Upload 9 files
Browse files- config.json +26 -0
- eval_results.txt +26 -0
- merges.txt +0 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +1 -0
- tokenizer_config.json +1 -0
- train_results.txt +3 -0
- trainer_state.json +437 -0
- vocab.json +0 -0
config.json
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "syncse-v3dsyn8105.san-roberta-large",
|
| 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": 1024,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 4096,
|
| 15 |
+
"layer_norm_eps": 1e-05,
|
| 16 |
+
"max_position_embeddings": 514,
|
| 17 |
+
"model_type": "roberta",
|
| 18 |
+
"num_attention_heads": 16,
|
| 19 |
+
"num_hidden_layers": 24,
|
| 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 |
+
}
|
eval_results.txt
ADDED
|
@@ -0,0 +1,26 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
epoch = 3.0
|
| 2 |
+
eval_avg_sts = 0.852688164945469
|
| 3 |
+
eval_sickr_spearman = 0.8305016923227837
|
| 4 |
+
eval_stsb_spearman = 0.8748746375681541
|
| 5 |
+
------ test ------
|
| 6 |
+
+-------+-------+-------+-------+-------+--------------+-----------------+-------+
|
| 7 |
+
| STS12 | STS13 | STS14 | STS15 | STS16 | STSBenchmark | SICKRelatedness | Avg. |
|
| 8 |
+
+-------+-------+-------+-------+-------+--------------+-----------------+-------+
|
| 9 |
+
| 77.30 | 86.62 | 81.69 | 86.02 | 84.17 | 85.79 | 81.62 | 83.32 |
|
| 10 |
+
+-------+-------+-------+-------+-------+--------------+-----------------+-------+
|
| 11 |
+
+------+------+------+------+------+------+------+------+
|
| 12 |
+
| MR | CR | SUBJ | MPQA | SST2 | TREC | MRPC | Avg. |
|
| 13 |
+
+------+------+------+------+------+------+------+------+
|
| 14 |
+
| 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 | 0.00 |
|
| 15 |
+
+------+------+------+------+------+------+------+------+
|
| 16 |
+
------ test ------
|
| 17 |
+
+-------+-------+-------+-------+-------+--------------+-----------------+-------+
|
| 18 |
+
| STS12 | STS13 | STS14 | STS15 | STS16 | STSBenchmark | SICKRelatedness | Avg. |
|
| 19 |
+
+-------+-------+-------+-------+-------+--------------+-----------------+-------+
|
| 20 |
+
| 77.30 | 86.62 | 81.69 | 86.02 | 84.17 | 85.79 | 81.62 | 83.32 |
|
| 21 |
+
+-------+-------+-------+-------+-------+--------------+-----------------+-------+
|
| 22 |
+
+-------+-------+-------+-------+-------+-------+-------+-------+
|
| 23 |
+
| MR | CR | SUBJ | MPQA | SST2 | TREC | MRPC | Avg. |
|
| 24 |
+
+-------+-------+-------+-------+-------+-------+-------+-------+
|
| 25 |
+
| 86.91 | 91.81 | 93.81 | 90.89 | 91.54 | 90.60 | 72.87 | 88.35 |
|
| 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 sha256:5c754596f03f3032ed4cd9c560fc29705db76dbd254fcb622c016d1578d1e747
|
| 3 |
+
size 1421605357
|
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": "<unk>", "bos_token": "<s>", "eos_token": "</s>", "add_prefix_space": false, "errors": "replace", "sep_token": "</s>", "cls_token": "<s>", "pad_token": "<pad>", "mask_token": "<mask>", "special_tokens_map_file": null, "name_or_path": "roberta-large"}
|
train_results.txt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
epoch = 3.0
|
| 2 |
+
train_runtime = 4896.9547
|
| 3 |
+
train_samples_per_second = 1.251
|
trainer_state.json
ADDED
|
@@ -0,0 +1,437 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.8748746375681541,
|
| 3 |
+
"best_model_checkpoint": "v3dsyn8105.san-roberta-large",
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"global_step": 6126,
|
| 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": 0.831522642680213,
|
| 13 |
+
"eval_sickr_spearman": 0.819481868688736,
|
| 14 |
+
"eval_stsb_spearman": 0.8435634166716901,
|
| 15 |
+
"step": 125
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.12,
|
| 19 |
+
"eval_avg_sts": 0.8394633788185508,
|
| 20 |
+
"eval_sickr_spearman": 0.8260827829593759,
|
| 21 |
+
"eval_stsb_spearman": 0.8528439746777257,
|
| 22 |
+
"step": 250
|
| 23 |
+
},
|
| 24 |
+
{
|
| 25 |
+
"epoch": 0.18,
|
| 26 |
+
"eval_avg_sts": 0.8387226540359579,
|
| 27 |
+
"eval_sickr_spearman": 0.8118421377561928,
|
| 28 |
+
"eval_stsb_spearman": 0.865603170315723,
|
| 29 |
+
"step": 375
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 0.24,
|
| 33 |
+
"learning_rate": 9.183806725432584e-06,
|
| 34 |
+
"loss": 0.4868,
|
| 35 |
+
"step": 500
|
| 36 |
+
},
|
| 37 |
+
{
|
| 38 |
+
"epoch": 0.24,
|
| 39 |
+
"eval_avg_sts": 0.8340637818765272,
|
| 40 |
+
"eval_sickr_spearman": 0.8097645044323858,
|
| 41 |
+
"eval_stsb_spearman": 0.8583630593206686,
|
| 42 |
+
"step": 500
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 0.31,
|
| 46 |
+
"eval_avg_sts": 0.8467054456929737,
|
| 47 |
+
"eval_sickr_spearman": 0.8272759235502001,
|
| 48 |
+
"eval_stsb_spearman": 0.8661349678357472,
|
| 49 |
+
"step": 625
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.37,
|
| 53 |
+
"eval_avg_sts": 0.8413222265544698,
|
| 54 |
+
"eval_sickr_spearman": 0.816061670017531,
|
| 55 |
+
"eval_stsb_spearman": 0.8665827830914085,
|
| 56 |
+
"step": 750
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.43,
|
| 60 |
+
"eval_avg_sts": 0.8498089066273828,
|
| 61 |
+
"eval_sickr_spearman": 0.8340720362170844,
|
| 62 |
+
"eval_stsb_spearman": 0.8655457770376812,
|
| 63 |
+
"step": 875
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 0.49,
|
| 67 |
+
"learning_rate": 8.367613450865165e-06,
|
| 68 |
+
"loss": 0.1238,
|
| 69 |
+
"step": 1000
|
| 70 |
+
},
|
| 71 |
+
{
|
| 72 |
+
"epoch": 0.49,
|
| 73 |
+
"eval_avg_sts": 0.8486804289849034,
|
| 74 |
+
"eval_sickr_spearman": 0.8251996831283348,
|
| 75 |
+
"eval_stsb_spearman": 0.8721611748414719,
|
| 76 |
+
"step": 1000
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.55,
|
| 80 |
+
"eval_avg_sts": 0.8467220562876994,
|
| 81 |
+
"eval_sickr_spearman": 0.8293082155142495,
|
| 82 |
+
"eval_stsb_spearman": 0.8641358970611491,
|
| 83 |
+
"step": 1125
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.61,
|
| 87 |
+
"eval_avg_sts": 0.8450297472977634,
|
| 88 |
+
"eval_sickr_spearman": 0.8284341455302786,
|
| 89 |
+
"eval_stsb_spearman": 0.8616253490652481,
|
| 90 |
+
"step": 1250
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.67,
|
| 94 |
+
"eval_avg_sts": 0.8399309524387217,
|
| 95 |
+
"eval_sickr_spearman": 0.817097460720044,
|
| 96 |
+
"eval_stsb_spearman": 0.8627644441573996,
|
| 97 |
+
"step": 1375
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.73,
|
| 101 |
+
"learning_rate": 7.551420176297748e-06,
|
| 102 |
+
"loss": 0.1051,
|
| 103 |
+
"step": 1500
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 0.73,
|
| 107 |
+
"eval_avg_sts": 0.8488287267471869,
|
| 108 |
+
"eval_sickr_spearman": 0.8298005823350916,
|
| 109 |
+
"eval_stsb_spearman": 0.8678568711592822,
|
| 110 |
+
"step": 1500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.8,
|
| 114 |
+
"eval_avg_sts": 0.8500693337499818,
|
| 115 |
+
"eval_sickr_spearman": 0.832001703620696,
|
| 116 |
+
"eval_stsb_spearman": 0.8681369638792675,
|
| 117 |
+
"step": 1625
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.86,
|
| 121 |
+
"eval_avg_sts": 0.8539883178184261,
|
| 122 |
+
"eval_sickr_spearman": 0.8361804094458117,
|
| 123 |
+
"eval_stsb_spearman": 0.8717962261910405,
|
| 124 |
+
"step": 1750
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.92,
|
| 128 |
+
"eval_avg_sts": 0.8506565279083476,
|
| 129 |
+
"eval_sickr_spearman": 0.8323338386871405,
|
| 130 |
+
"eval_stsb_spearman": 0.8689792171295547,
|
| 131 |
+
"step": 1875
|
| 132 |
+
},
|
| 133 |
+
{
|
| 134 |
+
"epoch": 0.98,
|
| 135 |
+
"learning_rate": 6.735226901730331e-06,
|
| 136 |
+
"loss": 0.0964,
|
| 137 |
+
"step": 2000
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.98,
|
| 141 |
+
"eval_avg_sts": 0.8566417773406937,
|
| 142 |
+
"eval_sickr_spearman": 0.8427579325700513,
|
| 143 |
+
"eval_stsb_spearman": 0.8705256221113361,
|
| 144 |
+
"step": 2000
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.04,
|
| 148 |
+
"eval_avg_sts": 0.8506860271225665,
|
| 149 |
+
"eval_sickr_spearman": 0.8358071597565362,
|
| 150 |
+
"eval_stsb_spearman": 0.8655648944885967,
|
| 151 |
+
"step": 2125
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.1,
|
| 155 |
+
"eval_avg_sts": 0.8526145284643478,
|
| 156 |
+
"eval_sickr_spearman": 0.8352194992304408,
|
| 157 |
+
"eval_stsb_spearman": 0.8700095576982549,
|
| 158 |
+
"step": 2250
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 1.16,
|
| 162 |
+
"eval_avg_sts": 0.8512789094461894,
|
| 163 |
+
"eval_sickr_spearman": 0.8339942258327547,
|
| 164 |
+
"eval_stsb_spearman": 0.8685635930596243,
|
| 165 |
+
"step": 2375
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 1.22,
|
| 169 |
+
"learning_rate": 5.919033627162912e-06,
|
| 170 |
+
"loss": 0.0724,
|
| 171 |
+
"step": 2500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 1.22,
|
| 175 |
+
"eval_avg_sts": 0.8542737971692321,
|
| 176 |
+
"eval_sickr_spearman": 0.8376304683982283,
|
| 177 |
+
"eval_stsb_spearman": 0.870917125940236,
|
| 178 |
+
"step": 2500
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 1.29,
|
| 182 |
+
"eval_avg_sts": 0.853215393568972,
|
| 183 |
+
"eval_sickr_spearman": 0.8382630380041683,
|
| 184 |
+
"eval_stsb_spearman": 0.8681677491337757,
|
| 185 |
+
"step": 2625
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.35,
|
| 189 |
+
"eval_avg_sts": 0.8512616991878239,
|
| 190 |
+
"eval_sickr_spearman": 0.8376803727126226,
|
| 191 |
+
"eval_stsb_spearman": 0.8648430256630253,
|
| 192 |
+
"step": 2750
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.41,
|
| 196 |
+
"eval_avg_sts": 0.8445713819963026,
|
| 197 |
+
"eval_sickr_spearman": 0.8245212438205214,
|
| 198 |
+
"eval_stsb_spearman": 0.8646215201720836,
|
| 199 |
+
"step": 2875
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.47,
|
| 203 |
+
"learning_rate": 5.102840352595495e-06,
|
| 204 |
+
"loss": 0.0672,
|
| 205 |
+
"step": 3000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 1.47,
|
| 209 |
+
"eval_avg_sts": 0.8547468779877831,
|
| 210 |
+
"eval_sickr_spearman": 0.840675015825086,
|
| 211 |
+
"eval_stsb_spearman": 0.8688187401504804,
|
| 212 |
+
"step": 3000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.53,
|
| 216 |
+
"eval_avg_sts": 0.8480351385120373,
|
| 217 |
+
"eval_sickr_spearman": 0.828842553985807,
|
| 218 |
+
"eval_stsb_spearman": 0.8672277230382677,
|
| 219 |
+
"step": 3125
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 1.59,
|
| 223 |
+
"eval_avg_sts": 0.851759527195628,
|
| 224 |
+
"eval_sickr_spearman": 0.8318612606800909,
|
| 225 |
+
"eval_stsb_spearman": 0.871657793711165,
|
| 226 |
+
"step": 3250
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 1.65,
|
| 230 |
+
"eval_avg_sts": 0.852688164945469,
|
| 231 |
+
"eval_sickr_spearman": 0.8305016923227837,
|
| 232 |
+
"eval_stsb_spearman": 0.8748746375681541,
|
| 233 |
+
"step": 3375
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.71,
|
| 237 |
+
"learning_rate": 4.286647078028077e-06,
|
| 238 |
+
"loss": 0.0671,
|
| 239 |
+
"step": 3500
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 1.71,
|
| 243 |
+
"eval_avg_sts": 0.8476247717318004,
|
| 244 |
+
"eval_sickr_spearman": 0.825738544055369,
|
| 245 |
+
"eval_stsb_spearman": 0.8695109994082318,
|
| 246 |
+
"step": 3500
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 1.78,
|
| 250 |
+
"eval_avg_sts": 0.8488707460006111,
|
| 251 |
+
"eval_sickr_spearman": 0.8274017650359679,
|
| 252 |
+
"eval_stsb_spearman": 0.8703397269652542,
|
| 253 |
+
"step": 3625
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 1.84,
|
| 257 |
+
"eval_avg_sts": 0.8474621405059352,
|
| 258 |
+
"eval_sickr_spearman": 0.8264578578305062,
|
| 259 |
+
"eval_stsb_spearman": 0.8684664231813641,
|
| 260 |
+
"step": 3750
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.9,
|
| 264 |
+
"eval_avg_sts": 0.8514977368080117,
|
| 265 |
+
"eval_sickr_spearman": 0.8317304719908749,
|
| 266 |
+
"eval_stsb_spearman": 0.8712650016251485,
|
| 267 |
+
"step": 3875
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 1.96,
|
| 271 |
+
"learning_rate": 3.47045380346066e-06,
|
| 272 |
+
"loss": 0.062,
|
| 273 |
+
"step": 4000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.96,
|
| 277 |
+
"eval_avg_sts": 0.8495823034187644,
|
| 278 |
+
"eval_sickr_spearman": 0.8280114718376232,
|
| 279 |
+
"eval_stsb_spearman": 0.8711531349999054,
|
| 280 |
+
"step": 4000
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 2.02,
|
| 284 |
+
"eval_avg_sts": 0.8472080808839515,
|
| 285 |
+
"eval_sickr_spearman": 0.8245909849798095,
|
| 286 |
+
"eval_stsb_spearman": 0.8698251767880936,
|
| 287 |
+
"step": 4125
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 2.08,
|
| 291 |
+
"eval_avg_sts": 0.8502145621787748,
|
| 292 |
+
"eval_sickr_spearman": 0.8296834344786841,
|
| 293 |
+
"eval_stsb_spearman": 0.8707456898788655,
|
| 294 |
+
"step": 4250
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 2.14,
|
| 298 |
+
"eval_avg_sts": 0.8491882134989206,
|
| 299 |
+
"eval_sickr_spearman": 0.8280905310305905,
|
| 300 |
+
"eval_stsb_spearman": 0.8702858959672507,
|
| 301 |
+
"step": 4375
|
| 302 |
+
},
|
| 303 |
+
{
|
| 304 |
+
"epoch": 2.2,
|
| 305 |
+
"learning_rate": 2.654260528893242e-06,
|
| 306 |
+
"loss": 0.0507,
|
| 307 |
+
"step": 4500
|
| 308 |
+
},
|
| 309 |
+
{
|
| 310 |
+
"epoch": 2.2,
|
| 311 |
+
"eval_avg_sts": 0.8489413736886298,
|
| 312 |
+
"eval_sickr_spearman": 0.8264288470452376,
|
| 313 |
+
"eval_stsb_spearman": 0.8714539003320219,
|
| 314 |
+
"step": 4500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 2.26,
|
| 318 |
+
"eval_avg_sts": 0.8403852153317499,
|
| 319 |
+
"eval_sickr_spearman": 0.8150179061521793,
|
| 320 |
+
"eval_stsb_spearman": 0.8657525245113207,
|
| 321 |
+
"step": 4625
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 2.33,
|
| 325 |
+
"eval_avg_sts": 0.8448990255323547,
|
| 326 |
+
"eval_sickr_spearman": 0.8213330834003639,
|
| 327 |
+
"eval_stsb_spearman": 0.8684649676643454,
|
| 328 |
+
"step": 4750
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 2.39,
|
| 332 |
+
"eval_avg_sts": 0.846744031918401,
|
| 333 |
+
"eval_sickr_spearman": 0.8236191236733109,
|
| 334 |
+
"eval_stsb_spearman": 0.8698689401634909,
|
| 335 |
+
"step": 4875
|
| 336 |
+
},
|
| 337 |
+
{
|
| 338 |
+
"epoch": 2.45,
|
| 339 |
+
"learning_rate": 1.8380672543258246e-06,
|
| 340 |
+
"loss": 0.0444,
|
| 341 |
+
"step": 5000
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 2.45,
|
| 345 |
+
"eval_avg_sts": 0.8446342020355491,
|
| 346 |
+
"eval_sickr_spearman": 0.8218428374799265,
|
| 347 |
+
"eval_stsb_spearman": 0.8674255665911716,
|
| 348 |
+
"step": 5000
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 2.51,
|
| 352 |
+
"eval_avg_sts": 0.8466793063787386,
|
| 353 |
+
"eval_sickr_spearman": 0.8251428143042321,
|
| 354 |
+
"eval_stsb_spearman": 0.8682157984532451,
|
| 355 |
+
"step": 5125
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 2.57,
|
| 359 |
+
"eval_avg_sts": 0.8472692519799312,
|
| 360 |
+
"eval_sickr_spearman": 0.8273726581884964,
|
| 361 |
+
"eval_stsb_spearman": 0.8671658457713661,
|
| 362 |
+
"step": 5250
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 2.63,
|
| 366 |
+
"eval_avg_sts": 0.8463050384868719,
|
| 367 |
+
"eval_sickr_spearman": 0.8245047211216268,
|
| 368 |
+
"eval_stsb_spearman": 0.868105355852117,
|
| 369 |
+
"step": 5375
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 2.69,
|
| 373 |
+
"learning_rate": 1.021873979758407e-06,
|
| 374 |
+
"loss": 0.0437,
|
| 375 |
+
"step": 5500
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 2.69,
|
| 379 |
+
"eval_avg_sts": 0.845129637101806,
|
| 380 |
+
"eval_sickr_spearman": 0.8231749320472111,
|
| 381 |
+
"eval_stsb_spearman": 0.8670843421564008,
|
| 382 |
+
"step": 5500
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 2.75,
|
| 386 |
+
"eval_avg_sts": 0.844980005172939,
|
| 387 |
+
"eval_sickr_spearman": 0.8226033139089961,
|
| 388 |
+
"eval_stsb_spearman": 0.8673566964368821,
|
| 389 |
+
"step": 5625
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 2.82,
|
| 393 |
+
"eval_avg_sts": 0.8454315705537849,
|
| 394 |
+
"eval_sickr_spearman": 0.8231090814071395,
|
| 395 |
+
"eval_stsb_spearman": 0.8677540597004305,
|
| 396 |
+
"step": 5750
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 2.88,
|
| 400 |
+
"eval_avg_sts": 0.8447869475256539,
|
| 401 |
+
"eval_sickr_spearman": 0.8219029724189271,
|
| 402 |
+
"eval_stsb_spearman": 0.8676709226323807,
|
| 403 |
+
"step": 5875
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 2.94,
|
| 407 |
+
"learning_rate": 2.0568070519098922e-07,
|
| 408 |
+
"loss": 0.0443,
|
| 409 |
+
"step": 6000
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 2.94,
|
| 413 |
+
"eval_avg_sts": 0.8453072401863975,
|
| 414 |
+
"eval_sickr_spearman": 0.8228684455889346,
|
| 415 |
+
"eval_stsb_spearman": 0.8677460347838605,
|
| 416 |
+
"step": 6000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 3.0,
|
| 420 |
+
"eval_avg_sts": 0.8456332694526848,
|
| 421 |
+
"eval_sickr_spearman": 0.8233002451908632,
|
| 422 |
+
"eval_stsb_spearman": 0.8679662937145064,
|
| 423 |
+
"step": 6125
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 3.0,
|
| 427 |
+
"step": 6126,
|
| 428 |
+
"train_runtime": 4896.9547,
|
| 429 |
+
"train_samples_per_second": 1.251
|
| 430 |
+
}
|
| 431 |
+
],
|
| 432 |
+
"max_steps": 6126,
|
| 433 |
+
"num_train_epochs": 3,
|
| 434 |
+
"total_flos": 213963637016494080,
|
| 435 |
+
"trial_name": null,
|
| 436 |
+
"trial_params": null
|
| 437 |
+
}
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|