Commit
•
1fcf832
1
Parent(s):
799510f
Upload trainer_state.json
Browse files- trainer_state.json +371 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,371 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.7991475759190197,
|
| 5 |
+
"eval_steps": 20.0,
|
| 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": 0.015982951518380393,
|
| 13 |
+
"grad_norm": 584.8897461802409,
|
| 14 |
+
"learning_rate": 3.194888178913738e-07,
|
| 15 |
+
"loss": 2.787,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.031965903036760786,
|
| 20 |
+
"grad_norm": 36.67515312992242,
|
| 21 |
+
"learning_rate": 6.389776357827476e-07,
|
| 22 |
+
"loss": 2.5634,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.04794885455514118,
|
| 27 |
+
"grad_norm": 10.762606870071352,
|
| 28 |
+
"learning_rate": 9.584664536741215e-07,
|
| 29 |
+
"loss": 2.0774,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.06393180607352157,
|
| 34 |
+
"grad_norm": 8.222655227801516,
|
| 35 |
+
"learning_rate": 1.2779552715654952e-06,
|
| 36 |
+
"loss": 1.7751,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.07991475759190197,
|
| 41 |
+
"grad_norm": 9.439874512071311,
|
| 42 |
+
"learning_rate": 1.5974440894568691e-06,
|
| 43 |
+
"loss": 1.6774,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.09589770911028236,
|
| 48 |
+
"grad_norm": 7.064833162397198,
|
| 49 |
+
"learning_rate": 1.916932907348243e-06,
|
| 50 |
+
"loss": 1.5861,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.11188066062866275,
|
| 55 |
+
"grad_norm": 10.311171109232708,
|
| 56 |
+
"learning_rate": 2.2364217252396165e-06,
|
| 57 |
+
"loss": 1.5787,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.12786361214704314,
|
| 62 |
+
"grad_norm": 9.640566718774098,
|
| 63 |
+
"learning_rate": 2.5559105431309904e-06,
|
| 64 |
+
"loss": 1.5406,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.14384656366542356,
|
| 69 |
+
"grad_norm": 7.158156117158717,
|
| 70 |
+
"learning_rate": 2.8753993610223648e-06,
|
| 71 |
+
"loss": 1.5008,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.15982951518380395,
|
| 76 |
+
"grad_norm": 6.670618524707829,
|
| 77 |
+
"learning_rate": 3.1948881789137383e-06,
|
| 78 |
+
"loss": 1.4853,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.17581246670218434,
|
| 83 |
+
"grad_norm": 6.9280590132917075,
|
| 84 |
+
"learning_rate": 3.514376996805112e-06,
|
| 85 |
+
"loss": 1.4688,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.19179541822056473,
|
| 90 |
+
"grad_norm": 6.984532799464284,
|
| 91 |
+
"learning_rate": 3.833865814696486e-06,
|
| 92 |
+
"loss": 1.4501,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.20777836973894512,
|
| 97 |
+
"grad_norm": 4.6035696826309165,
|
| 98 |
+
"learning_rate": 4.15335463258786e-06,
|
| 99 |
+
"loss": 1.4325,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.2237613212573255,
|
| 104 |
+
"grad_norm": 7.6475074547231054,
|
| 105 |
+
"learning_rate": 4.472843450479233e-06,
|
| 106 |
+
"loss": 1.4058,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.23974427277570592,
|
| 111 |
+
"grad_norm": 5.215774162230619,
|
| 112 |
+
"learning_rate": 4.792332268370608e-06,
|
| 113 |
+
"loss": 1.3873,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.2557272242940863,
|
| 118 |
+
"grad_norm": 5.284228741443329,
|
| 119 |
+
"learning_rate": 5.111821086261981e-06,
|
| 120 |
+
"loss": 1.3882,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.2717101758124667,
|
| 125 |
+
"grad_norm": 4.315985648750639,
|
| 126 |
+
"learning_rate": 5.431309904153355e-06,
|
| 127 |
+
"loss": 1.3684,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.2876931273308471,
|
| 132 |
+
"grad_norm": 4.037073137808611,
|
| 133 |
+
"learning_rate": 5.7507987220447296e-06,
|
| 134 |
+
"loss": 1.359,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.3036760788492275,
|
| 139 |
+
"grad_norm": 6.346300332686573,
|
| 140 |
+
"learning_rate": 6.070287539936103e-06,
|
| 141 |
+
"loss": 1.3435,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.3196590303676079,
|
| 146 |
+
"grad_norm": 4.63092372720661,
|
| 147 |
+
"learning_rate": 6.3897763578274765e-06,
|
| 148 |
+
"loss": 1.3468,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.33564198188598826,
|
| 153 |
+
"grad_norm": 5.057903719523363,
|
| 154 |
+
"learning_rate": 6.709265175718851e-06,
|
| 155 |
+
"loss": 1.3341,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.3516249334043687,
|
| 160 |
+
"grad_norm": 5.25790384573804,
|
| 161 |
+
"learning_rate": 7.028753993610224e-06,
|
| 162 |
+
"loss": 1.3233,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.3676078849227491,
|
| 167 |
+
"grad_norm": 3.3267127292309517,
|
| 168 |
+
"learning_rate": 7.348242811501598e-06,
|
| 169 |
+
"loss": 1.3025,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.38359083644112946,
|
| 174 |
+
"grad_norm": 4.174952594356688,
|
| 175 |
+
"learning_rate": 7.667731629392972e-06,
|
| 176 |
+
"loss": 1.3,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.3995737879595099,
|
| 181 |
+
"grad_norm": 2.7547093096631166,
|
| 182 |
+
"learning_rate": 7.987220447284347e-06,
|
| 183 |
+
"loss": 1.299,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.41555673947789024,
|
| 188 |
+
"grad_norm": 3.0076000117279023,
|
| 189 |
+
"learning_rate": 8.30670926517572e-06,
|
| 190 |
+
"loss": 1.2789,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.43153969099627065,
|
| 195 |
+
"grad_norm": 3.6118960867167695,
|
| 196 |
+
"learning_rate": 8.626198083067093e-06,
|
| 197 |
+
"loss": 1.289,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.447522642514651,
|
| 202 |
+
"grad_norm": 4.187154082722337,
|
| 203 |
+
"learning_rate": 8.945686900958466e-06,
|
| 204 |
+
"loss": 1.2772,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.46350559403303143,
|
| 209 |
+
"grad_norm": 3.3209936896861265,
|
| 210 |
+
"learning_rate": 9.265175718849841e-06,
|
| 211 |
+
"loss": 1.2714,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.47948854555141185,
|
| 216 |
+
"grad_norm": 3.4443001562081874,
|
| 217 |
+
"learning_rate": 9.584664536741216e-06,
|
| 218 |
+
"loss": 1.2615,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.4954714970697922,
|
| 223 |
+
"grad_norm": 3.4861656497713764,
|
| 224 |
+
"learning_rate": 9.904153354632589e-06,
|
| 225 |
+
"loss": 1.2756,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.5114544485881726,
|
| 230 |
+
"grad_norm": 3.2211114109625165,
|
| 231 |
+
"learning_rate": 9.999847101583393e-06,
|
| 232 |
+
"loss": 1.2434,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.527437400106553,
|
| 237 |
+
"grad_norm": 3.1006035508008485,
|
| 238 |
+
"learning_rate": 9.999098233890869e-06,
|
| 239 |
+
"loss": 1.2489,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.5434203516249334,
|
| 244 |
+
"grad_norm": 3.0983188498194516,
|
| 245 |
+
"learning_rate": 9.997725406892392e-06,
|
| 246 |
+
"loss": 1.2407,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.5594033031433138,
|
| 251 |
+
"grad_norm": 3.350092298430153,
|
| 252 |
+
"learning_rate": 9.995728791936505e-06,
|
| 253 |
+
"loss": 1.243,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.5753862546616942,
|
| 258 |
+
"grad_norm": 2.712885657555514,
|
| 259 |
+
"learning_rate": 9.993108638229449e-06,
|
| 260 |
+
"loss": 1.2677,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.5913692061800746,
|
| 265 |
+
"grad_norm": 2.6927845803955983,
|
| 266 |
+
"learning_rate": 9.989865272804064e-06,
|
| 267 |
+
"loss": 1.2272,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.607352157698455,
|
| 272 |
+
"grad_norm": 2.9181755886200675,
|
| 273 |
+
"learning_rate": 9.985999100478964e-06,
|
| 274 |
+
"loss": 1.2217,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.6233351092168353,
|
| 279 |
+
"grad_norm": 2.663582794855719,
|
| 280 |
+
"learning_rate": 9.981510603808024e-06,
|
| 281 |
+
"loss": 1.2225,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.6393180607352158,
|
| 286 |
+
"grad_norm": 2.81821319928569,
|
| 287 |
+
"learning_rate": 9.976400343020134e-06,
|
| 288 |
+
"loss": 1.2291,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.6553010122535962,
|
| 293 |
+
"grad_norm": 2.7190463266479363,
|
| 294 |
+
"learning_rate": 9.970668955949285e-06,
|
| 295 |
+
"loss": 1.2017,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.6712839637719765,
|
| 300 |
+
"grad_norm": 2.6121849551347056,
|
| 301 |
+
"learning_rate": 9.964317157954955e-06,
|
| 302 |
+
"loss": 1.1999,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.687266915290357,
|
| 307 |
+
"grad_norm": 2.8153860907477988,
|
| 308 |
+
"learning_rate": 9.95734574183282e-06,
|
| 309 |
+
"loss": 1.2119,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.7032498668087374,
|
| 314 |
+
"grad_norm": 2.8604210593081016,
|
| 315 |
+
"learning_rate": 9.949755577715806e-06,
|
| 316 |
+
"loss": 1.1973,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.7192328183271177,
|
| 321 |
+
"grad_norm": 2.683531260179304,
|
| 322 |
+
"learning_rate": 9.941547612965475e-06,
|
| 323 |
+
"loss": 1.2138,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.7352157698454982,
|
| 328 |
+
"grad_norm": 3.0411589457571377,
|
| 329 |
+
"learning_rate": 9.932722872053797e-06,
|
| 330 |
+
"loss": 1.2221,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.7511987213638786,
|
| 335 |
+
"grad_norm": 2.993913433024178,
|
| 336 |
+
"learning_rate": 9.923282456435262e-06,
|
| 337 |
+
"loss": 1.209,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.7671816728822589,
|
| 342 |
+
"grad_norm": 2.6707946625116072,
|
| 343 |
+
"learning_rate": 9.913227544409416e-06,
|
| 344 |
+
"loss": 1.1918,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.7831646244006393,
|
| 349 |
+
"grad_norm": 2.7683148756228033,
|
| 350 |
+
"learning_rate": 9.90255939097379e-06,
|
| 351 |
+
"loss": 1.1784,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.7991475759190197,
|
| 356 |
+
"grad_norm": 2.803374407208519,
|
| 357 |
+
"learning_rate": 9.891279327667252e-06,
|
| 358 |
+
"loss": 1.1769,
|
| 359 |
+
"step": 500
|
| 360 |
+
}
|
| 361 |
+
],
|
| 362 |
+
"logging_steps": 10,
|
| 363 |
+
"max_steps": 3125,
|
| 364 |
+
"num_input_tokens_seen": 0,
|
| 365 |
+
"num_train_epochs": 5,
|
| 366 |
+
"save_steps": 100,
|
| 367 |
+
"total_flos": 7.826917846385951e+18,
|
| 368 |
+
"train_batch_size": 2,
|
| 369 |
+
"trial_name": null,
|
| 370 |
+
"trial_params": null
|
| 371 |
+
}
|