Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +630 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,630 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9997049277072882,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 847,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.011802891708468575,
|
| 13 |
+
"grad_norm": 2.2601828575134277,
|
| 14 |
+
"learning_rate": 5.294117647058824e-06,
|
| 15 |
+
"loss": 3.4593,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.02360578341693715,
|
| 20 |
+
"grad_norm": 2.753499746322632,
|
| 21 |
+
"learning_rate": 1e-05,
|
| 22 |
+
"loss": 2.9512,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.03540867512540572,
|
| 27 |
+
"grad_norm": 1.42862069606781,
|
| 28 |
+
"learning_rate": 1.588235294117647e-05,
|
| 29 |
+
"loss": 2.2602,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.0472115668338743,
|
| 34 |
+
"grad_norm": 2.2324745655059814,
|
| 35 |
+
"learning_rate": 2.1764705882352943e-05,
|
| 36 |
+
"loss": 1.9542,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.05901445854234287,
|
| 41 |
+
"grad_norm": 0.9986198544502258,
|
| 42 |
+
"learning_rate": 2.7647058823529416e-05,
|
| 43 |
+
"loss": 1.7491,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.07081735025081144,
|
| 48 |
+
"grad_norm": 1.7015349864959717,
|
| 49 |
+
"learning_rate": 3.352941176470588e-05,
|
| 50 |
+
"loss": 1.7127,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.08262024195928003,
|
| 55 |
+
"grad_norm": 1.410741925239563,
|
| 56 |
+
"learning_rate": 3.9411764705882356e-05,
|
| 57 |
+
"loss": 1.5587,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.0944231336677486,
|
| 62 |
+
"grad_norm": 1.4795621633529663,
|
| 63 |
+
"learning_rate": 4.5294117647058826e-05,
|
| 64 |
+
"loss": 1.6199,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.10622602537621717,
|
| 69 |
+
"grad_norm": 1.2762919664382935,
|
| 70 |
+
"learning_rate": 4.999915012051437e-05,
|
| 71 |
+
"loss": 1.5481,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.11802891708468574,
|
| 76 |
+
"grad_norm": 1.1632708311080933,
|
| 77 |
+
"learning_rate": 4.996941040535653e-05,
|
| 78 |
+
"loss": 1.4648,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.1298318087931543,
|
| 83 |
+
"grad_norm": 1.950888752937317,
|
| 84 |
+
"learning_rate": 4.989723448187131e-05,
|
| 85 |
+
"loss": 1.5273,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.14163470050162288,
|
| 90 |
+
"grad_norm": 1.0861647129058838,
|
| 91 |
+
"learning_rate": 4.978274501505061e-05,
|
| 92 |
+
"loss": 1.5472,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.15343759221009148,
|
| 97 |
+
"grad_norm": 1.347829818725586,
|
| 98 |
+
"learning_rate": 4.962613658293158e-05,
|
| 99 |
+
"loss": 1.5428,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.16524048391856005,
|
| 104 |
+
"grad_norm": 1.014880657196045,
|
| 105 |
+
"learning_rate": 4.942767534590581e-05,
|
| 106 |
+
"loss": 1.3644,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.17704337562702863,
|
| 111 |
+
"grad_norm": 1.7476609945297241,
|
| 112 |
+
"learning_rate": 4.918769859437232e-05,
|
| 113 |
+
"loss": 1.4653,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.1888462673354972,
|
| 118 |
+
"grad_norm": 1.0597649812698364,
|
| 119 |
+
"learning_rate": 4.890661417550319e-05,
|
| 120 |
+
"loss": 1.5003,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.20064915904396577,
|
| 125 |
+
"grad_norm": 1.098489761352539,
|
| 126 |
+
"learning_rate": 4.8584899800095864e-05,
|
| 127 |
+
"loss": 1.462,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.21245205075243434,
|
| 132 |
+
"grad_norm": 0.8845277428627014,
|
| 133 |
+
"learning_rate": 4.822310223069039e-05,
|
| 134 |
+
"loss": 1.3805,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.2242549424609029,
|
| 139 |
+
"grad_norm": 1.181378960609436,
|
| 140 |
+
"learning_rate": 4.782183635233124e-05,
|
| 141 |
+
"loss": 1.4678,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.23605783416937148,
|
| 146 |
+
"grad_norm": 1.3490713834762573,
|
| 147 |
+
"learning_rate": 4.738178412755306e-05,
|
| 148 |
+
"loss": 1.378,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.24786072587784008,
|
| 153 |
+
"grad_norm": 1.1515178680419922,
|
| 154 |
+
"learning_rate": 4.690369343736636e-05,
|
| 155 |
+
"loss": 1.4216,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.2596636175863086,
|
| 160 |
+
"grad_norm": 1.3116216659545898,
|
| 161 |
+
"learning_rate": 4.6388376810212905e-05,
|
| 162 |
+
"loss": 1.3915,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.2714665092947772,
|
| 167 |
+
"grad_norm": 1.6775559186935425,
|
| 168 |
+
"learning_rate": 4.583671004105096e-05,
|
| 169 |
+
"loss": 1.4104,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.28326940100324577,
|
| 174 |
+
"grad_norm": 1.4933921098709106,
|
| 175 |
+
"learning_rate": 4.524963070291744e-05,
|
| 176 |
+
"loss": 1.411,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.29507229271171437,
|
| 181 |
+
"grad_norm": 1.057346224784851,
|
| 182 |
+
"learning_rate": 4.4628136553496375e-05,
|
| 183 |
+
"loss": 1.3628,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.30687518442018297,
|
| 188 |
+
"grad_norm": 1.389323353767395,
|
| 189 |
+
"learning_rate": 4.397328383940196e-05,
|
| 190 |
+
"loss": 1.331,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.3186780761286515,
|
| 195 |
+
"grad_norm": 1.2937726974487305,
|
| 196 |
+
"learning_rate": 4.328618550105802e-05,
|
| 197 |
+
"loss": 1.338,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.3304809678371201,
|
| 202 |
+
"grad_norm": 1.0137064456939697,
|
| 203 |
+
"learning_rate": 4.256800928122475e-05,
|
| 204 |
+
"loss": 1.4131,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.34228385954558865,
|
| 209 |
+
"grad_norm": 1.2065379619598389,
|
| 210 |
+
"learning_rate": 4.181997574038741e-05,
|
| 211 |
+
"loss": 1.3584,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.35408675125405725,
|
| 216 |
+
"grad_norm": 1.0482007265090942,
|
| 217 |
+
"learning_rate": 4.104335618237972e-05,
|
| 218 |
+
"loss": 1.3541,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.3658896429625258,
|
| 223 |
+
"grad_norm": 1.1176925897598267,
|
| 224 |
+
"learning_rate": 4.0239470493767704e-05,
|
| 225 |
+
"loss": 1.359,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.3776925346709944,
|
| 230 |
+
"grad_norm": 0.9922409653663635,
|
| 231 |
+
"learning_rate": 3.940968490066559e-05,
|
| 232 |
+
"loss": 1.261,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.389495426379463,
|
| 237 |
+
"grad_norm": 1.4820419549942017,
|
| 238 |
+
"learning_rate": 3.855540964679658e-05,
|
| 239 |
+
"loss": 1.2903,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.40129831808793154,
|
| 244 |
+
"grad_norm": 1.443935751914978,
|
| 245 |
+
"learning_rate": 3.767809659674433e-05,
|
| 246 |
+
"loss": 1.3593,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.41310120979640014,
|
| 251 |
+
"grad_norm": 2.1749682426452637,
|
| 252 |
+
"learning_rate": 3.677923676846864e-05,
|
| 253 |
+
"loss": 1.3608,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.4249041015048687,
|
| 258 |
+
"grad_norm": 1.4614121913909912,
|
| 259 |
+
"learning_rate": 3.586035779927896e-05,
|
| 260 |
+
"loss": 1.2742,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.4367069932133373,
|
| 265 |
+
"grad_norm": 1.1197657585144043,
|
| 266 |
+
"learning_rate": 3.492302134957218e-05,
|
| 267 |
+
"loss": 1.3217,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.4485098849218058,
|
| 272 |
+
"grad_norm": 1.1603928804397583,
|
| 273 |
+
"learning_rate": 3.396882044874736e-05,
|
| 274 |
+
"loss": 1.2824,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.4603127766302744,
|
| 279 |
+
"grad_norm": 1.6083821058273315,
|
| 280 |
+
"learning_rate": 3.2999376787807864e-05,
|
| 281 |
+
"loss": 1.344,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.47211566833874297,
|
| 286 |
+
"grad_norm": 1.56455397605896,
|
| 287 |
+
"learning_rate": 3.201633796325233e-05,
|
| 288 |
+
"loss": 1.3372,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.48391856004721157,
|
| 293 |
+
"grad_norm": 1.4750654697418213,
|
| 294 |
+
"learning_rate": 3.1021374676938584e-05,
|
| 295 |
+
"loss": 1.33,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.49572145175568016,
|
| 300 |
+
"grad_norm": 1.2510316371917725,
|
| 301 |
+
"learning_rate": 3.0016177896679255e-05,
|
| 302 |
+
"loss": 1.2919,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.5075243434641488,
|
| 307 |
+
"grad_norm": 1.2658268213272095,
|
| 308 |
+
"learning_rate": 2.9002455982394944e-05,
|
| 309 |
+
"loss": 1.2649,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.5193272351726173,
|
| 314 |
+
"grad_norm": 1.905948519706726,
|
| 315 |
+
"learning_rate": 2.798193178270889e-05,
|
| 316 |
+
"loss": 1.3047,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.5311301268810859,
|
| 321 |
+
"grad_norm": 1.2382662296295166,
|
| 322 |
+
"learning_rate": 2.695633970691786e-05,
|
| 323 |
+
"loss": 1.2862,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.5429330185895545,
|
| 328 |
+
"grad_norm": 1.2122917175292969,
|
| 329 |
+
"learning_rate": 2.592742277731513e-05,
|
| 330 |
+
"loss": 1.2843,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.554735910298023,
|
| 335 |
+
"grad_norm": 1.3638920783996582,
|
| 336 |
+
"learning_rate": 2.489692966687566e-05,
|
| 337 |
+
"loss": 1.2795,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.5665388020064915,
|
| 342 |
+
"grad_norm": 1.2353355884552002,
|
| 343 |
+
"learning_rate": 2.386661172733762e-05,
|
| 344 |
+
"loss": 1.1897,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.5783416937149601,
|
| 349 |
+
"grad_norm": 1.440238118171692,
|
| 350 |
+
"learning_rate": 2.2838220012731365e-05,
|
| 351 |
+
"loss": 1.3352,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.5901445854234287,
|
| 356 |
+
"grad_norm": 1.302875280380249,
|
| 357 |
+
"learning_rate": 2.1813502303414306e-05,
|
| 358 |
+
"loss": 1.2552,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.6019474771318973,
|
| 363 |
+
"grad_norm": 1.3485292196273804,
|
| 364 |
+
"learning_rate": 2.0794200135669584e-05,
|
| 365 |
+
"loss": 1.2573,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.6137503688403659,
|
| 370 |
+
"grad_norm": 2.1018223762512207,
|
| 371 |
+
"learning_rate": 1.9782045841916625e-05,
|
| 372 |
+
"loss": 1.2564,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.6255532605488344,
|
| 377 |
+
"grad_norm": 1.6336476802825928,
|
| 378 |
+
"learning_rate": 1.877875960656394e-05,
|
| 379 |
+
"loss": 1.1512,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.637356152257303,
|
| 384 |
+
"grad_norm": 1.4360566139221191,
|
| 385 |
+
"learning_rate": 1.7786046542507843e-05,
|
| 386 |
+
"loss": 1.2434,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.6491590439657716,
|
| 391 |
+
"grad_norm": 1.1216990947723389,
|
| 392 |
+
"learning_rate": 1.680559379324558e-05,
|
| 393 |
+
"loss": 1.325,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.6609619356742402,
|
| 398 |
+
"grad_norm": 1.6999801397323608,
|
| 399 |
+
"learning_rate": 1.583906766552799e-05,
|
| 400 |
+
"loss": 1.2197,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.6727648273827088,
|
| 405 |
+
"grad_norm": 1.4907481670379639,
|
| 406 |
+
"learning_rate": 1.4888110797424782e-05,
|
| 407 |
+
"loss": 1.2821,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.6845677190911773,
|
| 412 |
+
"grad_norm": 1.2150344848632812,
|
| 413 |
+
"learning_rate": 1.3954339366615334e-05,
|
| 414 |
+
"loss": 1.239,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.6963706107996459,
|
| 419 |
+
"grad_norm": 1.6709622144699097,
|
| 420 |
+
"learning_rate": 1.303934034364983e-05,
|
| 421 |
+
"loss": 1.2403,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.7081735025081145,
|
| 426 |
+
"grad_norm": 1.5160703659057617,
|
| 427 |
+
"learning_rate": 1.21446687948485e-05,
|
| 428 |
+
"loss": 1.2466,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.7199763942165831,
|
| 433 |
+
"grad_norm": 1.2667752504348755,
|
| 434 |
+
"learning_rate": 1.1271845239423196e-05,
|
| 435 |
+
"loss": 1.1662,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.7317792859250516,
|
| 440 |
+
"grad_norm": 1.685145616531372,
|
| 441 |
+
"learning_rate": 1.0422353065312573e-05,
|
| 442 |
+
"loss": 1.3161,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.7435821776335202,
|
| 447 |
+
"grad_norm": 1.5131856203079224,
|
| 448 |
+
"learning_rate": 9.59763600812305e-06,
|
| 449 |
+
"loss": 1.2608,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.7553850693419888,
|
| 454 |
+
"grad_norm": 1.2261701822280884,
|
| 455 |
+
"learning_rate": 8.79909569745987e-06,
|
| 456 |
+
"loss": 1.1507,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.7671879610504574,
|
| 461 |
+
"grad_norm": 1.2804995775222778,
|
| 462 |
+
"learning_rate": 8.028089274818624e-06,
|
| 463 |
+
"loss": 1.3008,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.778990852758926,
|
| 468 |
+
"grad_norm": 1.3678828477859497,
|
| 469 |
+
"learning_rate": 7.285927087085423e-06,
|
| 470 |
+
"loss": 1.272,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.7907937444673945,
|
| 475 |
+
"grad_norm": 1.3345593214035034,
|
| 476 |
+
"learning_rate": 6.5738704595659065e-06,
|
| 477 |
+
"loss": 1.1615,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.8025966361758631,
|
| 482 |
+
"grad_norm": 1.2585678100585938,
|
| 483 |
+
"learning_rate": 5.893129552327781e-06,
|
| 484 |
+
"loss": 1.1878,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.8143995278843317,
|
| 489 |
+
"grad_norm": 1.3462913036346436,
|
| 490 |
+
"learning_rate": 5.244861303500026e-06,
|
| 491 |
+
"loss": 1.2436,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.8262024195928003,
|
| 496 |
+
"grad_norm": 1.1118088960647583,
|
| 497 |
+
"learning_rate": 4.630167463024393e-06,
|
| 498 |
+
"loss": 1.0838,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.8380053113012688,
|
| 503 |
+
"grad_norm": 1.7299799919128418,
|
| 504 |
+
"learning_rate": 4.050092720200638e-06,
|
| 505 |
+
"loss": 1.1495,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.8498082030097374,
|
| 510 |
+
"grad_norm": 1.3773056268692017,
|
| 511 |
+
"learning_rate": 3.5056229282080077e-06,
|
| 512 |
+
"loss": 1.234,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.861611094718206,
|
| 517 |
+
"grad_norm": 1.2820888757705688,
|
| 518 |
+
"learning_rate": 2.997683428620296e-06,
|
| 519 |
+
"loss": 1.1803,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.8734139864266746,
|
| 524 |
+
"grad_norm": 1.3301385641098022,
|
| 525 |
+
"learning_rate": 2.527137478762037e-06,
|
| 526 |
+
"loss": 1.2197,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.8852168781351432,
|
| 531 |
+
"grad_norm": 1.7628834247589111,
|
| 532 |
+
"learning_rate": 2.094784784578707e-06,
|
| 533 |
+
"loss": 1.2354,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.8970197698436116,
|
| 538 |
+
"grad_norm": 1.2032676935195923,
|
| 539 |
+
"learning_rate": 1.7013601415141383e-06,
|
| 540 |
+
"loss": 1.1835,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.9088226615520802,
|
| 545 |
+
"grad_norm": 1.5983058214187622,
|
| 546 |
+
"learning_rate": 1.3475321857052386e-06,
|
| 547 |
+
"loss": 1.1651,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.9206255532605488,
|
| 552 |
+
"grad_norm": 1.0227899551391602,
|
| 553 |
+
"learning_rate": 1.03390225761624e-06,
|
| 554 |
+
"loss": 1.1662,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.9324284449690174,
|
| 559 |
+
"grad_norm": 1.352665901184082,
|
| 560 |
+
"learning_rate": 7.610033800438344e-07,
|
| 561 |
+
"loss": 1.1798,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.9442313366774859,
|
| 566 |
+
"grad_norm": 1.6476454734802246,
|
| 567 |
+
"learning_rate": 5.292993522301005e-07,
|
| 568 |
+
"loss": 1.2053,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.9560342283859545,
|
| 573 |
+
"grad_norm": 1.2775633335113525,
|
| 574 |
+
"learning_rate": 3.3918396162275214e-07,
|
| 575 |
+
"loss": 1.2049,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.9678371200944231,
|
| 580 |
+
"grad_norm": 1.4991925954818726,
|
| 581 |
+
"learning_rate": 1.9098031462242705e-07,
|
| 582 |
+
"loss": 1.2097,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.9796400118028917,
|
| 587 |
+
"grad_norm": 1.3501712083816528,
|
| 588 |
+
"learning_rate": 8.494028745434368e-08,
|
| 589 |
+
"loss": 1.2085,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.9914429035113603,
|
| 594 |
+
"grad_norm": 1.319488763809204,
|
| 595 |
+
"learning_rate": 2.124409809766692e-08,
|
| 596 |
+
"loss": 1.1854,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.9997049277072882,
|
| 601 |
+
"step": 847,
|
| 602 |
+
"total_flos": 1.2532647345436754e+18,
|
| 603 |
+
"train_loss": 1.3799798170537847,
|
| 604 |
+
"train_runtime": 10524.3823,
|
| 605 |
+
"train_samples_per_second": 2.576,
|
| 606 |
+
"train_steps_per_second": 0.08
|
| 607 |
+
}
|
| 608 |
+
],
|
| 609 |
+
"logging_steps": 10,
|
| 610 |
+
"max_steps": 847,
|
| 611 |
+
"num_input_tokens_seen": 0,
|
| 612 |
+
"num_train_epochs": 1,
|
| 613 |
+
"save_steps": 1000,
|
| 614 |
+
"stateful_callbacks": {
|
| 615 |
+
"TrainerControl": {
|
| 616 |
+
"args": {
|
| 617 |
+
"should_epoch_stop": false,
|
| 618 |
+
"should_evaluate": false,
|
| 619 |
+
"should_log": false,
|
| 620 |
+
"should_save": true,
|
| 621 |
+
"should_training_stop": true
|
| 622 |
+
},
|
| 623 |
+
"attributes": {}
|
| 624 |
+
}
|
| 625 |
+
},
|
| 626 |
+
"total_flos": 1.2532647345436754e+18,
|
| 627 |
+
"train_batch_size": 8,
|
| 628 |
+
"trial_name": null,
|
| 629 |
+
"trial_params": null
|
| 630 |
+
}
|