0.6177526936890713,
|
391 |
+
"eval_distillation_accuracy_factual": 0.8150333504361211,
|
392 |
+
"eval_distillation_f1_counterfactual": 0.5812759881530707,
|
393 |
+
"eval_distillation_f1_factual": 0.8048351370604768,
|
394 |
+
"eval_groundtruth_accuracy_counterfactual": 0.543099025141098,
|
395 |
+
"eval_groundtruth_f1_counterfactual": 0.502597773992042,
|
396 |
+
"eval_groundtruth_f1_factual": 0.5413490240682356,
|
397 |
+
"eval_icace_cosine": 0.4559078812599182,
|
398 |
+
"eval_icace_l2": 0.5052775740623474,
|
399 |
+
"eval_icace_normdiff": 0.34785109758377075,
|
400 |
+
"eval_loss": 7.774956703186035,
|
401 |
+
"eval_runtime": 5.8188,
|
402 |
+
"eval_samples_per_second": 669.896,
|
403 |
+
"eval_steps_per_second": 5.328,
|
404 |
+
"step": 600
|
405 |
+
},
|
406 |
+
{
|
407 |
+
"epoch": 4.03,
|
408 |
+
"learning_rate": 4.3284228769497406e-05,
|
409 |
+
"loss": 6.4975,
|
410 |
+
"step": 620
|
411 |
+
},
|
412 |
+
{
|
413 |
+
"epoch": 4.16,
|
414 |
+
"learning_rate": 4.306759098786828e-05,
|
415 |
+
"loss": 6.5642,
|
416 |
+
"step": 640
|
417 |
+
},
|
418 |
+
{
|
419 |
+
"epoch": 4.22,
|
420 |
+
"eval_distillation_accuracy_counterfactual": 0.6136480246280144,
|
421 |
+
"eval_distillation_accuracy_factual": 0.8183683940482298,
|
422 |
+
"eval_distillation_f1_counterfactual": 0.580211267327247,
|
423 |
+
"eval_distillation_f1_factual": 0.8068432506573439,
|
424 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5454079014879425,
|
425 |
+
"eval_groundtruth_f1_counterfactual": 0.5053418375187448,
|
426 |
+
"eval_groundtruth_f1_factual": 0.5441340640231574,
|
427 |
+
"eval_icace_cosine": 0.4580763578414917,
|
428 |
+
"eval_icace_l2": 0.5093169212341309,
|
429 |
+
"eval_icace_normdiff": 0.3487666845321655,
|
430 |
+
"eval_loss": 7.786745548248291,
|
431 |
+
"eval_runtime": 4.0712,
|
432 |
+
"eval_samples_per_second": 957.454,
|
433 |
+
"eval_steps_per_second": 7.614,
|
434 |
+
"step": 650
|
435 |
+
},
|
436 |
+
{
|
437 |
+
"epoch": 4.29,
|
438 |
+
"learning_rate": 4.285095320623917e-05,
|
439 |
+
"loss": 6.4044,
|
440 |
+
"step": 660
|
441 |
+
},
|
442 |
+
{
|
443 |
+
"epoch": 4.42,
|
444 |
+
"learning_rate": 4.2634315424610055e-05,
|
445 |
+
"loss": 6.4254,
|
446 |
+
"step": 680
|
447 |
+
},
|
448 |
+
{
|
449 |
+
"epoch": 4.55,
|
450 |
+
"learning_rate": 4.241767764298094e-05,
|
451 |
+
"loss": 6.393,
|
452 |
+
"step": 700
|
453 |
+
},
|
454 |
+
{
|
455 |
+
"epoch": 4.55,
|
456 |
+
"eval_distillation_accuracy_counterfactual": 0.6182657773217034,
|
457 |
+
"eval_distillation_accuracy_factual": 0.8160595177013853,
|
458 |
+
"eval_distillation_f1_counterfactual": 0.5808522807878681,
|
459 |
+
"eval_distillation_f1_factual": 0.8040768770971043,
|
460 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5454079014879425,
|
461 |
+
"eval_groundtruth_f1_counterfactual": 0.5006294773113318,
|
462 |
+
"eval_groundtruth_f1_factual": 0.5459691019650267,
|
463 |
+
"eval_icace_cosine": 0.4525621235370636,
|
464 |
+
"eval_icace_l2": 0.5032183527946472,
|
465 |
+
"eval_icace_normdiff": 0.3443147540092468,
|
466 |
+
"eval_loss": 7.749052047729492,
|
467 |
+
"eval_runtime": 5.3047,
|
468 |
+
"eval_samples_per_second": 734.818,
|
469 |
+
"eval_steps_per_second": 5.844,
|
470 |
+
"step": 700
|
471 |
+
},
|
472 |
+
{
|
473 |
+
"epoch": 4.68,
|
474 |
+
"learning_rate": 4.220103986135182e-05,
|
475 |
+
"loss": 6.3202,
|
476 |
+
"step": 720
|
477 |
+
},
|
478 |
+
{
|
479 |
+
"epoch": 4.81,
|
480 |
+
"learning_rate": 4.198440207972271e-05,
|
481 |
+
"loss": 6.6167,
|
482 |
+
"step": 740
|
483 |
+
},
|
484 |
+
{
|
485 |
+
"epoch": 4.87,
|
486 |
+
"eval_distillation_accuracy_counterfactual": 0.6182657773217034,
|
487 |
+
"eval_distillation_accuracy_factual": 0.8122113904566445,
|
488 |
+
"eval_distillation_f1_counterfactual": 0.5818063896260541,
|
489 |
+
"eval_distillation_f1_factual": 0.7988404215810908,
|
490 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5484864032837352,
|
491 |
+
"eval_groundtruth_f1_counterfactual": 0.5052584515576525,
|
492 |
+
"eval_groundtruth_f1_factual": 0.5394581551303677,
|
493 |
+
"eval_icace_cosine": 0.4518362283706665,
|
494 |
+
"eval_icace_l2": 0.501703679561615,
|
495 |
+
"eval_icace_normdiff": 0.3422965407371521,
|
496 |
+
"eval_loss": 7.7101054191589355,
|
497 |
+
"eval_runtime": 6.8348,
|
498 |
+
"eval_samples_per_second": 570.314,
|
499 |
+
"eval_steps_per_second": 4.536,
|
500 |
+
"step": 750
|
501 |
+
},
|
502 |
+
{
|
503 |
+
"epoch": 4.94,
|
504 |
+
"learning_rate": 4.1767764298093586e-05,
|
505 |
+
"loss": 6.4981,
|
506 |
+
"step": 760
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 5.06,
|
510 |
+
"learning_rate": 4.1551126516464476e-05,
|
511 |
+
"loss": 6.2109,
|
512 |
+
"step": 780
|
513 |
+
},
|
514 |
+
{
|
515 |
+
"epoch": 5.19,
|
516 |
+
"learning_rate": 4.133448873483536e-05,
|
517 |
+
"loss": 6.5075,
|
518 |
+
"step": 800
|
519 |
+
},
|
520 |
+
{
|
521 |
+
"epoch": 5.19,
|
522 |
+
"eval_distillation_accuracy_counterfactual": 0.6180092355053873,
|
523 |
+
"eval_distillation_accuracy_factual": 0.8142637249871729,
|
524 |
+
"eval_distillation_f1_counterfactual": 0.5811576499624775,
|
525 |
+
"eval_distillation_f1_factual": 0.8023242774635648,
|
526 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5443817342226783,
|
527 |
+
"eval_groundtruth_f1_counterfactual": 0.5018080158011544,
|
528 |
+
"eval_groundtruth_f1_factual": 0.5428002033480686,
|
529 |
+
"eval_icace_cosine": 0.45876333117485046,
|
530 |
+
"eval_icace_l2": 0.505964457988739,
|
531 |
+
"eval_icace_normdiff": 0.34437206387519836,
|
532 |
+
"eval_loss": 7.784791469573975,
|
533 |
+
"eval_runtime": 4.1015,
|
534 |
+
"eval_samples_per_second": 950.379,
|
535 |
+
"eval_steps_per_second": 7.558,
|
536 |
+
"step": 800
|
537 |
+
},
|
538 |
+
{
|
539 |
+
"epoch": 5.32,
|
540 |
+
"learning_rate": 4.111785095320624e-05,
|
541 |
+
"loss": 6.1528,
|
542 |
+
"step": 820
|
543 |
+
},
|
544 |
+
{
|
545 |
+
"epoch": 5.45,
|
546 |
+
"learning_rate": 4.0901213171577124e-05,
|
547 |
+
"loss": 6.2869,
|
548 |
+
"step": 840
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 5.52,
|
552 |
+
"eval_distillation_accuracy_counterfactual": 0.6185223191380195,
|
553 |
+
"eval_distillation_accuracy_factual": 0.8193945613134941,
|
554 |
+
"eval_distillation_f1_counterfactual": 0.5836247405075541,
|
555 |
+
"eval_distillation_f1_factual": 0.8091365156198146,
|
556 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5515649050795279,
|
557 |
+
"eval_groundtruth_f1_counterfactual": 0.510664962216415,
|
558 |
+
"eval_groundtruth_f1_factual": 0.5423106601987745,
|
559 |
+
"eval_icace_cosine": 0.45199427008628845,
|
560 |
+
"eval_icace_l2": 0.5019307732582092,
|
561 |
+
"eval_icace_normdiff": 0.3425298035144806,
|
562 |
+
"eval_loss": 7.68693733215332,
|
563 |
+
"eval_runtime": 4.883,
|
564 |
+
"eval_samples_per_second": 798.276,
|
565 |
+
"eval_steps_per_second": 6.349,
|
566 |
+
"step": 850
|
567 |
+
},
|
568 |
+
{
|
569 |
+
"epoch": 5.58,
|
570 |
+
"learning_rate": 4.068457538994801e-05,
|
571 |
+
"loss": 6.0441,
|
572 |
+
"step": 860
|
573 |
+
},
|
574 |
+
{
|
575 |
+
"epoch": 5.71,
|
576 |
+
"learning_rate": 4.04679376083189e-05,
|
577 |
+
"loss": 6.4014,
|
578 |
+
"step": 880
|
579 |
+
},
|
580 |
+
{
|
581 |
+
"epoch": 5.84,
|
582 |
+
"learning_rate": 4.025129982668977e-05,
|
583 |
+
"loss": 6.0569,
|
584 |
+
"step": 900
|
585 |
+
},
|
586 |
+
{
|
587 |
+
"epoch": 5.84,
|
588 |
+
"eval_distillation_accuracy_counterfactual": 0.6226269881990765,
|
589 |
+
"eval_distillation_accuracy_factual": 0.8114417650076963,
|
590 |
+
"eval_distillation_f1_counterfactual": 0.5854768643927275,
|
591 |
+
"eval_distillation_f1_factual": 0.7973005853765658,
|
592 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5461775269368907,
|
593 |
+
"eval_groundtruth_f1_counterfactual": 0.502247129912884,
|
594 |
+
"eval_groundtruth_f1_factual": 0.5450537518673406,
|
595 |
+
"eval_icace_cosine": 0.44710850715637207,
|
596 |
+
"eval_icace_l2": 0.4970768690109253,
|
597 |
+
"eval_icace_normdiff": 0.34042122960090637,
|
598 |
+
"eval_loss": 7.644977569580078,
|
599 |
+
"eval_runtime": 4.2109,
|
600 |
+
"eval_samples_per_second": 925.687,
|
601 |
+
"eval_steps_per_second": 7.362,
|
602 |
+
"step": 900
|
603 |
+
},
|
604 |
+
{
|
605 |
+
"epoch": 5.97,
|
606 |
+
"learning_rate": 4.003466204506066e-05,
|
607 |
+
"loss": 5.9713,
|
608 |
+
"step": 920
|
609 |
+
},
|
610 |
+
{
|
611 |
+
"epoch": 6.1,
|
612 |
+
"learning_rate": 3.9818024263431546e-05,
|
613 |
+
"loss": 6.0174,
|
614 |
+
"step": 940
|
615 |
+
},
|
616 |
+
{
|
617 |
+
"epoch": 6.17,
|
618 |
+
"eval_distillation_accuracy_counterfactual": 0.6180092355053873,
|
619 |
+
"eval_distillation_accuracy_factual": 0.8065674704976912,
|
620 |
+
"eval_distillation_f1_counterfactual": 0.5841502017308554,
|
621 |
+
"eval_distillation_f1_factual": 0.7977275175828281,
|
622 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5461775269368907,
|
623 |
+
"eval_groundtruth_f1_counterfactual": 0.506203917901584,
|
624 |
+
"eval_groundtruth_f1_factual": 0.5465942326831105,
|
625 |
+
"eval_icace_cosine": 0.45468205213546753,
|
626 |
+
"eval_icace_l2": 0.5059431791305542,
|
627 |
+
"eval_icace_normdiff": 0.3447358012199402,
|
628 |
+
"eval_loss": 7.76542854309082,
|
629 |
+
"eval_runtime": 4.3013,
|
630 |
+
"eval_samples_per_second": 906.232,
|
631 |
+
"eval_steps_per_second": 7.207,
|
632 |
+
"step": 950
|
633 |
+
},
|
634 |
+
{
|
635 |
+
"epoch": 6.23,
|
636 |
+
"learning_rate": 3.960138648180243e-05,
|
637 |
+
"loss": 6.0261,
|
638 |
+
"step": 960
|
639 |
+
},
|
640 |
+
{
|
641 |
+
"epoch": 6.36,
|
642 |
+
"learning_rate": 3.938474870017331e-05,
|
643 |
+
"loss": 6.0645,
|
644 |
+
"step": 980
|
645 |
+
},
|
646 |
+
{
|
647 |
+
"epoch": 6.49,
|
648 |
+
"learning_rate": 3.91681109185442e-05,
|
649 |
+
"loss": 6.0191,
|
650 |
+
"step": 1000
|
651 |
+
},
|
652 |
+
{
|
653 |
+
"epoch": 6.49,
|
654 |
+
"eval_distillation_accuracy_counterfactual": 0.620318111852232,
|
655 |
+
"eval_distillation_accuracy_factual": 0.8083632632119035,
|
656 |
+
"eval_distillation_f1_counterfactual": 0.5830477001223874,
|
657 |
+
"eval_distillation_f1_factual": 0.7963583724748586,
|
658 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5531041559774243,
|
659 |
+
"eval_groundtruth_f1_counterfactual": 0.5079499434851045,
|
660 |
+
"eval_groundtruth_f1_factual": 0.5433906032493001,
|
661 |
+
"eval_icace_cosine": 0.44436362385749817,
|
662 |
+
"eval_icace_l2": 0.49828970432281494,
|
663 |
+
"eval_icace_normdiff": 0.33891886472702026,
|
664 |
+
"eval_loss": 7.664845943450928,
|
665 |
+
"eval_runtime": 4.1469,
|
666 |
+
"eval_samples_per_second": 939.989,
|
667 |
+
"eval_steps_per_second": 7.476,
|
668 |
+
"step": 1000
|
669 |
+
},
|
670 |
+
{
|
671 |
+
"epoch": 6.62,
|
672 |
+
"learning_rate": 3.895147313691508e-05,
|
673 |
+
"loss": 5.9454,
|
674 |
+
"step": 1020
|
675 |
+
},
|
676 |
+
{
|
677 |
+
"epoch": 6.75,
|
678 |
+
"learning_rate": 3.873483535528597e-05,
|
679 |
+
"loss": 6.0488,
|
680 |
+
"step": 1040
|
681 |
+
},
|
682 |
+
{
|
683 |
+
"epoch": 6.82,
|
684 |
+
"eval_distillation_accuracy_counterfactual": 0.6228835300153925,
|
685 |
+
"eval_distillation_accuracy_factual": 0.810415597742432,
|
686 |
+
"eval_distillation_f1_counterfactual": 0.5845336847422479,
|
687 |
+
"eval_distillation_f1_factual": 0.7976470172715806,
|
688 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5525910723447922,
|
689 |
+
"eval_groundtruth_f1_counterfactual": 0.5059005142035983,
|
690 |
+
"eval_groundtruth_f1_factual": 0.540178117547913,
|
691 |
+
"eval_icace_cosine": 0.44367682933807373,
|
692 |
+
"eval_icace_l2": 0.49686795473098755,
|
693 |
+
"eval_icace_normdiff": 0.3366642892360687,
|
694 |
+
"eval_loss": 7.668118476867676,
|
695 |
+
"eval_runtime": 4.3685,
|
696 |
+
"eval_samples_per_second": 892.297,
|
697 |
+
"eval_steps_per_second": 7.096,
|
698 |
+
"step": 1050
|
699 |
+
},
|
700 |
+
{
|
701 |
+
"epoch": 6.88,
|
702 |
+
"learning_rate": 3.851819757365685e-05,
|
703 |
+
"loss": 5.9853,
|
704 |
+
"step": 1060
|
705 |
+
},
|
706 |
+
{
|
707 |
+
"epoch": 7.01,
|
708 |
+
"learning_rate": 3.830155979202773e-05,
|
709 |
+
"loss": 5.7619,
|
710 |
+
"step": 1080
|
711 |
+
},
|
712 |
+
{
|
713 |
+
"epoch": 7.14,
|
714 |
+
"learning_rate": 3.8084922010398616e-05,
|
715 |
+
"loss": 5.5946,
|
716 |
+
"step": 1100
|
717 |
+
},
|
718 |
+
{
|
719 |
+
"epoch": 7.14,
|
720 |
+
"eval_distillation_accuracy_counterfactual": 0.6200615700359159,
|
721 |
+
"eval_distillation_accuracy_factual": 0.8006670087224218,
|
722 |
+
"eval_distillation_f1_counterfactual": 0.5850719981039058,
|
723 |
+
"eval_distillation_f1_factual": 0.7887892378887658,
|
724 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5500256541816316,
|
725 |
+
"eval_groundtruth_f1_counterfactual": 0.50798102067776,
|
726 |
+
"eval_groundtruth_f1_factual": 0.5514350312905094,
|
727 |
+
"eval_icace_cosine": 0.445857971906662,
|
728 |
+
"eval_icace_l2": 0.5012825131416321,
|
729 |
+
"eval_icace_normdiff": 0.34079626202583313,
|
730 |
+
"eval_loss": 7.738803863525391,
|
731 |
+
"eval_runtime": 4.1651,
|
732 |
+
"eval_samples_per_second": 935.867,
|
733 |
+
"eval_steps_per_second": 7.443,
|
734 |
+
"step": 1100
|
735 |
+
},
|
736 |
+
{
|
737 |
+
"epoch": 7.27,
|
738 |
+
"learning_rate": 3.78682842287695e-05,
|
739 |
+
"loss": 5.8864,
|
740 |
+
"step": 1120
|
741 |
+
},
|
742 |
+
{
|
743 |
+
"epoch": 7.4,
|
744 |
+
"learning_rate": 3.765164644714038e-05,
|
745 |
+
"loss": 5.8432,
|
746 |
+
"step": 1140
|
747 |
+
},
|
748 |
+
{
|
749 |
+
"epoch": 7.47,
|
750 |
+
"eval_distillation_accuracy_counterfactual": 0.6213442791174961,
|
751 |
+
"eval_distillation_accuracy_factual": 0.7993842996408415,
|
752 |
+
"eval_distillation_f1_counterfactual": 0.5819867386627711,
|
753 |
+
"eval_distillation_f1_factual": 0.7847982271637772,
|
754 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5515649050795279,
|
755 |
+
"eval_groundtruth_f1_counterfactual": 0.5040470994435106,
|
756 |
+
"eval_groundtruth_f1_factual": 0.5443687788777846,
|
757 |
+
"eval_icace_cosine": 0.4456865191459656,
|
758 |
+
"eval_icace_l2": 0.4975723624229431,
|
759 |
+
"eval_icace_normdiff": 0.33811891078948975,
|
760 |
+
"eval_loss": 7.686666011810303,
|
761 |
+
"eval_runtime": 4.2581,
|
762 |
+
"eval_samples_per_second": 915.424,
|
763 |
+
"eval_steps_per_second": 7.28,
|
764 |
+
"step": 1150
|
765 |
+
},
|
766 |
+
{
|
767 |
+
"epoch": 7.53,
|
768 |
+
"learning_rate": 3.7435008665511264e-05,
|
769 |
+
"loss": 5.9408,
|
770 |
+
"step": 1160
|
771 |
+
},
|
772 |
+
{
|
773 |
+
"epoch": 7.66,
|
774 |
+
"learning_rate": 3.7218370883882154e-05,
|
775 |
+
"loss": 5.9549,
|
776 |
+
"step": 1180
|
777 |
+
},
|
778 |
+
{
|
779 |
+
"epoch": 7.79,
|
780 |
+
"learning_rate": 3.700173310225303e-05,
|
781 |
+
"loss": 5.5485,
|
782 |
+
"step": 1200
|
783 |
+
},
|
784 |
+
{
|
785 |
+
"epoch": 7.79,
|
786 |
+
"eval_distillation_accuracy_counterfactual": 0.6187788609543355,
|
787 |
+
"eval_distillation_accuracy_factual": 0.7981015905592611,
|
788 |
+
"eval_distillation_f1_counterfactual": 0.5834765365950793,
|
789 |
+
"eval_distillation_f1_factual": 0.7884169448594356,
|
790 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5502821959979477,
|
791 |
+
"eval_groundtruth_f1_counterfactual": 0.5082397669474391,
|
792 |
+
"eval_groundtruth_f1_factual": 0.5480615734783056,
|
793 |
+
"eval_icace_cosine": 0.45296600461006165,
|
794 |
+
"eval_icace_l2": 0.5046243667602539,
|
795 |
+
"eval_icace_normdiff": 0.34213510155677795,
|
796 |
+
"eval_loss": 7.746112823486328,
|
797 |
+
"eval_runtime": 4.1823,
|
798 |
+
"eval_samples_per_second": 932.018,
|
799 |
+
"eval_steps_per_second": 7.412,
|
800 |
+
"step": 1200
|
801 |
+
},
|
802 |
+
{
|
803 |
+
"epoch": 7.92,
|
804 |
+
"learning_rate": 3.678509532062392e-05,
|
805 |
+
"loss": 5.851,
|
806 |
+
"step": 1220
|
807 |
+
},
|
808 |
+
{
|
809 |
+
"epoch": 8.05,
|
810 |
+
"learning_rate": 3.65684575389948e-05,
|
811 |
+
"loss": 5.6669,
|
812 |
+
"step": 1240
|
813 |
+
},
|
814 |
+
{
|
815 |
+
"epoch": 8.12,
|
816 |
+
"eval_distillation_accuracy_counterfactual": 0.6236531554643406,
|
817 |
+
"eval_distillation_accuracy_factual": 0.8022062596203181,
|
818 |
+
"eval_distillation_f1_counterfactual": 0.5840941091107561,
|
819 |
+
"eval_distillation_f1_factual": 0.7870917202572578,
|
820 |
+
"eval_groundtruth_accuracy_counterfactual": 0.5556695741405849,
|
821 |
+
"eval_groundtruth_f1_counterfactual": 0.5089342422895711,
|
822 |
+
"eval_groundtruth_f1_factual": 0.5380085766933975,
|
823 |
+
"eval_icace_cosine": 0.4387560784816742,
|
824 |
+
"eval_icace_l2": 0.49367281794548035,
|
825 |
+
"eval_icace_normdiff": 0.33387720584869385,
|
826 |
+
"eval_loss": 7.6664533615112305,
|
827 |
+
"eval_runtime": 4.3535,
|
828 |
+
"eval_samples_per_second": 895.377,
|
829 |
+
"eval_steps_per_second": 7.121,
|
830 |
+
"step": 1250
|
831 |
+
}
|
832 |
+
],
|
833 |
+
"max_steps": 4616,
|
834 |
+
"num_train_epochs": 30,
|
835 |
+
"total_flos": 188664237047808.0,
|
836 |
+
"trial_name": null,
|
837 |
+
"trial_params": null
|
838 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:d757dde7d1b424fd93374400308effe2a857376e5320647a1d39b7f57d2feba1
|
3 |
+
size 3439
|
vocab.txt
ADDED
The diff for this file is too large to render.
See raw diff
|
|