Upload 8 files
Browse files- config.json +26 -0
- merges.txt +0 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +1 -0
- tokenizer_config.json +1 -0
- trainer_state.json +426 -0
- training_args.bin +3 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/data/lyhe/SentenceGen/RankCSE-master/checkpoints/multicse-roberta-base-uncased",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"RobertaForCL"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"bos_token_id": 0,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"gradient_checkpointing": false,
|
| 10 |
+
"hidden_act": "gelu",
|
| 11 |
+
"hidden_dropout_prob": 0.1,
|
| 12 |
+
"hidden_size": 768,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 3072,
|
| 15 |
+
"layer_norm_eps": 1e-05,
|
| 16 |
+
"max_position_embeddings": 514,
|
| 17 |
+
"model_type": "roberta",
|
| 18 |
+
"num_attention_heads": 12,
|
| 19 |
+
"num_hidden_layers": 12,
|
| 20 |
+
"pad_token_id": 1,
|
| 21 |
+
"position_embedding_type": "absolute",
|
| 22 |
+
"transformers_version": "4.2.1",
|
| 23 |
+
"type_vocab_size": 1,
|
| 24 |
+
"use_cache": true,
|
| 25 |
+
"vocab_size": 50265
|
| 26 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e2eb5f5ddf6d661e96aefe7eb2f1c74b7d2364583a7e74f2b1eedd2126eeb85e
|
| 3 |
+
size 498669738
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"bos_token": "<s>", "eos_token": "</s>", "unk_token": "<unk>", "sep_token": "</s>", "pad_token": "<pad>", "cls_token": "<s>", "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": false}}
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"unk_token": {"content": "<unk>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "bos_token": {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "eos_token": {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "add_prefix_space": false, "errors": "replace", "sep_token": {"content": "</s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "cls_token": {"content": "<s>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "pad_token": {"content": "<pad>", "single_word": false, "lstrip": false, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "mask_token": {"content": "<mask>", "single_word": false, "lstrip": true, "rstrip": false, "normalized": true, "__type": "AddedToken"}, "model_max_length": 512, "special_tokens_map_file": null, "name_or_path": "/data/lyhe/SentenceGen/RankCSE-master/checkpoints/multicse-roberta-base-uncased"}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,426 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8533187120426406,
|
| 3 |
+
"best_model_checkpoint": "checkpoints/longgen-soft_sort-multicse-roberta-base-uncased-lr:3e-6-es:25-dw:0.5",
|
| 4 |
+
"epoch": 0.08715682002116666,
|
| 5 |
+
"global_step": 1400,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.06,
|
| 12 |
+
"eval_avg_sts": 0.7933545386236226,
|
| 13 |
+
"eval_sickr_spearman": 0.7452889947949065,
|
| 14 |
+
"eval_stsb_spearman": 0.8414200824523387,
|
| 15 |
+
"step": 25
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.11,
|
| 19 |
+
"eval_avg_sts": 0.8246845923550631,
|
| 20 |
+
"eval_sickr_spearman": 0.7947764472597243,
|
| 21 |
+
"eval_stsb_spearman": 0.8545927374504019,
|
| 22 |
+
"step": 50
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"epoch": 0.17,
|
| 26 |
+
"eval_avg_sts": 0.831372297816624,
|
| 27 |
+
"eval_sickr_spearman": 0.7991409854163041,
|
| 28 |
+
"eval_stsb_spearman": 0.863603610216944,
|
| 29 |
+
"step": 75
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 0.22,
|
| 33 |
+
"eval_avg_sts": 0.8311033533020455,
|
| 34 |
+
"eval_sickr_spearman": 0.7968452428608664,
|
| 35 |
+
"eval_stsb_spearman": 0.8653614637432246,
|
| 36 |
+
"step": 100
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 0.28,
|
| 40 |
+
"eval_avg_sts": 0.8347302190626489,
|
| 41 |
+
"eval_sickr_spearman": 0.8026452865461272,
|
| 42 |
+
"eval_stsb_spearman": 0.8668151515791707,
|
| 43 |
+
"step": 125
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 0.33,
|
| 47 |
+
"eval_avg_sts": 0.8291379711550301,
|
| 48 |
+
"eval_sickr_spearman": 0.7935402227709101,
|
| 49 |
+
"eval_stsb_spearman": 0.8647357195391501,
|
| 50 |
+
"step": 150
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.39,
|
| 54 |
+
"eval_avg_sts": 0.8375071181348266,
|
| 55 |
+
"eval_sickr_spearman": 0.8121609682075389,
|
| 56 |
+
"eval_stsb_spearman": 0.8628532680621143,
|
| 57 |
+
"step": 175
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.44,
|
| 61 |
+
"eval_avg_sts": 0.8401116350055815,
|
| 62 |
+
"eval_sickr_spearman": 0.8101783404023766,
|
| 63 |
+
"eval_stsb_spearman": 0.8700449296087865,
|
| 64 |
+
"step": 200
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.5,
|
| 68 |
+
"eval_avg_sts": 0.8319177480289321,
|
| 69 |
+
"eval_sickr_spearman": 0.7987839222082131,
|
| 70 |
+
"eval_stsb_spearman": 0.8650515738496511,
|
| 71 |
+
"step": 225
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 0.55,
|
| 75 |
+
"eval_avg_sts": 0.8389465429116724,
|
| 76 |
+
"eval_sickr_spearman": 0.810887903863922,
|
| 77 |
+
"eval_stsb_spearman": 0.8670051819594228,
|
| 78 |
+
"step": 250
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 0.61,
|
| 82 |
+
"eval_avg_sts": 0.8427200181062976,
|
| 83 |
+
"eval_sickr_spearman": 0.8177831046930728,
|
| 84 |
+
"eval_stsb_spearman": 0.8676569315195224,
|
| 85 |
+
"step": 275
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.67,
|
| 89 |
+
"eval_avg_sts": 0.8463857988633654,
|
| 90 |
+
"eval_sickr_spearman": 0.8234845405270812,
|
| 91 |
+
"eval_stsb_spearman": 0.8692870571996496,
|
| 92 |
+
"step": 300
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.72,
|
| 96 |
+
"eval_avg_sts": 0.8427305224744921,
|
| 97 |
+
"eval_sickr_spearman": 0.8178719622307333,
|
| 98 |
+
"eval_stsb_spearman": 0.8675890827182509,
|
| 99 |
+
"step": 325
|
| 100 |
+
},
|
| 101 |
+
{
|
| 102 |
+
"epoch": 0.78,
|
| 103 |
+
"eval_avg_sts": 0.8421029268773901,
|
| 104 |
+
"eval_sickr_spearman": 0.8149265029661426,
|
| 105 |
+
"eval_stsb_spearman": 0.8692793507886375,
|
| 106 |
+
"step": 350
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 0.83,
|
| 110 |
+
"eval_avg_sts": 0.8477497237237577,
|
| 111 |
+
"eval_sickr_spearman": 0.8272267397023274,
|
| 112 |
+
"eval_stsb_spearman": 0.8682727077451879,
|
| 113 |
+
"step": 375
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.89,
|
| 117 |
+
"eval_avg_sts": 0.8442815051792503,
|
| 118 |
+
"eval_sickr_spearman": 0.821835440690305,
|
| 119 |
+
"eval_stsb_spearman": 0.8667275696681955,
|
| 120 |
+
"step": 400
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 0.94,
|
| 124 |
+
"eval_avg_sts": 0.8421091906691427,
|
| 125 |
+
"eval_sickr_spearman": 0.8156059028959847,
|
| 126 |
+
"eval_stsb_spearman": 0.8686124784423007,
|
| 127 |
+
"step": 425
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 1.0,
|
| 131 |
+
"eval_avg_sts": 0.8394955748362896,
|
| 132 |
+
"eval_sickr_spearman": 0.8169585547746848,
|
| 133 |
+
"eval_stsb_spearman": 0.8620325948978942,
|
| 134 |
+
"step": 450
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 1.05,
|
| 138 |
+
"eval_avg_sts": 0.8465826685429587,
|
| 139 |
+
"eval_sickr_spearman": 0.8285129165366373,
|
| 140 |
+
"eval_stsb_spearman": 0.86465242054928,
|
| 141 |
+
"step": 475
|
| 142 |
+
},
|
| 143 |
+
{
|
| 144 |
+
"epoch": 1.11,
|
| 145 |
+
"learning_rate": 3.152254249815226e-05,
|
| 146 |
+
"loss": 0.4875,
|
| 147 |
+
"step": 500
|
| 148 |
+
},
|
| 149 |
+
{
|
| 150 |
+
"epoch": 1.11,
|
| 151 |
+
"eval_avg_sts": 0.8509943622098309,
|
| 152 |
+
"eval_sickr_spearman": 0.8317867164106589,
|
| 153 |
+
"eval_stsb_spearman": 0.8702020080090029,
|
| 154 |
+
"step": 500
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.16,
|
| 158 |
+
"eval_avg_sts": 0.845221120650099,
|
| 159 |
+
"eval_sickr_spearman": 0.8256615501997637,
|
| 160 |
+
"eval_stsb_spearman": 0.8647806911004344,
|
| 161 |
+
"step": 525
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 1.22,
|
| 165 |
+
"eval_avg_sts": 0.84262665612863,
|
| 166 |
+
"eval_sickr_spearman": 0.8214538816204806,
|
| 167 |
+
"eval_stsb_spearman": 0.8637994306367793,
|
| 168 |
+
"step": 550
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.27,
|
| 172 |
+
"eval_avg_sts": 0.8424131248137818,
|
| 173 |
+
"eval_sickr_spearman": 0.8200806724303651,
|
| 174 |
+
"eval_stsb_spearman": 0.8647455771971985,
|
| 175 |
+
"step": 575
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.33,
|
| 179 |
+
"eval_avg_sts": 0.8456252822224128,
|
| 180 |
+
"eval_sickr_spearman": 0.8223322744035808,
|
| 181 |
+
"eval_stsb_spearman": 0.868918290041245,
|
| 182 |
+
"step": 600
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 1.39,
|
| 186 |
+
"eval_avg_sts": 0.8418597846725664,
|
| 187 |
+
"eval_sickr_spearman": 0.8187211040730578,
|
| 188 |
+
"eval_stsb_spearman": 0.8649984652720751,
|
| 189 |
+
"step": 625
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 1.44,
|
| 193 |
+
"eval_avg_sts": 0.8491621745286901,
|
| 194 |
+
"eval_sickr_spearman": 0.8251584244121993,
|
| 195 |
+
"eval_stsb_spearman": 0.8731659246451808,
|
| 196 |
+
"step": 650
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 1.5,
|
| 200 |
+
"eval_avg_sts": 0.8441242517008596,
|
| 201 |
+
"eval_sickr_spearman": 0.8207689100828717,
|
| 202 |
+
"eval_stsb_spearman": 0.8674795933188475,
|
| 203 |
+
"step": 675
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.55,
|
| 207 |
+
"eval_avg_sts": 0.8486831930599381,
|
| 208 |
+
"eval_sickr_spearman": 0.8272354813627892,
|
| 209 |
+
"eval_stsb_spearman": 0.8701309047570869,
|
| 210 |
+
"step": 700
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 1.61,
|
| 214 |
+
"eval_avg_sts": 0.8461637494650611,
|
| 215 |
+
"eval_sickr_spearman": 0.8234650879309988,
|
| 216 |
+
"eval_stsb_spearman": 0.8688624109991235,
|
| 217 |
+
"step": 725
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.66,
|
| 221 |
+
"eval_avg_sts": 0.8407491656841108,
|
| 222 |
+
"eval_sickr_spearman": 0.8176846409351248,
|
| 223 |
+
"eval_stsb_spearman": 0.863813690433097,
|
| 224 |
+
"step": 750
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.72,
|
| 228 |
+
"eval_avg_sts": 0.8437663221140155,
|
| 229 |
+
"eval_sickr_spearman": 0.8220878441283623,
|
| 230 |
+
"eval_stsb_spearman": 0.8654448000996685,
|
| 231 |
+
"step": 775
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.77,
|
| 235 |
+
"eval_avg_sts": 0.8461783409585929,
|
| 236 |
+
"eval_sickr_spearman": 0.8231578810062006,
|
| 237 |
+
"eval_stsb_spearman": 0.8691988009109852,
|
| 238 |
+
"step": 800
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 1.83,
|
| 242 |
+
"eval_avg_sts": 0.8405187952752955,
|
| 243 |
+
"eval_sickr_spearman": 0.8153428845845095,
|
| 244 |
+
"eval_stsb_spearman": 0.8656947059660813,
|
| 245 |
+
"step": 825
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.88,
|
| 249 |
+
"eval_avg_sts": 0.8427375515517908,
|
| 250 |
+
"eval_sickr_spearman": 0.8192698113758867,
|
| 251 |
+
"eval_stsb_spearman": 0.8662052917276947,
|
| 252 |
+
"step": 850
|
| 253 |
+
},
|
| 254 |
+
{
|
| 255 |
+
"epoch": 1.94,
|
| 256 |
+
"eval_avg_sts": 0.8433561035217969,
|
| 257 |
+
"eval_sickr_spearman": 0.8219217045484879,
|
| 258 |
+
"eval_stsb_spearman": 0.8647905024951058,
|
| 259 |
+
"step": 875
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 2.0,
|
| 263 |
+
"eval_avg_sts": 0.8422742315982205,
|
| 264 |
+
"eval_sickr_spearman": 0.8198740426319783,
|
| 265 |
+
"eval_stsb_spearman": 0.8646744205644626,
|
| 266 |
+
"step": 900
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 2.05,
|
| 270 |
+
"eval_avg_sts": 0.8450157787124789,
|
| 271 |
+
"eval_sickr_spearman": 0.8220855386354934,
|
| 272 |
+
"eval_stsb_spearman": 0.8679460187894644,
|
| 273 |
+
"step": 925
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 2.11,
|
| 277 |
+
"eval_avg_sts": 0.8433056158731305,
|
| 278 |
+
"eval_sickr_spearman": 0.8200347066662887,
|
| 279 |
+
"eval_stsb_spearman": 0.8665765250799722,
|
| 280 |
+
"step": 950
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 2.16,
|
| 284 |
+
"eval_avg_sts": 0.8406107657907074,
|
| 285 |
+
"eval_sickr_spearman": 0.8160642156659073,
|
| 286 |
+
"eval_stsb_spearman": 0.8651573159155074,
|
| 287 |
+
"step": 975
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 2.22,
|
| 291 |
+
"learning_rate": 1.3045084996304511e-05,
|
| 292 |
+
"loss": 0.2406,
|
| 293 |
+
"step": 1000
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 2.22,
|
| 297 |
+
"eval_avg_sts": 0.8441566677812831,
|
| 298 |
+
"eval_sickr_spearman": 0.819970921363579,
|
| 299 |
+
"eval_stsb_spearman": 0.8683424141989871,
|
| 300 |
+
"step": 1000
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 2.27,
|
| 304 |
+
"eval_avg_sts": 0.843133071876425,
|
| 305 |
+
"eval_sickr_spearman": 0.8194127519337666,
|
| 306 |
+
"eval_stsb_spearman": 0.8668533918190835,
|
| 307 |
+
"step": 1025
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 2.33,
|
| 311 |
+
"eval_avg_sts": 0.8427876550387465,
|
| 312 |
+
"eval_sickr_spearman": 0.818166056664839,
|
| 313 |
+
"eval_stsb_spearman": 0.8674092534126542,
|
| 314 |
+
"step": 1050
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 2.38,
|
| 318 |
+
"eval_avg_sts": 0.8430375455152233,
|
| 319 |
+
"eval_sickr_spearman": 0.8175366090804925,
|
| 320 |
+
"eval_stsb_spearman": 0.868538481949954,
|
| 321 |
+
"step": 1075
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 2.44,
|
| 325 |
+
"eval_avg_sts": 0.8419815162026707,
|
| 326 |
+
"eval_sickr_spearman": 0.8154516750292669,
|
| 327 |
+
"eval_stsb_spearman": 0.8685113573760747,
|
| 328 |
+
"step": 1100
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 2.49,
|
| 332 |
+
"eval_avg_sts": 0.8439982682941662,
|
| 333 |
+
"eval_sickr_spearman": 0.818244203266879,
|
| 334 |
+
"eval_stsb_spearman": 0.8697523333214534,
|
| 335 |
+
"step": 1125
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 2.55,
|
| 339 |
+
"eval_avg_sts": 0.8455002294336829,
|
| 340 |
+
"eval_sickr_spearman": 0.8215922111926224,
|
| 341 |
+
"eval_stsb_spearman": 0.8694082476747435,
|
| 342 |
+
"step": 1150
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 2.61,
|
| 346 |
+
"eval_avg_sts": 0.8415438728640847,
|
| 347 |
+
"eval_sickr_spearman": 0.8160743021972091,
|
| 348 |
+
"eval_stsb_spearman": 0.8670134435309604,
|
| 349 |
+
"step": 1175
|
| 350 |
+
},
|
| 351 |
+
{
|
| 352 |
+
"epoch": 2.66,
|
| 353 |
+
"eval_avg_sts": 0.8430556628773995,
|
| 354 |
+
"eval_sickr_spearman": 0.8172068275380184,
|
| 355 |
+
"eval_stsb_spearman": 0.8689044982167805,
|
| 356 |
+
"step": 1200
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 2.72,
|
| 360 |
+
"eval_avg_sts": 0.8426755323560923,
|
| 361 |
+
"eval_sickr_spearman": 0.8163431803030597,
|
| 362 |
+
"eval_stsb_spearman": 0.8690078844091248,
|
| 363 |
+
"step": 1225
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 2.77,
|
| 367 |
+
"eval_avg_sts": 0.8432975917391303,
|
| 368 |
+
"eval_sickr_spearman": 0.8183610149055763,
|
| 369 |
+
"eval_stsb_spearman": 0.8682341685726842,
|
| 370 |
+
"step": 1250
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 2.83,
|
| 374 |
+
"eval_avg_sts": 0.8435685574000784,
|
| 375 |
+
"eval_sickr_spearman": 0.8189177914334469,
|
| 376 |
+
"eval_stsb_spearman": 0.8682193233667098,
|
| 377 |
+
"step": 1275
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 2.88,
|
| 381 |
+
"eval_avg_sts": 0.8441294339312184,
|
| 382 |
+
"eval_sickr_spearman": 0.8195803804777856,
|
| 383 |
+
"eval_stsb_spearman": 0.8686784873846513,
|
| 384 |
+
"step": 1300
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 2.94,
|
| 388 |
+
"eval_avg_sts": 0.8438416918303956,
|
| 389 |
+
"eval_sickr_spearman": 0.8193592933178659,
|
| 390 |
+
"eval_stsb_spearman": 0.8683240903429252,
|
| 391 |
+
"step": 1325
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 2.99,
|
| 395 |
+
"eval_avg_sts": 0.8441351416286236,
|
| 396 |
+
"eval_sickr_spearman": 0.8197974330251845,
|
| 397 |
+
"eval_stsb_spearman": 0.8684728502320628,
|
| 398 |
+
"step": 1350
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 3.0,
|
| 402 |
+
"step": 1353,
|
| 403 |
+
"train_runtime": 2044.9965,
|
| 404 |
+
"train_samples_per_second": 0.662
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.09,
|
| 408 |
+
"eval_avg_sts": 0.8519068652662883,
|
| 409 |
+
"eval_sickr_spearman": 0.8324423889763906,
|
| 410 |
+
"eval_stsb_spearman": 0.8713713415561861,
|
| 411 |
+
"step": 1375
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 0.09,
|
| 415 |
+
"eval_avg_sts": 0.8533187120426406,
|
| 416 |
+
"eval_sickr_spearman": 0.8342147846505584,
|
| 417 |
+
"eval_stsb_spearman": 0.8724226394347229,
|
| 418 |
+
"step": 1400
|
| 419 |
+
}
|
| 420 |
+
],
|
| 421 |
+
"max_steps": 8032,
|
| 422 |
+
"num_train_epochs": 1,
|
| 423 |
+
"total_flos": 130663234395930240,
|
| 424 |
+
"trial_name": null,
|
| 425 |
+
"trial_params": null
|
| 426 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:852624934ea43863ac502b42d2f8edc7fc0ad7f1eab4a8a9103468c6222ec370
|
| 3 |
+
size 2744
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|