Jason Wu
commited on
Commit
•
48552b4
1
Parent(s):
15682b8
init
Browse files- config.json +28 -0
- pytorch_model.bin +3 -0
- special_tokens_map.json +1 -0
- tokenizer.json +0 -0
- tokenizer_config.json +1 -0
- trainer_state.json +654 -0
- training_args.bin +3 -0
- vocab.txt +0 -0
config.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "deepset/bert-large-uncased-whole-word-masking-squad2",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForQuestionAnswering"
|
| 5 |
+
],
|
| 6 |
+
"attention_probs_dropout_prob": 0.1,
|
| 7 |
+
"finetuning_task": "squad2",
|
| 8 |
+
"gradient_checkpointing": false,
|
| 9 |
+
"hidden_act": "gelu",
|
| 10 |
+
"hidden_dropout_prob": 0.1,
|
| 11 |
+
"hidden_size": 1024,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 4096,
|
| 14 |
+
"language": "english",
|
| 15 |
+
"layer_norm_eps": 1e-12,
|
| 16 |
+
"max_position_embeddings": 512,
|
| 17 |
+
"model_type": "bert",
|
| 18 |
+
"name": "Bert",
|
| 19 |
+
"num_attention_heads": 16,
|
| 20 |
+
"num_hidden_layers": 24,
|
| 21 |
+
"output_past": true,
|
| 22 |
+
"pad_token_id": 0,
|
| 23 |
+
"position_embedding_type": "absolute",
|
| 24 |
+
"transformers_version": "4.6.0.dev0",
|
| 25 |
+
"type_vocab_size": 2,
|
| 26 |
+
"use_cache": true,
|
| 27 |
+
"vocab_size": 30522
|
| 28 |
+
}
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:facb7a55e03b2347354430512861af8155dbad300b90e308dd157ad56b831c7e
|
| 3 |
+
size 1336470721
|
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.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
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, "special_tokens_map_file": "cache/ef9bc2ba6867b86932babe745cbccc9822ae43e5c873d03088aee4251bd45fff.dd8bd9bfd3664b530ea4e645105f557769387b3da9f79bdb55ed556bdd80611d", "name_or_path": "deepset/bert-large-uncased-whole-word-masking-squad2", "do_basic_tokenize": true, "never_split": null}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,654 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.999281732051852,
|
| 5 |
+
"global_step": 4565,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.05,
|
| 12 |
+
"learning_rate": 1.5e-06,
|
| 13 |
+
"loss": 2.2671,
|
| 14 |
+
"step": 50
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.11,
|
| 18 |
+
"learning_rate": 3e-06,
|
| 19 |
+
"loss": 1.8863,
|
| 20 |
+
"step": 100
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.16,
|
| 24 |
+
"learning_rate": 4.5e-06,
|
| 25 |
+
"loss": 1.8591,
|
| 26 |
+
"step": 150
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.22,
|
| 30 |
+
"learning_rate": 6e-06,
|
| 31 |
+
"loss": 1.7121,
|
| 32 |
+
"step": 200
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.27,
|
| 36 |
+
"learning_rate": 7.5e-06,
|
| 37 |
+
"loss": 1.6481,
|
| 38 |
+
"step": 250
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.33,
|
| 42 |
+
"learning_rate": 9e-06,
|
| 43 |
+
"loss": 1.606,
|
| 44 |
+
"step": 300
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.38,
|
| 48 |
+
"learning_rate": 1.05e-05,
|
| 49 |
+
"loss": 1.6154,
|
| 50 |
+
"step": 350
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.44,
|
| 54 |
+
"learning_rate": 1.2e-05,
|
| 55 |
+
"loss": 1.5695,
|
| 56 |
+
"step": 400
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.49,
|
| 60 |
+
"learning_rate": 1.3500000000000001e-05,
|
| 61 |
+
"loss": 1.4993,
|
| 62 |
+
"step": 450
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.55,
|
| 66 |
+
"learning_rate": 1.5e-05,
|
| 67 |
+
"loss": 1.4618,
|
| 68 |
+
"step": 500
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.6,
|
| 72 |
+
"learning_rate": 1.65e-05,
|
| 73 |
+
"loss": 1.4893,
|
| 74 |
+
"step": 550
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.66,
|
| 78 |
+
"learning_rate": 1.8e-05,
|
| 79 |
+
"loss": 1.4824,
|
| 80 |
+
"step": 600
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.71,
|
| 84 |
+
"learning_rate": 1.95e-05,
|
| 85 |
+
"loss": 1.4306,
|
| 86 |
+
"step": 650
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.77,
|
| 90 |
+
"learning_rate": 2.1e-05,
|
| 91 |
+
"loss": 1.4697,
|
| 92 |
+
"step": 700
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.82,
|
| 96 |
+
"learning_rate": 2.25e-05,
|
| 97 |
+
"loss": 1.4275,
|
| 98 |
+
"step": 750
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.88,
|
| 102 |
+
"learning_rate": 2.4e-05,
|
| 103 |
+
"loss": 1.4353,
|
| 104 |
+
"step": 800
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.93,
|
| 108 |
+
"learning_rate": 2.55e-05,
|
| 109 |
+
"loss": 1.4811,
|
| 110 |
+
"step": 850
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.99,
|
| 114 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 115 |
+
"loss": 1.4272,
|
| 116 |
+
"step": 900
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"HasAns_exact": 58.11939375193319,
|
| 120 |
+
"HasAns_f1": 72.62489483900757,
|
| 121 |
+
"HasAns_total": 3233,
|
| 122 |
+
"NoAns_exact": 76.07361963190183,
|
| 123 |
+
"NoAns_f1": 76.07361963190183,
|
| 124 |
+
"NoAns_total": 163,
|
| 125 |
+
"best_exact": 60.129564193168434,
|
| 126 |
+
"best_exact_thresh": 0.0,
|
| 127 |
+
"best_f1": 73.93883539885493,
|
| 128 |
+
"best_f1_thresh": 0.0,
|
| 129 |
+
"epoch": 1.0,
|
| 130 |
+
"exact": 58.9811542991755,
|
| 131 |
+
"f1": 72.79042550486203,
|
| 132 |
+
"step": 913,
|
| 133 |
+
"total": 3396
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 1.04,
|
| 137 |
+
"learning_rate": 2.8499999999999998e-05,
|
| 138 |
+
"loss": 1.168,
|
| 139 |
+
"step": 950
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 1.1,
|
| 143 |
+
"learning_rate": 3e-05,
|
| 144 |
+
"loss": 1.1045,
|
| 145 |
+
"step": 1000
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 1.15,
|
| 149 |
+
"learning_rate": 2.9579242636746143e-05,
|
| 150 |
+
"loss": 1.1772,
|
| 151 |
+
"step": 1050
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 1.2,
|
| 155 |
+
"learning_rate": 2.9158485273492286e-05,
|
| 156 |
+
"loss": 1.1478,
|
| 157 |
+
"step": 1100
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 1.26,
|
| 161 |
+
"learning_rate": 2.873772791023843e-05,
|
| 162 |
+
"loss": 1.2113,
|
| 163 |
+
"step": 1150
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 1.31,
|
| 167 |
+
"learning_rate": 2.831697054698457e-05,
|
| 168 |
+
"loss": 1.2509,
|
| 169 |
+
"step": 1200
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 1.37,
|
| 173 |
+
"learning_rate": 2.7896213183730714e-05,
|
| 174 |
+
"loss": 1.1653,
|
| 175 |
+
"step": 1250
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 1.42,
|
| 179 |
+
"learning_rate": 2.747545582047686e-05,
|
| 180 |
+
"loss": 1.1098,
|
| 181 |
+
"step": 1300
|
| 182 |
+
},
|
| 183 |
+
{
|
| 184 |
+
"epoch": 1.48,
|
| 185 |
+
"learning_rate": 2.7054698457223e-05,
|
| 186 |
+
"loss": 1.1876,
|
| 187 |
+
"step": 1350
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 1.53,
|
| 191 |
+
"learning_rate": 2.6633941093969145e-05,
|
| 192 |
+
"loss": 1.1803,
|
| 193 |
+
"step": 1400
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 1.59,
|
| 197 |
+
"learning_rate": 2.6213183730715288e-05,
|
| 198 |
+
"loss": 1.1392,
|
| 199 |
+
"step": 1450
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 1.64,
|
| 203 |
+
"learning_rate": 2.579242636746143e-05,
|
| 204 |
+
"loss": 1.158,
|
| 205 |
+
"step": 1500
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 1.7,
|
| 209 |
+
"learning_rate": 2.5371669004207573e-05,
|
| 210 |
+
"loss": 1.0864,
|
| 211 |
+
"step": 1550
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 1.75,
|
| 215 |
+
"learning_rate": 2.4950911640953716e-05,
|
| 216 |
+
"loss": 1.2314,
|
| 217 |
+
"step": 1600
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 1.81,
|
| 221 |
+
"learning_rate": 2.453015427769986e-05,
|
| 222 |
+
"loss": 1.174,
|
| 223 |
+
"step": 1650
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 1.86,
|
| 227 |
+
"learning_rate": 2.4109396914446005e-05,
|
| 228 |
+
"loss": 1.1833,
|
| 229 |
+
"step": 1700
|
| 230 |
+
},
|
| 231 |
+
{
|
| 232 |
+
"epoch": 1.92,
|
| 233 |
+
"learning_rate": 2.3688639551192144e-05,
|
| 234 |
+
"loss": 1.1636,
|
| 235 |
+
"step": 1750
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 1.97,
|
| 239 |
+
"learning_rate": 2.326788218793829e-05,
|
| 240 |
+
"loss": 1.1598,
|
| 241 |
+
"step": 1800
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"HasAns_exact": 60.19177234766471,
|
| 245 |
+
"HasAns_f1": 75.40873260282125,
|
| 246 |
+
"HasAns_total": 3233,
|
| 247 |
+
"NoAns_exact": 52.147239263803684,
|
| 248 |
+
"NoAns_f1": 52.147239263803684,
|
| 249 |
+
"NoAns_total": 163,
|
| 250 |
+
"best_exact": 62.102473498233216,
|
| 251 |
+
"best_exact_thresh": 0.0,
|
| 252 |
+
"best_f1": 76.58905550792727,
|
| 253 |
+
"best_f1_thresh": 0.0,
|
| 254 |
+
"epoch": 2.0,
|
| 255 |
+
"exact": 59.80565371024735,
|
| 256 |
+
"f1": 74.29223571994143,
|
| 257 |
+
"step": 1826,
|
| 258 |
+
"total": 3396
|
| 259 |
+
},
|
| 260 |
+
{
|
| 261 |
+
"epoch": 2.03,
|
| 262 |
+
"learning_rate": 2.2847124824684433e-05,
|
| 263 |
+
"loss": 0.962,
|
| 264 |
+
"step": 1850
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 2.08,
|
| 268 |
+
"learning_rate": 2.2426367461430575e-05,
|
| 269 |
+
"loss": 0.7194,
|
| 270 |
+
"step": 1900
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"epoch": 2.14,
|
| 274 |
+
"learning_rate": 2.2005610098176718e-05,
|
| 275 |
+
"loss": 0.7277,
|
| 276 |
+
"step": 1950
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 2.19,
|
| 280 |
+
"learning_rate": 2.158485273492286e-05,
|
| 281 |
+
"loss": 0.6882,
|
| 282 |
+
"step": 2000
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 2.25,
|
| 286 |
+
"learning_rate": 2.1164095371669003e-05,
|
| 287 |
+
"loss": 0.6856,
|
| 288 |
+
"step": 2050
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 2.3,
|
| 292 |
+
"learning_rate": 2.074333800841515e-05,
|
| 293 |
+
"loss": 0.7305,
|
| 294 |
+
"step": 2100
|
| 295 |
+
},
|
| 296 |
+
{
|
| 297 |
+
"epoch": 2.35,
|
| 298 |
+
"learning_rate": 2.032258064516129e-05,
|
| 299 |
+
"loss": 0.7234,
|
| 300 |
+
"step": 2150
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 2.41,
|
| 304 |
+
"learning_rate": 1.9901823281907435e-05,
|
| 305 |
+
"loss": 0.6732,
|
| 306 |
+
"step": 2200
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 2.46,
|
| 310 |
+
"learning_rate": 1.9481065918653574e-05,
|
| 311 |
+
"loss": 0.7105,
|
| 312 |
+
"step": 2250
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.52,
|
| 316 |
+
"learning_rate": 1.906030855539972e-05,
|
| 317 |
+
"loss": 0.7333,
|
| 318 |
+
"step": 2300
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 2.57,
|
| 322 |
+
"learning_rate": 1.8639551192145863e-05,
|
| 323 |
+
"loss": 0.6969,
|
| 324 |
+
"step": 2350
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 2.63,
|
| 328 |
+
"learning_rate": 1.8218793828892005e-05,
|
| 329 |
+
"loss": 0.7382,
|
| 330 |
+
"step": 2400
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 2.68,
|
| 334 |
+
"learning_rate": 1.7798036465638148e-05,
|
| 335 |
+
"loss": 0.7801,
|
| 336 |
+
"step": 2450
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 2.74,
|
| 340 |
+
"learning_rate": 1.7377279102384294e-05,
|
| 341 |
+
"loss": 0.7708,
|
| 342 |
+
"step": 2500
|
| 343 |
+
},
|
| 344 |
+
{
|
| 345 |
+
"epoch": 2.79,
|
| 346 |
+
"learning_rate": 1.6956521739130433e-05,
|
| 347 |
+
"loss": 0.7032,
|
| 348 |
+
"step": 2550
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 2.85,
|
| 352 |
+
"learning_rate": 1.653576437587658e-05,
|
| 353 |
+
"loss": 0.7041,
|
| 354 |
+
"step": 2600
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 2.9,
|
| 358 |
+
"learning_rate": 1.611500701262272e-05,
|
| 359 |
+
"loss": 0.7627,
|
| 360 |
+
"step": 2650
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 2.96,
|
| 364 |
+
"learning_rate": 1.5694249649368865e-05,
|
| 365 |
+
"loss": 0.7428,
|
| 366 |
+
"step": 2700
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"HasAns_exact": 59.44942777605939,
|
| 370 |
+
"HasAns_f1": 75.24381473873625,
|
| 371 |
+
"HasAns_total": 3233,
|
| 372 |
+
"NoAns_exact": 56.441717791411044,
|
| 373 |
+
"NoAns_f1": 56.441717791411044,
|
| 374 |
+
"NoAns_total": 163,
|
| 375 |
+
"best_exact": 61.39575971731449,
|
| 376 |
+
"best_exact_thresh": 0.0,
|
| 377 |
+
"best_f1": 76.43205331281922,
|
| 378 |
+
"best_f1_thresh": 0.0,
|
| 379 |
+
"epoch": 3.0,
|
| 380 |
+
"exact": 59.305064782096586,
|
| 381 |
+
"f1": 74.34135837760138,
|
| 382 |
+
"step": 2739,
|
| 383 |
+
"total": 3396
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 3.01,
|
| 387 |
+
"learning_rate": 1.5273492286115007e-05,
|
| 388 |
+
"loss": 0.6496,
|
| 389 |
+
"step": 2750
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 3.07,
|
| 393 |
+
"learning_rate": 1.4852734922861152e-05,
|
| 394 |
+
"loss": 0.4199,
|
| 395 |
+
"step": 2800
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 3.12,
|
| 399 |
+
"learning_rate": 1.4431977559607294e-05,
|
| 400 |
+
"loss": 0.4245,
|
| 401 |
+
"step": 2850
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 3.18,
|
| 405 |
+
"learning_rate": 1.4011220196353437e-05,
|
| 406 |
+
"loss": 0.4192,
|
| 407 |
+
"step": 2900
|
| 408 |
+
},
|
| 409 |
+
{
|
| 410 |
+
"epoch": 3.23,
|
| 411 |
+
"learning_rate": 1.359046283309958e-05,
|
| 412 |
+
"loss": 0.4115,
|
| 413 |
+
"step": 2950
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 3.29,
|
| 417 |
+
"learning_rate": 1.3169705469845722e-05,
|
| 418 |
+
"loss": 0.4392,
|
| 419 |
+
"step": 3000
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 3.34,
|
| 423 |
+
"learning_rate": 1.2748948106591867e-05,
|
| 424 |
+
"loss": 0.3749,
|
| 425 |
+
"step": 3050
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 3.4,
|
| 429 |
+
"learning_rate": 1.232819074333801e-05,
|
| 430 |
+
"loss": 0.4264,
|
| 431 |
+
"step": 3100
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 3.45,
|
| 435 |
+
"learning_rate": 1.1907433380084152e-05,
|
| 436 |
+
"loss": 0.4127,
|
| 437 |
+
"step": 3150
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 3.5,
|
| 441 |
+
"learning_rate": 1.1486676016830295e-05,
|
| 442 |
+
"loss": 0.4517,
|
| 443 |
+
"step": 3200
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 3.56,
|
| 447 |
+
"learning_rate": 1.1065918653576439e-05,
|
| 448 |
+
"loss": 0.4202,
|
| 449 |
+
"step": 3250
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 3.61,
|
| 453 |
+
"learning_rate": 1.0645161290322582e-05,
|
| 454 |
+
"loss": 0.413,
|
| 455 |
+
"step": 3300
|
| 456 |
+
},
|
| 457 |
+
{
|
| 458 |
+
"epoch": 3.67,
|
| 459 |
+
"learning_rate": 1.0224403927068724e-05,
|
| 460 |
+
"loss": 0.4274,
|
| 461 |
+
"step": 3350
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 3.72,
|
| 465 |
+
"learning_rate": 9.803646563814867e-06,
|
| 466 |
+
"loss": 0.3953,
|
| 467 |
+
"step": 3400
|
| 468 |
+
},
|
| 469 |
+
{
|
| 470 |
+
"epoch": 3.78,
|
| 471 |
+
"learning_rate": 9.382889200561011e-06,
|
| 472 |
+
"loss": 0.4115,
|
| 473 |
+
"step": 3450
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 3.83,
|
| 477 |
+
"learning_rate": 8.962131837307154e-06,
|
| 478 |
+
"loss": 0.4382,
|
| 479 |
+
"step": 3500
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 3.89,
|
| 483 |
+
"learning_rate": 8.541374474053297e-06,
|
| 484 |
+
"loss": 0.3753,
|
| 485 |
+
"step": 3550
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 3.94,
|
| 489 |
+
"learning_rate": 8.12061711079944e-06,
|
| 490 |
+
"loss": 0.4231,
|
| 491 |
+
"step": 3600
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 4.0,
|
| 495 |
+
"learning_rate": 7.699859747545584e-06,
|
| 496 |
+
"loss": 0.4801,
|
| 497 |
+
"step": 3650
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"HasAns_exact": 59.48035879987628,
|
| 501 |
+
"HasAns_f1": 75.10338801562317,
|
| 502 |
+
"HasAns_total": 3233,
|
| 503 |
+
"NoAns_exact": 65.03067484662577,
|
| 504 |
+
"NoAns_f1": 65.03067484662577,
|
| 505 |
+
"NoAns_total": 163,
|
| 506 |
+
"best_exact": 61.42520612485277,
|
| 507 |
+
"best_exact_thresh": 0.0,
|
| 508 |
+
"best_f1": 76.29836674161058,
|
| 509 |
+
"best_f1_thresh": 0.0,
|
| 510 |
+
"epoch": 4.0,
|
| 511 |
+
"exact": 59.74676089517079,
|
| 512 |
+
"f1": 74.61992151192865,
|
| 513 |
+
"step": 3652,
|
| 514 |
+
"total": 3396
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 4.05,
|
| 518 |
+
"learning_rate": 7.2791023842917254e-06,
|
| 519 |
+
"loss": 0.2907,
|
| 520 |
+
"step": 3700
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 4.11,
|
| 524 |
+
"learning_rate": 6.858345021037868e-06,
|
| 525 |
+
"loss": 0.2268,
|
| 526 |
+
"step": 3750
|
| 527 |
+
},
|
| 528 |
+
{
|
| 529 |
+
"epoch": 4.16,
|
| 530 |
+
"learning_rate": 6.437587657784012e-06,
|
| 531 |
+
"loss": 0.2669,
|
| 532 |
+
"step": 3800
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 4.22,
|
| 536 |
+
"learning_rate": 6.016830294530154e-06,
|
| 537 |
+
"loss": 0.2185,
|
| 538 |
+
"step": 3850
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 4.27,
|
| 542 |
+
"learning_rate": 5.596072931276297e-06,
|
| 543 |
+
"loss": 0.2734,
|
| 544 |
+
"step": 3900
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 4.33,
|
| 548 |
+
"learning_rate": 5.17531556802244e-06,
|
| 549 |
+
"loss": 0.2407,
|
| 550 |
+
"step": 3950
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 4.38,
|
| 554 |
+
"learning_rate": 4.754558204768583e-06,
|
| 555 |
+
"loss": 0.2524,
|
| 556 |
+
"step": 4000
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 4.44,
|
| 560 |
+
"learning_rate": 4.3338008415147266e-06,
|
| 561 |
+
"loss": 0.2545,
|
| 562 |
+
"step": 4050
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 4.49,
|
| 566 |
+
"learning_rate": 3.913043478260869e-06,
|
| 567 |
+
"loss": 0.2268,
|
| 568 |
+
"step": 4100
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 4.55,
|
| 572 |
+
"learning_rate": 3.4922861150070127e-06,
|
| 573 |
+
"loss": 0.2433,
|
| 574 |
+
"step": 4150
|
| 575 |
+
},
|
| 576 |
+
{
|
| 577 |
+
"epoch": 4.6,
|
| 578 |
+
"learning_rate": 3.071528751753156e-06,
|
| 579 |
+
"loss": 0.2661,
|
| 580 |
+
"step": 4200
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 4.65,
|
| 584 |
+
"learning_rate": 2.650771388499299e-06,
|
| 585 |
+
"loss": 0.2574,
|
| 586 |
+
"step": 4250
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 4.71,
|
| 590 |
+
"learning_rate": 2.230014025245442e-06,
|
| 591 |
+
"loss": 0.212,
|
| 592 |
+
"step": 4300
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 4.76,
|
| 596 |
+
"learning_rate": 1.8092566619915848e-06,
|
| 597 |
+
"loss": 0.2434,
|
| 598 |
+
"step": 4350
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 4.82,
|
| 602 |
+
"learning_rate": 1.388499298737728e-06,
|
| 603 |
+
"loss": 0.28,
|
| 604 |
+
"step": 4400
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 4.87,
|
| 608 |
+
"learning_rate": 9.67741935483871e-07,
|
| 609 |
+
"loss": 0.2294,
|
| 610 |
+
"step": 4450
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 4.93,
|
| 614 |
+
"learning_rate": 5.46984572230014e-07,
|
| 615 |
+
"loss": 0.2103,
|
| 616 |
+
"step": 4500
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 4.98,
|
| 620 |
+
"learning_rate": 1.2622720897615708e-07,
|
| 621 |
+
"loss": 0.2526,
|
| 622 |
+
"step": 4550
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"HasAns_exact": 57.96473863284875,
|
| 626 |
+
"HasAns_f1": 74.35908528391936,
|
| 627 |
+
"HasAns_total": 3233,
|
| 628 |
+
"NoAns_exact": 65.03067484662577,
|
| 629 |
+
"NoAns_f1": 65.03067484662577,
|
| 630 |
+
"NoAns_total": 163,
|
| 631 |
+
"best_exact": 59.98233215547703,
|
| 632 |
+
"best_exact_thresh": 0.0,
|
| 633 |
+
"best_f1": 75.58978878766521,
|
| 634 |
+
"best_f1_thresh": 0.0,
|
| 635 |
+
"epoch": 5.0,
|
| 636 |
+
"exact": 58.303886925795055,
|
| 637 |
+
"f1": 73.91134355798329,
|
| 638 |
+
"step": 4565,
|
| 639 |
+
"total": 3396
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 5.0,
|
| 643 |
+
"step": 4565,
|
| 644 |
+
"total_flos": 1.5001363873057997e+17,
|
| 645 |
+
"train_runtime": 22638.32,
|
| 646 |
+
"train_samples_per_second": 0.202
|
| 647 |
+
}
|
| 648 |
+
],
|
| 649 |
+
"max_steps": 4565,
|
| 650 |
+
"num_train_epochs": 5,
|
| 651 |
+
"total_flos": 1.5001363873057997e+17,
|
| 652 |
+
"trial_name": null,
|
| 653 |
+
"trial_params": null
|
| 654 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fb356b90ad8bddd41c1f2f53f7561cb1c308fa28e714dcae8faa9163fe75d960
|
| 3 |
+
size 2543
|
vocab.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|