Training in progress, step 1413, checkpoint
Browse files- checkpoint-1413/1_Pooling/config.json +10 -0
- checkpoint-1413/README.md +0 -0
- checkpoint-1413/added_tokens.json +3 -0
- checkpoint-1413/config.json +38 -0
- checkpoint-1413/config_sentence_transformers.json +10 -0
- checkpoint-1413/modules.json +14 -0
- checkpoint-1413/optimizer.pt +3 -0
- checkpoint-1413/pytorch_model.bin +3 -0
- checkpoint-1413/rng_state.pth +3 -0
- checkpoint-1413/scheduler.pt +3 -0
- checkpoint-1413/sentence_bert_config.json +4 -0
- checkpoint-1413/special_tokens_map.json +15 -0
- checkpoint-1413/spm.model +3 -0
- checkpoint-1413/tokenizer.json +0 -0
- checkpoint-1413/tokenizer_config.json +58 -0
- checkpoint-1413/trainer_state.json +2386 -0
- checkpoint-1413/training_args.bin +3 -0
checkpoint-1413/1_Pooling/config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"word_embedding_dimension": 1536,
|
| 3 |
+
"pooling_mode_cls_token": false,
|
| 4 |
+
"pooling_mode_mean_tokens": true,
|
| 5 |
+
"pooling_mode_max_tokens": false,
|
| 6 |
+
"pooling_mode_mean_sqrt_len_tokens": false,
|
| 7 |
+
"pooling_mode_weightedmean_tokens": false,
|
| 8 |
+
"pooling_mode_lasttoken": false,
|
| 9 |
+
"include_prompt": true
|
| 10 |
+
}
|
checkpoint-1413/README.md
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1413/added_tokens.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"[MASK]": 128000
|
| 3 |
+
}
|
checkpoint-1413/config.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "microsoft/deberta-v2-xlarge",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"DebertaV2Model"
|
| 5 |
+
],
|
| 6 |
+
"attention_head_size": 64,
|
| 7 |
+
"attention_probs_dropout_prob": 0.1,
|
| 8 |
+
"conv_act": "gelu",
|
| 9 |
+
"conv_kernel_size": 3,
|
| 10 |
+
"hidden_act": "gelu",
|
| 11 |
+
"hidden_dropout_prob": 0.1,
|
| 12 |
+
"hidden_size": 1536,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 6144,
|
| 15 |
+
"layer_norm_eps": 1e-07,
|
| 16 |
+
"max_position_embeddings": 512,
|
| 17 |
+
"max_relative_positions": -1,
|
| 18 |
+
"model_type": "deberta-v2",
|
| 19 |
+
"norm_rel_ebd": "layer_norm",
|
| 20 |
+
"num_attention_heads": 24,
|
| 21 |
+
"num_hidden_layers": 24,
|
| 22 |
+
"pad_token_id": 0,
|
| 23 |
+
"pooler_dropout": 0,
|
| 24 |
+
"pooler_hidden_act": "gelu",
|
| 25 |
+
"pooler_hidden_size": 1536,
|
| 26 |
+
"pos_att_type": [
|
| 27 |
+
"p2c",
|
| 28 |
+
"c2p"
|
| 29 |
+
],
|
| 30 |
+
"position_biased_input": false,
|
| 31 |
+
"position_buckets": 256,
|
| 32 |
+
"relative_attention": true,
|
| 33 |
+
"share_att_key": true,
|
| 34 |
+
"torch_dtype": "float32",
|
| 35 |
+
"transformers_version": "4.42.4",
|
| 36 |
+
"type_vocab_size": 0,
|
| 37 |
+
"vocab_size": 128100
|
| 38 |
+
}
|
checkpoint-1413/config_sentence_transformers.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"__version__": {
|
| 3 |
+
"sentence_transformers": "3.0.1",
|
| 4 |
+
"transformers": "4.42.4",
|
| 5 |
+
"pytorch": "2.4.0+cu121"
|
| 6 |
+
},
|
| 7 |
+
"prompts": {},
|
| 8 |
+
"default_prompt_name": null,
|
| 9 |
+
"similarity_fn_name": null
|
| 10 |
+
}
|
checkpoint-1413/modules.json
ADDED
|
@@ -0,0 +1,14 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"idx": 0,
|
| 4 |
+
"name": "0",
|
| 5 |
+
"path": "",
|
| 6 |
+
"type": "sentence_transformers.models.Transformer"
|
| 7 |
+
},
|
| 8 |
+
{
|
| 9 |
+
"idx": 1,
|
| 10 |
+
"name": "1",
|
| 11 |
+
"path": "1_Pooling",
|
| 12 |
+
"type": "sentence_transformers.models.Pooling"
|
| 13 |
+
}
|
| 14 |
+
]
|
checkpoint-1413/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f050b9d63994fc837693ead3d6362c264dc47651cff52f4585dc3fa955a8c1c0
|
| 3 |
+
size 7077084396
|
checkpoint-1413/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:33a374a5a57010930b77a804b02d5d76b11247800c9306196e68d0cd6deadb8f
|
| 3 |
+
size 3538506546
|
checkpoint-1413/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cdbc81179c5915b521503305fafa40e63a9238f25aa77dbc6d377a97f0c44b5f
|
| 3 |
+
size 14244
|
checkpoint-1413/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:80ae535c89c6643b66511848c7b4c02b660a3660b1072801516c52846f1d4546
|
| 3 |
+
size 1064
|
checkpoint-1413/sentence_bert_config.json
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"max_seq_length": 512,
|
| 3 |
+
"do_lower_case": false
|
| 4 |
+
}
|
checkpoint-1413/special_tokens_map.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "[CLS]",
|
| 3 |
+
"cls_token": "[CLS]",
|
| 4 |
+
"eos_token": "[SEP]",
|
| 5 |
+
"mask_token": "[MASK]",
|
| 6 |
+
"pad_token": "[PAD]",
|
| 7 |
+
"sep_token": "[SEP]",
|
| 8 |
+
"unk_token": {
|
| 9 |
+
"content": "[UNK]",
|
| 10 |
+
"lstrip": false,
|
| 11 |
+
"normalized": true,
|
| 12 |
+
"rstrip": false,
|
| 13 |
+
"single_word": false
|
| 14 |
+
}
|
| 15 |
+
}
|
checkpoint-1413/spm.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:5598d5e96f339a8d980c15f9afd405a2e5e1be7db41de3ed13b0f03fac1e8c17
|
| 3 |
+
size 2447305
|
checkpoint-1413/tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-1413/tokenizer_config.json
ADDED
|
@@ -0,0 +1,58 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"added_tokens_decoder": {
|
| 3 |
+
"0": {
|
| 4 |
+
"content": "[PAD]",
|
| 5 |
+
"lstrip": false,
|
| 6 |
+
"normalized": false,
|
| 7 |
+
"rstrip": false,
|
| 8 |
+
"single_word": false,
|
| 9 |
+
"special": true
|
| 10 |
+
},
|
| 11 |
+
"1": {
|
| 12 |
+
"content": "[CLS]",
|
| 13 |
+
"lstrip": false,
|
| 14 |
+
"normalized": false,
|
| 15 |
+
"rstrip": false,
|
| 16 |
+
"single_word": false,
|
| 17 |
+
"special": true
|
| 18 |
+
},
|
| 19 |
+
"2": {
|
| 20 |
+
"content": "[SEP]",
|
| 21 |
+
"lstrip": false,
|
| 22 |
+
"normalized": false,
|
| 23 |
+
"rstrip": false,
|
| 24 |
+
"single_word": false,
|
| 25 |
+
"special": true
|
| 26 |
+
},
|
| 27 |
+
"3": {
|
| 28 |
+
"content": "[UNK]",
|
| 29 |
+
"lstrip": false,
|
| 30 |
+
"normalized": true,
|
| 31 |
+
"rstrip": false,
|
| 32 |
+
"single_word": false,
|
| 33 |
+
"special": true
|
| 34 |
+
},
|
| 35 |
+
"128000": {
|
| 36 |
+
"content": "[MASK]",
|
| 37 |
+
"lstrip": false,
|
| 38 |
+
"normalized": false,
|
| 39 |
+
"rstrip": false,
|
| 40 |
+
"single_word": false,
|
| 41 |
+
"special": true
|
| 42 |
+
}
|
| 43 |
+
},
|
| 44 |
+
"bos_token": "[CLS]",
|
| 45 |
+
"clean_up_tokenization_spaces": true,
|
| 46 |
+
"cls_token": "[CLS]",
|
| 47 |
+
"do_lower_case": false,
|
| 48 |
+
"eos_token": "[SEP]",
|
| 49 |
+
"mask_token": "[MASK]",
|
| 50 |
+
"model_max_length": 1000000000000000019884624838656,
|
| 51 |
+
"pad_token": "[PAD]",
|
| 52 |
+
"sep_token": "[SEP]",
|
| 53 |
+
"sp_model_kwargs": {},
|
| 54 |
+
"split_by_punct": false,
|
| 55 |
+
"tokenizer_class": "DebertaV2Tokenizer",
|
| 56 |
+
"unk_token": "[UNK]",
|
| 57 |
+
"vocab_type": "spm"
|
| 58 |
+
}
|
checkpoint-1413/trainer_state.json
ADDED
|
@@ -0,0 +1,2386 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9034526854219949,
|
| 5 |
+
"eval_steps": 157,
|
| 6 |
+
"global_step": 1413,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.010230179028132993,
|
| 13 |
+
"grad_norm": 200.60667419433594,
|
| 14 |
+
"learning_rate": 2.129925452609159e-07,
|
| 15 |
+
"loss": 7.1882,
|
| 16 |
+
"step": 16
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.020460358056265986,
|
| 20 |
+
"grad_norm": 174.4742889404297,
|
| 21 |
+
"learning_rate": 5.324813631522898e-07,
|
| 22 |
+
"loss": 9.1489,
|
| 23 |
+
"step": 32
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.030690537084398978,
|
| 27 |
+
"grad_norm": 193.9903106689453,
|
| 28 |
+
"learning_rate": 8.732694355697552e-07,
|
| 29 |
+
"loss": 8.805,
|
| 30 |
+
"step": 48
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.04092071611253197,
|
| 34 |
+
"grad_norm": 3077.94775390625,
|
| 35 |
+
"learning_rate": 1.192758253461129e-06,
|
| 36 |
+
"loss": 5.7489,
|
| 37 |
+
"step": 64
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.05115089514066496,
|
| 41 |
+
"grad_norm": 80.12879943847656,
|
| 42 |
+
"learning_rate": 1.5335463258785943e-06,
|
| 43 |
+
"loss": 4.7163,
|
| 44 |
+
"step": 80
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.061381074168797956,
|
| 48 |
+
"grad_norm": 99.18599700927734,
|
| 49 |
+
"learning_rate": 1.8743343982960599e-06,
|
| 50 |
+
"loss": 3.176,
|
| 51 |
+
"step": 96
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.07161125319693094,
|
| 55 |
+
"grad_norm": 23.476715087890625,
|
| 56 |
+
"learning_rate": 2.1938232161874336e-06,
|
| 57 |
+
"loss": 2.034,
|
| 58 |
+
"step": 112
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.08184143222506395,
|
| 62 |
+
"grad_norm": 39.820777893066406,
|
| 63 |
+
"learning_rate": 2.534611288604899e-06,
|
| 64 |
+
"loss": 1.1278,
|
| 65 |
+
"step": 128
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.09207161125319693,
|
| 69 |
+
"grad_norm": 12.513270378112793,
|
| 70 |
+
"learning_rate": 2.8753993610223648e-06,
|
| 71 |
+
"loss": 0.7996,
|
| 72 |
+
"step": 144
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.100383631713555,
|
| 76 |
+
"eval_Qnli-dev_cosine_accuracy": 0.662109375,
|
| 77 |
+
"eval_Qnli-dev_cosine_accuracy_threshold": 0.8656402826309204,
|
| 78 |
+
"eval_Qnli-dev_cosine_ap": 0.6654869266843996,
|
| 79 |
+
"eval_Qnli-dev_cosine_f1": 0.6504559270516718,
|
| 80 |
+
"eval_Qnli-dev_cosine_f1_threshold": 0.6888570785522461,
|
| 81 |
+
"eval_Qnli-dev_cosine_precision": 0.5071090047393365,
|
| 82 |
+
"eval_Qnli-dev_cosine_recall": 0.9067796610169492,
|
| 83 |
+
"eval_Qnli-dev_dot_accuracy": 0.654296875,
|
| 84 |
+
"eval_Qnli-dev_dot_accuracy_threshold": 905.306640625,
|
| 85 |
+
"eval_Qnli-dev_dot_ap": 0.6262830221036416,
|
| 86 |
+
"eval_Qnli-dev_dot_f1": 0.65625,
|
| 87 |
+
"eval_Qnli-dev_dot_f1_threshold": 810.4493408203125,
|
| 88 |
+
"eval_Qnli-dev_dot_precision": 0.5558823529411765,
|
| 89 |
+
"eval_Qnli-dev_dot_recall": 0.8008474576271186,
|
| 90 |
+
"eval_Qnli-dev_euclidean_accuracy": 0.662109375,
|
| 91 |
+
"eval_Qnli-dev_euclidean_accuracy_threshold": 17.05492401123047,
|
| 92 |
+
"eval_Qnli-dev_euclidean_ap": 0.6671756125841228,
|
| 93 |
+
"eval_Qnli-dev_euclidean_f1": 0.6495176848874598,
|
| 94 |
+
"eval_Qnli-dev_euclidean_f1_threshold": 23.73143768310547,
|
| 95 |
+
"eval_Qnli-dev_euclidean_precision": 0.5233160621761658,
|
| 96 |
+
"eval_Qnli-dev_euclidean_recall": 0.8559322033898306,
|
| 97 |
+
"eval_Qnli-dev_manhattan_accuracy": 0.65625,
|
| 98 |
+
"eval_Qnli-dev_manhattan_accuracy_threshold": 526.1962890625,
|
| 99 |
+
"eval_Qnli-dev_manhattan_ap": 0.6662986170174698,
|
| 100 |
+
"eval_Qnli-dev_manhattan_f1": 0.6487341772151899,
|
| 101 |
+
"eval_Qnli-dev_manhattan_f1_threshold": 748.320068359375,
|
| 102 |
+
"eval_Qnli-dev_manhattan_precision": 0.5176767676767676,
|
| 103 |
+
"eval_Qnli-dev_manhattan_recall": 0.8686440677966102,
|
| 104 |
+
"eval_Qnli-dev_max_accuracy": 0.662109375,
|
| 105 |
+
"eval_Qnli-dev_max_accuracy_threshold": 905.306640625,
|
| 106 |
+
"eval_Qnli-dev_max_ap": 0.6671756125841228,
|
| 107 |
+
"eval_Qnli-dev_max_f1": 0.65625,
|
| 108 |
+
"eval_Qnli-dev_max_f1_threshold": 810.4493408203125,
|
| 109 |
+
"eval_Qnli-dev_max_precision": 0.5558823529411765,
|
| 110 |
+
"eval_Qnli-dev_max_recall": 0.9067796610169492,
|
| 111 |
+
"eval_allNLI-dev_cosine_accuracy": 0.6875,
|
| 112 |
+
"eval_allNLI-dev_cosine_accuracy_threshold": 0.9388645887374878,
|
| 113 |
+
"eval_allNLI-dev_cosine_ap": 0.5213973850248246,
|
| 114 |
+
"eval_allNLI-dev_cosine_f1": 0.5590361445783132,
|
| 115 |
+
"eval_allNLI-dev_cosine_f1_threshold": 0.7874077558517456,
|
| 116 |
+
"eval_allNLI-dev_cosine_precision": 0.4793388429752066,
|
| 117 |
+
"eval_allNLI-dev_cosine_recall": 0.6705202312138728,
|
| 118 |
+
"eval_allNLI-dev_dot_accuracy": 0.6953125,
|
| 119 |
+
"eval_allNLI-dev_dot_accuracy_threshold": 942.6085205078125,
|
| 120 |
+
"eval_allNLI-dev_dot_ap": 0.520777376545363,
|
| 121 |
+
"eval_allNLI-dev_dot_f1": 0.5538461538461539,
|
| 122 |
+
"eval_allNLI-dev_dot_f1_threshold": 817.1332397460938,
|
| 123 |
+
"eval_allNLI-dev_dot_precision": 0.44680851063829785,
|
| 124 |
+
"eval_allNLI-dev_dot_recall": 0.7283236994219653,
|
| 125 |
+
"eval_allNLI-dev_euclidean_accuracy": 0.6875,
|
| 126 |
+
"eval_allNLI-dev_euclidean_accuracy_threshold": 12.643478393554688,
|
| 127 |
+
"eval_allNLI-dev_euclidean_ap": 0.518083458930424,
|
| 128 |
+
"eval_allNLI-dev_euclidean_f1": 0.5605381165919282,
|
| 129 |
+
"eval_allNLI-dev_euclidean_f1_threshold": 22.352760314941406,
|
| 130 |
+
"eval_allNLI-dev_euclidean_precision": 0.45787545787545786,
|
| 131 |
+
"eval_allNLI-dev_euclidean_recall": 0.7225433526011561,
|
| 132 |
+
"eval_allNLI-dev_manhattan_accuracy": 0.6875,
|
| 133 |
+
"eval_allNLI-dev_manhattan_accuracy_threshold": 392.8375549316406,
|
| 134 |
+
"eval_allNLI-dev_manhattan_ap": 0.5229585168795372,
|
| 135 |
+
"eval_allNLI-dev_manhattan_f1": 0.5590909090909091,
|
| 136 |
+
"eval_allNLI-dev_manhattan_f1_threshold": 690.5735473632812,
|
| 137 |
+
"eval_allNLI-dev_manhattan_precision": 0.4606741573033708,
|
| 138 |
+
"eval_allNLI-dev_manhattan_recall": 0.7109826589595376,
|
| 139 |
+
"eval_allNLI-dev_max_accuracy": 0.6953125,
|
| 140 |
+
"eval_allNLI-dev_max_accuracy_threshold": 942.6085205078125,
|
| 141 |
+
"eval_allNLI-dev_max_ap": 0.5229585168795372,
|
| 142 |
+
"eval_allNLI-dev_max_f1": 0.5605381165919282,
|
| 143 |
+
"eval_allNLI-dev_max_f1_threshold": 817.1332397460938,
|
| 144 |
+
"eval_allNLI-dev_max_precision": 0.4793388429752066,
|
| 145 |
+
"eval_allNLI-dev_max_recall": 0.7283236994219653,
|
| 146 |
+
"eval_sequential_score": 0.6671756125841228,
|
| 147 |
+
"eval_sts-test_pearson_cosine": 0.8718645025760672,
|
| 148 |
+
"eval_sts-test_pearson_dot": 0.8442027854618407,
|
| 149 |
+
"eval_sts-test_pearson_euclidean": 0.8902978035400368,
|
| 150 |
+
"eval_sts-test_pearson_manhattan": 0.8909390707647828,
|
| 151 |
+
"eval_sts-test_pearson_max": 0.8909390707647828,
|
| 152 |
+
"eval_sts-test_spearman_cosine": 0.8775905391512535,
|
| 153 |
+
"eval_sts-test_spearman_dot": 0.8282565695883292,
|
| 154 |
+
"eval_sts-test_spearman_euclidean": 0.8785767674497972,
|
| 155 |
+
"eval_sts-test_spearman_manhattan": 0.8796654313856657,
|
| 156 |
+
"eval_sts-test_spearman_max": 0.8796654313856657,
|
| 157 |
+
"eval_vitaminc-pairs_loss": 3.82141375541687,
|
| 158 |
+
"eval_vitaminc-pairs_runtime": 4.6321,
|
| 159 |
+
"eval_vitaminc-pairs_samples_per_second": 27.633,
|
| 160 |
+
"eval_vitaminc-pairs_steps_per_second": 0.216,
|
| 161 |
+
"step": 157
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.100383631713555,
|
| 165 |
+
"eval_negation-triplets_loss": 1.2526816129684448,
|
| 166 |
+
"eval_negation-triplets_runtime": 3.4008,
|
| 167 |
+
"eval_negation-triplets_samples_per_second": 37.638,
|
| 168 |
+
"eval_negation-triplets_steps_per_second": 0.294,
|
| 169 |
+
"step": 157
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.100383631713555,
|
| 173 |
+
"eval_scitail-pairs-pos_loss": 0.07802588492631912,
|
| 174 |
+
"eval_scitail-pairs-pos_runtime": 2.7191,
|
| 175 |
+
"eval_scitail-pairs-pos_samples_per_second": 47.075,
|
| 176 |
+
"eval_scitail-pairs-pos_steps_per_second": 0.368,
|
| 177 |
+
"step": 157
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.100383631713555,
|
| 181 |
+
"eval_scitail-pairs-qa_loss": 0.11051780730485916,
|
| 182 |
+
"eval_scitail-pairs-qa_runtime": 2.1249,
|
| 183 |
+
"eval_scitail-pairs-qa_samples_per_second": 60.237,
|
| 184 |
+
"eval_scitail-pairs-qa_steps_per_second": 0.471,
|
| 185 |
+
"step": 157
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.100383631713555,
|
| 189 |
+
"eval_xsum-pairs_loss": 0.1444225311279297,
|
| 190 |
+
"eval_xsum-pairs_runtime": 3.075,
|
| 191 |
+
"eval_xsum-pairs_samples_per_second": 41.626,
|
| 192 |
+
"eval_xsum-pairs_steps_per_second": 0.325,
|
| 193 |
+
"step": 157
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.100383631713555,
|
| 197 |
+
"eval_sciq_pairs_loss": 0.10900198668241501,
|
| 198 |
+
"eval_sciq_pairs_runtime": 4.6201,
|
| 199 |
+
"eval_sciq_pairs_samples_per_second": 27.705,
|
| 200 |
+
"eval_sciq_pairs_steps_per_second": 0.216,
|
| 201 |
+
"step": 157
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.100383631713555,
|
| 205 |
+
"eval_qasc_pairs_loss": 0.8441975116729736,
|
| 206 |
+
"eval_qasc_pairs_runtime": 2.0926,
|
| 207 |
+
"eval_qasc_pairs_samples_per_second": 61.168,
|
| 208 |
+
"eval_qasc_pairs_steps_per_second": 0.478,
|
| 209 |
+
"step": 157
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.100383631713555,
|
| 213 |
+
"eval_openbookqa_pairs_loss": 0.7852775454521179,
|
| 214 |
+
"eval_openbookqa_pairs_runtime": 2.2378,
|
| 215 |
+
"eval_openbookqa_pairs_samples_per_second": 57.198,
|
| 216 |
+
"eval_openbookqa_pairs_steps_per_second": 0.447,
|
| 217 |
+
"step": 157
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.100383631713555,
|
| 221 |
+
"eval_msmarco_pairs_loss": 0.7973570823669434,
|
| 222 |
+
"eval_msmarco_pairs_runtime": 2.2477,
|
| 223 |
+
"eval_msmarco_pairs_samples_per_second": 56.948,
|
| 224 |
+
"eval_msmarco_pairs_steps_per_second": 0.445,
|
| 225 |
+
"step": 157
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.100383631713555,
|
| 229 |
+
"eval_nq_pairs_loss": 1.149829387664795,
|
| 230 |
+
"eval_nq_pairs_runtime": 3.1982,
|
| 231 |
+
"eval_nq_pairs_samples_per_second": 40.023,
|
| 232 |
+
"eval_nq_pairs_steps_per_second": 0.313,
|
| 233 |
+
"step": 157
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.100383631713555,
|
| 237 |
+
"eval_trivia_pairs_loss": 0.46642398834228516,
|
| 238 |
+
"eval_trivia_pairs_runtime": 3.2778,
|
| 239 |
+
"eval_trivia_pairs_samples_per_second": 39.05,
|
| 240 |
+
"eval_trivia_pairs_steps_per_second": 0.305,
|
| 241 |
+
"step": 157
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.100383631713555,
|
| 245 |
+
"eval_gooaq_pairs_loss": 0.7757962346076965,
|
| 246 |
+
"eval_gooaq_pairs_runtime": 2.1385,
|
| 247 |
+
"eval_gooaq_pairs_samples_per_second": 59.856,
|
| 248 |
+
"eval_gooaq_pairs_steps_per_second": 0.468,
|
| 249 |
+
"step": 157
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.100383631713555,
|
| 253 |
+
"eval_paws-pos_loss": 0.03750575706362724,
|
| 254 |
+
"eval_paws-pos_runtime": 2.2521,
|
| 255 |
+
"eval_paws-pos_samples_per_second": 56.836,
|
| 256 |
+
"eval_paws-pos_steps_per_second": 0.444,
|
| 257 |
+
"step": 157
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.100383631713555,
|
| 261 |
+
"eval_global_dataset_loss": 0.48674729466438293,
|
| 262 |
+
"eval_global_dataset_runtime": 8.6002,
|
| 263 |
+
"eval_global_dataset_samples_per_second": 37.79,
|
| 264 |
+
"eval_global_dataset_steps_per_second": 0.349,
|
| 265 |
+
"step": 157
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.10230179028132992,
|
| 269 |
+
"grad_norm": 45.195247650146484,
|
| 270 |
+
"learning_rate": 3.2161874334398298e-06,
|
| 271 |
+
"loss": 0.54,
|
| 272 |
+
"step": 160
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.11253196930946291,
|
| 276 |
+
"grad_norm": 64.59953308105469,
|
| 277 |
+
"learning_rate": 3.556975505857295e-06,
|
| 278 |
+
"loss": 0.6267,
|
| 279 |
+
"step": 176
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.12276214833759591,
|
| 283 |
+
"grad_norm": 108.30097961425781,
|
| 284 |
+
"learning_rate": 3.8977635782747605e-06,
|
| 285 |
+
"loss": 0.401,
|
| 286 |
+
"step": 192
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 0.1329923273657289,
|
| 290 |
+
"grad_norm": 10.606786727905273,
|
| 291 |
+
"learning_rate": 4.238551650692226e-06,
|
| 292 |
+
"loss": 0.455,
|
| 293 |
+
"step": 208
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 0.1432225063938619,
|
| 297 |
+
"grad_norm": 10.502503395080566,
|
| 298 |
+
"learning_rate": 4.579339723109691e-06,
|
| 299 |
+
"loss": 0.308,
|
| 300 |
+
"step": 224
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.1534526854219949,
|
| 304 |
+
"grad_norm": 10.550800323486328,
|
| 305 |
+
"learning_rate": 4.920127795527157e-06,
|
| 306 |
+
"loss": 0.2808,
|
| 307 |
+
"step": 240
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 0.1636828644501279,
|
| 311 |
+
"grad_norm": 0.047671813517808914,
|
| 312 |
+
"learning_rate": 5.260915867944623e-06,
|
| 313 |
+
"loss": 0.319,
|
| 314 |
+
"step": 256
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.17391304347826086,
|
| 318 |
+
"grad_norm": 4.752868175506592,
|
| 319 |
+
"learning_rate": 5.6017039403620875e-06,
|
| 320 |
+
"loss": 0.3241,
|
| 321 |
+
"step": 272
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.18414322250639387,
|
| 325 |
+
"grad_norm": 3.744504690170288,
|
| 326 |
+
"learning_rate": 5.942492012779553e-06,
|
| 327 |
+
"loss": 0.2181,
|
| 328 |
+
"step": 288
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.19437340153452684,
|
| 332 |
+
"grad_norm": 0.10304329544305801,
|
| 333 |
+
"learning_rate": 6.283280085197019e-06,
|
| 334 |
+
"loss": 0.3247,
|
| 335 |
+
"step": 304
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 0.20076726342711,
|
| 339 |
+
"eval_Qnli-dev_cosine_accuracy": 0.68359375,
|
| 340 |
+
"eval_Qnli-dev_cosine_accuracy_threshold": 0.728395938873291,
|
| 341 |
+
"eval_Qnli-dev_cosine_ap": 0.7208629559865216,
|
| 342 |
+
"eval_Qnli-dev_cosine_f1": 0.6887661141804787,
|
| 343 |
+
"eval_Qnli-dev_cosine_f1_threshold": 0.671984076499939,
|
| 344 |
+
"eval_Qnli-dev_cosine_precision": 0.6091205211726385,
|
| 345 |
+
"eval_Qnli-dev_cosine_recall": 0.7923728813559322,
|
| 346 |
+
"eval_Qnli-dev_dot_accuracy": 0.681640625,
|
| 347 |
+
"eval_Qnli-dev_dot_accuracy_threshold": 713.8800048828125,
|
| 348 |
+
"eval_Qnli-dev_dot_ap": 0.6851873907589782,
|
| 349 |
+
"eval_Qnli-dev_dot_f1": 0.6788685524126457,
|
| 350 |
+
"eval_Qnli-dev_dot_f1_threshold": 571.084228515625,
|
| 351 |
+
"eval_Qnli-dev_dot_precision": 0.5589041095890411,
|
| 352 |
+
"eval_Qnli-dev_dot_recall": 0.864406779661017,
|
| 353 |
+
"eval_Qnli-dev_euclidean_accuracy": 0.689453125,
|
| 354 |
+
"eval_Qnli-dev_euclidean_accuracy_threshold": 23.01095962524414,
|
| 355 |
+
"eval_Qnli-dev_euclidean_ap": 0.7234217762808675,
|
| 356 |
+
"eval_Qnli-dev_euclidean_f1": 0.6831858407079647,
|
| 357 |
+
"eval_Qnli-dev_euclidean_f1_threshold": 25.087703704833984,
|
| 358 |
+
"eval_Qnli-dev_euclidean_precision": 0.5866261398176292,
|
| 359 |
+
"eval_Qnli-dev_euclidean_recall": 0.8177966101694916,
|
| 360 |
+
"eval_Qnli-dev_manhattan_accuracy": 0.6875,
|
| 361 |
+
"eval_Qnli-dev_manhattan_accuracy_threshold": 715.2478637695312,
|
| 362 |
+
"eval_Qnli-dev_manhattan_ap": 0.7243486016703601,
|
| 363 |
+
"eval_Qnli-dev_manhattan_f1": 0.6844919786096257,
|
| 364 |
+
"eval_Qnli-dev_manhattan_f1_threshold": 780.095947265625,
|
| 365 |
+
"eval_Qnli-dev_manhattan_precision": 0.5907692307692308,
|
| 366 |
+
"eval_Qnli-dev_manhattan_recall": 0.8135593220338984,
|
| 367 |
+
"eval_Qnli-dev_max_accuracy": 0.689453125,
|
| 368 |
+
"eval_Qnli-dev_max_accuracy_threshold": 715.2478637695312,
|
| 369 |
+
"eval_Qnli-dev_max_ap": 0.7243486016703601,
|
| 370 |
+
"eval_Qnli-dev_max_f1": 0.6887661141804787,
|
| 371 |
+
"eval_Qnli-dev_max_f1_threshold": 780.095947265625,
|
| 372 |
+
"eval_Qnli-dev_max_precision": 0.6091205211726385,
|
| 373 |
+
"eval_Qnli-dev_max_recall": 0.864406779661017,
|
| 374 |
+
"eval_allNLI-dev_cosine_accuracy": 0.720703125,
|
| 375 |
+
"eval_allNLI-dev_cosine_accuracy_threshold": 0.8353008031845093,
|
| 376 |
+
"eval_allNLI-dev_cosine_ap": 0.5917629293541244,
|
| 377 |
+
"eval_allNLI-dev_cosine_f1": 0.6000000000000001,
|
| 378 |
+
"eval_allNLI-dev_cosine_f1_threshold": 0.7087839245796204,
|
| 379 |
+
"eval_allNLI-dev_cosine_precision": 0.5019455252918288,
|
| 380 |
+
"eval_allNLI-dev_cosine_recall": 0.7456647398843931,
|
| 381 |
+
"eval_allNLI-dev_dot_accuracy": 0.71484375,
|
| 382 |
+
"eval_allNLI-dev_dot_accuracy_threshold": 817.4105224609375,
|
| 383 |
+
"eval_allNLI-dev_dot_ap": 0.5858245181656638,
|
| 384 |
+
"eval_allNLI-dev_dot_f1": 0.5980392156862746,
|
| 385 |
+
"eval_allNLI-dev_dot_f1_threshold": 674.387939453125,
|
| 386 |
+
"eval_allNLI-dev_dot_precision": 0.5191489361702127,
|
| 387 |
+
"eval_allNLI-dev_dot_recall": 0.7052023121387283,
|
| 388 |
+
"eval_allNLI-dev_euclidean_accuracy": 0.716796875,
|
| 389 |
+
"eval_allNLI-dev_euclidean_accuracy_threshold": 17.109304428100586,
|
| 390 |
+
"eval_allNLI-dev_euclidean_ap": 0.5912447191211185,
|
| 391 |
+
"eval_allNLI-dev_euclidean_f1": 0.6013986013986015,
|
| 392 |
+
"eval_allNLI-dev_euclidean_f1_threshold": 23.201446533203125,
|
| 393 |
+
"eval_allNLI-dev_euclidean_precision": 0.50390625,
|
| 394 |
+
"eval_allNLI-dev_euclidean_recall": 0.7456647398843931,
|
| 395 |
+
"eval_allNLI-dev_manhattan_accuracy": 0.71875,
|
| 396 |
+
"eval_allNLI-dev_manhattan_accuracy_threshold": 552.2703857421875,
|
| 397 |
+
"eval_allNLI-dev_manhattan_ap": 0.5910112781514775,
|
| 398 |
+
"eval_allNLI-dev_manhattan_f1": 0.5981735159817351,
|
| 399 |
+
"eval_allNLI-dev_manhattan_f1_threshold": 731.6765747070312,
|
| 400 |
+
"eval_allNLI-dev_manhattan_precision": 0.49433962264150944,
|
| 401 |
+
"eval_allNLI-dev_manhattan_recall": 0.7572254335260116,
|
| 402 |
+
"eval_allNLI-dev_max_accuracy": 0.720703125,
|
| 403 |
+
"eval_allNLI-dev_max_accuracy_threshold": 817.4105224609375,
|
| 404 |
+
"eval_allNLI-dev_max_ap": 0.5917629293541244,
|
| 405 |
+
"eval_allNLI-dev_max_f1": 0.6013986013986015,
|
| 406 |
+
"eval_allNLI-dev_max_f1_threshold": 731.6765747070312,
|
| 407 |
+
"eval_allNLI-dev_max_precision": 0.5191489361702127,
|
| 408 |
+
"eval_allNLI-dev_max_recall": 0.7572254335260116,
|
| 409 |
+
"eval_sequential_score": 0.7243486016703601,
|
| 410 |
+
"eval_sts-test_pearson_cosine": 0.91107907465776,
|
| 411 |
+
"eval_sts-test_pearson_dot": 0.8802333333831622,
|
| 412 |
+
"eval_sts-test_pearson_euclidean": 0.92175721802385,
|
| 413 |
+
"eval_sts-test_pearson_manhattan": 0.9207411620779591,
|
| 414 |
+
"eval_sts-test_pearson_max": 0.92175721802385,
|
| 415 |
+
"eval_sts-test_spearman_cosine": 0.9170956996722621,
|
| 416 |
+
"eval_sts-test_spearman_dot": 0.8733774316475602,
|
| 417 |
+
"eval_sts-test_spearman_euclidean": 0.9165704988685727,
|
| 418 |
+
"eval_sts-test_spearman_manhattan": 0.9153493264477239,
|
| 419 |
+
"eval_sts-test_spearman_max": 0.9170956996722621,
|
| 420 |
+
"eval_vitaminc-pairs_loss": 3.3627231121063232,
|
| 421 |
+
"eval_vitaminc-pairs_runtime": 4.4746,
|
| 422 |
+
"eval_vitaminc-pairs_samples_per_second": 28.606,
|
| 423 |
+
"eval_vitaminc-pairs_steps_per_second": 0.223,
|
| 424 |
+
"step": 314
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 0.20076726342711,
|
| 428 |
+
"eval_negation-triplets_loss": 0.7009554505348206,
|
| 429 |
+
"eval_negation-triplets_runtime": 3.3424,
|
| 430 |
+
"eval_negation-triplets_samples_per_second": 38.296,
|
| 431 |
+
"eval_negation-triplets_steps_per_second": 0.299,
|
| 432 |
+
"step": 314
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.20076726342711,
|
| 436 |
+
"eval_scitail-pairs-pos_loss": 0.027083460241556168,
|
| 437 |
+
"eval_scitail-pairs-pos_runtime": 2.6253,
|
| 438 |
+
"eval_scitail-pairs-pos_samples_per_second": 48.757,
|
| 439 |
+
"eval_scitail-pairs-pos_steps_per_second": 0.381,
|
| 440 |
+
"step": 314
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.20076726342711,
|
| 444 |
+
"eval_scitail-pairs-qa_loss": 0.0013978304341435432,
|
| 445 |
+
"eval_scitail-pairs-qa_runtime": 2.0929,
|
| 446 |
+
"eval_scitail-pairs-qa_samples_per_second": 61.159,
|
| 447 |
+
"eval_scitail-pairs-qa_steps_per_second": 0.478,
|
| 448 |
+
"step": 314
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.20076726342711,
|
| 452 |
+
"eval_xsum-pairs_loss": 0.03819597512483597,
|
| 453 |
+
"eval_xsum-pairs_runtime": 3.0764,
|
| 454 |
+
"eval_xsum-pairs_samples_per_second": 41.607,
|
| 455 |
+
"eval_xsum-pairs_steps_per_second": 0.325,
|
| 456 |
+
"step": 314
|
| 457 |
+
},
|
| 458 |
+
{
|
| 459 |
+
"epoch": 0.20076726342711,
|
| 460 |
+
"eval_sciq_pairs_loss": 0.04831065982580185,
|
| 461 |
+
"eval_sciq_pairs_runtime": 4.6239,
|
| 462 |
+
"eval_sciq_pairs_samples_per_second": 27.682,
|
| 463 |
+
"eval_sciq_pairs_steps_per_second": 0.216,
|
| 464 |
+
"step": 314
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.20076726342711,
|
| 468 |
+
"eval_qasc_pairs_loss": 0.11294831335544586,
|
| 469 |
+
"eval_qasc_pairs_runtime": 2.1013,
|
| 470 |
+
"eval_qasc_pairs_samples_per_second": 60.915,
|
| 471 |
+
"eval_qasc_pairs_steps_per_second": 0.476,
|
| 472 |
+
"step": 314
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.20076726342711,
|
| 476 |
+
"eval_openbookqa_pairs_loss": 0.4045051634311676,
|
| 477 |
+
"eval_openbookqa_pairs_runtime": 2.2313,
|
| 478 |
+
"eval_openbookqa_pairs_samples_per_second": 57.367,
|
| 479 |
+
"eval_openbookqa_pairs_steps_per_second": 0.448,
|
| 480 |
+
"step": 314
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.20076726342711,
|
| 484 |
+
"eval_msmarco_pairs_loss": 0.17428961396217346,
|
| 485 |
+
"eval_msmarco_pairs_runtime": 2.2379,
|
| 486 |
+
"eval_msmarco_pairs_samples_per_second": 57.197,
|
| 487 |
+
"eval_msmarco_pairs_steps_per_second": 0.447,
|
| 488 |
+
"step": 314
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.20076726342711,
|
| 492 |
+
"eval_nq_pairs_loss": 0.18139179050922394,
|
| 493 |
+
"eval_nq_pairs_runtime": 3.1825,
|
| 494 |
+
"eval_nq_pairs_samples_per_second": 40.22,
|
| 495 |
+
"eval_nq_pairs_steps_per_second": 0.314,
|
| 496 |
+
"step": 314
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 0.20076726342711,
|
| 500 |
+
"eval_trivia_pairs_loss": 0.10668718814849854,
|
| 501 |
+
"eval_trivia_pairs_runtime": 3.2891,
|
| 502 |
+
"eval_trivia_pairs_samples_per_second": 38.917,
|
| 503 |
+
"eval_trivia_pairs_steps_per_second": 0.304,
|
| 504 |
+
"step": 314
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 0.20076726342711,
|
| 508 |
+
"eval_gooaq_pairs_loss": 0.2038198560476303,
|
| 509 |
+
"eval_gooaq_pairs_runtime": 2.209,
|
| 510 |
+
"eval_gooaq_pairs_samples_per_second": 57.945,
|
| 511 |
+
"eval_gooaq_pairs_steps_per_second": 0.453,
|
| 512 |
+
"step": 314
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.20076726342711,
|
| 516 |
+
"eval_paws-pos_loss": 0.02337946929037571,
|
| 517 |
+
"eval_paws-pos_runtime": 2.289,
|
| 518 |
+
"eval_paws-pos_samples_per_second": 55.92,
|
| 519 |
+
"eval_paws-pos_steps_per_second": 0.437,
|
| 520 |
+
"step": 314
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.20076726342711,
|
| 524 |
+
"eval_global_dataset_loss": 0.23699811100959778,
|
| 525 |
+
"eval_global_dataset_runtime": 8.6273,
|
| 526 |
+
"eval_global_dataset_samples_per_second": 37.671,
|
| 527 |
+
"eval_global_dataset_steps_per_second": 0.348,
|
| 528 |
+
"step": 314
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.20460358056265984,
|
| 532 |
+
"grad_norm": 7.919135570526123,
|
| 533 |
+
"learning_rate": 6.624068157614484e-06,
|
| 534 |
+
"loss": 0.233,
|
| 535 |
+
"step": 320
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.21483375959079284,
|
| 539 |
+
"grad_norm": 8.169248580932617,
|
| 540 |
+
"learning_rate": 6.96485623003195e-06,
|
| 541 |
+
"loss": 0.1946,
|
| 542 |
+
"step": 336
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.22506393861892582,
|
| 546 |
+
"grad_norm": 4.551661968231201,
|
| 547 |
+
"learning_rate": 7.305644302449415e-06,
|
| 548 |
+
"loss": 0.2393,
|
| 549 |
+
"step": 352
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.23529411764705882,
|
| 553 |
+
"grad_norm": 49.21135330200195,
|
| 554 |
+
"learning_rate": 7.64643237486688e-06,
|
| 555 |
+
"loss": 0.1476,
|
| 556 |
+
"step": 368
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.24552429667519182,
|
| 560 |
+
"grad_norm": 12.119938850402832,
|
| 561 |
+
"learning_rate": 7.987220447284347e-06,
|
| 562 |
+
"loss": 0.1976,
|
| 563 |
+
"step": 384
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.2557544757033248,
|
| 567 |
+
"grad_norm": 6.922146797180176,
|
| 568 |
+
"learning_rate": 8.32800851970181e-06,
|
| 569 |
+
"loss": 0.1139,
|
| 570 |
+
"step": 400
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.2659846547314578,
|
| 574 |
+
"grad_norm": 1.659225583076477,
|
| 575 |
+
"learning_rate": 8.668796592119276e-06,
|
| 576 |
+
"loss": 0.1986,
|
| 577 |
+
"step": 416
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.27621483375959077,
|
| 581 |
+
"grad_norm": 11.489442825317383,
|
| 582 |
+
"learning_rate": 9.009584664536743e-06,
|
| 583 |
+
"loss": 0.2405,
|
| 584 |
+
"step": 432
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.2864450127877238,
|
| 588 |
+
"grad_norm": 6.8070149421691895,
|
| 589 |
+
"learning_rate": 9.350372736954207e-06,
|
| 590 |
+
"loss": 0.1519,
|
| 591 |
+
"step": 448
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.2966751918158568,
|
| 595 |
+
"grad_norm": 8.569122314453125,
|
| 596 |
+
"learning_rate": 9.691160809371672e-06,
|
| 597 |
+
"loss": 0.1508,
|
| 598 |
+
"step": 464
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.30115089514066495,
|
| 602 |
+
"eval_Qnli-dev_cosine_accuracy": 0.689453125,
|
| 603 |
+
"eval_Qnli-dev_cosine_accuracy_threshold": 0.7045497894287109,
|
| 604 |
+
"eval_Qnli-dev_cosine_ap": 0.7262372250492908,
|
| 605 |
+
"eval_Qnli-dev_cosine_f1": 0.693103448275862,
|
| 606 |
+
"eval_Qnli-dev_cosine_f1_threshold": 0.6277430057525635,
|
| 607 |
+
"eval_Qnli-dev_cosine_precision": 0.5843023255813954,
|
| 608 |
+
"eval_Qnli-dev_cosine_recall": 0.8516949152542372,
|
| 609 |
+
"eval_Qnli-dev_dot_accuracy": 0.671875,
|
| 610 |
+
"eval_Qnli-dev_dot_accuracy_threshold": 652.669677734375,
|
| 611 |
+
"eval_Qnli-dev_dot_ap": 0.6928244257428712,
|
| 612 |
+
"eval_Qnli-dev_dot_f1": 0.6923076923076922,
|
| 613 |
+
"eval_Qnli-dev_dot_f1_threshold": 571.573486328125,
|
| 614 |
+
"eval_Qnli-dev_dot_precision": 0.5892857142857143,
|
| 615 |
+
"eval_Qnli-dev_dot_recall": 0.8389830508474576,
|
| 616 |
+
"eval_Qnli-dev_euclidean_accuracy": 0.689453125,
|
| 617 |
+
"eval_Qnli-dev_euclidean_accuracy_threshold": 23.253864288330078,
|
| 618 |
+
"eval_Qnli-dev_euclidean_ap": 0.7293266479878985,
|
| 619 |
+
"eval_Qnli-dev_euclidean_f1": 0.6928571428571428,
|
| 620 |
+
"eval_Qnli-dev_euclidean_f1_threshold": 25.396923065185547,
|
| 621 |
+
"eval_Qnli-dev_euclidean_precision": 0.5987654320987654,
|
| 622 |
+
"eval_Qnli-dev_euclidean_recall": 0.8220338983050848,
|
| 623 |
+
"eval_Qnli-dev_manhattan_accuracy": 0.6875,
|
| 624 |
+
"eval_Qnli-dev_manhattan_accuracy_threshold": 704.4737548828125,
|
| 625 |
+
"eval_Qnli-dev_manhattan_ap": 0.7283792835739776,
|
| 626 |
+
"eval_Qnli-dev_manhattan_f1": 0.6925675675675677,
|
| 627 |
+
"eval_Qnli-dev_manhattan_f1_threshold": 820.7425537109375,
|
| 628 |
+
"eval_Qnli-dev_manhattan_precision": 0.5758426966292135,
|
| 629 |
+
"eval_Qnli-dev_manhattan_recall": 0.8686440677966102,
|
| 630 |
+
"eval_Qnli-dev_max_accuracy": 0.689453125,
|
| 631 |
+
"eval_Qnli-dev_max_accuracy_threshold": 704.4737548828125,
|
| 632 |
+
"eval_Qnli-dev_max_ap": 0.7293266479878985,
|
| 633 |
+
"eval_Qnli-dev_max_f1": 0.693103448275862,
|
| 634 |
+
"eval_Qnli-dev_max_f1_threshold": 820.7425537109375,
|
| 635 |
+
"eval_Qnli-dev_max_precision": 0.5987654320987654,
|
| 636 |
+
"eval_Qnli-dev_max_recall": 0.8686440677966102,
|
| 637 |
+
"eval_allNLI-dev_cosine_accuracy": 0.716796875,
|
| 638 |
+
"eval_allNLI-dev_cosine_accuracy_threshold": 0.805169939994812,
|
| 639 |
+
"eval_allNLI-dev_cosine_ap": 0.5991158471614924,
|
| 640 |
+
"eval_allNLI-dev_cosine_f1": 0.6095617529880478,
|
| 641 |
+
"eval_allNLI-dev_cosine_f1_threshold": 0.6292704343795776,
|
| 642 |
+
"eval_allNLI-dev_cosine_precision": 0.46504559270516715,
|
| 643 |
+
"eval_allNLI-dev_cosine_recall": 0.884393063583815,
|
| 644 |
+
"eval_allNLI-dev_dot_accuracy": 0.7265625,
|
| 645 |
+
"eval_allNLI-dev_dot_accuracy_threshold": 685.67724609375,
|
| 646 |
+
"eval_allNLI-dev_dot_ap": 0.6022852154957883,
|
| 647 |
+
"eval_allNLI-dev_dot_f1": 0.603112840466926,
|
| 648 |
+
"eval_allNLI-dev_dot_f1_threshold": 566.2506103515625,
|
| 649 |
+
"eval_allNLI-dev_dot_precision": 0.45454545454545453,
|
| 650 |
+
"eval_allNLI-dev_dot_recall": 0.8959537572254336,
|
| 651 |
+
"eval_allNLI-dev_euclidean_accuracy": 0.71484375,
|
| 652 |
+
"eval_allNLI-dev_euclidean_accuracy_threshold": 20.33218765258789,
|
| 653 |
+
"eval_allNLI-dev_euclidean_ap": 0.5957908065503393,
|
| 654 |
+
"eval_allNLI-dev_euclidean_f1": 0.6140724946695097,
|
| 655 |
+
"eval_allNLI-dev_euclidean_f1_threshold": 25.308191299438477,
|
| 656 |
+
"eval_allNLI-dev_euclidean_precision": 0.4864864864864865,
|
| 657 |
+
"eval_allNLI-dev_euclidean_recall": 0.8323699421965318,
|
| 658 |
+
"eval_allNLI-dev_manhattan_accuracy": 0.716796875,
|
| 659 |
+
"eval_allNLI-dev_manhattan_accuracy_threshold": 638.1854248046875,
|
| 660 |
+
"eval_allNLI-dev_manhattan_ap": 0.5968486499581244,
|
| 661 |
+
"eval_allNLI-dev_manhattan_f1": 0.616052060737527,
|
| 662 |
+
"eval_allNLI-dev_manhattan_f1_threshold": 783.5068359375,
|
| 663 |
+
"eval_allNLI-dev_manhattan_precision": 0.4930555555555556,
|
| 664 |
+
"eval_allNLI-dev_manhattan_recall": 0.8208092485549133,
|
| 665 |
+
"eval_allNLI-dev_max_accuracy": 0.7265625,
|
| 666 |
+
"eval_allNLI-dev_max_accuracy_threshold": 685.67724609375,
|
| 667 |
+
"eval_allNLI-dev_max_ap": 0.6022852154957883,
|
| 668 |
+
"eval_allNLI-dev_max_f1": 0.616052060737527,
|
| 669 |
+
"eval_allNLI-dev_max_f1_threshold": 783.5068359375,
|
| 670 |
+
"eval_allNLI-dev_max_precision": 0.4930555555555556,
|
| 671 |
+
"eval_allNLI-dev_max_recall": 0.8959537572254336,
|
| 672 |
+
"eval_sequential_score": 0.7293266479878985,
|
| 673 |
+
"eval_sts-test_pearson_cosine": 0.9163990296196276,
|
| 674 |
+
"eval_sts-test_pearson_dot": 0.9067925652682547,
|
| 675 |
+
"eval_sts-test_pearson_euclidean": 0.9258813754113262,
|
| 676 |
+
"eval_sts-test_pearson_manhattan": 0.9250575950562571,
|
| 677 |
+
"eval_sts-test_pearson_max": 0.9258813754113262,
|
| 678 |
+
"eval_sts-test_spearman_cosine": 0.9200439127727275,
|
| 679 |
+
"eval_sts-test_spearman_dot": 0.9030931642366772,
|
| 680 |
+
"eval_sts-test_spearman_euclidean": 0.920441080217576,
|
| 681 |
+
"eval_sts-test_spearman_manhattan": 0.9196469690841858,
|
| 682 |
+
"eval_sts-test_spearman_max": 0.920441080217576,
|
| 683 |
+
"eval_vitaminc-pairs_loss": 2.4166860580444336,
|
| 684 |
+
"eval_vitaminc-pairs_runtime": 4.558,
|
| 685 |
+
"eval_vitaminc-pairs_samples_per_second": 28.082,
|
| 686 |
+
"eval_vitaminc-pairs_steps_per_second": 0.219,
|
| 687 |
+
"step": 471
|
| 688 |
+
},
|
| 689 |
+
{
|
| 690 |
+
"epoch": 0.30115089514066495,
|
| 691 |
+
"eval_negation-triplets_loss": 0.6827869415283203,
|
| 692 |
+
"eval_negation-triplets_runtime": 3.2935,
|
| 693 |
+
"eval_negation-triplets_samples_per_second": 38.864,
|
| 694 |
+
"eval_negation-triplets_steps_per_second": 0.304,
|
| 695 |
+
"step": 471
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.30115089514066495,
|
| 699 |
+
"eval_scitail-pairs-pos_loss": 0.0062057748436927795,
|
| 700 |
+
"eval_scitail-pairs-pos_runtime": 2.638,
|
| 701 |
+
"eval_scitail-pairs-pos_samples_per_second": 48.521,
|
| 702 |
+
"eval_scitail-pairs-pos_steps_per_second": 0.379,
|
| 703 |
+
"step": 471
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.30115089514066495,
|
| 707 |
+
"eval_scitail-pairs-qa_loss": 0.0004358906007837504,
|
| 708 |
+
"eval_scitail-pairs-qa_runtime": 2.0989,
|
| 709 |
+
"eval_scitail-pairs-qa_samples_per_second": 60.985,
|
| 710 |
+
"eval_scitail-pairs-qa_steps_per_second": 0.476,
|
| 711 |
+
"step": 471
|
| 712 |
+
},
|
| 713 |
+
{
|
| 714 |
+
"epoch": 0.30115089514066495,
|
| 715 |
+
"eval_xsum-pairs_loss": 0.03286559134721756,
|
| 716 |
+
"eval_xsum-pairs_runtime": 3.0772,
|
| 717 |
+
"eval_xsum-pairs_samples_per_second": 41.596,
|
| 718 |
+
"eval_xsum-pairs_steps_per_second": 0.325,
|
| 719 |
+
"step": 471
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 0.30115089514066495,
|
| 723 |
+
"eval_sciq_pairs_loss": 0.04141838476061821,
|
| 724 |
+
"eval_sciq_pairs_runtime": 4.69,
|
| 725 |
+
"eval_sciq_pairs_samples_per_second": 27.292,
|
| 726 |
+
"eval_sciq_pairs_steps_per_second": 0.213,
|
| 727 |
+
"step": 471
|
| 728 |
+
},
|
| 729 |
+
{
|
| 730 |
+
"epoch": 0.30115089514066495,
|
| 731 |
+
"eval_qasc_pairs_loss": 0.03906266763806343,
|
| 732 |
+
"eval_qasc_pairs_runtime": 2.104,
|
| 733 |
+
"eval_qasc_pairs_samples_per_second": 60.836,
|
| 734 |
+
"eval_qasc_pairs_steps_per_second": 0.475,
|
| 735 |
+
"step": 471
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.30115089514066495,
|
| 739 |
+
"eval_openbookqa_pairs_loss": 0.38474294543266296,
|
| 740 |
+
"eval_openbookqa_pairs_runtime": 2.2386,
|
| 741 |
+
"eval_openbookqa_pairs_samples_per_second": 57.18,
|
| 742 |
+
"eval_openbookqa_pairs_steps_per_second": 0.447,
|
| 743 |
+
"step": 471
|
| 744 |
+
},
|
| 745 |
+
{
|
| 746 |
+
"epoch": 0.30115089514066495,
|
| 747 |
+
"eval_msmarco_pairs_loss": 0.08647920191287994,
|
| 748 |
+
"eval_msmarco_pairs_runtime": 2.2858,
|
| 749 |
+
"eval_msmarco_pairs_samples_per_second": 55.999,
|
| 750 |
+
"eval_msmarco_pairs_steps_per_second": 0.437,
|
| 751 |
+
"step": 471
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.30115089514066495,
|
| 755 |
+
"eval_nq_pairs_loss": 0.10326902568340302,
|
| 756 |
+
"eval_nq_pairs_runtime": 3.1905,
|
| 757 |
+
"eval_nq_pairs_samples_per_second": 40.12,
|
| 758 |
+
"eval_nq_pairs_steps_per_second": 0.313,
|
| 759 |
+
"step": 471
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 0.30115089514066495,
|
| 763 |
+
"eval_trivia_pairs_loss": 0.09608602523803711,
|
| 764 |
+
"eval_trivia_pairs_runtime": 3.256,
|
| 765 |
+
"eval_trivia_pairs_samples_per_second": 39.312,
|
| 766 |
+
"eval_trivia_pairs_steps_per_second": 0.307,
|
| 767 |
+
"step": 471
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.30115089514066495,
|
| 771 |
+
"eval_gooaq_pairs_loss": 0.12804344296455383,
|
| 772 |
+
"eval_gooaq_pairs_runtime": 2.1753,
|
| 773 |
+
"eval_gooaq_pairs_samples_per_second": 58.841,
|
| 774 |
+
"eval_gooaq_pairs_steps_per_second": 0.46,
|
| 775 |
+
"step": 471
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 0.30115089514066495,
|
| 779 |
+
"eval_paws-pos_loss": 0.025196708738803864,
|
| 780 |
+
"eval_paws-pos_runtime": 2.3068,
|
| 781 |
+
"eval_paws-pos_samples_per_second": 55.488,
|
| 782 |
+
"eval_paws-pos_steps_per_second": 0.434,
|
| 783 |
+
"step": 471
|
| 784 |
+
},
|
| 785 |
+
{
|
| 786 |
+
"epoch": 0.30115089514066495,
|
| 787 |
+
"eval_global_dataset_loss": 0.17351965606212616,
|
| 788 |
+
"eval_global_dataset_runtime": 8.6124,
|
| 789 |
+
"eval_global_dataset_samples_per_second": 37.736,
|
| 790 |
+
"eval_global_dataset_steps_per_second": 0.348,
|
| 791 |
+
"step": 471
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 0.3069053708439898,
|
| 795 |
+
"grad_norm": 0.09573666751384735,
|
| 796 |
+
"learning_rate": 1.0031948881789138e-05,
|
| 797 |
+
"loss": 0.1457,
|
| 798 |
+
"step": 480
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 0.3171355498721228,
|
| 802 |
+
"grad_norm": 2.005824565887451,
|
| 803 |
+
"learning_rate": 1.0372736954206603e-05,
|
| 804 |
+
"loss": 0.1086,
|
| 805 |
+
"step": 496
|
| 806 |
+
},
|
| 807 |
+
{
|
| 808 |
+
"epoch": 0.3273657289002558,
|
| 809 |
+
"grad_norm": 10.186065673828125,
|
| 810 |
+
"learning_rate": 1.071352502662407e-05,
|
| 811 |
+
"loss": 0.1412,
|
| 812 |
+
"step": 512
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.3375959079283887,
|
| 816 |
+
"grad_norm": 29.2017822265625,
|
| 817 |
+
"learning_rate": 1.1054313099041534e-05,
|
| 818 |
+
"loss": 0.1538,
|
| 819 |
+
"step": 528
|
| 820 |
+
},
|
| 821 |
+
{
|
| 822 |
+
"epoch": 0.34782608695652173,
|
| 823 |
+
"grad_norm": 7.944432735443115,
|
| 824 |
+
"learning_rate": 1.1395101171459e-05,
|
| 825 |
+
"loss": 0.1013,
|
| 826 |
+
"step": 544
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"epoch": 0.35805626598465473,
|
| 830 |
+
"grad_norm": 5.722199440002441,
|
| 831 |
+
"learning_rate": 1.1735889243876466e-05,
|
| 832 |
+
"loss": 0.1007,
|
| 833 |
+
"step": 560
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.36828644501278773,
|
| 837 |
+
"grad_norm": 10.10983943939209,
|
| 838 |
+
"learning_rate": 1.207667731629393e-05,
|
| 839 |
+
"loss": 0.0853,
|
| 840 |
+
"step": 576
|
| 841 |
+
},
|
| 842 |
+
{
|
| 843 |
+
"epoch": 0.37851662404092073,
|
| 844 |
+
"grad_norm": 0.017955342307686806,
|
| 845 |
+
"learning_rate": 1.2417465388711395e-05,
|
| 846 |
+
"loss": 0.0696,
|
| 847 |
+
"step": 592
|
| 848 |
+
},
|
| 849 |
+
{
|
| 850 |
+
"epoch": 0.3887468030690537,
|
| 851 |
+
"grad_norm": 87.81786346435547,
|
| 852 |
+
"learning_rate": 1.2758253461128862e-05,
|
| 853 |
+
"loss": 0.1468,
|
| 854 |
+
"step": 608
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.3989769820971867,
|
| 858 |
+
"grad_norm": 4.879648208618164,
|
| 859 |
+
"learning_rate": 1.3099041533546326e-05,
|
| 860 |
+
"loss": 0.1314,
|
| 861 |
+
"step": 624
|
| 862 |
+
},
|
| 863 |
+
{
|
| 864 |
+
"epoch": 0.40153452685422,
|
| 865 |
+
"eval_Qnli-dev_cosine_accuracy": 0.69140625,
|
| 866 |
+
"eval_Qnli-dev_cosine_accuracy_threshold": 0.723533034324646,
|
| 867 |
+
"eval_Qnli-dev_cosine_ap": 0.7379282328356602,
|
| 868 |
+
"eval_Qnli-dev_cosine_f1": 0.68824531516184,
|
| 869 |
+
"eval_Qnli-dev_cosine_f1_threshold": 0.6101292967796326,
|
| 870 |
+
"eval_Qnli-dev_cosine_precision": 0.5754985754985755,
|
| 871 |
+
"eval_Qnli-dev_cosine_recall": 0.8559322033898306,
|
| 872 |
+
"eval_Qnli-dev_dot_accuracy": 0.67578125,
|
| 873 |
+
"eval_Qnli-dev_dot_accuracy_threshold": 669.32177734375,
|
| 874 |
+
"eval_Qnli-dev_dot_ap": 0.7085572115324772,
|
| 875 |
+
"eval_Qnli-dev_dot_f1": 0.6891651865008882,
|
| 876 |
+
"eval_Qnli-dev_dot_f1_threshold": 557.8812255859375,
|
| 877 |
+
"eval_Qnli-dev_dot_precision": 0.5932721712538226,
|
| 878 |
+
"eval_Qnli-dev_dot_recall": 0.8220338983050848,
|
| 879 |
+
"eval_Qnli-dev_euclidean_accuracy": 0.69921875,
|
| 880 |
+
"eval_Qnli-dev_euclidean_accuracy_threshold": 24.064918518066406,
|
| 881 |
+
"eval_Qnli-dev_euclidean_ap": 0.7427699086648181,
|
| 882 |
+
"eval_Qnli-dev_euclidean_f1": 0.6867469879518073,
|
| 883 |
+
"eval_Qnli-dev_euclidean_f1_threshold": 24.312030792236328,
|
| 884 |
+
"eval_Qnli-dev_euclidean_precision": 0.6526717557251909,
|
| 885 |
+
"eval_Qnli-dev_euclidean_recall": 0.7245762711864406,
|
| 886 |
+
"eval_Qnli-dev_manhattan_accuracy": 0.697265625,
|
| 887 |
+
"eval_Qnli-dev_manhattan_accuracy_threshold": 756.5322875976562,
|
| 888 |
+
"eval_Qnli-dev_manhattan_ap": 0.7423808621909044,
|
| 889 |
+
"eval_Qnli-dev_manhattan_f1": 0.6892655367231638,
|
| 890 |
+
"eval_Qnli-dev_manhattan_f1_threshold": 785.015380859375,
|
| 891 |
+
"eval_Qnli-dev_manhattan_precision": 0.6203389830508474,
|
| 892 |
+
"eval_Qnli-dev_manhattan_recall": 0.7754237288135594,
|
| 893 |
+
"eval_Qnli-dev_max_accuracy": 0.69921875,
|
| 894 |
+
"eval_Qnli-dev_max_accuracy_threshold": 756.5322875976562,
|
| 895 |
+
"eval_Qnli-dev_max_ap": 0.7427699086648181,
|
| 896 |
+
"eval_Qnli-dev_max_f1": 0.6892655367231638,
|
| 897 |
+
"eval_Qnli-dev_max_f1_threshold": 785.015380859375,
|
| 898 |
+
"eval_Qnli-dev_max_precision": 0.6526717557251909,
|
| 899 |
+
"eval_Qnli-dev_max_recall": 0.8559322033898306,
|
| 900 |
+
"eval_allNLI-dev_cosine_accuracy": 0.716796875,
|
| 901 |
+
"eval_allNLI-dev_cosine_accuracy_threshold": 0.822735607624054,
|
| 902 |
+
"eval_allNLI-dev_cosine_ap": 0.6013654435047096,
|
| 903 |
+
"eval_allNLI-dev_cosine_f1": 0.6241758241758242,
|
| 904 |
+
"eval_allNLI-dev_cosine_f1_threshold": 0.6583813428878784,
|
| 905 |
+
"eval_allNLI-dev_cosine_precision": 0.5035460992907801,
|
| 906 |
+
"eval_allNLI-dev_cosine_recall": 0.8208092485549133,
|
| 907 |
+
"eval_allNLI-dev_dot_accuracy": 0.71484375,
|
| 908 |
+
"eval_allNLI-dev_dot_accuracy_threshold": 780.966796875,
|
| 909 |
+
"eval_allNLI-dev_dot_ap": 0.6035715304630727,
|
| 910 |
+
"eval_allNLI-dev_dot_f1": 0.6181015452538632,
|
| 911 |
+
"eval_allNLI-dev_dot_f1_threshold": 605.0504150390625,
|
| 912 |
+
"eval_allNLI-dev_dot_precision": 0.5,
|
| 913 |
+
"eval_allNLI-dev_dot_recall": 0.8092485549132948,
|
| 914 |
+
"eval_allNLI-dev_euclidean_accuracy": 0.71875,
|
| 915 |
+
"eval_allNLI-dev_euclidean_accuracy_threshold": 19.111133575439453,
|
| 916 |
+
"eval_allNLI-dev_euclidean_ap": 0.5987461911075059,
|
| 917 |
+
"eval_allNLI-dev_euclidean_f1": 0.6140724946695097,
|
| 918 |
+
"eval_allNLI-dev_euclidean_f1_threshold": 25.491098403930664,
|
| 919 |
+
"eval_allNLI-dev_euclidean_precision": 0.4864864864864865,
|
| 920 |
+
"eval_allNLI-dev_euclidean_recall": 0.8323699421965318,
|
| 921 |
+
"eval_allNLI-dev_manhattan_accuracy": 0.71875,
|
| 922 |
+
"eval_allNLI-dev_manhattan_accuracy_threshold": 601.7767333984375,
|
| 923 |
+
"eval_allNLI-dev_manhattan_ap": 0.5998202267289205,
|
| 924 |
+
"eval_allNLI-dev_manhattan_f1": 0.6160919540229884,
|
| 925 |
+
"eval_allNLI-dev_manhattan_f1_threshold": 765.3077392578125,
|
| 926 |
+
"eval_allNLI-dev_manhattan_precision": 0.5114503816793893,
|
| 927 |
+
"eval_allNLI-dev_manhattan_recall": 0.7745664739884393,
|
| 928 |
+
"eval_allNLI-dev_max_accuracy": 0.71875,
|
| 929 |
+
"eval_allNLI-dev_max_accuracy_threshold": 780.966796875,
|
| 930 |
+
"eval_allNLI-dev_max_ap": 0.6035715304630727,
|
| 931 |
+
"eval_allNLI-dev_max_f1": 0.6241758241758242,
|
| 932 |
+
"eval_allNLI-dev_max_f1_threshold": 765.3077392578125,
|
| 933 |
+
"eval_allNLI-dev_max_precision": 0.5114503816793893,
|
| 934 |
+
"eval_allNLI-dev_max_recall": 0.8323699421965318,
|
| 935 |
+
"eval_sequential_score": 0.7427699086648181,
|
| 936 |
+
"eval_sts-test_pearson_cosine": 0.9073197584934252,
|
| 937 |
+
"eval_sts-test_pearson_dot": 0.8968254992482173,
|
| 938 |
+
"eval_sts-test_pearson_euclidean": 0.9181231094029128,
|
| 939 |
+
"eval_sts-test_pearson_manhattan": 0.9171411172664841,
|
| 940 |
+
"eval_sts-test_pearson_max": 0.9181231094029128,
|
| 941 |
+
"eval_sts-test_spearman_cosine": 0.9153029193896612,
|
| 942 |
+
"eval_sts-test_spearman_dot": 0.897675084268759,
|
| 943 |
+
"eval_sts-test_spearman_euclidean": 0.9150786708188288,
|
| 944 |
+
"eval_sts-test_spearman_manhattan": 0.9142644663690737,
|
| 945 |
+
"eval_sts-test_spearman_max": 0.9153029193896612,
|
| 946 |
+
"eval_vitaminc-pairs_loss": 3.235485076904297,
|
| 947 |
+
"eval_vitaminc-pairs_runtime": 4.4491,
|
| 948 |
+
"eval_vitaminc-pairs_samples_per_second": 28.77,
|
| 949 |
+
"eval_vitaminc-pairs_steps_per_second": 0.225,
|
| 950 |
+
"step": 628
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.40153452685422,
|
| 954 |
+
"eval_negation-triplets_loss": 0.6763876080513,
|
| 955 |
+
"eval_negation-triplets_runtime": 3.2743,
|
| 956 |
+
"eval_negation-triplets_samples_per_second": 39.092,
|
| 957 |
+
"eval_negation-triplets_steps_per_second": 0.305,
|
| 958 |
+
"step": 628
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 0.40153452685422,
|
| 962 |
+
"eval_scitail-pairs-pos_loss": 0.009911377914249897,
|
| 963 |
+
"eval_scitail-pairs-pos_runtime": 2.585,
|
| 964 |
+
"eval_scitail-pairs-pos_samples_per_second": 49.516,
|
| 965 |
+
"eval_scitail-pairs-pos_steps_per_second": 0.387,
|
| 966 |
+
"step": 628
|
| 967 |
+
},
|
| 968 |
+
{
|
| 969 |
+
"epoch": 0.40153452685422,
|
| 970 |
+
"eval_scitail-pairs-qa_loss": 0.0001384670613333583,
|
| 971 |
+
"eval_scitail-pairs-qa_runtime": 2.0603,
|
| 972 |
+
"eval_scitail-pairs-qa_samples_per_second": 62.128,
|
| 973 |
+
"eval_scitail-pairs-qa_steps_per_second": 0.485,
|
| 974 |
+
"step": 628
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.40153452685422,
|
| 978 |
+
"eval_xsum-pairs_loss": 0.017213236540555954,
|
| 979 |
+
"eval_xsum-pairs_runtime": 3.042,
|
| 980 |
+
"eval_xsum-pairs_samples_per_second": 42.078,
|
| 981 |
+
"eval_xsum-pairs_steps_per_second": 0.329,
|
| 982 |
+
"step": 628
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.40153452685422,
|
| 986 |
+
"eval_sciq_pairs_loss": 0.029452303424477577,
|
| 987 |
+
"eval_sciq_pairs_runtime": 4.5793,
|
| 988 |
+
"eval_sciq_pairs_samples_per_second": 27.952,
|
| 989 |
+
"eval_sciq_pairs_steps_per_second": 0.218,
|
| 990 |
+
"step": 628
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 0.40153452685422,
|
| 994 |
+
"eval_qasc_pairs_loss": 0.0720791220664978,
|
| 995 |
+
"eval_qasc_pairs_runtime": 2.0524,
|
| 996 |
+
"eval_qasc_pairs_samples_per_second": 62.366,
|
| 997 |
+
"eval_qasc_pairs_steps_per_second": 0.487,
|
| 998 |
+
"step": 628
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.40153452685422,
|
| 1002 |
+
"eval_openbookqa_pairs_loss": 0.3721720278263092,
|
| 1003 |
+
"eval_openbookqa_pairs_runtime": 2.1848,
|
| 1004 |
+
"eval_openbookqa_pairs_samples_per_second": 58.586,
|
| 1005 |
+
"eval_openbookqa_pairs_steps_per_second": 0.458,
|
| 1006 |
+
"step": 628
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.40153452685422,
|
| 1010 |
+
"eval_msmarco_pairs_loss": 0.09755530953407288,
|
| 1011 |
+
"eval_msmarco_pairs_runtime": 2.2104,
|
| 1012 |
+
"eval_msmarco_pairs_samples_per_second": 57.907,
|
| 1013 |
+
"eval_msmarco_pairs_steps_per_second": 0.452,
|
| 1014 |
+
"step": 628
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 0.40153452685422,
|
| 1018 |
+
"eval_nq_pairs_loss": 0.08931688219308853,
|
| 1019 |
+
"eval_nq_pairs_runtime": 3.1702,
|
| 1020 |
+
"eval_nq_pairs_samples_per_second": 40.376,
|
| 1021 |
+
"eval_nq_pairs_steps_per_second": 0.315,
|
| 1022 |
+
"step": 628
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 0.40153452685422,
|
| 1026 |
+
"eval_trivia_pairs_loss": 0.0646594688296318,
|
| 1027 |
+
"eval_trivia_pairs_runtime": 3.2357,
|
| 1028 |
+
"eval_trivia_pairs_samples_per_second": 39.558,
|
| 1029 |
+
"eval_trivia_pairs_steps_per_second": 0.309,
|
| 1030 |
+
"step": 628
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 0.40153452685422,
|
| 1034 |
+
"eval_gooaq_pairs_loss": 0.1432056725025177,
|
| 1035 |
+
"eval_gooaq_pairs_runtime": 2.1108,
|
| 1036 |
+
"eval_gooaq_pairs_samples_per_second": 60.641,
|
| 1037 |
+
"eval_gooaq_pairs_steps_per_second": 0.474,
|
| 1038 |
+
"step": 628
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.40153452685422,
|
| 1042 |
+
"eval_paws-pos_loss": 0.02637900970876217,
|
| 1043 |
+
"eval_paws-pos_runtime": 2.1958,
|
| 1044 |
+
"eval_paws-pos_samples_per_second": 58.294,
|
| 1045 |
+
"eval_paws-pos_steps_per_second": 0.455,
|
| 1046 |
+
"step": 628
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.40153452685422,
|
| 1050 |
+
"eval_global_dataset_loss": 0.2361193746328354,
|
| 1051 |
+
"eval_global_dataset_runtime": 8.5048,
|
| 1052 |
+
"eval_global_dataset_samples_per_second": 38.214,
|
| 1053 |
+
"eval_global_dataset_steps_per_second": 0.353,
|
| 1054 |
+
"step": 628
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 0.4092071611253197,
|
| 1058 |
+
"grad_norm": 10.842616081237793,
|
| 1059 |
+
"learning_rate": 1.3439829605963792e-05,
|
| 1060 |
+
"loss": 0.149,
|
| 1061 |
+
"step": 640
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 0.4194373401534527,
|
| 1065 |
+
"grad_norm": 10.382250785827637,
|
| 1066 |
+
"learning_rate": 1.3780617678381259e-05,
|
| 1067 |
+
"loss": 0.1402,
|
| 1068 |
+
"step": 656
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 0.4296675191815857,
|
| 1072 |
+
"grad_norm": 11.57950496673584,
|
| 1073 |
+
"learning_rate": 1.4121405750798722e-05,
|
| 1074 |
+
"loss": 0.1056,
|
| 1075 |
+
"step": 672
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 0.4398976982097187,
|
| 1079 |
+
"grad_norm": 5.777312278747559,
|
| 1080 |
+
"learning_rate": 1.4462193823216188e-05,
|
| 1081 |
+
"loss": 0.0932,
|
| 1082 |
+
"step": 688
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.45012787723785164,
|
| 1086 |
+
"grad_norm": 0.2746630311012268,
|
| 1087 |
+
"learning_rate": 1.4802981895633655e-05,
|
| 1088 |
+
"loss": 0.0534,
|
| 1089 |
+
"step": 704
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.46035805626598464,
|
| 1093 |
+
"grad_norm": 0.021142609417438507,
|
| 1094 |
+
"learning_rate": 1.5143769968051119e-05,
|
| 1095 |
+
"loss": 0.2175,
|
| 1096 |
+
"step": 720
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 0.47058823529411764,
|
| 1100 |
+
"grad_norm": 0.00019583643006626517,
|
| 1101 |
+
"learning_rate": 1.5484558040468584e-05,
|
| 1102 |
+
"loss": 0.1107,
|
| 1103 |
+
"step": 736
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 0.48081841432225064,
|
| 1107 |
+
"grad_norm": 18.342988967895508,
|
| 1108 |
+
"learning_rate": 1.582534611288605e-05,
|
| 1109 |
+
"loss": 0.2301,
|
| 1110 |
+
"step": 752
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 0.49104859335038364,
|
| 1114 |
+
"grad_norm": 4.310429096221924,
|
| 1115 |
+
"learning_rate": 1.6166134185303515e-05,
|
| 1116 |
+
"loss": 0.2317,
|
| 1117 |
+
"step": 768
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 0.5012787723785166,
|
| 1121 |
+
"grad_norm": 4.9908766746521,
|
| 1122 |
+
"learning_rate": 1.6506922257720982e-05,
|
| 1123 |
+
"loss": 0.1084,
|
| 1124 |
+
"step": 784
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.5019181585677749,
|
| 1128 |
+
"eval_Qnli-dev_cosine_accuracy": 0.69921875,
|
| 1129 |
+
"eval_Qnli-dev_cosine_accuracy_threshold": 0.692398190498352,
|
| 1130 |
+
"eval_Qnli-dev_cosine_ap": 0.7373451033369327,
|
| 1131 |
+
"eval_Qnli-dev_cosine_f1": 0.7100175746924429,
|
| 1132 |
+
"eval_Qnli-dev_cosine_f1_threshold": 0.6075993776321411,
|
| 1133 |
+
"eval_Qnli-dev_cosine_precision": 0.6066066066066066,
|
| 1134 |
+
"eval_Qnli-dev_cosine_recall": 0.8559322033898306,
|
| 1135 |
+
"eval_Qnli-dev_dot_accuracy": 0.68359375,
|
| 1136 |
+
"eval_Qnli-dev_dot_accuracy_threshold": 593.7620849609375,
|
| 1137 |
+
"eval_Qnli-dev_dot_ap": 0.7049377659461418,
|
| 1138 |
+
"eval_Qnli-dev_dot_f1": 0.7008849557522123,
|
| 1139 |
+
"eval_Qnli-dev_dot_f1_threshold": 556.5305786132812,
|
| 1140 |
+
"eval_Qnli-dev_dot_precision": 0.601823708206687,
|
| 1141 |
+
"eval_Qnli-dev_dot_recall": 0.8389830508474576,
|
| 1142 |
+
"eval_Qnli-dev_euclidean_accuracy": 0.705078125,
|
| 1143 |
+
"eval_Qnli-dev_euclidean_accuracy_threshold": 24.380191802978516,
|
| 1144 |
+
"eval_Qnli-dev_euclidean_ap": 0.7421878778893789,
|
| 1145 |
+
"eval_Qnli-dev_euclidean_f1": 0.7024952015355086,
|
| 1146 |
+
"eval_Qnli-dev_euclidean_f1_threshold": 25.691165924072266,
|
| 1147 |
+
"eval_Qnli-dev_euclidean_precision": 0.6421052631578947,
|
| 1148 |
+
"eval_Qnli-dev_euclidean_recall": 0.7754237288135594,
|
| 1149 |
+
"eval_Qnli-dev_manhattan_accuracy": 0.705078125,
|
| 1150 |
+
"eval_Qnli-dev_manhattan_accuracy_threshold": 755.8370361328125,
|
| 1151 |
+
"eval_Qnli-dev_manhattan_ap": 0.7420425528192431,
|
| 1152 |
+
"eval_Qnli-dev_manhattan_f1": 0.7012987012987013,
|
| 1153 |
+
"eval_Qnli-dev_manhattan_f1_threshold": 872.5838623046875,
|
| 1154 |
+
"eval_Qnli-dev_manhattan_precision": 0.5684210526315789,
|
| 1155 |
+
"eval_Qnli-dev_manhattan_recall": 0.9152542372881356,
|
| 1156 |
+
"eval_Qnli-dev_max_accuracy": 0.705078125,
|
| 1157 |
+
"eval_Qnli-dev_max_accuracy_threshold": 755.8370361328125,
|
| 1158 |
+
"eval_Qnli-dev_max_ap": 0.7421878778893789,
|
| 1159 |
+
"eval_Qnli-dev_max_f1": 0.7100175746924429,
|
| 1160 |
+
"eval_Qnli-dev_max_f1_threshold": 872.5838623046875,
|
| 1161 |
+
"eval_Qnli-dev_max_precision": 0.6421052631578947,
|
| 1162 |
+
"eval_Qnli-dev_max_recall": 0.9152542372881356,
|
| 1163 |
+
"eval_allNLI-dev_cosine_accuracy": 0.72265625,
|
| 1164 |
+
"eval_allNLI-dev_cosine_accuracy_threshold": 0.8044645190238953,
|
| 1165 |
+
"eval_allNLI-dev_cosine_ap": 0.6029077511537212,
|
| 1166 |
+
"eval_allNLI-dev_cosine_f1": 0.6193548387096773,
|
| 1167 |
+
"eval_allNLI-dev_cosine_f1_threshold": 0.6631404161453247,
|
| 1168 |
+
"eval_allNLI-dev_cosine_precision": 0.4931506849315068,
|
| 1169 |
+
"eval_allNLI-dev_cosine_recall": 0.8323699421965318,
|
| 1170 |
+
"eval_allNLI-dev_dot_accuracy": 0.716796875,
|
| 1171 |
+
"eval_allNLI-dev_dot_accuracy_threshold": 788.6072998046875,
|
| 1172 |
+
"eval_allNLI-dev_dot_ap": 0.5937600770830995,
|
| 1173 |
+
"eval_allNLI-dev_dot_f1": 0.6299559471365639,
|
| 1174 |
+
"eval_allNLI-dev_dot_f1_threshold": 632.650634765625,
|
| 1175 |
+
"eval_allNLI-dev_dot_precision": 0.5088967971530249,
|
| 1176 |
+
"eval_allNLI-dev_dot_recall": 0.8265895953757225,
|
| 1177 |
+
"eval_allNLI-dev_euclidean_accuracy": 0.71875,
|
| 1178 |
+
"eval_allNLI-dev_euclidean_accuracy_threshold": 19.220054626464844,
|
| 1179 |
+
"eval_allNLI-dev_euclidean_ap": 0.6000308340748244,
|
| 1180 |
+
"eval_allNLI-dev_euclidean_f1": 0.6140350877192983,
|
| 1181 |
+
"eval_allNLI-dev_euclidean_f1_threshold": 24.999332427978516,
|
| 1182 |
+
"eval_allNLI-dev_euclidean_precision": 0.49469964664310956,
|
| 1183 |
+
"eval_allNLI-dev_euclidean_recall": 0.8092485549132948,
|
| 1184 |
+
"eval_allNLI-dev_manhattan_accuracy": 0.724609375,
|
| 1185 |
+
"eval_allNLI-dev_manhattan_accuracy_threshold": 601.0503540039062,
|
| 1186 |
+
"eval_allNLI-dev_manhattan_ap": 0.5994463454270988,
|
| 1187 |
+
"eval_allNLI-dev_manhattan_f1": 0.6137339055793991,
|
| 1188 |
+
"eval_allNLI-dev_manhattan_f1_threshold": 785.221435546875,
|
| 1189 |
+
"eval_allNLI-dev_manhattan_precision": 0.4880546075085324,
|
| 1190 |
+
"eval_allNLI-dev_manhattan_recall": 0.8265895953757225,
|
| 1191 |
+
"eval_allNLI-dev_max_accuracy": 0.724609375,
|
| 1192 |
+
"eval_allNLI-dev_max_accuracy_threshold": 788.6072998046875,
|
| 1193 |
+
"eval_allNLI-dev_max_ap": 0.6029077511537212,
|
| 1194 |
+
"eval_allNLI-dev_max_f1": 0.6299559471365639,
|
| 1195 |
+
"eval_allNLI-dev_max_f1_threshold": 785.221435546875,
|
| 1196 |
+
"eval_allNLI-dev_max_precision": 0.5088967971530249,
|
| 1197 |
+
"eval_allNLI-dev_max_recall": 0.8323699421965318,
|
| 1198 |
+
"eval_sequential_score": 0.7421878778893789,
|
| 1199 |
+
"eval_sts-test_pearson_cosine": 0.9043291123731831,
|
| 1200 |
+
"eval_sts-test_pearson_dot": 0.8879215841003365,
|
| 1201 |
+
"eval_sts-test_pearson_euclidean": 0.9229060512014216,
|
| 1202 |
+
"eval_sts-test_pearson_manhattan": 0.9223094827934384,
|
| 1203 |
+
"eval_sts-test_pearson_max": 0.9229060512014216,
|
| 1204 |
+
"eval_sts-test_spearman_cosine": 0.9171074245027466,
|
| 1205 |
+
"eval_sts-test_spearman_dot": 0.8910804041363384,
|
| 1206 |
+
"eval_sts-test_spearman_euclidean": 0.9186099481039627,
|
| 1207 |
+
"eval_sts-test_spearman_manhattan": 0.9180492147987285,
|
| 1208 |
+
"eval_sts-test_spearman_max": 0.9186099481039627,
|
| 1209 |
+
"eval_vitaminc-pairs_loss": 3.900864362716675,
|
| 1210 |
+
"eval_vitaminc-pairs_runtime": 4.4894,
|
| 1211 |
+
"eval_vitaminc-pairs_samples_per_second": 28.512,
|
| 1212 |
+
"eval_vitaminc-pairs_steps_per_second": 0.223,
|
| 1213 |
+
"step": 785
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 0.5019181585677749,
|
| 1217 |
+
"eval_negation-triplets_loss": 0.633728563785553,
|
| 1218 |
+
"eval_negation-triplets_runtime": 3.4005,
|
| 1219 |
+
"eval_negation-triplets_samples_per_second": 37.641,
|
| 1220 |
+
"eval_negation-triplets_steps_per_second": 0.294,
|
| 1221 |
+
"step": 785
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.5019181585677749,
|
| 1225 |
+
"eval_scitail-pairs-pos_loss": 0.012064059264957905,
|
| 1226 |
+
"eval_scitail-pairs-pos_runtime": 2.6615,
|
| 1227 |
+
"eval_scitail-pairs-pos_samples_per_second": 48.092,
|
| 1228 |
+
"eval_scitail-pairs-pos_steps_per_second": 0.376,
|
| 1229 |
+
"step": 785
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 0.5019181585677749,
|
| 1233 |
+
"eval_scitail-pairs-qa_loss": 7.389660459011793e-05,
|
| 1234 |
+
"eval_scitail-pairs-qa_runtime": 2.0934,
|
| 1235 |
+
"eval_scitail-pairs-qa_samples_per_second": 61.146,
|
| 1236 |
+
"eval_scitail-pairs-qa_steps_per_second": 0.478,
|
| 1237 |
+
"step": 785
|
| 1238 |
+
},
|
| 1239 |
+
{
|
| 1240 |
+
"epoch": 0.5019181585677749,
|
| 1241 |
+
"eval_xsum-pairs_loss": 0.009806518442928791,
|
| 1242 |
+
"eval_xsum-pairs_runtime": 3.0655,
|
| 1243 |
+
"eval_xsum-pairs_samples_per_second": 41.755,
|
| 1244 |
+
"eval_xsum-pairs_steps_per_second": 0.326,
|
| 1245 |
+
"step": 785
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.5019181585677749,
|
| 1249 |
+
"eval_sciq_pairs_loss": 0.03242116793990135,
|
| 1250 |
+
"eval_sciq_pairs_runtime": 4.6405,
|
| 1251 |
+
"eval_sciq_pairs_samples_per_second": 27.583,
|
| 1252 |
+
"eval_sciq_pairs_steps_per_second": 0.215,
|
| 1253 |
+
"step": 785
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 0.5019181585677749,
|
| 1257 |
+
"eval_qasc_pairs_loss": 0.07289526611566544,
|
| 1258 |
+
"eval_qasc_pairs_runtime": 2.0997,
|
| 1259 |
+
"eval_qasc_pairs_samples_per_second": 60.962,
|
| 1260 |
+
"eval_qasc_pairs_steps_per_second": 0.476,
|
| 1261 |
+
"step": 785
|
| 1262 |
+
},
|
| 1263 |
+
{
|
| 1264 |
+
"epoch": 0.5019181585677749,
|
| 1265 |
+
"eval_openbookqa_pairs_loss": 0.41281530261039734,
|
| 1266 |
+
"eval_openbookqa_pairs_runtime": 2.2362,
|
| 1267 |
+
"eval_openbookqa_pairs_samples_per_second": 57.24,
|
| 1268 |
+
"eval_openbookqa_pairs_steps_per_second": 0.447,
|
| 1269 |
+
"step": 785
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 0.5019181585677749,
|
| 1273 |
+
"eval_msmarco_pairs_loss": 0.11559046804904938,
|
| 1274 |
+
"eval_msmarco_pairs_runtime": 2.2303,
|
| 1275 |
+
"eval_msmarco_pairs_samples_per_second": 57.393,
|
| 1276 |
+
"eval_msmarco_pairs_steps_per_second": 0.448,
|
| 1277 |
+
"step": 785
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 0.5019181585677749,
|
| 1281 |
+
"eval_nq_pairs_loss": 0.10707583278417587,
|
| 1282 |
+
"eval_nq_pairs_runtime": 3.2248,
|
| 1283 |
+
"eval_nq_pairs_samples_per_second": 39.693,
|
| 1284 |
+
"eval_nq_pairs_steps_per_second": 0.31,
|
| 1285 |
+
"step": 785
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 0.5019181585677749,
|
| 1289 |
+
"eval_trivia_pairs_loss": 0.09321507811546326,
|
| 1290 |
+
"eval_trivia_pairs_runtime": 3.2802,
|
| 1291 |
+
"eval_trivia_pairs_samples_per_second": 39.022,
|
| 1292 |
+
"eval_trivia_pairs_steps_per_second": 0.305,
|
| 1293 |
+
"step": 785
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 0.5019181585677749,
|
| 1297 |
+
"eval_gooaq_pairs_loss": 0.1236836165189743,
|
| 1298 |
+
"eval_gooaq_pairs_runtime": 2.166,
|
| 1299 |
+
"eval_gooaq_pairs_samples_per_second": 59.094,
|
| 1300 |
+
"eval_gooaq_pairs_steps_per_second": 0.462,
|
| 1301 |
+
"step": 785
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 0.5019181585677749,
|
| 1305 |
+
"eval_paws-pos_loss": 0.025379396975040436,
|
| 1306 |
+
"eval_paws-pos_runtime": 2.2622,
|
| 1307 |
+
"eval_paws-pos_samples_per_second": 56.581,
|
| 1308 |
+
"eval_paws-pos_steps_per_second": 0.442,
|
| 1309 |
+
"step": 785
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 0.5019181585677749,
|
| 1313 |
+
"eval_global_dataset_loss": 0.2525935471057892,
|
| 1314 |
+
"eval_global_dataset_runtime": 8.5173,
|
| 1315 |
+
"eval_global_dataset_samples_per_second": 38.158,
|
| 1316 |
+
"eval_global_dataset_steps_per_second": 0.352,
|
| 1317 |
+
"step": 785
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 0.5115089514066496,
|
| 1321 |
+
"grad_norm": 2.781236171722412,
|
| 1322 |
+
"learning_rate": 1.6847710330138445e-05,
|
| 1323 |
+
"loss": 0.0799,
|
| 1324 |
+
"step": 800
|
| 1325 |
+
},
|
| 1326 |
+
{
|
| 1327 |
+
"epoch": 0.5217391304347826,
|
| 1328 |
+
"grad_norm": 4.296014785766602,
|
| 1329 |
+
"learning_rate": 1.7188498402555913e-05,
|
| 1330 |
+
"loss": 0.1466,
|
| 1331 |
+
"step": 816
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 0.5319693094629157,
|
| 1335 |
+
"grad_norm": 0.6779389977455139,
|
| 1336 |
+
"learning_rate": 1.7529286474973376e-05,
|
| 1337 |
+
"loss": 0.1208,
|
| 1338 |
+
"step": 832
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 0.5421994884910486,
|
| 1342 |
+
"grad_norm": 5.031887054443359,
|
| 1343 |
+
"learning_rate": 1.7870074547390843e-05,
|
| 1344 |
+
"loss": 0.1021,
|
| 1345 |
+
"step": 848
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 0.5524296675191815,
|
| 1349 |
+
"grad_norm": 3.492619276046753,
|
| 1350 |
+
"learning_rate": 1.8210862619808307e-05,
|
| 1351 |
+
"loss": 0.1391,
|
| 1352 |
+
"step": 864
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 0.5626598465473146,
|
| 1356 |
+
"grad_norm": 3.4289321899414062,
|
| 1357 |
+
"learning_rate": 1.8551650692225774e-05,
|
| 1358 |
+
"loss": 0.185,
|
| 1359 |
+
"step": 880
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 0.5728900255754475,
|
| 1363 |
+
"grad_norm": 0.48259153962135315,
|
| 1364 |
+
"learning_rate": 1.8892438764643238e-05,
|
| 1365 |
+
"loss": 0.1108,
|
| 1366 |
+
"step": 896
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"epoch": 0.5831202046035806,
|
| 1370 |
+
"grad_norm": 1.390660047531128,
|
| 1371 |
+
"learning_rate": 1.9233226837060705e-05,
|
| 1372 |
+
"loss": 0.0926,
|
| 1373 |
+
"step": 912
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 0.5933503836317136,
|
| 1377 |
+
"grad_norm": 11.370463371276855,
|
| 1378 |
+
"learning_rate": 1.957401490947817e-05,
|
| 1379 |
+
"loss": 0.157,
|
| 1380 |
+
"step": 928
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 0.6023017902813299,
|
| 1384 |
+
"eval_Qnli-dev_cosine_accuracy": 0.6796875,
|
| 1385 |
+
"eval_Qnli-dev_cosine_accuracy_threshold": 0.7110848426818848,
|
| 1386 |
+
"eval_Qnli-dev_cosine_ap": 0.7098259706381642,
|
| 1387 |
+
"eval_Qnli-dev_cosine_f1": 0.6848,
|
| 1388 |
+
"eval_Qnli-dev_cosine_f1_threshold": 0.5396614074707031,
|
| 1389 |
+
"eval_Qnli-dev_cosine_precision": 0.5501285347043702,
|
| 1390 |
+
"eval_Qnli-dev_cosine_recall": 0.9067796610169492,
|
| 1391 |
+
"eval_Qnli-dev_dot_accuracy": 0.66015625,
|
| 1392 |
+
"eval_Qnli-dev_dot_accuracy_threshold": 612.0003662109375,
|
| 1393 |
+
"eval_Qnli-dev_dot_ap": 0.6677871203032961,
|
| 1394 |
+
"eval_Qnli-dev_dot_f1": 0.6847133757961782,
|
| 1395 |
+
"eval_Qnli-dev_dot_f1_threshold": 489.4056091308594,
|
| 1396 |
+
"eval_Qnli-dev_dot_precision": 0.548469387755102,
|
| 1397 |
+
"eval_Qnli-dev_dot_recall": 0.9110169491525424,
|
| 1398 |
+
"eval_Qnli-dev_euclidean_accuracy": 0.689453125,
|
| 1399 |
+
"eval_Qnli-dev_euclidean_accuracy_threshold": 22.8848876953125,
|
| 1400 |
+
"eval_Qnli-dev_euclidean_ap": 0.7192898440461163,
|
| 1401 |
+
"eval_Qnli-dev_euclidean_f1": 0.6879194630872484,
|
| 1402 |
+
"eval_Qnli-dev_euclidean_f1_threshold": 28.238075256347656,
|
| 1403 |
+
"eval_Qnli-dev_euclidean_precision": 0.5694444444444444,
|
| 1404 |
+
"eval_Qnli-dev_euclidean_recall": 0.8686440677966102,
|
| 1405 |
+
"eval_Qnli-dev_manhattan_accuracy": 0.681640625,
|
| 1406 |
+
"eval_Qnli-dev_manhattan_accuracy_threshold": 722.6270751953125,
|
| 1407 |
+
"eval_Qnli-dev_manhattan_ap": 0.7187658029633445,
|
| 1408 |
+
"eval_Qnli-dev_manhattan_f1": 0.6911519198664441,
|
| 1409 |
+
"eval_Qnli-dev_manhattan_f1_threshold": 881.6773681640625,
|
| 1410 |
+
"eval_Qnli-dev_manhattan_precision": 0.5702479338842975,
|
| 1411 |
+
"eval_Qnli-dev_manhattan_recall": 0.8771186440677966,
|
| 1412 |
+
"eval_Qnli-dev_max_accuracy": 0.689453125,
|
| 1413 |
+
"eval_Qnli-dev_max_accuracy_threshold": 722.6270751953125,
|
| 1414 |
+
"eval_Qnli-dev_max_ap": 0.7192898440461163,
|
| 1415 |
+
"eval_Qnli-dev_max_f1": 0.6911519198664441,
|
| 1416 |
+
"eval_Qnli-dev_max_f1_threshold": 881.6773681640625,
|
| 1417 |
+
"eval_Qnli-dev_max_precision": 0.5702479338842975,
|
| 1418 |
+
"eval_Qnli-dev_max_recall": 0.9110169491525424,
|
| 1419 |
+
"eval_allNLI-dev_cosine_accuracy": 0.720703125,
|
| 1420 |
+
"eval_allNLI-dev_cosine_accuracy_threshold": 0.8192055225372314,
|
| 1421 |
+
"eval_allNLI-dev_cosine_ap": 0.6064453769503819,
|
| 1422 |
+
"eval_allNLI-dev_cosine_f1": 0.6255924170616114,
|
| 1423 |
+
"eval_allNLI-dev_cosine_f1_threshold": 0.6535474061965942,
|
| 1424 |
+
"eval_allNLI-dev_cosine_precision": 0.5301204819277109,
|
| 1425 |
+
"eval_allNLI-dev_cosine_recall": 0.7630057803468208,
|
| 1426 |
+
"eval_allNLI-dev_dot_accuracy": 0.71875,
|
| 1427 |
+
"eval_allNLI-dev_dot_accuracy_threshold": 818.2044067382812,
|
| 1428 |
+
"eval_allNLI-dev_dot_ap": 0.6065404649281851,
|
| 1429 |
+
"eval_allNLI-dev_dot_f1": 0.6232558139534884,
|
| 1430 |
+
"eval_allNLI-dev_dot_f1_threshold": 617.462158203125,
|
| 1431 |
+
"eval_allNLI-dev_dot_precision": 0.5214007782101168,
|
| 1432 |
+
"eval_allNLI-dev_dot_recall": 0.7745664739884393,
|
| 1433 |
+
"eval_allNLI-dev_euclidean_accuracy": 0.71484375,
|
| 1434 |
+
"eval_allNLI-dev_euclidean_accuracy_threshold": 18.268875122070312,
|
| 1435 |
+
"eval_allNLI-dev_euclidean_ap": 0.6020313861913525,
|
| 1436 |
+
"eval_allNLI-dev_euclidean_f1": 0.6205128205128205,
|
| 1437 |
+
"eval_allNLI-dev_euclidean_f1_threshold": 24.51853370666504,
|
| 1438 |
+
"eval_allNLI-dev_euclidean_precision": 0.5576036866359447,
|
| 1439 |
+
"eval_allNLI-dev_euclidean_recall": 0.6994219653179191,
|
| 1440 |
+
"eval_allNLI-dev_manhattan_accuracy": 0.720703125,
|
| 1441 |
+
"eval_allNLI-dev_manhattan_accuracy_threshold": 748.0535888671875,
|
| 1442 |
+
"eval_allNLI-dev_manhattan_ap": 0.6018544507527778,
|
| 1443 |
+
"eval_allNLI-dev_manhattan_f1": 0.62,
|
| 1444 |
+
"eval_allNLI-dev_manhattan_f1_threshold": 774.2725219726562,
|
| 1445 |
+
"eval_allNLI-dev_manhattan_precision": 0.5462555066079295,
|
| 1446 |
+
"eval_allNLI-dev_manhattan_recall": 0.7167630057803468,
|
| 1447 |
+
"eval_allNLI-dev_max_accuracy": 0.720703125,
|
| 1448 |
+
"eval_allNLI-dev_max_accuracy_threshold": 818.2044067382812,
|
| 1449 |
+
"eval_allNLI-dev_max_ap": 0.6065404649281851,
|
| 1450 |
+
"eval_allNLI-dev_max_f1": 0.6255924170616114,
|
| 1451 |
+
"eval_allNLI-dev_max_f1_threshold": 774.2725219726562,
|
| 1452 |
+
"eval_allNLI-dev_max_precision": 0.5576036866359447,
|
| 1453 |
+
"eval_allNLI-dev_max_recall": 0.7745664739884393,
|
| 1454 |
+
"eval_sequential_score": 0.7192898440461163,
|
| 1455 |
+
"eval_sts-test_pearson_cosine": 0.917151106536728,
|
| 1456 |
+
"eval_sts-test_pearson_dot": 0.9064091595160165,
|
| 1457 |
+
"eval_sts-test_pearson_euclidean": 0.9276686994987783,
|
| 1458 |
+
"eval_sts-test_pearson_manhattan": 0.9272998345544423,
|
| 1459 |
+
"eval_sts-test_pearson_max": 0.9276686994987783,
|
| 1460 |
+
"eval_sts-test_spearman_cosine": 0.9217884511955483,
|
| 1461 |
+
"eval_sts-test_spearman_dot": 0.9049310090394994,
|
| 1462 |
+
"eval_sts-test_spearman_euclidean": 0.9226553278799612,
|
| 1463 |
+
"eval_sts-test_spearman_manhattan": 0.9221132558507271,
|
| 1464 |
+
"eval_sts-test_spearman_max": 0.9226553278799612,
|
| 1465 |
+
"eval_vitaminc-pairs_loss": 2.814775228500366,
|
| 1466 |
+
"eval_vitaminc-pairs_runtime": 4.4847,
|
| 1467 |
+
"eval_vitaminc-pairs_samples_per_second": 28.542,
|
| 1468 |
+
"eval_vitaminc-pairs_steps_per_second": 0.223,
|
| 1469 |
+
"step": 942
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 0.6023017902813299,
|
| 1473 |
+
"eval_negation-triplets_loss": 0.5488928556442261,
|
| 1474 |
+
"eval_negation-triplets_runtime": 3.2998,
|
| 1475 |
+
"eval_negation-triplets_samples_per_second": 38.79,
|
| 1476 |
+
"eval_negation-triplets_steps_per_second": 0.303,
|
| 1477 |
+
"step": 942
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 0.6023017902813299,
|
| 1481 |
+
"eval_scitail-pairs-pos_loss": 0.0049793594516813755,
|
| 1482 |
+
"eval_scitail-pairs-pos_runtime": 2.6898,
|
| 1483 |
+
"eval_scitail-pairs-pos_samples_per_second": 47.588,
|
| 1484 |
+
"eval_scitail-pairs-pos_steps_per_second": 0.372,
|
| 1485 |
+
"step": 942
|
| 1486 |
+
},
|
| 1487 |
+
{
|
| 1488 |
+
"epoch": 0.6023017902813299,
|
| 1489 |
+
"eval_scitail-pairs-qa_loss": 0.0024724495597183704,
|
| 1490 |
+
"eval_scitail-pairs-qa_runtime": 2.0719,
|
| 1491 |
+
"eval_scitail-pairs-qa_samples_per_second": 61.778,
|
| 1492 |
+
"eval_scitail-pairs-qa_steps_per_second": 0.483,
|
| 1493 |
+
"step": 942
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 0.6023017902813299,
|
| 1497 |
+
"eval_xsum-pairs_loss": 0.01770014315843582,
|
| 1498 |
+
"eval_xsum-pairs_runtime": 3.0541,
|
| 1499 |
+
"eval_xsum-pairs_samples_per_second": 41.911,
|
| 1500 |
+
"eval_xsum-pairs_steps_per_second": 0.327,
|
| 1501 |
+
"step": 942
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 0.6023017902813299,
|
| 1505 |
+
"eval_sciq_pairs_loss": 0.030735233798623085,
|
| 1506 |
+
"eval_sciq_pairs_runtime": 4.5905,
|
| 1507 |
+
"eval_sciq_pairs_samples_per_second": 27.883,
|
| 1508 |
+
"eval_sciq_pairs_steps_per_second": 0.218,
|
| 1509 |
+
"step": 942
|
| 1510 |
+
},
|
| 1511 |
+
{
|
| 1512 |
+
"epoch": 0.6023017902813299,
|
| 1513 |
+
"eval_qasc_pairs_loss": 0.04597622901201248,
|
| 1514 |
+
"eval_qasc_pairs_runtime": 2.1506,
|
| 1515 |
+
"eval_qasc_pairs_samples_per_second": 59.518,
|
| 1516 |
+
"eval_qasc_pairs_steps_per_second": 0.465,
|
| 1517 |
+
"step": 942
|
| 1518 |
+
},
|
| 1519 |
+
{
|
| 1520 |
+
"epoch": 0.6023017902813299,
|
| 1521 |
+
"eval_openbookqa_pairs_loss": 0.3580770492553711,
|
| 1522 |
+
"eval_openbookqa_pairs_runtime": 2.2167,
|
| 1523 |
+
"eval_openbookqa_pairs_samples_per_second": 57.743,
|
| 1524 |
+
"eval_openbookqa_pairs_steps_per_second": 0.451,
|
| 1525 |
+
"step": 942
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 0.6023017902813299,
|
| 1529 |
+
"eval_msmarco_pairs_loss": 0.08442261815071106,
|
| 1530 |
+
"eval_msmarco_pairs_runtime": 2.2308,
|
| 1531 |
+
"eval_msmarco_pairs_samples_per_second": 57.377,
|
| 1532 |
+
"eval_msmarco_pairs_steps_per_second": 0.448,
|
| 1533 |
+
"step": 942
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 0.6023017902813299,
|
| 1537 |
+
"eval_nq_pairs_loss": 0.12354034185409546,
|
| 1538 |
+
"eval_nq_pairs_runtime": 3.1805,
|
| 1539 |
+
"eval_nq_pairs_samples_per_second": 40.245,
|
| 1540 |
+
"eval_nq_pairs_steps_per_second": 0.314,
|
| 1541 |
+
"step": 942
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 0.6023017902813299,
|
| 1545 |
+
"eval_trivia_pairs_loss": 0.08584598451852798,
|
| 1546 |
+
"eval_trivia_pairs_runtime": 3.2829,
|
| 1547 |
+
"eval_trivia_pairs_samples_per_second": 38.989,
|
| 1548 |
+
"eval_trivia_pairs_steps_per_second": 0.305,
|
| 1549 |
+
"step": 942
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 0.6023017902813299,
|
| 1553 |
+
"eval_gooaq_pairs_loss": 0.1563483476638794,
|
| 1554 |
+
"eval_gooaq_pairs_runtime": 2.1458,
|
| 1555 |
+
"eval_gooaq_pairs_samples_per_second": 59.651,
|
| 1556 |
+
"eval_gooaq_pairs_steps_per_second": 0.466,
|
| 1557 |
+
"step": 942
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 0.6023017902813299,
|
| 1561 |
+
"eval_paws-pos_loss": 0.026773599907755852,
|
| 1562 |
+
"eval_paws-pos_runtime": 2.2412,
|
| 1563 |
+
"eval_paws-pos_samples_per_second": 57.112,
|
| 1564 |
+
"eval_paws-pos_steps_per_second": 0.446,
|
| 1565 |
+
"step": 942
|
| 1566 |
+
},
|
| 1567 |
+
{
|
| 1568 |
+
"epoch": 0.6023017902813299,
|
| 1569 |
+
"eval_global_dataset_loss": 0.15439902245998383,
|
| 1570 |
+
"eval_global_dataset_runtime": 8.5471,
|
| 1571 |
+
"eval_global_dataset_samples_per_second": 38.025,
|
| 1572 |
+
"eval_global_dataset_steps_per_second": 0.351,
|
| 1573 |
+
"step": 942
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 0.6035805626598465,
|
| 1577 |
+
"grad_norm": 5.923606872558594,
|
| 1578 |
+
"learning_rate": 1.9914802981895636e-05,
|
| 1579 |
+
"loss": 0.126,
|
| 1580 |
+
"step": 944
|
| 1581 |
+
},
|
| 1582 |
+
{
|
| 1583 |
+
"epoch": 0.6138107416879796,
|
| 1584 |
+
"grad_norm": 2.5291152000427246,
|
| 1585 |
+
"learning_rate": 1.9998887778598045e-05,
|
| 1586 |
+
"loss": 0.1416,
|
| 1587 |
+
"step": 960
|
| 1588 |
+
},
|
| 1589 |
+
{
|
| 1590 |
+
"epoch": 0.6240409207161125,
|
| 1591 |
+
"grad_norm": 6.44370174407959,
|
| 1592 |
+
"learning_rate": 1.9993945237540454e-05,
|
| 1593 |
+
"loss": 0.157,
|
| 1594 |
+
"step": 976
|
| 1595 |
+
},
|
| 1596 |
+
{
|
| 1597 |
+
"epoch": 0.6342710997442456,
|
| 1598 |
+
"grad_norm": 5.631255149841309,
|
| 1599 |
+
"learning_rate": 1.9985051399986202e-05,
|
| 1600 |
+
"loss": 0.076,
|
| 1601 |
+
"step": 992
|
| 1602 |
+
},
|
| 1603 |
+
{
|
| 1604 |
+
"epoch": 0.6445012787723785,
|
| 1605 |
+
"grad_norm": 3.087242364883423,
|
| 1606 |
+
"learning_rate": 1.9972210955346275e-05,
|
| 1607 |
+
"loss": 0.0956,
|
| 1608 |
+
"step": 1008
|
| 1609 |
+
},
|
| 1610 |
+
{
|
| 1611 |
+
"epoch": 0.6547314578005116,
|
| 1612 |
+
"grad_norm": 5.553103446960449,
|
| 1613 |
+
"learning_rate": 1.995543067394022e-05,
|
| 1614 |
+
"loss": 0.1297,
|
| 1615 |
+
"step": 1024
|
| 1616 |
+
},
|
| 1617 |
+
{
|
| 1618 |
+
"epoch": 0.6649616368286445,
|
| 1619 |
+
"grad_norm": 4.881236553192139,
|
| 1620 |
+
"learning_rate": 1.9934719403426386e-05,
|
| 1621 |
+
"loss": 0.1673,
|
| 1622 |
+
"step": 1040
|
| 1623 |
+
},
|
| 1624 |
+
{
|
| 1625 |
+
"epoch": 0.6751918158567775,
|
| 1626 |
+
"grad_norm": 9.578516006469727,
|
| 1627 |
+
"learning_rate": 1.9910088064136855e-05,
|
| 1628 |
+
"loss": 0.0801,
|
| 1629 |
+
"step": 1056
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"epoch": 0.6854219948849105,
|
| 1633 |
+
"grad_norm": 2.1746671199798584,
|
| 1634 |
+
"learning_rate": 1.9881549643319548e-05,
|
| 1635 |
+
"loss": 0.1508,
|
| 1636 |
+
"step": 1072
|
| 1637 |
+
},
|
| 1638 |
+
{
|
| 1639 |
+
"epoch": 0.6956521739130435,
|
| 1640 |
+
"grad_norm": 1.0966675281524658,
|
| 1641 |
+
"learning_rate": 1.9849119188290472e-05,
|
| 1642 |
+
"loss": 0.082,
|
| 1643 |
+
"step": 1088
|
| 1644 |
+
},
|
| 1645 |
+
{
|
| 1646 |
+
"epoch": 0.7026854219948849,
|
| 1647 |
+
"eval_Qnli-dev_cosine_accuracy": 0.701171875,
|
| 1648 |
+
"eval_Qnli-dev_cosine_accuracy_threshold": 0.6949167251586914,
|
| 1649 |
+
"eval_Qnli-dev_cosine_ap": 0.7347660327067432,
|
| 1650 |
+
"eval_Qnli-dev_cosine_f1": 0.6996336996336996,
|
| 1651 |
+
"eval_Qnli-dev_cosine_f1_threshold": 0.6012573838233948,
|
| 1652 |
+
"eval_Qnli-dev_cosine_precision": 0.6161290322580645,
|
| 1653 |
+
"eval_Qnli-dev_cosine_recall": 0.809322033898305,
|
| 1654 |
+
"eval_Qnli-dev_dot_accuracy": 0.681640625,
|
| 1655 |
+
"eval_Qnli-dev_dot_accuracy_threshold": 601.9920654296875,
|
| 1656 |
+
"eval_Qnli-dev_dot_ap": 0.6929187579311147,
|
| 1657 |
+
"eval_Qnli-dev_dot_f1": 0.6769759450171821,
|
| 1658 |
+
"eval_Qnli-dev_dot_f1_threshold": 541.3872680664062,
|
| 1659 |
+
"eval_Qnli-dev_dot_precision": 0.569364161849711,
|
| 1660 |
+
"eval_Qnli-dev_dot_recall": 0.8347457627118644,
|
| 1661 |
+
"eval_Qnli-dev_euclidean_accuracy": 0.705078125,
|
| 1662 |
+
"eval_Qnli-dev_euclidean_accuracy_threshold": 24.72307586669922,
|
| 1663 |
+
"eval_Qnli-dev_euclidean_ap": 0.7433711154783218,
|
| 1664 |
+
"eval_Qnli-dev_euclidean_f1": 0.6951871657754012,
|
| 1665 |
+
"eval_Qnli-dev_euclidean_f1_threshold": 27.680713653564453,
|
| 1666 |
+
"eval_Qnli-dev_euclidean_precision": 0.6,
|
| 1667 |
+
"eval_Qnli-dev_euclidean_recall": 0.826271186440678,
|
| 1668 |
+
"eval_Qnli-dev_manhattan_accuracy": 0.705078125,
|
| 1669 |
+
"eval_Qnli-dev_manhattan_accuracy_threshold": 771.5457763671875,
|
| 1670 |
+
"eval_Qnli-dev_manhattan_ap": 0.7417925203500755,
|
| 1671 |
+
"eval_Qnli-dev_manhattan_f1": 0.7015706806282722,
|
| 1672 |
+
"eval_Qnli-dev_manhattan_f1_threshold": 873.9434814453125,
|
| 1673 |
+
"eval_Qnli-dev_manhattan_precision": 0.5964391691394659,
|
| 1674 |
+
"eval_Qnli-dev_manhattan_recall": 0.8516949152542372,
|
| 1675 |
+
"eval_Qnli-dev_max_accuracy": 0.705078125,
|
| 1676 |
+
"eval_Qnli-dev_max_accuracy_threshold": 771.5457763671875,
|
| 1677 |
+
"eval_Qnli-dev_max_ap": 0.7433711154783218,
|
| 1678 |
+
"eval_Qnli-dev_max_f1": 0.7015706806282722,
|
| 1679 |
+
"eval_Qnli-dev_max_f1_threshold": 873.9434814453125,
|
| 1680 |
+
"eval_Qnli-dev_max_precision": 0.6161290322580645,
|
| 1681 |
+
"eval_Qnli-dev_max_recall": 0.8516949152542372,
|
| 1682 |
+
"eval_allNLI-dev_cosine_accuracy": 0.71484375,
|
| 1683 |
+
"eval_allNLI-dev_cosine_accuracy_threshold": 0.7252209186553955,
|
| 1684 |
+
"eval_allNLI-dev_cosine_ap": 0.5940767502246462,
|
| 1685 |
+
"eval_allNLI-dev_cosine_f1": 0.6111111111111112,
|
| 1686 |
+
"eval_allNLI-dev_cosine_f1_threshold": 0.6290402412414551,
|
| 1687 |
+
"eval_allNLI-dev_cosine_precision": 0.4847457627118644,
|
| 1688 |
+
"eval_allNLI-dev_cosine_recall": 0.8265895953757225,
|
| 1689 |
+
"eval_allNLI-dev_dot_accuracy": 0.712890625,
|
| 1690 |
+
"eval_allNLI-dev_dot_accuracy_threshold": 814.8841552734375,
|
| 1691 |
+
"eval_allNLI-dev_dot_ap": 0.5890724491285201,
|
| 1692 |
+
"eval_allNLI-dev_dot_f1": 0.6238938053097345,
|
| 1693 |
+
"eval_allNLI-dev_dot_f1_threshold": 624.3211669921875,
|
| 1694 |
+
"eval_allNLI-dev_dot_precision": 0.5053763440860215,
|
| 1695 |
+
"eval_allNLI-dev_dot_recall": 0.815028901734104,
|
| 1696 |
+
"eval_allNLI-dev_euclidean_accuracy": 0.712890625,
|
| 1697 |
+
"eval_allNLI-dev_euclidean_accuracy_threshold": 21.297210693359375,
|
| 1698 |
+
"eval_allNLI-dev_euclidean_ap": 0.5885663962477001,
|
| 1699 |
+
"eval_allNLI-dev_euclidean_f1": 0.6053169734151329,
|
| 1700 |
+
"eval_allNLI-dev_euclidean_f1_threshold": 27.594894409179688,
|
| 1701 |
+
"eval_allNLI-dev_euclidean_precision": 0.46835443037974683,
|
| 1702 |
+
"eval_allNLI-dev_euclidean_recall": 0.8554913294797688,
|
| 1703 |
+
"eval_allNLI-dev_manhattan_accuracy": 0.7109375,
|
| 1704 |
+
"eval_allNLI-dev_manhattan_accuracy_threshold": 655.72802734375,
|
| 1705 |
+
"eval_allNLI-dev_manhattan_ap": 0.5903691578356358,
|
| 1706 |
+
"eval_allNLI-dev_manhattan_f1": 0.6157024793388429,
|
| 1707 |
+
"eval_allNLI-dev_manhattan_f1_threshold": 858.11962890625,
|
| 1708 |
+
"eval_allNLI-dev_manhattan_precision": 0.4790996784565916,
|
| 1709 |
+
"eval_allNLI-dev_manhattan_recall": 0.861271676300578,
|
| 1710 |
+
"eval_allNLI-dev_max_accuracy": 0.71484375,
|
| 1711 |
+
"eval_allNLI-dev_max_accuracy_threshold": 814.8841552734375,
|
| 1712 |
+
"eval_allNLI-dev_max_ap": 0.5940767502246462,
|
| 1713 |
+
"eval_allNLI-dev_max_f1": 0.6238938053097345,
|
| 1714 |
+
"eval_allNLI-dev_max_f1_threshold": 858.11962890625,
|
| 1715 |
+
"eval_allNLI-dev_max_precision": 0.5053763440860215,
|
| 1716 |
+
"eval_allNLI-dev_max_recall": 0.861271676300578,
|
| 1717 |
+
"eval_sequential_score": 0.7433711154783218,
|
| 1718 |
+
"eval_sts-test_pearson_cosine": 0.9045402318307136,
|
| 1719 |
+
"eval_sts-test_pearson_dot": 0.8756384139446844,
|
| 1720 |
+
"eval_sts-test_pearson_euclidean": 0.9164304447060423,
|
| 1721 |
+
"eval_sts-test_pearson_manhattan": 0.9157247908119455,
|
| 1722 |
+
"eval_sts-test_pearson_max": 0.9164304447060423,
|
| 1723 |
+
"eval_sts-test_spearman_cosine": 0.9084125676732273,
|
| 1724 |
+
"eval_sts-test_spearman_dot": 0.8658350988039152,
|
| 1725 |
+
"eval_sts-test_spearman_euclidean": 0.9116794097547996,
|
| 1726 |
+
"eval_sts-test_spearman_manhattan": 0.9112685036727037,
|
| 1727 |
+
"eval_sts-test_spearman_max": 0.9116794097547996,
|
| 1728 |
+
"eval_vitaminc-pairs_loss": 4.510133266448975,
|
| 1729 |
+
"eval_vitaminc-pairs_runtime": 4.4836,
|
| 1730 |
+
"eval_vitaminc-pairs_samples_per_second": 28.548,
|
| 1731 |
+
"eval_vitaminc-pairs_steps_per_second": 0.223,
|
| 1732 |
+
"step": 1099
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 0.7026854219948849,
|
| 1736 |
+
"eval_negation-triplets_loss": 0.6731557846069336,
|
| 1737 |
+
"eval_negation-triplets_runtime": 3.3513,
|
| 1738 |
+
"eval_negation-triplets_samples_per_second": 38.194,
|
| 1739 |
+
"eval_negation-triplets_steps_per_second": 0.298,
|
| 1740 |
+
"step": 1099
|
| 1741 |
+
},
|
| 1742 |
+
{
|
| 1743 |
+
"epoch": 0.7026854219948849,
|
| 1744 |
+
"eval_scitail-pairs-pos_loss": 0.026047488674521446,
|
| 1745 |
+
"eval_scitail-pairs-pos_runtime": 2.666,
|
| 1746 |
+
"eval_scitail-pairs-pos_samples_per_second": 48.012,
|
| 1747 |
+
"eval_scitail-pairs-pos_steps_per_second": 0.375,
|
| 1748 |
+
"step": 1099
|
| 1749 |
+
},
|
| 1750 |
+
{
|
| 1751 |
+
"epoch": 0.7026854219948849,
|
| 1752 |
+
"eval_scitail-pairs-qa_loss": 6.3366285758093e-05,
|
| 1753 |
+
"eval_scitail-pairs-qa_runtime": 2.1132,
|
| 1754 |
+
"eval_scitail-pairs-qa_samples_per_second": 60.571,
|
| 1755 |
+
"eval_scitail-pairs-qa_steps_per_second": 0.473,
|
| 1756 |
+
"step": 1099
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 0.7026854219948849,
|
| 1760 |
+
"eval_xsum-pairs_loss": 0.010141758248209953,
|
| 1761 |
+
"eval_xsum-pairs_runtime": 3.077,
|
| 1762 |
+
"eval_xsum-pairs_samples_per_second": 41.599,
|
| 1763 |
+
"eval_xsum-pairs_steps_per_second": 0.325,
|
| 1764 |
+
"step": 1099
|
| 1765 |
+
},
|
| 1766 |
+
{
|
| 1767 |
+
"epoch": 0.7026854219948849,
|
| 1768 |
+
"eval_sciq_pairs_loss": 0.03606099635362625,
|
| 1769 |
+
"eval_sciq_pairs_runtime": 4.6134,
|
| 1770 |
+
"eval_sciq_pairs_samples_per_second": 27.745,
|
| 1771 |
+
"eval_sciq_pairs_steps_per_second": 0.217,
|
| 1772 |
+
"step": 1099
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 0.7026854219948849,
|
| 1776 |
+
"eval_qasc_pairs_loss": 0.0794580727815628,
|
| 1777 |
+
"eval_qasc_pairs_runtime": 2.0891,
|
| 1778 |
+
"eval_qasc_pairs_samples_per_second": 61.27,
|
| 1779 |
+
"eval_qasc_pairs_steps_per_second": 0.479,
|
| 1780 |
+
"step": 1099
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 0.7026854219948849,
|
| 1784 |
+
"eval_openbookqa_pairs_loss": 0.4053290784358978,
|
| 1785 |
+
"eval_openbookqa_pairs_runtime": 2.2366,
|
| 1786 |
+
"eval_openbookqa_pairs_samples_per_second": 57.229,
|
| 1787 |
+
"eval_openbookqa_pairs_steps_per_second": 0.447,
|
| 1788 |
+
"step": 1099
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 0.7026854219948849,
|
| 1792 |
+
"eval_msmarco_pairs_loss": 0.08789154142141342,
|
| 1793 |
+
"eval_msmarco_pairs_runtime": 2.2396,
|
| 1794 |
+
"eval_msmarco_pairs_samples_per_second": 57.153,
|
| 1795 |
+
"eval_msmarco_pairs_steps_per_second": 0.447,
|
| 1796 |
+
"step": 1099
|
| 1797 |
+
},
|
| 1798 |
+
{
|
| 1799 |
+
"epoch": 0.7026854219948849,
|
| 1800 |
+
"eval_nq_pairs_loss": 0.07979367673397064,
|
| 1801 |
+
"eval_nq_pairs_runtime": 3.1969,
|
| 1802 |
+
"eval_nq_pairs_samples_per_second": 40.039,
|
| 1803 |
+
"eval_nq_pairs_steps_per_second": 0.313,
|
| 1804 |
+
"step": 1099
|
| 1805 |
+
},
|
| 1806 |
+
{
|
| 1807 |
+
"epoch": 0.7026854219948849,
|
| 1808 |
+
"eval_trivia_pairs_loss": 0.08963163197040558,
|
| 1809 |
+
"eval_trivia_pairs_runtime": 3.2784,
|
| 1810 |
+
"eval_trivia_pairs_samples_per_second": 39.044,
|
| 1811 |
+
"eval_trivia_pairs_steps_per_second": 0.305,
|
| 1812 |
+
"step": 1099
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 0.7026854219948849,
|
| 1816 |
+
"eval_gooaq_pairs_loss": 0.1373358964920044,
|
| 1817 |
+
"eval_gooaq_pairs_runtime": 2.1632,
|
| 1818 |
+
"eval_gooaq_pairs_samples_per_second": 59.172,
|
| 1819 |
+
"eval_gooaq_pairs_steps_per_second": 0.462,
|
| 1820 |
+
"step": 1099
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 0.7026854219948849,
|
| 1824 |
+
"eval_paws-pos_loss": 0.025509441271424294,
|
| 1825 |
+
"eval_paws-pos_runtime": 2.2614,
|
| 1826 |
+
"eval_paws-pos_samples_per_second": 56.603,
|
| 1827 |
+
"eval_paws-pos_steps_per_second": 0.442,
|
| 1828 |
+
"step": 1099
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 0.7026854219948849,
|
| 1832 |
+
"eval_global_dataset_loss": 0.3096943199634552,
|
| 1833 |
+
"eval_global_dataset_runtime": 8.5451,
|
| 1834 |
+
"eval_global_dataset_samples_per_second": 38.034,
|
| 1835 |
+
"eval_global_dataset_steps_per_second": 0.351,
|
| 1836 |
+
"step": 1099
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 0.7058823529411765,
|
| 1840 |
+
"grad_norm": 3.7510690689086914,
|
| 1841 |
+
"learning_rate": 1.981281379849982e-05,
|
| 1842 |
+
"loss": 0.112,
|
| 1843 |
+
"step": 1104
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 0.7161125319693095,
|
| 1847 |
+
"grad_norm": 4.916189670562744,
|
| 1848 |
+
"learning_rate": 1.9772652616516024e-05,
|
| 1849 |
+
"loss": 0.0565,
|
| 1850 |
+
"step": 1120
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 0.7263427109974424,
|
| 1854 |
+
"grad_norm": 2.7884182929992676,
|
| 1855 |
+
"learning_rate": 1.972865681793253e-05,
|
| 1856 |
+
"loss": 0.1297,
|
| 1857 |
+
"step": 1136
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 0.7365728900255755,
|
| 1861 |
+
"grad_norm": 1.4771391153335571,
|
| 1862 |
+
"learning_rate": 1.9680849600202654e-05,
|
| 1863 |
+
"loss": 0.1792,
|
| 1864 |
+
"step": 1152
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 0.7468030690537084,
|
| 1868 |
+
"grad_norm": 3.511519193649292,
|
| 1869 |
+
"learning_rate": 1.9629256170408374e-05,
|
| 1870 |
+
"loss": 0.1376,
|
| 1871 |
+
"step": 1168
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 0.7570332480818415,
|
| 1875 |
+
"grad_norm": 3.235853910446167,
|
| 1876 |
+
"learning_rate": 1.957390373196951e-05,
|
| 1877 |
+
"loss": 0.1362,
|
| 1878 |
+
"step": 1184
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 0.7672634271099744,
|
| 1882 |
+
"grad_norm": 5.856909275054932,
|
| 1883 |
+
"learning_rate": 1.95148214703003e-05,
|
| 1884 |
+
"loss": 0.1589,
|
| 1885 |
+
"step": 1200
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 0.7774936061381074,
|
| 1889 |
+
"grad_norm": 0.06289539486169815,
|
| 1890 |
+
"learning_rate": 1.945204053742097e-05,
|
| 1891 |
+
"loss": 0.0846,
|
| 1892 |
+
"step": 1216
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 0.7877237851662404,
|
| 1896 |
+
"grad_norm": 1.4921785805199761e-05,
|
| 1897 |
+
"learning_rate": 1.9385594035532334e-05,
|
| 1898 |
+
"loss": 0.1241,
|
| 1899 |
+
"step": 1232
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 0.7979539641943734,
|
| 1903 |
+
"grad_norm": 2.955247640609741,
|
| 1904 |
+
"learning_rate": 1.9315516999562157e-05,
|
| 1905 |
+
"loss": 0.1532,
|
| 1906 |
+
"step": 1248
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 0.80306905370844,
|
| 1910 |
+
"eval_Qnli-dev_cosine_accuracy": 0.69140625,
|
| 1911 |
+
"eval_Qnli-dev_cosine_accuracy_threshold": 0.751553475856781,
|
| 1912 |
+
"eval_Qnli-dev_cosine_ap": 0.7397067201271598,
|
| 1913 |
+
"eval_Qnli-dev_cosine_f1": 0.6996587030716723,
|
| 1914 |
+
"eval_Qnli-dev_cosine_f1_threshold": 0.6113905906677246,
|
| 1915 |
+
"eval_Qnli-dev_cosine_precision": 0.5857142857142857,
|
| 1916 |
+
"eval_Qnli-dev_cosine_recall": 0.8686440677966102,
|
| 1917 |
+
"eval_Qnli-dev_dot_accuracy": 0.671875,
|
| 1918 |
+
"eval_Qnli-dev_dot_accuracy_threshold": 701.7063598632812,
|
| 1919 |
+
"eval_Qnli-dev_dot_ap": 0.6973190002082854,
|
| 1920 |
+
"eval_Qnli-dev_dot_f1": 0.6950819672131149,
|
| 1921 |
+
"eval_Qnli-dev_dot_f1_threshold": 549.8828735351562,
|
| 1922 |
+
"eval_Qnli-dev_dot_precision": 0.5668449197860963,
|
| 1923 |
+
"eval_Qnli-dev_dot_recall": 0.8983050847457628,
|
| 1924 |
+
"eval_Qnli-dev_euclidean_accuracy": 0.703125,
|
| 1925 |
+
"eval_Qnli-dev_euclidean_accuracy_threshold": 24.13254165649414,
|
| 1926 |
+
"eval_Qnli-dev_euclidean_ap": 0.7479263198995789,
|
| 1927 |
+
"eval_Qnli-dev_euclidean_f1": 0.7001733102253034,
|
| 1928 |
+
"eval_Qnli-dev_euclidean_f1_threshold": 26.746252059936523,
|
| 1929 |
+
"eval_Qnli-dev_euclidean_precision": 0.592375366568915,
|
| 1930 |
+
"eval_Qnli-dev_euclidean_recall": 0.8559322033898306,
|
| 1931 |
+
"eval_Qnli-dev_manhattan_accuracy": 0.703125,
|
| 1932 |
+
"eval_Qnli-dev_manhattan_accuracy_threshold": 746.0479736328125,
|
| 1933 |
+
"eval_Qnli-dev_manhattan_ap": 0.7475255901877714,
|
| 1934 |
+
"eval_Qnli-dev_manhattan_f1": 0.7038327526132405,
|
| 1935 |
+
"eval_Qnli-dev_manhattan_f1_threshold": 831.5120849609375,
|
| 1936 |
+
"eval_Qnli-dev_manhattan_precision": 0.5976331360946746,
|
| 1937 |
+
"eval_Qnli-dev_manhattan_recall": 0.8559322033898306,
|
| 1938 |
+
"eval_Qnli-dev_max_accuracy": 0.703125,
|
| 1939 |
+
"eval_Qnli-dev_max_accuracy_threshold": 746.0479736328125,
|
| 1940 |
+
"eval_Qnli-dev_max_ap": 0.7479263198995789,
|
| 1941 |
+
"eval_Qnli-dev_max_f1": 0.7038327526132405,
|
| 1942 |
+
"eval_Qnli-dev_max_f1_threshold": 831.5120849609375,
|
| 1943 |
+
"eval_Qnli-dev_max_precision": 0.5976331360946746,
|
| 1944 |
+
"eval_Qnli-dev_max_recall": 0.8983050847457628,
|
| 1945 |
+
"eval_allNLI-dev_cosine_accuracy": 0.728515625,
|
| 1946 |
+
"eval_allNLI-dev_cosine_accuracy_threshold": 0.7869724035263062,
|
| 1947 |
+
"eval_allNLI-dev_cosine_ap": 0.5886410776508728,
|
| 1948 |
+
"eval_allNLI-dev_cosine_f1": 0.6026871401151632,
|
| 1949 |
+
"eval_allNLI-dev_cosine_f1_threshold": 0.6058770418167114,
|
| 1950 |
+
"eval_allNLI-dev_cosine_precision": 0.4511494252873563,
|
| 1951 |
+
"eval_allNLI-dev_cosine_recall": 0.9075144508670521,
|
| 1952 |
+
"eval_allNLI-dev_dot_accuracy": 0.7265625,
|
| 1953 |
+
"eval_allNLI-dev_dot_accuracy_threshold": 785.6377563476562,
|
| 1954 |
+
"eval_allNLI-dev_dot_ap": 0.5906938241820221,
|
| 1955 |
+
"eval_allNLI-dev_dot_f1": 0.6022471910112359,
|
| 1956 |
+
"eval_allNLI-dev_dot_f1_threshold": 636.974365234375,
|
| 1957 |
+
"eval_allNLI-dev_dot_precision": 0.49264705882352944,
|
| 1958 |
+
"eval_allNLI-dev_dot_recall": 0.7745664739884393,
|
| 1959 |
+
"eval_allNLI-dev_euclidean_accuracy": 0.728515625,
|
| 1960 |
+
"eval_allNLI-dev_euclidean_accuracy_threshold": 20.30437469482422,
|
| 1961 |
+
"eval_allNLI-dev_euclidean_ap": 0.5863882590299698,
|
| 1962 |
+
"eval_allNLI-dev_euclidean_f1": 0.6101694915254238,
|
| 1963 |
+
"eval_allNLI-dev_euclidean_f1_threshold": 24.397754669189453,
|
| 1964 |
+
"eval_allNLI-dev_euclidean_precision": 0.525,
|
| 1965 |
+
"eval_allNLI-dev_euclidean_recall": 0.7283236994219653,
|
| 1966 |
+
"eval_allNLI-dev_manhattan_accuracy": 0.7265625,
|
| 1967 |
+
"eval_allNLI-dev_manhattan_accuracy_threshold": 632.682861328125,
|
| 1968 |
+
"eval_allNLI-dev_manhattan_ap": 0.5887339373924899,
|
| 1969 |
+
"eval_allNLI-dev_manhattan_f1": 0.6124401913875598,
|
| 1970 |
+
"eval_allNLI-dev_manhattan_f1_threshold": 763.71728515625,
|
| 1971 |
+
"eval_allNLI-dev_manhattan_precision": 0.5224489795918368,
|
| 1972 |
+
"eval_allNLI-dev_manhattan_recall": 0.7398843930635838,
|
| 1973 |
+
"eval_allNLI-dev_max_accuracy": 0.728515625,
|
| 1974 |
+
"eval_allNLI-dev_max_accuracy_threshold": 785.6377563476562,
|
| 1975 |
+
"eval_allNLI-dev_max_ap": 0.5906938241820221,
|
| 1976 |
+
"eval_allNLI-dev_max_f1": 0.6124401913875598,
|
| 1977 |
+
"eval_allNLI-dev_max_f1_threshold": 763.71728515625,
|
| 1978 |
+
"eval_allNLI-dev_max_precision": 0.525,
|
| 1979 |
+
"eval_allNLI-dev_max_recall": 0.9075144508670521,
|
| 1980 |
+
"eval_sequential_score": 0.7479263198995789,
|
| 1981 |
+
"eval_sts-test_pearson_cosine": 0.9080505623602911,
|
| 1982 |
+
"eval_sts-test_pearson_dot": 0.8941914869262664,
|
| 1983 |
+
"eval_sts-test_pearson_euclidean": 0.9250463167676135,
|
| 1984 |
+
"eval_sts-test_pearson_manhattan": 0.9243739071859142,
|
| 1985 |
+
"eval_sts-test_pearson_max": 0.9250463167676135,
|
| 1986 |
+
"eval_sts-test_spearman_cosine": 0.9193773427343025,
|
| 1987 |
+
"eval_sts-test_spearman_dot": 0.8898582919390006,
|
| 1988 |
+
"eval_sts-test_spearman_euclidean": 0.9199829615547123,
|
| 1989 |
+
"eval_sts-test_spearman_manhattan": 0.9193662444214774,
|
| 1990 |
+
"eval_sts-test_spearman_max": 0.9199829615547123,
|
| 1991 |
+
"eval_vitaminc-pairs_loss": 3.124298095703125,
|
| 1992 |
+
"eval_vitaminc-pairs_runtime": 4.4686,
|
| 1993 |
+
"eval_vitaminc-pairs_samples_per_second": 28.644,
|
| 1994 |
+
"eval_vitaminc-pairs_steps_per_second": 0.224,
|
| 1995 |
+
"step": 1256
|
| 1996 |
+
},
|
| 1997 |
+
{
|
| 1998 |
+
"epoch": 0.80306905370844,
|
| 1999 |
+
"eval_negation-triplets_loss": 0.5495098233222961,
|
| 2000 |
+
"eval_negation-triplets_runtime": 3.2625,
|
| 2001 |
+
"eval_negation-triplets_samples_per_second": 39.233,
|
| 2002 |
+
"eval_negation-triplets_steps_per_second": 0.307,
|
| 2003 |
+
"step": 1256
|
| 2004 |
+
},
|
| 2005 |
+
{
|
| 2006 |
+
"epoch": 0.80306905370844,
|
| 2007 |
+
"eval_scitail-pairs-pos_loss": 0.011394428089261055,
|
| 2008 |
+
"eval_scitail-pairs-pos_runtime": 2.6344,
|
| 2009 |
+
"eval_scitail-pairs-pos_samples_per_second": 48.587,
|
| 2010 |
+
"eval_scitail-pairs-pos_steps_per_second": 0.38,
|
| 2011 |
+
"step": 1256
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 0.80306905370844,
|
| 2015 |
+
"eval_scitail-pairs-qa_loss": 0.00027512505766935647,
|
| 2016 |
+
"eval_scitail-pairs-qa_runtime": 2.0567,
|
| 2017 |
+
"eval_scitail-pairs-qa_samples_per_second": 62.235,
|
| 2018 |
+
"eval_scitail-pairs-qa_steps_per_second": 0.486,
|
| 2019 |
+
"step": 1256
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 0.80306905370844,
|
| 2023 |
+
"eval_xsum-pairs_loss": 0.021254997700452805,
|
| 2024 |
+
"eval_xsum-pairs_runtime": 3.0587,
|
| 2025 |
+
"eval_xsum-pairs_samples_per_second": 41.848,
|
| 2026 |
+
"eval_xsum-pairs_steps_per_second": 0.327,
|
| 2027 |
+
"step": 1256
|
| 2028 |
+
},
|
| 2029 |
+
{
|
| 2030 |
+
"epoch": 0.80306905370844,
|
| 2031 |
+
"eval_sciq_pairs_loss": 0.030362099409103394,
|
| 2032 |
+
"eval_sciq_pairs_runtime": 4.6386,
|
| 2033 |
+
"eval_sciq_pairs_samples_per_second": 27.594,
|
| 2034 |
+
"eval_sciq_pairs_steps_per_second": 0.216,
|
| 2035 |
+
"step": 1256
|
| 2036 |
+
},
|
| 2037 |
+
{
|
| 2038 |
+
"epoch": 0.80306905370844,
|
| 2039 |
+
"eval_qasc_pairs_loss": 0.03284536302089691,
|
| 2040 |
+
"eval_qasc_pairs_runtime": 2.093,
|
| 2041 |
+
"eval_qasc_pairs_samples_per_second": 61.156,
|
| 2042 |
+
"eval_qasc_pairs_steps_per_second": 0.478,
|
| 2043 |
+
"step": 1256
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 0.80306905370844,
|
| 2047 |
+
"eval_openbookqa_pairs_loss": 0.36659273505210876,
|
| 2048 |
+
"eval_openbookqa_pairs_runtime": 2.2191,
|
| 2049 |
+
"eval_openbookqa_pairs_samples_per_second": 57.681,
|
| 2050 |
+
"eval_openbookqa_pairs_steps_per_second": 0.451,
|
| 2051 |
+
"step": 1256
|
| 2052 |
+
},
|
| 2053 |
+
{
|
| 2054 |
+
"epoch": 0.80306905370844,
|
| 2055 |
+
"eval_msmarco_pairs_loss": 0.10601727664470673,
|
| 2056 |
+
"eval_msmarco_pairs_runtime": 2.2441,
|
| 2057 |
+
"eval_msmarco_pairs_samples_per_second": 57.038,
|
| 2058 |
+
"eval_msmarco_pairs_steps_per_second": 0.446,
|
| 2059 |
+
"step": 1256
|
| 2060 |
+
},
|
| 2061 |
+
{
|
| 2062 |
+
"epoch": 0.80306905370844,
|
| 2063 |
+
"eval_nq_pairs_loss": 0.13051600754261017,
|
| 2064 |
+
"eval_nq_pairs_runtime": 3.176,
|
| 2065 |
+
"eval_nq_pairs_samples_per_second": 40.303,
|
| 2066 |
+
"eval_nq_pairs_steps_per_second": 0.315,
|
| 2067 |
+
"step": 1256
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 0.80306905370844,
|
| 2071 |
+
"eval_trivia_pairs_loss": 0.06672928482294083,
|
| 2072 |
+
"eval_trivia_pairs_runtime": 3.2525,
|
| 2073 |
+
"eval_trivia_pairs_samples_per_second": 39.354,
|
| 2074 |
+
"eval_trivia_pairs_steps_per_second": 0.307,
|
| 2075 |
+
"step": 1256
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 0.80306905370844,
|
| 2079 |
+
"eval_gooaq_pairs_loss": 0.21090617775917053,
|
| 2080 |
+
"eval_gooaq_pairs_runtime": 2.1187,
|
| 2081 |
+
"eval_gooaq_pairs_samples_per_second": 60.415,
|
| 2082 |
+
"eval_gooaq_pairs_steps_per_second": 0.472,
|
| 2083 |
+
"step": 1256
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 0.80306905370844,
|
| 2087 |
+
"eval_paws-pos_loss": 0.027658751234412193,
|
| 2088 |
+
"eval_paws-pos_runtime": 2.1768,
|
| 2089 |
+
"eval_paws-pos_samples_per_second": 58.802,
|
| 2090 |
+
"eval_paws-pos_steps_per_second": 0.459,
|
| 2091 |
+
"step": 1256
|
| 2092 |
+
},
|
| 2093 |
+
{
|
| 2094 |
+
"epoch": 0.80306905370844,
|
| 2095 |
+
"eval_global_dataset_loss": 0.21396204829216003,
|
| 2096 |
+
"eval_global_dataset_runtime": 8.4805,
|
| 2097 |
+
"eval_global_dataset_samples_per_second": 38.323,
|
| 2098 |
+
"eval_global_dataset_steps_per_second": 0.354,
|
| 2099 |
+
"step": 1256
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 0.8081841432225064,
|
| 2103 |
+
"grad_norm": 5.305633544921875,
|
| 2104 |
+
"learning_rate": 1.9241846378692478e-05,
|
| 2105 |
+
"loss": 0.0859,
|
| 2106 |
+
"step": 1264
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 0.8184143222506394,
|
| 2110 |
+
"grad_norm": 6.407199382781982,
|
| 2111 |
+
"learning_rate": 1.9164621016877572e-05,
|
| 2112 |
+
"loss": 0.0872,
|
| 2113 |
+
"step": 1280
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 0.8286445012787724,
|
| 2117 |
+
"grad_norm": 9.240785598754883,
|
| 2118 |
+
"learning_rate": 1.908388163236289e-05,
|
| 2119 |
+
"loss": 0.0685,
|
| 2120 |
+
"step": 1296
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 0.8388746803069054,
|
| 2124 |
+
"grad_norm": 0.3949645757675171,
|
| 2125 |
+
"learning_rate": 1.899967079621575e-05,
|
| 2126 |
+
"loss": 0.0729,
|
| 2127 |
+
"step": 1312
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 0.8491048593350383,
|
| 2131 |
+
"grad_norm": 4.485825061798096,
|
| 2132 |
+
"learning_rate": 1.8912032909879084e-05,
|
| 2133 |
+
"loss": 0.0679,
|
| 2134 |
+
"step": 1328
|
| 2135 |
+
},
|
| 2136 |
+
{
|
| 2137 |
+
"epoch": 0.8593350383631714,
|
| 2138 |
+
"grad_norm": 8.746214866638184,
|
| 2139 |
+
"learning_rate": 1.882101418176013e-05,
|
| 2140 |
+
"loss": 0.0752,
|
| 2141 |
+
"step": 1344
|
| 2142 |
+
},
|
| 2143 |
+
{
|
| 2144 |
+
"epoch": 0.8695652173913043,
|
| 2145 |
+
"grad_norm": 1.009152889251709,
|
| 2146 |
+
"learning_rate": 1.872666260286631e-05,
|
| 2147 |
+
"loss": 0.1651,
|
| 2148 |
+
"step": 1360
|
| 2149 |
+
},
|
| 2150 |
+
{
|
| 2151 |
+
"epoch": 0.8797953964194374,
|
| 2152 |
+
"grad_norm": 6.190288066864014,
|
| 2153 |
+
"learning_rate": 1.8629027921501253e-05,
|
| 2154 |
+
"loss": 0.0975,
|
| 2155 |
+
"step": 1376
|
| 2156 |
+
},
|
| 2157 |
+
{
|
| 2158 |
+
"epoch": 0.8900255754475703,
|
| 2159 |
+
"grad_norm": 17.912038803100586,
|
| 2160 |
+
"learning_rate": 1.8528161617034276e-05,
|
| 2161 |
+
"loss": 0.166,
|
| 2162 |
+
"step": 1392
|
| 2163 |
+
},
|
| 2164 |
+
{
|
| 2165 |
+
"epoch": 0.9002557544757033,
|
| 2166 |
+
"grad_norm": 10.208046913146973,
|
| 2167 |
+
"learning_rate": 1.8424116872757063e-05,
|
| 2168 |
+
"loss": 0.079,
|
| 2169 |
+
"step": 1408
|
| 2170 |
+
},
|
| 2171 |
+
{
|
| 2172 |
+
"epoch": 0.9034526854219949,
|
| 2173 |
+
"eval_Qnli-dev_cosine_accuracy": 0.701171875,
|
| 2174 |
+
"eval_Qnli-dev_cosine_accuracy_threshold": 0.6952118873596191,
|
| 2175 |
+
"eval_Qnli-dev_cosine_ap": 0.7283975483790786,
|
| 2176 |
+
"eval_Qnli-dev_cosine_f1": 0.6923076923076924,
|
| 2177 |
+
"eval_Qnli-dev_cosine_f1_threshold": 0.5710936784744263,
|
| 2178 |
+
"eval_Qnli-dev_cosine_precision": 0.5718232044198895,
|
| 2179 |
+
"eval_Qnli-dev_cosine_recall": 0.8771186440677966,
|
| 2180 |
+
"eval_Qnli-dev_dot_accuracy": 0.673828125,
|
| 2181 |
+
"eval_Qnli-dev_dot_accuracy_threshold": 626.558837890625,
|
| 2182 |
+
"eval_Qnli-dev_dot_ap": 0.6907790387670252,
|
| 2183 |
+
"eval_Qnli-dev_dot_f1": 0.6821963394342762,
|
| 2184 |
+
"eval_Qnli-dev_dot_f1_threshold": 524.71044921875,
|
| 2185 |
+
"eval_Qnli-dev_dot_precision": 0.5616438356164384,
|
| 2186 |
+
"eval_Qnli-dev_dot_recall": 0.8686440677966102,
|
| 2187 |
+
"eval_Qnli-dev_euclidean_accuracy": 0.705078125,
|
| 2188 |
+
"eval_Qnli-dev_euclidean_accuracy_threshold": 25.063961029052734,
|
| 2189 |
+
"eval_Qnli-dev_euclidean_ap": 0.7370960778450586,
|
| 2190 |
+
"eval_Qnli-dev_euclidean_f1": 0.693200663349917,
|
| 2191 |
+
"eval_Qnli-dev_euclidean_f1_threshold": 28.465593338012695,
|
| 2192 |
+
"eval_Qnli-dev_euclidean_precision": 0.5694822888283378,
|
| 2193 |
+
"eval_Qnli-dev_euclidean_recall": 0.885593220338983,
|
| 2194 |
+
"eval_Qnli-dev_manhattan_accuracy": 0.703125,
|
| 2195 |
+
"eval_Qnli-dev_manhattan_accuracy_threshold": 774.7894897460938,
|
| 2196 |
+
"eval_Qnli-dev_manhattan_ap": 0.7379094600871705,
|
| 2197 |
+
"eval_Qnli-dev_manhattan_f1": 0.6941580756013745,
|
| 2198 |
+
"eval_Qnli-dev_manhattan_f1_threshold": 870.0184326171875,
|
| 2199 |
+
"eval_Qnli-dev_manhattan_precision": 0.5838150289017341,
|
| 2200 |
+
"eval_Qnli-dev_manhattan_recall": 0.8559322033898306,
|
| 2201 |
+
"eval_Qnli-dev_max_accuracy": 0.705078125,
|
| 2202 |
+
"eval_Qnli-dev_max_accuracy_threshold": 774.7894897460938,
|
| 2203 |
+
"eval_Qnli-dev_max_ap": 0.7379094600871705,
|
| 2204 |
+
"eval_Qnli-dev_max_f1": 0.6941580756013745,
|
| 2205 |
+
"eval_Qnli-dev_max_f1_threshold": 870.0184326171875,
|
| 2206 |
+
"eval_Qnli-dev_max_precision": 0.5838150289017341,
|
| 2207 |
+
"eval_Qnli-dev_max_recall": 0.885593220338983,
|
| 2208 |
+
"eval_allNLI-dev_cosine_accuracy": 0.72265625,
|
| 2209 |
+
"eval_allNLI-dev_cosine_accuracy_threshold": 0.7867391109466553,
|
| 2210 |
+
"eval_allNLI-dev_cosine_ap": 0.6041209614947041,
|
| 2211 |
+
"eval_allNLI-dev_cosine_f1": 0.64,
|
| 2212 |
+
"eval_allNLI-dev_cosine_f1_threshold": 0.6645647287368774,
|
| 2213 |
+
"eval_allNLI-dev_cosine_precision": 0.51985559566787,
|
| 2214 |
+
"eval_allNLI-dev_cosine_recall": 0.8323699421965318,
|
| 2215 |
+
"eval_allNLI-dev_dot_accuracy": 0.716796875,
|
| 2216 |
+
"eval_allNLI-dev_dot_accuracy_threshold": 743.5255126953125,
|
| 2217 |
+
"eval_allNLI-dev_dot_ap": 0.6040927239029834,
|
| 2218 |
+
"eval_allNLI-dev_dot_f1": 0.6311300639658849,
|
| 2219 |
+
"eval_allNLI-dev_dot_f1_threshold": 632.1693115234375,
|
| 2220 |
+
"eval_allNLI-dev_dot_precision": 0.5,
|
| 2221 |
+
"eval_allNLI-dev_dot_recall": 0.8554913294797688,
|
| 2222 |
+
"eval_allNLI-dev_euclidean_accuracy": 0.72265625,
|
| 2223 |
+
"eval_allNLI-dev_euclidean_accuracy_threshold": 20.951969146728516,
|
| 2224 |
+
"eval_allNLI-dev_euclidean_ap": 0.6039261955905633,
|
| 2225 |
+
"eval_allNLI-dev_euclidean_f1": 0.6391304347826088,
|
| 2226 |
+
"eval_allNLI-dev_euclidean_f1_threshold": 25.697128295898438,
|
| 2227 |
+
"eval_allNLI-dev_euclidean_precision": 0.5121951219512195,
|
| 2228 |
+
"eval_allNLI-dev_euclidean_recall": 0.8497109826589595,
|
| 2229 |
+
"eval_allNLI-dev_manhattan_accuracy": 0.724609375,
|
| 2230 |
+
"eval_allNLI-dev_manhattan_accuracy_threshold": 655.82666015625,
|
| 2231 |
+
"eval_allNLI-dev_manhattan_ap": 0.6051707157459539,
|
| 2232 |
+
"eval_allNLI-dev_manhattan_f1": 0.6408602150537634,
|
| 2233 |
+
"eval_allNLI-dev_manhattan_f1_threshold": 802.6712646484375,
|
| 2234 |
+
"eval_allNLI-dev_manhattan_precision": 0.5102739726027398,
|
| 2235 |
+
"eval_allNLI-dev_manhattan_recall": 0.861271676300578,
|
| 2236 |
+
"eval_allNLI-dev_max_accuracy": 0.724609375,
|
| 2237 |
+
"eval_allNLI-dev_max_accuracy_threshold": 743.5255126953125,
|
| 2238 |
+
"eval_allNLI-dev_max_ap": 0.6051707157459539,
|
| 2239 |
+
"eval_allNLI-dev_max_f1": 0.6408602150537634,
|
| 2240 |
+
"eval_allNLI-dev_max_f1_threshold": 802.6712646484375,
|
| 2241 |
+
"eval_allNLI-dev_max_precision": 0.51985559566787,
|
| 2242 |
+
"eval_allNLI-dev_max_recall": 0.861271676300578,
|
| 2243 |
+
"eval_sequential_score": 0.7379094600871705,
|
| 2244 |
+
"eval_sts-test_pearson_cosine": 0.8937187089587758,
|
| 2245 |
+
"eval_sts-test_pearson_dot": 0.8780160036144846,
|
| 2246 |
+
"eval_sts-test_pearson_euclidean": 0.9143032007413827,
|
| 2247 |
+
"eval_sts-test_pearson_manhattan": 0.9135829801788513,
|
| 2248 |
+
"eval_sts-test_pearson_max": 0.9143032007413827,
|
| 2249 |
+
"eval_sts-test_spearman_cosine": 0.9102164462686534,
|
| 2250 |
+
"eval_sts-test_spearman_dot": 0.8867396212838953,
|
| 2251 |
+
"eval_sts-test_spearman_euclidean": 0.9118095911741873,
|
| 2252 |
+
"eval_sts-test_spearman_manhattan": 0.9116914478441138,
|
| 2253 |
+
"eval_sts-test_spearman_max": 0.9118095911741873,
|
| 2254 |
+
"eval_vitaminc-pairs_loss": 3.493912935256958,
|
| 2255 |
+
"eval_vitaminc-pairs_runtime": 4.593,
|
| 2256 |
+
"eval_vitaminc-pairs_samples_per_second": 27.868,
|
| 2257 |
+
"eval_vitaminc-pairs_steps_per_second": 0.218,
|
| 2258 |
+
"step": 1413
|
| 2259 |
+
},
|
| 2260 |
+
{
|
| 2261 |
+
"epoch": 0.9034526854219949,
|
| 2262 |
+
"eval_negation-triplets_loss": 0.5945984125137329,
|
| 2263 |
+
"eval_negation-triplets_runtime": 3.4606,
|
| 2264 |
+
"eval_negation-triplets_samples_per_second": 36.988,
|
| 2265 |
+
"eval_negation-triplets_steps_per_second": 0.289,
|
| 2266 |
+
"step": 1413
|
| 2267 |
+
},
|
| 2268 |
+
{
|
| 2269 |
+
"epoch": 0.9034526854219949,
|
| 2270 |
+
"eval_scitail-pairs-pos_loss": 0.014124380424618721,
|
| 2271 |
+
"eval_scitail-pairs-pos_runtime": 2.7213,
|
| 2272 |
+
"eval_scitail-pairs-pos_samples_per_second": 47.037,
|
| 2273 |
+
"eval_scitail-pairs-pos_steps_per_second": 0.367,
|
| 2274 |
+
"step": 1413
|
| 2275 |
+
},
|
| 2276 |
+
{
|
| 2277 |
+
"epoch": 0.9034526854219949,
|
| 2278 |
+
"eval_scitail-pairs-qa_loss": 0.0002780009526759386,
|
| 2279 |
+
"eval_scitail-pairs-qa_runtime": 2.1293,
|
| 2280 |
+
"eval_scitail-pairs-qa_samples_per_second": 60.112,
|
| 2281 |
+
"eval_scitail-pairs-qa_steps_per_second": 0.47,
|
| 2282 |
+
"step": 1413
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 0.9034526854219949,
|
| 2286 |
+
"eval_xsum-pairs_loss": 0.006070319563150406,
|
| 2287 |
+
"eval_xsum-pairs_runtime": 3.0868,
|
| 2288 |
+
"eval_xsum-pairs_samples_per_second": 41.467,
|
| 2289 |
+
"eval_xsum-pairs_steps_per_second": 0.324,
|
| 2290 |
+
"step": 1413
|
| 2291 |
+
},
|
| 2292 |
+
{
|
| 2293 |
+
"epoch": 0.9034526854219949,
|
| 2294 |
+
"eval_sciq_pairs_loss": 0.03563125431537628,
|
| 2295 |
+
"eval_sciq_pairs_runtime": 4.7217,
|
| 2296 |
+
"eval_sciq_pairs_samples_per_second": 27.109,
|
| 2297 |
+
"eval_sciq_pairs_steps_per_second": 0.212,
|
| 2298 |
+
"step": 1413
|
| 2299 |
+
},
|
| 2300 |
+
{
|
| 2301 |
+
"epoch": 0.9034526854219949,
|
| 2302 |
+
"eval_qasc_pairs_loss": 0.0521722249686718,
|
| 2303 |
+
"eval_qasc_pairs_runtime": 2.1625,
|
| 2304 |
+
"eval_qasc_pairs_samples_per_second": 59.191,
|
| 2305 |
+
"eval_qasc_pairs_steps_per_second": 0.462,
|
| 2306 |
+
"step": 1413
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.9034526854219949,
|
| 2310 |
+
"eval_openbookqa_pairs_loss": 0.42811527848243713,
|
| 2311 |
+
"eval_openbookqa_pairs_runtime": 2.2676,
|
| 2312 |
+
"eval_openbookqa_pairs_samples_per_second": 56.447,
|
| 2313 |
+
"eval_openbookqa_pairs_steps_per_second": 0.441,
|
| 2314 |
+
"step": 1413
|
| 2315 |
+
},
|
| 2316 |
+
{
|
| 2317 |
+
"epoch": 0.9034526854219949,
|
| 2318 |
+
"eval_msmarco_pairs_loss": 0.06089657172560692,
|
| 2319 |
+
"eval_msmarco_pairs_runtime": 2.3099,
|
| 2320 |
+
"eval_msmarco_pairs_samples_per_second": 55.413,
|
| 2321 |
+
"eval_msmarco_pairs_steps_per_second": 0.433,
|
| 2322 |
+
"step": 1413
|
| 2323 |
+
},
|
| 2324 |
+
{
|
| 2325 |
+
"epoch": 0.9034526854219949,
|
| 2326 |
+
"eval_nq_pairs_loss": 0.07195077836513519,
|
| 2327 |
+
"eval_nq_pairs_runtime": 3.2241,
|
| 2328 |
+
"eval_nq_pairs_samples_per_second": 39.701,
|
| 2329 |
+
"eval_nq_pairs_steps_per_second": 0.31,
|
| 2330 |
+
"step": 1413
|
| 2331 |
+
},
|
| 2332 |
+
{
|
| 2333 |
+
"epoch": 0.9034526854219949,
|
| 2334 |
+
"eval_trivia_pairs_loss": 0.07842885702848434,
|
| 2335 |
+
"eval_trivia_pairs_runtime": 3.318,
|
| 2336 |
+
"eval_trivia_pairs_samples_per_second": 38.578,
|
| 2337 |
+
"eval_trivia_pairs_steps_per_second": 0.301,
|
| 2338 |
+
"step": 1413
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 0.9034526854219949,
|
| 2342 |
+
"eval_gooaq_pairs_loss": 0.22998635470867157,
|
| 2343 |
+
"eval_gooaq_pairs_runtime": 2.1845,
|
| 2344 |
+
"eval_gooaq_pairs_samples_per_second": 58.594,
|
| 2345 |
+
"eval_gooaq_pairs_steps_per_second": 0.458,
|
| 2346 |
+
"step": 1413
|
| 2347 |
+
},
|
| 2348 |
+
{
|
| 2349 |
+
"epoch": 0.9034526854219949,
|
| 2350 |
+
"eval_paws-pos_loss": 0.026847416535019875,
|
| 2351 |
+
"eval_paws-pos_runtime": 2.257,
|
| 2352 |
+
"eval_paws-pos_samples_per_second": 56.711,
|
| 2353 |
+
"eval_paws-pos_steps_per_second": 0.443,
|
| 2354 |
+
"step": 1413
|
| 2355 |
+
},
|
| 2356 |
+
{
|
| 2357 |
+
"epoch": 0.9034526854219949,
|
| 2358 |
+
"eval_global_dataset_loss": 0.2313106507062912,
|
| 2359 |
+
"eval_global_dataset_runtime": 8.5928,
|
| 2360 |
+
"eval_global_dataset_samples_per_second": 37.822,
|
| 2361 |
+
"eval_global_dataset_steps_per_second": 0.349,
|
| 2362 |
+
"step": 1413
|
| 2363 |
+
}
|
| 2364 |
+
],
|
| 2365 |
+
"logging_steps": 16,
|
| 2366 |
+
"max_steps": 3128,
|
| 2367 |
+
"num_input_tokens_seen": 0,
|
| 2368 |
+
"num_train_epochs": 2,
|
| 2369 |
+
"save_steps": 157,
|
| 2370 |
+
"stateful_callbacks": {
|
| 2371 |
+
"TrainerControl": {
|
| 2372 |
+
"args": {
|
| 2373 |
+
"should_epoch_stop": false,
|
| 2374 |
+
"should_evaluate": false,
|
| 2375 |
+
"should_log": false,
|
| 2376 |
+
"should_save": true,
|
| 2377 |
+
"should_training_stop": false
|
| 2378 |
+
},
|
| 2379 |
+
"attributes": {}
|
| 2380 |
+
}
|
| 2381 |
+
},
|
| 2382 |
+
"total_flos": 0.0,
|
| 2383 |
+
"train_batch_size": 64,
|
| 2384 |
+
"trial_name": null,
|
| 2385 |
+
"trial_params": null
|
| 2386 |
+
}
|
checkpoint-1413/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ae5d47b05b44fe0990966324c4a681d83a347e47ed07c3ca3b6e4479e0e161a9
|
| 3 |
+
size 5560
|