Upload 13 files
Browse files- optimizer.pt +3 -0
- scheduler.pt +3 -0
- trainer_state.json +352 -0
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8f00c447f1af4db5a1d601b1e404555837a49338bbd2abcb75a62b8961b0e398
|
| 3 |
+
size 655344209
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8be544c7e85bb8fbfe3e36bd703b7d2fa8697e70210232aadbe33cf7f9d4a1ce
|
| 3 |
+
size 623
|
trainer_state.json
ADDED
|
@@ -0,0 +1,352 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.9577464788732395,
|
| 5 |
+
"global_step": 420,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.07,
|
| 12 |
+
"eval_loss": 3.769569158554077,
|
| 13 |
+
"eval_runtime": 27.5829,
|
| 14 |
+
"eval_samples_per_second": 80.34,
|
| 15 |
+
"eval_steps_per_second": 1.269,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.14,
|
| 20 |
+
"eval_loss": 3.7680795192718506,
|
| 21 |
+
"eval_runtime": 27.3249,
|
| 22 |
+
"eval_samples_per_second": 81.098,
|
| 23 |
+
"eval_steps_per_second": 1.281,
|
| 24 |
+
"step": 20
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.21,
|
| 28 |
+
"eval_loss": 3.7672388553619385,
|
| 29 |
+
"eval_runtime": 27.4916,
|
| 30 |
+
"eval_samples_per_second": 80.606,
|
| 31 |
+
"eval_steps_per_second": 1.273,
|
| 32 |
+
"step": 30
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.28,
|
| 36 |
+
"eval_loss": 3.765839099884033,
|
| 37 |
+
"eval_runtime": 27.4005,
|
| 38 |
+
"eval_samples_per_second": 80.874,
|
| 39 |
+
"eval_steps_per_second": 1.277,
|
| 40 |
+
"step": 40
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 0.35,
|
| 44 |
+
"eval_loss": 3.763303518295288,
|
| 45 |
+
"eval_runtime": 27.5296,
|
| 46 |
+
"eval_samples_per_second": 80.495,
|
| 47 |
+
"eval_steps_per_second": 1.271,
|
| 48 |
+
"step": 50
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.42,
|
| 52 |
+
"eval_loss": 3.7637836933135986,
|
| 53 |
+
"eval_runtime": 27.328,
|
| 54 |
+
"eval_samples_per_second": 81.089,
|
| 55 |
+
"eval_steps_per_second": 1.281,
|
| 56 |
+
"step": 60
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.49,
|
| 60 |
+
"eval_loss": 3.760821580886841,
|
| 61 |
+
"eval_runtime": 27.5647,
|
| 62 |
+
"eval_samples_per_second": 80.393,
|
| 63 |
+
"eval_steps_per_second": 1.27,
|
| 64 |
+
"step": 70
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.56,
|
| 68 |
+
"eval_loss": 3.7594783306121826,
|
| 69 |
+
"eval_runtime": 27.3075,
|
| 70 |
+
"eval_samples_per_second": 81.15,
|
| 71 |
+
"eval_steps_per_second": 1.282,
|
| 72 |
+
"step": 80
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.63,
|
| 76 |
+
"eval_loss": 3.7594757080078125,
|
| 77 |
+
"eval_runtime": 27.4379,
|
| 78 |
+
"eval_samples_per_second": 80.764,
|
| 79 |
+
"eval_steps_per_second": 1.276,
|
| 80 |
+
"step": 90
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.7,
|
| 84 |
+
"eval_loss": 3.7583823204040527,
|
| 85 |
+
"eval_runtime": 27.5694,
|
| 86 |
+
"eval_samples_per_second": 80.379,
|
| 87 |
+
"eval_steps_per_second": 1.27,
|
| 88 |
+
"step": 100
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 0.77,
|
| 92 |
+
"eval_loss": 3.7562084197998047,
|
| 93 |
+
"eval_runtime": 27.3431,
|
| 94 |
+
"eval_samples_per_second": 81.044,
|
| 95 |
+
"eval_steps_per_second": 1.28,
|
| 96 |
+
"step": 110
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.85,
|
| 100 |
+
"eval_loss": 3.755629301071167,
|
| 101 |
+
"eval_runtime": 27.5374,
|
| 102 |
+
"eval_samples_per_second": 80.472,
|
| 103 |
+
"eval_steps_per_second": 1.271,
|
| 104 |
+
"step": 120
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.92,
|
| 108 |
+
"eval_loss": 3.7544658184051514,
|
| 109 |
+
"eval_runtime": 27.3291,
|
| 110 |
+
"eval_samples_per_second": 81.086,
|
| 111 |
+
"eval_steps_per_second": 1.281,
|
| 112 |
+
"step": 130
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.99,
|
| 116 |
+
"eval_loss": 3.7536823749542236,
|
| 117 |
+
"eval_runtime": 27.3625,
|
| 118 |
+
"eval_samples_per_second": 80.987,
|
| 119 |
+
"eval_steps_per_second": 1.279,
|
| 120 |
+
"step": 140
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"epoch": 1.06,
|
| 124 |
+
"eval_loss": 3.7540736198425293,
|
| 125 |
+
"eval_runtime": 27.6591,
|
| 126 |
+
"eval_samples_per_second": 80.118,
|
| 127 |
+
"eval_steps_per_second": 1.265,
|
| 128 |
+
"step": 150
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.13,
|
| 132 |
+
"eval_loss": 3.7542569637298584,
|
| 133 |
+
"eval_runtime": 27.4748,
|
| 134 |
+
"eval_samples_per_second": 80.656,
|
| 135 |
+
"eval_steps_per_second": 1.274,
|
| 136 |
+
"step": 160
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.2,
|
| 140 |
+
"eval_loss": 3.752318859100342,
|
| 141 |
+
"eval_runtime": 27.6148,
|
| 142 |
+
"eval_samples_per_second": 80.247,
|
| 143 |
+
"eval_steps_per_second": 1.267,
|
| 144 |
+
"step": 170
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 1.27,
|
| 148 |
+
"eval_loss": 3.7517406940460205,
|
| 149 |
+
"eval_runtime": 27.3496,
|
| 150 |
+
"eval_samples_per_second": 81.025,
|
| 151 |
+
"eval_steps_per_second": 1.28,
|
| 152 |
+
"step": 180
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 1.34,
|
| 156 |
+
"eval_loss": 3.751498222351074,
|
| 157 |
+
"eval_runtime": 27.5367,
|
| 158 |
+
"eval_samples_per_second": 80.474,
|
| 159 |
+
"eval_steps_per_second": 1.271,
|
| 160 |
+
"step": 190
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 1.41,
|
| 164 |
+
"eval_loss": 3.7509350776672363,
|
| 165 |
+
"eval_runtime": 27.354,
|
| 166 |
+
"eval_samples_per_second": 81.012,
|
| 167 |
+
"eval_steps_per_second": 1.28,
|
| 168 |
+
"step": 200
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 1.48,
|
| 172 |
+
"eval_loss": 3.750244140625,
|
| 173 |
+
"eval_runtime": 27.3514,
|
| 174 |
+
"eval_samples_per_second": 81.02,
|
| 175 |
+
"eval_steps_per_second": 1.28,
|
| 176 |
+
"step": 210
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 1.55,
|
| 180 |
+
"eval_loss": 3.750495433807373,
|
| 181 |
+
"eval_runtime": 27.5933,
|
| 182 |
+
"eval_samples_per_second": 80.309,
|
| 183 |
+
"eval_steps_per_second": 1.268,
|
| 184 |
+
"step": 220
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 1.62,
|
| 188 |
+
"eval_loss": 3.7491180896759033,
|
| 189 |
+
"eval_runtime": 27.3222,
|
| 190 |
+
"eval_samples_per_second": 81.106,
|
| 191 |
+
"eval_steps_per_second": 1.281,
|
| 192 |
+
"step": 230
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 1.69,
|
| 196 |
+
"eval_loss": 3.747929096221924,
|
| 197 |
+
"eval_runtime": 27.6237,
|
| 198 |
+
"eval_samples_per_second": 80.221,
|
| 199 |
+
"eval_steps_per_second": 1.267,
|
| 200 |
+
"step": 240
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 1.76,
|
| 204 |
+
"eval_loss": 3.749088764190674,
|
| 205 |
+
"eval_runtime": 27.2827,
|
| 206 |
+
"eval_samples_per_second": 81.224,
|
| 207 |
+
"eval_steps_per_second": 1.283,
|
| 208 |
+
"step": 250
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 1.83,
|
| 212 |
+
"eval_loss": 3.748143434524536,
|
| 213 |
+
"eval_runtime": 27.5685,
|
| 214 |
+
"eval_samples_per_second": 80.381,
|
| 215 |
+
"eval_steps_per_second": 1.27,
|
| 216 |
+
"step": 260
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 1.9,
|
| 220 |
+
"eval_loss": 3.747373580932617,
|
| 221 |
+
"eval_runtime": 27.3404,
|
| 222 |
+
"eval_samples_per_second": 81.052,
|
| 223 |
+
"eval_steps_per_second": 1.28,
|
| 224 |
+
"step": 270
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.97,
|
| 228 |
+
"eval_loss": 3.7469279766082764,
|
| 229 |
+
"eval_runtime": 27.5885,
|
| 230 |
+
"eval_samples_per_second": 80.323,
|
| 231 |
+
"eval_steps_per_second": 1.269,
|
| 232 |
+
"step": 280
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 2.04,
|
| 236 |
+
"eval_loss": 3.7472589015960693,
|
| 237 |
+
"eval_runtime": 27.3166,
|
| 238 |
+
"eval_samples_per_second": 81.123,
|
| 239 |
+
"eval_steps_per_second": 1.281,
|
| 240 |
+
"step": 290
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 2.11,
|
| 244 |
+
"eval_loss": 3.7470862865448,
|
| 245 |
+
"eval_runtime": 27.2994,
|
| 246 |
+
"eval_samples_per_second": 81.174,
|
| 247 |
+
"eval_steps_per_second": 1.282,
|
| 248 |
+
"step": 300
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 2.18,
|
| 252 |
+
"eval_loss": 3.746790885925293,
|
| 253 |
+
"eval_runtime": 27.3969,
|
| 254 |
+
"eval_samples_per_second": 80.885,
|
| 255 |
+
"eval_steps_per_second": 1.278,
|
| 256 |
+
"step": 310
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 2.25,
|
| 260 |
+
"eval_loss": 3.747112989425659,
|
| 261 |
+
"eval_runtime": 27.3496,
|
| 262 |
+
"eval_samples_per_second": 81.025,
|
| 263 |
+
"eval_steps_per_second": 1.28,
|
| 264 |
+
"step": 320
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 2.32,
|
| 268 |
+
"eval_loss": 3.747150421142578,
|
| 269 |
+
"eval_runtime": 27.5712,
|
| 270 |
+
"eval_samples_per_second": 80.374,
|
| 271 |
+
"eval_steps_per_second": 1.269,
|
| 272 |
+
"step": 330
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 2.39,
|
| 276 |
+
"eval_loss": 3.7463109493255615,
|
| 277 |
+
"eval_runtime": 27.3541,
|
| 278 |
+
"eval_samples_per_second": 81.012,
|
| 279 |
+
"eval_steps_per_second": 1.28,
|
| 280 |
+
"step": 340
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 2.46,
|
| 284 |
+
"eval_loss": 3.745861530303955,
|
| 285 |
+
"eval_runtime": 27.5761,
|
| 286 |
+
"eval_samples_per_second": 80.36,
|
| 287 |
+
"eval_steps_per_second": 1.269,
|
| 288 |
+
"step": 350
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 2.54,
|
| 292 |
+
"eval_loss": 3.7457809448242188,
|
| 293 |
+
"eval_runtime": 27.4138,
|
| 294 |
+
"eval_samples_per_second": 80.835,
|
| 295 |
+
"eval_steps_per_second": 1.277,
|
| 296 |
+
"step": 360
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 2.61,
|
| 300 |
+
"eval_loss": 3.745532751083374,
|
| 301 |
+
"eval_runtime": 27.5336,
|
| 302 |
+
"eval_samples_per_second": 80.483,
|
| 303 |
+
"eval_steps_per_second": 1.271,
|
| 304 |
+
"step": 370
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 2.68,
|
| 308 |
+
"eval_loss": 3.7454142570495605,
|
| 309 |
+
"eval_runtime": 27.3417,
|
| 310 |
+
"eval_samples_per_second": 81.048,
|
| 311 |
+
"eval_steps_per_second": 1.28,
|
| 312 |
+
"step": 380
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 2.75,
|
| 316 |
+
"eval_loss": 3.7453861236572266,
|
| 317 |
+
"eval_runtime": 27.3314,
|
| 318 |
+
"eval_samples_per_second": 81.079,
|
| 319 |
+
"eval_steps_per_second": 1.281,
|
| 320 |
+
"step": 390
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 2.82,
|
| 324 |
+
"eval_loss": 3.7453415393829346,
|
| 325 |
+
"eval_runtime": 27.5694,
|
| 326 |
+
"eval_samples_per_second": 80.379,
|
| 327 |
+
"eval_steps_per_second": 1.27,
|
| 328 |
+
"step": 400
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 2.89,
|
| 332 |
+
"eval_loss": 3.745169162750244,
|
| 333 |
+
"eval_runtime": 27.3841,
|
| 334 |
+
"eval_samples_per_second": 80.923,
|
| 335 |
+
"eval_steps_per_second": 1.278,
|
| 336 |
+
"step": 410
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 2.96,
|
| 340 |
+
"eval_loss": 3.745060682296753,
|
| 341 |
+
"eval_runtime": 27.5883,
|
| 342 |
+
"eval_samples_per_second": 80.324,
|
| 343 |
+
"eval_steps_per_second": 1.269,
|
| 344 |
+
"step": 420
|
| 345 |
+
}
|
| 346 |
+
],
|
| 347 |
+
"max_steps": 426,
|
| 348 |
+
"num_train_epochs": 3,
|
| 349 |
+
"total_flos": 875997356359680.0,
|
| 350 |
+
"trial_name": null,
|
| 351 |
+
"trial_params": null
|
| 352 |
+
}
|