Training in progress, step 500, checkpoint
Browse files- last-checkpoint/config.json +1 -1
- last-checkpoint/model.safetensors +2 -2
- last-checkpoint/optimizer.pt +2 -2
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/tokenizer.json +0 -0
- last-checkpoint/tokenizer_config.json +4 -4
- last-checkpoint/trainer_state.json +5 -649
- last-checkpoint/training_args.bin +1 -1
- last-checkpoint/vocab.txt +0 -0
last-checkpoint/config.json
CHANGED
|
@@ -27,5 +27,5 @@
|
|
| 27 |
"transformers_version": "4.41.2",
|
| 28 |
"type_vocab_size": 2,
|
| 29 |
"use_cache": true,
|
| 30 |
-
"vocab_size":
|
| 31 |
}
|
|
|
|
| 27 |
"transformers_version": "4.41.2",
|
| 28 |
"type_vocab_size": 2,
|
| 29 |
"use_cache": true,
|
| 30 |
+
"vocab_size": 176008
|
| 31 |
}
|
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ac2d545e95d53dbd847e3dfb313fc59d64125bf026261ddc71edbe677f2edb81
|
| 3 |
+
size 885595848
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dbc5e79a5f32a4ee4d255a20aeadc3cfbad6eb4cc0d92611370c0dfe67fc9dd5
|
| 3 |
+
size 1771313210
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fd920835b7e1d3f942ead294d5d63bc784d9e6ec27e5553642cbea5f2c9e77c4
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a5252eff43b3c3312cd40cbc7cbc932038114747237577193618542ca149ff6e
|
| 3 |
size 1064
|
last-checkpoint/tokenizer.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
last-checkpoint/tokenizer_config.json
CHANGED
|
@@ -8,7 +8,7 @@
|
|
| 8 |
"single_word": false,
|
| 9 |
"special": true
|
| 10 |
},
|
| 11 |
-
"
|
| 12 |
"content": "[UNK]",
|
| 13 |
"lstrip": false,
|
| 14 |
"normalized": false,
|
|
@@ -16,7 +16,7 @@
|
|
| 16 |
"single_word": false,
|
| 17 |
"special": true
|
| 18 |
},
|
| 19 |
-
"
|
| 20 |
"content": "[CLS]",
|
| 21 |
"lstrip": false,
|
| 22 |
"normalized": false,
|
|
@@ -24,7 +24,7 @@
|
|
| 24 |
"single_word": false,
|
| 25 |
"special": true
|
| 26 |
},
|
| 27 |
-
"
|
| 28 |
"content": "[SEP]",
|
| 29 |
"lstrip": false,
|
| 30 |
"normalized": false,
|
|
@@ -32,7 +32,7 @@
|
|
| 32 |
"single_word": false,
|
| 33 |
"special": true
|
| 34 |
},
|
| 35 |
-
"
|
| 36 |
"content": "[MASK]",
|
| 37 |
"lstrip": false,
|
| 38 |
"normalized": false,
|
|
|
|
| 8 |
"single_word": false,
|
| 9 |
"special": true
|
| 10 |
},
|
| 11 |
+
"1": {
|
| 12 |
"content": "[UNK]",
|
| 13 |
"lstrip": false,
|
| 14 |
"normalized": false,
|
|
|
|
| 16 |
"single_word": false,
|
| 17 |
"special": true
|
| 18 |
},
|
| 19 |
+
"2": {
|
| 20 |
"content": "[CLS]",
|
| 21 |
"lstrip": false,
|
| 22 |
"normalized": false,
|
|
|
|
| 24 |
"single_word": false,
|
| 25 |
"special": true
|
| 26 |
},
|
| 27 |
+
"3": {
|
| 28 |
"content": "[SEP]",
|
| 29 |
"lstrip": false,
|
| 30 |
"normalized": false,
|
|
|
|
| 32 |
"single_word": false,
|
| 33 |
"special": true
|
| 34 |
},
|
| 35 |
+
"4": {
|
| 36 |
"content": "[MASK]",
|
| 37 |
"lstrip": false,
|
| 38 |
"normalized": false,
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,663 +1,19 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.014542070209114969,
|
| 13 |
-
"grad_norm":
|
| 14 |
"learning_rate": 1.9903052865272566e-05,
|
| 15 |
-
"loss":
|
| 16 |
"step": 500
|
| 17 |
-
},
|
| 18 |
-
{
|
| 19 |
-
"epoch": 0.029084140418229938,
|
| 20 |
-
"grad_norm": 16.82585906982422,
|
| 21 |
-
"learning_rate": 1.9806105730545135e-05,
|
| 22 |
-
"loss": 0.7196,
|
| 23 |
-
"step": 1000
|
| 24 |
-
},
|
| 25 |
-
{
|
| 26 |
-
"epoch": 0.04362621062734491,
|
| 27 |
-
"grad_norm": 7.649247646331787,
|
| 28 |
-
"learning_rate": 1.9709158595817703e-05,
|
| 29 |
-
"loss": 0.6657,
|
| 30 |
-
"step": 1500
|
| 31 |
-
},
|
| 32 |
-
{
|
| 33 |
-
"epoch": 0.058168280836459876,
|
| 34 |
-
"grad_norm": 15.57618522644043,
|
| 35 |
-
"learning_rate": 1.961221146109027e-05,
|
| 36 |
-
"loss": 0.6333,
|
| 37 |
-
"step": 2000
|
| 38 |
-
},
|
| 39 |
-
{
|
| 40 |
-
"epoch": 0.07271035104557484,
|
| 41 |
-
"grad_norm": 8.144426345825195,
|
| 42 |
-
"learning_rate": 1.9515264326362836e-05,
|
| 43 |
-
"loss": 0.6274,
|
| 44 |
-
"step": 2500
|
| 45 |
-
},
|
| 46 |
-
{
|
| 47 |
-
"epoch": 0.08725242125468982,
|
| 48 |
-
"grad_norm": 5.530069828033447,
|
| 49 |
-
"learning_rate": 1.9418317191635404e-05,
|
| 50 |
-
"loss": 0.5996,
|
| 51 |
-
"step": 3000
|
| 52 |
-
},
|
| 53 |
-
{
|
| 54 |
-
"epoch": 0.10179449146380479,
|
| 55 |
-
"grad_norm": 17.099382400512695,
|
| 56 |
-
"learning_rate": 1.932137005690797e-05,
|
| 57 |
-
"loss": 0.5851,
|
| 58 |
-
"step": 3500
|
| 59 |
-
},
|
| 60 |
-
{
|
| 61 |
-
"epoch": 0.11633656167291975,
|
| 62 |
-
"grad_norm": 2.200979709625244,
|
| 63 |
-
"learning_rate": 1.9224422922180537e-05,
|
| 64 |
-
"loss": 0.6288,
|
| 65 |
-
"step": 4000
|
| 66 |
-
},
|
| 67 |
-
{
|
| 68 |
-
"epoch": 0.13087863188203472,
|
| 69 |
-
"grad_norm": 13.900425910949707,
|
| 70 |
-
"learning_rate": 1.91274757874531e-05,
|
| 71 |
-
"loss": 0.6023,
|
| 72 |
-
"step": 4500
|
| 73 |
-
},
|
| 74 |
-
{
|
| 75 |
-
"epoch": 0.14542070209114968,
|
| 76 |
-
"grad_norm": 2.3364081382751465,
|
| 77 |
-
"learning_rate": 1.903052865272567e-05,
|
| 78 |
-
"loss": 0.5997,
|
| 79 |
-
"step": 5000
|
| 80 |
-
},
|
| 81 |
-
{
|
| 82 |
-
"epoch": 0.15996277230026468,
|
| 83 |
-
"grad_norm": 60.922637939453125,
|
| 84 |
-
"learning_rate": 1.8933581517998238e-05,
|
| 85 |
-
"loss": 0.5695,
|
| 86 |
-
"step": 5500
|
| 87 |
-
},
|
| 88 |
-
{
|
| 89 |
-
"epoch": 0.17450484250937964,
|
| 90 |
-
"grad_norm": 5.009929656982422,
|
| 91 |
-
"learning_rate": 1.8836634383270806e-05,
|
| 92 |
-
"loss": 0.5434,
|
| 93 |
-
"step": 6000
|
| 94 |
-
},
|
| 95 |
-
{
|
| 96 |
-
"epoch": 0.1890469127184946,
|
| 97 |
-
"grad_norm": 10.972416877746582,
|
| 98 |
-
"learning_rate": 1.873968724854337e-05,
|
| 99 |
-
"loss": 0.5832,
|
| 100 |
-
"step": 6500
|
| 101 |
-
},
|
| 102 |
-
{
|
| 103 |
-
"epoch": 0.20358898292760957,
|
| 104 |
-
"grad_norm": 3.8107187747955322,
|
| 105 |
-
"learning_rate": 1.864274011381594e-05,
|
| 106 |
-
"loss": 0.5418,
|
| 107 |
-
"step": 7000
|
| 108 |
-
},
|
| 109 |
-
{
|
| 110 |
-
"epoch": 0.21813105313672454,
|
| 111 |
-
"grad_norm": 1.1021103858947754,
|
| 112 |
-
"learning_rate": 1.8545792979088504e-05,
|
| 113 |
-
"loss": 0.5869,
|
| 114 |
-
"step": 7500
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.2326731233458395,
|
| 118 |
-
"grad_norm": 4.626884460449219,
|
| 119 |
-
"learning_rate": 1.844884584436107e-05,
|
| 120 |
-
"loss": 0.5532,
|
| 121 |
-
"step": 8000
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 0.24721519355495447,
|
| 125 |
-
"grad_norm": 2.25512957572937,
|
| 126 |
-
"learning_rate": 1.8351898709633636e-05,
|
| 127 |
-
"loss": 0.5484,
|
| 128 |
-
"step": 8500
|
| 129 |
-
},
|
| 130 |
-
{
|
| 131 |
-
"epoch": 0.26175726376406944,
|
| 132 |
-
"grad_norm": 10.878036499023438,
|
| 133 |
-
"learning_rate": 1.8254951574906205e-05,
|
| 134 |
-
"loss": 0.517,
|
| 135 |
-
"step": 9000
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 0.2762993339731844,
|
| 139 |
-
"grad_norm": 0.1631862372159958,
|
| 140 |
-
"learning_rate": 1.8158004440178773e-05,
|
| 141 |
-
"loss": 0.5495,
|
| 142 |
-
"step": 9500
|
| 143 |
-
},
|
| 144 |
-
{
|
| 145 |
-
"epoch": 0.29084140418229937,
|
| 146 |
-
"grad_norm": 4.393772125244141,
|
| 147 |
-
"learning_rate": 1.806105730545134e-05,
|
| 148 |
-
"loss": 0.5365,
|
| 149 |
-
"step": 10000
|
| 150 |
-
},
|
| 151 |
-
{
|
| 152 |
-
"epoch": 0.3053834743914144,
|
| 153 |
-
"grad_norm": 3.9716999530792236,
|
| 154 |
-
"learning_rate": 1.7964110170723906e-05,
|
| 155 |
-
"loss": 0.5463,
|
| 156 |
-
"step": 10500
|
| 157 |
-
},
|
| 158 |
-
{
|
| 159 |
-
"epoch": 0.31992554460052935,
|
| 160 |
-
"grad_norm": 16.515634536743164,
|
| 161 |
-
"learning_rate": 1.7867163035996474e-05,
|
| 162 |
-
"loss": 0.5367,
|
| 163 |
-
"step": 11000
|
| 164 |
-
},
|
| 165 |
-
{
|
| 166 |
-
"epoch": 0.3344676148096443,
|
| 167 |
-
"grad_norm": 8.901307106018066,
|
| 168 |
-
"learning_rate": 1.777021590126904e-05,
|
| 169 |
-
"loss": 0.5279,
|
| 170 |
-
"step": 11500
|
| 171 |
-
},
|
| 172 |
-
{
|
| 173 |
-
"epoch": 0.3490096850187593,
|
| 174 |
-
"grad_norm": 5.7714524269104,
|
| 175 |
-
"learning_rate": 1.7673268766541607e-05,
|
| 176 |
-
"loss": 0.5694,
|
| 177 |
-
"step": 12000
|
| 178 |
-
},
|
| 179 |
-
{
|
| 180 |
-
"epoch": 0.36355175522787425,
|
| 181 |
-
"grad_norm": 14.238178253173828,
|
| 182 |
-
"learning_rate": 1.757632163181417e-05,
|
| 183 |
-
"loss": 0.5453,
|
| 184 |
-
"step": 12500
|
| 185 |
-
},
|
| 186 |
-
{
|
| 187 |
-
"epoch": 0.3780938254369892,
|
| 188 |
-
"grad_norm": 2.3930513858795166,
|
| 189 |
-
"learning_rate": 1.747937449708674e-05,
|
| 190 |
-
"loss": 0.5055,
|
| 191 |
-
"step": 13000
|
| 192 |
-
},
|
| 193 |
-
{
|
| 194 |
-
"epoch": 0.3926358956461042,
|
| 195 |
-
"grad_norm": 0.9438181519508362,
|
| 196 |
-
"learning_rate": 1.7382427362359308e-05,
|
| 197 |
-
"loss": 0.5561,
|
| 198 |
-
"step": 13500
|
| 199 |
-
},
|
| 200 |
-
{
|
| 201 |
-
"epoch": 0.40717796585521915,
|
| 202 |
-
"grad_norm": 19.889507293701172,
|
| 203 |
-
"learning_rate": 1.7285480227631876e-05,
|
| 204 |
-
"loss": 0.5267,
|
| 205 |
-
"step": 14000
|
| 206 |
-
},
|
| 207 |
-
{
|
| 208 |
-
"epoch": 0.4217200360643341,
|
| 209 |
-
"grad_norm": 9.12895393371582,
|
| 210 |
-
"learning_rate": 1.718853309290444e-05,
|
| 211 |
-
"loss": 0.5353,
|
| 212 |
-
"step": 14500
|
| 213 |
-
},
|
| 214 |
-
{
|
| 215 |
-
"epoch": 0.4362621062734491,
|
| 216 |
-
"grad_norm": 6.552937030792236,
|
| 217 |
-
"learning_rate": 1.709158595817701e-05,
|
| 218 |
-
"loss": 0.5231,
|
| 219 |
-
"step": 15000
|
| 220 |
-
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.45080417648256405,
|
| 223 |
-
"grad_norm": 6.750959396362305,
|
| 224 |
-
"learning_rate": 1.6994638823449574e-05,
|
| 225 |
-
"loss": 0.4972,
|
| 226 |
-
"step": 15500
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 0.465346246691679,
|
| 230 |
-
"grad_norm": 10.956033706665039,
|
| 231 |
-
"learning_rate": 1.689769168872214e-05,
|
| 232 |
-
"loss": 0.4949,
|
| 233 |
-
"step": 16000
|
| 234 |
-
},
|
| 235 |
-
{
|
| 236 |
-
"epoch": 0.479888316900794,
|
| 237 |
-
"grad_norm": 3.459519863128662,
|
| 238 |
-
"learning_rate": 1.6800744553994706e-05,
|
| 239 |
-
"loss": 0.5219,
|
| 240 |
-
"step": 16500
|
| 241 |
-
},
|
| 242 |
-
{
|
| 243 |
-
"epoch": 0.49443038710990894,
|
| 244 |
-
"grad_norm": 7.870626926422119,
|
| 245 |
-
"learning_rate": 1.6703797419267275e-05,
|
| 246 |
-
"loss": 0.5049,
|
| 247 |
-
"step": 17000
|
| 248 |
-
},
|
| 249 |
-
{
|
| 250 |
-
"epoch": 0.508972457319024,
|
| 251 |
-
"grad_norm": 9.513204574584961,
|
| 252 |
-
"learning_rate": 1.6606850284539843e-05,
|
| 253 |
-
"loss": 0.5365,
|
| 254 |
-
"step": 17500
|
| 255 |
-
},
|
| 256 |
-
{
|
| 257 |
-
"epoch": 0.5235145275281389,
|
| 258 |
-
"grad_norm": 4.7530951499938965,
|
| 259 |
-
"learning_rate": 1.650990314981241e-05,
|
| 260 |
-
"loss": 0.4979,
|
| 261 |
-
"step": 18000
|
| 262 |
-
},
|
| 263 |
-
{
|
| 264 |
-
"epoch": 0.5380565977372539,
|
| 265 |
-
"grad_norm": 4.865274906158447,
|
| 266 |
-
"learning_rate": 1.6412956015084976e-05,
|
| 267 |
-
"loss": 0.5273,
|
| 268 |
-
"step": 18500
|
| 269 |
-
},
|
| 270 |
-
{
|
| 271 |
-
"epoch": 0.5525986679463688,
|
| 272 |
-
"grad_norm": 2.539562940597534,
|
| 273 |
-
"learning_rate": 1.6316008880357544e-05,
|
| 274 |
-
"loss": 0.5029,
|
| 275 |
-
"step": 19000
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 0.5671407381554838,
|
| 279 |
-
"grad_norm": 2.071009874343872,
|
| 280 |
-
"learning_rate": 1.621906174563011e-05,
|
| 281 |
-
"loss": 0.5166,
|
| 282 |
-
"step": 19500
|
| 283 |
-
},
|
| 284 |
-
{
|
| 285 |
-
"epoch": 0.5816828083645987,
|
| 286 |
-
"grad_norm": 7.213927268981934,
|
| 287 |
-
"learning_rate": 1.6122114610902677e-05,
|
| 288 |
-
"loss": 0.5628,
|
| 289 |
-
"step": 20000
|
| 290 |
-
},
|
| 291 |
-
{
|
| 292 |
-
"epoch": 0.5962248785737138,
|
| 293 |
-
"grad_norm": 19.586095809936523,
|
| 294 |
-
"learning_rate": 1.602516747617524e-05,
|
| 295 |
-
"loss": 0.5259,
|
| 296 |
-
"step": 20500
|
| 297 |
-
},
|
| 298 |
-
{
|
| 299 |
-
"epoch": 0.6107669487828288,
|
| 300 |
-
"grad_norm": 2.6470067501068115,
|
| 301 |
-
"learning_rate": 1.592822034144781e-05,
|
| 302 |
-
"loss": 0.5044,
|
| 303 |
-
"step": 21000
|
| 304 |
-
},
|
| 305 |
-
{
|
| 306 |
-
"epoch": 0.6253090189919437,
|
| 307 |
-
"grad_norm": 8.12119197845459,
|
| 308 |
-
"learning_rate": 1.5831273206720378e-05,
|
| 309 |
-
"loss": 0.4909,
|
| 310 |
-
"step": 21500
|
| 311 |
-
},
|
| 312 |
-
{
|
| 313 |
-
"epoch": 0.6398510892010587,
|
| 314 |
-
"grad_norm": 11.704862594604492,
|
| 315 |
-
"learning_rate": 1.5734326071992943e-05,
|
| 316 |
-
"loss": 0.5103,
|
| 317 |
-
"step": 22000
|
| 318 |
-
},
|
| 319 |
-
{
|
| 320 |
-
"epoch": 0.6543931594101736,
|
| 321 |
-
"grad_norm": 5.466031551361084,
|
| 322 |
-
"learning_rate": 1.563737893726551e-05,
|
| 323 |
-
"loss": 0.5097,
|
| 324 |
-
"step": 22500
|
| 325 |
-
},
|
| 326 |
-
{
|
| 327 |
-
"epoch": 0.6689352296192886,
|
| 328 |
-
"grad_norm": 1.2860121726989746,
|
| 329 |
-
"learning_rate": 1.5540431802538075e-05,
|
| 330 |
-
"loss": 0.4858,
|
| 331 |
-
"step": 23000
|
| 332 |
-
},
|
| 333 |
-
{
|
| 334 |
-
"epoch": 0.6834772998284036,
|
| 335 |
-
"grad_norm": 5.133608341217041,
|
| 336 |
-
"learning_rate": 1.5443484667810644e-05,
|
| 337 |
-
"loss": 0.4814,
|
| 338 |
-
"step": 23500
|
| 339 |
-
},
|
| 340 |
-
{
|
| 341 |
-
"epoch": 0.6980193700375186,
|
| 342 |
-
"grad_norm": 10.405769348144531,
|
| 343 |
-
"learning_rate": 1.5346537533083212e-05,
|
| 344 |
-
"loss": 0.4985,
|
| 345 |
-
"step": 24000
|
| 346 |
-
},
|
| 347 |
-
{
|
| 348 |
-
"epoch": 0.7125614402466335,
|
| 349 |
-
"grad_norm": 3.4990031719207764,
|
| 350 |
-
"learning_rate": 1.5249590398355778e-05,
|
| 351 |
-
"loss": 0.493,
|
| 352 |
-
"step": 24500
|
| 353 |
-
},
|
| 354 |
-
{
|
| 355 |
-
"epoch": 0.7271035104557485,
|
| 356 |
-
"grad_norm": 5.292512893676758,
|
| 357 |
-
"learning_rate": 1.5152643263628345e-05,
|
| 358 |
-
"loss": 0.4807,
|
| 359 |
-
"step": 25000
|
| 360 |
-
},
|
| 361 |
-
{
|
| 362 |
-
"epoch": 0.7416455806648634,
|
| 363 |
-
"grad_norm": 17.39272117614746,
|
| 364 |
-
"learning_rate": 1.5055696128900911e-05,
|
| 365 |
-
"loss": 0.4966,
|
| 366 |
-
"step": 25500
|
| 367 |
-
},
|
| 368 |
-
{
|
| 369 |
-
"epoch": 0.7561876508739784,
|
| 370 |
-
"grad_norm": 18.466636657714844,
|
| 371 |
-
"learning_rate": 1.4958748994173478e-05,
|
| 372 |
-
"loss": 0.4851,
|
| 373 |
-
"step": 26000
|
| 374 |
-
},
|
| 375 |
-
{
|
| 376 |
-
"epoch": 0.7707297210830933,
|
| 377 |
-
"grad_norm": 2.161870241165161,
|
| 378 |
-
"learning_rate": 1.4861801859446046e-05,
|
| 379 |
-
"loss": 0.4922,
|
| 380 |
-
"step": 26500
|
| 381 |
-
},
|
| 382 |
-
{
|
| 383 |
-
"epoch": 0.7852717912922084,
|
| 384 |
-
"grad_norm": 1.0825892686843872,
|
| 385 |
-
"learning_rate": 1.4764854724718612e-05,
|
| 386 |
-
"loss": 0.4987,
|
| 387 |
-
"step": 27000
|
| 388 |
-
},
|
| 389 |
-
{
|
| 390 |
-
"epoch": 0.7998138615013233,
|
| 391 |
-
"grad_norm": 11.1233491897583,
|
| 392 |
-
"learning_rate": 1.4667907589991179e-05,
|
| 393 |
-
"loss": 0.4728,
|
| 394 |
-
"step": 27500
|
| 395 |
-
},
|
| 396 |
-
{
|
| 397 |
-
"epoch": 0.8143559317104383,
|
| 398 |
-
"grad_norm": 8.120223045349121,
|
| 399 |
-
"learning_rate": 1.4570960455263745e-05,
|
| 400 |
-
"loss": 0.5043,
|
| 401 |
-
"step": 28000
|
| 402 |
-
},
|
| 403 |
-
{
|
| 404 |
-
"epoch": 0.8288980019195533,
|
| 405 |
-
"grad_norm": 6.8163933753967285,
|
| 406 |
-
"learning_rate": 1.4474013320536313e-05,
|
| 407 |
-
"loss": 0.5224,
|
| 408 |
-
"step": 28500
|
| 409 |
-
},
|
| 410 |
-
{
|
| 411 |
-
"epoch": 0.8434400721286682,
|
| 412 |
-
"grad_norm": 15.210949897766113,
|
| 413 |
-
"learning_rate": 1.437706618580888e-05,
|
| 414 |
-
"loss": 0.4532,
|
| 415 |
-
"step": 29000
|
| 416 |
-
},
|
| 417 |
-
{
|
| 418 |
-
"epoch": 0.8579821423377832,
|
| 419 |
-
"grad_norm": 3.8793272972106934,
|
| 420 |
-
"learning_rate": 1.4280119051081446e-05,
|
| 421 |
-
"loss": 0.4864,
|
| 422 |
-
"step": 29500
|
| 423 |
-
},
|
| 424 |
-
{
|
| 425 |
-
"epoch": 0.8725242125468982,
|
| 426 |
-
"grad_norm": 8.277094841003418,
|
| 427 |
-
"learning_rate": 1.4183171916354013e-05,
|
| 428 |
-
"loss": 0.5122,
|
| 429 |
-
"step": 30000
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 0.8870662827560132,
|
| 433 |
-
"grad_norm": 6.275518417358398,
|
| 434 |
-
"learning_rate": 1.408622478162658e-05,
|
| 435 |
-
"loss": 0.49,
|
| 436 |
-
"step": 30500
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 0.9016083529651281,
|
| 440 |
-
"grad_norm": 8.79964542388916,
|
| 441 |
-
"learning_rate": 1.3989277646899147e-05,
|
| 442 |
-
"loss": 0.4991,
|
| 443 |
-
"step": 31000
|
| 444 |
-
},
|
| 445 |
-
{
|
| 446 |
-
"epoch": 0.9161504231742431,
|
| 447 |
-
"grad_norm": 23.397424697875977,
|
| 448 |
-
"learning_rate": 1.3892330512171715e-05,
|
| 449 |
-
"loss": 0.4966,
|
| 450 |
-
"step": 31500
|
| 451 |
-
},
|
| 452 |
-
{
|
| 453 |
-
"epoch": 0.930692493383358,
|
| 454 |
-
"grad_norm": 9.809805870056152,
|
| 455 |
-
"learning_rate": 1.379538337744428e-05,
|
| 456 |
-
"loss": 0.5091,
|
| 457 |
-
"step": 32000
|
| 458 |
-
},
|
| 459 |
-
{
|
| 460 |
-
"epoch": 0.945234563592473,
|
| 461 |
-
"grad_norm": 1.723449468612671,
|
| 462 |
-
"learning_rate": 1.3698436242716848e-05,
|
| 463 |
-
"loss": 0.4963,
|
| 464 |
-
"step": 32500
|
| 465 |
-
},
|
| 466 |
-
{
|
| 467 |
-
"epoch": 0.959776633801588,
|
| 468 |
-
"grad_norm": 9.49936580657959,
|
| 469 |
-
"learning_rate": 1.3601489107989415e-05,
|
| 470 |
-
"loss": 0.4739,
|
| 471 |
-
"step": 33000
|
| 472 |
-
},
|
| 473 |
-
{
|
| 474 |
-
"epoch": 0.974318704010703,
|
| 475 |
-
"grad_norm": 5.465103626251221,
|
| 476 |
-
"learning_rate": 1.3504541973261983e-05,
|
| 477 |
-
"loss": 0.4947,
|
| 478 |
-
"step": 33500
|
| 479 |
-
},
|
| 480 |
-
{
|
| 481 |
-
"epoch": 0.9888607742198179,
|
| 482 |
-
"grad_norm": 5.015740871429443,
|
| 483 |
-
"learning_rate": 1.3407594838534548e-05,
|
| 484 |
-
"loss": 0.4692,
|
| 485 |
-
"step": 34000
|
| 486 |
-
},
|
| 487 |
-
{
|
| 488 |
-
"epoch": 1.003402844428933,
|
| 489 |
-
"grad_norm": 11.569725036621094,
|
| 490 |
-
"learning_rate": 1.3310647703807116e-05,
|
| 491 |
-
"loss": 0.4667,
|
| 492 |
-
"step": 34500
|
| 493 |
-
},
|
| 494 |
-
{
|
| 495 |
-
"epoch": 1.017944914638048,
|
| 496 |
-
"grad_norm": 6.356573581695557,
|
| 497 |
-
"learning_rate": 1.3213700569079682e-05,
|
| 498 |
-
"loss": 0.4682,
|
| 499 |
-
"step": 35000
|
| 500 |
-
},
|
| 501 |
-
{
|
| 502 |
-
"epoch": 1.032486984847163,
|
| 503 |
-
"grad_norm": 5.768576622009277,
|
| 504 |
-
"learning_rate": 1.311675343435225e-05,
|
| 505 |
-
"loss": 0.486,
|
| 506 |
-
"step": 35500
|
| 507 |
-
},
|
| 508 |
-
{
|
| 509 |
-
"epoch": 1.0470290550562777,
|
| 510 |
-
"grad_norm": 5.155892372131348,
|
| 511 |
-
"learning_rate": 1.3019806299624815e-05,
|
| 512 |
-
"loss": 0.4968,
|
| 513 |
-
"step": 36000
|
| 514 |
-
},
|
| 515 |
-
{
|
| 516 |
-
"epoch": 1.0615711252653928,
|
| 517 |
-
"grad_norm": 3.3541529178619385,
|
| 518 |
-
"learning_rate": 1.2922859164897383e-05,
|
| 519 |
-
"loss": 0.4944,
|
| 520 |
-
"step": 36500
|
| 521 |
-
},
|
| 522 |
-
{
|
| 523 |
-
"epoch": 1.0761131954745078,
|
| 524 |
-
"grad_norm": 8.195282936096191,
|
| 525 |
-
"learning_rate": 1.282591203016995e-05,
|
| 526 |
-
"loss": 0.4467,
|
| 527 |
-
"step": 37000
|
| 528 |
-
},
|
| 529 |
-
{
|
| 530 |
-
"epoch": 1.0906552656836228,
|
| 531 |
-
"grad_norm": 3.4823594093322754,
|
| 532 |
-
"learning_rate": 1.2728964895442518e-05,
|
| 533 |
-
"loss": 0.4544,
|
| 534 |
-
"step": 37500
|
| 535 |
-
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 1.1051973358927376,
|
| 538 |
-
"grad_norm": 2.6191506385803223,
|
| 539 |
-
"learning_rate": 1.2632017760715083e-05,
|
| 540 |
-
"loss": 0.4807,
|
| 541 |
-
"step": 38000
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 1.1197394061018526,
|
| 545 |
-
"grad_norm": 3.6867098808288574,
|
| 546 |
-
"learning_rate": 1.253507062598765e-05,
|
| 547 |
-
"loss": 0.5017,
|
| 548 |
-
"step": 38500
|
| 549 |
-
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 1.1342814763109677,
|
| 552 |
-
"grad_norm": 4.41229772567749,
|
| 553 |
-
"learning_rate": 1.2438123491260217e-05,
|
| 554 |
-
"loss": 0.4981,
|
| 555 |
-
"step": 39000
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 1.1488235465200827,
|
| 559 |
-
"grad_norm": 6.181690692901611,
|
| 560 |
-
"learning_rate": 1.2341176356532785e-05,
|
| 561 |
-
"loss": 0.4632,
|
| 562 |
-
"step": 39500
|
| 563 |
-
},
|
| 564 |
-
{
|
| 565 |
-
"epoch": 1.1633656167291975,
|
| 566 |
-
"grad_norm": 1.7811199426651,
|
| 567 |
-
"learning_rate": 1.224422922180535e-05,
|
| 568 |
-
"loss": 0.4935,
|
| 569 |
-
"step": 40000
|
| 570 |
-
},
|
| 571 |
-
{
|
| 572 |
-
"epoch": 1.1779076869383125,
|
| 573 |
-
"grad_norm": 7.902093410491943,
|
| 574 |
-
"learning_rate": 1.2147282087077918e-05,
|
| 575 |
-
"loss": 0.467,
|
| 576 |
-
"step": 40500
|
| 577 |
-
},
|
| 578 |
-
{
|
| 579 |
-
"epoch": 1.1924497571474275,
|
| 580 |
-
"grad_norm": 7.348107814788818,
|
| 581 |
-
"learning_rate": 1.2050334952350485e-05,
|
| 582 |
-
"loss": 0.4732,
|
| 583 |
-
"step": 41000
|
| 584 |
-
},
|
| 585 |
-
{
|
| 586 |
-
"epoch": 1.2069918273565425,
|
| 587 |
-
"grad_norm": 1.9056262969970703,
|
| 588 |
-
"learning_rate": 1.1953387817623053e-05,
|
| 589 |
-
"loss": 0.4689,
|
| 590 |
-
"step": 41500
|
| 591 |
-
},
|
| 592 |
-
{
|
| 593 |
-
"epoch": 1.2215338975656573,
|
| 594 |
-
"grad_norm": 3.8321168422698975,
|
| 595 |
-
"learning_rate": 1.1856440682895618e-05,
|
| 596 |
-
"loss": 0.4957,
|
| 597 |
-
"step": 42000
|
| 598 |
-
},
|
| 599 |
-
{
|
| 600 |
-
"epoch": 1.2360759677747724,
|
| 601 |
-
"grad_norm": 0.5265329480171204,
|
| 602 |
-
"learning_rate": 1.1759493548168184e-05,
|
| 603 |
-
"loss": 0.4738,
|
| 604 |
-
"step": 42500
|
| 605 |
-
},
|
| 606 |
-
{
|
| 607 |
-
"epoch": 1.2506180379838874,
|
| 608 |
-
"grad_norm": 7.372343063354492,
|
| 609 |
-
"learning_rate": 1.1662546413440752e-05,
|
| 610 |
-
"loss": 0.485,
|
| 611 |
-
"step": 43000
|
| 612 |
-
},
|
| 613 |
-
{
|
| 614 |
-
"epoch": 1.2651601081930024,
|
| 615 |
-
"grad_norm": 2.4991230964660645,
|
| 616 |
-
"learning_rate": 1.1565599278713317e-05,
|
| 617 |
-
"loss": 0.4497,
|
| 618 |
-
"step": 43500
|
| 619 |
-
},
|
| 620 |
-
{
|
| 621 |
-
"epoch": 1.2797021784021174,
|
| 622 |
-
"grad_norm": 11.783917427062988,
|
| 623 |
-
"learning_rate": 1.1468652143985885e-05,
|
| 624 |
-
"loss": 0.5148,
|
| 625 |
-
"step": 44000
|
| 626 |
-
},
|
| 627 |
-
{
|
| 628 |
-
"epoch": 1.2942442486112322,
|
| 629 |
-
"grad_norm": 12.756231307983398,
|
| 630 |
-
"learning_rate": 1.1371705009258452e-05,
|
| 631 |
-
"loss": 0.4676,
|
| 632 |
-
"step": 44500
|
| 633 |
-
},
|
| 634 |
-
{
|
| 635 |
-
"epoch": 1.3087863188203472,
|
| 636 |
-
"grad_norm": 4.692300319671631,
|
| 637 |
-
"learning_rate": 1.127475787453102e-05,
|
| 638 |
-
"loss": 0.4647,
|
| 639 |
-
"step": 45000
|
| 640 |
-
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 1.3233283890294623,
|
| 643 |
-
"grad_norm": 7.03782844543457,
|
| 644 |
-
"learning_rate": 1.1177810739803584e-05,
|
| 645 |
-
"loss": 0.4572,
|
| 646 |
-
"step": 45500
|
| 647 |
-
},
|
| 648 |
-
{
|
| 649 |
-
"epoch": 1.3378704592385773,
|
| 650 |
-
"grad_norm": 3.1515519618988037,
|
| 651 |
-
"learning_rate": 1.1080863605076153e-05,
|
| 652 |
-
"loss": 0.5162,
|
| 653 |
-
"step": 46000
|
| 654 |
-
},
|
| 655 |
-
{
|
| 656 |
-
"epoch": 1.352412529447692,
|
| 657 |
-
"grad_norm": 13.067205429077148,
|
| 658 |
-
"learning_rate": 1.0983916470348719e-05,
|
| 659 |
-
"loss": 0.5158,
|
| 660 |
-
"step": 46500
|
| 661 |
}
|
| 662 |
],
|
| 663 |
"logging_steps": 500,
|
|
@@ -677,7 +33,7 @@
|
|
| 677 |
"attributes": {}
|
| 678 |
}
|
| 679 |
},
|
| 680 |
-
"total_flos":
|
| 681 |
"train_batch_size": 16,
|
| 682 |
"trial_name": null,
|
| 683 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.014542070209114969,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 500,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.014542070209114969,
|
| 13 |
+
"grad_norm": 12.663599014282227,
|
| 14 |
"learning_rate": 1.9903052865272566e-05,
|
| 15 |
+
"loss": 7.4704,
|
| 16 |
"step": 500
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
}
|
| 18 |
],
|
| 19 |
"logging_steps": 500,
|
|
|
|
| 33 |
"attributes": {}
|
| 34 |
}
|
| 35 |
},
|
| 36 |
+
"total_flos": 268694085430272.0,
|
| 37 |
"train_batch_size": 16,
|
| 38 |
"trial_name": null,
|
| 39 |
"trial_params": null
|
last-checkpoint/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1cf779800eed062a2e03f99378369c4384c63e953f13f7419f13d9c53267c737
|
| 3 |
size 5176
|
last-checkpoint/vocab.txt
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|