Hugging Face's logo

bobox
/
XLMRoBERTaM3-CustomPoolinMRL-v1.07-2048MLP-step1-checkpoints-tmp

PyTorch
TensorBoard
xlm-roberta
Model card Files
xet
Metrics Community
bobox commited on
Commit
70d027b
·
verified ·
1 Parent(s): 2bddd48

Training in progress, step 462, checkpoint

Browse files
Files changed (19) hide show
  1. .gitattributes +1 -0
  2. checkpoint-462/1_AdvancedWeightedPooling/config.json +16 -0
  3. checkpoint-462/1_AdvancedWeightedPooling/pytorch_model.bin +3 -0
  4. checkpoint-462/README.md +0 -0
  5. checkpoint-462/config.json +27 -0
  6. checkpoint-462/config_sentence_transformers.json +10 -0
  7. checkpoint-462/modules.json +14 -0
  8. checkpoint-462/optimizer.pt +3 -0
  9. checkpoint-462/pytorch_model.bin +3 -0
  10. checkpoint-462/rng_state.pth +3 -0
  11. checkpoint-462/scaler.pt +3 -0
  12. checkpoint-462/scheduler.pt +3 -0
  13. checkpoint-462/sentence_bert_config.json +4 -0
  14. checkpoint-462/sentencepiece.bpe.model +3 -0
  15. checkpoint-462/special_tokens_map.json +51 -0
  16. checkpoint-462/tokenizer.json +3 -0
  17. checkpoint-462/tokenizer_config.json +56 -0
  18. checkpoint-462/trainer_state.json +3442 -0
  19. checkpoint-462/training_args.bin +3 -0
.gitattributes CHANGED
@@ -35,3 +35,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
  tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
  checkpoint-231/tokenizer.json filter=lfs diff=lfs merge=lfs -text
 
 
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
  tokenizer.json filter=lfs diff=lfs merge=lfs -text
37
  checkpoint-231/tokenizer.json filter=lfs diff=lfs merge=lfs -text
38
+ checkpoint-462/tokenizer.json filter=lfs diff=lfs merge=lfs -text
checkpoint-462/1_AdvancedWeightedPooling/config.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "embed_dim": 1024,
3
+ "num_heads": 4,
4
+ "dropout": 0.1,
5
+ "bias": true,
6
+ "use_layernorm": true,
7
+ "use_MLP": true,
8
+ "MLP_h_size": 2048,
9
+ "MLP_output_size": 2048,
10
+ "use_residual_MLP": "no",
11
+ "MLP_type": "swiglu_h+swiglu_d",
12
+ "h2_size": 2048,
13
+ "ignore_cls_as_kv": true,
14
+ "expand_emb_dim_to": 0,
15
+ "compress_output_dim_to": 0
16
+ }
checkpoint-462/1_AdvancedWeightedPooling/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:edc368638f42f9bfb02aa222287a23bf7065260f585931e24805e7183dd71bde
3
+ size 83965726
checkpoint-462/README.md ADDED
The diff for this file is too large to render. See raw diff
 
checkpoint-462/config.json ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "architectures": [
3
+ "XLMRobertaModel"
4
+ ],
5
+ "attention_probs_dropout_prob": 0.1,
6
+ "bos_token_id": 0,
7
+ "classifier_dropout": null,
8
+ "eos_token_id": 2,
9
+ "hidden_act": "gelu",
10
+ "hidden_dropout_prob": 0.1,
11
+ "hidden_size": 1024,
12
+ "initializer_range": 0.02,
13
+ "intermediate_size": 4096,
14
+ "layer_norm_eps": 1e-05,
15
+ "max_position_embeddings": 8194,
16
+ "model_type": "xlm-roberta",
17
+ "num_attention_heads": 16,
18
+ "num_hidden_layers": 24,
19
+ "output_past": true,
20
+ "pad_token_id": 1,
21
+ "position_embedding_type": "absolute",
22
+ "torch_dtype": "float32",
23
+ "transformers_version": "4.51.1",
24
+ "type_vocab_size": 1,
25
+ "use_cache": true,
26
+ "vocab_size": 250002
27
+ }
checkpoint-462/config_sentence_transformers.json ADDED
@@ -0,0 +1,10 @@
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "__version__": {
3
+ "sentence_transformers": "4.1.0",
4
+ "transformers": "4.51.1",
5
+ "pytorch": "2.5.1+cu124"
6
+ },
7
+ "prompts": {},
8
+ "default_prompt_name": null,
9
+ "similarity_fn_name": "cosine"
10
+ }
checkpoint-462/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_AdvancedWeightedPooling",
12
+ "type": "__main__.AdvancedWeightedPooling"
13
+ }
14
+ ]
checkpoint-462/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:570047832bdf8f1e8838bd090519a6a44974f6dee205ad3d1a3c35344bd0d72d
3
+ size 167933422
checkpoint-462/pytorch_model.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0e969b03be79a3f792216f1febef058e86c2b7f27374fb1e044d87938ceea1ba
3
+ size 2271151270
checkpoint-462/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:608f3352765b20626d784e9c55dcd2e4ec71e0792f539f3238e4a2061ca53a2c
3
+ size 14244
checkpoint-462/scaler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:13d25576324708607f9af66da50abc00c1d3793fcced667aa67d6adfe5eec34e
3
+ size 988
checkpoint-462/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d261dfa604f6227d4dfa171f9043bff415bed98871470feb4c9f394bef7fd1f
3
+ size 1000
checkpoint-462/sentence_bert_config.json ADDED
@@ -0,0 +1,4 @@
 
 
 
 
 
1
+ {
2
+ "max_seq_length": 512,
3
+ "do_lower_case": false
4
+ }
checkpoint-462/sentencepiece.bpe.model ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cfc8146abe2a0488e9e2a0c56de7952f7c11ab059eca145a0a727afce0db2865
3
+ size 5069051
checkpoint-462/special_tokens_map.json ADDED
@@ -0,0 +1,51 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "bos_token": {
3
+ "content": "<s>",
4
+ "lstrip": false,
5
+ "normalized": false,
6
+ "rstrip": false,
7
+ "single_word": false
8
+ },
9
+ "cls_token": {
10
+ "content": "<s>",
11
+ "lstrip": false,
12
+ "normalized": false,
13
+ "rstrip": false,
14
+ "single_word": false
15
+ },
16
+ "eos_token": {
17
+ "content": "</s>",
18
+ "lstrip": false,
19
+ "normalized": false,
20
+ "rstrip": false,
21
+ "single_word": false
22
+ },
23
+ "mask_token": {
24
+ "content": "<mask>",
25
+ "lstrip": true,
26
+ "normalized": false,
27
+ "rstrip": false,
28
+ "single_word": false
29
+ },
30
+ "pad_token": {
31
+ "content": "<pad>",
32
+ "lstrip": false,
33
+ "normalized": false,
34
+ "rstrip": false,
35
+ "single_word": false
36
+ },
37
+ "sep_token": {
38
+ "content": "</s>",
39
+ "lstrip": false,
40
+ "normalized": false,
41
+ "rstrip": false,
42
+ "single_word": false
43
+ },
44
+ "unk_token": {
45
+ "content": "<unk>",
46
+ "lstrip": false,
47
+ "normalized": false,
48
+ "rstrip": false,
49
+ "single_word": false
50
+ }
51
+ }
checkpoint-462/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d9a6af42442a3e3e9f05f618eae0bb2d98ca4f6a6406cb80ef7a4fa865204d61
3
+ size 17083052
checkpoint-462/tokenizer_config.json ADDED
@@ -0,0 +1,56 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "added_tokens_decoder": {
3
+ "0": {
4
+ "content": "<s>",
5
+ "lstrip": false,
6
+ "normalized": false,
7
+ "rstrip": false,
8
+ "single_word": false,
9
+ "special": true
10
+ },
11
+ "1": {
12
+ "content": "<pad>",
13
+ "lstrip": false,
14
+ "normalized": false,
15
+ "rstrip": false,
16
+ "single_word": false,
17
+ "special": true
18
+ },
19
+ "2": {
20
+ "content": "</s>",
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": false,
31
+ "rstrip": false,
32
+ "single_word": false,
33
+ "special": true
34
+ },
35
+ "250001": {
36
+ "content": "<mask>",
37
+ "lstrip": true,
38
+ "normalized": false,
39
+ "rstrip": false,
40
+ "single_word": false,
41
+ "special": true
42
+ }
43
+ },
44
+ "bos_token": "<s>",
45
+ "clean_up_tokenization_spaces": true,
46
+ "cls_token": "<s>",
47
+ "eos_token": "</s>",
48
+ "extra_special_tokens": {},
49
+ "mask_token": "<mask>",
50
+ "model_max_length": 8192,
51
+ "pad_token": "<pad>",
52
+ "sep_token": "</s>",
53
+ "sp_model_kwargs": {},
54
+ "tokenizer_class": "XLMRobertaTokenizer",
55
+ "unk_token": "<unk>"
56
+ }
checkpoint-462/trainer_state.json ADDED
@@ -0,0 +1,3442 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.6,
6
+ "eval_steps": 77,
7
+ "global_step": 462,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0012987012987012987,
14
+ "grad_norm": 0.011165409348905087,
15
+ "learning_rate": 0.0,
16
+ "loss": 11.01,
17
+ "step": 1
18
+ },
19
+ {
20
+ "epoch": 0.0025974025974025974,
21
+ "grad_norm": 0.011394021101295948,
22
+ "learning_rate": 2.1645021645021646e-07,
23
+ "loss": 11.01,
24
+ "step": 2
25
+ },
26
+ {
27
+ "epoch": 0.003896103896103896,
28
+ "grad_norm": 0.012114278972148895,
29
+ "learning_rate": 4.329004329004329e-07,
30
+ "loss": 11.0099,
31
+ "step": 3
32
+ },
33
+ {
34
+ "epoch": 0.005194805194805195,
35
+ "grad_norm": 0.011571550741791725,
36
+ "learning_rate": 6.493506493506494e-07,
37
+ "loss": 11.0097,
38
+ "step": 4
39
+ },
40
+ {
41
+ "epoch": 0.006493506493506494,
42
+ "grad_norm": 0.012371700257062912,
43
+ "learning_rate": 8.658008658008658e-07,
44
+ "loss": 11.0099,
45
+ "step": 5
46
+ },
47
+ {
48
+ "epoch": 0.007792207792207792,
49
+ "grad_norm": 0.012187338434159756,
50
+ "learning_rate": 1.0822510822510822e-06,
51
+ "loss": 11.0099,
52
+ "step": 6
53
+ },
54
+ {
55
+ "epoch": 0.00909090909090909,
56
+ "grad_norm": 0.012087634764611721,
57
+ "learning_rate": 1.2987012987012988e-06,
58
+ "loss": 11.0099,
59
+ "step": 7
60
+ },
61
+ {
62
+ "epoch": 0.01038961038961039,
63
+ "grad_norm": 0.011910870671272278,
64
+ "learning_rate": 1.5151515151515152e-06,
65
+ "loss": 11.0099,
66
+ "step": 8
67
+ },
68
+ {
69
+ "epoch": 0.011688311688311689,
70
+ "grad_norm": 0.01282087154686451,
71
+ "learning_rate": 1.7316017316017317e-06,
72
+ "loss": 11.0099,
73
+ "step": 9
74
+ },
75
+ {
76
+ "epoch": 0.012987012987012988,
77
+ "grad_norm": 0.01281706988811493,
78
+ "learning_rate": 1.948051948051948e-06,
79
+ "loss": 11.0099,
80
+ "step": 10
81
+ },
82
+ {
83
+ "epoch": 0.014285714285714285,
84
+ "grad_norm": 0.013667456805706024,
85
+ "learning_rate": 2.1645021645021643e-06,
86
+ "loss": 11.0097,
87
+ "step": 11
88
+ },
89
+ {
90
+ "epoch": 0.015584415584415584,
91
+ "grad_norm": 0.01482036616653204,
92
+ "learning_rate": 2.3809523809523808e-06,
93
+ "loss": 11.0097,
94
+ "step": 12
95
+ },
96
+ {
97
+ "epoch": 0.016883116883116882,
98
+ "grad_norm": 0.014698008075356483,
99
+ "learning_rate": 2.5974025974025976e-06,
100
+ "loss": 11.0097,
101
+ "step": 13
102
+ },
103
+ {
104
+ "epoch": 0.01818181818181818,
105
+ "grad_norm": 0.014225203543901443,
106
+ "learning_rate": 2.813852813852814e-06,
107
+ "loss": 11.0097,
108
+ "step": 14
109
+ },
110
+ {
111
+ "epoch": 0.01948051948051948,
112
+ "grad_norm": 0.016268573701381683,
113
+ "learning_rate": 3.0303030303030305e-06,
114
+ "loss": 11.0096,
115
+ "step": 15
116
+ },
117
+ {
118
+ "epoch": 0.02077922077922078,
119
+ "grad_norm": 0.017191331833600998,
120
+ "learning_rate": 3.2467532467532465e-06,
121
+ "loss": 11.0095,
122
+ "step": 16
123
+ },
124
+ {
125
+ "epoch": 0.02207792207792208,
126
+ "grad_norm": 0.016200311481952667,
127
+ "learning_rate": 3.4632034632034634e-06,
128
+ "loss": 11.0096,
129
+ "step": 17
130
+ },
131
+ {
132
+ "epoch": 0.023376623376623377,
133
+ "grad_norm": 0.019106043502688408,
134
+ "learning_rate": 3.67965367965368e-06,
135
+ "loss": 11.0094,
136
+ "step": 18
137
+ },
138
+ {
139
+ "epoch": 0.024675324675324677,
140
+ "grad_norm": 0.017216097563505173,
141
+ "learning_rate": 3.896103896103896e-06,
142
+ "loss": 11.0095,
143
+ "step": 19
144
+ },
145
+ {
146
+ "epoch": 0.025974025974025976,
147
+ "grad_norm": 0.020744245499372482,
148
+ "learning_rate": 4.112554112554113e-06,
149
+ "loss": 11.0093,
150
+ "step": 20
151
+ },
152
+ {
153
+ "epoch": 0.02727272727272727,
154
+ "grad_norm": 0.02340623177587986,
155
+ "learning_rate": 4.329004329004329e-06,
156
+ "loss": 11.0091,
157
+ "step": 21
158
+ },
159
+ {
160
+ "epoch": 0.02857142857142857,
161
+ "grad_norm": 0.023833388462662697,
162
+ "learning_rate": 4.5454545454545455e-06,
163
+ "loss": 11.0091,
164
+ "step": 22
165
+ },
166
+ {
167
+ "epoch": 0.02987012987012987,
168
+ "grad_norm": 0.027158446609973907,
169
+ "learning_rate": 4.7619047619047615e-06,
170
+ "loss": 11.0089,
171
+ "step": 23
172
+ },
173
+ {
174
+ "epoch": 0.03116883116883117,
175
+ "grad_norm": 0.029171699658036232,
176
+ "learning_rate": 4.978354978354978e-06,
177
+ "loss": 11.0088,
178
+ "step": 24
179
+ },
180
+ {
181
+ "epoch": 0.032467532467532464,
182
+ "grad_norm": 0.03013516403734684,
183
+ "learning_rate": 5.194805194805195e-06,
184
+ "loss": 11.0087,
185
+ "step": 25
186
+ },
187
+ {
188
+ "epoch": 0.033766233766233764,
189
+ "grad_norm": 0.035284992307424545,
190
+ "learning_rate": 5.411255411255411e-06,
191
+ "loss": 11.0083,
192
+ "step": 26
193
+ },
194
+ {
195
+ "epoch": 0.03506493506493506,
196
+ "grad_norm": 0.03512846305966377,
197
+ "learning_rate": 5.627705627705628e-06,
198
+ "loss": 11.0083,
199
+ "step": 27
200
+ },
201
+ {
202
+ "epoch": 0.03636363636363636,
203
+ "grad_norm": 0.044411927461624146,
204
+ "learning_rate": 5.844155844155844e-06,
205
+ "loss": 11.0077,
206
+ "step": 28
207
+ },
208
+ {
209
+ "epoch": 0.03766233766233766,
210
+ "grad_norm": 0.04630338028073311,
211
+ "learning_rate": 6.060606060606061e-06,
212
+ "loss": 11.0075,
213
+ "step": 29
214
+ },
215
+ {
216
+ "epoch": 0.03896103896103896,
217
+ "grad_norm": 0.05681515112519264,
218
+ "learning_rate": 6.277056277056277e-06,
219
+ "loss": 11.0053,
220
+ "step": 30
221
+ },
222
+ {
223
+ "epoch": 0.04025974025974026,
224
+ "grad_norm": 0.05574386194348335,
225
+ "learning_rate": 6.493506493506493e-06,
226
+ "loss": 11.007,
227
+ "step": 31
228
+ },
229
+ {
230
+ "epoch": 0.04155844155844156,
231
+ "grad_norm": 0.07001659274101257,
232
+ "learning_rate": 6.709956709956711e-06,
233
+ "loss": 11.006,
234
+ "step": 32
235
+ },
236
+ {
237
+ "epoch": 0.04285714285714286,
238
+ "grad_norm": 0.07613399624824524,
239
+ "learning_rate": 6.926406926406927e-06,
240
+ "loss": 11.0056,
241
+ "step": 33
242
+ },
243
+ {
244
+ "epoch": 0.04415584415584416,
245
+ "grad_norm": 0.0930153876543045,
246
+ "learning_rate": 7.142857142857143e-06,
247
+ "loss": 11.0045,
248
+ "step": 34
249
+ },
250
+ {
251
+ "epoch": 0.045454545454545456,
252
+ "grad_norm": 0.11075685918331146,
253
+ "learning_rate": 7.35930735930736e-06,
254
+ "loss": 11.0035,
255
+ "step": 35
256
+ },
257
+ {
258
+ "epoch": 0.046753246753246755,
259
+ "grad_norm": 0.1421954482793808,
260
+ "learning_rate": 7.5757575757575764e-06,
261
+ "loss": 11.0017,
262
+ "step": 36
263
+ },
264
+ {
265
+ "epoch": 0.048051948051948054,
266
+ "grad_norm": 0.165273517370224,
267
+ "learning_rate": 7.792207792207792e-06,
268
+ "loss": 11.0001,
269
+ "step": 37
270
+ },
271
+ {
272
+ "epoch": 0.04935064935064935,
273
+ "grad_norm": 0.2073814570903778,
274
+ "learning_rate": 8.008658008658008e-06,
275
+ "loss": 10.9974,
276
+ "step": 38
277
+ },
278
+ {
279
+ "epoch": 0.05064935064935065,
280
+ "grad_norm": 0.240090012550354,
281
+ "learning_rate": 8.225108225108225e-06,
282
+ "loss": 10.9955,
283
+ "step": 39
284
+ },
285
+ {
286
+ "epoch": 0.05194805194805195,
287
+ "grad_norm": 0.2896934747695923,
288
+ "learning_rate": 8.441558441558442e-06,
289
+ "loss": 10.9926,
290
+ "step": 40
291
+ },
292
+ {
293
+ "epoch": 0.053246753246753244,
294
+ "grad_norm": 0.3400895595550537,
295
+ "learning_rate": 8.658008658008657e-06,
296
+ "loss": 10.9886,
297
+ "step": 41
298
+ },
299
+ {
300
+ "epoch": 0.05454545454545454,
301
+ "grad_norm": 0.4271922707557678,
302
+ "learning_rate": 8.874458874458876e-06,
303
+ "loss": 10.983,
304
+ "step": 42
305
+ },
306
+ {
307
+ "epoch": 0.05584415584415584,
308
+ "grad_norm": 0.5322330594062805,
309
+ "learning_rate": 9.090909090909091e-06,
310
+ "loss": 10.977,
311
+ "step": 43
312
+ },
313
+ {
314
+ "epoch": 0.05714285714285714,
315
+ "grad_norm": 0.6354700922966003,
316
+ "learning_rate": 9.307359307359308e-06,
317
+ "loss": 10.9703,
318
+ "step": 44
319
+ },
320
+ {
321
+ "epoch": 0.05844155844155844,
322
+ "grad_norm": 0.8793215155601501,
323
+ "learning_rate": 9.523809523809523e-06,
324
+ "loss": 10.9547,
325
+ "step": 45
326
+ },
327
+ {
328
+ "epoch": 0.05974025974025974,
329
+ "grad_norm": 1.008970856666565,
330
+ "learning_rate": 9.740259740259742e-06,
331
+ "loss": 10.9435,
332
+ "step": 46
333
+ },
334
+ {
335
+ "epoch": 0.06103896103896104,
336
+ "grad_norm": 1.2915687561035156,
337
+ "learning_rate": 9.956709956709957e-06,
338
+ "loss": 10.9263,
339
+ "step": 47
340
+ },
341
+ {
342
+ "epoch": 0.06233766233766234,
343
+ "grad_norm": 1.5786389112472534,
344
+ "learning_rate": 1.0173160173160174e-05,
345
+ "loss": 10.9014,
346
+ "step": 48
347
+ },
348
+ {
349
+ "epoch": 0.06363636363636363,
350
+ "grad_norm": 2.4445056915283203,
351
+ "learning_rate": 1.038961038961039e-05,
352
+ "loss": 10.8546,
353
+ "step": 49
354
+ },
355
+ {
356
+ "epoch": 0.06493506493506493,
357
+ "grad_norm": 2.5852527618408203,
358
+ "learning_rate": 1.0606060606060607e-05,
359
+ "loss": 10.8134,
360
+ "step": 50
361
+ },
362
+ {
363
+ "epoch": 0.06623376623376623,
364
+ "grad_norm": 2.70867919921875,
365
+ "learning_rate": 1.0822510822510823e-05,
366
+ "loss": 10.7805,
367
+ "step": 51
368
+ },
369
+ {
370
+ "epoch": 0.06753246753246753,
371
+ "grad_norm": 2.796766757965088,
372
+ "learning_rate": 1.103896103896104e-05,
373
+ "loss": 10.7495,
374
+ "step": 52
375
+ },
376
+ {
377
+ "epoch": 0.06883116883116883,
378
+ "grad_norm": 2.8984150886535645,
379
+ "learning_rate": 1.1255411255411256e-05,
380
+ "loss": 10.6997,
381
+ "step": 53
382
+ },
383
+ {
384
+ "epoch": 0.07012987012987013,
385
+ "grad_norm": 3.7665364742279053,
386
+ "learning_rate": 1.1471861471861473e-05,
387
+ "loss": 10.5774,
388
+ "step": 54
389
+ },
390
+ {
391
+ "epoch": 0.07142857142857142,
392
+ "grad_norm": 4.654474258422852,
393
+ "learning_rate": 1.1688311688311688e-05,
394
+ "loss": 10.5443,
395
+ "step": 55
396
+ },
397
+ {
398
+ "epoch": 0.07272727272727272,
399
+ "grad_norm": 5.21140718460083,
400
+ "learning_rate": 1.1904761904761905e-05,
401
+ "loss": 10.4336,
402
+ "step": 56
403
+ },
404
+ {
405
+ "epoch": 0.07402597402597402,
406
+ "grad_norm": 5.744906425476074,
407
+ "learning_rate": 1.2121212121212122e-05,
408
+ "loss": 10.3595,
409
+ "step": 57
410
+ },
411
+ {
412
+ "epoch": 0.07532467532467532,
413
+ "grad_norm": 7.051860332489014,
414
+ "learning_rate": 1.2337662337662339e-05,
415
+ "loss": 10.2175,
416
+ "step": 58
417
+ },
418
+ {
419
+ "epoch": 0.07662337662337662,
420
+ "grad_norm": 11.466012954711914,
421
+ "learning_rate": 1.2554112554112554e-05,
422
+ "loss": 10.667,
423
+ "step": 59
424
+ },
425
+ {
426
+ "epoch": 0.07792207792207792,
427
+ "grad_norm": 7.726934432983398,
428
+ "learning_rate": 1.2770562770562773e-05,
429
+ "loss": 10.0476,
430
+ "step": 60
431
+ },
432
+ {
433
+ "epoch": 0.07922077922077922,
434
+ "grad_norm": 8.461792945861816,
435
+ "learning_rate": 1.2987012987012986e-05,
436
+ "loss": 10.0367,
437
+ "step": 61
438
+ },
439
+ {
440
+ "epoch": 0.08051948051948052,
441
+ "grad_norm": 9.503273010253906,
442
+ "learning_rate": 1.3203463203463205e-05,
443
+ "loss": 10.0531,
444
+ "step": 62
445
+ },
446
+ {
447
+ "epoch": 0.08181818181818182,
448
+ "grad_norm": 9.650754928588867,
449
+ "learning_rate": 1.3419913419913421e-05,
450
+ "loss": 9.7963,
451
+ "step": 63
452
+ },
453
+ {
454
+ "epoch": 0.08311688311688312,
455
+ "grad_norm": 11.210810661315918,
456
+ "learning_rate": 1.3636363636363637e-05,
457
+ "loss": 9.4349,
458
+ "step": 64
459
+ },
460
+ {
461
+ "epoch": 0.08441558441558442,
462
+ "grad_norm": 11.565103530883789,
463
+ "learning_rate": 1.3852813852813853e-05,
464
+ "loss": 9.5733,
465
+ "step": 65
466
+ },
467
+ {
468
+ "epoch": 0.08571428571428572,
469
+ "grad_norm": 13.807746887207031,
470
+ "learning_rate": 1.406926406926407e-05,
471
+ "loss": 9.0604,
472
+ "step": 66
473
+ },
474
+ {
475
+ "epoch": 0.08701298701298701,
476
+ "grad_norm": 14.853765487670898,
477
+ "learning_rate": 1.4285714285714285e-05,
478
+ "loss": 8.7291,
479
+ "step": 67
480
+ },
481
+ {
482
+ "epoch": 0.08831168831168831,
483
+ "grad_norm": 15.971939086914062,
484
+ "learning_rate": 1.4502164502164502e-05,
485
+ "loss": 8.38,
486
+ "step": 68
487
+ },
488
+ {
489
+ "epoch": 0.08961038961038961,
490
+ "grad_norm": 16.851449966430664,
491
+ "learning_rate": 1.471861471861472e-05,
492
+ "loss": 8.2196,
493
+ "step": 69
494
+ },
495
+ {
496
+ "epoch": 0.09090909090909091,
497
+ "grad_norm": 18.741369247436523,
498
+ "learning_rate": 1.4935064935064936e-05,
499
+ "loss": 7.8955,
500
+ "step": 70
501
+ },
502
+ {
503
+ "epoch": 0.09220779220779221,
504
+ "grad_norm": 18.92623519897461,
505
+ "learning_rate": 1.5151515151515153e-05,
506
+ "loss": 7.2905,
507
+ "step": 71
508
+ },
509
+ {
510
+ "epoch": 0.09350649350649351,
511
+ "grad_norm": 20.634117126464844,
512
+ "learning_rate": 1.5367965367965366e-05,
513
+ "loss": 7.3614,
514
+ "step": 72
515
+ },
516
+ {
517
+ "epoch": 0.09480519480519481,
518
+ "grad_norm": 20.857202529907227,
519
+ "learning_rate": 1.5584415584415583e-05,
520
+ "loss": 6.6031,
521
+ "step": 73
522
+ },
523
+ {
524
+ "epoch": 0.09610389610389611,
525
+ "grad_norm": 22.039953231811523,
526
+ "learning_rate": 1.5800865800865803e-05,
527
+ "loss": 5.9224,
528
+ "step": 74
529
+ },
530
+ {
531
+ "epoch": 0.09740259740259741,
532
+ "grad_norm": 22.81475257873535,
533
+ "learning_rate": 1.6017316017316017e-05,
534
+ "loss": 5.7627,
535
+ "step": 75
536
+ },
537
+ {
538
+ "epoch": 0.0987012987012987,
539
+ "grad_norm": 22.588449478149414,
540
+ "learning_rate": 1.6233766233766234e-05,
541
+ "loss": 5.2194,
542
+ "step": 76
543
+ },
544
+ {
545
+ "epoch": 0.1,
546
+ "grad_norm": 24.180652618408203,
547
+ "learning_rate": 1.645021645021645e-05,
548
+ "loss": 5.3339,
549
+ "step": 77
550
+ },
551
+ {
552
+ "epoch": 0.1,
553
+ "eval_allNLI--triplets-1024_cosine_accuracy": 0.9453125,
554
+ "eval_allNLI--triplets-128_cosine_accuracy": 0.90625,
555
+ "eval_allNLI--triplets-256_cosine_accuracy": 0.9296875,
556
+ "eval_allNLI--triplets-32_cosine_accuracy": 0.9140625,
557
+ "eval_allNLI--triplets-512_cosine_accuracy": 0.9375,
558
+ "eval_allNLI-triplets_cosine_accuracy": 0.9375,
559
+ "eval_global_dataset_loss": 1.779486060142517,
560
+ "eval_global_dataset_runtime": 113.4191,
561
+ "eval_global_dataset_samples_per_second": 9.443,
562
+ "eval_global_dataset_steps_per_second": 0.079,
563
+ "eval_sequential_score": 0.9140625,
564
+ "eval_sts-test-1024_pearson_cosine": 0.8570094674494524,
565
+ "eval_sts-test-1024_spearman_cosine": 0.9032869635214775,
566
+ "eval_sts-test-128_pearson_cosine": 0.8454774972598778,
567
+ "eval_sts-test-128_spearman_cosine": 0.8941657316165149,
568
+ "eval_sts-test-256_pearson_cosine": 0.846547357013354,
569
+ "eval_sts-test-256_spearman_cosine": 0.8961152942356522,
570
+ "eval_sts-test-32_pearson_cosine": 0.8158296423835969,
571
+ "eval_sts-test-32_spearman_cosine": 0.8781449456856982,
572
+ "eval_sts-test-512_pearson_cosine": 0.8548761563192739,
573
+ "eval_sts-test-512_spearman_cosine": 0.9013753538201782,
574
+ "eval_sts-test-64_pearson_cosine": 0.8326302932611978,
575
+ "eval_sts-test-64_spearman_cosine": 0.8899336232311702,
576
+ "eval_sts-test_pearson_cosine": 0.8635863698279433,
577
+ "eval_sts-test_spearman_cosine": 0.906139519148046,
578
+ "step": 77
579
+ },
580
+ {
581
+ "epoch": 0.1012987012987013,
582
+ "grad_norm": 25.592409133911133,
583
+ "learning_rate": 1.6666666666666667e-05,
584
+ "loss": 5.0749,
585
+ "step": 78
586
+ },
587
+ {
588
+ "epoch": 0.1025974025974026,
589
+ "grad_norm": 25.978023529052734,
590
+ "learning_rate": 1.6883116883116884e-05,
591
+ "loss": 4.7881,
592
+ "step": 79
593
+ },
594
+ {
595
+ "epoch": 0.1038961038961039,
596
+ "grad_norm": 29.861833572387695,
597
+ "learning_rate": 1.70995670995671e-05,
598
+ "loss": 4.2653,
599
+ "step": 80
600
+ },
601
+ {
602
+ "epoch": 0.10519480519480519,
603
+ "grad_norm": 28.283815383911133,
604
+ "learning_rate": 1.7316017316017315e-05,
605
+ "loss": 4.0863,
606
+ "step": 81
607
+ },
608
+ {
609
+ "epoch": 0.10649350649350649,
610
+ "grad_norm": 29.428701400756836,
611
+ "learning_rate": 1.7532467532467535e-05,
612
+ "loss": 4.2845,
613
+ "step": 82
614
+ },
615
+ {
616
+ "epoch": 0.10779220779220779,
617
+ "grad_norm": 27.637096405029297,
618
+ "learning_rate": 1.7748917748917752e-05,
619
+ "loss": 3.8356,
620
+ "step": 83
621
+ },
622
+ {
623
+ "epoch": 0.10909090909090909,
624
+ "grad_norm": 28.08953857421875,
625
+ "learning_rate": 1.7965367965367965e-05,
626
+ "loss": 3.5652,
627
+ "step": 84
628
+ },
629
+ {
630
+ "epoch": 0.11038961038961038,
631
+ "grad_norm": 28.57408332824707,
632
+ "learning_rate": 1.8181818181818182e-05,
633
+ "loss": 4.1484,
634
+ "step": 85
635
+ },
636
+ {
637
+ "epoch": 0.11168831168831168,
638
+ "grad_norm": 26.506061553955078,
639
+ "learning_rate": 1.83982683982684e-05,
640
+ "loss": 3.4415,
641
+ "step": 86
642
+ },
643
+ {
644
+ "epoch": 0.11298701298701298,
645
+ "grad_norm": 26.066709518432617,
646
+ "learning_rate": 1.8614718614718616e-05,
647
+ "loss": 3.129,
648
+ "step": 87
649
+ },
650
+ {
651
+ "epoch": 0.11428571428571428,
652
+ "grad_norm": 27.7386531829834,
653
+ "learning_rate": 1.8831168831168833e-05,
654
+ "loss": 3.5604,
655
+ "step": 88
656
+ },
657
+ {
658
+ "epoch": 0.11558441558441558,
659
+ "grad_norm": 26.726911544799805,
660
+ "learning_rate": 1.9047619047619046e-05,
661
+ "loss": 3.6036,
662
+ "step": 89
663
+ },
664
+ {
665
+ "epoch": 0.11688311688311688,
666
+ "grad_norm": 22.78415870666504,
667
+ "learning_rate": 1.9264069264069266e-05,
668
+ "loss": 2.8276,
669
+ "step": 90
670
+ },
671
+ {
672
+ "epoch": 0.11818181818181818,
673
+ "grad_norm": 23.38465118408203,
674
+ "learning_rate": 1.9480519480519483e-05,
675
+ "loss": 3.4752,
676
+ "step": 91
677
+ },
678
+ {
679
+ "epoch": 0.11948051948051948,
680
+ "grad_norm": 24.04189109802246,
681
+ "learning_rate": 1.9696969696969697e-05,
682
+ "loss": 3.3796,
683
+ "step": 92
684
+ },
685
+ {
686
+ "epoch": 0.12077922077922078,
687
+ "grad_norm": 20.70046043395996,
688
+ "learning_rate": 1.9913419913419914e-05,
689
+ "loss": 2.7567,
690
+ "step": 93
691
+ },
692
+ {
693
+ "epoch": 0.12207792207792208,
694
+ "grad_norm": 22.61043930053711,
695
+ "learning_rate": 2.012987012987013e-05,
696
+ "loss": 3.2728,
697
+ "step": 94
698
+ },
699
+ {
700
+ "epoch": 0.12337662337662338,
701
+ "grad_norm": 21.59564208984375,
702
+ "learning_rate": 2.0346320346320347e-05,
703
+ "loss": 2.8496,
704
+ "step": 95
705
+ },
706
+ {
707
+ "epoch": 0.12467532467532468,
708
+ "grad_norm": 19.719247817993164,
709
+ "learning_rate": 2.0562770562770564e-05,
710
+ "loss": 2.6072,
711
+ "step": 96
712
+ },
713
+ {
714
+ "epoch": 0.12597402597402596,
715
+ "grad_norm": 19.73863410949707,
716
+ "learning_rate": 2.077922077922078e-05,
717
+ "loss": 2.5959,
718
+ "step": 97
719
+ },
720
+ {
721
+ "epoch": 0.12727272727272726,
722
+ "grad_norm": 20.621767044067383,
723
+ "learning_rate": 2.0995670995670998e-05,
724
+ "loss": 2.4787,
725
+ "step": 98
726
+ },
727
+ {
728
+ "epoch": 0.12857142857142856,
729
+ "grad_norm": 18.53163719177246,
730
+ "learning_rate": 2.1212121212121215e-05,
731
+ "loss": 2.1368,
732
+ "step": 99
733
+ },
734
+ {
735
+ "epoch": 0.12987012987012986,
736
+ "grad_norm": 19.333633422851562,
737
+ "learning_rate": 2.1428571428571428e-05,
738
+ "loss": 2.233,
739
+ "step": 100
740
+ },
741
+ {
742
+ "epoch": 0.13116883116883116,
743
+ "grad_norm": 20.839101791381836,
744
+ "learning_rate": 2.1645021645021645e-05,
745
+ "loss": 2.4691,
746
+ "step": 101
747
+ },
748
+ {
749
+ "epoch": 0.13246753246753246,
750
+ "grad_norm": 20.799968719482422,
751
+ "learning_rate": 2.1861471861471862e-05,
752
+ "loss": 2.3472,
753
+ "step": 102
754
+ },
755
+ {
756
+ "epoch": 0.13376623376623376,
757
+ "grad_norm": 18.40387725830078,
758
+ "learning_rate": 2.207792207792208e-05,
759
+ "loss": 2.0578,
760
+ "step": 103
761
+ },
762
+ {
763
+ "epoch": 0.13506493506493505,
764
+ "grad_norm": 19.460140228271484,
765
+ "learning_rate": 2.2294372294372296e-05,
766
+ "loss": 2.3437,
767
+ "step": 104
768
+ },
769
+ {
770
+ "epoch": 0.13636363636363635,
771
+ "grad_norm": 20.304363250732422,
772
+ "learning_rate": 2.2510822510822512e-05,
773
+ "loss": 2.4822,
774
+ "step": 105
775
+ },
776
+ {
777
+ "epoch": 0.13766233766233765,
778
+ "grad_norm": 19.452438354492188,
779
+ "learning_rate": 2.272727272727273e-05,
780
+ "loss": 2.2261,
781
+ "step": 106
782
+ },
783
+ {
784
+ "epoch": 0.13896103896103895,
785
+ "grad_norm": 18.387685775756836,
786
+ "learning_rate": 2.2943722943722946e-05,
787
+ "loss": 2.3113,
788
+ "step": 107
789
+ },
790
+ {
791
+ "epoch": 0.14025974025974025,
792
+ "grad_norm": 19.084980010986328,
793
+ "learning_rate": 2.3160173160173163e-05,
794
+ "loss": 2.2162,
795
+ "step": 108
796
+ },
797
+ {
798
+ "epoch": 0.14155844155844155,
799
+ "grad_norm": 18.559663772583008,
800
+ "learning_rate": 2.3376623376623376e-05,
801
+ "loss": 2.0638,
802
+ "step": 109
803
+ },
804
+ {
805
+ "epoch": 0.14285714285714285,
806
+ "grad_norm": 17.65582847595215,
807
+ "learning_rate": 2.3593073593073593e-05,
808
+ "loss": 1.9822,
809
+ "step": 110
810
+ },
811
+ {
812
+ "epoch": 0.14415584415584415,
813
+ "grad_norm": 17.26728057861328,
814
+ "learning_rate": 2.380952380952381e-05,
815
+ "loss": 1.9165,
816
+ "step": 111
817
+ },
818
+ {
819
+ "epoch": 0.14545454545454545,
820
+ "grad_norm": 17.52690887451172,
821
+ "learning_rate": 2.4025974025974027e-05,
822
+ "loss": 2.0633,
823
+ "step": 112
824
+ },
825
+ {
826
+ "epoch": 0.14675324675324675,
827
+ "grad_norm": 17.66120147705078,
828
+ "learning_rate": 2.4242424242424244e-05,
829
+ "loss": 1.9012,
830
+ "step": 113
831
+ },
832
+ {
833
+ "epoch": 0.14805194805194805,
834
+ "grad_norm": 17.794565200805664,
835
+ "learning_rate": 2.4458874458874457e-05,
836
+ "loss": 1.8559,
837
+ "step": 114
838
+ },
839
+ {
840
+ "epoch": 0.14935064935064934,
841
+ "grad_norm": 18.78278923034668,
842
+ "learning_rate": 2.4675324675324678e-05,
843
+ "loss": 2.1804,
844
+ "step": 115
845
+ },
846
+ {
847
+ "epoch": 0.15064935064935064,
848
+ "grad_norm": 17.93096351623535,
849
+ "learning_rate": 2.4891774891774894e-05,
850
+ "loss": 1.9728,
851
+ "step": 116
852
+ },
853
+ {
854
+ "epoch": 0.15194805194805194,
855
+ "grad_norm": 16.865169525146484,
856
+ "learning_rate": 2.5108225108225108e-05,
857
+ "loss": 1.795,
858
+ "step": 117
859
+ },
860
+ {
861
+ "epoch": 0.15324675324675324,
862
+ "grad_norm": 16.265134811401367,
863
+ "learning_rate": 2.5324675324675325e-05,
864
+ "loss": 1.5135,
865
+ "step": 118
866
+ },
867
+ {
868
+ "epoch": 0.15454545454545454,
869
+ "grad_norm": 18.573184967041016,
870
+ "learning_rate": 2.5541125541125545e-05,
871
+ "loss": 1.8776,
872
+ "step": 119
873
+ },
874
+ {
875
+ "epoch": 0.15584415584415584,
876
+ "grad_norm": 18.620136260986328,
877
+ "learning_rate": 2.575757575757576e-05,
878
+ "loss": 1.9858,
879
+ "step": 120
880
+ },
881
+ {
882
+ "epoch": 0.15714285714285714,
883
+ "grad_norm": 17.044523239135742,
884
+ "learning_rate": 2.5974025974025972e-05,
885
+ "loss": 1.6143,
886
+ "step": 121
887
+ },
888
+ {
889
+ "epoch": 0.15844155844155844,
890
+ "grad_norm": 17.63654136657715,
891
+ "learning_rate": 2.6190476190476192e-05,
892
+ "loss": 1.7321,
893
+ "step": 122
894
+ },
895
+ {
896
+ "epoch": 0.15974025974025974,
897
+ "grad_norm": 17.7145938873291,
898
+ "learning_rate": 2.640692640692641e-05,
899
+ "loss": 1.7272,
900
+ "step": 123
901
+ },
902
+ {
903
+ "epoch": 0.16103896103896104,
904
+ "grad_norm": 16.036033630371094,
905
+ "learning_rate": 2.6623376623376623e-05,
906
+ "loss": 1.5508,
907
+ "step": 124
908
+ },
909
+ {
910
+ "epoch": 0.16233766233766234,
911
+ "grad_norm": 16.252811431884766,
912
+ "learning_rate": 2.6839826839826843e-05,
913
+ "loss": 1.2961,
914
+ "step": 125
915
+ },
916
+ {
917
+ "epoch": 0.16363636363636364,
918
+ "grad_norm": 17.2477970123291,
919
+ "learning_rate": 2.7056277056277056e-05,
920
+ "loss": 1.6425,
921
+ "step": 126
922
+ },
923
+ {
924
+ "epoch": 0.16493506493506493,
925
+ "grad_norm": 16.966596603393555,
926
+ "learning_rate": 2.7272727272727273e-05,
927
+ "loss": 1.5193,
928
+ "step": 127
929
+ },
930
+ {
931
+ "epoch": 0.16623376623376623,
932
+ "grad_norm": 16.886764526367188,
933
+ "learning_rate": 2.7489177489177493e-05,
934
+ "loss": 1.6626,
935
+ "step": 128
936
+ },
937
+ {
938
+ "epoch": 0.16753246753246753,
939
+ "grad_norm": 18.556442260742188,
940
+ "learning_rate": 2.7705627705627707e-05,
941
+ "loss": 2.0871,
942
+ "step": 129
943
+ },
944
+ {
945
+ "epoch": 0.16883116883116883,
946
+ "grad_norm": 15.574275970458984,
947
+ "learning_rate": 2.792207792207792e-05,
948
+ "loss": 1.5114,
949
+ "step": 130
950
+ },
951
+ {
952
+ "epoch": 0.17012987012987013,
953
+ "grad_norm": 16.09569549560547,
954
+ "learning_rate": 2.813852813852814e-05,
955
+ "loss": 1.4414,
956
+ "step": 131
957
+ },
958
+ {
959
+ "epoch": 0.17142857142857143,
960
+ "grad_norm": 16.894790649414062,
961
+ "learning_rate": 2.8354978354978357e-05,
962
+ "loss": 1.5323,
963
+ "step": 132
964
+ },
965
+ {
966
+ "epoch": 0.17272727272727273,
967
+ "grad_norm": 14.77379035949707,
968
+ "learning_rate": 2.857142857142857e-05,
969
+ "loss": 1.2476,
970
+ "step": 133
971
+ },
972
+ {
973
+ "epoch": 0.17402597402597403,
974
+ "grad_norm": 16.156719207763672,
975
+ "learning_rate": 2.878787878787879e-05,
976
+ "loss": 1.4671,
977
+ "step": 134
978
+ },
979
+ {
980
+ "epoch": 0.17532467532467533,
981
+ "grad_norm": 16.27682113647461,
982
+ "learning_rate": 2.9004329004329005e-05,
983
+ "loss": 1.5581,
984
+ "step": 135
985
+ },
986
+ {
987
+ "epoch": 0.17662337662337663,
988
+ "grad_norm": 16.400676727294922,
989
+ "learning_rate": 2.922077922077922e-05,
990
+ "loss": 1.5237,
991
+ "step": 136
992
+ },
993
+ {
994
+ "epoch": 0.17792207792207793,
995
+ "grad_norm": 15.038251876831055,
996
+ "learning_rate": 2.943722943722944e-05,
997
+ "loss": 1.2613,
998
+ "step": 137
999
+ },
1000
+ {
1001
+ "epoch": 0.17922077922077922,
1002
+ "grad_norm": 15.061074256896973,
1003
+ "learning_rate": 2.9653679653679655e-05,
1004
+ "loss": 1.2949,
1005
+ "step": 138
1006
+ },
1007
+ {
1008
+ "epoch": 0.18051948051948052,
1009
+ "grad_norm": 15.902937889099121,
1010
+ "learning_rate": 2.9870129870129872e-05,
1011
+ "loss": 1.3592,
1012
+ "step": 139
1013
+ },
1014
+ {
1015
+ "epoch": 0.18181818181818182,
1016
+ "grad_norm": 15.730782508850098,
1017
+ "learning_rate": 3.0086580086580092e-05,
1018
+ "loss": 1.3918,
1019
+ "step": 140
1020
+ },
1021
+ {
1022
+ "epoch": 0.18311688311688312,
1023
+ "grad_norm": 13.916067123413086,
1024
+ "learning_rate": 3.0303030303030306e-05,
1025
+ "loss": 1.0371,
1026
+ "step": 141
1027
+ },
1028
+ {
1029
+ "epoch": 0.18441558441558442,
1030
+ "grad_norm": 16.503026962280273,
1031
+ "learning_rate": 3.051948051948052e-05,
1032
+ "loss": 1.5028,
1033
+ "step": 142
1034
+ },
1035
+ {
1036
+ "epoch": 0.18571428571428572,
1037
+ "grad_norm": 14.891773223876953,
1038
+ "learning_rate": 3.073593073593073e-05,
1039
+ "loss": 1.2347,
1040
+ "step": 143
1041
+ },
1042
+ {
1043
+ "epoch": 0.18701298701298702,
1044
+ "grad_norm": 14.444226264953613,
1045
+ "learning_rate": 3.095238095238095e-05,
1046
+ "loss": 1.2731,
1047
+ "step": 144
1048
+ },
1049
+ {
1050
+ "epoch": 0.18831168831168832,
1051
+ "grad_norm": 14.655235290527344,
1052
+ "learning_rate": 3.1168831168831166e-05,
1053
+ "loss": 1.2688,
1054
+ "step": 145
1055
+ },
1056
+ {
1057
+ "epoch": 0.18961038961038962,
1058
+ "grad_norm": 15.224197387695312,
1059
+ "learning_rate": 3.1385281385281387e-05,
1060
+ "loss": 1.3688,
1061
+ "step": 146
1062
+ },
1063
+ {
1064
+ "epoch": 0.19090909090909092,
1065
+ "grad_norm": 17.209598541259766,
1066
+ "learning_rate": 3.160173160173161e-05,
1067
+ "loss": 1.8782,
1068
+ "step": 147
1069
+ },
1070
+ {
1071
+ "epoch": 0.19220779220779222,
1072
+ "grad_norm": 12.981212615966797,
1073
+ "learning_rate": 3.181818181818182e-05,
1074
+ "loss": 1.0597,
1075
+ "step": 148
1076
+ },
1077
+ {
1078
+ "epoch": 0.19350649350649352,
1079
+ "grad_norm": 14.368124961853027,
1080
+ "learning_rate": 3.2034632034632034e-05,
1081
+ "loss": 1.229,
1082
+ "step": 149
1083
+ },
1084
+ {
1085
+ "epoch": 0.19480519480519481,
1086
+ "grad_norm": 13.885086059570312,
1087
+ "learning_rate": 3.2251082251082254e-05,
1088
+ "loss": 1.1233,
1089
+ "step": 150
1090
+ },
1091
+ {
1092
+ "epoch": 0.1961038961038961,
1093
+ "grad_norm": 14.75069522857666,
1094
+ "learning_rate": 3.246753246753247e-05,
1095
+ "loss": 1.2579,
1096
+ "step": 151
1097
+ },
1098
+ {
1099
+ "epoch": 0.1974025974025974,
1100
+ "grad_norm": 14.663084030151367,
1101
+ "learning_rate": 3.268398268398268e-05,
1102
+ "loss": 1.1547,
1103
+ "step": 152
1104
+ },
1105
+ {
1106
+ "epoch": 0.1987012987012987,
1107
+ "grad_norm": 16.447593688964844,
1108
+ "learning_rate": 3.29004329004329e-05,
1109
+ "loss": 1.3986,
1110
+ "step": 153
1111
+ },
1112
+ {
1113
+ "epoch": 0.2,
1114
+ "grad_norm": 16.009765625,
1115
+ "learning_rate": 3.311688311688312e-05,
1116
+ "loss": 1.3071,
1117
+ "step": 154
1118
+ },
1119
+ {
1120
+ "epoch": 0.2,
1121
+ "eval_allNLI--triplets-1024_cosine_accuracy": 0.9375,
1122
+ "eval_allNLI--triplets-128_cosine_accuracy": 0.90625,
1123
+ "eval_allNLI--triplets-256_cosine_accuracy": 0.90625,
1124
+ "eval_allNLI--triplets-32_cosine_accuracy": 0.9140625,
1125
+ "eval_allNLI--triplets-512_cosine_accuracy": 0.921875,
1126
+ "eval_allNLI-triplets_cosine_accuracy": 0.9453125,
1127
+ "eval_global_dataset_loss": 0.8410314917564392,
1128
+ "eval_global_dataset_runtime": 113.466,
1129
+ "eval_global_dataset_samples_per_second": 9.439,
1130
+ "eval_global_dataset_steps_per_second": 0.079,
1131
+ "eval_sequential_score": 0.9140625,
1132
+ "eval_sts-test-1024_pearson_cosine": 0.8637686212043777,
1133
+ "eval_sts-test-1024_spearman_cosine": 0.9158389959804479,
1134
+ "eval_sts-test-128_pearson_cosine": 0.8587735042324499,
1135
+ "eval_sts-test-128_spearman_cosine": 0.9079734327816954,
1136
+ "eval_sts-test-256_pearson_cosine": 0.8614854438059253,
1137
+ "eval_sts-test-256_spearman_cosine": 0.9113397849846575,
1138
+ "eval_sts-test-32_pearson_cosine": 0.8204607604412608,
1139
+ "eval_sts-test-32_spearman_cosine": 0.8866968838230891,
1140
+ "eval_sts-test-512_pearson_cosine": 0.8636960327159487,
1141
+ "eval_sts-test-512_spearman_cosine": 0.9151565595521514,
1142
+ "eval_sts-test-64_pearson_cosine": 0.8415761749347184,
1143
+ "eval_sts-test-64_spearman_cosine": 0.9004612613745113,
1144
+ "eval_sts-test_pearson_cosine": 0.8701225080029533,
1145
+ "eval_sts-test_spearman_cosine": 0.9182271654330917,
1146
+ "step": 154
1147
+ },
1148
+ {
1149
+ "epoch": 0.2012987012987013,
1150
+ "grad_norm": 16.777559280395508,
1151
+ "learning_rate": 3.3333333333333335e-05,
1152
+ "loss": 1.5306,
1153
+ "step": 155
1154
+ },
1155
+ {
1156
+ "epoch": 0.2025974025974026,
1157
+ "grad_norm": 14.096006393432617,
1158
+ "learning_rate": 3.3549783549783555e-05,
1159
+ "loss": 1.0775,
1160
+ "step": 156
1161
+ },
1162
+ {
1163
+ "epoch": 0.2038961038961039,
1164
+ "grad_norm": 15.965473175048828,
1165
+ "learning_rate": 3.376623376623377e-05,
1166
+ "loss": 1.2954,
1167
+ "step": 157
1168
+ },
1169
+ {
1170
+ "epoch": 0.2051948051948052,
1171
+ "grad_norm": 13.835360527038574,
1172
+ "learning_rate": 3.398268398268398e-05,
1173
+ "loss": 1.1233,
1174
+ "step": 158
1175
+ },
1176
+ {
1177
+ "epoch": 0.2064935064935065,
1178
+ "grad_norm": 14.972121238708496,
1179
+ "learning_rate": 3.41991341991342e-05,
1180
+ "loss": 1.2492,
1181
+ "step": 159
1182
+ },
1183
+ {
1184
+ "epoch": 0.2077922077922078,
1185
+ "grad_norm": 14.487752914428711,
1186
+ "learning_rate": 3.4415584415584416e-05,
1187
+ "loss": 1.2897,
1188
+ "step": 160
1189
+ },
1190
+ {
1191
+ "epoch": 0.20909090909090908,
1192
+ "grad_norm": 13.836915016174316,
1193
+ "learning_rate": 3.463203463203463e-05,
1194
+ "loss": 1.0752,
1195
+ "step": 161
1196
+ },
1197
+ {
1198
+ "epoch": 0.21038961038961038,
1199
+ "grad_norm": 15.030878067016602,
1200
+ "learning_rate": 3.484848484848485e-05,
1201
+ "loss": 1.269,
1202
+ "step": 162
1203
+ },
1204
+ {
1205
+ "epoch": 0.21168831168831168,
1206
+ "grad_norm": 14.309725761413574,
1207
+ "learning_rate": 3.506493506493507e-05,
1208
+ "loss": 1.1273,
1209
+ "step": 163
1210
+ },
1211
+ {
1212
+ "epoch": 0.21298701298701297,
1213
+ "grad_norm": 14.790210723876953,
1214
+ "learning_rate": 3.528138528138528e-05,
1215
+ "loss": 1.2208,
1216
+ "step": 164
1217
+ },
1218
+ {
1219
+ "epoch": 0.21428571428571427,
1220
+ "grad_norm": 14.586909294128418,
1221
+ "learning_rate": 3.5497835497835503e-05,
1222
+ "loss": 1.2356,
1223
+ "step": 165
1224
+ },
1225
+ {
1226
+ "epoch": 0.21558441558441557,
1227
+ "grad_norm": 13.816767692565918,
1228
+ "learning_rate": 3.571428571428572e-05,
1229
+ "loss": 1.1039,
1230
+ "step": 166
1231
+ },
1232
+ {
1233
+ "epoch": 0.21688311688311687,
1234
+ "grad_norm": 13.635117530822754,
1235
+ "learning_rate": 3.593073593073593e-05,
1236
+ "loss": 1.0146,
1237
+ "step": 167
1238
+ },
1239
+ {
1240
+ "epoch": 0.21818181818181817,
1241
+ "grad_norm": 14.119977951049805,
1242
+ "learning_rate": 3.6147186147186144e-05,
1243
+ "loss": 1.3097,
1244
+ "step": 168
1245
+ },
1246
+ {
1247
+ "epoch": 0.21948051948051947,
1248
+ "grad_norm": 14.68720531463623,
1249
+ "learning_rate": 3.6363636363636364e-05,
1250
+ "loss": 1.2947,
1251
+ "step": 169
1252
+ },
1253
+ {
1254
+ "epoch": 0.22077922077922077,
1255
+ "grad_norm": 14.55096435546875,
1256
+ "learning_rate": 3.6580086580086584e-05,
1257
+ "loss": 1.291,
1258
+ "step": 170
1259
+ },
1260
+ {
1261
+ "epoch": 0.22207792207792207,
1262
+ "grad_norm": 13.37848949432373,
1263
+ "learning_rate": 3.67965367965368e-05,
1264
+ "loss": 1.0295,
1265
+ "step": 171
1266
+ },
1267
+ {
1268
+ "epoch": 0.22337662337662337,
1269
+ "grad_norm": 12.699470520019531,
1270
+ "learning_rate": 3.701298701298702e-05,
1271
+ "loss": 1.0948,
1272
+ "step": 172
1273
+ },
1274
+ {
1275
+ "epoch": 0.22467532467532467,
1276
+ "grad_norm": 13.583250999450684,
1277
+ "learning_rate": 3.722943722943723e-05,
1278
+ "loss": 1.1393,
1279
+ "step": 173
1280
+ },
1281
+ {
1282
+ "epoch": 0.22597402597402597,
1283
+ "grad_norm": 12.980260848999023,
1284
+ "learning_rate": 3.7445887445887445e-05,
1285
+ "loss": 1.0063,
1286
+ "step": 174
1287
+ },
1288
+ {
1289
+ "epoch": 0.22727272727272727,
1290
+ "grad_norm": 14.338798522949219,
1291
+ "learning_rate": 3.7662337662337665e-05,
1292
+ "loss": 1.1769,
1293
+ "step": 175
1294
+ },
1295
+ {
1296
+ "epoch": 0.22857142857142856,
1297
+ "grad_norm": 14.311129570007324,
1298
+ "learning_rate": 3.787878787878788e-05,
1299
+ "loss": 1.1731,
1300
+ "step": 176
1301
+ },
1302
+ {
1303
+ "epoch": 0.22987012987012986,
1304
+ "grad_norm": 12.44087028503418,
1305
+ "learning_rate": 3.809523809523809e-05,
1306
+ "loss": 0.7476,
1307
+ "step": 177
1308
+ },
1309
+ {
1310
+ "epoch": 0.23116883116883116,
1311
+ "grad_norm": 14.917692184448242,
1312
+ "learning_rate": 3.831168831168831e-05,
1313
+ "loss": 1.022,
1314
+ "step": 178
1315
+ },
1316
+ {
1317
+ "epoch": 0.23246753246753246,
1318
+ "grad_norm": 13.485183715820312,
1319
+ "learning_rate": 3.852813852813853e-05,
1320
+ "loss": 0.9579,
1321
+ "step": 179
1322
+ },
1323
+ {
1324
+ "epoch": 0.23376623376623376,
1325
+ "grad_norm": 13.994976997375488,
1326
+ "learning_rate": 3.8744588744588746e-05,
1327
+ "loss": 1.0753,
1328
+ "step": 180
1329
+ },
1330
+ {
1331
+ "epoch": 0.23506493506493506,
1332
+ "grad_norm": 15.207565307617188,
1333
+ "learning_rate": 3.8961038961038966e-05,
1334
+ "loss": 1.2243,
1335
+ "step": 181
1336
+ },
1337
+ {
1338
+ "epoch": 0.23636363636363636,
1339
+ "grad_norm": 15.602945327758789,
1340
+ "learning_rate": 3.917748917748918e-05,
1341
+ "loss": 1.2154,
1342
+ "step": 182
1343
+ },
1344
+ {
1345
+ "epoch": 0.23766233766233766,
1346
+ "grad_norm": 12.87066650390625,
1347
+ "learning_rate": 3.939393939393939e-05,
1348
+ "loss": 0.8147,
1349
+ "step": 183
1350
+ },
1351
+ {
1352
+ "epoch": 0.23896103896103896,
1353
+ "grad_norm": 14.42218017578125,
1354
+ "learning_rate": 3.9610389610389614e-05,
1355
+ "loss": 1.1086,
1356
+ "step": 184
1357
+ },
1358
+ {
1359
+ "epoch": 0.24025974025974026,
1360
+ "grad_norm": 14.212447166442871,
1361
+ "learning_rate": 3.982683982683983e-05,
1362
+ "loss": 1.0155,
1363
+ "step": 185
1364
+ },
1365
+ {
1366
+ "epoch": 0.24155844155844156,
1367
+ "grad_norm": 15.133934020996094,
1368
+ "learning_rate": 4.004329004329004e-05,
1369
+ "loss": 1.1898,
1370
+ "step": 186
1371
+ },
1372
+ {
1373
+ "epoch": 0.24285714285714285,
1374
+ "grad_norm": 13.727578163146973,
1375
+ "learning_rate": 4.025974025974026e-05,
1376
+ "loss": 1.11,
1377
+ "step": 187
1378
+ },
1379
+ {
1380
+ "epoch": 0.24415584415584415,
1381
+ "grad_norm": 15.756597518920898,
1382
+ "learning_rate": 4.047619047619048e-05,
1383
+ "loss": 1.3128,
1384
+ "step": 188
1385
+ },
1386
+ {
1387
+ "epoch": 0.24545454545454545,
1388
+ "grad_norm": 13.591508865356445,
1389
+ "learning_rate": 4.0692640692640695e-05,
1390
+ "loss": 1.0642,
1391
+ "step": 189
1392
+ },
1393
+ {
1394
+ "epoch": 0.24675324675324675,
1395
+ "grad_norm": 13.608804702758789,
1396
+ "learning_rate": 4.0909090909090915e-05,
1397
+ "loss": 0.8932,
1398
+ "step": 190
1399
+ },
1400
+ {
1401
+ "epoch": 0.24805194805194805,
1402
+ "grad_norm": 13.768484115600586,
1403
+ "learning_rate": 4.112554112554113e-05,
1404
+ "loss": 1.1683,
1405
+ "step": 191
1406
+ },
1407
+ {
1408
+ "epoch": 0.24935064935064935,
1409
+ "grad_norm": 13.267477989196777,
1410
+ "learning_rate": 4.134199134199134e-05,
1411
+ "loss": 1.0554,
1412
+ "step": 192
1413
+ },
1414
+ {
1415
+ "epoch": 0.2506493506493506,
1416
+ "grad_norm": 14.142014503479004,
1417
+ "learning_rate": 4.155844155844156e-05,
1418
+ "loss": 1.2186,
1419
+ "step": 193
1420
+ },
1421
+ {
1422
+ "epoch": 0.2519480519480519,
1423
+ "grad_norm": 13.442983627319336,
1424
+ "learning_rate": 4.1774891774891775e-05,
1425
+ "loss": 1.027,
1426
+ "step": 194
1427
+ },
1428
+ {
1429
+ "epoch": 0.2532467532467532,
1430
+ "grad_norm": 12.284928321838379,
1431
+ "learning_rate": 4.1991341991341996e-05,
1432
+ "loss": 0.799,
1433
+ "step": 195
1434
+ },
1435
+ {
1436
+ "epoch": 0.2545454545454545,
1437
+ "grad_norm": 14.540982246398926,
1438
+ "learning_rate": 4.220779220779221e-05,
1439
+ "loss": 1.099,
1440
+ "step": 196
1441
+ },
1442
+ {
1443
+ "epoch": 0.2558441558441558,
1444
+ "grad_norm": 12.08164119720459,
1445
+ "learning_rate": 4.242424242424243e-05,
1446
+ "loss": 0.7717,
1447
+ "step": 197
1448
+ },
1449
+ {
1450
+ "epoch": 0.2571428571428571,
1451
+ "grad_norm": 14.687990188598633,
1452
+ "learning_rate": 4.264069264069264e-05,
1453
+ "loss": 1.1011,
1454
+ "step": 198
1455
+ },
1456
+ {
1457
+ "epoch": 0.2584415584415584,
1458
+ "grad_norm": 13.88438606262207,
1459
+ "learning_rate": 4.2857142857142856e-05,
1460
+ "loss": 1.0083,
1461
+ "step": 199
1462
+ },
1463
+ {
1464
+ "epoch": 0.2597402597402597,
1465
+ "grad_norm": 13.972654342651367,
1466
+ "learning_rate": 4.3073593073593077e-05,
1467
+ "loss": 1.1488,
1468
+ "step": 200
1469
+ },
1470
+ {
1471
+ "epoch": 0.261038961038961,
1472
+ "grad_norm": 14.187093734741211,
1473
+ "learning_rate": 4.329004329004329e-05,
1474
+ "loss": 1.0453,
1475
+ "step": 201
1476
+ },
1477
+ {
1478
+ "epoch": 0.2623376623376623,
1479
+ "grad_norm": 14.916047096252441,
1480
+ "learning_rate": 4.3506493506493503e-05,
1481
+ "loss": 1.2942,
1482
+ "step": 202
1483
+ },
1484
+ {
1485
+ "epoch": 0.2636363636363636,
1486
+ "grad_norm": 13.239340782165527,
1487
+ "learning_rate": 4.3722943722943724e-05,
1488
+ "loss": 1.0279,
1489
+ "step": 203
1490
+ },
1491
+ {
1492
+ "epoch": 0.2649350649350649,
1493
+ "grad_norm": 12.49507999420166,
1494
+ "learning_rate": 4.3939393939393944e-05,
1495
+ "loss": 0.9493,
1496
+ "step": 204
1497
+ },
1498
+ {
1499
+ "epoch": 0.2662337662337662,
1500
+ "grad_norm": 14.923321723937988,
1501
+ "learning_rate": 4.415584415584416e-05,
1502
+ "loss": 1.2698,
1503
+ "step": 205
1504
+ },
1505
+ {
1506
+ "epoch": 0.2675324675324675,
1507
+ "grad_norm": 13.260164260864258,
1508
+ "learning_rate": 4.437229437229438e-05,
1509
+ "loss": 1.066,
1510
+ "step": 206
1511
+ },
1512
+ {
1513
+ "epoch": 0.2688311688311688,
1514
+ "grad_norm": 13.298815727233887,
1515
+ "learning_rate": 4.458874458874459e-05,
1516
+ "loss": 1.0352,
1517
+ "step": 207
1518
+ },
1519
+ {
1520
+ "epoch": 0.2701298701298701,
1521
+ "grad_norm": 12.438820838928223,
1522
+ "learning_rate": 4.4805194805194805e-05,
1523
+ "loss": 0.8514,
1524
+ "step": 208
1525
+ },
1526
+ {
1527
+ "epoch": 0.2714285714285714,
1528
+ "grad_norm": 13.921923637390137,
1529
+ "learning_rate": 4.5021645021645025e-05,
1530
+ "loss": 1.094,
1531
+ "step": 209
1532
+ },
1533
+ {
1534
+ "epoch": 0.2727272727272727,
1535
+ "grad_norm": 13.15014934539795,
1536
+ "learning_rate": 4.523809523809524e-05,
1537
+ "loss": 1.0087,
1538
+ "step": 210
1539
+ },
1540
+ {
1541
+ "epoch": 0.274025974025974,
1542
+ "grad_norm": 12.404814720153809,
1543
+ "learning_rate": 4.545454545454546e-05,
1544
+ "loss": 0.9035,
1545
+ "step": 211
1546
+ },
1547
+ {
1548
+ "epoch": 0.2753246753246753,
1549
+ "grad_norm": 13.696551322937012,
1550
+ "learning_rate": 4.567099567099568e-05,
1551
+ "loss": 0.9475,
1552
+ "step": 212
1553
+ },
1554
+ {
1555
+ "epoch": 0.2766233766233766,
1556
+ "grad_norm": 13.52684211730957,
1557
+ "learning_rate": 4.588744588744589e-05,
1558
+ "loss": 0.9838,
1559
+ "step": 213
1560
+ },
1561
+ {
1562
+ "epoch": 0.2779220779220779,
1563
+ "grad_norm": 14.523106575012207,
1564
+ "learning_rate": 4.6103896103896106e-05,
1565
+ "loss": 1.1285,
1566
+ "step": 214
1567
+ },
1568
+ {
1569
+ "epoch": 0.2792207792207792,
1570
+ "grad_norm": 12.710862159729004,
1571
+ "learning_rate": 4.6320346320346326e-05,
1572
+ "loss": 0.7406,
1573
+ "step": 215
1574
+ },
1575
+ {
1576
+ "epoch": 0.2805194805194805,
1577
+ "grad_norm": 13.21323299407959,
1578
+ "learning_rate": 4.653679653679654e-05,
1579
+ "loss": 0.9309,
1580
+ "step": 216
1581
+ },
1582
+ {
1583
+ "epoch": 0.2818181818181818,
1584
+ "grad_norm": 13.333338737487793,
1585
+ "learning_rate": 4.675324675324675e-05,
1586
+ "loss": 0.8669,
1587
+ "step": 217
1588
+ },
1589
+ {
1590
+ "epoch": 0.2831168831168831,
1591
+ "grad_norm": 14.696738243103027,
1592
+ "learning_rate": 4.696969696969697e-05,
1593
+ "loss": 1.0496,
1594
+ "step": 218
1595
+ },
1596
+ {
1597
+ "epoch": 0.2844155844155844,
1598
+ "grad_norm": 14.372692108154297,
1599
+ "learning_rate": 4.718614718614719e-05,
1600
+ "loss": 0.9547,
1601
+ "step": 219
1602
+ },
1603
+ {
1604
+ "epoch": 0.2857142857142857,
1605
+ "grad_norm": 11.897599220275879,
1606
+ "learning_rate": 4.740259740259741e-05,
1607
+ "loss": 0.7442,
1608
+ "step": 220
1609
+ },
1610
+ {
1611
+ "epoch": 0.287012987012987,
1612
+ "grad_norm": 12.831521034240723,
1613
+ "learning_rate": 4.761904761904762e-05,
1614
+ "loss": 0.6949,
1615
+ "step": 221
1616
+ },
1617
+ {
1618
+ "epoch": 0.2883116883116883,
1619
+ "grad_norm": 12.80624008178711,
1620
+ "learning_rate": 4.783549783549784e-05,
1621
+ "loss": 0.958,
1622
+ "step": 222
1623
+ },
1624
+ {
1625
+ "epoch": 0.2896103896103896,
1626
+ "grad_norm": 11.822639465332031,
1627
+ "learning_rate": 4.8051948051948054e-05,
1628
+ "loss": 0.7528,
1629
+ "step": 223
1630
+ },
1631
+ {
1632
+ "epoch": 0.2909090909090909,
1633
+ "grad_norm": 12.583551406860352,
1634
+ "learning_rate": 4.826839826839827e-05,
1635
+ "loss": 0.8338,
1636
+ "step": 224
1637
+ },
1638
+ {
1639
+ "epoch": 0.2922077922077922,
1640
+ "grad_norm": 14.761630058288574,
1641
+ "learning_rate": 4.848484848484849e-05,
1642
+ "loss": 1.108,
1643
+ "step": 225
1644
+ },
1645
+ {
1646
+ "epoch": 0.2935064935064935,
1647
+ "grad_norm": 14.278302192687988,
1648
+ "learning_rate": 4.87012987012987e-05,
1649
+ "loss": 1.1113,
1650
+ "step": 226
1651
+ },
1652
+ {
1653
+ "epoch": 0.2948051948051948,
1654
+ "grad_norm": 12.455058097839355,
1655
+ "learning_rate": 4.8917748917748915e-05,
1656
+ "loss": 0.8191,
1657
+ "step": 227
1658
+ },
1659
+ {
1660
+ "epoch": 0.2961038961038961,
1661
+ "grad_norm": 12.987092018127441,
1662
+ "learning_rate": 4.9134199134199135e-05,
1663
+ "loss": 0.8167,
1664
+ "step": 228
1665
+ },
1666
+ {
1667
+ "epoch": 0.2974025974025974,
1668
+ "grad_norm": 11.661968231201172,
1669
+ "learning_rate": 4.9350649350649355e-05,
1670
+ "loss": 0.8165,
1671
+ "step": 229
1672
+ },
1673
+ {
1674
+ "epoch": 0.2987012987012987,
1675
+ "grad_norm": 12.673922538757324,
1676
+ "learning_rate": 4.956709956709957e-05,
1677
+ "loss": 0.8569,
1678
+ "step": 230
1679
+ },
1680
+ {
1681
+ "epoch": 0.3,
1682
+ "grad_norm": 13.494362831115723,
1683
+ "learning_rate": 4.978354978354979e-05,
1684
+ "loss": 0.943,
1685
+ "step": 231
1686
+ },
1687
+ {
1688
+ "epoch": 0.3,
1689
+ "eval_allNLI--triplets-1024_cosine_accuracy": 0.953125,
1690
+ "eval_allNLI--triplets-128_cosine_accuracy": 0.8984375,
1691
+ "eval_allNLI--triplets-256_cosine_accuracy": 0.921875,
1692
+ "eval_allNLI--triplets-32_cosine_accuracy": 0.9140625,
1693
+ "eval_allNLI--triplets-512_cosine_accuracy": 0.9375,
1694
+ "eval_allNLI-triplets_cosine_accuracy": 0.9453125,
1695
+ "eval_global_dataset_loss": 0.677043080329895,
1696
+ "eval_global_dataset_runtime": 113.6829,
1697
+ "eval_global_dataset_samples_per_second": 9.421,
1698
+ "eval_global_dataset_steps_per_second": 0.079,
1699
+ "eval_sequential_score": 0.9140625,
1700
+ "eval_sts-test-1024_pearson_cosine": 0.8787763074934598,
1701
+ "eval_sts-test-1024_spearman_cosine": 0.9210331750283494,
1702
+ "eval_sts-test-128_pearson_cosine": 0.8722131705651872,
1703
+ "eval_sts-test-128_spearman_cosine": 0.9158898385684954,
1704
+ "eval_sts-test-256_pearson_cosine": 0.8732650189534636,
1705
+ "eval_sts-test-256_spearman_cosine": 0.9174469823390422,
1706
+ "eval_sts-test-32_pearson_cosine": 0.8348374873613509,
1707
+ "eval_sts-test-32_spearman_cosine": 0.8928345151210231,
1708
+ "eval_sts-test-512_pearson_cosine": 0.8784119512407846,
1709
+ "eval_sts-test-512_spearman_cosine": 0.9211549108025473,
1710
+ "eval_sts-test-64_pearson_cosine": 0.8606305478646585,
1711
+ "eval_sts-test-64_spearman_cosine": 0.9075380483657413,
1712
+ "eval_sts-test_pearson_cosine": 0.8837500780432396,
1713
+ "eval_sts-test_spearman_cosine": 0.9228735334971034,
1714
+ "step": 231
1715
+ },
1716
+ {
1717
+ "epoch": 0.3012987012987013,
1718
+ "grad_norm": 12.865336418151855,
1719
+ "learning_rate": 5e-05,
1720
+ "loss": 0.8316,
1721
+ "step": 232
1722
+ },
1723
+ {
1724
+ "epoch": 0.3025974025974026,
1725
+ "grad_norm": 11.271172523498535,
1726
+ "learning_rate": 5.0216450216450216e-05,
1727
+ "loss": 0.6811,
1728
+ "step": 233
1729
+ },
1730
+ {
1731
+ "epoch": 0.3038961038961039,
1732
+ "grad_norm": 11.870718955993652,
1733
+ "learning_rate": 5.043290043290043e-05,
1734
+ "loss": 0.7579,
1735
+ "step": 234
1736
+ },
1737
+ {
1738
+ "epoch": 0.3051948051948052,
1739
+ "grad_norm": 12.892427444458008,
1740
+ "learning_rate": 5.064935064935065e-05,
1741
+ "loss": 0.8088,
1742
+ "step": 235
1743
+ },
1744
+ {
1745
+ "epoch": 0.3064935064935065,
1746
+ "grad_norm": 13.148665428161621,
1747
+ "learning_rate": 5.086580086580087e-05,
1748
+ "loss": 0.8242,
1749
+ "step": 236
1750
+ },
1751
+ {
1752
+ "epoch": 0.3077922077922078,
1753
+ "grad_norm": 13.16065788269043,
1754
+ "learning_rate": 5.108225108225109e-05,
1755
+ "loss": 0.9401,
1756
+ "step": 237
1757
+ },
1758
+ {
1759
+ "epoch": 0.3090909090909091,
1760
+ "grad_norm": 13.374222755432129,
1761
+ "learning_rate": 5.1298701298701304e-05,
1762
+ "loss": 0.9815,
1763
+ "step": 238
1764
+ },
1765
+ {
1766
+ "epoch": 0.3103896103896104,
1767
+ "grad_norm": 11.324379920959473,
1768
+ "learning_rate": 5.151515151515152e-05,
1769
+ "loss": 0.6338,
1770
+ "step": 239
1771
+ },
1772
+ {
1773
+ "epoch": 0.3116883116883117,
1774
+ "grad_norm": 14.477818489074707,
1775
+ "learning_rate": 5.173160173160173e-05,
1776
+ "loss": 1.1544,
1777
+ "step": 240
1778
+ },
1779
+ {
1780
+ "epoch": 0.312987012987013,
1781
+ "grad_norm": 13.306774139404297,
1782
+ "learning_rate": 5.1948051948051944e-05,
1783
+ "loss": 0.7693,
1784
+ "step": 241
1785
+ },
1786
+ {
1787
+ "epoch": 0.3142857142857143,
1788
+ "grad_norm": 12.13784408569336,
1789
+ "learning_rate": 5.216450216450217e-05,
1790
+ "loss": 0.745,
1791
+ "step": 242
1792
+ },
1793
+ {
1794
+ "epoch": 0.3155844155844156,
1795
+ "grad_norm": 12.215293884277344,
1796
+ "learning_rate": 5.2380952380952384e-05,
1797
+ "loss": 0.722,
1798
+ "step": 243
1799
+ },
1800
+ {
1801
+ "epoch": 0.3168831168831169,
1802
+ "grad_norm": 13.042692184448242,
1803
+ "learning_rate": 5.25974025974026e-05,
1804
+ "loss": 0.7834,
1805
+ "step": 244
1806
+ },
1807
+ {
1808
+ "epoch": 0.3181818181818182,
1809
+ "grad_norm": 12.925629615783691,
1810
+ "learning_rate": 5.281385281385282e-05,
1811
+ "loss": 0.8727,
1812
+ "step": 245
1813
+ },
1814
+ {
1815
+ "epoch": 0.3194805194805195,
1816
+ "grad_norm": 11.345806121826172,
1817
+ "learning_rate": 5.303030303030303e-05,
1818
+ "loss": 0.657,
1819
+ "step": 246
1820
+ },
1821
+ {
1822
+ "epoch": 0.3207792207792208,
1823
+ "grad_norm": 14.048829078674316,
1824
+ "learning_rate": 5.3246753246753245e-05,
1825
+ "loss": 0.9334,
1826
+ "step": 247
1827
+ },
1828
+ {
1829
+ "epoch": 0.3220779220779221,
1830
+ "grad_norm": 14.103421211242676,
1831
+ "learning_rate": 5.346320346320347e-05,
1832
+ "loss": 0.9641,
1833
+ "step": 248
1834
+ },
1835
+ {
1836
+ "epoch": 0.3233766233766234,
1837
+ "grad_norm": 13.120491027832031,
1838
+ "learning_rate": 5.3679653679653686e-05,
1839
+ "loss": 0.9539,
1840
+ "step": 249
1841
+ },
1842
+ {
1843
+ "epoch": 0.3246753246753247,
1844
+ "grad_norm": 12.34601879119873,
1845
+ "learning_rate": 5.38961038961039e-05,
1846
+ "loss": 0.916,
1847
+ "step": 250
1848
+ },
1849
+ {
1850
+ "epoch": 0.32597402597402597,
1851
+ "grad_norm": 11.876169204711914,
1852
+ "learning_rate": 5.411255411255411e-05,
1853
+ "loss": 0.7535,
1854
+ "step": 251
1855
+ },
1856
+ {
1857
+ "epoch": 0.32727272727272727,
1858
+ "grad_norm": 13.08942985534668,
1859
+ "learning_rate": 5.4329004329004326e-05,
1860
+ "loss": 1.0744,
1861
+ "step": 252
1862
+ },
1863
+ {
1864
+ "epoch": 0.32857142857142857,
1865
+ "grad_norm": 11.773645401000977,
1866
+ "learning_rate": 5.4545454545454546e-05,
1867
+ "loss": 0.7207,
1868
+ "step": 253
1869
+ },
1870
+ {
1871
+ "epoch": 0.32987012987012987,
1872
+ "grad_norm": 13.143170356750488,
1873
+ "learning_rate": 5.4761904761904766e-05,
1874
+ "loss": 0.9337,
1875
+ "step": 254
1876
+ },
1877
+ {
1878
+ "epoch": 0.33116883116883117,
1879
+ "grad_norm": 11.403676986694336,
1880
+ "learning_rate": 5.497835497835499e-05,
1881
+ "loss": 0.7114,
1882
+ "step": 255
1883
+ },
1884
+ {
1885
+ "epoch": 0.33246753246753247,
1886
+ "grad_norm": 12.294326782226562,
1887
+ "learning_rate": 5.51948051948052e-05,
1888
+ "loss": 0.6995,
1889
+ "step": 256
1890
+ },
1891
+ {
1892
+ "epoch": 0.33376623376623377,
1893
+ "grad_norm": 13.27484130859375,
1894
+ "learning_rate": 5.5411255411255414e-05,
1895
+ "loss": 0.8138,
1896
+ "step": 257
1897
+ },
1898
+ {
1899
+ "epoch": 0.33506493506493507,
1900
+ "grad_norm": 13.444418907165527,
1901
+ "learning_rate": 5.562770562770563e-05,
1902
+ "loss": 1.0225,
1903
+ "step": 258
1904
+ },
1905
+ {
1906
+ "epoch": 0.33636363636363636,
1907
+ "grad_norm": 13.552763938903809,
1908
+ "learning_rate": 5.584415584415584e-05,
1909
+ "loss": 0.9528,
1910
+ "step": 259
1911
+ },
1912
+ {
1913
+ "epoch": 0.33766233766233766,
1914
+ "grad_norm": 13.628582954406738,
1915
+ "learning_rate": 5.606060606060606e-05,
1916
+ "loss": 0.9607,
1917
+ "step": 260
1918
+ },
1919
+ {
1920
+ "epoch": 0.33896103896103896,
1921
+ "grad_norm": 13.577703475952148,
1922
+ "learning_rate": 5.627705627705628e-05,
1923
+ "loss": 0.971,
1924
+ "step": 261
1925
+ },
1926
+ {
1927
+ "epoch": 0.34025974025974026,
1928
+ "grad_norm": 11.462188720703125,
1929
+ "learning_rate": 5.64935064935065e-05,
1930
+ "loss": 0.7819,
1931
+ "step": 262
1932
+ },
1933
+ {
1934
+ "epoch": 0.34155844155844156,
1935
+ "grad_norm": 10.696789741516113,
1936
+ "learning_rate": 5.6709956709956715e-05,
1937
+ "loss": 0.6537,
1938
+ "step": 263
1939
+ },
1940
+ {
1941
+ "epoch": 0.34285714285714286,
1942
+ "grad_norm": 12.647172927856445,
1943
+ "learning_rate": 5.692640692640693e-05,
1944
+ "loss": 1.0185,
1945
+ "step": 264
1946
+ },
1947
+ {
1948
+ "epoch": 0.34415584415584416,
1949
+ "grad_norm": 13.042900085449219,
1950
+ "learning_rate": 5.714285714285714e-05,
1951
+ "loss": 1.0832,
1952
+ "step": 265
1953
+ },
1954
+ {
1955
+ "epoch": 0.34545454545454546,
1956
+ "grad_norm": 9.650392532348633,
1957
+ "learning_rate": 5.7359307359307355e-05,
1958
+ "loss": 0.533,
1959
+ "step": 266
1960
+ },
1961
+ {
1962
+ "epoch": 0.34675324675324676,
1963
+ "grad_norm": 12.420098304748535,
1964
+ "learning_rate": 5.757575757575758e-05,
1965
+ "loss": 0.9556,
1966
+ "step": 267
1967
+ },
1968
+ {
1969
+ "epoch": 0.34805194805194806,
1970
+ "grad_norm": 11.905637741088867,
1971
+ "learning_rate": 5.7792207792207796e-05,
1972
+ "loss": 0.8863,
1973
+ "step": 268
1974
+ },
1975
+ {
1976
+ "epoch": 0.34935064935064936,
1977
+ "grad_norm": 13.118106842041016,
1978
+ "learning_rate": 5.800865800865801e-05,
1979
+ "loss": 1.0875,
1980
+ "step": 269
1981
+ },
1982
+ {
1983
+ "epoch": 0.35064935064935066,
1984
+ "grad_norm": 12.499323844909668,
1985
+ "learning_rate": 5.822510822510823e-05,
1986
+ "loss": 0.9424,
1987
+ "step": 270
1988
+ },
1989
+ {
1990
+ "epoch": 0.35194805194805195,
1991
+ "grad_norm": 11.671981811523438,
1992
+ "learning_rate": 5.844155844155844e-05,
1993
+ "loss": 0.7181,
1994
+ "step": 271
1995
+ },
1996
+ {
1997
+ "epoch": 0.35324675324675325,
1998
+ "grad_norm": 12.202942848205566,
1999
+ "learning_rate": 5.8658008658008656e-05,
2000
+ "loss": 0.9829,
2001
+ "step": 272
2002
+ },
2003
+ {
2004
+ "epoch": 0.35454545454545455,
2005
+ "grad_norm": 12.920580863952637,
2006
+ "learning_rate": 5.887445887445888e-05,
2007
+ "loss": 0.8581,
2008
+ "step": 273
2009
+ },
2010
+ {
2011
+ "epoch": 0.35584415584415585,
2012
+ "grad_norm": 10.84870433807373,
2013
+ "learning_rate": 5.90909090909091e-05,
2014
+ "loss": 0.669,
2015
+ "step": 274
2016
+ },
2017
+ {
2018
+ "epoch": 0.35714285714285715,
2019
+ "grad_norm": 13.038583755493164,
2020
+ "learning_rate": 5.930735930735931e-05,
2021
+ "loss": 1.0623,
2022
+ "step": 275
2023
+ },
2024
+ {
2025
+ "epoch": 0.35844155844155845,
2026
+ "grad_norm": 11.414346694946289,
2027
+ "learning_rate": 5.9523809523809524e-05,
2028
+ "loss": 0.8738,
2029
+ "step": 276
2030
+ },
2031
+ {
2032
+ "epoch": 0.35974025974025975,
2033
+ "grad_norm": 12.988877296447754,
2034
+ "learning_rate": 5.9740259740259744e-05,
2035
+ "loss": 1.0384,
2036
+ "step": 277
2037
+ },
2038
+ {
2039
+ "epoch": 0.36103896103896105,
2040
+ "grad_norm": 12.147465705871582,
2041
+ "learning_rate": 5.995670995670996e-05,
2042
+ "loss": 0.9253,
2043
+ "step": 278
2044
+ },
2045
+ {
2046
+ "epoch": 0.36233766233766235,
2047
+ "grad_norm": 9.99681568145752,
2048
+ "learning_rate": 6.0173160173160184e-05,
2049
+ "loss": 0.6211,
2050
+ "step": 279
2051
+ },
2052
+ {
2053
+ "epoch": 0.36363636363636365,
2054
+ "grad_norm": 10.887341499328613,
2055
+ "learning_rate": 6.03896103896104e-05,
2056
+ "loss": 0.7466,
2057
+ "step": 280
2058
+ },
2059
+ {
2060
+ "epoch": 0.36493506493506495,
2061
+ "grad_norm": 11.995635032653809,
2062
+ "learning_rate": 6.060606060606061e-05,
2063
+ "loss": 0.8584,
2064
+ "step": 281
2065
+ },
2066
+ {
2067
+ "epoch": 0.36623376623376624,
2068
+ "grad_norm": 11.818434715270996,
2069
+ "learning_rate": 6.0822510822510825e-05,
2070
+ "loss": 0.8048,
2071
+ "step": 282
2072
+ },
2073
+ {
2074
+ "epoch": 0.36753246753246754,
2075
+ "grad_norm": 11.93957805633545,
2076
+ "learning_rate": 6.103896103896104e-05,
2077
+ "loss": 0.7548,
2078
+ "step": 283
2079
+ },
2080
+ {
2081
+ "epoch": 0.36883116883116884,
2082
+ "grad_norm": 12.328822135925293,
2083
+ "learning_rate": 6.125541125541126e-05,
2084
+ "loss": 0.7822,
2085
+ "step": 284
2086
+ },
2087
+ {
2088
+ "epoch": 0.37012987012987014,
2089
+ "grad_norm": 13.202193260192871,
2090
+ "learning_rate": 6.147186147186147e-05,
2091
+ "loss": 1.0389,
2092
+ "step": 285
2093
+ },
2094
+ {
2095
+ "epoch": 0.37142857142857144,
2096
+ "grad_norm": 10.988061904907227,
2097
+ "learning_rate": 6.16883116883117e-05,
2098
+ "loss": 0.7156,
2099
+ "step": 286
2100
+ },
2101
+ {
2102
+ "epoch": 0.37272727272727274,
2103
+ "grad_norm": 11.593925476074219,
2104
+ "learning_rate": 6.19047619047619e-05,
2105
+ "loss": 0.7989,
2106
+ "step": 287
2107
+ },
2108
+ {
2109
+ "epoch": 0.37402597402597404,
2110
+ "grad_norm": 14.6128568649292,
2111
+ "learning_rate": 6.212121212121213e-05,
2112
+ "loss": 1.0917,
2113
+ "step": 288
2114
+ },
2115
+ {
2116
+ "epoch": 0.37532467532467534,
2117
+ "grad_norm": 12.564693450927734,
2118
+ "learning_rate": 6.233766233766233e-05,
2119
+ "loss": 0.9575,
2120
+ "step": 289
2121
+ },
2122
+ {
2123
+ "epoch": 0.37662337662337664,
2124
+ "grad_norm": 12.380928039550781,
2125
+ "learning_rate": 6.255411255411255e-05,
2126
+ "loss": 0.9086,
2127
+ "step": 290
2128
+ },
2129
+ {
2130
+ "epoch": 0.37792207792207794,
2131
+ "grad_norm": 13.53039836883545,
2132
+ "learning_rate": 6.277056277056277e-05,
2133
+ "loss": 1.0582,
2134
+ "step": 291
2135
+ },
2136
+ {
2137
+ "epoch": 0.37922077922077924,
2138
+ "grad_norm": 12.092082023620605,
2139
+ "learning_rate": 6.2987012987013e-05,
2140
+ "loss": 0.879,
2141
+ "step": 292
2142
+ },
2143
+ {
2144
+ "epoch": 0.38051948051948054,
2145
+ "grad_norm": 10.496182441711426,
2146
+ "learning_rate": 6.320346320346321e-05,
2147
+ "loss": 0.6524,
2148
+ "step": 293
2149
+ },
2150
+ {
2151
+ "epoch": 0.38181818181818183,
2152
+ "grad_norm": 11.645306587219238,
2153
+ "learning_rate": 6.341991341991342e-05,
2154
+ "loss": 0.767,
2155
+ "step": 294
2156
+ },
2157
+ {
2158
+ "epoch": 0.38311688311688313,
2159
+ "grad_norm": 12.280860900878906,
2160
+ "learning_rate": 6.363636363636364e-05,
2161
+ "loss": 0.8842,
2162
+ "step": 295
2163
+ },
2164
+ {
2165
+ "epoch": 0.38441558441558443,
2166
+ "grad_norm": 12.687444686889648,
2167
+ "learning_rate": 6.385281385281385e-05,
2168
+ "loss": 0.9167,
2169
+ "step": 296
2170
+ },
2171
+ {
2172
+ "epoch": 0.38571428571428573,
2173
+ "grad_norm": 12.968870162963867,
2174
+ "learning_rate": 6.406926406926407e-05,
2175
+ "loss": 1.0106,
2176
+ "step": 297
2177
+ },
2178
+ {
2179
+ "epoch": 0.38701298701298703,
2180
+ "grad_norm": 10.828211784362793,
2181
+ "learning_rate": 6.428571428571429e-05,
2182
+ "loss": 0.8014,
2183
+ "step": 298
2184
+ },
2185
+ {
2186
+ "epoch": 0.38831168831168833,
2187
+ "grad_norm": 10.606407165527344,
2188
+ "learning_rate": 6.450216450216451e-05,
2189
+ "loss": 0.7783,
2190
+ "step": 299
2191
+ },
2192
+ {
2193
+ "epoch": 0.38961038961038963,
2194
+ "grad_norm": 11.974769592285156,
2195
+ "learning_rate": 6.471861471861473e-05,
2196
+ "loss": 0.8608,
2197
+ "step": 300
2198
+ },
2199
+ {
2200
+ "epoch": 0.39090909090909093,
2201
+ "grad_norm": 10.047250747680664,
2202
+ "learning_rate": 6.493506493506494e-05,
2203
+ "loss": 0.6383,
2204
+ "step": 301
2205
+ },
2206
+ {
2207
+ "epoch": 0.3922077922077922,
2208
+ "grad_norm": 12.191349983215332,
2209
+ "learning_rate": 6.515151515151516e-05,
2210
+ "loss": 0.9668,
2211
+ "step": 302
2212
+ },
2213
+ {
2214
+ "epoch": 0.3935064935064935,
2215
+ "grad_norm": 13.38426685333252,
2216
+ "learning_rate": 6.536796536796536e-05,
2217
+ "loss": 0.9447,
2218
+ "step": 303
2219
+ },
2220
+ {
2221
+ "epoch": 0.3948051948051948,
2222
+ "grad_norm": 10.779813766479492,
2223
+ "learning_rate": 6.55844155844156e-05,
2224
+ "loss": 0.6609,
2225
+ "step": 304
2226
+ },
2227
+ {
2228
+ "epoch": 0.3961038961038961,
2229
+ "grad_norm": 11.49394702911377,
2230
+ "learning_rate": 6.58008658008658e-05,
2231
+ "loss": 0.8104,
2232
+ "step": 305
2233
+ },
2234
+ {
2235
+ "epoch": 0.3974025974025974,
2236
+ "grad_norm": 10.99763298034668,
2237
+ "learning_rate": 6.601731601731602e-05,
2238
+ "loss": 0.7231,
2239
+ "step": 306
2240
+ },
2241
+ {
2242
+ "epoch": 0.3987012987012987,
2243
+ "grad_norm": 12.296140670776367,
2244
+ "learning_rate": 6.623376623376624e-05,
2245
+ "loss": 0.853,
2246
+ "step": 307
2247
+ },
2248
+ {
2249
+ "epoch": 0.4,
2250
+ "grad_norm": 12.42627239227295,
2251
+ "learning_rate": 6.645021645021645e-05,
2252
+ "loss": 0.8893,
2253
+ "step": 308
2254
+ },
2255
+ {
2256
+ "epoch": 0.4,
2257
+ "eval_allNLI--triplets-1024_cosine_accuracy": 0.9453125,
2258
+ "eval_allNLI--triplets-128_cosine_accuracy": 0.9140625,
2259
+ "eval_allNLI--triplets-256_cosine_accuracy": 0.9296875,
2260
+ "eval_allNLI--triplets-32_cosine_accuracy": 0.9140625,
2261
+ "eval_allNLI--triplets-512_cosine_accuracy": 0.9375,
2262
+ "eval_allNLI-triplets_cosine_accuracy": 0.9453125,
2263
+ "eval_global_dataset_loss": 0.6458740234375,
2264
+ "eval_global_dataset_runtime": 113.11,
2265
+ "eval_global_dataset_samples_per_second": 9.469,
2266
+ "eval_global_dataset_steps_per_second": 0.08,
2267
+ "eval_sequential_score": 0.9140625,
2268
+ "eval_sts-test-1024_pearson_cosine": 0.8917026424281715,
2269
+ "eval_sts-test-1024_spearman_cosine": 0.9280852568183454,
2270
+ "eval_sts-test-128_pearson_cosine": 0.8799111388722691,
2271
+ "eval_sts-test-128_spearman_cosine": 0.9174530691277519,
2272
+ "eval_sts-test-256_pearson_cosine": 0.8869619950989827,
2273
+ "eval_sts-test-256_spearman_cosine": 0.924964524442147,
2274
+ "eval_sts-test-32_pearson_cosine": 0.8506411556887592,
2275
+ "eval_sts-test-32_spearman_cosine": 0.9038834688150462,
2276
+ "eval_sts-test-512_pearson_cosine": 0.8886416592593962,
2277
+ "eval_sts-test-512_spearman_cosine": 0.9265875487493185,
2278
+ "eval_sts-test-64_pearson_cosine": 0.8675142483501502,
2279
+ "eval_sts-test-64_spearman_cosine": 0.9149750300300387,
2280
+ "eval_sts-test_pearson_cosine": 0.8953751075507514,
2281
+ "eval_sts-test_spearman_cosine": 0.9303824824867356,
2282
+ "step": 308
2283
+ },
2284
+ {
2285
+ "epoch": 0.4012987012987013,
2286
+ "grad_norm": 10.81874942779541,
2287
+ "learning_rate": 6.666666666666667e-05,
2288
+ "loss": 0.6674,
2289
+ "step": 309
2290
+ },
2291
+ {
2292
+ "epoch": 0.4025974025974026,
2293
+ "grad_norm": 10.319266319274902,
2294
+ "learning_rate": 6.688311688311688e-05,
2295
+ "loss": 0.7047,
2296
+ "step": 310
2297
+ },
2298
+ {
2299
+ "epoch": 0.4038961038961039,
2300
+ "grad_norm": 11.703706741333008,
2301
+ "learning_rate": 6.709956709956711e-05,
2302
+ "loss": 0.778,
2303
+ "step": 311
2304
+ },
2305
+ {
2306
+ "epoch": 0.4051948051948052,
2307
+ "grad_norm": 9.888168334960938,
2308
+ "learning_rate": 6.731601731601732e-05,
2309
+ "loss": 0.6232,
2310
+ "step": 312
2311
+ },
2312
+ {
2313
+ "epoch": 0.4064935064935065,
2314
+ "grad_norm": 11.571825981140137,
2315
+ "learning_rate": 6.753246753246754e-05,
2316
+ "loss": 0.8989,
2317
+ "step": 313
2318
+ },
2319
+ {
2320
+ "epoch": 0.4077922077922078,
2321
+ "grad_norm": 12.945980072021484,
2322
+ "learning_rate": 6.774891774891774e-05,
2323
+ "loss": 0.9614,
2324
+ "step": 314
2325
+ },
2326
+ {
2327
+ "epoch": 0.4090909090909091,
2328
+ "grad_norm": 12.342436790466309,
2329
+ "learning_rate": 6.796536796536796e-05,
2330
+ "loss": 0.9365,
2331
+ "step": 315
2332
+ },
2333
+ {
2334
+ "epoch": 0.4103896103896104,
2335
+ "grad_norm": 11.575722694396973,
2336
+ "learning_rate": 6.818181818181818e-05,
2337
+ "loss": 0.8101,
2338
+ "step": 316
2339
+ },
2340
+ {
2341
+ "epoch": 0.4116883116883117,
2342
+ "grad_norm": 11.095999717712402,
2343
+ "learning_rate": 6.83982683982684e-05,
2344
+ "loss": 0.8199,
2345
+ "step": 317
2346
+ },
2347
+ {
2348
+ "epoch": 0.412987012987013,
2349
+ "grad_norm": 11.149064064025879,
2350
+ "learning_rate": 6.861471861471862e-05,
2351
+ "loss": 0.8435,
2352
+ "step": 318
2353
+ },
2354
+ {
2355
+ "epoch": 0.4142857142857143,
2356
+ "grad_norm": 11.510180473327637,
2357
+ "learning_rate": 6.883116883116883e-05,
2358
+ "loss": 0.9404,
2359
+ "step": 319
2360
+ },
2361
+ {
2362
+ "epoch": 0.4155844155844156,
2363
+ "grad_norm": 10.658203125,
2364
+ "learning_rate": 6.904761904761905e-05,
2365
+ "loss": 0.6967,
2366
+ "step": 320
2367
+ },
2368
+ {
2369
+ "epoch": 0.41688311688311686,
2370
+ "grad_norm": 11.118837356567383,
2371
+ "learning_rate": 6.926406926406926e-05,
2372
+ "loss": 0.8238,
2373
+ "step": 321
2374
+ },
2375
+ {
2376
+ "epoch": 0.41818181818181815,
2377
+ "grad_norm": 11.655184745788574,
2378
+ "learning_rate": 6.948051948051948e-05,
2379
+ "loss": 0.9732,
2380
+ "step": 322
2381
+ },
2382
+ {
2383
+ "epoch": 0.41948051948051945,
2384
+ "grad_norm": 10.72670841217041,
2385
+ "learning_rate": 6.96969696969697e-05,
2386
+ "loss": 0.6953,
2387
+ "step": 323
2388
+ },
2389
+ {
2390
+ "epoch": 0.42077922077922075,
2391
+ "grad_norm": 10.018539428710938,
2392
+ "learning_rate": 6.991341991341992e-05,
2393
+ "loss": 0.6633,
2394
+ "step": 324
2395
+ },
2396
+ {
2397
+ "epoch": 0.42207792207792205,
2398
+ "grad_norm": 11.938111305236816,
2399
+ "learning_rate": 7.012987012987014e-05,
2400
+ "loss": 0.9486,
2401
+ "step": 325
2402
+ },
2403
+ {
2404
+ "epoch": 0.42337662337662335,
2405
+ "grad_norm": 12.473401069641113,
2406
+ "learning_rate": 7.034632034632035e-05,
2407
+ "loss": 0.9642,
2408
+ "step": 326
2409
+ },
2410
+ {
2411
+ "epoch": 0.42467532467532465,
2412
+ "grad_norm": 11.210603713989258,
2413
+ "learning_rate": 7.056277056277057e-05,
2414
+ "loss": 0.7443,
2415
+ "step": 327
2416
+ },
2417
+ {
2418
+ "epoch": 0.42597402597402595,
2419
+ "grad_norm": 12.498502731323242,
2420
+ "learning_rate": 7.077922077922077e-05,
2421
+ "loss": 1.0133,
2422
+ "step": 328
2423
+ },
2424
+ {
2425
+ "epoch": 0.42727272727272725,
2426
+ "grad_norm": 13.254681587219238,
2427
+ "learning_rate": 7.099567099567101e-05,
2428
+ "loss": 1.0489,
2429
+ "step": 329
2430
+ },
2431
+ {
2432
+ "epoch": 0.42857142857142855,
2433
+ "grad_norm": 9.72973918914795,
2434
+ "learning_rate": 7.121212121212121e-05,
2435
+ "loss": 0.5865,
2436
+ "step": 330
2437
+ },
2438
+ {
2439
+ "epoch": 0.42987012987012985,
2440
+ "grad_norm": 10.99699592590332,
2441
+ "learning_rate": 7.142857142857143e-05,
2442
+ "loss": 0.7757,
2443
+ "step": 331
2444
+ },
2445
+ {
2446
+ "epoch": 0.43116883116883115,
2447
+ "grad_norm": 12.112956047058105,
2448
+ "learning_rate": 7.164502164502165e-05,
2449
+ "loss": 0.9716,
2450
+ "step": 332
2451
+ },
2452
+ {
2453
+ "epoch": 0.43246753246753245,
2454
+ "grad_norm": 10.795978546142578,
2455
+ "learning_rate": 7.186147186147186e-05,
2456
+ "loss": 0.7289,
2457
+ "step": 333
2458
+ },
2459
+ {
2460
+ "epoch": 0.43376623376623374,
2461
+ "grad_norm": 11.796692848205566,
2462
+ "learning_rate": 7.207792207792208e-05,
2463
+ "loss": 0.864,
2464
+ "step": 334
2465
+ },
2466
+ {
2467
+ "epoch": 0.43506493506493504,
2468
+ "grad_norm": 10.812421798706055,
2469
+ "learning_rate": 7.229437229437229e-05,
2470
+ "loss": 0.706,
2471
+ "step": 335
2472
+ },
2473
+ {
2474
+ "epoch": 0.43636363636363634,
2475
+ "grad_norm": 10.89258098602295,
2476
+ "learning_rate": 7.251082251082252e-05,
2477
+ "loss": 0.7336,
2478
+ "step": 336
2479
+ },
2480
+ {
2481
+ "epoch": 0.43766233766233764,
2482
+ "grad_norm": 10.730430603027344,
2483
+ "learning_rate": 7.272727272727273e-05,
2484
+ "loss": 0.6631,
2485
+ "step": 337
2486
+ },
2487
+ {
2488
+ "epoch": 0.43896103896103894,
2489
+ "grad_norm": 11.678521156311035,
2490
+ "learning_rate": 7.294372294372295e-05,
2491
+ "loss": 0.8824,
2492
+ "step": 338
2493
+ },
2494
+ {
2495
+ "epoch": 0.44025974025974024,
2496
+ "grad_norm": 10.951140403747559,
2497
+ "learning_rate": 7.316017316017317e-05,
2498
+ "loss": 0.7209,
2499
+ "step": 339
2500
+ },
2501
+ {
2502
+ "epoch": 0.44155844155844154,
2503
+ "grad_norm": 10.380850791931152,
2504
+ "learning_rate": 7.337662337662338e-05,
2505
+ "loss": 0.6866,
2506
+ "step": 340
2507
+ },
2508
+ {
2509
+ "epoch": 0.44285714285714284,
2510
+ "grad_norm": 10.9320707321167,
2511
+ "learning_rate": 7.35930735930736e-05,
2512
+ "loss": 0.8544,
2513
+ "step": 341
2514
+ },
2515
+ {
2516
+ "epoch": 0.44415584415584414,
2517
+ "grad_norm": 11.206174850463867,
2518
+ "learning_rate": 7.380952380952382e-05,
2519
+ "loss": 0.7208,
2520
+ "step": 342
2521
+ },
2522
+ {
2523
+ "epoch": 0.44545454545454544,
2524
+ "grad_norm": 12.502874374389648,
2525
+ "learning_rate": 7.402597402597404e-05,
2526
+ "loss": 0.9237,
2527
+ "step": 343
2528
+ },
2529
+ {
2530
+ "epoch": 0.44675324675324674,
2531
+ "grad_norm": 10.985047340393066,
2532
+ "learning_rate": 7.424242424242424e-05,
2533
+ "loss": 0.8569,
2534
+ "step": 344
2535
+ },
2536
+ {
2537
+ "epoch": 0.44805194805194803,
2538
+ "grad_norm": 11.576885223388672,
2539
+ "learning_rate": 7.445887445887446e-05,
2540
+ "loss": 0.9145,
2541
+ "step": 345
2542
+ },
2543
+ {
2544
+ "epoch": 0.44935064935064933,
2545
+ "grad_norm": 10.66639518737793,
2546
+ "learning_rate": 7.467532467532467e-05,
2547
+ "loss": 0.8172,
2548
+ "step": 346
2549
+ },
2550
+ {
2551
+ "epoch": 0.45064935064935063,
2552
+ "grad_norm": 10.709420204162598,
2553
+ "learning_rate": 7.489177489177489e-05,
2554
+ "loss": 0.8755,
2555
+ "step": 347
2556
+ },
2557
+ {
2558
+ "epoch": 0.45194805194805193,
2559
+ "grad_norm": 9.81844425201416,
2560
+ "learning_rate": 7.510822510822511e-05,
2561
+ "loss": 0.7049,
2562
+ "step": 348
2563
+ },
2564
+ {
2565
+ "epoch": 0.45324675324675323,
2566
+ "grad_norm": 10.78894329071045,
2567
+ "learning_rate": 7.532467532467533e-05,
2568
+ "loss": 0.8571,
2569
+ "step": 349
2570
+ },
2571
+ {
2572
+ "epoch": 0.45454545454545453,
2573
+ "grad_norm": 10.1858549118042,
2574
+ "learning_rate": 7.554112554112555e-05,
2575
+ "loss": 0.71,
2576
+ "step": 350
2577
+ },
2578
+ {
2579
+ "epoch": 0.45584415584415583,
2580
+ "grad_norm": 10.635689735412598,
2581
+ "learning_rate": 7.575757575757576e-05,
2582
+ "loss": 0.7568,
2583
+ "step": 351
2584
+ },
2585
+ {
2586
+ "epoch": 0.45714285714285713,
2587
+ "grad_norm": 9.103358268737793,
2588
+ "learning_rate": 7.597402597402598e-05,
2589
+ "loss": 0.5732,
2590
+ "step": 352
2591
+ },
2592
+ {
2593
+ "epoch": 0.4584415584415584,
2594
+ "grad_norm": 11.04738712310791,
2595
+ "learning_rate": 7.619047619047618e-05,
2596
+ "loss": 0.7801,
2597
+ "step": 353
2598
+ },
2599
+ {
2600
+ "epoch": 0.4597402597402597,
2601
+ "grad_norm": 10.552362442016602,
2602
+ "learning_rate": 7.640692640692642e-05,
2603
+ "loss": 0.6895,
2604
+ "step": 354
2605
+ },
2606
+ {
2607
+ "epoch": 0.461038961038961,
2608
+ "grad_norm": 11.45176887512207,
2609
+ "learning_rate": 7.662337662337662e-05,
2610
+ "loss": 0.8078,
2611
+ "step": 355
2612
+ },
2613
+ {
2614
+ "epoch": 0.4623376623376623,
2615
+ "grad_norm": 12.02926254272461,
2616
+ "learning_rate": 7.683982683982685e-05,
2617
+ "loss": 0.9609,
2618
+ "step": 356
2619
+ },
2620
+ {
2621
+ "epoch": 0.4636363636363636,
2622
+ "grad_norm": 12.51374626159668,
2623
+ "learning_rate": 7.705627705627707e-05,
2624
+ "loss": 0.9131,
2625
+ "step": 357
2626
+ },
2627
+ {
2628
+ "epoch": 0.4649350649350649,
2629
+ "grad_norm": 11.659627914428711,
2630
+ "learning_rate": 7.727272727272727e-05,
2631
+ "loss": 0.8842,
2632
+ "step": 358
2633
+ },
2634
+ {
2635
+ "epoch": 0.4662337662337662,
2636
+ "grad_norm": 10.37750244140625,
2637
+ "learning_rate": 7.748917748917749e-05,
2638
+ "loss": 0.7314,
2639
+ "step": 359
2640
+ },
2641
+ {
2642
+ "epoch": 0.4675324675324675,
2643
+ "grad_norm": 10.31059455871582,
2644
+ "learning_rate": 7.770562770562771e-05,
2645
+ "loss": 0.7894,
2646
+ "step": 360
2647
+ },
2648
+ {
2649
+ "epoch": 0.4688311688311688,
2650
+ "grad_norm": 10.418132781982422,
2651
+ "learning_rate": 7.792207792207793e-05,
2652
+ "loss": 0.7982,
2653
+ "step": 361
2654
+ },
2655
+ {
2656
+ "epoch": 0.4701298701298701,
2657
+ "grad_norm": 10.784585952758789,
2658
+ "learning_rate": 7.813852813852814e-05,
2659
+ "loss": 0.8729,
2660
+ "step": 362
2661
+ },
2662
+ {
2663
+ "epoch": 0.4714285714285714,
2664
+ "grad_norm": 11.513368606567383,
2665
+ "learning_rate": 7.835497835497836e-05,
2666
+ "loss": 0.9753,
2667
+ "step": 363
2668
+ },
2669
+ {
2670
+ "epoch": 0.4727272727272727,
2671
+ "grad_norm": 9.88028335571289,
2672
+ "learning_rate": 7.857142857142858e-05,
2673
+ "loss": 0.6694,
2674
+ "step": 364
2675
+ },
2676
+ {
2677
+ "epoch": 0.474025974025974,
2678
+ "grad_norm": 10.181724548339844,
2679
+ "learning_rate": 7.878787878787879e-05,
2680
+ "loss": 0.7563,
2681
+ "step": 365
2682
+ },
2683
+ {
2684
+ "epoch": 0.4753246753246753,
2685
+ "grad_norm": 10.725600242614746,
2686
+ "learning_rate": 7.900432900432901e-05,
2687
+ "loss": 0.7814,
2688
+ "step": 366
2689
+ },
2690
+ {
2691
+ "epoch": 0.4766233766233766,
2692
+ "grad_norm": 9.441679000854492,
2693
+ "learning_rate": 7.922077922077923e-05,
2694
+ "loss": 0.5552,
2695
+ "step": 367
2696
+ },
2697
+ {
2698
+ "epoch": 0.4779220779220779,
2699
+ "grad_norm": 12.737554550170898,
2700
+ "learning_rate": 7.943722943722945e-05,
2701
+ "loss": 1.0459,
2702
+ "step": 368
2703
+ },
2704
+ {
2705
+ "epoch": 0.4792207792207792,
2706
+ "grad_norm": 10.601099014282227,
2707
+ "learning_rate": 7.965367965367965e-05,
2708
+ "loss": 0.9098,
2709
+ "step": 369
2710
+ },
2711
+ {
2712
+ "epoch": 0.4805194805194805,
2713
+ "grad_norm": 10.562568664550781,
2714
+ "learning_rate": 7.987012987012987e-05,
2715
+ "loss": 0.7817,
2716
+ "step": 370
2717
+ },
2718
+ {
2719
+ "epoch": 0.4818181818181818,
2720
+ "grad_norm": 9.046710968017578,
2721
+ "learning_rate": 8.008658008658008e-05,
2722
+ "loss": 0.6203,
2723
+ "step": 371
2724
+ },
2725
+ {
2726
+ "epoch": 0.4831168831168831,
2727
+ "grad_norm": 10.283613204956055,
2728
+ "learning_rate": 8.03030303030303e-05,
2729
+ "loss": 0.7885,
2730
+ "step": 372
2731
+ },
2732
+ {
2733
+ "epoch": 0.4844155844155844,
2734
+ "grad_norm": 10.934900283813477,
2735
+ "learning_rate": 8.051948051948052e-05,
2736
+ "loss": 0.7767,
2737
+ "step": 373
2738
+ },
2739
+ {
2740
+ "epoch": 0.4857142857142857,
2741
+ "grad_norm": 10.616456985473633,
2742
+ "learning_rate": 8.073593073593074e-05,
2743
+ "loss": 0.8126,
2744
+ "step": 374
2745
+ },
2746
+ {
2747
+ "epoch": 0.487012987012987,
2748
+ "grad_norm": 10.01803970336914,
2749
+ "learning_rate": 8.095238095238096e-05,
2750
+ "loss": 0.7252,
2751
+ "step": 375
2752
+ },
2753
+ {
2754
+ "epoch": 0.4883116883116883,
2755
+ "grad_norm": 10.306028366088867,
2756
+ "learning_rate": 8.116883116883117e-05,
2757
+ "loss": 0.8657,
2758
+ "step": 376
2759
+ },
2760
+ {
2761
+ "epoch": 0.4896103896103896,
2762
+ "grad_norm": 10.669329643249512,
2763
+ "learning_rate": 8.138528138528139e-05,
2764
+ "loss": 0.8491,
2765
+ "step": 377
2766
+ },
2767
+ {
2768
+ "epoch": 0.4909090909090909,
2769
+ "grad_norm": 10.165072441101074,
2770
+ "learning_rate": 8.16017316017316e-05,
2771
+ "loss": 0.784,
2772
+ "step": 378
2773
+ },
2774
+ {
2775
+ "epoch": 0.4922077922077922,
2776
+ "grad_norm": 8.929415702819824,
2777
+ "learning_rate": 8.181818181818183e-05,
2778
+ "loss": 0.6108,
2779
+ "step": 379
2780
+ },
2781
+ {
2782
+ "epoch": 0.4935064935064935,
2783
+ "grad_norm": 11.677674293518066,
2784
+ "learning_rate": 8.203463203463204e-05,
2785
+ "loss": 1.0215,
2786
+ "step": 380
2787
+ },
2788
+ {
2789
+ "epoch": 0.4948051948051948,
2790
+ "grad_norm": 10.382437705993652,
2791
+ "learning_rate": 8.225108225108226e-05,
2792
+ "loss": 0.8208,
2793
+ "step": 381
2794
+ },
2795
+ {
2796
+ "epoch": 0.4961038961038961,
2797
+ "grad_norm": 10.082671165466309,
2798
+ "learning_rate": 8.246753246753248e-05,
2799
+ "loss": 0.7456,
2800
+ "step": 382
2801
+ },
2802
+ {
2803
+ "epoch": 0.4974025974025974,
2804
+ "grad_norm": 10.928060531616211,
2805
+ "learning_rate": 8.268398268398268e-05,
2806
+ "loss": 1.0181,
2807
+ "step": 383
2808
+ },
2809
+ {
2810
+ "epoch": 0.4987012987012987,
2811
+ "grad_norm": 10.922670364379883,
2812
+ "learning_rate": 8.29004329004329e-05,
2813
+ "loss": 0.8263,
2814
+ "step": 384
2815
+ },
2816
+ {
2817
+ "epoch": 0.5,
2818
+ "grad_norm": 10.033293724060059,
2819
+ "learning_rate": 8.311688311688312e-05,
2820
+ "loss": 0.7218,
2821
+ "step": 385
2822
+ },
2823
+ {
2824
+ "epoch": 0.5,
2825
+ "eval_allNLI--triplets-1024_cosine_accuracy": 0.953125,
2826
+ "eval_allNLI--triplets-128_cosine_accuracy": 0.921875,
2827
+ "eval_allNLI--triplets-256_cosine_accuracy": 0.9375,
2828
+ "eval_allNLI--triplets-32_cosine_accuracy": 0.90625,
2829
+ "eval_allNLI--triplets-512_cosine_accuracy": 0.9375,
2830
+ "eval_allNLI-triplets_cosine_accuracy": 0.9609375,
2831
+ "eval_global_dataset_loss": 0.6583967208862305,
2832
+ "eval_global_dataset_runtime": 112.7848,
2833
+ "eval_global_dataset_samples_per_second": 9.496,
2834
+ "eval_global_dataset_steps_per_second": 0.08,
2835
+ "eval_sequential_score": 0.90625,
2836
+ "eval_sts-test-1024_pearson_cosine": 0.8804763387747968,
2837
+ "eval_sts-test-1024_spearman_cosine": 0.9192536844466944,
2838
+ "eval_sts-test-128_pearson_cosine": 0.867682985932934,
2839
+ "eval_sts-test-128_spearman_cosine": 0.9096662761358334,
2840
+ "eval_sts-test-256_pearson_cosine": 0.8723111264791312,
2841
+ "eval_sts-test-256_spearman_cosine": 0.9130394312202947,
2842
+ "eval_sts-test-32_pearson_cosine": 0.8338645489107269,
2843
+ "eval_sts-test-32_spearman_cosine": 0.8955438521897122,
2844
+ "eval_sts-test-512_pearson_cosine": 0.8794209596972049,
2845
+ "eval_sts-test-512_spearman_cosine": 0.9186070526578676,
2846
+ "eval_sts-test-64_pearson_cosine": 0.8606438099441043,
2847
+ "eval_sts-test-64_spearman_cosine": 0.9104955115859569,
2848
+ "eval_sts-test_pearson_cosine": 0.8855632236901727,
2849
+ "eval_sts-test_spearman_cosine": 0.9217009315544635,
2850
+ "step": 385
2851
+ },
2852
+ {
2853
+ "epoch": 0.5012987012987012,
2854
+ "grad_norm": 9.382086753845215,
2855
+ "learning_rate": 8.333333333333334e-05,
2856
+ "loss": 0.6829,
2857
+ "step": 386
2858
+ },
2859
+ {
2860
+ "epoch": 0.5025974025974026,
2861
+ "grad_norm": 10.253658294677734,
2862
+ "learning_rate": 8.354978354978355e-05,
2863
+ "loss": 0.8304,
2864
+ "step": 387
2865
+ },
2866
+ {
2867
+ "epoch": 0.5038961038961038,
2868
+ "grad_norm": 9.05147647857666,
2869
+ "learning_rate": 8.376623376623377e-05,
2870
+ "loss": 0.6935,
2871
+ "step": 388
2872
+ },
2873
+ {
2874
+ "epoch": 0.5051948051948052,
2875
+ "grad_norm": 10.728251457214355,
2876
+ "learning_rate": 8.398268398268399e-05,
2877
+ "loss": 0.9569,
2878
+ "step": 389
2879
+ },
2880
+ {
2881
+ "epoch": 0.5064935064935064,
2882
+ "grad_norm": 9.828920364379883,
2883
+ "learning_rate": 8.41991341991342e-05,
2884
+ "loss": 0.7174,
2885
+ "step": 390
2886
+ },
2887
+ {
2888
+ "epoch": 0.5077922077922078,
2889
+ "grad_norm": 10.326533317565918,
2890
+ "learning_rate": 8.441558441558442e-05,
2891
+ "loss": 0.8543,
2892
+ "step": 391
2893
+ },
2894
+ {
2895
+ "epoch": 0.509090909090909,
2896
+ "grad_norm": 10.311280250549316,
2897
+ "learning_rate": 8.463203463203464e-05,
2898
+ "loss": 0.8769,
2899
+ "step": 392
2900
+ },
2901
+ {
2902
+ "epoch": 0.5103896103896104,
2903
+ "grad_norm": 8.989762306213379,
2904
+ "learning_rate": 8.484848484848486e-05,
2905
+ "loss": 0.5945,
2906
+ "step": 393
2907
+ },
2908
+ {
2909
+ "epoch": 0.5116883116883116,
2910
+ "grad_norm": 9.76150894165039,
2911
+ "learning_rate": 8.506493506493507e-05,
2912
+ "loss": 0.7787,
2913
+ "step": 394
2914
+ },
2915
+ {
2916
+ "epoch": 0.512987012987013,
2917
+ "grad_norm": 11.173674583435059,
2918
+ "learning_rate": 8.528138528138529e-05,
2919
+ "loss": 0.9674,
2920
+ "step": 395
2921
+ },
2922
+ {
2923
+ "epoch": 0.5142857142857142,
2924
+ "grad_norm": 8.645306587219238,
2925
+ "learning_rate": 8.549783549783549e-05,
2926
+ "loss": 0.7011,
2927
+ "step": 396
2928
+ },
2929
+ {
2930
+ "epoch": 0.5155844155844156,
2931
+ "grad_norm": 11.850281715393066,
2932
+ "learning_rate": 8.571428571428571e-05,
2933
+ "loss": 0.9993,
2934
+ "step": 397
2935
+ },
2936
+ {
2937
+ "epoch": 0.5168831168831168,
2938
+ "grad_norm": 8.96986198425293,
2939
+ "learning_rate": 8.593073593073593e-05,
2940
+ "loss": 0.6368,
2941
+ "step": 398
2942
+ },
2943
+ {
2944
+ "epoch": 0.5181818181818182,
2945
+ "grad_norm": 9.281974792480469,
2946
+ "learning_rate": 8.614718614718615e-05,
2947
+ "loss": 0.6236,
2948
+ "step": 399
2949
+ },
2950
+ {
2951
+ "epoch": 0.5194805194805194,
2952
+ "grad_norm": 9.10240650177002,
2953
+ "learning_rate": 8.636363636363637e-05,
2954
+ "loss": 0.6543,
2955
+ "step": 400
2956
+ },
2957
+ {
2958
+ "epoch": 0.5207792207792208,
2959
+ "grad_norm": 9.484097480773926,
2960
+ "learning_rate": 8.658008658008658e-05,
2961
+ "loss": 0.6927,
2962
+ "step": 401
2963
+ },
2964
+ {
2965
+ "epoch": 0.522077922077922,
2966
+ "grad_norm": 12.794879913330078,
2967
+ "learning_rate": 8.67965367965368e-05,
2968
+ "loss": 1.1117,
2969
+ "step": 402
2970
+ },
2971
+ {
2972
+ "epoch": 0.5233766233766234,
2973
+ "grad_norm": 10.31647777557373,
2974
+ "learning_rate": 8.701298701298701e-05,
2975
+ "loss": 0.8847,
2976
+ "step": 403
2977
+ },
2978
+ {
2979
+ "epoch": 0.5246753246753246,
2980
+ "grad_norm": 11.432804107666016,
2981
+ "learning_rate": 8.722943722943724e-05,
2982
+ "loss": 0.9657,
2983
+ "step": 404
2984
+ },
2985
+ {
2986
+ "epoch": 0.525974025974026,
2987
+ "grad_norm": 11.18320369720459,
2988
+ "learning_rate": 8.744588744588745e-05,
2989
+ "loss": 0.8722,
2990
+ "step": 405
2991
+ },
2992
+ {
2993
+ "epoch": 0.5272727272727272,
2994
+ "grad_norm": 9.664923667907715,
2995
+ "learning_rate": 8.766233766233767e-05,
2996
+ "loss": 0.7465,
2997
+ "step": 406
2998
+ },
2999
+ {
3000
+ "epoch": 0.5285714285714286,
3001
+ "grad_norm": 8.084552764892578,
3002
+ "learning_rate": 8.787878787878789e-05,
3003
+ "loss": 0.5488,
3004
+ "step": 407
3005
+ },
3006
+ {
3007
+ "epoch": 0.5298701298701298,
3008
+ "grad_norm": 9.583297729492188,
3009
+ "learning_rate": 8.80952380952381e-05,
3010
+ "loss": 0.6888,
3011
+ "step": 408
3012
+ },
3013
+ {
3014
+ "epoch": 0.5311688311688312,
3015
+ "grad_norm": 10.47243595123291,
3016
+ "learning_rate": 8.831168831168831e-05,
3017
+ "loss": 0.7561,
3018
+ "step": 409
3019
+ },
3020
+ {
3021
+ "epoch": 0.5324675324675324,
3022
+ "grad_norm": 9.173786163330078,
3023
+ "learning_rate": 8.852813852813854e-05,
3024
+ "loss": 0.5411,
3025
+ "step": 410
3026
+ },
3027
+ {
3028
+ "epoch": 0.5337662337662338,
3029
+ "grad_norm": 10.696986198425293,
3030
+ "learning_rate": 8.874458874458876e-05,
3031
+ "loss": 0.7663,
3032
+ "step": 411
3033
+ },
3034
+ {
3035
+ "epoch": 0.535064935064935,
3036
+ "grad_norm": 12.49531078338623,
3037
+ "learning_rate": 8.896103896103896e-05,
3038
+ "loss": 1.0137,
3039
+ "step": 412
3040
+ },
3041
+ {
3042
+ "epoch": 0.5363636363636364,
3043
+ "grad_norm": 9.877222061157227,
3044
+ "learning_rate": 8.917748917748918e-05,
3045
+ "loss": 0.667,
3046
+ "step": 413
3047
+ },
3048
+ {
3049
+ "epoch": 0.5376623376623376,
3050
+ "grad_norm": 9.855676651000977,
3051
+ "learning_rate": 8.93939393939394e-05,
3052
+ "loss": 0.7111,
3053
+ "step": 414
3054
+ },
3055
+ {
3056
+ "epoch": 0.538961038961039,
3057
+ "grad_norm": 11.593993186950684,
3058
+ "learning_rate": 8.961038961038961e-05,
3059
+ "loss": 0.8531,
3060
+ "step": 415
3061
+ },
3062
+ {
3063
+ "epoch": 0.5402597402597402,
3064
+ "grad_norm": 10.321015357971191,
3065
+ "learning_rate": 8.982683982683983e-05,
3066
+ "loss": 0.8561,
3067
+ "step": 416
3068
+ },
3069
+ {
3070
+ "epoch": 0.5415584415584416,
3071
+ "grad_norm": 8.318321228027344,
3072
+ "learning_rate": 9.004329004329005e-05,
3073
+ "loss": 0.612,
3074
+ "step": 417
3075
+ },
3076
+ {
3077
+ "epoch": 0.5428571428571428,
3078
+ "grad_norm": 10.25160026550293,
3079
+ "learning_rate": 9.025974025974027e-05,
3080
+ "loss": 0.7229,
3081
+ "step": 418
3082
+ },
3083
+ {
3084
+ "epoch": 0.5441558441558442,
3085
+ "grad_norm": 9.849309921264648,
3086
+ "learning_rate": 9.047619047619048e-05,
3087
+ "loss": 0.8387,
3088
+ "step": 419
3089
+ },
3090
+ {
3091
+ "epoch": 0.5454545454545454,
3092
+ "grad_norm": 9.49033260345459,
3093
+ "learning_rate": 9.06926406926407e-05,
3094
+ "loss": 0.7439,
3095
+ "step": 420
3096
+ },
3097
+ {
3098
+ "epoch": 0.5467532467532468,
3099
+ "grad_norm": 8.027220726013184,
3100
+ "learning_rate": 9.090909090909092e-05,
3101
+ "loss": 0.5846,
3102
+ "step": 421
3103
+ },
3104
+ {
3105
+ "epoch": 0.548051948051948,
3106
+ "grad_norm": 8.108675003051758,
3107
+ "learning_rate": 9.112554112554112e-05,
3108
+ "loss": 0.5976,
3109
+ "step": 422
3110
+ },
3111
+ {
3112
+ "epoch": 0.5493506493506494,
3113
+ "grad_norm": 9.802972793579102,
3114
+ "learning_rate": 9.134199134199136e-05,
3115
+ "loss": 0.7378,
3116
+ "step": 423
3117
+ },
3118
+ {
3119
+ "epoch": 0.5506493506493506,
3120
+ "grad_norm": 10.613837242126465,
3121
+ "learning_rate": 9.155844155844156e-05,
3122
+ "loss": 0.8756,
3123
+ "step": 424
3124
+ },
3125
+ {
3126
+ "epoch": 0.551948051948052,
3127
+ "grad_norm": 10.099786758422852,
3128
+ "learning_rate": 9.177489177489178e-05,
3129
+ "loss": 0.6755,
3130
+ "step": 425
3131
+ },
3132
+ {
3133
+ "epoch": 0.5532467532467532,
3134
+ "grad_norm": 9.524248123168945,
3135
+ "learning_rate": 9.199134199134199e-05,
3136
+ "loss": 0.6566,
3137
+ "step": 426
3138
+ },
3139
+ {
3140
+ "epoch": 0.5545454545454546,
3141
+ "grad_norm": 9.682258605957031,
3142
+ "learning_rate": 9.220779220779221e-05,
3143
+ "loss": 0.5624,
3144
+ "step": 427
3145
+ },
3146
+ {
3147
+ "epoch": 0.5558441558441558,
3148
+ "grad_norm": 9.256147384643555,
3149
+ "learning_rate": 9.242424242424242e-05,
3150
+ "loss": 0.5887,
3151
+ "step": 428
3152
+ },
3153
+ {
3154
+ "epoch": 0.5571428571428572,
3155
+ "grad_norm": 11.301840782165527,
3156
+ "learning_rate": 9.264069264069265e-05,
3157
+ "loss": 0.9906,
3158
+ "step": 429
3159
+ },
3160
+ {
3161
+ "epoch": 0.5584415584415584,
3162
+ "grad_norm": 10.433177947998047,
3163
+ "learning_rate": 9.285714285714286e-05,
3164
+ "loss": 0.735,
3165
+ "step": 430
3166
+ },
3167
+ {
3168
+ "epoch": 0.5597402597402598,
3169
+ "grad_norm": 11.203728675842285,
3170
+ "learning_rate": 9.307359307359308e-05,
3171
+ "loss": 0.8759,
3172
+ "step": 431
3173
+ },
3174
+ {
3175
+ "epoch": 0.561038961038961,
3176
+ "grad_norm": 10.540921211242676,
3177
+ "learning_rate": 9.32900432900433e-05,
3178
+ "loss": 0.8514,
3179
+ "step": 432
3180
+ },
3181
+ {
3182
+ "epoch": 0.5623376623376624,
3183
+ "grad_norm": 9.670626640319824,
3184
+ "learning_rate": 9.35064935064935e-05,
3185
+ "loss": 0.7531,
3186
+ "step": 433
3187
+ },
3188
+ {
3189
+ "epoch": 0.5636363636363636,
3190
+ "grad_norm": 9.004586219787598,
3191
+ "learning_rate": 9.372294372294373e-05,
3192
+ "loss": 0.6816,
3193
+ "step": 434
3194
+ },
3195
+ {
3196
+ "epoch": 0.564935064935065,
3197
+ "grad_norm": 9.889915466308594,
3198
+ "learning_rate": 9.393939393939395e-05,
3199
+ "loss": 0.8911,
3200
+ "step": 435
3201
+ },
3202
+ {
3203
+ "epoch": 0.5662337662337662,
3204
+ "grad_norm": 9.216524124145508,
3205
+ "learning_rate": 9.415584415584417e-05,
3206
+ "loss": 0.7555,
3207
+ "step": 436
3208
+ },
3209
+ {
3210
+ "epoch": 0.5675324675324676,
3211
+ "grad_norm": 9.710390090942383,
3212
+ "learning_rate": 9.437229437229437e-05,
3213
+ "loss": 0.8079,
3214
+ "step": 437
3215
+ },
3216
+ {
3217
+ "epoch": 0.5688311688311688,
3218
+ "grad_norm": 8.755694389343262,
3219
+ "learning_rate": 9.45887445887446e-05,
3220
+ "loss": 0.6203,
3221
+ "step": 438
3222
+ },
3223
+ {
3224
+ "epoch": 0.5701298701298702,
3225
+ "grad_norm": 9.18278694152832,
3226
+ "learning_rate": 9.480519480519481e-05,
3227
+ "loss": 0.6348,
3228
+ "step": 439
3229
+ },
3230
+ {
3231
+ "epoch": 0.5714285714285714,
3232
+ "grad_norm": 8.409049987792969,
3233
+ "learning_rate": 9.502164502164502e-05,
3234
+ "loss": 0.5986,
3235
+ "step": 440
3236
+ },
3237
+ {
3238
+ "epoch": 0.5727272727272728,
3239
+ "grad_norm": 10.397635459899902,
3240
+ "learning_rate": 9.523809523809524e-05,
3241
+ "loss": 0.8259,
3242
+ "step": 441
3243
+ },
3244
+ {
3245
+ "epoch": 0.574025974025974,
3246
+ "grad_norm": 9.205062866210938,
3247
+ "learning_rate": 9.545454545454546e-05,
3248
+ "loss": 0.7051,
3249
+ "step": 442
3250
+ },
3251
+ {
3252
+ "epoch": 0.5753246753246753,
3253
+ "grad_norm": 9.241551399230957,
3254
+ "learning_rate": 9.567099567099568e-05,
3255
+ "loss": 0.591,
3256
+ "step": 443
3257
+ },
3258
+ {
3259
+ "epoch": 0.5766233766233766,
3260
+ "grad_norm": 9.01000690460205,
3261
+ "learning_rate": 9.588744588744589e-05,
3262
+ "loss": 0.7197,
3263
+ "step": 444
3264
+ },
3265
+ {
3266
+ "epoch": 0.577922077922078,
3267
+ "grad_norm": 8.933067321777344,
3268
+ "learning_rate": 9.610389610389611e-05,
3269
+ "loss": 0.6624,
3270
+ "step": 445
3271
+ },
3272
+ {
3273
+ "epoch": 0.5792207792207792,
3274
+ "grad_norm": 10.92744255065918,
3275
+ "learning_rate": 9.632034632034633e-05,
3276
+ "loss": 0.9108,
3277
+ "step": 446
3278
+ },
3279
+ {
3280
+ "epoch": 0.5805194805194805,
3281
+ "grad_norm": 9.574989318847656,
3282
+ "learning_rate": 9.653679653679654e-05,
3283
+ "loss": 0.7401,
3284
+ "step": 447
3285
+ },
3286
+ {
3287
+ "epoch": 0.5818181818181818,
3288
+ "grad_norm": 9.929916381835938,
3289
+ "learning_rate": 9.675324675324677e-05,
3290
+ "loss": 0.7475,
3291
+ "step": 448
3292
+ },
3293
+ {
3294
+ "epoch": 0.5831168831168831,
3295
+ "grad_norm": 10.541657447814941,
3296
+ "learning_rate": 9.696969696969698e-05,
3297
+ "loss": 1.0367,
3298
+ "step": 449
3299
+ },
3300
+ {
3301
+ "epoch": 0.5844155844155844,
3302
+ "grad_norm": 9.241097450256348,
3303
+ "learning_rate": 9.71861471861472e-05,
3304
+ "loss": 0.819,
3305
+ "step": 450
3306
+ },
3307
+ {
3308
+ "epoch": 0.5857142857142857,
3309
+ "grad_norm": 10.230196952819824,
3310
+ "learning_rate": 9.74025974025974e-05,
3311
+ "loss": 0.9914,
3312
+ "step": 451
3313
+ },
3314
+ {
3315
+ "epoch": 0.587012987012987,
3316
+ "grad_norm": 7.502562999725342,
3317
+ "learning_rate": 9.761904761904762e-05,
3318
+ "loss": 0.5852,
3319
+ "step": 452
3320
+ },
3321
+ {
3322
+ "epoch": 0.5883116883116883,
3323
+ "grad_norm": 9.432659149169922,
3324
+ "learning_rate": 9.783549783549783e-05,
3325
+ "loss": 0.8283,
3326
+ "step": 453
3327
+ },
3328
+ {
3329
+ "epoch": 0.5896103896103896,
3330
+ "grad_norm": 8.886880874633789,
3331
+ "learning_rate": 9.805194805194806e-05,
3332
+ "loss": 0.9477,
3333
+ "step": 454
3334
+ },
3335
+ {
3336
+ "epoch": 0.5909090909090909,
3337
+ "grad_norm": 8.474105834960938,
3338
+ "learning_rate": 9.826839826839827e-05,
3339
+ "loss": 0.7091,
3340
+ "step": 455
3341
+ },
3342
+ {
3343
+ "epoch": 0.5922077922077922,
3344
+ "grad_norm": 9.925548553466797,
3345
+ "learning_rate": 9.848484848484849e-05,
3346
+ "loss": 0.958,
3347
+ "step": 456
3348
+ },
3349
+ {
3350
+ "epoch": 0.5935064935064935,
3351
+ "grad_norm": 7.961300849914551,
3352
+ "learning_rate": 9.870129870129871e-05,
3353
+ "loss": 0.6203,
3354
+ "step": 457
3355
+ },
3356
+ {
3357
+ "epoch": 0.5948051948051948,
3358
+ "grad_norm": 8.644482612609863,
3359
+ "learning_rate": 9.891774891774892e-05,
3360
+ "loss": 0.6183,
3361
+ "step": 458
3362
+ },
3363
+ {
3364
+ "epoch": 0.5961038961038961,
3365
+ "grad_norm": 8.991077423095703,
3366
+ "learning_rate": 9.913419913419914e-05,
3367
+ "loss": 0.7303,
3368
+ "step": 459
3369
+ },
3370
+ {
3371
+ "epoch": 0.5974025974025974,
3372
+ "grad_norm": 10.966276168823242,
3373
+ "learning_rate": 9.935064935064936e-05,
3374
+ "loss": 0.9273,
3375
+ "step": 460
3376
+ },
3377
+ {
3378
+ "epoch": 0.5987012987012987,
3379
+ "grad_norm": 10.023669242858887,
3380
+ "learning_rate": 9.956709956709958e-05,
3381
+ "loss": 0.7091,
3382
+ "step": 461
3383
+ },
3384
+ {
3385
+ "epoch": 0.6,
3386
+ "grad_norm": 11.322511672973633,
3387
+ "learning_rate": 9.978354978354978e-05,
3388
+ "loss": 0.9906,
3389
+ "step": 462
3390
+ },
3391
+ {
3392
+ "epoch": 0.6,
3393
+ "eval_allNLI--triplets-1024_cosine_accuracy": 0.9453125,
3394
+ "eval_allNLI--triplets-128_cosine_accuracy": 0.9296875,
3395
+ "eval_allNLI--triplets-256_cosine_accuracy": 0.9375,
3396
+ "eval_allNLI--triplets-32_cosine_accuracy": 0.921875,
3397
+ "eval_allNLI--triplets-512_cosine_accuracy": 0.9453125,
3398
+ "eval_allNLI-triplets_cosine_accuracy": 0.9453125,
3399
+ "eval_global_dataset_loss": 0.6539728045463562,
3400
+ "eval_global_dataset_runtime": 113.7266,
3401
+ "eval_global_dataset_samples_per_second": 9.417,
3402
+ "eval_global_dataset_steps_per_second": 0.079,
3403
+ "eval_sequential_score": 0.921875,
3404
+ "eval_sts-test-1024_pearson_cosine": 0.8797687244140173,
3405
+ "eval_sts-test-1024_spearman_cosine": 0.9263218783244517,
3406
+ "eval_sts-test-128_pearson_cosine": 0.865709098903634,
3407
+ "eval_sts-test-128_spearman_cosine": 0.9134647903371974,
3408
+ "eval_sts-test-256_pearson_cosine": 0.873391317761029,
3409
+ "eval_sts-test-256_spearman_cosine": 0.9216776585388081,
3410
+ "eval_sts-test-32_pearson_cosine": 0.835229116152323,
3411
+ "eval_sts-test-32_spearman_cosine": 0.8990595097392636,
3412
+ "eval_sts-test-512_pearson_cosine": 0.8784096835085939,
3413
+ "eval_sts-test-512_spearman_cosine": 0.9279259261727042,
3414
+ "eval_sts-test-64_pearson_cosine": 0.8518886914166497,
3415
+ "eval_sts-test-64_spearman_cosine": 0.9082587957582707,
3416
+ "eval_sts-test_pearson_cosine": 0.8813813147500178,
3417
+ "eval_sts-test_spearman_cosine": 0.9273104444202163,
3418
+ "step": 462
3419
+ }
3420
+ ],
3421
+ "logging_steps": 1,
3422
+ "max_steps": 2310,
3423
+ "num_input_tokens_seen": 0,
3424
+ "num_train_epochs": 3,
3425
+ "save_steps": 231,
3426
+ "stateful_callbacks": {
3427
+ "TrainerControl": {
3428
+ "args": {
3429
+ "should_epoch_stop": false,
3430
+ "should_evaluate": false,
3431
+ "should_log": false,
3432
+ "should_save": true,
3433
+ "should_training_stop": false
3434
+ },
3435
+ "attributes": {}
3436
+ }
3437
+ },
3438
+ "total_flos": 0.0,
3439
+ "train_batch_size": 192,
3440
+ "trial_name": null,
3441
+ "trial_params": null
3442
+ }
checkpoint-462/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9fadd592bad93c127ab2cc1e37bb17c0477dfea67bb61a8baf05d9892c2d39d8
3
+ size 5880