Hugging Face's logo

bobox
/
DeBERTa-small-ST-v1-test

PyTorch
TensorBoard
deberta-v2
Model card Files Metrics Community
bobox commited on
Commit
6b2e73a
•
1 Parent(s): 6d2927f

Training in progress, step 12, checkpoint

Browse files
Files changed (17) hide show
  1. checkpoint-12/1_Pooling/config.json +10 -0
  2. checkpoint-12/README.md +0 -0
  3. checkpoint-12/added_tokens.json +3 -0
  4. checkpoint-12/config.json +35 -0
  5. checkpoint-12/config_sentence_transformers.json +10 -0
  6. checkpoint-12/modules.json +14 -0
  7. checkpoint-12/optimizer.pt +3 -0
  8. checkpoint-12/pytorch_model.bin +3 -0
  9. checkpoint-12/rng_state.pth +3 -0
  10. checkpoint-12/scheduler.pt +3 -0
  11. checkpoint-12/sentence_bert_config.json +4 -0
  12. checkpoint-12/special_tokens_map.json +15 -0
  13. checkpoint-12/spm.model +3 -0
  14. checkpoint-12/tokenizer.json +0 -0
  15. checkpoint-12/tokenizer_config.json +58 -0
  16. checkpoint-12/trainer_state.json +999 -0
  17. checkpoint-12/training_args.bin +3 -0
checkpoint-12/1_Pooling/config.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "word_embedding_dimension": 768,
3
+ "pooling_mode_cls_token": false,
4
+ "pooling_mode_mean_tokens": true,
5
+ "pooling_mode_max_tokens": false,
6
+ "pooling_mode_mean_sqrt_len_tokens": false,
7
+ "pooling_mode_weightedmean_tokens": false,
8
+ "pooling_mode_lasttoken": false,
9
+ "include_prompt": true
10
+ }
checkpoint-12/README.md ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-12/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ {
2
+ "[MASK]": 128000
3
+ }
checkpoint-12/config.json ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "_name_or_path": "microsoft/deberta-v3-small",
3
+ "architectures": [
4
+ "DebertaV2Model"
5
+ ],
6
+ "attention_probs_dropout_prob": 0.1,
7
+ "hidden_act": "gelu",
8
+ "hidden_dropout_prob": 0.1,
9
+ "hidden_size": 768,
10
+ "initializer_range": 0.02,
11
+ "intermediate_size": 3072,
12
+ "layer_norm_eps": 1e-07,
13
+ "max_position_embeddings": 512,
14
+ "max_relative_positions": -1,
15
+ "model_type": "deberta-v2",
16
+ "norm_rel_ebd": "layer_norm",
17
+ "num_attention_heads": 12,
18
+ "num_hidden_layers": 6,
19
+ "pad_token_id": 0,
20
+ "pooler_dropout": 0,
21
+ "pooler_hidden_act": "gelu",
22
+ "pooler_hidden_size": 768,
23
+ "pos_att_type": [
24
+ "p2c",
25
+ "c2p"
26
+ ],
27
+ "position_biased_input": false,
28
+ "position_buckets": 256,
29
+ "relative_attention": true,
30
+ "share_att_key": true,
31
+ "torch_dtype": "float32",
32
+ "transformers_version": "4.42.3",
33
+ "type_vocab_size": 0,
34
+ "vocab_size": 128100
35
+ }
checkpoint-12/config_sentence_transformers.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "3.0.1",
4
+ "transformers": "4.42.3",
5
+ "pytorch": "2.1.2"
6
+ },
7
+ "prompts": {},
8
+ "default_prompt_name": null,
9
+ "similarity_fn_name": null
10
+ }
checkpoint-12/modules.json ADDED
@@ -0,0 +1,14 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "idx": 0,
4
+ "name": "0",
5
+ "path": "",
6
+ "type": "sentence_transformers.models.Transformer"
7
+ },
8
+ {
9
+ "idx": 1,
10
+ "name": "1",
11
+ "path": "1_Pooling",
12
+ "type": "sentence_transformers.models.Pooling"
13
+ }
14
+ ]
checkpoint-12/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:240cfed1b06cf1384ac1cd0909772f908dd3680319ee129881c4b5b80ef584fb
3
+ size 1130520122
checkpoint-12/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6778088791fae1e7292de4d6f29fca6861fbb98ab65d211b5e737e21e431c189
3
+ size 565251810
checkpoint-12/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a5a74b0d974998758fffd2a36972af2b011f363b677acea5988bd5bf61032a2e
3
+ size 14244
checkpoint-12/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7969a6d98ccf86631987f7ac53e84c055aba94996ce4fb9a2e5399b2ecf1dbb2
3
+ size 1064
checkpoint-12/sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 512,
3
+ "do_lower_case": false
4
+ }
checkpoint-12/special_tokens_map.json ADDED
@@ -0,0 +1,15 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": "[CLS]",
3
+ "cls_token": "[CLS]",
4
+ "eos_token": "[SEP]",
5
+ "mask_token": "[MASK]",
6
+ "pad_token": "[PAD]",
7
+ "sep_token": "[SEP]",
8
+ "unk_token": {
9
+ "content": "[UNK]",
10
+ "lstrip": false,
11
+ "normalized": true,
12
+ "rstrip": false,
13
+ "single_word": false
14
+ }
15
+ }
checkpoint-12/spm.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c679fbf93643d19aab7ee10c0b99e460bdbc02fedf34b92b05af343b4af586fd
3
+ size 2464616
checkpoint-12/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-12/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
+ "1": {
12
+ "content": "[CLS]",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "[SEP]",
21
+ "lstrip": false,
22
+ "normalized": false,
23
+ "rstrip": false,
24
+ "single_word": false,
25
+ "special": true
26
+ },
27
+ "3": {
28
+ "content": "[UNK]",
29
+ "lstrip": false,
30
+ "normalized": true,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "128000": {
36
+ "content": "[MASK]",
37
+ "lstrip": false,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "[CLS]",
45
+ "clean_up_tokenization_spaces": true,
46
+ "cls_token": "[CLS]",
47
+ "do_lower_case": false,
48
+ "eos_token": "[SEP]",
49
+ "mask_token": "[MASK]",
50
+ "model_max_length": 1000000000000000019884624838656,
51
+ "pad_token": "[PAD]",
52
+ "sep_token": "[SEP]",
53
+ "sp_model_kwargs": {},
54
+ "split_by_punct": false,
55
+ "tokenizer_class": "DebertaV2Tokenizer",
56
+ "unk_token": "[UNK]",
57
+ "vocab_type": "spm"
58
+ }
checkpoint-12/trainer_state.json ADDED
@@ -0,0 +1,999 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_metric": null,
3
+ "best_model_checkpoint": null,
4
+ "epoch": 0.3488372093023256,
5
+ "eval_steps": 2,
6
+ "global_step": 12,
7
+ "is_hyper_param_search": false,
8
+ "is_local_process_zero": true,
9
+ "is_world_process_zero": true,
10
+ "log_history": [
11
+ {
12
+ "epoch": 0.029069767441860465,
13
+ "grad_norm": 13.73514461517334,
14
+ "learning_rate": 3.3333333333333333e-06,
15
+ "loss": 6.7536,
16
+ "step": 1
17
+ },
18
+ {
19
+ "epoch": 0.05813953488372093,
20
+ "grad_norm": 16.892961502075195,
21
+ "learning_rate": 6.666666666666667e-06,
22
+ "loss": 6.6203,
23
+ "step": 2
24
+ },
25
+ {
26
+ "epoch": 0.05813953488372093,
27
+ "eval_NLI-v2_cosine_accuracy": 1.0,
28
+ "eval_NLI-v2_dot_accuracy": 0.109375,
29
+ "eval_NLI-v2_euclidean_accuracy": 1.0,
30
+ "eval_NLI-v2_manhattan_accuracy": 1.0,
31
+ "eval_NLI-v2_max_accuracy": 1.0,
32
+ "eval_VitaminC_cosine_accuracy": 0.55078125,
33
+ "eval_VitaminC_cosine_accuracy_threshold": 0.9469717741012573,
34
+ "eval_VitaminC_cosine_ap": 0.5165514227338435,
35
+ "eval_VitaminC_cosine_f1": 0.6525198938992042,
36
+ "eval_VitaminC_cosine_f1_threshold": 0.4987494945526123,
37
+ "eval_VitaminC_cosine_precision": 0.484251968503937,
38
+ "eval_VitaminC_cosine_recall": 1.0,
39
+ "eval_VitaminC_dot_accuracy": 0.55078125,
40
+ "eval_VitaminC_dot_accuracy_threshold": 417.45166015625,
41
+ "eval_VitaminC_dot_ap": 0.5125282416460941,
42
+ "eval_VitaminC_dot_f1": 0.6525198938992042,
43
+ "eval_VitaminC_dot_f1_threshold": 200.84530639648438,
44
+ "eval_VitaminC_dot_precision": 0.484251968503937,
45
+ "eval_VitaminC_dot_recall": 1.0,
46
+ "eval_VitaminC_euclidean_accuracy": 0.546875,
47
+ "eval_VitaminC_euclidean_accuracy_threshold": 6.841136932373047,
48
+ "eval_VitaminC_euclidean_ap": 0.512994913068431,
49
+ "eval_VitaminC_euclidean_f1": 0.6525198938992042,
50
+ "eval_VitaminC_euclidean_f1_threshold": 20.22179412841797,
51
+ "eval_VitaminC_euclidean_precision": 0.484251968503937,
52
+ "eval_VitaminC_euclidean_recall": 1.0,
53
+ "eval_VitaminC_manhattan_accuracy": 0.546875,
54
+ "eval_VitaminC_manhattan_accuracy_threshold": 117.1143569946289,
55
+ "eval_VitaminC_manhattan_ap": 0.5177961641566705,
56
+ "eval_VitaminC_manhattan_f1": 0.6542553191489362,
57
+ "eval_VitaminC_manhattan_f1_threshold": 292.3347473144531,
58
+ "eval_VitaminC_manhattan_precision": 0.48616600790513836,
59
+ "eval_VitaminC_manhattan_recall": 1.0,
60
+ "eval_VitaminC_max_accuracy": 0.55078125,
61
+ "eval_VitaminC_max_accuracy_threshold": 417.45166015625,
62
+ "eval_VitaminC_max_ap": 0.5177961641566705,
63
+ "eval_VitaminC_max_f1": 0.6542553191489362,
64
+ "eval_VitaminC_max_f1_threshold": 292.3347473144531,
65
+ "eval_VitaminC_max_precision": 0.48616600790513836,
66
+ "eval_VitaminC_max_recall": 1.0,
67
+ "eval_sequential_score": 0.5177961641566705,
68
+ "eval_sts-test_pearson_cosine": 0.017178505918243114,
69
+ "eval_sts-test_pearson_dot": 0.16228524441844774,
70
+ "eval_sts-test_pearson_euclidean": 0.024391561236282323,
71
+ "eval_sts-test_pearson_manhattan": 0.059272519732015624,
72
+ "eval_sts-test_pearson_max": 0.16228524441844774,
73
+ "eval_sts-test_spearman_cosine": 0.07115563415775981,
74
+ "eval_sts-test_spearman_dot": 0.1967074668301984,
75
+ "eval_sts-test_spearman_euclidean": 0.05117306486959643,
76
+ "eval_sts-test_spearman_manhattan": 0.07641354546391785,
77
+ "eval_sts-test_spearman_max": 0.1967074668301984,
78
+ "eval_vitaminc-pairs_loss": 2.72658371925354,
79
+ "eval_vitaminc-pairs_runtime": 1.4719,
80
+ "eval_vitaminc-pairs_samples_per_second": 73.373,
81
+ "eval_vitaminc-pairs_steps_per_second": 1.359,
82
+ "step": 2
83
+ },
84
+ {
85
+ "epoch": 0.05813953488372093,
86
+ "eval_negation-triplets_loss": 5.094177722930908,
87
+ "eval_negation-triplets_runtime": 0.3027,
88
+ "eval_negation-triplets_samples_per_second": 211.46,
89
+ "eval_negation-triplets_steps_per_second": 3.304,
90
+ "step": 2
91
+ },
92
+ {
93
+ "epoch": 0.05813953488372093,
94
+ "eval_scitail-pairs-pos_loss": 1.9183871746063232,
95
+ "eval_scitail-pairs-pos_runtime": 0.3785,
96
+ "eval_scitail-pairs-pos_samples_per_second": 142.654,
97
+ "eval_scitail-pairs-pos_steps_per_second": 2.642,
98
+ "step": 2
99
+ },
100
+ {
101
+ "epoch": 0.05813953488372093,
102
+ "eval_xsum-pairs_loss": 6.3277668952941895,
103
+ "eval_xsum-pairs_runtime": 3.4205,
104
+ "eval_xsum-pairs_samples_per_second": 37.422,
105
+ "eval_xsum-pairs_steps_per_second": 0.585,
106
+ "step": 2
107
+ },
108
+ {
109
+ "epoch": 0.05813953488372093,
110
+ "eval_sciq_pairs_loss": 0.36539140343666077,
111
+ "eval_sciq_pairs_runtime": 3.4462,
112
+ "eval_sciq_pairs_samples_per_second": 37.143,
113
+ "eval_sciq_pairs_steps_per_second": 0.58,
114
+ "step": 2
115
+ },
116
+ {
117
+ "epoch": 0.05813953488372093,
118
+ "eval_qasc_pairs_loss": 3.9689104557037354,
119
+ "eval_qasc_pairs_runtime": 0.6477,
120
+ "eval_qasc_pairs_samples_per_second": 197.636,
121
+ "eval_qasc_pairs_steps_per_second": 3.088,
122
+ "step": 2
123
+ },
124
+ {
125
+ "epoch": 0.05813953488372093,
126
+ "eval_openbookqa_pairs_loss": 4.7439374923706055,
127
+ "eval_openbookqa_pairs_runtime": 0.5759,
128
+ "eval_openbookqa_pairs_samples_per_second": 222.248,
129
+ "eval_openbookqa_pairs_steps_per_second": 3.473,
130
+ "step": 2
131
+ },
132
+ {
133
+ "epoch": 0.05813953488372093,
134
+ "eval_msmarco_pairs_loss": 10.51363754272461,
135
+ "eval_msmarco_pairs_runtime": 1.3036,
136
+ "eval_msmarco_pairs_samples_per_second": 98.189,
137
+ "eval_msmarco_pairs_steps_per_second": 1.534,
138
+ "step": 2
139
+ },
140
+ {
141
+ "epoch": 0.05813953488372093,
142
+ "eval_nq_pairs_loss": 4.969011306762695,
143
+ "eval_nq_pairs_runtime": 2.5922,
144
+ "eval_nq_pairs_samples_per_second": 49.378,
145
+ "eval_nq_pairs_steps_per_second": 0.772,
146
+ "step": 2
147
+ },
148
+ {
149
+ "epoch": 0.05813953488372093,
150
+ "eval_trivia_pairs_loss": 3.8609507083892822,
151
+ "eval_trivia_pairs_runtime": 4.3666,
152
+ "eval_trivia_pairs_samples_per_second": 29.314,
153
+ "eval_trivia_pairs_steps_per_second": 0.458,
154
+ "step": 2
155
+ },
156
+ {
157
+ "epoch": 0.05813953488372093,
158
+ "eval_gooaq_pairs_loss": 8.041110038757324,
159
+ "eval_gooaq_pairs_runtime": 0.9119,
160
+ "eval_gooaq_pairs_samples_per_second": 140.361,
161
+ "eval_gooaq_pairs_steps_per_second": 2.193,
162
+ "step": 2
163
+ },
164
+ {
165
+ "epoch": 0.05813953488372093,
166
+ "eval_paws-pos_loss": 2.21897029876709,
167
+ "eval_paws-pos_runtime": 0.6905,
168
+ "eval_paws-pos_samples_per_second": 185.364,
169
+ "eval_paws-pos_steps_per_second": 2.896,
170
+ "step": 2
171
+ },
172
+ {
173
+ "epoch": 0.0872093023255814,
174
+ "grad_norm": 15.801941871643066,
175
+ "learning_rate": 1e-05,
176
+ "loss": 6.7963,
177
+ "step": 3
178
+ },
179
+ {
180
+ "epoch": 0.11627906976744186,
181
+ "grad_norm": 11.94517993927002,
182
+ "learning_rate": 1.3333333333333333e-05,
183
+ "loss": 6.4488,
184
+ "step": 4
185
+ },
186
+ {
187
+ "epoch": 0.11627906976744186,
188
+ "eval_NLI-v2_cosine_accuracy": 1.0,
189
+ "eval_NLI-v2_dot_accuracy": 0.109375,
190
+ "eval_NLI-v2_euclidean_accuracy": 1.0,
191
+ "eval_NLI-v2_manhattan_accuracy": 1.0,
192
+ "eval_NLI-v2_max_accuracy": 1.0,
193
+ "eval_VitaminC_cosine_accuracy": 0.55078125,
194
+ "eval_VitaminC_cosine_accuracy_threshold": 0.9437637329101562,
195
+ "eval_VitaminC_cosine_ap": 0.5161989667198088,
196
+ "eval_VitaminC_cosine_f1": 0.6525198938992042,
197
+ "eval_VitaminC_cosine_f1_threshold": 0.56722092628479,
198
+ "eval_VitaminC_cosine_precision": 0.484251968503937,
199
+ "eval_VitaminC_cosine_recall": 1.0,
200
+ "eval_VitaminC_dot_accuracy": 0.546875,
201
+ "eval_VitaminC_dot_accuracy_threshold": 416.9976806640625,
202
+ "eval_VitaminC_dot_ap": 0.515337582265173,
203
+ "eval_VitaminC_dot_f1": 0.6507936507936508,
204
+ "eval_VitaminC_dot_f1_threshold": 223.1047821044922,
205
+ "eval_VitaminC_dot_precision": 0.4823529411764706,
206
+ "eval_VitaminC_dot_recall": 1.0,
207
+ "eval_VitaminC_euclidean_accuracy": 0.55078125,
208
+ "eval_VitaminC_euclidean_accuracy_threshold": 6.841610908508301,
209
+ "eval_VitaminC_euclidean_ap": 0.5167890569390883,
210
+ "eval_VitaminC_euclidean_f1": 0.6542553191489362,
211
+ "eval_VitaminC_euclidean_f1_threshold": 18.548978805541992,
212
+ "eval_VitaminC_euclidean_precision": 0.48616600790513836,
213
+ "eval_VitaminC_euclidean_recall": 1.0,
214
+ "eval_VitaminC_manhattan_accuracy": 0.54296875,
215
+ "eval_VitaminC_manhattan_accuracy_threshold": 110.40855407714844,
216
+ "eval_VitaminC_manhattan_ap": 0.5219582103185639,
217
+ "eval_VitaminC_manhattan_f1": 0.6542553191489362,
218
+ "eval_VitaminC_manhattan_f1_threshold": 282.48602294921875,
219
+ "eval_VitaminC_manhattan_precision": 0.48616600790513836,
220
+ "eval_VitaminC_manhattan_recall": 1.0,
221
+ "eval_VitaminC_max_accuracy": 0.55078125,
222
+ "eval_VitaminC_max_accuracy_threshold": 416.9976806640625,
223
+ "eval_VitaminC_max_ap": 0.5219582103185639,
224
+ "eval_VitaminC_max_f1": 0.6542553191489362,
225
+ "eval_VitaminC_max_f1_threshold": 282.48602294921875,
226
+ "eval_VitaminC_max_precision": 0.48616600790513836,
227
+ "eval_VitaminC_max_recall": 1.0,
228
+ "eval_sequential_score": 0.5219582103185639,
229
+ "eval_sts-test_pearson_cosine": 0.025276809579094173,
230
+ "eval_sts-test_pearson_dot": 0.21470147681711343,
231
+ "eval_sts-test_pearson_euclidean": 0.027783806662194415,
232
+ "eval_sts-test_pearson_manhattan": 0.059736177006118926,
233
+ "eval_sts-test_pearson_max": 0.21470147681711343,
234
+ "eval_sts-test_spearman_cosine": 0.0776783582433888,
235
+ "eval_sts-test_spearman_dot": 0.2491347682204262,
236
+ "eval_sts-test_spearman_euclidean": 0.054640386155863654,
237
+ "eval_sts-test_spearman_manhattan": 0.07764379287610436,
238
+ "eval_sts-test_spearman_max": 0.2491347682204262,
239
+ "eval_vitaminc-pairs_loss": 2.7045021057128906,
240
+ "eval_vitaminc-pairs_runtime": 1.4394,
241
+ "eval_vitaminc-pairs_samples_per_second": 75.033,
242
+ "eval_vitaminc-pairs_steps_per_second": 1.389,
243
+ "step": 4
244
+ },
245
+ {
246
+ "epoch": 0.11627906976744186,
247
+ "eval_negation-triplets_loss": 5.081584453582764,
248
+ "eval_negation-triplets_runtime": 0.2952,
249
+ "eval_negation-triplets_samples_per_second": 216.79,
250
+ "eval_negation-triplets_steps_per_second": 3.387,
251
+ "step": 4
252
+ },
253
+ {
254
+ "epoch": 0.11627906976744186,
255
+ "eval_scitail-pairs-pos_loss": 1.9186776876449585,
256
+ "eval_scitail-pairs-pos_runtime": 0.3702,
257
+ "eval_scitail-pairs-pos_samples_per_second": 145.879,
258
+ "eval_scitail-pairs-pos_steps_per_second": 2.701,
259
+ "step": 4
260
+ },
261
+ {
262
+ "epoch": 0.11627906976744186,
263
+ "eval_xsum-pairs_loss": 6.199001312255859,
264
+ "eval_xsum-pairs_runtime": 3.4227,
265
+ "eval_xsum-pairs_samples_per_second": 37.398,
266
+ "eval_xsum-pairs_steps_per_second": 0.584,
267
+ "step": 4
268
+ },
269
+ {
270
+ "epoch": 0.11627906976744186,
271
+ "eval_sciq_pairs_loss": 0.3414052426815033,
272
+ "eval_sciq_pairs_runtime": 3.4403,
273
+ "eval_sciq_pairs_samples_per_second": 37.206,
274
+ "eval_sciq_pairs_steps_per_second": 0.581,
275
+ "step": 4
276
+ },
277
+ {
278
+ "epoch": 0.11627906976744186,
279
+ "eval_qasc_pairs_loss": 3.662200927734375,
280
+ "eval_qasc_pairs_runtime": 0.6486,
281
+ "eval_qasc_pairs_samples_per_second": 197.349,
282
+ "eval_qasc_pairs_steps_per_second": 3.084,
283
+ "step": 4
284
+ },
285
+ {
286
+ "epoch": 0.11627906976744186,
287
+ "eval_openbookqa_pairs_loss": 4.650759220123291,
288
+ "eval_openbookqa_pairs_runtime": 0.5722,
289
+ "eval_openbookqa_pairs_samples_per_second": 223.702,
290
+ "eval_openbookqa_pairs_steps_per_second": 3.495,
291
+ "step": 4
292
+ },
293
+ {
294
+ "epoch": 0.11627906976744186,
295
+ "eval_msmarco_pairs_loss": 9.487885475158691,
296
+ "eval_msmarco_pairs_runtime": 1.2978,
297
+ "eval_msmarco_pairs_samples_per_second": 98.63,
298
+ "eval_msmarco_pairs_steps_per_second": 1.541,
299
+ "step": 4
300
+ },
301
+ {
302
+ "epoch": 0.11627906976744186,
303
+ "eval_nq_pairs_loss": 4.4714131355285645,
304
+ "eval_nq_pairs_runtime": 2.569,
305
+ "eval_nq_pairs_samples_per_second": 49.824,
306
+ "eval_nq_pairs_steps_per_second": 0.779,
307
+ "step": 4
308
+ },
309
+ {
310
+ "epoch": 0.11627906976744186,
311
+ "eval_trivia_pairs_loss": 3.5246145725250244,
312
+ "eval_trivia_pairs_runtime": 4.3557,
313
+ "eval_trivia_pairs_samples_per_second": 29.387,
314
+ "eval_trivia_pairs_steps_per_second": 0.459,
315
+ "step": 4
316
+ },
317
+ {
318
+ "epoch": 0.11627906976744186,
319
+ "eval_gooaq_pairs_loss": 7.395092010498047,
320
+ "eval_gooaq_pairs_runtime": 0.9109,
321
+ "eval_gooaq_pairs_samples_per_second": 140.521,
322
+ "eval_gooaq_pairs_steps_per_second": 2.196,
323
+ "step": 4
324
+ },
325
+ {
326
+ "epoch": 0.11627906976744186,
327
+ "eval_paws-pos_loss": 2.233201742172241,
328
+ "eval_paws-pos_runtime": 0.6853,
329
+ "eval_paws-pos_samples_per_second": 186.773,
330
+ "eval_paws-pos_steps_per_second": 2.918,
331
+ "step": 4
332
+ },
333
+ {
334
+ "epoch": 0.14534883720930233,
335
+ "grad_norm": 6.336462497711182,
336
+ "learning_rate": 1.6666666666666667e-05,
337
+ "loss": 6.5567,
338
+ "step": 5
339
+ },
340
+ {
341
+ "epoch": 0.1744186046511628,
342
+ "grad_norm": 30.24117660522461,
343
+ "learning_rate": 2e-05,
344
+ "loss": 7.994,
345
+ "step": 6
346
+ },
347
+ {
348
+ "epoch": 0.1744186046511628,
349
+ "eval_NLI-v2_cosine_accuracy": 1.0,
350
+ "eval_NLI-v2_dot_accuracy": 0.125,
351
+ "eval_NLI-v2_euclidean_accuracy": 1.0,
352
+ "eval_NLI-v2_manhattan_accuracy": 1.0,
353
+ "eval_NLI-v2_max_accuracy": 1.0,
354
+ "eval_VitaminC_cosine_accuracy": 0.546875,
355
+ "eval_VitaminC_cosine_accuracy_threshold": 0.9505010843276978,
356
+ "eval_VitaminC_cosine_ap": 0.5220332354916479,
357
+ "eval_VitaminC_cosine_f1": 0.6542553191489362,
358
+ "eval_VitaminC_cosine_f1_threshold": 0.6742120385169983,
359
+ "eval_VitaminC_cosine_precision": 0.48616600790513836,
360
+ "eval_VitaminC_cosine_recall": 1.0,
361
+ "eval_VitaminC_dot_accuracy": 0.55859375,
362
+ "eval_VitaminC_dot_accuracy_threshold": 410.96356201171875,
363
+ "eval_VitaminC_dot_ap": 0.5098545013010067,
364
+ "eval_VitaminC_dot_f1": 0.6507936507936508,
365
+ "eval_VitaminC_dot_f1_threshold": 259.4541015625,
366
+ "eval_VitaminC_dot_precision": 0.4823529411764706,
367
+ "eval_VitaminC_dot_recall": 1.0,
368
+ "eval_VitaminC_euclidean_accuracy": 0.55078125,
369
+ "eval_VitaminC_euclidean_accuracy_threshold": 6.85032844543457,
370
+ "eval_VitaminC_euclidean_ap": 0.5187640034678729,
371
+ "eval_VitaminC_euclidean_f1": 0.6522911051212938,
372
+ "eval_VitaminC_euclidean_f1_threshold": 15.907812118530273,
373
+ "eval_VitaminC_euclidean_precision": 0.4879032258064516,
374
+ "eval_VitaminC_euclidean_recall": 0.983739837398374,
375
+ "eval_VitaminC_manhattan_accuracy": 0.54296875,
376
+ "eval_VitaminC_manhattan_accuracy_threshold": 128.04562377929688,
377
+ "eval_VitaminC_manhattan_ap": 0.5216728184773658,
378
+ "eval_VitaminC_manhattan_f1": 0.6542553191489362,
379
+ "eval_VitaminC_manhattan_f1_threshold": 266.50201416015625,
380
+ "eval_VitaminC_manhattan_precision": 0.48616600790513836,
381
+ "eval_VitaminC_manhattan_recall": 1.0,
382
+ "eval_VitaminC_max_accuracy": 0.55859375,
383
+ "eval_VitaminC_max_accuracy_threshold": 410.96356201171875,
384
+ "eval_VitaminC_max_ap": 0.5220332354916479,
385
+ "eval_VitaminC_max_f1": 0.6542553191489362,
386
+ "eval_VitaminC_max_f1_threshold": 266.50201416015625,
387
+ "eval_VitaminC_max_precision": 0.4879032258064516,
388
+ "eval_VitaminC_max_recall": 1.0,
389
+ "eval_sequential_score": 0.5220332354916479,
390
+ "eval_sts-test_pearson_cosine": 0.03816033089318139,
391
+ "eval_sts-test_pearson_dot": 0.2714501319216333,
392
+ "eval_sts-test_pearson_euclidean": 0.03534045207172642,
393
+ "eval_sts-test_pearson_manhattan": 0.061786120590017285,
394
+ "eval_sts-test_pearson_max": 0.2714501319216333,
395
+ "eval_sts-test_spearman_cosine": 0.09093735567962451,
396
+ "eval_sts-test_spearman_dot": 0.29608861437488704,
397
+ "eval_sts-test_spearman_euclidean": 0.06076512778358849,
398
+ "eval_sts-test_spearman_manhattan": 0.08153826351124775,
399
+ "eval_sts-test_spearman_max": 0.29608861437488704,
400
+ "eval_vitaminc-pairs_loss": 2.6921818256378174,
401
+ "eval_vitaminc-pairs_runtime": 1.4315,
402
+ "eval_vitaminc-pairs_samples_per_second": 75.443,
403
+ "eval_vitaminc-pairs_steps_per_second": 1.397,
404
+ "step": 6
405
+ },
406
+ {
407
+ "epoch": 0.1744186046511628,
408
+ "eval_negation-triplets_loss": 5.068139553070068,
409
+ "eval_negation-triplets_runtime": 0.2981,
410
+ "eval_negation-triplets_samples_per_second": 214.716,
411
+ "eval_negation-triplets_steps_per_second": 3.355,
412
+ "step": 6
413
+ },
414
+ {
415
+ "epoch": 0.1744186046511628,
416
+ "eval_scitail-pairs-pos_loss": 1.9319576025009155,
417
+ "eval_scitail-pairs-pos_runtime": 0.3653,
418
+ "eval_scitail-pairs-pos_samples_per_second": 147.818,
419
+ "eval_scitail-pairs-pos_steps_per_second": 2.737,
420
+ "step": 6
421
+ },
422
+ {
423
+ "epoch": 0.1744186046511628,
424
+ "eval_xsum-pairs_loss": 6.079549789428711,
425
+ "eval_xsum-pairs_runtime": 3.4013,
426
+ "eval_xsum-pairs_samples_per_second": 37.633,
427
+ "eval_xsum-pairs_steps_per_second": 0.588,
428
+ "step": 6
429
+ },
430
+ {
431
+ "epoch": 0.1744186046511628,
432
+ "eval_sciq_pairs_loss": 0.32075461745262146,
433
+ "eval_sciq_pairs_runtime": 3.442,
434
+ "eval_sciq_pairs_samples_per_second": 37.188,
435
+ "eval_sciq_pairs_steps_per_second": 0.581,
436
+ "step": 6
437
+ },
438
+ {
439
+ "epoch": 0.1744186046511628,
440
+ "eval_qasc_pairs_loss": 3.36326265335083,
441
+ "eval_qasc_pairs_runtime": 0.6495,
442
+ "eval_qasc_pairs_samples_per_second": 197.087,
443
+ "eval_qasc_pairs_steps_per_second": 3.079,
444
+ "step": 6
445
+ },
446
+ {
447
+ "epoch": 0.1744186046511628,
448
+ "eval_openbookqa_pairs_loss": 4.481137275695801,
449
+ "eval_openbookqa_pairs_runtime": 0.5729,
450
+ "eval_openbookqa_pairs_samples_per_second": 223.432,
451
+ "eval_openbookqa_pairs_steps_per_second": 3.491,
452
+ "step": 6
453
+ },
454
+ {
455
+ "epoch": 0.1744186046511628,
456
+ "eval_msmarco_pairs_loss": 8.048794746398926,
457
+ "eval_msmarco_pairs_runtime": 1.2936,
458
+ "eval_msmarco_pairs_samples_per_second": 98.947,
459
+ "eval_msmarco_pairs_steps_per_second": 1.546,
460
+ "step": 6
461
+ },
462
+ {
463
+ "epoch": 0.1744186046511628,
464
+ "eval_nq_pairs_loss": 3.7927472591400146,
465
+ "eval_nq_pairs_runtime": 2.5733,
466
+ "eval_nq_pairs_samples_per_second": 49.741,
467
+ "eval_nq_pairs_steps_per_second": 0.777,
468
+ "step": 6
469
+ },
470
+ {
471
+ "epoch": 0.1744186046511628,
472
+ "eval_trivia_pairs_loss": 3.284496545791626,
473
+ "eval_trivia_pairs_runtime": 4.355,
474
+ "eval_trivia_pairs_samples_per_second": 29.391,
475
+ "eval_trivia_pairs_steps_per_second": 0.459,
476
+ "step": 6
477
+ },
478
+ {
479
+ "epoch": 0.1744186046511628,
480
+ "eval_gooaq_pairs_loss": 6.677769184112549,
481
+ "eval_gooaq_pairs_runtime": 0.9178,
482
+ "eval_gooaq_pairs_samples_per_second": 139.457,
483
+ "eval_gooaq_pairs_steps_per_second": 2.179,
484
+ "step": 6
485
+ },
486
+ {
487
+ "epoch": 0.1744186046511628,
488
+ "eval_paws-pos_loss": 2.2626476287841797,
489
+ "eval_paws-pos_runtime": 0.6901,
490
+ "eval_paws-pos_samples_per_second": 185.475,
491
+ "eval_paws-pos_steps_per_second": 2.898,
492
+ "step": 6
493
+ },
494
+ {
495
+ "epoch": 0.20348837209302326,
496
+ "grad_norm": 11.503045082092285,
497
+ "learning_rate": 2.3333333333333336e-05,
498
+ "loss": 7.1037,
499
+ "step": 7
500
+ },
501
+ {
502
+ "epoch": 0.23255813953488372,
503
+ "grad_norm": 6.524661540985107,
504
+ "learning_rate": 2.6666666666666667e-05,
505
+ "loss": 6.6239,
506
+ "step": 8
507
+ },
508
+ {
509
+ "epoch": 0.23255813953488372,
510
+ "eval_NLI-v2_cosine_accuracy": 1.0,
511
+ "eval_NLI-v2_dot_accuracy": 0.140625,
512
+ "eval_NLI-v2_euclidean_accuracy": 1.0,
513
+ "eval_NLI-v2_manhattan_accuracy": 1.0,
514
+ "eval_NLI-v2_max_accuracy": 1.0,
515
+ "eval_VitaminC_cosine_accuracy": 0.55078125,
516
+ "eval_VitaminC_cosine_accuracy_threshold": 0.9556466341018677,
517
+ "eval_VitaminC_cosine_ap": 0.5262234678756146,
518
+ "eval_VitaminC_cosine_f1": 0.6542553191489362,
519
+ "eval_VitaminC_cosine_f1_threshold": 0.7482036352157593,
520
+ "eval_VitaminC_cosine_precision": 0.48616600790513836,
521
+ "eval_VitaminC_cosine_recall": 1.0,
522
+ "eval_VitaminC_dot_accuracy": 0.54296875,
523
+ "eval_VitaminC_dot_accuracy_threshold": 429.2105712890625,
524
+ "eval_VitaminC_dot_ap": 0.5105928204741528,
525
+ "eval_VitaminC_dot_f1": 0.6525198938992042,
526
+ "eval_VitaminC_dot_f1_threshold": 304.46807861328125,
527
+ "eval_VitaminC_dot_precision": 0.484251968503937,
528
+ "eval_VitaminC_dot_recall": 1.0,
529
+ "eval_VitaminC_euclidean_accuracy": 0.55078125,
530
+ "eval_VitaminC_euclidean_accuracy_threshold": 6.457396030426025,
531
+ "eval_VitaminC_euclidean_ap": 0.5262113804807371,
532
+ "eval_VitaminC_euclidean_f1": 0.6505376344086021,
533
+ "eval_VitaminC_euclidean_f1_threshold": 14.363262176513672,
534
+ "eval_VitaminC_euclidean_precision": 0.4859437751004016,
535
+ "eval_VitaminC_euclidean_recall": 0.983739837398374,
536
+ "eval_VitaminC_manhattan_accuracy": 0.546875,
537
+ "eval_VitaminC_manhattan_accuracy_threshold": 144.2735137939453,
538
+ "eval_VitaminC_manhattan_ap": 0.5216238817948092,
539
+ "eval_VitaminC_manhattan_f1": 0.6542553191489362,
540
+ "eval_VitaminC_manhattan_f1_threshold": 248.78077697753906,
541
+ "eval_VitaminC_manhattan_precision": 0.48616600790513836,
542
+ "eval_VitaminC_manhattan_recall": 1.0,
543
+ "eval_VitaminC_max_accuracy": 0.55078125,
544
+ "eval_VitaminC_max_accuracy_threshold": 429.2105712890625,
545
+ "eval_VitaminC_max_ap": 0.5262234678756146,
546
+ "eval_VitaminC_max_f1": 0.6542553191489362,
547
+ "eval_VitaminC_max_f1_threshold": 304.46807861328125,
548
+ "eval_VitaminC_max_precision": 0.48616600790513836,
549
+ "eval_VitaminC_max_recall": 1.0,
550
+ "eval_sequential_score": 0.5262234678756146,
551
+ "eval_sts-test_pearson_cosine": 0.04822016371804072,
552
+ "eval_sts-test_pearson_dot": 0.2933829804301914,
553
+ "eval_sts-test_pearson_euclidean": 0.0440427995340741,
554
+ "eval_sts-test_pearson_manhattan": 0.06814345626491929,
555
+ "eval_sts-test_pearson_max": 0.2933829804301914,
556
+ "eval_sts-test_spearman_cosine": 0.10311449730221736,
557
+ "eval_sts-test_spearman_dot": 0.30840433621335606,
558
+ "eval_sts-test_spearman_euclidean": 0.06897119310369988,
559
+ "eval_sts-test_spearman_manhattan": 0.08808657402390177,
560
+ "eval_sts-test_spearman_max": 0.30840433621335606,
561
+ "eval_vitaminc-pairs_loss": 2.7006633281707764,
562
+ "eval_vitaminc-pairs_runtime": 1.435,
563
+ "eval_vitaminc-pairs_samples_per_second": 75.262,
564
+ "eval_vitaminc-pairs_steps_per_second": 1.394,
565
+ "step": 8
566
+ },
567
+ {
568
+ "epoch": 0.23255813953488372,
569
+ "eval_negation-triplets_loss": 5.0855937004089355,
570
+ "eval_negation-triplets_runtime": 0.2946,
571
+ "eval_negation-triplets_samples_per_second": 217.208,
572
+ "eval_negation-triplets_steps_per_second": 3.394,
573
+ "step": 8
574
+ },
575
+ {
576
+ "epoch": 0.23255813953488372,
577
+ "eval_scitail-pairs-pos_loss": 1.9701284170150757,
578
+ "eval_scitail-pairs-pos_runtime": 0.3657,
579
+ "eval_scitail-pairs-pos_samples_per_second": 147.656,
580
+ "eval_scitail-pairs-pos_steps_per_second": 2.734,
581
+ "step": 8
582
+ },
583
+ {
584
+ "epoch": 0.23255813953488372,
585
+ "eval_xsum-pairs_loss": 6.050904273986816,
586
+ "eval_xsum-pairs_runtime": 3.4161,
587
+ "eval_xsum-pairs_samples_per_second": 37.47,
588
+ "eval_xsum-pairs_steps_per_second": 0.585,
589
+ "step": 8
590
+ },
591
+ {
592
+ "epoch": 0.23255813953488372,
593
+ "eval_sciq_pairs_loss": 0.31550008058547974,
594
+ "eval_sciq_pairs_runtime": 3.4195,
595
+ "eval_sciq_pairs_samples_per_second": 37.432,
596
+ "eval_sciq_pairs_steps_per_second": 0.585,
597
+ "step": 8
598
+ },
599
+ {
600
+ "epoch": 0.23255813953488372,
601
+ "eval_qasc_pairs_loss": 3.174623727798462,
602
+ "eval_qasc_pairs_runtime": 0.6393,
603
+ "eval_qasc_pairs_samples_per_second": 200.217,
604
+ "eval_qasc_pairs_steps_per_second": 3.128,
605
+ "step": 8
606
+ },
607
+ {
608
+ "epoch": 0.23255813953488372,
609
+ "eval_openbookqa_pairs_loss": 4.325982570648193,
610
+ "eval_openbookqa_pairs_runtime": 0.5723,
611
+ "eval_openbookqa_pairs_samples_per_second": 223.64,
612
+ "eval_openbookqa_pairs_steps_per_second": 3.494,
613
+ "step": 8
614
+ },
615
+ {
616
+ "epoch": 0.23255813953488372,
617
+ "eval_msmarco_pairs_loss": 6.989845275878906,
618
+ "eval_msmarco_pairs_runtime": 1.2986,
619
+ "eval_msmarco_pairs_samples_per_second": 98.569,
620
+ "eval_msmarco_pairs_steps_per_second": 1.54,
621
+ "step": 8
622
+ },
623
+ {
624
+ "epoch": 0.23255813953488372,
625
+ "eval_nq_pairs_loss": 3.3526666164398193,
626
+ "eval_nq_pairs_runtime": 2.5766,
627
+ "eval_nq_pairs_samples_per_second": 49.677,
628
+ "eval_nq_pairs_steps_per_second": 0.776,
629
+ "step": 8
630
+ },
631
+ {
632
+ "epoch": 0.23255813953488372,
633
+ "eval_trivia_pairs_loss": 3.2416629791259766,
634
+ "eval_trivia_pairs_runtime": 4.3655,
635
+ "eval_trivia_pairs_samples_per_second": 29.321,
636
+ "eval_trivia_pairs_steps_per_second": 0.458,
637
+ "step": 8
638
+ },
639
+ {
640
+ "epoch": 0.23255813953488372,
641
+ "eval_gooaq_pairs_loss": 6.288425922393799,
642
+ "eval_gooaq_pairs_runtime": 0.9111,
643
+ "eval_gooaq_pairs_samples_per_second": 140.492,
644
+ "eval_gooaq_pairs_steps_per_second": 2.195,
645
+ "step": 8
646
+ },
647
+ {
648
+ "epoch": 0.23255813953488372,
649
+ "eval_paws-pos_loss": 2.351145029067993,
650
+ "eval_paws-pos_runtime": 0.6907,
651
+ "eval_paws-pos_samples_per_second": 185.325,
652
+ "eval_paws-pos_steps_per_second": 2.896,
653
+ "step": 8
654
+ },
655
+ {
656
+ "epoch": 0.2616279069767442,
657
+ "grad_norm": 15.498976707458496,
658
+ "learning_rate": 3.0000000000000004e-05,
659
+ "loss": 6.7359,
660
+ "step": 9
661
+ },
662
+ {
663
+ "epoch": 0.29069767441860467,
664
+ "grad_norm": 17.9885311126709,
665
+ "learning_rate": 3.3333333333333335e-05,
666
+ "loss": 7.0187,
667
+ "step": 10
668
+ },
669
+ {
670
+ "epoch": 0.29069767441860467,
671
+ "eval_NLI-v2_cosine_accuracy": 1.0,
672
+ "eval_NLI-v2_dot_accuracy": 0.140625,
673
+ "eval_NLI-v2_euclidean_accuracy": 1.0,
674
+ "eval_NLI-v2_manhattan_accuracy": 1.0,
675
+ "eval_NLI-v2_max_accuracy": 1.0,
676
+ "eval_VitaminC_cosine_accuracy": 0.55078125,
677
+ "eval_VitaminC_cosine_accuracy_threshold": 0.9581937789916992,
678
+ "eval_VitaminC_cosine_ap": 0.5266754197706615,
679
+ "eval_VitaminC_cosine_f1": 0.6507936507936508,
680
+ "eval_VitaminC_cosine_f1_threshold": 0.7856193780899048,
681
+ "eval_VitaminC_cosine_precision": 0.4823529411764706,
682
+ "eval_VitaminC_cosine_recall": 1.0,
683
+ "eval_VitaminC_dot_accuracy": 0.54296875,
684
+ "eval_VitaminC_dot_accuracy_threshold": 461.7385559082031,
685
+ "eval_VitaminC_dot_ap": 0.5148613370532991,
686
+ "eval_VitaminC_dot_f1": 0.6542553191489362,
687
+ "eval_VitaminC_dot_f1_threshold": 349.2696838378906,
688
+ "eval_VitaminC_dot_precision": 0.48616600790513836,
689
+ "eval_VitaminC_dot_recall": 1.0,
690
+ "eval_VitaminC_euclidean_accuracy": 0.55078125,
691
+ "eval_VitaminC_euclidean_accuracy_threshold": 6.300814628601074,
692
+ "eval_VitaminC_euclidean_ap": 0.5255273000837065,
693
+ "eval_VitaminC_euclidean_f1": 0.6472148541114058,
694
+ "eval_VitaminC_euclidean_f1_threshold": 14.141785621643066,
695
+ "eval_VitaminC_euclidean_precision": 0.48031496062992124,
696
+ "eval_VitaminC_euclidean_recall": 0.991869918699187,
697
+ "eval_VitaminC_manhattan_accuracy": 0.546875,
698
+ "eval_VitaminC_manhattan_accuracy_threshold": 111.04672241210938,
699
+ "eval_VitaminC_manhattan_ap": 0.5200203121459024,
700
+ "eval_VitaminC_manhattan_f1": 0.6542553191489362,
701
+ "eval_VitaminC_manhattan_f1_threshold": 232.8947296142578,
702
+ "eval_VitaminC_manhattan_precision": 0.48616600790513836,
703
+ "eval_VitaminC_manhattan_recall": 1.0,
704
+ "eval_VitaminC_max_accuracy": 0.55078125,
705
+ "eval_VitaminC_max_accuracy_threshold": 461.7385559082031,
706
+ "eval_VitaminC_max_ap": 0.5266754197706615,
707
+ "eval_VitaminC_max_f1": 0.6542553191489362,
708
+ "eval_VitaminC_max_f1_threshold": 349.2696838378906,
709
+ "eval_VitaminC_max_precision": 0.48616600790513836,
710
+ "eval_VitaminC_max_recall": 1.0,
711
+ "eval_sequential_score": 0.5266754197706615,
712
+ "eval_sts-test_pearson_cosine": 0.05719150926706661,
713
+ "eval_sts-test_pearson_dot": 0.2946717285740077,
714
+ "eval_sts-test_pearson_euclidean": 0.05619658108650874,
715
+ "eval_sts-test_pearson_manhattan": 0.07947531332122215,
716
+ "eval_sts-test_pearson_max": 0.2946717285740077,
717
+ "eval_sts-test_spearman_cosine": 0.11293393987132996,
718
+ "eval_sts-test_spearman_dot": 0.3047454601059247,
719
+ "eval_sts-test_spearman_euclidean": 0.07970793034154534,
720
+ "eval_sts-test_spearman_manhattan": 0.09745056138447042,
721
+ "eval_sts-test_spearman_max": 0.3047454601059247,
722
+ "eval_vitaminc-pairs_loss": 2.715181589126587,
723
+ "eval_vitaminc-pairs_runtime": 1.4378,
724
+ "eval_vitaminc-pairs_samples_per_second": 75.113,
725
+ "eval_vitaminc-pairs_steps_per_second": 1.391,
726
+ "step": 10
727
+ },
728
+ {
729
+ "epoch": 0.29069767441860467,
730
+ "eval_negation-triplets_loss": 5.137111186981201,
731
+ "eval_negation-triplets_runtime": 0.296,
732
+ "eval_negation-triplets_samples_per_second": 216.18,
733
+ "eval_negation-triplets_steps_per_second": 3.378,
734
+ "step": 10
735
+ },
736
+ {
737
+ "epoch": 0.29069767441860467,
738
+ "eval_scitail-pairs-pos_loss": 2.0431957244873047,
739
+ "eval_scitail-pairs-pos_runtime": 0.3627,
740
+ "eval_scitail-pairs-pos_samples_per_second": 148.876,
741
+ "eval_scitail-pairs-pos_steps_per_second": 2.757,
742
+ "step": 10
743
+ },
744
+ {
745
+ "epoch": 0.29069767441860467,
746
+ "eval_xsum-pairs_loss": 5.958922863006592,
747
+ "eval_xsum-pairs_runtime": 3.4184,
748
+ "eval_xsum-pairs_samples_per_second": 37.445,
749
+ "eval_xsum-pairs_steps_per_second": 0.585,
750
+ "step": 10
751
+ },
752
+ {
753
+ "epoch": 0.29069767441860467,
754
+ "eval_sciq_pairs_loss": 0.3122929334640503,
755
+ "eval_sciq_pairs_runtime": 3.4301,
756
+ "eval_sciq_pairs_samples_per_second": 37.316,
757
+ "eval_sciq_pairs_steps_per_second": 0.583,
758
+ "step": 10
759
+ },
760
+ {
761
+ "epoch": 0.29069767441860467,
762
+ "eval_qasc_pairs_loss": 3.0287587642669678,
763
+ "eval_qasc_pairs_runtime": 0.6451,
764
+ "eval_qasc_pairs_samples_per_second": 198.424,
765
+ "eval_qasc_pairs_steps_per_second": 3.1,
766
+ "step": 10
767
+ },
768
+ {
769
+ "epoch": 0.29069767441860467,
770
+ "eval_openbookqa_pairs_loss": 4.213818073272705,
771
+ "eval_openbookqa_pairs_runtime": 0.5726,
772
+ "eval_openbookqa_pairs_samples_per_second": 223.54,
773
+ "eval_openbookqa_pairs_steps_per_second": 3.493,
774
+ "step": 10
775
+ },
776
+ {
777
+ "epoch": 0.29069767441860467,
778
+ "eval_msmarco_pairs_loss": 6.443023204803467,
779
+ "eval_msmarco_pairs_runtime": 1.2926,
780
+ "eval_msmarco_pairs_samples_per_second": 99.023,
781
+ "eval_msmarco_pairs_steps_per_second": 1.547,
782
+ "step": 10
783
+ },
784
+ {
785
+ "epoch": 0.29069767441860467,
786
+ "eval_nq_pairs_loss": 3.1351583003997803,
787
+ "eval_nq_pairs_runtime": 2.579,
788
+ "eval_nq_pairs_samples_per_second": 49.632,
789
+ "eval_nq_pairs_steps_per_second": 0.776,
790
+ "step": 10
791
+ },
792
+ {
793
+ "epoch": 0.29069767441860467,
794
+ "eval_trivia_pairs_loss": 3.116779088973999,
795
+ "eval_trivia_pairs_runtime": 4.3604,
796
+ "eval_trivia_pairs_samples_per_second": 29.355,
797
+ "eval_trivia_pairs_steps_per_second": 0.459,
798
+ "step": 10
799
+ },
800
+ {
801
+ "epoch": 0.29069767441860467,
802
+ "eval_gooaq_pairs_loss": 6.086311340332031,
803
+ "eval_gooaq_pairs_runtime": 0.9098,
804
+ "eval_gooaq_pairs_samples_per_second": 140.686,
805
+ "eval_gooaq_pairs_steps_per_second": 2.198,
806
+ "step": 10
807
+ },
808
+ {
809
+ "epoch": 0.29069767441860467,
810
+ "eval_paws-pos_loss": 2.509471893310547,
811
+ "eval_paws-pos_runtime": 0.6868,
812
+ "eval_paws-pos_samples_per_second": 186.374,
813
+ "eval_paws-pos_steps_per_second": 2.912,
814
+ "step": 10
815
+ },
816
+ {
817
+ "epoch": 0.31976744186046513,
818
+ "grad_norm": 8.210014343261719,
819
+ "learning_rate": 3.6666666666666666e-05,
820
+ "loss": 6.4394,
821
+ "step": 11
822
+ },
823
+ {
824
+ "epoch": 0.3488372093023256,
825
+ "grad_norm": 4.276110649108887,
826
+ "learning_rate": 4e-05,
827
+ "loss": 6.2215,
828
+ "step": 12
829
+ },
830
+ {
831
+ "epoch": 0.3488372093023256,
832
+ "eval_NLI-v2_cosine_accuracy": 1.0,
833
+ "eval_NLI-v2_dot_accuracy": 0.125,
834
+ "eval_NLI-v2_euclidean_accuracy": 1.0,
835
+ "eval_NLI-v2_manhattan_accuracy": 1.0,
836
+ "eval_NLI-v2_max_accuracy": 1.0,
837
+ "eval_VitaminC_cosine_accuracy": 0.54296875,
838
+ "eval_VitaminC_cosine_accuracy_threshold": 0.967127799987793,
839
+ "eval_VitaminC_cosine_ap": 0.5255548283042585,
840
+ "eval_VitaminC_cosine_f1": 0.648936170212766,
841
+ "eval_VitaminC_cosine_f1_threshold": 0.8157126903533936,
842
+ "eval_VitaminC_cosine_precision": 0.48221343873517786,
843
+ "eval_VitaminC_cosine_recall": 0.991869918699187,
844
+ "eval_VitaminC_dot_accuracy": 0.54296875,
845
+ "eval_VitaminC_dot_accuracy_threshold": 494.9104309082031,
846
+ "eval_VitaminC_dot_ap": 0.5147267697897451,
847
+ "eval_VitaminC_dot_f1": 0.6525198938992042,
848
+ "eval_VitaminC_dot_f1_threshold": 383.80987548828125,
849
+ "eval_VitaminC_dot_precision": 0.484251968503937,
850
+ "eval_VitaminC_dot_recall": 1.0,
851
+ "eval_VitaminC_euclidean_accuracy": 0.546875,
852
+ "eval_VitaminC_euclidean_accuracy_threshold": 5.804706573486328,
853
+ "eval_VitaminC_euclidean_ap": 0.5251569661235761,
854
+ "eval_VitaminC_euclidean_f1": 0.648936170212766,
855
+ "eval_VitaminC_euclidean_f1_threshold": 13.634963035583496,
856
+ "eval_VitaminC_euclidean_precision": 0.48221343873517786,
857
+ "eval_VitaminC_euclidean_recall": 0.991869918699187,
858
+ "eval_VitaminC_manhattan_accuracy": 0.55078125,
859
+ "eval_VitaminC_manhattan_accuracy_threshold": 115.37125396728516,
860
+ "eval_VitaminC_manhattan_ap": 0.5211715241173163,
861
+ "eval_VitaminC_manhattan_f1": 0.6525198938992042,
862
+ "eval_VitaminC_manhattan_f1_threshold": 222.44180297851562,
863
+ "eval_VitaminC_manhattan_precision": 0.484251968503937,
864
+ "eval_VitaminC_manhattan_recall": 1.0,
865
+ "eval_VitaminC_max_accuracy": 0.55078125,
866
+ "eval_VitaminC_max_accuracy_threshold": 494.9104309082031,
867
+ "eval_VitaminC_max_ap": 0.5255548283042585,
868
+ "eval_VitaminC_max_f1": 0.6525198938992042,
869
+ "eval_VitaminC_max_f1_threshold": 383.80987548828125,
870
+ "eval_VitaminC_max_precision": 0.484251968503937,
871
+ "eval_VitaminC_max_recall": 1.0,
872
+ "eval_sequential_score": 0.5255548283042585,
873
+ "eval_sts-test_pearson_cosine": 0.07362980702318303,
874
+ "eval_sts-test_pearson_dot": 0.2873931216538933,
875
+ "eval_sts-test_pearson_euclidean": 0.0777294747568082,
876
+ "eval_sts-test_pearson_manhattan": 0.09990133962066747,
877
+ "eval_sts-test_pearson_max": 0.2873931216538933,
878
+ "eval_sts-test_spearman_cosine": 0.1307277513542105,
879
+ "eval_sts-test_spearman_dot": 0.29102261593628775,
880
+ "eval_sts-test_spearman_euclidean": 0.10450013684970334,
881
+ "eval_sts-test_spearman_manhattan": 0.1162952592302797,
882
+ "eval_sts-test_spearman_max": 0.29102261593628775,
883
+ "eval_vitaminc-pairs_loss": 2.7209765911102295,
884
+ "eval_vitaminc-pairs_runtime": 1.4394,
885
+ "eval_vitaminc-pairs_samples_per_second": 75.03,
886
+ "eval_vitaminc-pairs_steps_per_second": 1.389,
887
+ "step": 12
888
+ },
889
+ {
890
+ "epoch": 0.3488372093023256,
891
+ "eval_negation-triplets_loss": 5.181106090545654,
892
+ "eval_negation-triplets_runtime": 0.2954,
893
+ "eval_negation-triplets_samples_per_second": 216.692,
894
+ "eval_negation-triplets_steps_per_second": 3.386,
895
+ "step": 12
896
+ },
897
+ {
898
+ "epoch": 0.3488372093023256,
899
+ "eval_scitail-pairs-pos_loss": 2.0938117504119873,
900
+ "eval_scitail-pairs-pos_runtime": 0.3664,
901
+ "eval_scitail-pairs-pos_samples_per_second": 147.368,
902
+ "eval_scitail-pairs-pos_steps_per_second": 2.729,
903
+ "step": 12
904
+ },
905
+ {
906
+ "epoch": 0.3488372093023256,
907
+ "eval_xsum-pairs_loss": 5.798585891723633,
908
+ "eval_xsum-pairs_runtime": 3.4113,
909
+ "eval_xsum-pairs_samples_per_second": 37.522,
910
+ "eval_xsum-pairs_steps_per_second": 0.586,
911
+ "step": 12
912
+ },
913
+ {
914
+ "epoch": 0.3488372093023256,
915
+ "eval_sciq_pairs_loss": 0.30859243869781494,
916
+ "eval_sciq_pairs_runtime": 3.4356,
917
+ "eval_sciq_pairs_samples_per_second": 37.256,
918
+ "eval_sciq_pairs_steps_per_second": 0.582,
919
+ "step": 12
920
+ },
921
+ {
922
+ "epoch": 0.3488372093023256,
923
+ "eval_qasc_pairs_loss": 2.919435739517212,
924
+ "eval_qasc_pairs_runtime": 0.6393,
925
+ "eval_qasc_pairs_samples_per_second": 200.207,
926
+ "eval_qasc_pairs_steps_per_second": 3.128,
927
+ "step": 12
928
+ },
929
+ {
930
+ "epoch": 0.3488372093023256,
931
+ "eval_openbookqa_pairs_loss": 4.164936542510986,
932
+ "eval_openbookqa_pairs_runtime": 0.5772,
933
+ "eval_openbookqa_pairs_samples_per_second": 221.764,
934
+ "eval_openbookqa_pairs_steps_per_second": 3.465,
935
+ "step": 12
936
+ },
937
+ {
938
+ "epoch": 0.3488372093023256,
939
+ "eval_msmarco_pairs_loss": 6.209752082824707,
940
+ "eval_msmarco_pairs_runtime": 1.296,
941
+ "eval_msmarco_pairs_samples_per_second": 98.767,
942
+ "eval_msmarco_pairs_steps_per_second": 1.543,
943
+ "step": 12
944
+ },
945
+ {
946
+ "epoch": 0.3488372093023256,
947
+ "eval_nq_pairs_loss": 3.008612632751465,
948
+ "eval_nq_pairs_runtime": 2.5746,
949
+ "eval_nq_pairs_samples_per_second": 49.716,
950
+ "eval_nq_pairs_steps_per_second": 0.777,
951
+ "step": 12
952
+ },
953
+ {
954
+ "epoch": 0.3488372093023256,
955
+ "eval_trivia_pairs_loss": 2.97579288482666,
956
+ "eval_trivia_pairs_runtime": 4.3688,
957
+ "eval_trivia_pairs_samples_per_second": 29.299,
958
+ "eval_trivia_pairs_steps_per_second": 0.458,
959
+ "step": 12
960
+ },
961
+ {
962
+ "epoch": 0.3488372093023256,
963
+ "eval_gooaq_pairs_loss": 5.966132640838623,
964
+ "eval_gooaq_pairs_runtime": 0.9077,
965
+ "eval_gooaq_pairs_samples_per_second": 141.017,
966
+ "eval_gooaq_pairs_steps_per_second": 2.203,
967
+ "step": 12
968
+ },
969
+ {
970
+ "epoch": 0.3488372093023256,
971
+ "eval_paws-pos_loss": 2.638349771499634,
972
+ "eval_paws-pos_runtime": 0.6823,
973
+ "eval_paws-pos_samples_per_second": 187.606,
974
+ "eval_paws-pos_steps_per_second": 2.931,
975
+ "step": 12
976
+ }
977
+ ],
978
+ "logging_steps": 1,
979
+ "max_steps": 34,
980
+ "num_input_tokens_seen": 0,
981
+ "num_train_epochs": 1,
982
+ "save_steps": 2,
983
+ "stateful_callbacks": {
984
+ "TrainerControl": {
985
+ "args": {
986
+ "should_epoch_stop": false,
987
+ "should_evaluate": false,
988
+ "should_log": false,
989
+ "should_save": true,
990
+ "should_training_stop": false
991
+ },
992
+ "attributes": {}
993
+ }
994
+ },
995
+ "total_flos": 0.0,
996
+ "train_batch_size": 224,
997
+ "trial_name": null,
998
+ "trial_params": null
999
+ }
checkpoint-12/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edceb7011ecd574f3e3fa5380c854a14e3b5904426b5550bdb62acd6cda68870
3
+ size 5688