End of training
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_exact_match": 56.09375,
|
| 4 |
+
"eval_f1": 69.86920835599972,
|
| 5 |
+
"eval_runtime": 45.3133,
|
| 6 |
+
"eval_samples": 4052,
|
| 7 |
+
"eval_samples_per_second": 89.422,
|
| 8 |
+
"eval_steps_per_second": 3.2,
|
| 9 |
+
"total_flos": 3.780898745780736e+16,
|
| 10 |
+
"train_loss": 1.4068151059358016,
|
| 11 |
+
"train_runtime": 4389.3737,
|
| 12 |
+
"train_samples": 19293,
|
| 13 |
+
"train_samples_per_second": 43.954,
|
| 14 |
+
"train_steps_per_second": 1.572
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_exact_match": 56.09375,
|
| 4 |
+
"eval_f1": 69.86920835599972,
|
| 5 |
+
"eval_runtime": 45.3133,
|
| 6 |
+
"eval_samples": 4052,
|
| 7 |
+
"eval_samples_per_second": 89.422,
|
| 8 |
+
"eval_steps_per_second": 3.2
|
| 9 |
+
}
|
runs/May09_17-45-46_ml-machine/events.out.tfevents.1715281307.ml-machine.68332.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7b42d5297799129d41d8f60249f8508847232d9129a3b132366b3e5772cb3a24
|
| 3 |
+
size 412
|
train_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"total_flos": 3.780898745780736e+16,
|
| 4 |
+
"train_loss": 1.4068151059358016,
|
| 5 |
+
"train_runtime": 4389.3737,
|
| 6 |
+
"train_samples": 19293,
|
| 7 |
+
"train_samples_per_second": 43.954,
|
| 8 |
+
"train_steps_per_second": 1.572
|
| 9 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,280 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 69.86920835599972,
|
| 3 |
+
"best_model_checkpoint": "/root/turkic_qa/en_uzn_models/en_uzn_xlm_roberta_base_model/checkpoint-6900",
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 6900,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 1.0,
|
| 13 |
+
"step": 690,
|
| 14 |
+
"train_exact_match": 23.076923076923077,
|
| 15 |
+
"train_f1": 37.566079561862445,
|
| 16 |
+
"train_runtime": 14.0841,
|
| 17 |
+
"train_samples_per_second": 89.604,
|
| 18 |
+
"train_steps_per_second": 3.266
|
| 19 |
+
},
|
| 20 |
+
{
|
| 21 |
+
"epoch": 1.0,
|
| 22 |
+
"grad_norm": 116.39849853515625,
|
| 23 |
+
"learning_rate": 5e-06,
|
| 24 |
+
"loss": 4.5006,
|
| 25 |
+
"step": 690
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 1.0,
|
| 29 |
+
"eval_exact_match": 21.46875,
|
| 30 |
+
"eval_f1": 34.27477969652227,
|
| 31 |
+
"eval_runtime": 44.8553,
|
| 32 |
+
"eval_samples_per_second": 90.335,
|
| 33 |
+
"eval_steps_per_second": 3.233,
|
| 34 |
+
"step": 690
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 2.0,
|
| 38 |
+
"step": 1380,
|
| 39 |
+
"train_exact_match": 51.04895104895105,
|
| 40 |
+
"train_f1": 64.28880003049197,
|
| 41 |
+
"train_runtime": 14.6272,
|
| 42 |
+
"train_samples_per_second": 87.03,
|
| 43 |
+
"train_steps_per_second": 3.145
|
| 44 |
+
},
|
| 45 |
+
{
|
| 46 |
+
"epoch": 2.0,
|
| 47 |
+
"grad_norm": 242.94326782226562,
|
| 48 |
+
"learning_rate": 1e-05,
|
| 49 |
+
"loss": 2.1628,
|
| 50 |
+
"step": 1380
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 2.0,
|
| 54 |
+
"eval_exact_match": 45.46875,
|
| 55 |
+
"eval_f1": 60.28283429101314,
|
| 56 |
+
"eval_runtime": 46.3266,
|
| 57 |
+
"eval_samples_per_second": 87.466,
|
| 58 |
+
"eval_steps_per_second": 3.13,
|
| 59 |
+
"step": 1380
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 3.0,
|
| 63 |
+
"step": 2070,
|
| 64 |
+
"train_exact_match": 57.142857142857146,
|
| 65 |
+
"train_f1": 71.92623291529159,
|
| 66 |
+
"train_runtime": 14.5466,
|
| 67 |
+
"train_samples_per_second": 89.024,
|
| 68 |
+
"train_steps_per_second": 3.231
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 3.0,
|
| 72 |
+
"grad_norm": 18.747507095336914,
|
| 73 |
+
"learning_rate": 8.750000000000001e-06,
|
| 74 |
+
"loss": 1.4946,
|
| 75 |
+
"step": 2070
|
| 76 |
+
},
|
| 77 |
+
{
|
| 78 |
+
"epoch": 3.0,
|
| 79 |
+
"eval_exact_match": 52.40625,
|
| 80 |
+
"eval_f1": 67.09147147337285,
|
| 81 |
+
"eval_runtime": 45.3354,
|
| 82 |
+
"eval_samples_per_second": 89.378,
|
| 83 |
+
"eval_steps_per_second": 3.198,
|
| 84 |
+
"step": 2070
|
| 85 |
+
},
|
| 86 |
+
{
|
| 87 |
+
"epoch": 4.0,
|
| 88 |
+
"step": 2760,
|
| 89 |
+
"train_exact_match": 64.13586413586414,
|
| 90 |
+
"train_f1": 78.3731582090412,
|
| 91 |
+
"train_runtime": 14.5689,
|
| 92 |
+
"train_samples_per_second": 88.888,
|
| 93 |
+
"train_steps_per_second": 3.226
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 4.0,
|
| 97 |
+
"grad_norm": 269.3212890625,
|
| 98 |
+
"learning_rate": 7.500000000000001e-06,
|
| 99 |
+
"loss": 1.2186,
|
| 100 |
+
"step": 2760
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 4.0,
|
| 104 |
+
"eval_exact_match": 53.96875,
|
| 105 |
+
"eval_f1": 68.38825732557214,
|
| 106 |
+
"eval_runtime": 45.3562,
|
| 107 |
+
"eval_samples_per_second": 89.337,
|
| 108 |
+
"eval_steps_per_second": 3.197,
|
| 109 |
+
"step": 2760
|
| 110 |
+
},
|
| 111 |
+
{
|
| 112 |
+
"epoch": 5.0,
|
| 113 |
+
"step": 3450,
|
| 114 |
+
"train_exact_match": 68.33166833166833,
|
| 115 |
+
"train_f1": 82.36153774583077,
|
| 116 |
+
"train_runtime": 13.878,
|
| 117 |
+
"train_samples_per_second": 89.278,
|
| 118 |
+
"train_steps_per_second": 3.243
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 5.0,
|
| 122 |
+
"grad_norm": 259.8070068359375,
|
| 123 |
+
"learning_rate": 6.25e-06,
|
| 124 |
+
"loss": 1.0359,
|
| 125 |
+
"step": 3450
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 5.0,
|
| 129 |
+
"eval_exact_match": 54.96875,
|
| 130 |
+
"eval_f1": 69.02164576043303,
|
| 131 |
+
"eval_runtime": 45.1308,
|
| 132 |
+
"eval_samples_per_second": 89.783,
|
| 133 |
+
"eval_steps_per_second": 3.213,
|
| 134 |
+
"step": 3450
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 6.0,
|
| 138 |
+
"step": 4140,
|
| 139 |
+
"train_exact_match": 71.92807192807193,
|
| 140 |
+
"train_f1": 83.477467208777,
|
| 141 |
+
"train_runtime": 14.0926,
|
| 142 |
+
"train_samples_per_second": 89.55,
|
| 143 |
+
"train_steps_per_second": 3.264
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 6.0,
|
| 147 |
+
"grad_norm": 39.62363052368164,
|
| 148 |
+
"learning_rate": 5e-06,
|
| 149 |
+
"loss": 0.8945,
|
| 150 |
+
"step": 4140
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 6.0,
|
| 154 |
+
"eval_exact_match": 55.125,
|
| 155 |
+
"eval_f1": 69.56052410993578,
|
| 156 |
+
"eval_runtime": 44.9507,
|
| 157 |
+
"eval_samples_per_second": 90.143,
|
| 158 |
+
"eval_steps_per_second": 3.226,
|
| 159 |
+
"step": 4140
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 7.0,
|
| 163 |
+
"step": 4830,
|
| 164 |
+
"train_exact_match": 75.62437562437563,
|
| 165 |
+
"train_f1": 86.36686612245309,
|
| 166 |
+
"train_runtime": 14.0278,
|
| 167 |
+
"train_samples_per_second": 88.823,
|
| 168 |
+
"train_steps_per_second": 3.208
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 7.0,
|
| 172 |
+
"grad_norm": 122.51107788085938,
|
| 173 |
+
"learning_rate": 3.7500000000000005e-06,
|
| 174 |
+
"loss": 0.7916,
|
| 175 |
+
"step": 4830
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 7.0,
|
| 179 |
+
"eval_exact_match": 55.4375,
|
| 180 |
+
"eval_f1": 69.46738777882736,
|
| 181 |
+
"eval_runtime": 45.1053,
|
| 182 |
+
"eval_samples_per_second": 89.834,
|
| 183 |
+
"eval_steps_per_second": 3.215,
|
| 184 |
+
"step": 4830
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 8.0,
|
| 188 |
+
"step": 5520,
|
| 189 |
+
"train_exact_match": 75.92407592407592,
|
| 190 |
+
"train_f1": 86.72722282271204,
|
| 191 |
+
"train_runtime": 14.2506,
|
| 192 |
+
"train_samples_per_second": 89.75,
|
| 193 |
+
"train_steps_per_second": 3.228
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 8.0,
|
| 197 |
+
"grad_norm": 202.73541259765625,
|
| 198 |
+
"learning_rate": 2.5e-06,
|
| 199 |
+
"loss": 0.7119,
|
| 200 |
+
"step": 5520
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 8.0,
|
| 204 |
+
"eval_exact_match": 56.0,
|
| 205 |
+
"eval_f1": 69.69766614851221,
|
| 206 |
+
"eval_runtime": 45.2302,
|
| 207 |
+
"eval_samples_per_second": 89.586,
|
| 208 |
+
"eval_steps_per_second": 3.206,
|
| 209 |
+
"step": 5520
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 9.0,
|
| 213 |
+
"step": 6210,
|
| 214 |
+
"train_exact_match": 77.32267732267732,
|
| 215 |
+
"train_f1": 88.72058931170982,
|
| 216 |
+
"train_runtime": 14.7856,
|
| 217 |
+
"train_samples_per_second": 87.653,
|
| 218 |
+
"train_steps_per_second": 3.179
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 9.0,
|
| 222 |
+
"grad_norm": 5.172795295715332,
|
| 223 |
+
"learning_rate": 1.25e-06,
|
| 224 |
+
"loss": 0.6445,
|
| 225 |
+
"step": 6210
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 9.0,
|
| 229 |
+
"eval_exact_match": 55.90625,
|
| 230 |
+
"eval_f1": 69.57114295943043,
|
| 231 |
+
"eval_runtime": 45.9508,
|
| 232 |
+
"eval_samples_per_second": 88.181,
|
| 233 |
+
"eval_steps_per_second": 3.156,
|
| 234 |
+
"step": 6210
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 10.0,
|
| 238 |
+
"step": 6900,
|
| 239 |
+
"train_exact_match": 80.81918081918081,
|
| 240 |
+
"train_f1": 90.37663245353303,
|
| 241 |
+
"train_runtime": 14.2722,
|
| 242 |
+
"train_samples_per_second": 88.634,
|
| 243 |
+
"train_steps_per_second": 3.223
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 10.0,
|
| 247 |
+
"grad_norm": 6.113234996795654,
|
| 248 |
+
"learning_rate": 0.0,
|
| 249 |
+
"loss": 0.6132,
|
| 250 |
+
"step": 6900
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 10.0,
|
| 254 |
+
"eval_exact_match": 56.09375,
|
| 255 |
+
"eval_f1": 69.86920835599972,
|
| 256 |
+
"eval_runtime": 44.9954,
|
| 257 |
+
"eval_samples_per_second": 90.054,
|
| 258 |
+
"eval_steps_per_second": 3.223,
|
| 259 |
+
"step": 6900
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 10.0,
|
| 263 |
+
"step": 6900,
|
| 264 |
+
"total_flos": 3.780898745780736e+16,
|
| 265 |
+
"train_loss": 1.4068151059358016,
|
| 266 |
+
"train_runtime": 4389.3737,
|
| 267 |
+
"train_samples_per_second": 43.954,
|
| 268 |
+
"train_steps_per_second": 1.572
|
| 269 |
+
}
|
| 270 |
+
],
|
| 271 |
+
"logging_steps": 500,
|
| 272 |
+
"max_steps": 6900,
|
| 273 |
+
"num_input_tokens_seen": 0,
|
| 274 |
+
"num_train_epochs": 10,
|
| 275 |
+
"save_steps": 500,
|
| 276 |
+
"total_flos": 3.780898745780736e+16,
|
| 277 |
+
"train_batch_size": 28,
|
| 278 |
+
"trial_name": null,
|
| 279 |
+
"trial_params": null
|
| 280 |
+
}
|