Upload folder using huggingface_hub
Browse files- config.json +30 -0
- optimizer.pt +3 -0
- pytorch_model.bin +3 -0
- scheduler.pt +3 -0
- trainer_state.json +280 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "bert-base-uncased",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"BertForSequenceClassification"
|
| 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 |
+
"id2label": {
|
| 12 |
+
"0": "LABEL_0"
|
| 13 |
+
},
|
| 14 |
+
"initializer_range": 0.02,
|
| 15 |
+
"intermediate_size": 3072,
|
| 16 |
+
"label2id": {
|
| 17 |
+
"LABEL_0": 0
|
| 18 |
+
},
|
| 19 |
+
"layer_norm_eps": 1e-12,
|
| 20 |
+
"max_position_embeddings": 512,
|
| 21 |
+
"model_type": "bert",
|
| 22 |
+
"num_attention_heads": 12,
|
| 23 |
+
"num_hidden_layers": 12,
|
| 24 |
+
"pad_token_id": 0,
|
| 25 |
+
"position_embedding_type": "absolute",
|
| 26 |
+
"transformers_version": "4.6.0.dev0",
|
| 27 |
+
"type_vocab_size": 2,
|
| 28 |
+
"use_cache": true,
|
| 29 |
+
"vocab_size": 30522
|
| 30 |
+
}
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3bad7c0b0c2129abdb0fdedc2439e7e9041849b18ac807a4cee63c7c17d5f29f
|
| 3 |
+
size 875990225
|
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:440ae8efd24ba003843a5063a9f6b765c75b727d85aa9825e707d5d4f742cbe6
|
| 3 |
+
size 438021385
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b2b673570378c3a376ad3c1ed4c05609e0d0b151b3c2d9225d4d9311690e00bd
|
| 3 |
+
size 623
|
trainer_state.json
ADDED
|
@@ -0,0 +1,280 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.0637156292796135,
|
| 3 |
+
"best_model_checkpoint": "training_output/reranker_focal_meta_answer/checkpoint-15132",
|
| 4 |
+
"epoch": 12.0,
|
| 5 |
+
"global_step": 15132,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.4,
|
| 12 |
+
"learning_rate": 4.960348929421095e-05,
|
| 13 |
+
"loss": 0.0291,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.79,
|
| 18 |
+
"learning_rate": 4.920697858842189e-05,
|
| 19 |
+
"loss": 0.0124,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 1.0,
|
| 24 |
+
"eval_loss": 0.07028326392173767,
|
| 25 |
+
"eval_runtime": 70.3628,
|
| 26 |
+
"eval_samples_per_second": 561.22,
|
| 27 |
+
"step": 1261
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 1.19,
|
| 31 |
+
"learning_rate": 4.881046788263283e-05,
|
| 32 |
+
"loss": 0.0106,
|
| 33 |
+
"step": 1500
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 1.59,
|
| 37 |
+
"learning_rate": 4.841395717684378e-05,
|
| 38 |
+
"loss": 0.01,
|
| 39 |
+
"step": 2000
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 1.98,
|
| 43 |
+
"learning_rate": 4.8017446471054725e-05,
|
| 44 |
+
"loss": 0.0097,
|
| 45 |
+
"step": 2500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 2.0,
|
| 49 |
+
"eval_loss": 0.06599269062280655,
|
| 50 |
+
"eval_runtime": 70.3845,
|
| 51 |
+
"eval_samples_per_second": 561.047,
|
| 52 |
+
"step": 2522
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 2.38,
|
| 56 |
+
"learning_rate": 4.7620935765265664e-05,
|
| 57 |
+
"loss": 0.0088,
|
| 58 |
+
"step": 3000
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 2.78,
|
| 62 |
+
"learning_rate": 4.722442505947661e-05,
|
| 63 |
+
"loss": 0.009,
|
| 64 |
+
"step": 3500
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 3.0,
|
| 68 |
+
"eval_loss": 0.0710427388548851,
|
| 69 |
+
"eval_runtime": 70.4108,
|
| 70 |
+
"eval_samples_per_second": 560.838,
|
| 71 |
+
"step": 3783
|
| 72 |
+
},
|
| 73 |
+
{
|
| 74 |
+
"epoch": 3.17,
|
| 75 |
+
"learning_rate": 4.682791435368755e-05,
|
| 76 |
+
"loss": 0.0085,
|
| 77 |
+
"step": 4000
|
| 78 |
+
},
|
| 79 |
+
{
|
| 80 |
+
"epoch": 3.57,
|
| 81 |
+
"learning_rate": 4.6431403647898495e-05,
|
| 82 |
+
"loss": 0.008,
|
| 83 |
+
"step": 4500
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 3.97,
|
| 87 |
+
"learning_rate": 4.6034892942109434e-05,
|
| 88 |
+
"loss": 0.0078,
|
| 89 |
+
"step": 5000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 4.0,
|
| 93 |
+
"eval_loss": 0.07019173353910446,
|
| 94 |
+
"eval_runtime": 70.4591,
|
| 95 |
+
"eval_samples_per_second": 560.452,
|
| 96 |
+
"step": 5044
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 4.36,
|
| 100 |
+
"learning_rate": 4.563838223632038e-05,
|
| 101 |
+
"loss": 0.007,
|
| 102 |
+
"step": 5500
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 4.76,
|
| 106 |
+
"learning_rate": 4.5241871530531326e-05,
|
| 107 |
+
"loss": 0.0071,
|
| 108 |
+
"step": 6000
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 5.0,
|
| 112 |
+
"eval_loss": 0.06962954998016357,
|
| 113 |
+
"eval_runtime": 70.4328,
|
| 114 |
+
"eval_samples_per_second": 560.662,
|
| 115 |
+
"step": 6305
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 5.15,
|
| 119 |
+
"learning_rate": 4.484536082474227e-05,
|
| 120 |
+
"loss": 0.0066,
|
| 121 |
+
"step": 6500
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 5.55,
|
| 125 |
+
"learning_rate": 4.444885011895321e-05,
|
| 126 |
+
"loss": 0.0062,
|
| 127 |
+
"step": 7000
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 5.95,
|
| 131 |
+
"learning_rate": 4.405233941316416e-05,
|
| 132 |
+
"loss": 0.0063,
|
| 133 |
+
"step": 7500
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 6.0,
|
| 137 |
+
"eval_loss": 0.07188139855861664,
|
| 138 |
+
"eval_runtime": 70.4837,
|
| 139 |
+
"eval_samples_per_second": 560.257,
|
| 140 |
+
"step": 7566
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 6.34,
|
| 144 |
+
"learning_rate": 4.36558287073751e-05,
|
| 145 |
+
"loss": 0.0056,
|
| 146 |
+
"step": 8000
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 6.74,
|
| 150 |
+
"learning_rate": 4.325931800158605e-05,
|
| 151 |
+
"loss": 0.0054,
|
| 152 |
+
"step": 8500
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 7.0,
|
| 156 |
+
"eval_loss": 0.06544803828001022,
|
| 157 |
+
"eval_runtime": 70.4941,
|
| 158 |
+
"eval_samples_per_second": 560.175,
|
| 159 |
+
"step": 8827
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 7.14,
|
| 163 |
+
"learning_rate": 4.286280729579699e-05,
|
| 164 |
+
"loss": 0.0053,
|
| 165 |
+
"step": 9000
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 7.53,
|
| 169 |
+
"learning_rate": 4.2466296590007934e-05,
|
| 170 |
+
"loss": 0.0048,
|
| 171 |
+
"step": 9500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 7.93,
|
| 175 |
+
"learning_rate": 4.206978588421887e-05,
|
| 176 |
+
"loss": 0.0049,
|
| 177 |
+
"step": 10000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 8.0,
|
| 181 |
+
"eval_loss": 0.06971824169158936,
|
| 182 |
+
"eval_runtime": 70.4814,
|
| 183 |
+
"eval_samples_per_second": 560.276,
|
| 184 |
+
"step": 10088
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 8.33,
|
| 188 |
+
"learning_rate": 4.167327517842982e-05,
|
| 189 |
+
"loss": 0.0045,
|
| 190 |
+
"step": 10500
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 8.72,
|
| 194 |
+
"learning_rate": 4.127676447264076e-05,
|
| 195 |
+
"loss": 0.0043,
|
| 196 |
+
"step": 11000
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 9.0,
|
| 200 |
+
"eval_loss": 0.06489751487970352,
|
| 201 |
+
"eval_runtime": 70.4921,
|
| 202 |
+
"eval_samples_per_second": 560.19,
|
| 203 |
+
"step": 11349
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 9.12,
|
| 207 |
+
"learning_rate": 4.0880253766851704e-05,
|
| 208 |
+
"loss": 0.0043,
|
| 209 |
+
"step": 11500
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 9.52,
|
| 213 |
+
"learning_rate": 4.048374306106265e-05,
|
| 214 |
+
"loss": 0.0039,
|
| 215 |
+
"step": 12000
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 9.91,
|
| 219 |
+
"learning_rate": 4.0087232355273596e-05,
|
| 220 |
+
"loss": 0.0042,
|
| 221 |
+
"step": 12500
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 10.0,
|
| 225 |
+
"eval_loss": 0.06486855447292328,
|
| 226 |
+
"eval_runtime": 70.4903,
|
| 227 |
+
"eval_samples_per_second": 560.205,
|
| 228 |
+
"step": 12610
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 10.31,
|
| 232 |
+
"learning_rate": 3.9690721649484535e-05,
|
| 233 |
+
"loss": 0.0037,
|
| 234 |
+
"step": 13000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 10.71,
|
| 238 |
+
"learning_rate": 3.929421094369548e-05,
|
| 239 |
+
"loss": 0.0037,
|
| 240 |
+
"step": 13500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 11.0,
|
| 244 |
+
"eval_loss": 0.06568709015846252,
|
| 245 |
+
"eval_runtime": 70.5212,
|
| 246 |
+
"eval_samples_per_second": 559.96,
|
| 247 |
+
"step": 13871
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 11.1,
|
| 251 |
+
"learning_rate": 3.889770023790643e-05,
|
| 252 |
+
"loss": 0.0036,
|
| 253 |
+
"step": 14000
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 11.5,
|
| 257 |
+
"learning_rate": 3.850118953211737e-05,
|
| 258 |
+
"loss": 0.0033,
|
| 259 |
+
"step": 14500
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 11.9,
|
| 263 |
+
"learning_rate": 3.810467882632831e-05,
|
| 264 |
+
"loss": 0.0034,
|
| 265 |
+
"step": 15000
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 12.0,
|
| 269 |
+
"eval_loss": 0.0637156292796135,
|
| 270 |
+
"eval_runtime": 70.5037,
|
| 271 |
+
"eval_samples_per_second": 560.098,
|
| 272 |
+
"step": 15132
|
| 273 |
+
}
|
| 274 |
+
],
|
| 275 |
+
"max_steps": 63050,
|
| 276 |
+
"num_train_epochs": 50,
|
| 277 |
+
"total_flos": 8.14042362618455e+16,
|
| 278 |
+
"trial_name": null,
|
| 279 |
+
"trial_params": null
|
| 280 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e47d37c7f5716898c67a25f964154498891c64d9833262d7f7fc28cc46c8e681
|
| 3 |
+
size 2351
|