Training checkpoint at step 1000
Browse files- trainer_state.json +394 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,394 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 1000,
|
| 3 |
+
"best_metric": 2.493894577026367,
|
| 4 |
+
"best_model_checkpoint": "../runs/karpathy/fineweb-edu-100b-shuffle/meta-llama/Llama-3.2-1B/linear_adamw_wd1e-03_7x1024_mem32_bs64_hf_armt_dmem64/run_30/checkpoint-1000",
|
| 5 |
+
"epoch": 0.02,
|
| 6 |
+
"eval_steps": 100,
|
| 7 |
+
"global_step": 1000,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0005,
|
| 14 |
+
"grad_norm": 1.374199618670654,
|
| 15 |
+
"learning_rate": 4.8e-08,
|
| 16 |
+
"loss": 3.8192,
|
| 17 |
+
"step": 25
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.001,
|
| 21 |
+
"grad_norm": 1.0626796430084513,
|
| 22 |
+
"learning_rate": 9.8e-08,
|
| 23 |
+
"loss": 3.7481,
|
| 24 |
+
"step": 50
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.0015,
|
| 28 |
+
"grad_norm": 0.554900729210382,
|
| 29 |
+
"learning_rate": 1.4800000000000003e-07,
|
| 30 |
+
"loss": 3.5249,
|
| 31 |
+
"step": 75
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.002,
|
| 35 |
+
"grad_norm": 0.407340177021375,
|
| 36 |
+
"learning_rate": 1.9800000000000003e-07,
|
| 37 |
+
"loss": 3.3424,
|
| 38 |
+
"step": 100
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.002,
|
| 42 |
+
"eval_loss": 3.072638988494873,
|
| 43 |
+
"eval_runtime": 33.0855,
|
| 44 |
+
"eval_samples_per_second": 3.536,
|
| 45 |
+
"eval_steps_per_second": 1.783,
|
| 46 |
+
"step": 100
|
| 47 |
+
},
|
| 48 |
+
{
|
| 49 |
+
"epoch": 0.0025,
|
| 50 |
+
"grad_norm": 0.22554288925827917,
|
| 51 |
+
"learning_rate": 2.48e-07,
|
| 52 |
+
"loss": 3.1284,
|
| 53 |
+
"step": 125
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 0.003,
|
| 57 |
+
"grad_norm": 0.14377524138618566,
|
| 58 |
+
"learning_rate": 2.9800000000000005e-07,
|
| 59 |
+
"loss": 2.9699,
|
| 60 |
+
"step": 150
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 0.0035,
|
| 64 |
+
"grad_norm": 0.1273100096023732,
|
| 65 |
+
"learning_rate": 3.48e-07,
|
| 66 |
+
"loss": 2.8784,
|
| 67 |
+
"step": 175
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 0.004,
|
| 71 |
+
"grad_norm": 0.08052145135761532,
|
| 72 |
+
"learning_rate": 3.9800000000000004e-07,
|
| 73 |
+
"loss": 2.808,
|
| 74 |
+
"step": 200
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.004,
|
| 78 |
+
"eval_loss": 2.7464711666107178,
|
| 79 |
+
"eval_runtime": 33.2528,
|
| 80 |
+
"eval_samples_per_second": 3.519,
|
| 81 |
+
"eval_steps_per_second": 1.774,
|
| 82 |
+
"step": 200
|
| 83 |
+
},
|
| 84 |
+
{
|
| 85 |
+
"epoch": 0.0045,
|
| 86 |
+
"grad_norm": 0.09601426553193865,
|
| 87 |
+
"learning_rate": 4.4800000000000004e-07,
|
| 88 |
+
"loss": 2.7735,
|
| 89 |
+
"step": 225
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.005,
|
| 93 |
+
"grad_norm": 0.06772677130883735,
|
| 94 |
+
"learning_rate": 4.98e-07,
|
| 95 |
+
"loss": 2.7358,
|
| 96 |
+
"step": 250
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.0055,
|
| 100 |
+
"grad_norm": 0.0617749171010752,
|
| 101 |
+
"learning_rate": 5.480000000000001e-07,
|
| 102 |
+
"loss": 2.7143,
|
| 103 |
+
"step": 275
|
| 104 |
+
},
|
| 105 |
+
{
|
| 106 |
+
"epoch": 0.006,
|
| 107 |
+
"grad_norm": 0.06081364438555446,
|
| 108 |
+
"learning_rate": 5.98e-07,
|
| 109 |
+
"loss": 2.695,
|
| 110 |
+
"step": 300
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.006,
|
| 114 |
+
"eval_loss": 2.658987045288086,
|
| 115 |
+
"eval_runtime": 33.098,
|
| 116 |
+
"eval_samples_per_second": 3.535,
|
| 117 |
+
"eval_steps_per_second": 1.783,
|
| 118 |
+
"step": 300
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.0065,
|
| 122 |
+
"grad_norm": 0.055229056926588416,
|
| 123 |
+
"learning_rate": 6.48e-07,
|
| 124 |
+
"loss": 2.6775,
|
| 125 |
+
"step": 325
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.007,
|
| 129 |
+
"grad_norm": 0.05285547880508365,
|
| 130 |
+
"learning_rate": 6.98e-07,
|
| 131 |
+
"loss": 2.6498,
|
| 132 |
+
"step": 350
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 0.0075,
|
| 136 |
+
"grad_norm": 0.05321418646538081,
|
| 137 |
+
"learning_rate": 7.480000000000001e-07,
|
| 138 |
+
"loss": 2.6375,
|
| 139 |
+
"step": 375
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 0.008,
|
| 143 |
+
"grad_norm": 0.046163922554101317,
|
| 144 |
+
"learning_rate": 7.98e-07,
|
| 145 |
+
"loss": 2.6273,
|
| 146 |
+
"step": 400
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.008,
|
| 150 |
+
"eval_loss": 2.5985612869262695,
|
| 151 |
+
"eval_runtime": 33.8223,
|
| 152 |
+
"eval_samples_per_second": 3.459,
|
| 153 |
+
"eval_steps_per_second": 1.744,
|
| 154 |
+
"step": 400
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 0.0085,
|
| 158 |
+
"grad_norm": 0.05184119325112733,
|
| 159 |
+
"learning_rate": 8.480000000000001e-07,
|
| 160 |
+
"loss": 2.6037,
|
| 161 |
+
"step": 425
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.009,
|
| 165 |
+
"grad_norm": 0.036013105129600216,
|
| 166 |
+
"learning_rate": 8.980000000000001e-07,
|
| 167 |
+
"loss": 2.594,
|
| 168 |
+
"step": 450
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 0.0095,
|
| 172 |
+
"grad_norm": 0.03061363860030697,
|
| 173 |
+
"learning_rate": 9.480000000000001e-07,
|
| 174 |
+
"loss": 2.5746,
|
| 175 |
+
"step": 475
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.01,
|
| 179 |
+
"grad_norm": 0.036970324693471876,
|
| 180 |
+
"learning_rate": 9.98e-07,
|
| 181 |
+
"loss": 2.5827,
|
| 182 |
+
"step": 500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.01,
|
| 186 |
+
"eval_loss": 2.56477952003479,
|
| 187 |
+
"eval_runtime": 33.0671,
|
| 188 |
+
"eval_samples_per_second": 3.538,
|
| 189 |
+
"eval_steps_per_second": 1.784,
|
| 190 |
+
"step": 500
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.0105,
|
| 194 |
+
"grad_norm": 0.03044033343054017,
|
| 195 |
+
"learning_rate": 1.0480000000000002e-06,
|
| 196 |
+
"loss": 2.5717,
|
| 197 |
+
"step": 525
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 0.011,
|
| 201 |
+
"grad_norm": 0.03340669717167394,
|
| 202 |
+
"learning_rate": 1.0980000000000001e-06,
|
| 203 |
+
"loss": 2.5613,
|
| 204 |
+
"step": 550
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.0115,
|
| 208 |
+
"grad_norm": 0.029909971322257495,
|
| 209 |
+
"learning_rate": 1.148e-06,
|
| 210 |
+
"loss": 2.5661,
|
| 211 |
+
"step": 575
|
| 212 |
+
},
|
| 213 |
+
{
|
| 214 |
+
"epoch": 0.012,
|
| 215 |
+
"grad_norm": 0.028901093383770705,
|
| 216 |
+
"learning_rate": 1.1980000000000002e-06,
|
| 217 |
+
"loss": 2.561,
|
| 218 |
+
"step": 600
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.012,
|
| 222 |
+
"eval_loss": 2.5446865558624268,
|
| 223 |
+
"eval_runtime": 33.1445,
|
| 224 |
+
"eval_samples_per_second": 3.53,
|
| 225 |
+
"eval_steps_per_second": 1.78,
|
| 226 |
+
"step": 600
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.0125,
|
| 230 |
+
"grad_norm": 0.0338582199400455,
|
| 231 |
+
"learning_rate": 1.248e-06,
|
| 232 |
+
"loss": 2.5392,
|
| 233 |
+
"step": 625
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.013,
|
| 237 |
+
"grad_norm": 0.031107046362937877,
|
| 238 |
+
"learning_rate": 1.2980000000000001e-06,
|
| 239 |
+
"loss": 2.5378,
|
| 240 |
+
"step": 650
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.0135,
|
| 244 |
+
"grad_norm": 0.027777474622611625,
|
| 245 |
+
"learning_rate": 1.348e-06,
|
| 246 |
+
"loss": 2.5421,
|
| 247 |
+
"step": 675
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.014,
|
| 251 |
+
"grad_norm": 0.032848272502404616,
|
| 252 |
+
"learning_rate": 1.3980000000000002e-06,
|
| 253 |
+
"loss": 2.5345,
|
| 254 |
+
"step": 700
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.014,
|
| 258 |
+
"eval_loss": 2.5294137001037598,
|
| 259 |
+
"eval_runtime": 33.222,
|
| 260 |
+
"eval_samples_per_second": 3.522,
|
| 261 |
+
"eval_steps_per_second": 1.776,
|
| 262 |
+
"step": 700
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.0145,
|
| 266 |
+
"grad_norm": 0.028260965292318807,
|
| 267 |
+
"learning_rate": 1.4480000000000002e-06,
|
| 268 |
+
"loss": 2.5342,
|
| 269 |
+
"step": 725
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.015,
|
| 273 |
+
"grad_norm": 0.02748431921263886,
|
| 274 |
+
"learning_rate": 1.498e-06,
|
| 275 |
+
"loss": 2.5188,
|
| 276 |
+
"step": 750
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.0155,
|
| 280 |
+
"grad_norm": 0.02859453618814513,
|
| 281 |
+
"learning_rate": 1.548e-06,
|
| 282 |
+
"loss": 2.5245,
|
| 283 |
+
"step": 775
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.016,
|
| 287 |
+
"grad_norm": 0.03906649589898274,
|
| 288 |
+
"learning_rate": 1.5980000000000002e-06,
|
| 289 |
+
"loss": 2.5142,
|
| 290 |
+
"step": 800
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.016,
|
| 294 |
+
"eval_loss": 2.5156726837158203,
|
| 295 |
+
"eval_runtime": 33.2465,
|
| 296 |
+
"eval_samples_per_second": 3.519,
|
| 297 |
+
"eval_steps_per_second": 1.775,
|
| 298 |
+
"step": 800
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 0.0165,
|
| 302 |
+
"grad_norm": 0.03055728445213633,
|
| 303 |
+
"learning_rate": 1.6480000000000001e-06,
|
| 304 |
+
"loss": 2.504,
|
| 305 |
+
"step": 825
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.017,
|
| 309 |
+
"grad_norm": 0.03831919004049627,
|
| 310 |
+
"learning_rate": 1.6980000000000003e-06,
|
| 311 |
+
"loss": 2.5096,
|
| 312 |
+
"step": 850
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 0.0175,
|
| 316 |
+
"grad_norm": 0.04714764947462498,
|
| 317 |
+
"learning_rate": 1.7480000000000002e-06,
|
| 318 |
+
"loss": 2.5057,
|
| 319 |
+
"step": 875
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 0.018,
|
| 323 |
+
"grad_norm": 0.04480333925801958,
|
| 324 |
+
"learning_rate": 1.798e-06,
|
| 325 |
+
"loss": 2.4949,
|
| 326 |
+
"step": 900
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.018,
|
| 330 |
+
"eval_loss": 2.503145456314087,
|
| 331 |
+
"eval_runtime": 33.3398,
|
| 332 |
+
"eval_samples_per_second": 3.509,
|
| 333 |
+
"eval_steps_per_second": 1.77,
|
| 334 |
+
"step": 900
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 0.0185,
|
| 338 |
+
"grad_norm": 0.04514734300904146,
|
| 339 |
+
"learning_rate": 1.8480000000000001e-06,
|
| 340 |
+
"loss": 2.5044,
|
| 341 |
+
"step": 925
|
| 342 |
+
},
|
| 343 |
+
{
|
| 344 |
+
"epoch": 0.019,
|
| 345 |
+
"grad_norm": 0.03664477032679196,
|
| 346 |
+
"learning_rate": 1.898e-06,
|
| 347 |
+
"loss": 2.4857,
|
| 348 |
+
"step": 950
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 0.0195,
|
| 352 |
+
"grad_norm": 0.03891788038244039,
|
| 353 |
+
"learning_rate": 1.9480000000000002e-06,
|
| 354 |
+
"loss": 2.4954,
|
| 355 |
+
"step": 975
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 0.02,
|
| 359 |
+
"grad_norm": 0.041701680819843504,
|
| 360 |
+
"learning_rate": 1.998e-06,
|
| 361 |
+
"loss": 2.4935,
|
| 362 |
+
"step": 1000
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.02,
|
| 366 |
+
"eval_loss": 2.493894577026367,
|
| 367 |
+
"eval_runtime": 33.4036,
|
| 368 |
+
"eval_samples_per_second": 3.503,
|
| 369 |
+
"eval_steps_per_second": 1.766,
|
| 370 |
+
"step": 1000
|
| 371 |
+
}
|
| 372 |
+
],
|
| 373 |
+
"logging_steps": 25,
|
| 374 |
+
"max_steps": 50000,
|
| 375 |
+
"num_input_tokens_seen": 0,
|
| 376 |
+
"num_train_epochs": 9223372036854775807,
|
| 377 |
+
"save_steps": 1000,
|
| 378 |
+
"stateful_callbacks": {
|
| 379 |
+
"TrainerControl": {
|
| 380 |
+
"args": {
|
| 381 |
+
"should_epoch_stop": false,
|
| 382 |
+
"should_evaluate": false,
|
| 383 |
+
"should_log": false,
|
| 384 |
+
"should_save": true,
|
| 385 |
+
"should_training_stop": false
|
| 386 |
+
},
|
| 387 |
+
"attributes": {}
|
| 388 |
+
}
|
| 389 |
+
},
|
| 390 |
+
"total_flos": 2.7853019251369574e+18,
|
| 391 |
+
"train_batch_size": 1,
|
| 392 |
+
"trial_name": null,
|
| 393 |
+
"trial_params": null
|
| 394 |
+
}
|