End of training
Browse files- all_results.json +7 -0
- train_results.json +7 -0
- trainer_state.json +310 -0
all_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"train_loss": 1.997600625588558,
|
| 4 |
+
"train_runtime": 109033.1022,
|
| 5 |
+
"train_samples_per_second": 16.274,
|
| 6 |
+
"train_steps_per_second": 0.17
|
| 7 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 4.0,
|
| 3 |
+
"train_loss": 1.997600625588558,
|
| 4 |
+
"train_runtime": 109033.1022,
|
| 5 |
+
"train_samples_per_second": 16.274,
|
| 6 |
+
"train_steps_per_second": 0.17
|
| 7 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,310 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 4.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 18484,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.11,
|
| 13 |
+
"learning_rate": 1.945899156026834e-05,
|
| 14 |
+
"loss": 2.5051,
|
| 15 |
+
"step": 500
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.22,
|
| 19 |
+
"learning_rate": 1.8917983120536682e-05,
|
| 20 |
+
"loss": 2.3264,
|
| 21 |
+
"step": 1000
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.32,
|
| 25 |
+
"learning_rate": 1.837697468080502e-05,
|
| 26 |
+
"loss": 2.2657,
|
| 27 |
+
"step": 1500
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 0.43,
|
| 31 |
+
"learning_rate": 1.7835966241073363e-05,
|
| 32 |
+
"loss": 2.2257,
|
| 33 |
+
"step": 2000
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.54,
|
| 37 |
+
"learning_rate": 1.7294957801341702e-05,
|
| 38 |
+
"loss": 2.1931,
|
| 39 |
+
"step": 2500
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"epoch": 0.65,
|
| 43 |
+
"learning_rate": 1.6753949361610044e-05,
|
| 44 |
+
"loss": 2.1735,
|
| 45 |
+
"step": 3000
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.76,
|
| 49 |
+
"learning_rate": 1.6212940921878382e-05,
|
| 50 |
+
"loss": 2.1473,
|
| 51 |
+
"step": 3500
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.87,
|
| 55 |
+
"learning_rate": 1.5671932482146724e-05,
|
| 56 |
+
"loss": 2.1401,
|
| 57 |
+
"step": 4000
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.97,
|
| 61 |
+
"learning_rate": 1.5130924042415063e-05,
|
| 62 |
+
"loss": 2.1171,
|
| 63 |
+
"step": 4500
|
| 64 |
+
},
|
| 65 |
+
{
|
| 66 |
+
"epoch": 1.0,
|
| 67 |
+
"eval_f1": 0.8773,
|
| 68 |
+
"eval_gen_len": 16.4704,
|
| 69 |
+
"eval_loss": 2.013514280319214,
|
| 70 |
+
"eval_precision": 0.8836,
|
| 71 |
+
"eval_recall": 0.8717,
|
| 72 |
+
"eval_rouge1": 0.3138,
|
| 73 |
+
"eval_rouge2": 0.1556,
|
| 74 |
+
"eval_rougeL": 0.2853,
|
| 75 |
+
"eval_rougeLsum": 0.2853,
|
| 76 |
+
"eval_runtime": 1599.8646,
|
| 77 |
+
"eval_samples_per_second": 6.251,
|
| 78 |
+
"eval_steps_per_second": 0.391,
|
| 79 |
+
"step": 4621
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.08,
|
| 83 |
+
"learning_rate": 1.4589915602683402e-05,
|
| 84 |
+
"loss": 2.0523,
|
| 85 |
+
"step": 5000
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 1.19,
|
| 89 |
+
"learning_rate": 1.4048907162951744e-05,
|
| 90 |
+
"loss": 2.0306,
|
| 91 |
+
"step": 5500
|
| 92 |
+
},
|
| 93 |
+
{
|
| 94 |
+
"epoch": 1.3,
|
| 95 |
+
"learning_rate": 1.3507898723220082e-05,
|
| 96 |
+
"loss": 2.0262,
|
| 97 |
+
"step": 6000
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 1.41,
|
| 101 |
+
"learning_rate": 1.2966890283488423e-05,
|
| 102 |
+
"loss": 2.0169,
|
| 103 |
+
"step": 6500
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 1.51,
|
| 107 |
+
"learning_rate": 1.2425881843756765e-05,
|
| 108 |
+
"loss": 2.0101,
|
| 109 |
+
"step": 7000
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 1.62,
|
| 113 |
+
"learning_rate": 1.1884873404025104e-05,
|
| 114 |
+
"loss": 1.9955,
|
| 115 |
+
"step": 7500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 1.73,
|
| 119 |
+
"learning_rate": 1.1343864964293442e-05,
|
| 120 |
+
"loss": 1.9943,
|
| 121 |
+
"step": 8000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.84,
|
| 125 |
+
"learning_rate": 1.0802856524561784e-05,
|
| 126 |
+
"loss": 1.9863,
|
| 127 |
+
"step": 8500
|
| 128 |
+
},
|
| 129 |
+
{
|
| 130 |
+
"epoch": 1.95,
|
| 131 |
+
"learning_rate": 1.0261848084830125e-05,
|
| 132 |
+
"loss": 1.9804,
|
| 133 |
+
"step": 9000
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"epoch": 2.0,
|
| 137 |
+
"eval_f1": 0.8775,
|
| 138 |
+
"eval_gen_len": 16.2207,
|
| 139 |
+
"eval_loss": 1.9440101385116577,
|
| 140 |
+
"eval_precision": 0.8831,
|
| 141 |
+
"eval_recall": 0.8725,
|
| 142 |
+
"eval_rouge1": 0.3147,
|
| 143 |
+
"eval_rouge2": 0.1581,
|
| 144 |
+
"eval_rougeL": 0.2864,
|
| 145 |
+
"eval_rougeLsum": 0.2866,
|
| 146 |
+
"eval_runtime": 1574.2726,
|
| 147 |
+
"eval_samples_per_second": 6.352,
|
| 148 |
+
"eval_steps_per_second": 0.397,
|
| 149 |
+
"step": 9242
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 2.06,
|
| 153 |
+
"learning_rate": 9.720839645098465e-06,
|
| 154 |
+
"loss": 1.9455,
|
| 155 |
+
"step": 9500
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 2.16,
|
| 159 |
+
"learning_rate": 9.179831205366805e-06,
|
| 160 |
+
"loss": 1.9171,
|
| 161 |
+
"step": 10000
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 2.27,
|
| 165 |
+
"learning_rate": 8.638822765635144e-06,
|
| 166 |
+
"loss": 1.9211,
|
| 167 |
+
"step": 10500
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 2.38,
|
| 171 |
+
"learning_rate": 8.097814325903484e-06,
|
| 172 |
+
"loss": 1.9122,
|
| 173 |
+
"step": 11000
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 2.49,
|
| 177 |
+
"learning_rate": 7.5568058861718255e-06,
|
| 178 |
+
"loss": 1.9126,
|
| 179 |
+
"step": 11500
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 2.6,
|
| 183 |
+
"learning_rate": 7.015797446440165e-06,
|
| 184 |
+
"loss": 1.9065,
|
| 185 |
+
"step": 12000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 2.71,
|
| 189 |
+
"learning_rate": 6.474789006708505e-06,
|
| 190 |
+
"loss": 1.8955,
|
| 191 |
+
"step": 12500
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 2.81,
|
| 195 |
+
"learning_rate": 5.933780566976846e-06,
|
| 196 |
+
"loss": 1.899,
|
| 197 |
+
"step": 13000
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 2.92,
|
| 201 |
+
"learning_rate": 5.392772127245185e-06,
|
| 202 |
+
"loss": 1.8971,
|
| 203 |
+
"step": 13500
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 3.0,
|
| 207 |
+
"eval_f1": 0.8792,
|
| 208 |
+
"eval_gen_len": 15.4676,
|
| 209 |
+
"eval_loss": 1.9157460927963257,
|
| 210 |
+
"eval_precision": 0.8857,
|
| 211 |
+
"eval_recall": 0.8733,
|
| 212 |
+
"eval_rouge1": 0.3209,
|
| 213 |
+
"eval_rouge2": 0.1638,
|
| 214 |
+
"eval_rougeL": 0.2925,
|
| 215 |
+
"eval_rougeLsum": 0.2926,
|
| 216 |
+
"eval_runtime": 1567.1343,
|
| 217 |
+
"eval_samples_per_second": 6.381,
|
| 218 |
+
"eval_steps_per_second": 0.399,
|
| 219 |
+
"step": 13863
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 3.03,
|
| 223 |
+
"learning_rate": 4.8517636875135255e-06,
|
| 224 |
+
"loss": 1.8862,
|
| 225 |
+
"step": 14000
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 3.14,
|
| 229 |
+
"learning_rate": 4.310755247781866e-06,
|
| 230 |
+
"loss": 1.8611,
|
| 231 |
+
"step": 14500
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 3.25,
|
| 235 |
+
"learning_rate": 3.769746808050206e-06,
|
| 236 |
+
"loss": 1.8474,
|
| 237 |
+
"step": 15000
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 3.35,
|
| 241 |
+
"learning_rate": 3.2287383683185457e-06,
|
| 242 |
+
"loss": 1.8523,
|
| 243 |
+
"step": 15500
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 3.46,
|
| 247 |
+
"learning_rate": 2.687729928586886e-06,
|
| 248 |
+
"loss": 1.8462,
|
| 249 |
+
"step": 16000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 3.57,
|
| 253 |
+
"learning_rate": 2.1467214888552264e-06,
|
| 254 |
+
"loss": 1.8455,
|
| 255 |
+
"step": 16500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 3.68,
|
| 259 |
+
"learning_rate": 1.6057130491235664e-06,
|
| 260 |
+
"loss": 1.8456,
|
| 261 |
+
"step": 17000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 3.79,
|
| 265 |
+
"learning_rate": 1.0647046093919065e-06,
|
| 266 |
+
"loss": 1.8425,
|
| 267 |
+
"step": 17500
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 3.9,
|
| 271 |
+
"learning_rate": 5.236961696602468e-07,
|
| 272 |
+
"loss": 1.8449,
|
| 273 |
+
"step": 18000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 4.0,
|
| 277 |
+
"eval_f1": 0.8795,
|
| 278 |
+
"eval_gen_len": 15.7946,
|
| 279 |
+
"eval_loss": 1.9021437168121338,
|
| 280 |
+
"eval_precision": 0.8858,
|
| 281 |
+
"eval_recall": 0.8739,
|
| 282 |
+
"eval_rouge1": 0.3236,
|
| 283 |
+
"eval_rouge2": 0.1651,
|
| 284 |
+
"eval_rougeL": 0.2953,
|
| 285 |
+
"eval_rougeLsum": 0.2953,
|
| 286 |
+
"eval_runtime": 1613.4446,
|
| 287 |
+
"eval_samples_per_second": 6.198,
|
| 288 |
+
"eval_steps_per_second": 0.387,
|
| 289 |
+
"step": 18484
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 4.0,
|
| 293 |
+
"step": 18484,
|
| 294 |
+
"total_flos": 3.845275732313899e+18,
|
| 295 |
+
"train_loss": 1.997600625588558,
|
| 296 |
+
"train_runtime": 109033.1022,
|
| 297 |
+
"train_samples_per_second": 16.274,
|
| 298 |
+
"train_steps_per_second": 0.17
|
| 299 |
+
}
|
| 300 |
+
],
|
| 301 |
+
"logging_steps": 500,
|
| 302 |
+
"max_steps": 18484,
|
| 303 |
+
"num_input_tokens_seen": 0,
|
| 304 |
+
"num_train_epochs": 4,
|
| 305 |
+
"save_steps": 500,
|
| 306 |
+
"total_flos": 3.845275732313899e+18,
|
| 307 |
+
"train_batch_size": 24,
|
| 308 |
+
"trial_name": null,
|
| 309 |
+
"trial_params": null
|
| 310 |
+
}
|