Commit
·
96d71cc
1
Parent(s):
d491d2f
init
Browse files- config.json +24 -0
- eval_results.txt +12 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- scheduler.pt +3 -0
- special_tokens_map.json +1 -0
- tokenizer_config.json +1 -0
- train_results.txt +3 -0
- trainer_state.json +2465 -0
- training_args.bin +3 -0
- vocab.txt +0 -0
config.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "condenser-bert-base-uncased_mix_0.00001_cls_output_dir",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForCL"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"gradient_checkpointing": false,
|
| 8 |
+
"hidden_act": "gelu",
|
| 9 |
+
"hidden_dropout_prob": 0.1,
|
| 10 |
+
"hidden_size": 768,
|
| 11 |
+
"initializer_range": 0.02,
|
| 12 |
+
"intermediate_size": 3072,
|
| 13 |
+
"layer_norm_eps": 1e-12,
|
| 14 |
+
"max_position_embeddings": 512,
|
| 15 |
+
"model_type": "bert",
|
| 16 |
+
"num_attention_heads": 12,
|
| 17 |
+
"num_hidden_layers": 12,
|
| 18 |
+
"pad_token_id": 0,
|
| 19 |
+
"position_embedding_type": "absolute",
|
| 20 |
+
"transformers_version": "4.2.1",
|
| 21 |
+
"type_vocab_size": 2,
|
| 22 |
+
"use_cache": true,
|
| 23 |
+
"vocab_size": 30522
|
| 24 |
+
}
|
eval_results.txt
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
epoch = 2.0
|
| 2 |
+
eval_CR = 85.66
|
| 3 |
+
eval_MPQA = 88.14
|
| 4 |
+
eval_MR = 79.86
|
| 5 |
+
eval_MRPC = 75.17
|
| 6 |
+
eval_SST2 = 85.44
|
| 7 |
+
eval_SUBJ = 94.37
|
| 8 |
+
eval_TREC = 77.73
|
| 9 |
+
eval_avg_sts = 0.8077588380658864
|
| 10 |
+
eval_avg_transfer = 83.76714285714286
|
| 11 |
+
eval_sickr_spearman = 0.7571885560828946
|
| 12 |
+
eval_stsb_spearman = 0.8583291200488784
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2bd99e42f53761469aa17e488799db23a3c269f6f8b68ee78bf8328c0e18aac6
|
| 3 |
+
size 2067367739
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0ce87a1350c4b43fc28c2628b2c093ee80e4e00b31a4c80be47c11f23229eebe
|
| 3 |
+
size 1572106385
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0860173bd14d0ff91fc823ff03ea3804cf1630ba377c53353277d58ebad94505
|
| 3 |
+
size 623
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]"}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"do_lower_case": true, "unk_token": "[UNK]", "sep_token": "[SEP]", "pad_token": "[PAD]", "cls_token": "[CLS]", "mask_token": "[MASK]", "tokenize_chinese_chars": true, "strip_accents": null, "do_basic_tokenize": true, "never_split": null, "model_max_length": 512, "name_or_path": "condenser-bert-base-uncased", "special_tokens_map_file": "condenser-bert-base-uncased/special_tokens_map.json"}
|
train_results.txt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
epoch = 2.0
|
| 2 |
+
train_runtime = 14615.2615
|
| 3 |
+
train_samples_per_second = 2.138
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2465 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8583291200488784,
|
| 3 |
+
"best_model_checkpoint": "condenser-bert-base-uncased_mix_0.00001_cls_output_dir",
|
| 4 |
+
"epoch": 2.0,
|
| 5 |
+
"global_step": 31252,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"electra_acc": 97.8496,
|
| 12 |
+
"electra_fix_acc": 0.9447,
|
| 13 |
+
"electra_rep_acc": 0.0242,
|
| 14 |
+
"epoch": 0.0,
|
| 15 |
+
"learning_rate": 6.999776014335083e-06,
|
| 16 |
+
"loss": 0.0046,
|
| 17 |
+
"neg_sim": -0.0126,
|
| 18 |
+
"pos_sim": 0.6661,
|
| 19 |
+
"step": 1
|
| 20 |
+
},
|
| 21 |
+
{
|
| 22 |
+
"epoch": 0.01,
|
| 23 |
+
"eval_avg_sts": 0.6467281681849183,
|
| 24 |
+
"eval_sickr_spearman": 0.6493190116334423,
|
| 25 |
+
"eval_stsb_spearman": 0.6441373247363943,
|
| 26 |
+
"step": 125
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.02,
|
| 30 |
+
"eval_avg_sts": 0.6766102197245321,
|
| 31 |
+
"eval_sickr_spearman": 0.6747404807226937,
|
| 32 |
+
"eval_stsb_spearman": 0.6784799587263706,
|
| 33 |
+
"step": 250
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.02,
|
| 37 |
+
"eval_avg_sts": 0.6927318121134369,
|
| 38 |
+
"eval_sickr_spearman": 0.6849395969264682,
|
| 39 |
+
"eval_stsb_spearman": 0.7005240273004057,
|
| 40 |
+
"step": 375
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"electra_acc": 111.3766,
|
| 44 |
+
"electra_fix_acc": 0.9699,
|
| 45 |
+
"electra_rep_acc": 0.4695,
|
| 46 |
+
"epoch": 0.03,
|
| 47 |
+
"learning_rate": 6.888007167541277e-06,
|
| 48 |
+
"loss": 0.0027,
|
| 49 |
+
"neg_sim": -0.0136,
|
| 50 |
+
"pos_sim": 0.7499,
|
| 51 |
+
"step": 500
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.03,
|
| 55 |
+
"eval_avg_sts": 0.70406785938114,
|
| 56 |
+
"eval_sickr_spearman": 0.6912377231336422,
|
| 57 |
+
"eval_stsb_spearman": 0.716897995628638,
|
| 58 |
+
"step": 500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.04,
|
| 62 |
+
"eval_avg_sts": 0.7079843162896757,
|
| 63 |
+
"eval_sickr_spearman": 0.6933601214439895,
|
| 64 |
+
"eval_stsb_spearman": 0.7226085111353618,
|
| 65 |
+
"step": 625
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.05,
|
| 69 |
+
"eval_avg_sts": 0.721992790338854,
|
| 70 |
+
"eval_sickr_spearman": 0.7041983394834997,
|
| 71 |
+
"eval_stsb_spearman": 0.7397872411942082,
|
| 72 |
+
"step": 750
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.06,
|
| 76 |
+
"eval_avg_sts": 0.7371963659388359,
|
| 77 |
+
"eval_sickr_spearman": 0.7161540010979675,
|
| 78 |
+
"eval_stsb_spearman": 0.7582387307797042,
|
| 79 |
+
"step": 875
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"electra_acc": 115.0045,
|
| 83 |
+
"electra_fix_acc": 0.9686,
|
| 84 |
+
"electra_rep_acc": 0.6116,
|
| 85 |
+
"epoch": 0.06,
|
| 86 |
+
"learning_rate": 6.776014335082554e-06,
|
| 87 |
+
"loss": 0.002,
|
| 88 |
+
"neg_sim": -0.0139,
|
| 89 |
+
"pos_sim": 0.7675,
|
| 90 |
+
"step": 1000
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.06,
|
| 94 |
+
"eval_avg_sts": 0.7419491613050363,
|
| 95 |
+
"eval_sickr_spearman": 0.7199929829426116,
|
| 96 |
+
"eval_stsb_spearman": 0.763905339667461,
|
| 97 |
+
"step": 1000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.07,
|
| 101 |
+
"eval_avg_sts": 0.7503468807257706,
|
| 102 |
+
"eval_sickr_spearman": 0.7253389886259777,
|
| 103 |
+
"eval_stsb_spearman": 0.7753547728255635,
|
| 104 |
+
"step": 1125
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.08,
|
| 108 |
+
"eval_avg_sts": 0.755381635152361,
|
| 109 |
+
"eval_sickr_spearman": 0.7283938147085431,
|
| 110 |
+
"eval_stsb_spearman": 0.782369455596179,
|
| 111 |
+
"step": 1250
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.09,
|
| 115 |
+
"eval_avg_sts": 0.7557951064861299,
|
| 116 |
+
"eval_sickr_spearman": 0.7286210018183454,
|
| 117 |
+
"eval_stsb_spearman": 0.7829692111539145,
|
| 118 |
+
"step": 1375
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"electra_acc": 115.5115,
|
| 122 |
+
"electra_fix_acc": 0.9698,
|
| 123 |
+
"electra_rep_acc": 0.6302,
|
| 124 |
+
"epoch": 0.1,
|
| 125 |
+
"learning_rate": 6.6640215026238325e-06,
|
| 126 |
+
"loss": 0.0017,
|
| 127 |
+
"neg_sim": -0.014,
|
| 128 |
+
"pos_sim": 0.7738,
|
| 129 |
+
"step": 1500
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.1,
|
| 133 |
+
"eval_avg_sts": 0.7624381952103951,
|
| 134 |
+
"eval_sickr_spearman": 0.7329619086730178,
|
| 135 |
+
"eval_stsb_spearman": 0.7919144817477726,
|
| 136 |
+
"step": 1500
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.1,
|
| 140 |
+
"eval_avg_sts": 0.765407531009575,
|
| 141 |
+
"eval_sickr_spearman": 0.7350173996590618,
|
| 142 |
+
"eval_stsb_spearman": 0.7957976623600883,
|
| 143 |
+
"step": 1625
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.11,
|
| 147 |
+
"eval_avg_sts": 0.7663640946087169,
|
| 148 |
+
"eval_sickr_spearman": 0.7356288836114704,
|
| 149 |
+
"eval_stsb_spearman": 0.7970993056059634,
|
| 150 |
+
"step": 1750
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.12,
|
| 154 |
+
"eval_avg_sts": 0.7654280480014255,
|
| 155 |
+
"eval_sickr_spearman": 0.7361625571795493,
|
| 156 |
+
"eval_stsb_spearman": 0.7946935388233016,
|
| 157 |
+
"step": 1875
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"electra_acc": 115.9922,
|
| 161 |
+
"electra_fix_acc": 0.971,
|
| 162 |
+
"electra_rep_acc": 0.641,
|
| 163 |
+
"epoch": 0.13,
|
| 164 |
+
"learning_rate": 6.55202867016511e-06,
|
| 165 |
+
"loss": 0.0016,
|
| 166 |
+
"neg_sim": -0.014,
|
| 167 |
+
"pos_sim": 0.7751,
|
| 168 |
+
"step": 2000
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 0.13,
|
| 172 |
+
"eval_avg_sts": 0.7672864050175499,
|
| 173 |
+
"eval_sickr_spearman": 0.7369789418106929,
|
| 174 |
+
"eval_stsb_spearman": 0.7975938682244069,
|
| 175 |
+
"step": 2000
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.14,
|
| 179 |
+
"eval_avg_sts": 0.7747771973615925,
|
| 180 |
+
"eval_sickr_spearman": 0.7388607043028352,
|
| 181 |
+
"eval_stsb_spearman": 0.8106936904203497,
|
| 182 |
+
"step": 2125
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.14,
|
| 186 |
+
"eval_avg_sts": 0.7774171758966744,
|
| 187 |
+
"eval_sickr_spearman": 0.7346681655205053,
|
| 188 |
+
"eval_stsb_spearman": 0.8201661862728435,
|
| 189 |
+
"step": 2250
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.15,
|
| 193 |
+
"eval_avg_sts": 0.7804372234935246,
|
| 194 |
+
"eval_sickr_spearman": 0.7373037281186174,
|
| 195 |
+
"eval_stsb_spearman": 0.8235707188684319,
|
| 196 |
+
"step": 2375
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"electra_acc": 116.1787,
|
| 200 |
+
"electra_fix_acc": 0.971,
|
| 201 |
+
"electra_rep_acc": 0.6492,
|
| 202 |
+
"epoch": 0.16,
|
| 203 |
+
"learning_rate": 6.440035837706387e-06,
|
| 204 |
+
"loss": 0.0018,
|
| 205 |
+
"neg_sim": -0.0139,
|
| 206 |
+
"pos_sim": 0.7692,
|
| 207 |
+
"step": 2500
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.16,
|
| 211 |
+
"eval_avg_sts": 0.7810767988630156,
|
| 212 |
+
"eval_sickr_spearman": 0.7360565525386754,
|
| 213 |
+
"eval_stsb_spearman": 0.8260970451873557,
|
| 214 |
+
"step": 2500
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 0.17,
|
| 218 |
+
"eval_avg_sts": 0.7810534850927242,
|
| 219 |
+
"eval_sickr_spearman": 0.737071930023077,
|
| 220 |
+
"eval_stsb_spearman": 0.8250350401623715,
|
| 221 |
+
"step": 2625
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 0.18,
|
| 225 |
+
"eval_avg_sts": 0.7840637132723827,
|
| 226 |
+
"eval_sickr_spearman": 0.7369239942306477,
|
| 227 |
+
"eval_stsb_spearman": 0.8312034323141176,
|
| 228 |
+
"step": 2750
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 0.18,
|
| 232 |
+
"eval_avg_sts": 0.7807650924553761,
|
| 233 |
+
"eval_sickr_spearman": 0.736852331827302,
|
| 234 |
+
"eval_stsb_spearman": 0.8246778530834501,
|
| 235 |
+
"step": 2875
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"electra_acc": 116.4021,
|
| 239 |
+
"electra_fix_acc": 0.9711,
|
| 240 |
+
"electra_rep_acc": 0.6575,
|
| 241 |
+
"epoch": 0.19,
|
| 242 |
+
"learning_rate": 6.328043005247664e-06,
|
| 243 |
+
"loss": 0.0017,
|
| 244 |
+
"neg_sim": -0.014,
|
| 245 |
+
"pos_sim": 0.778,
|
| 246 |
+
"step": 3000
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.19,
|
| 250 |
+
"eval_avg_sts": 0.7812090218351604,
|
| 251 |
+
"eval_sickr_spearman": 0.7374819235049528,
|
| 252 |
+
"eval_stsb_spearman": 0.824936120165368,
|
| 253 |
+
"step": 3000
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.2,
|
| 257 |
+
"eval_avg_sts": 0.7792184022991786,
|
| 258 |
+
"eval_sickr_spearman": 0.735666443932795,
|
| 259 |
+
"eval_stsb_spearman": 0.8227703606655621,
|
| 260 |
+
"step": 3125
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.21,
|
| 264 |
+
"eval_avg_sts": 0.7868425974055346,
|
| 265 |
+
"eval_sickr_spearman": 0.7406432345394065,
|
| 266 |
+
"eval_stsb_spearman": 0.8330419602716627,
|
| 267 |
+
"step": 3250
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 0.22,
|
| 271 |
+
"eval_avg_sts": 0.7879152712258332,
|
| 272 |
+
"eval_sickr_spearman": 0.74201269730361,
|
| 273 |
+
"eval_stsb_spearman": 0.8338178451480563,
|
| 274 |
+
"step": 3375
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"electra_acc": 116.4857,
|
| 278 |
+
"electra_fix_acc": 0.9706,
|
| 279 |
+
"electra_rep_acc": 0.6652,
|
| 280 |
+
"epoch": 0.22,
|
| 281 |
+
"learning_rate": 6.2160501727889415e-06,
|
| 282 |
+
"loss": 0.0016,
|
| 283 |
+
"neg_sim": -0.014,
|
| 284 |
+
"pos_sim": 0.7831,
|
| 285 |
+
"step": 3500
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.22,
|
| 289 |
+
"eval_avg_sts": 0.7874505134915601,
|
| 290 |
+
"eval_sickr_spearman": 0.7415818102926086,
|
| 291 |
+
"eval_stsb_spearman": 0.8333192166905118,
|
| 292 |
+
"step": 3500
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.23,
|
| 296 |
+
"eval_avg_sts": 0.7885881180801638,
|
| 297 |
+
"eval_sickr_spearman": 0.745498890708191,
|
| 298 |
+
"eval_stsb_spearman": 0.8316773454521367,
|
| 299 |
+
"step": 3625
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 0.24,
|
| 303 |
+
"eval_avg_sts": 0.786897125097908,
|
| 304 |
+
"eval_sickr_spearman": 0.7427211560498221,
|
| 305 |
+
"eval_stsb_spearman": 0.8310730941459938,
|
| 306 |
+
"step": 3750
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.25,
|
| 310 |
+
"eval_avg_sts": 0.7884858023873396,
|
| 311 |
+
"eval_sickr_spearman": 0.7450858232358233,
|
| 312 |
+
"eval_stsb_spearman": 0.831885781538856,
|
| 313 |
+
"step": 3875
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"electra_acc": 116.6361,
|
| 317 |
+
"electra_fix_acc": 0.9706,
|
| 318 |
+
"electra_rep_acc": 0.6699,
|
| 319 |
+
"epoch": 0.26,
|
| 320 |
+
"learning_rate": 6.104057340330219e-06,
|
| 321 |
+
"loss": 0.0016,
|
| 322 |
+
"neg_sim": -0.014,
|
| 323 |
+
"pos_sim": 0.7846,
|
| 324 |
+
"step": 4000
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.26,
|
| 328 |
+
"eval_avg_sts": 0.7883005639004375,
|
| 329 |
+
"eval_sickr_spearman": 0.7467304081490642,
|
| 330 |
+
"eval_stsb_spearman": 0.8298707196518108,
|
| 331 |
+
"step": 4000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.26,
|
| 335 |
+
"eval_avg_sts": 0.7890018606938185,
|
| 336 |
+
"eval_sickr_spearman": 0.7466418868291138,
|
| 337 |
+
"eval_stsb_spearman": 0.8313618345585232,
|
| 338 |
+
"step": 4125
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.27,
|
| 342 |
+
"eval_avg_sts": 0.794939161048242,
|
| 343 |
+
"eval_sickr_spearman": 0.7542289276433795,
|
| 344 |
+
"eval_stsb_spearman": 0.8356493944531046,
|
| 345 |
+
"step": 4250
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.28,
|
| 349 |
+
"eval_avg_sts": 0.7950124813232355,
|
| 350 |
+
"eval_sickr_spearman": 0.7546040025145097,
|
| 351 |
+
"eval_stsb_spearman": 0.8354209601319615,
|
| 352 |
+
"step": 4375
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"electra_acc": 116.6274,
|
| 356 |
+
"electra_fix_acc": 0.9702,
|
| 357 |
+
"electra_rep_acc": 0.6733,
|
| 358 |
+
"epoch": 0.29,
|
| 359 |
+
"learning_rate": 5.992064507871497e-06,
|
| 360 |
+
"loss": 0.0016,
|
| 361 |
+
"neg_sim": -0.0141,
|
| 362 |
+
"pos_sim": 0.7871,
|
| 363 |
+
"step": 4500
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 0.29,
|
| 367 |
+
"eval_avg_sts": 0.7938800981934206,
|
| 368 |
+
"eval_sickr_spearman": 0.7533463081233527,
|
| 369 |
+
"eval_stsb_spearman": 0.8344138882634886,
|
| 370 |
+
"step": 4500
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.3,
|
| 374 |
+
"eval_avg_sts": 0.794615358586916,
|
| 375 |
+
"eval_sickr_spearman": 0.7546161063520722,
|
| 376 |
+
"eval_stsb_spearman": 0.8346146108217597,
|
| 377 |
+
"step": 4625
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.3,
|
| 381 |
+
"eval_avg_sts": 0.7945912313688637,
|
| 382 |
+
"eval_sickr_spearman": 0.7546232629861865,
|
| 383 |
+
"eval_stsb_spearman": 0.8345591997515407,
|
| 384 |
+
"step": 4750
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.31,
|
| 388 |
+
"eval_avg_sts": 0.7926504221766928,
|
| 389 |
+
"eval_sickr_spearman": 0.7513844777851131,
|
| 390 |
+
"eval_stsb_spearman": 0.8339163665682726,
|
| 391 |
+
"step": 4875
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"electra_acc": 116.73,
|
| 395 |
+
"electra_fix_acc": 0.9701,
|
| 396 |
+
"electra_rep_acc": 0.6783,
|
| 397 |
+
"epoch": 0.32,
|
| 398 |
+
"learning_rate": 5.880071675412773e-06,
|
| 399 |
+
"loss": 0.0015,
|
| 400 |
+
"neg_sim": -0.0141,
|
| 401 |
+
"pos_sim": 0.7904,
|
| 402 |
+
"step": 5000
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.32,
|
| 406 |
+
"eval_avg_sts": 0.7934518878705803,
|
| 407 |
+
"eval_sickr_spearman": 0.7517016271479089,
|
| 408 |
+
"eval_stsb_spearman": 0.8352021485932517,
|
| 409 |
+
"step": 5000
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.33,
|
| 413 |
+
"eval_avg_sts": 0.7947730330766363,
|
| 414 |
+
"eval_sickr_spearman": 0.7522364054313209,
|
| 415 |
+
"eval_stsb_spearman": 0.8373096607219516,
|
| 416 |
+
"step": 5125
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.34,
|
| 420 |
+
"eval_avg_sts": 0.7957221976017312,
|
| 421 |
+
"eval_sickr_spearman": 0.7467118201128077,
|
| 422 |
+
"eval_stsb_spearman": 0.8447325750906549,
|
| 423 |
+
"step": 5250
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.34,
|
| 427 |
+
"eval_avg_sts": 0.7897362921108364,
|
| 428 |
+
"eval_sickr_spearman": 0.7382444172402828,
|
| 429 |
+
"eval_stsb_spearman": 0.8412281669813901,
|
| 430 |
+
"step": 5375
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"electra_acc": 116.8512,
|
| 434 |
+
"electra_fix_acc": 0.9704,
|
| 435 |
+
"electra_rep_acc": 0.6807,
|
| 436 |
+
"epoch": 0.35,
|
| 437 |
+
"learning_rate": 5.7680788429540505e-06,
|
| 438 |
+
"loss": 0.0016,
|
| 439 |
+
"neg_sim": -0.014,
|
| 440 |
+
"pos_sim": 0.7834,
|
| 441 |
+
"step": 5500
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.35,
|
| 445 |
+
"eval_avg_sts": 0.791664811568233,
|
| 446 |
+
"eval_sickr_spearman": 0.7412088968210431,
|
| 447 |
+
"eval_stsb_spearman": 0.8421207263154228,
|
| 448 |
+
"step": 5500
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.36,
|
| 452 |
+
"eval_avg_sts": 0.791979531720516,
|
| 453 |
+
"eval_sickr_spearman": 0.7422650046794644,
|
| 454 |
+
"eval_stsb_spearman": 0.8416940587615678,
|
| 455 |
+
"step": 5625
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 0.37,
|
| 459 |
+
"eval_avg_sts": 0.7909414570864208,
|
| 460 |
+
"eval_sickr_spearman": 0.7407966939085013,
|
| 461 |
+
"eval_stsb_spearman": 0.8410862202643404,
|
| 462 |
+
"step": 5750
|
| 463 |
+
},
|
| 464 |
+
{
|
| 465 |
+
"epoch": 0.38,
|
| 466 |
+
"eval_avg_sts": 0.7916033250892136,
|
| 467 |
+
"eval_sickr_spearman": 0.7420070776647416,
|
| 468 |
+
"eval_stsb_spearman": 0.8411995725136856,
|
| 469 |
+
"step": 5875
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"electra_acc": 116.9386,
|
| 473 |
+
"electra_fix_acc": 0.9702,
|
| 474 |
+
"electra_rep_acc": 0.6839,
|
| 475 |
+
"epoch": 0.38,
|
| 476 |
+
"learning_rate": 5.656086010495328e-06,
|
| 477 |
+
"loss": 0.0014,
|
| 478 |
+
"neg_sim": -0.014,
|
| 479 |
+
"pos_sim": 0.7867,
|
| 480 |
+
"step": 6000
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.38,
|
| 484 |
+
"eval_avg_sts": 0.7939838516948429,
|
| 485 |
+
"eval_sickr_spearman": 0.7457015819562598,
|
| 486 |
+
"eval_stsb_spearman": 0.8422661214334259,
|
| 487 |
+
"step": 6000
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 0.39,
|
| 491 |
+
"eval_avg_sts": 0.7897949803868374,
|
| 492 |
+
"eval_sickr_spearman": 0.7381373559151772,
|
| 493 |
+
"eval_stsb_spearman": 0.8414526048584976,
|
| 494 |
+
"step": 6125
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.4,
|
| 498 |
+
"eval_avg_sts": 0.792929873578434,
|
| 499 |
+
"eval_sickr_spearman": 0.7426651998166468,
|
| 500 |
+
"eval_stsb_spearman": 0.8431945473402213,
|
| 501 |
+
"step": 6250
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.41,
|
| 505 |
+
"eval_avg_sts": 0.7941689576028486,
|
| 506 |
+
"eval_sickr_spearman": 0.7447726604544468,
|
| 507 |
+
"eval_stsb_spearman": 0.8435652547512503,
|
| 508 |
+
"step": 6375
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"electra_acc": 116.9876,
|
| 512 |
+
"electra_fix_acc": 0.9702,
|
| 513 |
+
"electra_rep_acc": 0.6869,
|
| 514 |
+
"epoch": 0.42,
|
| 515 |
+
"learning_rate": 5.544093178036605e-06,
|
| 516 |
+
"loss": 0.0014,
|
| 517 |
+
"neg_sim": -0.0141,
|
| 518 |
+
"pos_sim": 0.7923,
|
| 519 |
+
"step": 6500
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 0.42,
|
| 523 |
+
"eval_avg_sts": 0.7943810788058159,
|
| 524 |
+
"eval_sickr_spearman": 0.7449998475642491,
|
| 525 |
+
"eval_stsb_spearman": 0.8437623100473826,
|
| 526 |
+
"step": 6500
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 0.42,
|
| 530 |
+
"eval_avg_sts": 0.7943037355857338,
|
| 531 |
+
"eval_sickr_spearman": 0.7452223756572118,
|
| 532 |
+
"eval_stsb_spearman": 0.8433850955142557,
|
| 533 |
+
"step": 6625
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.43,
|
| 537 |
+
"eval_avg_sts": 0.7941830032512052,
|
| 538 |
+
"eval_sickr_spearman": 0.7457148385402568,
|
| 539 |
+
"eval_stsb_spearman": 0.8426511679621534,
|
| 540 |
+
"step": 6750
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.44,
|
| 544 |
+
"eval_avg_sts": 0.7945109790018022,
|
| 545 |
+
"eval_sickr_spearman": 0.7458948591084468,
|
| 546 |
+
"eval_stsb_spearman": 0.8431270988951577,
|
| 547 |
+
"step": 6875
|
| 548 |
+
},
|
| 549 |
+
{
|
| 550 |
+
"electra_acc": 117.061,
|
| 551 |
+
"electra_fix_acc": 0.9704,
|
| 552 |
+
"electra_rep_acc": 0.6883,
|
| 553 |
+
"epoch": 0.45,
|
| 554 |
+
"learning_rate": 5.432100345577883e-06,
|
| 555 |
+
"loss": 0.0014,
|
| 556 |
+
"neg_sim": -0.0142,
|
| 557 |
+
"pos_sim": 0.8,
|
| 558 |
+
"step": 7000
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 0.45,
|
| 562 |
+
"eval_avg_sts": 0.7988457805161364,
|
| 563 |
+
"eval_sickr_spearman": 0.7515535952932767,
|
| 564 |
+
"eval_stsb_spearman": 0.846137965738996,
|
| 565 |
+
"step": 7000
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 0.46,
|
| 569 |
+
"eval_avg_sts": 0.7981158834848314,
|
| 570 |
+
"eval_sickr_spearman": 0.75075815222236,
|
| 571 |
+
"eval_stsb_spearman": 0.845473614747303,
|
| 572 |
+
"step": 7125
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.46,
|
| 576 |
+
"eval_avg_sts": 0.7984443557563039,
|
| 577 |
+
"eval_sickr_spearman": 0.7501957080245197,
|
| 578 |
+
"eval_stsb_spearman": 0.8466930034880882,
|
| 579 |
+
"step": 7250
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 0.47,
|
| 583 |
+
"eval_avg_sts": 0.7981127107574888,
|
| 584 |
+
"eval_sickr_spearman": 0.750318523550897,
|
| 585 |
+
"eval_stsb_spearman": 0.8459068979640807,
|
| 586 |
+
"step": 7375
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"electra_acc": 117.069,
|
| 590 |
+
"electra_fix_acc": 0.9701,
|
| 591 |
+
"electra_rep_acc": 0.6888,
|
| 592 |
+
"epoch": 0.48,
|
| 593 |
+
"learning_rate": 5.32010751311916e-06,
|
| 594 |
+
"loss": 0.0015,
|
| 595 |
+
"neg_sim": -0.0142,
|
| 596 |
+
"pos_sim": 0.8001,
|
| 597 |
+
"step": 7500
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.48,
|
| 601 |
+
"eval_avg_sts": 0.7955868437200273,
|
| 602 |
+
"eval_sickr_spearman": 0.7487511246176671,
|
| 603 |
+
"eval_stsb_spearman": 0.8424225628223875,
|
| 604 |
+
"step": 7500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.49,
|
| 608 |
+
"eval_avg_sts": 0.7949766204622221,
|
| 609 |
+
"eval_sickr_spearman": 0.7481150007102207,
|
| 610 |
+
"eval_stsb_spearman": 0.8418382402142234,
|
| 611 |
+
"step": 7625
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.5,
|
| 615 |
+
"eval_avg_sts": 0.7950165968598919,
|
| 616 |
+
"eval_sickr_spearman": 0.7485309980797762,
|
| 617 |
+
"eval_stsb_spearman": 0.8415021956400077,
|
| 618 |
+
"step": 7750
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.5,
|
| 622 |
+
"eval_avg_sts": 0.7952250496118516,
|
| 623 |
+
"eval_sickr_spearman": 0.7486211044260741,
|
| 624 |
+
"eval_stsb_spearman": 0.8418289947976292,
|
| 625 |
+
"step": 7875
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"electra_acc": 117.1619,
|
| 629 |
+
"electra_fix_acc": 0.9703,
|
| 630 |
+
"electra_rep_acc": 0.6937,
|
| 631 |
+
"epoch": 0.51,
|
| 632 |
+
"learning_rate": 5.2081146806604376e-06,
|
| 633 |
+
"loss": 0.0014,
|
| 634 |
+
"neg_sim": -0.0142,
|
| 635 |
+
"pos_sim": 0.7987,
|
| 636 |
+
"step": 8000
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 0.51,
|
| 640 |
+
"eval_avg_sts": 0.7952021451003002,
|
| 641 |
+
"eval_sickr_spearman": 0.7491326836874914,
|
| 642 |
+
"eval_stsb_spearman": 0.8412716065131091,
|
| 643 |
+
"step": 8000
|
| 644 |
+
},
|
| 645 |
+
{
|
| 646 |
+
"epoch": 0.52,
|
| 647 |
+
"eval_avg_sts": 0.7951490971039081,
|
| 648 |
+
"eval_sickr_spearman": 0.749455164502547,
|
| 649 |
+
"eval_stsb_spearman": 0.8408430297052691,
|
| 650 |
+
"step": 8125
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.53,
|
| 654 |
+
"eval_avg_sts": 0.7950116620178889,
|
| 655 |
+
"eval_sickr_spearman": 0.749741093649408,
|
| 656 |
+
"eval_stsb_spearman": 0.8402822303863696,
|
| 657 |
+
"step": 8250
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.54,
|
| 661 |
+
"eval_avg_sts": 0.7951408099684255,
|
| 662 |
+
"eval_sickr_spearman": 0.7501430179062422,
|
| 663 |
+
"eval_stsb_spearman": 0.8401386020306089,
|
| 664 |
+
"step": 8375
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"electra_acc": 117.2641,
|
| 668 |
+
"electra_fix_acc": 0.9702,
|
| 669 |
+
"electra_rep_acc": 0.6954,
|
| 670 |
+
"epoch": 0.54,
|
| 671 |
+
"learning_rate": 5.096121848201715e-06,
|
| 672 |
+
"loss": 0.0014,
|
| 673 |
+
"neg_sim": -0.0142,
|
| 674 |
+
"pos_sim": 0.806,
|
| 675 |
+
"step": 8500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.54,
|
| 679 |
+
"eval_avg_sts": 0.7949691120514141,
|
| 680 |
+
"eval_sickr_spearman": 0.7501533445930513,
|
| 681 |
+
"eval_stsb_spearman": 0.8397848795097769,
|
| 682 |
+
"step": 8500
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.55,
|
| 686 |
+
"eval_avg_sts": 0.7947062336358759,
|
| 687 |
+
"eval_sickr_spearman": 0.7503012323543793,
|
| 688 |
+
"eval_stsb_spearman": 0.8391112349173724,
|
| 689 |
+
"step": 8625
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.56,
|
| 693 |
+
"eval_avg_sts": 0.7944181183521635,
|
| 694 |
+
"eval_sickr_spearman": 0.7493049712483499,
|
| 695 |
+
"eval_stsb_spearman": 0.8395312654559769,
|
| 696 |
+
"step": 8750
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.57,
|
| 700 |
+
"eval_avg_sts": 0.7950044953143297,
|
| 701 |
+
"eval_sickr_spearman": 0.7496116017999308,
|
| 702 |
+
"eval_stsb_spearman": 0.8403973888287286,
|
| 703 |
+
"step": 8875
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"electra_acc": 117.2706,
|
| 707 |
+
"electra_fix_acc": 0.9705,
|
| 708 |
+
"electra_rep_acc": 0.6956,
|
| 709 |
+
"epoch": 0.58,
|
| 710 |
+
"learning_rate": 4.984129015742992e-06,
|
| 711 |
+
"loss": 0.0014,
|
| 712 |
+
"neg_sim": -0.0143,
|
| 713 |
+
"pos_sim": 0.8093,
|
| 714 |
+
"step": 9000
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 0.58,
|
| 718 |
+
"eval_avg_sts": 0.7950848432290827,
|
| 719 |
+
"eval_sickr_spearman": 0.7493472386176154,
|
| 720 |
+
"eval_stsb_spearman": 0.8408224478405498,
|
| 721 |
+
"step": 9000
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 0.58,
|
| 725 |
+
"eval_avg_sts": 0.794847808534162,
|
| 726 |
+
"eval_sickr_spearman": 0.7489566016696193,
|
| 727 |
+
"eval_stsb_spearman": 0.8407390153987048,
|
| 728 |
+
"step": 9125
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.59,
|
| 732 |
+
"eval_avg_sts": 0.795644348129942,
|
| 733 |
+
"eval_sickr_spearman": 0.7493907547955183,
|
| 734 |
+
"eval_stsb_spearman": 0.8418979414643657,
|
| 735 |
+
"step": 9250
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.6,
|
| 739 |
+
"eval_avg_sts": 0.7966154624139274,
|
| 740 |
+
"eval_sickr_spearman": 0.7486175020934662,
|
| 741 |
+
"eval_stsb_spearman": 0.8446134227343888,
|
| 742 |
+
"step": 9375
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"electra_acc": 117.2471,
|
| 746 |
+
"electra_fix_acc": 0.9703,
|
| 747 |
+
"electra_rep_acc": 0.6966,
|
| 748 |
+
"epoch": 0.61,
|
| 749 |
+
"learning_rate": 4.872136183284269e-06,
|
| 750 |
+
"loss": 0.0015,
|
| 751 |
+
"neg_sim": -0.0143,
|
| 752 |
+
"pos_sim": 0.8085,
|
| 753 |
+
"step": 9500
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.61,
|
| 757 |
+
"eval_avg_sts": 0.7962741237962319,
|
| 758 |
+
"eval_sickr_spearman": 0.7489400789707246,
|
| 759 |
+
"eval_stsb_spearman": 0.8436081686217393,
|
| 760 |
+
"step": 9500
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.62,
|
| 764 |
+
"eval_avg_sts": 0.7966016709380292,
|
| 765 |
+
"eval_sickr_spearman": 0.7497793264061527,
|
| 766 |
+
"eval_stsb_spearman": 0.8434240154699058,
|
| 767 |
+
"step": 9625
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.62,
|
| 771 |
+
"eval_avg_sts": 0.7969314966558516,
|
| 772 |
+
"eval_sickr_spearman": 0.750271933382502,
|
| 773 |
+
"eval_stsb_spearman": 0.8435910599292011,
|
| 774 |
+
"step": 9750
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.63,
|
| 778 |
+
"eval_avg_sts": 0.7965405773738363,
|
| 779 |
+
"eval_sickr_spearman": 0.7509994124448837,
|
| 780 |
+
"eval_stsb_spearman": 0.842081742302789,
|
| 781 |
+
"step": 9875
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"electra_acc": 117.3599,
|
| 785 |
+
"electra_fix_acc": 0.9704,
|
| 786 |
+
"electra_rep_acc": 0.6998,
|
| 787 |
+
"epoch": 0.64,
|
| 788 |
+
"learning_rate": 4.760143350825547e-06,
|
| 789 |
+
"loss": 0.0014,
|
| 790 |
+
"neg_sim": -0.0143,
|
| 791 |
+
"pos_sim": 0.8077,
|
| 792 |
+
"step": 10000
|
| 793 |
+
},
|
| 794 |
+
{
|
| 795 |
+
"epoch": 0.64,
|
| 796 |
+
"eval_avg_sts": 0.7964394231656002,
|
| 797 |
+
"eval_sickr_spearman": 0.7509571931067197,
|
| 798 |
+
"eval_stsb_spearman": 0.8419216532244808,
|
| 799 |
+
"step": 10000
|
| 800 |
+
},
|
| 801 |
+
{
|
| 802 |
+
"epoch": 0.65,
|
| 803 |
+
"eval_avg_sts": 0.7967246593169031,
|
| 804 |
+
"eval_sickr_spearman": 0.7512494143278691,
|
| 805 |
+
"eval_stsb_spearman": 0.842199904305937,
|
| 806 |
+
"step": 10125
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.66,
|
| 810 |
+
"eval_avg_sts": 0.798107636355013,
|
| 811 |
+
"eval_sickr_spearman": 0.750279041985515,
|
| 812 |
+
"eval_stsb_spearman": 0.845936230724511,
|
| 813 |
+
"step": 10250
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"epoch": 0.66,
|
| 817 |
+
"eval_avg_sts": 0.7980710535062169,
|
| 818 |
+
"eval_sickr_spearman": 0.7505661719099119,
|
| 819 |
+
"eval_stsb_spearman": 0.8455759351025218,
|
| 820 |
+
"step": 10375
|
| 821 |
+
},
|
| 822 |
+
{
|
| 823 |
+
"electra_acc": 117.3933,
|
| 824 |
+
"electra_fix_acc": 0.9703,
|
| 825 |
+
"electra_rep_acc": 0.7006,
|
| 826 |
+
"epoch": 0.67,
|
| 827 |
+
"learning_rate": 4.648150518366825e-06,
|
| 828 |
+
"loss": 0.0014,
|
| 829 |
+
"neg_sim": -0.0143,
|
| 830 |
+
"pos_sim": 0.8081,
|
| 831 |
+
"step": 10500
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.67,
|
| 835 |
+
"eval_avg_sts": 0.7985924819917696,
|
| 836 |
+
"eval_sickr_spearman": 0.7519674312632674,
|
| 837 |
+
"eval_stsb_spearman": 0.8452175327202718,
|
| 838 |
+
"step": 10500
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 0.68,
|
| 842 |
+
"eval_avg_sts": 0.7991692973398108,
|
| 843 |
+
"eval_sickr_spearman": 0.7515644683598212,
|
| 844 |
+
"eval_stsb_spearman": 0.8467741263198004,
|
| 845 |
+
"step": 10625
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 0.69,
|
| 849 |
+
"eval_avg_sts": 0.7987755187970038,
|
| 850 |
+
"eval_sickr_spearman": 0.751040334943309,
|
| 851 |
+
"eval_stsb_spearman": 0.8465107026506987,
|
| 852 |
+
"step": 10750
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 0.7,
|
| 856 |
+
"eval_avg_sts": 0.7986197258898549,
|
| 857 |
+
"eval_sickr_spearman": 0.7514156980010478,
|
| 858 |
+
"eval_stsb_spearman": 0.8458237537786619,
|
| 859 |
+
"step": 10875
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"electra_acc": 117.4302,
|
| 863 |
+
"electra_fix_acc": 0.9705,
|
| 864 |
+
"electra_rep_acc": 0.7012,
|
| 865 |
+
"epoch": 0.7,
|
| 866 |
+
"learning_rate": 4.536157685908102e-06,
|
| 867 |
+
"loss": 0.0013,
|
| 868 |
+
"neg_sim": -0.0143,
|
| 869 |
+
"pos_sim": 0.8102,
|
| 870 |
+
"step": 11000
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.7,
|
| 874 |
+
"eval_avg_sts": 0.7996251236145968,
|
| 875 |
+
"eval_sickr_spearman": 0.7529341052108109,
|
| 876 |
+
"eval_stsb_spearman": 0.8463161420183827,
|
| 877 |
+
"step": 11000
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.71,
|
| 881 |
+
"eval_avg_sts": 0.7994727880295167,
|
| 882 |
+
"eval_sickr_spearman": 0.752963500244891,
|
| 883 |
+
"eval_stsb_spearman": 0.8459820758141423,
|
| 884 |
+
"step": 11125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.72,
|
| 888 |
+
"eval_avg_sts": 0.7988993639342742,
|
| 889 |
+
"eval_sickr_spearman": 0.7527381383169434,
|
| 890 |
+
"eval_stsb_spearman": 0.845060589551605,
|
| 891 |
+
"step": 11250
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.73,
|
| 895 |
+
"eval_avg_sts": 0.7949700279790515,
|
| 896 |
+
"eval_sickr_spearman": 0.7504287549286974,
|
| 897 |
+
"eval_stsb_spearman": 0.8395113010294055,
|
| 898 |
+
"step": 11375
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"electra_acc": 117.3738,
|
| 902 |
+
"electra_fix_acc": 0.9702,
|
| 903 |
+
"electra_rep_acc": 0.6995,
|
| 904 |
+
"epoch": 0.74,
|
| 905 |
+
"learning_rate": 4.424164853449379e-06,
|
| 906 |
+
"loss": 0.0014,
|
| 907 |
+
"neg_sim": -0.0143,
|
| 908 |
+
"pos_sim": 0.8157,
|
| 909 |
+
"step": 11500
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 0.74,
|
| 913 |
+
"eval_avg_sts": 0.7945615778286725,
|
| 914 |
+
"eval_sickr_spearman": 0.7499855239246266,
|
| 915 |
+
"eval_stsb_spearman": 0.8391376317327186,
|
| 916 |
+
"step": 11500
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 0.74,
|
| 920 |
+
"eval_avg_sts": 0.7934292933863014,
|
| 921 |
+
"eval_sickr_spearman": 0.7480520799673368,
|
| 922 |
+
"eval_stsb_spearman": 0.838806506805266,
|
| 923 |
+
"step": 11625
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 0.75,
|
| 927 |
+
"eval_avg_sts": 0.7937723514656398,
|
| 928 |
+
"eval_sickr_spearman": 0.7481406973494901,
|
| 929 |
+
"eval_stsb_spearman": 0.8394040055817897,
|
| 930 |
+
"step": 11750
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 0.76,
|
| 934 |
+
"eval_avg_sts": 0.7945405318685601,
|
| 935 |
+
"eval_sickr_spearman": 0.7490017989360725,
|
| 936 |
+
"eval_stsb_spearman": 0.8400792648010476,
|
| 937 |
+
"step": 11875
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"electra_acc": 117.5412,
|
| 941 |
+
"electra_fix_acc": 0.9707,
|
| 942 |
+
"electra_rep_acc": 0.7042,
|
| 943 |
+
"epoch": 0.77,
|
| 944 |
+
"learning_rate": 4.312172020990656e-06,
|
| 945 |
+
"loss": 0.0016,
|
| 946 |
+
"neg_sim": -0.0143,
|
| 947 |
+
"pos_sim": 0.812,
|
| 948 |
+
"step": 12000
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 0.77,
|
| 952 |
+
"eval_avg_sts": 0.7960514566703415,
|
| 953 |
+
"eval_sickr_spearman": 0.7499365802322611,
|
| 954 |
+
"eval_stsb_spearman": 0.8421663331084219,
|
| 955 |
+
"step": 12000
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 0.78,
|
| 959 |
+
"eval_avg_sts": 0.7951893349270129,
|
| 960 |
+
"eval_sickr_spearman": 0.7490124138094904,
|
| 961 |
+
"eval_stsb_spearman": 0.8413662560445356,
|
| 962 |
+
"step": 12125
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.78,
|
| 966 |
+
"eval_avg_sts": 0.7937236608939515,
|
| 967 |
+
"eval_sickr_spearman": 0.7449763603556459,
|
| 968 |
+
"eval_stsb_spearman": 0.8424709614322572,
|
| 969 |
+
"step": 12250
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.79,
|
| 973 |
+
"eval_avg_sts": 0.7934627772608434,
|
| 974 |
+
"eval_sickr_spearman": 0.746114793521932,
|
| 975 |
+
"eval_stsb_spearman": 0.8408107609997548,
|
| 976 |
+
"step": 12375
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"electra_acc": 117.4273,
|
| 980 |
+
"electra_fix_acc": 0.9702,
|
| 981 |
+
"electra_rep_acc": 0.7033,
|
| 982 |
+
"epoch": 0.8,
|
| 983 |
+
"learning_rate": 4.2001791885319345e-06,
|
| 984 |
+
"loss": 0.0016,
|
| 985 |
+
"neg_sim": -0.0143,
|
| 986 |
+
"pos_sim": 0.8084,
|
| 987 |
+
"step": 12500
|
| 988 |
+
},
|
| 989 |
+
{
|
| 990 |
+
"epoch": 0.8,
|
| 991 |
+
"eval_avg_sts": 0.793506856374318,
|
| 992 |
+
"eval_sickr_spearman": 0.7463023069419462,
|
| 993 |
+
"eval_stsb_spearman": 0.8407114058066898,
|
| 994 |
+
"step": 12500
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 0.81,
|
| 998 |
+
"eval_avg_sts": 0.7929774718082235,
|
| 999 |
+
"eval_sickr_spearman": 0.7465257476258366,
|
| 1000 |
+
"eval_stsb_spearman": 0.8394291959906105,
|
| 1001 |
+
"step": 12625
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.82,
|
| 1005 |
+
"eval_avg_sts": 0.7907041469753885,
|
| 1006 |
+
"eval_sickr_spearman": 0.7438328358926072,
|
| 1007 |
+
"eval_stsb_spearman": 0.8375754580581698,
|
| 1008 |
+
"step": 12750
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 0.82,
|
| 1012 |
+
"eval_avg_sts": 0.7919377238754862,
|
| 1013 |
+
"eval_sickr_spearman": 0.744254116683321,
|
| 1014 |
+
"eval_stsb_spearman": 0.8396213310676515,
|
| 1015 |
+
"step": 12875
|
| 1016 |
+
},
|
| 1017 |
+
{
|
| 1018 |
+
"electra_acc": 117.4657,
|
| 1019 |
+
"electra_fix_acc": 0.9701,
|
| 1020 |
+
"electra_rep_acc": 0.7049,
|
| 1021 |
+
"epoch": 0.83,
|
| 1022 |
+
"learning_rate": 4.088186356073212e-06,
|
| 1023 |
+
"loss": 0.0014,
|
| 1024 |
+
"neg_sim": -0.0143,
|
| 1025 |
+
"pos_sim": 0.8091,
|
| 1026 |
+
"step": 13000
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 0.83,
|
| 1030 |
+
"eval_avg_sts": 0.7950524625796482,
|
| 1031 |
+
"eval_sickr_spearman": 0.7459233415515996,
|
| 1032 |
+
"eval_stsb_spearman": 0.8441815836076968,
|
| 1033 |
+
"step": 13000
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 0.84,
|
| 1037 |
+
"eval_avg_sts": 0.7953046984165892,
|
| 1038 |
+
"eval_sickr_spearman": 0.7460563877025831,
|
| 1039 |
+
"eval_stsb_spearman": 0.8445530091305953,
|
| 1040 |
+
"step": 13125
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.85,
|
| 1044 |
+
"eval_avg_sts": 0.7960686224710865,
|
| 1045 |
+
"eval_sickr_spearman": 0.745027801665286,
|
| 1046 |
+
"eval_stsb_spearman": 0.8471094432768869,
|
| 1047 |
+
"step": 13250
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.86,
|
| 1051 |
+
"eval_avg_sts": 0.7964731409284048,
|
| 1052 |
+
"eval_sickr_spearman": 0.7453551336215867,
|
| 1053 |
+
"eval_stsb_spearman": 0.847591148235223,
|
| 1054 |
+
"step": 13375
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"electra_acc": 117.5481,
|
| 1058 |
+
"electra_fix_acc": 0.9705,
|
| 1059 |
+
"electra_rep_acc": 0.7049,
|
| 1060 |
+
"epoch": 0.86,
|
| 1061 |
+
"learning_rate": 3.976193523614489e-06,
|
| 1062 |
+
"loss": 0.0014,
|
| 1063 |
+
"neg_sim": -0.0143,
|
| 1064 |
+
"pos_sim": 0.8115,
|
| 1065 |
+
"step": 13500
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.86,
|
| 1069 |
+
"eval_avg_sts": 0.7966766934403904,
|
| 1070 |
+
"eval_sickr_spearman": 0.7463476483017039,
|
| 1071 |
+
"eval_stsb_spearman": 0.8470057385790768,
|
| 1072 |
+
"step": 13500
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 0.87,
|
| 1076 |
+
"eval_avg_sts": 0.7966025495961442,
|
| 1077 |
+
"eval_sickr_spearman": 0.7463816062904207,
|
| 1078 |
+
"eval_stsb_spearman": 0.8468234929018676,
|
| 1079 |
+
"step": 13625
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 0.88,
|
| 1083 |
+
"eval_avg_sts": 0.7967818015292757,
|
| 1084 |
+
"eval_sickr_spearman": 0.746880793527667,
|
| 1085 |
+
"eval_stsb_spearman": 0.8466828095308844,
|
| 1086 |
+
"step": 13750
|
| 1087 |
+
},
|
| 1088 |
+
{
|
| 1089 |
+
"epoch": 0.89,
|
| 1090 |
+
"eval_avg_sts": 0.7969535576693652,
|
| 1091 |
+
"eval_sickr_spearman": 0.7471762808637142,
|
| 1092 |
+
"eval_stsb_spearman": 0.8467308344750161,
|
| 1093 |
+
"step": 13875
|
| 1094 |
+
},
|
| 1095 |
+
{
|
| 1096 |
+
"electra_acc": 117.5482,
|
| 1097 |
+
"electra_fix_acc": 0.9705,
|
| 1098 |
+
"electra_rep_acc": 0.7068,
|
| 1099 |
+
"epoch": 0.9,
|
| 1100 |
+
"learning_rate": 3.864200691155766e-06,
|
| 1101 |
+
"loss": 0.0013,
|
| 1102 |
+
"neg_sim": -0.0143,
|
| 1103 |
+
"pos_sim": 0.8142,
|
| 1104 |
+
"step": 14000
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 0.9,
|
| 1108 |
+
"eval_avg_sts": 0.7965629720194758,
|
| 1109 |
+
"eval_sickr_spearman": 0.7468308350906325,
|
| 1110 |
+
"eval_stsb_spearman": 0.8462951089483192,
|
| 1111 |
+
"step": 14000
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 0.9,
|
| 1115 |
+
"eval_avg_sts": 0.7965266307017043,
|
| 1116 |
+
"eval_sickr_spearman": 0.7447568101909722,
|
| 1117 |
+
"eval_stsb_spearman": 0.8482964512124362,
|
| 1118 |
+
"step": 14125
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.91,
|
| 1122 |
+
"eval_avg_sts": 0.7965480662045883,
|
| 1123 |
+
"eval_sickr_spearman": 0.7448633912050633,
|
| 1124 |
+
"eval_stsb_spearman": 0.8482327412041132,
|
| 1125 |
+
"step": 14250
|
| 1126 |
+
},
|
| 1127 |
+
{
|
| 1128 |
+
"epoch": 0.92,
|
| 1129 |
+
"eval_avg_sts": 0.7964806384496487,
|
| 1130 |
+
"eval_sickr_spearman": 0.745111231688484,
|
| 1131 |
+
"eval_stsb_spearman": 0.8478500452108133,
|
| 1132 |
+
"step": 14375
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"electra_acc": 117.5208,
|
| 1136 |
+
"electra_fix_acc": 0.9701,
|
| 1137 |
+
"electra_rep_acc": 0.7071,
|
| 1138 |
+
"epoch": 0.93,
|
| 1139 |
+
"learning_rate": 3.7522078586970435e-06,
|
| 1140 |
+
"loss": 0.0013,
|
| 1141 |
+
"neg_sim": -0.0143,
|
| 1142 |
+
"pos_sim": 0.8155,
|
| 1143 |
+
"step": 14500
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.93,
|
| 1147 |
+
"eval_avg_sts": 0.7971949399590785,
|
| 1148 |
+
"eval_sickr_spearman": 0.7457434650767139,
|
| 1149 |
+
"eval_stsb_spearman": 0.8486464148414431,
|
| 1150 |
+
"step": 14500
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.94,
|
| 1154 |
+
"eval_avg_sts": 0.7970029281099968,
|
| 1155 |
+
"eval_sickr_spearman": 0.7458370776934168,
|
| 1156 |
+
"eval_stsb_spearman": 0.8481687785265768,
|
| 1157 |
+
"step": 14625
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 0.94,
|
| 1161 |
+
"eval_avg_sts": 0.7991675032273975,
|
| 1162 |
+
"eval_sickr_spearman": 0.7483727836316391,
|
| 1163 |
+
"eval_stsb_spearman": 0.8499622228231559,
|
| 1164 |
+
"step": 14750
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.95,
|
| 1168 |
+
"eval_avg_sts": 0.79901349058513,
|
| 1169 |
+
"eval_sickr_spearman": 0.748446895621158,
|
| 1170 |
+
"eval_stsb_spearman": 0.8495800855491022,
|
| 1171 |
+
"step": 14875
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"electra_acc": 117.6139,
|
| 1175 |
+
"electra_fix_acc": 0.9705,
|
| 1176 |
+
"electra_rep_acc": 0.7091,
|
| 1177 |
+
"epoch": 0.96,
|
| 1178 |
+
"learning_rate": 3.6402150262383203e-06,
|
| 1179 |
+
"loss": 0.0014,
|
| 1180 |
+
"neg_sim": -0.0143,
|
| 1181 |
+
"pos_sim": 0.8156,
|
| 1182 |
+
"step": 15000
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 0.96,
|
| 1186 |
+
"eval_avg_sts": 0.79914478688965,
|
| 1187 |
+
"eval_sickr_spearman": 0.7468535118620501,
|
| 1188 |
+
"eval_stsb_spearman": 0.8514360619172497,
|
| 1189 |
+
"step": 15000
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 0.97,
|
| 1193 |
+
"eval_avg_sts": 0.7991571351120899,
|
| 1194 |
+
"eval_sickr_spearman": 0.7471076924508606,
|
| 1195 |
+
"eval_stsb_spearman": 0.8512065777733191,
|
| 1196 |
+
"step": 15125
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.98,
|
| 1200 |
+
"eval_avg_sts": 0.8012474062999755,
|
| 1201 |
+
"eval_sickr_spearman": 0.7502938355647578,
|
| 1202 |
+
"eval_stsb_spearman": 0.8522009770351932,
|
| 1203 |
+
"step": 15250
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 0.98,
|
| 1207 |
+
"eval_avg_sts": 0.7993104516704408,
|
| 1208 |
+
"eval_sickr_spearman": 0.7512376467080167,
|
| 1209 |
+
"eval_stsb_spearman": 0.8473832566328647,
|
| 1210 |
+
"step": 15375
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"electra_acc": 117.5407,
|
| 1214 |
+
"electra_fix_acc": 0.9702,
|
| 1215 |
+
"electra_rep_acc": 0.7085,
|
| 1216 |
+
"epoch": 0.99,
|
| 1217 |
+
"learning_rate": 3.5282221937795976e-06,
|
| 1218 |
+
"loss": 0.0015,
|
| 1219 |
+
"neg_sim": -0.0143,
|
| 1220 |
+
"pos_sim": 0.8109,
|
| 1221 |
+
"step": 15500
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 0.99,
|
| 1225 |
+
"eval_avg_sts": 0.7991830405974262,
|
| 1226 |
+
"eval_sickr_spearman": 0.7510609883169274,
|
| 1227 |
+
"eval_stsb_spearman": 0.847305092877925,
|
| 1228 |
+
"step": 15500
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 1.0,
|
| 1232 |
+
"eval_avg_sts": 0.7993142209919086,
|
| 1233 |
+
"eval_sickr_spearman": 0.750953302587503,
|
| 1234 |
+
"eval_stsb_spearman": 0.8476751393963143,
|
| 1235 |
+
"step": 15625
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 1.01,
|
| 1239 |
+
"eval_avg_sts": 0.79998042307191,
|
| 1240 |
+
"eval_sickr_spearman": 0.7521459148362114,
|
| 1241 |
+
"eval_stsb_spearman": 0.8478149313076085,
|
| 1242 |
+
"step": 15750
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.02,
|
| 1246 |
+
"eval_avg_sts": 0.8000408218838944,
|
| 1247 |
+
"eval_sickr_spearman": 0.7522030718469228,
|
| 1248 |
+
"eval_stsb_spearman": 0.847878571920866,
|
| 1249 |
+
"step": 15875
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"electra_acc": NaN,
|
| 1253 |
+
"electra_fix_acc": NaN,
|
| 1254 |
+
"electra_rep_acc": NaN,
|
| 1255 |
+
"epoch": 1.02,
|
| 1256 |
+
"learning_rate": 3.4162293613208757e-06,
|
| 1257 |
+
"loss": 0.0013,
|
| 1258 |
+
"neg_sim": NaN,
|
| 1259 |
+
"pos_sim": 0.8075,
|
| 1260 |
+
"step": 16000
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 1.02,
|
| 1264 |
+
"eval_avg_sts": 0.7998575815186055,
|
| 1265 |
+
"eval_sickr_spearman": 0.7520637336216508,
|
| 1266 |
+
"eval_stsb_spearman": 0.8476514294155602,
|
| 1267 |
+
"step": 16000
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 1.03,
|
| 1271 |
+
"eval_avg_sts": 0.7998976758754202,
|
| 1272 |
+
"eval_sickr_spearman": 0.752151678568384,
|
| 1273 |
+
"eval_stsb_spearman": 0.8476436731824563,
|
| 1274 |
+
"step": 16125
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 1.04,
|
| 1278 |
+
"eval_avg_sts": 0.7993375115090717,
|
| 1279 |
+
"eval_sickr_spearman": 0.7512065225542848,
|
| 1280 |
+
"eval_stsb_spearman": 0.8474685004638588,
|
| 1281 |
+
"step": 16250
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 1.05,
|
| 1285 |
+
"eval_avg_sts": 0.8005700206739319,
|
| 1286 |
+
"eval_sickr_spearman": 0.7500328285448524,
|
| 1287 |
+
"eval_stsb_spearman": 0.8511072128030113,
|
| 1288 |
+
"step": 16375
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"electra_acc": NaN,
|
| 1292 |
+
"electra_fix_acc": NaN,
|
| 1293 |
+
"electra_rep_acc": NaN,
|
| 1294 |
+
"epoch": 1.06,
|
| 1295 |
+
"learning_rate": 3.304236528862153e-06,
|
| 1296 |
+
"loss": 0.0014,
|
| 1297 |
+
"neg_sim": NaN,
|
| 1298 |
+
"pos_sim": 0.8139,
|
| 1299 |
+
"step": 16500
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 1.06,
|
| 1303 |
+
"eval_avg_sts": 0.800932566485014,
|
| 1304 |
+
"eval_sickr_spearman": 0.7492046823085472,
|
| 1305 |
+
"eval_stsb_spearman": 0.8526604506614809,
|
| 1306 |
+
"step": 16500
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 1.06,
|
| 1310 |
+
"eval_avg_sts": 0.8020790163363096,
|
| 1311 |
+
"eval_sickr_spearman": 0.7523164732774181,
|
| 1312 |
+
"eval_stsb_spearman": 0.851841559395201,
|
| 1313 |
+
"step": 16625
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 1.07,
|
| 1317 |
+
"eval_avg_sts": 0.8034376489964099,
|
| 1318 |
+
"eval_sickr_spearman": 0.7546485273455429,
|
| 1319 |
+
"eval_stsb_spearman": 0.8522267706472768,
|
| 1320 |
+
"step": 16750
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 1.08,
|
| 1324 |
+
"eval_avg_sts": 0.8035121251749653,
|
| 1325 |
+
"eval_sickr_spearman": 0.7551173589266803,
|
| 1326 |
+
"eval_stsb_spearman": 0.8519068914232504,
|
| 1327 |
+
"step": 16875
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"electra_acc": NaN,
|
| 1331 |
+
"electra_fix_acc": NaN,
|
| 1332 |
+
"electra_rep_acc": NaN,
|
| 1333 |
+
"epoch": 1.09,
|
| 1334 |
+
"learning_rate": 3.19224369640343e-06,
|
| 1335 |
+
"loss": 0.0015,
|
| 1336 |
+
"neg_sim": NaN,
|
| 1337 |
+
"pos_sim": 0.8123,
|
| 1338 |
+
"step": 17000
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 1.09,
|
| 1342 |
+
"eval_avg_sts": 0.8034498780822419,
|
| 1343 |
+
"eval_sickr_spearman": 0.755224660407293,
|
| 1344 |
+
"eval_stsb_spearman": 0.8516750957571909,
|
| 1345 |
+
"step": 17000
|
| 1346 |
+
},
|
| 1347 |
+
{
|
| 1348 |
+
"epoch": 1.1,
|
| 1349 |
+
"eval_avg_sts": 0.8032831913725695,
|
| 1350 |
+
"eval_sickr_spearman": 0.7549221124993344,
|
| 1351 |
+
"eval_stsb_spearman": 0.8516442702458048,
|
| 1352 |
+
"step": 17125
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 1.1,
|
| 1356 |
+
"eval_avg_sts": 0.8025478170596181,
|
| 1357 |
+
"eval_sickr_spearman": 0.7531311768200115,
|
| 1358 |
+
"eval_stsb_spearman": 0.8519644572992248,
|
| 1359 |
+
"step": 17250
|
| 1360 |
+
},
|
| 1361 |
+
{
|
| 1362 |
+
"epoch": 1.11,
|
| 1363 |
+
"eval_avg_sts": 0.8023474610208348,
|
| 1364 |
+
"eval_sickr_spearman": 0.7529551428332407,
|
| 1365 |
+
"eval_stsb_spearman": 0.8517397792084288,
|
| 1366 |
+
"step": 17375
|
| 1367 |
+
},
|
| 1368 |
+
{
|
| 1369 |
+
"electra_acc": NaN,
|
| 1370 |
+
"electra_fix_acc": NaN,
|
| 1371 |
+
"electra_rep_acc": NaN,
|
| 1372 |
+
"epoch": 1.12,
|
| 1373 |
+
"learning_rate": 3.0802508639447074e-06,
|
| 1374 |
+
"loss": 0.0015,
|
| 1375 |
+
"neg_sim": NaN,
|
| 1376 |
+
"pos_sim": 0.8124,
|
| 1377 |
+
"step": 17500
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 1.12,
|
| 1381 |
+
"eval_avg_sts": 0.8036377678015303,
|
| 1382 |
+
"eval_sickr_spearman": 0.7540948728392658,
|
| 1383 |
+
"eval_stsb_spearman": 0.8531806627637948,
|
| 1384 |
+
"step": 17500
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 1.13,
|
| 1388 |
+
"eval_avg_sts": 0.803436785093961,
|
| 1389 |
+
"eval_sickr_spearman": 0.7537820462755993,
|
| 1390 |
+
"eval_stsb_spearman": 0.8530915239123227,
|
| 1391 |
+
"step": 17625
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 1.14,
|
| 1395 |
+
"eval_avg_sts": 0.8034543332861956,
|
| 1396 |
+
"eval_sickr_spearman": 0.753641363179487,
|
| 1397 |
+
"eval_stsb_spearman": 0.8532673033929044,
|
| 1398 |
+
"step": 17750
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 1.14,
|
| 1402 |
+
"eval_avg_sts": 0.8031959700775069,
|
| 1403 |
+
"eval_sickr_spearman": 0.7533919857008204,
|
| 1404 |
+
"eval_stsb_spearman": 0.8529999544541933,
|
| 1405 |
+
"step": 17875
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"electra_acc": NaN,
|
| 1409 |
+
"electra_fix_acc": NaN,
|
| 1410 |
+
"electra_rep_acc": NaN,
|
| 1411 |
+
"epoch": 1.15,
|
| 1412 |
+
"learning_rate": 2.9682580314859847e-06,
|
| 1413 |
+
"loss": 0.0013,
|
| 1414 |
+
"neg_sim": NaN,
|
| 1415 |
+
"pos_sim": 0.8138,
|
| 1416 |
+
"step": 18000
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.15,
|
| 1420 |
+
"eval_avg_sts": 0.8033449509317736,
|
| 1421 |
+
"eval_sickr_spearman": 0.7535802676184579,
|
| 1422 |
+
"eval_stsb_spearman": 0.8531096342450895,
|
| 1423 |
+
"step": 18000
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 1.16,
|
| 1427 |
+
"eval_avg_sts": 0.803034323911963,
|
| 1428 |
+
"eval_sickr_spearman": 0.7532389586116386,
|
| 1429 |
+
"eval_stsb_spearman": 0.8528296892122876,
|
| 1430 |
+
"step": 18125
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 1.17,
|
| 1434 |
+
"eval_avg_sts": 0.8030336070807917,
|
| 1435 |
+
"eval_sickr_spearman": 0.753408940679628,
|
| 1436 |
+
"eval_stsb_spearman": 0.8526582734819556,
|
| 1437 |
+
"step": 18250
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 1.18,
|
| 1441 |
+
"eval_avg_sts": 0.8042466753465274,
|
| 1442 |
+
"eval_sickr_spearman": 0.754791275779017,
|
| 1443 |
+
"eval_stsb_spearman": 0.8537020749140378,
|
| 1444 |
+
"step": 18375
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"electra_acc": NaN,
|
| 1448 |
+
"electra_fix_acc": NaN,
|
| 1449 |
+
"electra_rep_acc": NaN,
|
| 1450 |
+
"epoch": 1.18,
|
| 1451 |
+
"learning_rate": 2.8562651990272623e-06,
|
| 1452 |
+
"loss": 0.0013,
|
| 1453 |
+
"neg_sim": NaN,
|
| 1454 |
+
"pos_sim": 0.8154,
|
| 1455 |
+
"step": 18500
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 1.18,
|
| 1459 |
+
"eval_avg_sts": 0.8041866669158593,
|
| 1460 |
+
"eval_sickr_spearman": 0.7544020317329624,
|
| 1461 |
+
"eval_stsb_spearman": 0.8539713020987562,
|
| 1462 |
+
"step": 18500
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 1.19,
|
| 1466 |
+
"eval_avg_sts": 0.8040941868331957,
|
| 1467 |
+
"eval_sickr_spearman": 0.7544721091099607,
|
| 1468 |
+
"eval_stsb_spearman": 0.8537162645564307,
|
| 1469 |
+
"step": 18625
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 1.2,
|
| 1473 |
+
"eval_avg_sts": 0.8041380482719174,
|
| 1474 |
+
"eval_sickr_spearman": 0.7546320526777496,
|
| 1475 |
+
"eval_stsb_spearman": 0.8536440438660853,
|
| 1476 |
+
"step": 18750
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 1.21,
|
| 1480 |
+
"eval_avg_sts": 0.8038765751251531,
|
| 1481 |
+
"eval_sickr_spearman": 0.7544602934590069,
|
| 1482 |
+
"eval_stsb_spearman": 0.8532928567912992,
|
| 1483 |
+
"step": 18875
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"electra_acc": NaN,
|
| 1487 |
+
"electra_fix_acc": NaN,
|
| 1488 |
+
"electra_rep_acc": NaN,
|
| 1489 |
+
"epoch": 1.22,
|
| 1490 |
+
"learning_rate": 2.7442723665685396e-06,
|
| 1491 |
+
"loss": 0.0014,
|
| 1492 |
+
"neg_sim": NaN,
|
| 1493 |
+
"pos_sim": 0.8165,
|
| 1494 |
+
"step": 19000
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 1.22,
|
| 1498 |
+
"eval_avg_sts": 0.8066267695432612,
|
| 1499 |
+
"eval_sickr_spearman": 0.7576009030887407,
|
| 1500 |
+
"eval_stsb_spearman": 0.8556526359977817,
|
| 1501 |
+
"step": 19000
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 1.22,
|
| 1505 |
+
"eval_avg_sts": 0.8079773606909346,
|
| 1506 |
+
"eval_sickr_spearman": 0.7579651709620474,
|
| 1507 |
+
"eval_stsb_spearman": 0.8579895504198217,
|
| 1508 |
+
"step": 19125
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 1.23,
|
| 1512 |
+
"eval_avg_sts": 0.8076160011490612,
|
| 1513 |
+
"eval_sickr_spearman": 0.7573030141976216,
|
| 1514 |
+
"eval_stsb_spearman": 0.8579289881005008,
|
| 1515 |
+
"step": 19250
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 1.24,
|
| 1519 |
+
"eval_avg_sts": 0.8076237482649742,
|
| 1520 |
+
"eval_sickr_spearman": 0.7578992242597729,
|
| 1521 |
+
"eval_stsb_spearman": 0.8573482722701754,
|
| 1522 |
+
"step": 19375
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"electra_acc": NaN,
|
| 1526 |
+
"electra_fix_acc": NaN,
|
| 1527 |
+
"electra_rep_acc": NaN,
|
| 1528 |
+
"epoch": 1.25,
|
| 1529 |
+
"learning_rate": 2.632279534109817e-06,
|
| 1530 |
+
"loss": 0.0014,
|
| 1531 |
+
"neg_sim": NaN,
|
| 1532 |
+
"pos_sim": 0.817,
|
| 1533 |
+
"step": 19500
|
| 1534 |
+
},
|
| 1535 |
+
{
|
| 1536 |
+
"epoch": 1.25,
|
| 1537 |
+
"eval_avg_sts": 0.8074421455549705,
|
| 1538 |
+
"eval_sickr_spearman": 0.7572703530486437,
|
| 1539 |
+
"eval_stsb_spearman": 0.8576139380612975,
|
| 1540 |
+
"step": 19500
|
| 1541 |
+
},
|
| 1542 |
+
{
|
| 1543 |
+
"epoch": 1.26,
|
| 1544 |
+
"eval_avg_sts": 0.8072772138963991,
|
| 1545 |
+
"eval_sickr_spearman": 0.7571532532233376,
|
| 1546 |
+
"eval_stsb_spearman": 0.8574011745694605,
|
| 1547 |
+
"step": 19625
|
| 1548 |
+
},
|
| 1549 |
+
{
|
| 1550 |
+
"epoch": 1.26,
|
| 1551 |
+
"eval_avg_sts": 0.8073302425810744,
|
| 1552 |
+
"eval_sickr_spearman": 0.7571811592932731,
|
| 1553 |
+
"eval_stsb_spearman": 0.8574793258688757,
|
| 1554 |
+
"step": 19750
|
| 1555 |
+
},
|
| 1556 |
+
{
|
| 1557 |
+
"epoch": 1.27,
|
| 1558 |
+
"eval_avg_sts": 0.8072437293168806,
|
| 1559 |
+
"eval_sickr_spearman": 0.7571653090297986,
|
| 1560 |
+
"eval_stsb_spearman": 0.8573221496039627,
|
| 1561 |
+
"step": 19875
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"electra_acc": NaN,
|
| 1565 |
+
"electra_fix_acc": NaN,
|
| 1566 |
+
"electra_rep_acc": NaN,
|
| 1567 |
+
"epoch": 1.28,
|
| 1568 |
+
"learning_rate": 2.5202867016510945e-06,
|
| 1569 |
+
"loss": 0.0013,
|
| 1570 |
+
"neg_sim": NaN,
|
| 1571 |
+
"pos_sim": 0.819,
|
| 1572 |
+
"step": 20000
|
| 1573 |
+
},
|
| 1574 |
+
{
|
| 1575 |
+
"epoch": 1.28,
|
| 1576 |
+
"eval_avg_sts": 0.8077588380658864,
|
| 1577 |
+
"eval_sickr_spearman": 0.7571885560828946,
|
| 1578 |
+
"eval_stsb_spearman": 0.8583291200488784,
|
| 1579 |
+
"step": 20000
|
| 1580 |
+
},
|
| 1581 |
+
{
|
| 1582 |
+
"epoch": 1.29,
|
| 1583 |
+
"eval_avg_sts": 0.8073697057890559,
|
| 1584 |
+
"eval_sickr_spearman": 0.7569122331563212,
|
| 1585 |
+
"eval_stsb_spearman": 0.8578271784217905,
|
| 1586 |
+
"step": 20125
|
| 1587 |
+
},
|
| 1588 |
+
{
|
| 1589 |
+
"epoch": 1.3,
|
| 1590 |
+
"eval_avg_sts": 0.8072112242255804,
|
| 1591 |
+
"eval_sickr_spearman": 0.7568430683702502,
|
| 1592 |
+
"eval_stsb_spearman": 0.8575793800809108,
|
| 1593 |
+
"step": 20250
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 1.3,
|
| 1597 |
+
"eval_avg_sts": 0.8070283089413652,
|
| 1598 |
+
"eval_sickr_spearman": 0.7565875429105995,
|
| 1599 |
+
"eval_stsb_spearman": 0.8574690749721309,
|
| 1600 |
+
"step": 20375
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"electra_acc": NaN,
|
| 1604 |
+
"electra_fix_acc": NaN,
|
| 1605 |
+
"electra_rep_acc": NaN,
|
| 1606 |
+
"epoch": 1.31,
|
| 1607 |
+
"learning_rate": 2.4082938691923718e-06,
|
| 1608 |
+
"loss": 0.0013,
|
| 1609 |
+
"neg_sim": NaN,
|
| 1610 |
+
"pos_sim": 0.8202,
|
| 1611 |
+
"step": 20500
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 1.31,
|
| 1615 |
+
"eval_avg_sts": 0.806707361091437,
|
| 1616 |
+
"eval_sickr_spearman": 0.7562538228178075,
|
| 1617 |
+
"eval_stsb_spearman": 0.8571608993650666,
|
| 1618 |
+
"step": 20500
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 1.32,
|
| 1622 |
+
"eval_avg_sts": 0.8065118723390399,
|
| 1623 |
+
"eval_sickr_spearman": 0.7561714014477396,
|
| 1624 |
+
"eval_stsb_spearman": 0.85685234323034,
|
| 1625 |
+
"step": 20625
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 1.33,
|
| 1629 |
+
"eval_avg_sts": 0.8067084472431667,
|
| 1630 |
+
"eval_sickr_spearman": 0.756075627431472,
|
| 1631 |
+
"eval_stsb_spearman": 0.8573412670548614,
|
| 1632 |
+
"step": 20750
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 1.34,
|
| 1636 |
+
"eval_avg_sts": 0.8066874969794837,
|
| 1637 |
+
"eval_sickr_spearman": 0.7560982500802493,
|
| 1638 |
+
"eval_stsb_spearman": 0.857276743878718,
|
| 1639 |
+
"step": 20875
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"electra_acc": NaN,
|
| 1643 |
+
"electra_fix_acc": NaN,
|
| 1644 |
+
"electra_rep_acc": NaN,
|
| 1645 |
+
"epoch": 1.34,
|
| 1646 |
+
"learning_rate": 2.296301036733649e-06,
|
| 1647 |
+
"loss": 0.0013,
|
| 1648 |
+
"neg_sim": NaN,
|
| 1649 |
+
"pos_sim": 0.8217,
|
| 1650 |
+
"step": 21000
|
| 1651 |
+
},
|
| 1652 |
+
{
|
| 1653 |
+
"epoch": 1.34,
|
| 1654 |
+
"eval_avg_sts": 0.80597633722049,
|
| 1655 |
+
"eval_sickr_spearman": 0.7559074225142358,
|
| 1656 |
+
"eval_stsb_spearman": 0.8560452519267441,
|
| 1657 |
+
"step": 21000
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 1.35,
|
| 1661 |
+
"eval_avg_sts": 0.8058007667522757,
|
| 1662 |
+
"eval_sickr_spearman": 0.7558859045807915,
|
| 1663 |
+
"eval_stsb_spearman": 0.8557156289237601,
|
| 1664 |
+
"step": 21125
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 1.36,
|
| 1668 |
+
"eval_avg_sts": 0.8052139487149964,
|
| 1669 |
+
"eval_sickr_spearman": 0.7551075125508855,
|
| 1670 |
+
"eval_stsb_spearman": 0.8553203848791073,
|
| 1671 |
+
"step": 21250
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"epoch": 1.37,
|
| 1675 |
+
"eval_avg_sts": 0.8050337032021436,
|
| 1676 |
+
"eval_sickr_spearman": 0.7547609681540094,
|
| 1677 |
+
"eval_stsb_spearman": 0.8553064382502779,
|
| 1678 |
+
"step": 21375
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"electra_acc": NaN,
|
| 1682 |
+
"electra_fix_acc": NaN,
|
| 1683 |
+
"electra_rep_acc": NaN,
|
| 1684 |
+
"epoch": 1.38,
|
| 1685 |
+
"learning_rate": 2.1843082042749267e-06,
|
| 1686 |
+
"loss": 0.0014,
|
| 1687 |
+
"neg_sim": NaN,
|
| 1688 |
+
"pos_sim": 0.8233,
|
| 1689 |
+
"step": 21500
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 1.38,
|
| 1693 |
+
"eval_avg_sts": 0.8048704441406708,
|
| 1694 |
+
"eval_sickr_spearman": 0.7537258018558153,
|
| 1695 |
+
"eval_stsb_spearman": 0.8560150864255263,
|
| 1696 |
+
"step": 21500
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 1.38,
|
| 1700 |
+
"eval_avg_sts": 0.8048149155952563,
|
| 1701 |
+
"eval_sickr_spearman": 0.7537628338350241,
|
| 1702 |
+
"eval_stsb_spearman": 0.8558669973554885,
|
| 1703 |
+
"step": 21625
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 1.39,
|
| 1707 |
+
"eval_avg_sts": 0.8049705872600981,
|
| 1708 |
+
"eval_sickr_spearman": 0.7530663348330701,
|
| 1709 |
+
"eval_stsb_spearman": 0.8568748396871261,
|
| 1710 |
+
"step": 21750
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 1.4,
|
| 1714 |
+
"eval_avg_sts": 0.8048803781742058,
|
| 1715 |
+
"eval_sickr_spearman": 0.7529998117575781,
|
| 1716 |
+
"eval_stsb_spearman": 0.8567609445908335,
|
| 1717 |
+
"step": 21875
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"electra_acc": NaN,
|
| 1721 |
+
"electra_fix_acc": NaN,
|
| 1722 |
+
"electra_rep_acc": NaN,
|
| 1723 |
+
"epoch": 1.41,
|
| 1724 |
+
"learning_rate": 2.072315371816204e-06,
|
| 1725 |
+
"loss": 0.0014,
|
| 1726 |
+
"neg_sim": NaN,
|
| 1727 |
+
"pos_sim": 0.8232,
|
| 1728 |
+
"step": 22000
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 1.41,
|
| 1732 |
+
"eval_avg_sts": 0.8048396134318049,
|
| 1733 |
+
"eval_sickr_spearman": 0.7529523570293574,
|
| 1734 |
+
"eval_stsb_spearman": 0.8567268698342524,
|
| 1735 |
+
"step": 22000
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 1.42,
|
| 1739 |
+
"eval_avg_sts": 0.8052282014222374,
|
| 1740 |
+
"eval_sickr_spearman": 0.754343433789208,
|
| 1741 |
+
"eval_stsb_spearman": 0.8561129690552667,
|
| 1742 |
+
"step": 22125
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 1.42,
|
| 1746 |
+
"eval_avg_sts": 0.8052272637457882,
|
| 1747 |
+
"eval_sickr_spearman": 0.7546001119952933,
|
| 1748 |
+
"eval_stsb_spearman": 0.855854415496283,
|
| 1749 |
+
"step": 22250
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 1.43,
|
| 1753 |
+
"eval_avg_sts": 0.8050641056280918,
|
| 1754 |
+
"eval_sickr_spearman": 0.7546412266181242,
|
| 1755 |
+
"eval_stsb_spearman": 0.8554869846380595,
|
| 1756 |
+
"step": 22375
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"electra_acc": NaN,
|
| 1760 |
+
"electra_fix_acc": NaN,
|
| 1761 |
+
"electra_rep_acc": NaN,
|
| 1762 |
+
"epoch": 1.44,
|
| 1763 |
+
"learning_rate": 1.9603225393574807e-06,
|
| 1764 |
+
"loss": 0.0014,
|
| 1765 |
+
"neg_sim": NaN,
|
| 1766 |
+
"pos_sim": 0.8233,
|
| 1767 |
+
"step": 22500
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 1.44,
|
| 1771 |
+
"eval_avg_sts": 0.8049566807526001,
|
| 1772 |
+
"eval_sickr_spearman": 0.7546373841300092,
|
| 1773 |
+
"eval_stsb_spearman": 0.8552759773751909,
|
| 1774 |
+
"step": 22500
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 1.45,
|
| 1778 |
+
"eval_avg_sts": 0.8022531756326772,
|
| 1779 |
+
"eval_sickr_spearman": 0.7524902498024213,
|
| 1780 |
+
"eval_stsb_spearman": 0.8520161014629332,
|
| 1781 |
+
"step": 22625
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 1.46,
|
| 1785 |
+
"eval_avg_sts": 0.80224163626539,
|
| 1786 |
+
"eval_sickr_spearman": 0.7525507689902333,
|
| 1787 |
+
"eval_stsb_spearman": 0.8519325035405467,
|
| 1788 |
+
"step": 22750
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 1.46,
|
| 1792 |
+
"eval_avg_sts": 0.8021754313869522,
|
| 1793 |
+
"eval_sickr_spearman": 0.7525084055587649,
|
| 1794 |
+
"eval_stsb_spearman": 0.8518424572151397,
|
| 1795 |
+
"step": 22875
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"electra_acc": NaN,
|
| 1799 |
+
"electra_fix_acc": NaN,
|
| 1800 |
+
"electra_rep_acc": NaN,
|
| 1801 |
+
"epoch": 1.47,
|
| 1802 |
+
"learning_rate": 1.8483297068987584e-06,
|
| 1803 |
+
"loss": 0.0015,
|
| 1804 |
+
"neg_sim": NaN,
|
| 1805 |
+
"pos_sim": 0.8241,
|
| 1806 |
+
"step": 23000
|
| 1807 |
+
},
|
| 1808 |
+
{
|
| 1809 |
+
"epoch": 1.47,
|
| 1810 |
+
"eval_avg_sts": 0.8024969817513419,
|
| 1811 |
+
"eval_sickr_spearman": 0.752065078492491,
|
| 1812 |
+
"eval_stsb_spearman": 0.8529288850101928,
|
| 1813 |
+
"step": 23000
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 1.48,
|
| 1817 |
+
"eval_avg_sts": 0.8024605472785751,
|
| 1818 |
+
"eval_sickr_spearman": 0.7519645974282826,
|
| 1819 |
+
"eval_stsb_spearman": 0.8529564971288677,
|
| 1820 |
+
"step": 23125
|
| 1821 |
+
},
|
| 1822 |
+
{
|
| 1823 |
+
"epoch": 1.49,
|
| 1824 |
+
"eval_avg_sts": 0.8024320134196474,
|
| 1825 |
+
"eval_sickr_spearman": 0.751913444305251,
|
| 1826 |
+
"eval_stsb_spearman": 0.8529505825340439,
|
| 1827 |
+
"step": 23250
|
| 1828 |
+
},
|
| 1829 |
+
{
|
| 1830 |
+
"epoch": 1.5,
|
| 1831 |
+
"eval_avg_sts": 0.8024048624411102,
|
| 1832 |
+
"eval_sickr_spearman": 0.7518156049516216,
|
| 1833 |
+
"eval_stsb_spearman": 0.8529941199305989,
|
| 1834 |
+
"step": 23375
|
| 1835 |
+
},
|
| 1836 |
+
{
|
| 1837 |
+
"electra_acc": NaN,
|
| 1838 |
+
"electra_fix_acc": NaN,
|
| 1839 |
+
"electra_rep_acc": NaN,
|
| 1840 |
+
"epoch": 1.5,
|
| 1841 |
+
"learning_rate": 1.7363368744400357e-06,
|
| 1842 |
+
"loss": 0.0013,
|
| 1843 |
+
"neg_sim": NaN,
|
| 1844 |
+
"pos_sim": 0.8244,
|
| 1845 |
+
"step": 23500
|
| 1846 |
+
},
|
| 1847 |
+
{
|
| 1848 |
+
"epoch": 1.5,
|
| 1849 |
+
"eval_avg_sts": 0.802379657071767,
|
| 1850 |
+
"eval_sickr_spearman": 0.7517114254926023,
|
| 1851 |
+
"eval_stsb_spearman": 0.8530478886509316,
|
| 1852 |
+
"step": 23500
|
| 1853 |
+
},
|
| 1854 |
+
{
|
| 1855 |
+
"epoch": 1.51,
|
| 1856 |
+
"eval_avg_sts": 0.8022866209294395,
|
| 1857 |
+
"eval_sickr_spearman": 0.7517222805215273,
|
| 1858 |
+
"eval_stsb_spearman": 0.8528509613373516,
|
| 1859 |
+
"step": 23625
|
| 1860 |
+
},
|
| 1861 |
+
{
|
| 1862 |
+
"epoch": 1.52,
|
| 1863 |
+
"eval_avg_sts": 0.8029982809362839,
|
| 1864 |
+
"eval_sickr_spearman": 0.7519777099189752,
|
| 1865 |
+
"eval_stsb_spearman": 0.8540188519535926,
|
| 1866 |
+
"step": 23750
|
| 1867 |
+
},
|
| 1868 |
+
{
|
| 1869 |
+
"epoch": 1.53,
|
| 1870 |
+
"eval_avg_sts": 0.8029407406070306,
|
| 1871 |
+
"eval_sickr_spearman": 0.7520159907068213,
|
| 1872 |
+
"eval_stsb_spearman": 0.8538654905072398,
|
| 1873 |
+
"step": 23875
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"electra_acc": NaN,
|
| 1877 |
+
"electra_fix_acc": NaN,
|
| 1878 |
+
"electra_rep_acc": NaN,
|
| 1879 |
+
"epoch": 1.54,
|
| 1880 |
+
"learning_rate": 1.6243440419813131e-06,
|
| 1881 |
+
"loss": 0.0014,
|
| 1882 |
+
"neg_sim": NaN,
|
| 1883 |
+
"pos_sim": 0.8241,
|
| 1884 |
+
"step": 24000
|
| 1885 |
+
},
|
| 1886 |
+
{
|
| 1887 |
+
"epoch": 1.54,
|
| 1888 |
+
"eval_avg_sts": 0.8026597102203472,
|
| 1889 |
+
"eval_sickr_spearman": 0.7517222805215273,
|
| 1890 |
+
"eval_stsb_spearman": 0.8535971399191672,
|
| 1891 |
+
"step": 24000
|
| 1892 |
+
},
|
| 1893 |
+
{
|
| 1894 |
+
"epoch": 1.54,
|
| 1895 |
+
"eval_avg_sts": 0.8025235294672179,
|
| 1896 |
+
"eval_sickr_spearman": 0.7517699273741539,
|
| 1897 |
+
"eval_stsb_spearman": 0.8532771315602818,
|
| 1898 |
+
"step": 24125
|
| 1899 |
+
},
|
| 1900 |
+
{
|
| 1901 |
+
"epoch": 1.55,
|
| 1902 |
+
"eval_avg_sts": 0.8024910901785833,
|
| 1903 |
+
"eval_sickr_spearman": 0.7517942311114816,
|
| 1904 |
+
"eval_stsb_spearman": 0.8531879492456851,
|
| 1905 |
+
"step": 24250
|
| 1906 |
+
},
|
| 1907 |
+
{
|
| 1908 |
+
"epoch": 1.56,
|
| 1909 |
+
"eval_avg_sts": 0.8022646880892426,
|
| 1910 |
+
"eval_sickr_spearman": 0.7516456709147336,
|
| 1911 |
+
"eval_stsb_spearman": 0.8528837052637515,
|
| 1912 |
+
"step": 24375
|
| 1913 |
+
},
|
| 1914 |
+
{
|
| 1915 |
+
"electra_acc": NaN,
|
| 1916 |
+
"electra_fix_acc": NaN,
|
| 1917 |
+
"electra_rep_acc": NaN,
|
| 1918 |
+
"epoch": 1.57,
|
| 1919 |
+
"learning_rate": 1.5123512095225906e-06,
|
| 1920 |
+
"loss": 0.0013,
|
| 1921 |
+
"neg_sim": NaN,
|
| 1922 |
+
"pos_sim": 0.8242,
|
| 1923 |
+
"step": 24500
|
| 1924 |
+
},
|
| 1925 |
+
{
|
| 1926 |
+
"epoch": 1.57,
|
| 1927 |
+
"eval_avg_sts": 0.8021439174552474,
|
| 1928 |
+
"eval_sickr_spearman": 0.7515729518271562,
|
| 1929 |
+
"eval_stsb_spearman": 0.8527148830833385,
|
| 1930 |
+
"step": 24500
|
| 1931 |
+
},
|
| 1932 |
+
{
|
| 1933 |
+
"epoch": 1.58,
|
| 1934 |
+
"eval_avg_sts": 0.8023944530487281,
|
| 1935 |
+
"eval_sickr_spearman": 0.7522279999885693,
|
| 1936 |
+
"eval_stsb_spearman": 0.8525609061088869,
|
| 1937 |
+
"step": 24625
|
| 1938 |
+
},
|
| 1939 |
+
{
|
| 1940 |
+
"epoch": 1.58,
|
| 1941 |
+
"eval_avg_sts": 0.8023338277994898,
|
| 1942 |
+
"eval_sickr_spearman": 0.7522989419253933,
|
| 1943 |
+
"eval_stsb_spearman": 0.8523687136735865,
|
| 1944 |
+
"step": 24750
|
| 1945 |
+
},
|
| 1946 |
+
{
|
| 1947 |
+
"epoch": 1.59,
|
| 1948 |
+
"eval_avg_sts": 0.8021448888713669,
|
| 1949 |
+
"eval_sickr_spearman": 0.7524111906094542,
|
| 1950 |
+
"eval_stsb_spearman": 0.8518785871332797,
|
| 1951 |
+
"step": 24875
|
| 1952 |
+
},
|
| 1953 |
+
{
|
| 1954 |
+
"electra_acc": NaN,
|
| 1955 |
+
"electra_fix_acc": NaN,
|
| 1956 |
+
"electra_rep_acc": NaN,
|
| 1957 |
+
"epoch": 1.6,
|
| 1958 |
+
"learning_rate": 1.4003583770638678e-06,
|
| 1959 |
+
"loss": 0.0014,
|
| 1960 |
+
"neg_sim": NaN,
|
| 1961 |
+
"pos_sim": 0.825,
|
| 1962 |
+
"step": 25000
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 1.6,
|
| 1966 |
+
"eval_avg_sts": 0.8020245919486174,
|
| 1967 |
+
"eval_sickr_spearman": 0.7524232944470166,
|
| 1968 |
+
"eval_stsb_spearman": 0.8516258894502182,
|
| 1969 |
+
"step": 25000
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 1.61,
|
| 1973 |
+
"eval_avg_sts": 0.8027255384898491,
|
| 1974 |
+
"eval_sickr_spearman": 0.753412110732323,
|
| 1975 |
+
"eval_stsb_spearman": 0.8520389662473753,
|
| 1976 |
+
"step": 25125
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 1.62,
|
| 1980 |
+
"eval_avg_sts": 0.8025519191789285,
|
| 1981 |
+
"eval_sickr_spearman": 0.7524277133083489,
|
| 1982 |
+
"eval_stsb_spearman": 0.8526761250495083,
|
| 1983 |
+
"step": 25250
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 1.62,
|
| 1987 |
+
"eval_avg_sts": 0.8020490496082332,
|
| 1988 |
+
"eval_sickr_spearman": 0.7522735334727325,
|
| 1989 |
+
"eval_stsb_spearman": 0.8518245657437339,
|
| 1990 |
+
"step": 25375
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"electra_acc": NaN,
|
| 1994 |
+
"electra_fix_acc": NaN,
|
| 1995 |
+
"electra_rep_acc": NaN,
|
| 1996 |
+
"epoch": 1.63,
|
| 1997 |
+
"learning_rate": 1.2883655446051453e-06,
|
| 1998 |
+
"loss": 0.0014,
|
| 1999 |
+
"neg_sim": NaN,
|
| 2000 |
+
"pos_sim": 0.8247,
|
| 2001 |
+
"step": 25500
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 1.63,
|
| 2005 |
+
"eval_avg_sts": 0.8019816352585948,
|
| 2006 |
+
"eval_sickr_spearman": 0.7522344361561619,
|
| 2007 |
+
"eval_stsb_spearman": 0.8517288343610276,
|
| 2008 |
+
"step": 25500
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 1.64,
|
| 2012 |
+
"eval_avg_sts": 0.8013792896117395,
|
| 2013 |
+
"eval_sickr_spearman": 0.7513900974239813,
|
| 2014 |
+
"eval_stsb_spearman": 0.8513684817994976,
|
| 2015 |
+
"step": 25625
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 1.65,
|
| 2019 |
+
"eval_avg_sts": 0.8012957662635658,
|
| 2020 |
+
"eval_sickr_spearman": 0.7513326522266613,
|
| 2021 |
+
"eval_stsb_spearman": 0.8512588803004703,
|
| 2022 |
+
"step": 25750
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 1.66,
|
| 2026 |
+
"eval_avg_sts": 0.8012601775924122,
|
| 2027 |
+
"eval_sickr_spearman": 0.751396101311661,
|
| 2028 |
+
"eval_stsb_spearman": 0.8511242538731633,
|
| 2029 |
+
"step": 25875
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"electra_acc": NaN,
|
| 2033 |
+
"electra_fix_acc": NaN,
|
| 2034 |
+
"electra_rep_acc": NaN,
|
| 2035 |
+
"epoch": 1.66,
|
| 2036 |
+
"learning_rate": 1.1763727121464225e-06,
|
| 2037 |
+
"loss": 0.0014,
|
| 2038 |
+
"neg_sim": NaN,
|
| 2039 |
+
"pos_sim": 0.825,
|
| 2040 |
+
"step": 26000
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 1.66,
|
| 2044 |
+
"eval_avg_sts": 0.80119947875736,
|
| 2045 |
+
"eval_sickr_spearman": 0.7513846699095188,
|
| 2046 |
+
"eval_stsb_spearman": 0.8510142876052013,
|
| 2047 |
+
"step": 26000
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 1.67,
|
| 2051 |
+
"eval_avg_sts": 0.8011593246270685,
|
| 2052 |
+
"eval_sickr_spearman": 0.751327464867706,
|
| 2053 |
+
"eval_stsb_spearman": 0.8509911843864311,
|
| 2054 |
+
"step": 26125
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 1.68,
|
| 2058 |
+
"eval_avg_sts": 0.8011629421191884,
|
| 2059 |
+
"eval_sickr_spearman": 0.7513375033679065,
|
| 2060 |
+
"eval_stsb_spearman": 0.8509883808704702,
|
| 2061 |
+
"step": 26250
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 1.69,
|
| 2065 |
+
"eval_avg_sts": 0.801390135441459,
|
| 2066 |
+
"eval_sickr_spearman": 0.7513261680279671,
|
| 2067 |
+
"eval_stsb_spearman": 0.8514541028549507,
|
| 2068 |
+
"step": 26375
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"electra_acc": NaN,
|
| 2072 |
+
"electra_fix_acc": NaN,
|
| 2073 |
+
"electra_rep_acc": NaN,
|
| 2074 |
+
"epoch": 1.7,
|
| 2075 |
+
"learning_rate": 1.0643798796877e-06,
|
| 2076 |
+
"loss": 0.0013,
|
| 2077 |
+
"neg_sim": NaN,
|
| 2078 |
+
"pos_sim": 0.8251,
|
| 2079 |
+
"step": 26500
|
| 2080 |
+
},
|
| 2081 |
+
{
|
| 2082 |
+
"epoch": 1.7,
|
| 2083 |
+
"eval_avg_sts": 0.8011107242729139,
|
| 2084 |
+
"eval_sickr_spearman": 0.7510607481614202,
|
| 2085 |
+
"eval_stsb_spearman": 0.8511607003844076,
|
| 2086 |
+
"step": 26500
|
| 2087 |
+
},
|
| 2088 |
+
{
|
| 2089 |
+
"epoch": 1.7,
|
| 2090 |
+
"eval_avg_sts": 0.8017416851794723,
|
| 2091 |
+
"eval_sickr_spearman": 0.7518608022180748,
|
| 2092 |
+
"eval_stsb_spearman": 0.8516225681408698,
|
| 2093 |
+
"step": 26625
|
| 2094 |
+
},
|
| 2095 |
+
{
|
| 2096 |
+
"epoch": 1.71,
|
| 2097 |
+
"eval_avg_sts": 0.8012848966551385,
|
| 2098 |
+
"eval_sickr_spearman": 0.7518499471891499,
|
| 2099 |
+
"eval_stsb_spearman": 0.8507198461211269,
|
| 2100 |
+
"step": 26750
|
| 2101 |
+
},
|
| 2102 |
+
{
|
| 2103 |
+
"epoch": 1.72,
|
| 2104 |
+
"eval_avg_sts": 0.8018531908928693,
|
| 2105 |
+
"eval_sickr_spearman": 0.7526851600120571,
|
| 2106 |
+
"eval_stsb_spearman": 0.8510212217736816,
|
| 2107 |
+
"step": 26875
|
| 2108 |
+
},
|
| 2109 |
+
{
|
| 2110 |
+
"electra_acc": NaN,
|
| 2111 |
+
"electra_fix_acc": NaN,
|
| 2112 |
+
"electra_rep_acc": NaN,
|
| 2113 |
+
"epoch": 1.73,
|
| 2114 |
+
"learning_rate": 9.523870472289774e-07,
|
| 2115 |
+
"loss": 0.0015,
|
| 2116 |
+
"neg_sim": NaN,
|
| 2117 |
+
"pos_sim": 0.8252,
|
| 2118 |
+
"step": 27000
|
| 2119 |
+
},
|
| 2120 |
+
{
|
| 2121 |
+
"epoch": 1.73,
|
| 2122 |
+
"eval_avg_sts": 0.8022264378652083,
|
| 2123 |
+
"eval_sickr_spearman": 0.7527636908629084,
|
| 2124 |
+
"eval_stsb_spearman": 0.8516891848675082,
|
| 2125 |
+
"step": 27000
|
| 2126 |
+
},
|
| 2127 |
+
{
|
| 2128 |
+
"epoch": 1.74,
|
| 2129 |
+
"eval_avg_sts": 0.8022941628228328,
|
| 2130 |
+
"eval_sickr_spearman": 0.7528165731055918,
|
| 2131 |
+
"eval_stsb_spearman": 0.851771752540074,
|
| 2132 |
+
"step": 27125
|
| 2133 |
+
},
|
| 2134 |
+
{
|
| 2135 |
+
"epoch": 1.74,
|
| 2136 |
+
"eval_avg_sts": 0.802330897941351,
|
| 2137 |
+
"eval_sickr_spearman": 0.7527780521622385,
|
| 2138 |
+
"eval_stsb_spearman": 0.8518837437204637,
|
| 2139 |
+
"step": 27250
|
| 2140 |
+
},
|
| 2141 |
+
{
|
| 2142 |
+
"epoch": 1.75,
|
| 2143 |
+
"eval_avg_sts": 0.8023123636818033,
|
| 2144 |
+
"eval_sickr_spearman": 0.75282886906756,
|
| 2145 |
+
"eval_stsb_spearman": 0.8517958582960464,
|
| 2146 |
+
"step": 27375
|
| 2147 |
+
},
|
| 2148 |
+
{
|
| 2149 |
+
"electra_acc": NaN,
|
| 2150 |
+
"electra_fix_acc": NaN,
|
| 2151 |
+
"electra_rep_acc": NaN,
|
| 2152 |
+
"epoch": 1.76,
|
| 2153 |
+
"learning_rate": 8.403942147702547e-07,
|
| 2154 |
+
"loss": 0.0013,
|
| 2155 |
+
"neg_sim": NaN,
|
| 2156 |
+
"pos_sim": 0.8254,
|
| 2157 |
+
"step": 27500
|
| 2158 |
+
},
|
| 2159 |
+
{
|
| 2160 |
+
"epoch": 1.76,
|
| 2161 |
+
"eval_avg_sts": 0.8022805236271966,
|
| 2162 |
+
"eval_sickr_spearman": 0.7531723875050453,
|
| 2163 |
+
"eval_stsb_spearman": 0.8513886597493479,
|
| 2164 |
+
"step": 27500
|
| 2165 |
+
},
|
| 2166 |
+
{
|
| 2167 |
+
"epoch": 1.77,
|
| 2168 |
+
"eval_avg_sts": 0.8021820813348178,
|
| 2169 |
+
"eval_sickr_spearman": 0.7531067289893795,
|
| 2170 |
+
"eval_stsb_spearman": 0.8512574336802562,
|
| 2171 |
+
"step": 27625
|
| 2172 |
+
},
|
| 2173 |
+
{
|
| 2174 |
+
"epoch": 1.78,
|
| 2175 |
+
"eval_avg_sts": 0.8021045835263574,
|
| 2176 |
+
"eval_sickr_spearman": 0.7530147494301255,
|
| 2177 |
+
"eval_stsb_spearman": 0.8511944176225893,
|
| 2178 |
+
"step": 27750
|
| 2179 |
+
},
|
| 2180 |
+
{
|
| 2181 |
+
"epoch": 1.78,
|
| 2182 |
+
"eval_avg_sts": 0.8021931793671637,
|
| 2183 |
+
"eval_sickr_spearman": 0.7531494766696594,
|
| 2184 |
+
"eval_stsb_spearman": 0.8512368820646681,
|
| 2185 |
+
"step": 27875
|
| 2186 |
+
},
|
| 2187 |
+
{
|
| 2188 |
+
"electra_acc": NaN,
|
| 2189 |
+
"electra_fix_acc": NaN,
|
| 2190 |
+
"electra_rep_acc": NaN,
|
| 2191 |
+
"epoch": 1.79,
|
| 2192 |
+
"learning_rate": 7.284013823115321e-07,
|
| 2193 |
+
"loss": 0.0014,
|
| 2194 |
+
"neg_sim": NaN,
|
| 2195 |
+
"pos_sim": 0.8249,
|
| 2196 |
+
"step": 28000
|
| 2197 |
+
},
|
| 2198 |
+
{
|
| 2199 |
+
"epoch": 1.79,
|
| 2200 |
+
"eval_avg_sts": 0.8019519952437764,
|
| 2201 |
+
"eval_sickr_spearman": 0.752846640575092,
|
| 2202 |
+
"eval_stsb_spearman": 0.8510573499124608,
|
| 2203 |
+
"step": 28000
|
| 2204 |
+
},
|
| 2205 |
+
{
|
| 2206 |
+
"epoch": 1.8,
|
| 2207 |
+
"eval_avg_sts": 0.8020597161190861,
|
| 2208 |
+
"eval_sickr_spearman": 0.7531391019517488,
|
| 2209 |
+
"eval_stsb_spearman": 0.8509803302864235,
|
| 2210 |
+
"step": 28125
|
| 2211 |
+
},
|
| 2212 |
+
{
|
| 2213 |
+
"epoch": 1.81,
|
| 2214 |
+
"eval_avg_sts": 0.8017160897849254,
|
| 2215 |
+
"eval_sickr_spearman": 0.7524687799000784,
|
| 2216 |
+
"eval_stsb_spearman": 0.8509633996697724,
|
| 2217 |
+
"step": 28250
|
| 2218 |
+
},
|
| 2219 |
+
{
|
| 2220 |
+
"epoch": 1.82,
|
| 2221 |
+
"eval_avg_sts": 0.8019391531942626,
|
| 2222 |
+
"eval_sickr_spearman": 0.7527104243714136,
|
| 2223 |
+
"eval_stsb_spearman": 0.8511678820171117,
|
| 2224 |
+
"step": 28375
|
| 2225 |
+
},
|
| 2226 |
+
{
|
| 2227 |
+
"electra_acc": NaN,
|
| 2228 |
+
"electra_fix_acc": NaN,
|
| 2229 |
+
"electra_rep_acc": NaN,
|
| 2230 |
+
"epoch": 1.82,
|
| 2231 |
+
"learning_rate": 6.164085498528094e-07,
|
| 2232 |
+
"loss": 0.0014,
|
| 2233 |
+
"neg_sim": NaN,
|
| 2234 |
+
"pos_sim": 0.8248,
|
| 2235 |
+
"step": 28500
|
| 2236 |
+
},
|
| 2237 |
+
{
|
| 2238 |
+
"epoch": 1.82,
|
| 2239 |
+
"eval_avg_sts": 0.8019393193841213,
|
| 2240 |
+
"eval_sickr_spearman": 0.7527119133355582,
|
| 2241 |
+
"eval_stsb_spearman": 0.8511667254326846,
|
| 2242 |
+
"step": 28500
|
| 2243 |
+
},
|
| 2244 |
+
{
|
| 2245 |
+
"epoch": 1.83,
|
| 2246 |
+
"eval_avg_sts": 0.8019492124313456,
|
| 2247 |
+
"eval_sickr_spearman": 0.7531468829901817,
|
| 2248 |
+
"eval_stsb_spearman": 0.8507515418725097,
|
| 2249 |
+
"step": 28625
|
| 2250 |
+
},
|
| 2251 |
+
{
|
| 2252 |
+
"epoch": 1.84,
|
| 2253 |
+
"eval_avg_sts": 0.8018476761350394,
|
| 2254 |
+
"eval_sickr_spearman": 0.753012059688445,
|
| 2255 |
+
"eval_stsb_spearman": 0.8506832925816338,
|
| 2256 |
+
"step": 28750
|
| 2257 |
+
},
|
| 2258 |
+
{
|
| 2259 |
+
"epoch": 1.85,
|
| 2260 |
+
"eval_avg_sts": 0.8021637081346856,
|
| 2261 |
+
"eval_sickr_spearman": 0.7534332924480571,
|
| 2262 |
+
"eval_stsb_spearman": 0.8508941238213141,
|
| 2263 |
+
"step": 28875
|
| 2264 |
+
},
|
| 2265 |
+
{
|
| 2266 |
+
"electra_acc": NaN,
|
| 2267 |
+
"electra_fix_acc": NaN,
|
| 2268 |
+
"electra_rep_acc": NaN,
|
| 2269 |
+
"epoch": 1.86,
|
| 2270 |
+
"learning_rate": 5.044157173940868e-07,
|
| 2271 |
+
"loss": 0.0014,
|
| 2272 |
+
"neg_sim": NaN,
|
| 2273 |
+
"pos_sim": 0.8253,
|
| 2274 |
+
"step": 29000
|
| 2275 |
+
},
|
| 2276 |
+
{
|
| 2277 |
+
"epoch": 1.86,
|
| 2278 |
+
"eval_avg_sts": 0.8021424201726441,
|
| 2279 |
+
"eval_sickr_spearman": 0.7534233980411609,
|
| 2280 |
+
"eval_stsb_spearman": 0.8508614423041275,
|
| 2281 |
+
"step": 29000
|
| 2282 |
+
},
|
| 2283 |
+
{
|
| 2284 |
+
"epoch": 1.86,
|
| 2285 |
+
"eval_avg_sts": 0.8021059378722697,
|
| 2286 |
+
"eval_sickr_spearman": 0.753409276897338,
|
| 2287 |
+
"eval_stsb_spearman": 0.8508025988472012,
|
| 2288 |
+
"step": 29125
|
| 2289 |
+
},
|
| 2290 |
+
{
|
| 2291 |
+
"epoch": 1.87,
|
| 2292 |
+
"eval_avg_sts": 0.802109297825347,
|
| 2293 |
+
"eval_sickr_spearman": 0.7534563954078488,
|
| 2294 |
+
"eval_stsb_spearman": 0.8507622002428452,
|
| 2295 |
+
"step": 29250
|
| 2296 |
+
},
|
| 2297 |
+
{
|
| 2298 |
+
"epoch": 1.88,
|
| 2299 |
+
"eval_avg_sts": 0.802140899314784,
|
| 2300 |
+
"eval_sickr_spearman": 0.7535362711295405,
|
| 2301 |
+
"eval_stsb_spearman": 0.8507455275000276,
|
| 2302 |
+
"step": 29375
|
| 2303 |
+
},
|
| 2304 |
+
{
|
| 2305 |
+
"electra_acc": NaN,
|
| 2306 |
+
"electra_fix_acc": NaN,
|
| 2307 |
+
"electra_rep_acc": NaN,
|
| 2308 |
+
"epoch": 1.89,
|
| 2309 |
+
"learning_rate": 3.9242288493536413e-07,
|
| 2310 |
+
"loss": 0.0013,
|
| 2311 |
+
"neg_sim": NaN,
|
| 2312 |
+
"pos_sim": 0.8247,
|
| 2313 |
+
"step": 29500
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 1.89,
|
| 2317 |
+
"eval_avg_sts": 0.8021494286648734,
|
| 2318 |
+
"eval_sickr_spearman": 0.7535442923234806,
|
| 2319 |
+
"eval_stsb_spearman": 0.850754565006266,
|
| 2320 |
+
"step": 29500
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 1.9,
|
| 2324 |
+
"eval_avg_sts": 0.802139809482882,
|
| 2325 |
+
"eval_sickr_spearman": 0.7535292585887305,
|
| 2326 |
+
"eval_stsb_spearman": 0.8507503603770336,
|
| 2327 |
+
"step": 29625
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 1.9,
|
| 2331 |
+
"eval_avg_sts": 0.8021667358363417,
|
| 2332 |
+
"eval_sickr_spearman": 0.7535602866802595,
|
| 2333 |
+
"eval_stsb_spearman": 0.8507731849924238,
|
| 2334 |
+
"step": 29750
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 1.91,
|
| 2338 |
+
"eval_avg_sts": 0.802187610188689,
|
| 2339 |
+
"eval_sickr_spearman": 0.7535883368434994,
|
| 2340 |
+
"eval_stsb_spearman": 0.8507868835338787,
|
| 2341 |
+
"step": 29875
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"electra_acc": NaN,
|
| 2345 |
+
"electra_fix_acc": NaN,
|
| 2346 |
+
"electra_rep_acc": NaN,
|
| 2347 |
+
"epoch": 1.92,
|
| 2348 |
+
"learning_rate": 2.804300524766415e-07,
|
| 2349 |
+
"loss": 0.0013,
|
| 2350 |
+
"neg_sim": NaN,
|
| 2351 |
+
"pos_sim": 0.8251,
|
| 2352 |
+
"step": 30000
|
| 2353 |
+
},
|
| 2354 |
+
{
|
| 2355 |
+
"epoch": 1.92,
|
| 2356 |
+
"eval_avg_sts": 0.802156902456721,
|
| 2357 |
+
"eval_sickr_spearman": 0.7535403537731628,
|
| 2358 |
+
"eval_stsb_spearman": 0.8507734511402791,
|
| 2359 |
+
"step": 30000
|
| 2360 |
+
},
|
| 2361 |
+
{
|
| 2362 |
+
"epoch": 1.93,
|
| 2363 |
+
"eval_avg_sts": 0.8024228445195005,
|
| 2364 |
+
"eval_sickr_spearman": 0.7538071665416515,
|
| 2365 |
+
"eval_stsb_spearman": 0.8510385224973495,
|
| 2366 |
+
"step": 30125
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 1.94,
|
| 2370 |
+
"eval_avg_sts": 0.802384678105279,
|
| 2371 |
+
"eval_sickr_spearman": 0.753726522322337,
|
| 2372 |
+
"eval_stsb_spearman": 0.8510428338882212,
|
| 2373 |
+
"step": 30250
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 1.94,
|
| 2377 |
+
"eval_avg_sts": 0.8023422930470364,
|
| 2378 |
+
"eval_sickr_spearman": 0.7536587504382077,
|
| 2379 |
+
"eval_stsb_spearman": 0.8510258356558651,
|
| 2380 |
+
"step": 30375
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"electra_acc": NaN,
|
| 2384 |
+
"electra_fix_acc": NaN,
|
| 2385 |
+
"electra_rep_acc": NaN,
|
| 2386 |
+
"epoch": 1.95,
|
| 2387 |
+
"learning_rate": 1.6843722001791884e-07,
|
| 2388 |
+
"loss": 0.0014,
|
| 2389 |
+
"neg_sim": NaN,
|
| 2390 |
+
"pos_sim": 0.8246,
|
| 2391 |
+
"step": 30500
|
| 2392 |
+
},
|
| 2393 |
+
{
|
| 2394 |
+
"epoch": 1.95,
|
| 2395 |
+
"eval_avg_sts": 0.8023481067748586,
|
| 2396 |
+
"eval_sickr_spearman": 0.7536835344865497,
|
| 2397 |
+
"eval_stsb_spearman": 0.8510126790631675,
|
| 2398 |
+
"step": 30500
|
| 2399 |
+
},
|
| 2400 |
+
{
|
| 2401 |
+
"epoch": 1.96,
|
| 2402 |
+
"eval_avg_sts": 0.8023432918069697,
|
| 2403 |
+
"eval_sickr_spearman": 0.7536910273383741,
|
| 2404 |
+
"eval_stsb_spearman": 0.8509955562755652,
|
| 2405 |
+
"step": 30625
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 1.97,
|
| 2409 |
+
"eval_avg_sts": 0.8023505057059448,
|
| 2410 |
+
"eval_sickr_spearman": 0.7537099996234421,
|
| 2411 |
+
"eval_stsb_spearman": 0.8509910117884474,
|
| 2412 |
+
"step": 30750
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 1.98,
|
| 2416 |
+
"eval_avg_sts": 0.8023412928143316,
|
| 2417 |
+
"eval_sickr_spearman": 0.7536968391016481,
|
| 2418 |
+
"eval_stsb_spearman": 0.8509857465270153,
|
| 2419 |
+
"step": 30875
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"electra_acc": NaN,
|
| 2423 |
+
"electra_fix_acc": NaN,
|
| 2424 |
+
"electra_rep_acc": NaN,
|
| 2425 |
+
"epoch": 1.98,
|
| 2426 |
+
"learning_rate": 5.64443875591962e-08,
|
| 2427 |
+
"loss": 0.0013,
|
| 2428 |
+
"neg_sim": NaN,
|
| 2429 |
+
"pos_sim": 0.8246,
|
| 2430 |
+
"step": 31000
|
| 2431 |
+
},
|
| 2432 |
+
{
|
| 2433 |
+
"epoch": 1.98,
|
| 2434 |
+
"eval_avg_sts": 0.8023353086079636,
|
| 2435 |
+
"eval_sickr_spearman": 0.7536697975915385,
|
| 2436 |
+
"eval_stsb_spearman": 0.8510008196243888,
|
| 2437 |
+
"step": 31000
|
| 2438 |
+
},
|
| 2439 |
+
{
|
| 2440 |
+
"epoch": 1.99,
|
| 2441 |
+
"eval_avg_sts": 0.8023565101430139,
|
| 2442 |
+
"eval_sickr_spearman": 0.7536807486826663,
|
| 2443 |
+
"eval_stsb_spearman": 0.8510322716033614,
|
| 2444 |
+
"step": 31125
|
| 2445 |
+
},
|
| 2446 |
+
{
|
| 2447 |
+
"epoch": 2.0,
|
| 2448 |
+
"eval_avg_sts": 0.802324478822219,
|
| 2449 |
+
"eval_sickr_spearman": 0.7536380010023865,
|
| 2450 |
+
"eval_stsb_spearman": 0.8510109566420515,
|
| 2451 |
+
"step": 31250
|
| 2452 |
+
},
|
| 2453 |
+
{
|
| 2454 |
+
"epoch": 2.0,
|
| 2455 |
+
"step": 31252,
|
| 2456 |
+
"train_runtime": 14615.2615,
|
| 2457 |
+
"train_samples_per_second": 2.138
|
| 2458 |
+
}
|
| 2459 |
+
],
|
| 2460 |
+
"max_steps": 31252,
|
| 2461 |
+
"num_train_epochs": 2,
|
| 2462 |
+
"total_flos": 603571104573277440,
|
| 2463 |
+
"trial_name": null,
|
| 2464 |
+
"trial_params": null
|
| 2465 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:38b425a4de30cc200a1563c2b661ff08d680b7ef9e3b162261a336d7f27932d7
|
| 3 |
+
size 2095
|
vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|