Fathom: upload expert-e4-forensics/training_log.json
Browse files
adapters/expert-e4-forensics/training_log.json
ADDED
|
@@ -0,0 +1,611 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"loss": 1.9463,
|
| 4 |
+
"grad_norm": 0.9043406844139099,
|
| 5 |
+
"learning_rate": 1.8e-05,
|
| 6 |
+
"entropy": 1.4898367643356323,
|
| 7 |
+
"num_tokens": 239592.0,
|
| 8 |
+
"mean_token_accuracy": 0.5726580291986465,
|
| 9 |
+
"epoch": 0.016680567139282735,
|
| 10 |
+
"step": 10
|
| 11 |
+
},
|
| 12 |
+
{
|
| 13 |
+
"loss": 1.8033,
|
| 14 |
+
"grad_norm": 0.6490057706832886,
|
| 15 |
+
"learning_rate": 3.8e-05,
|
| 16 |
+
"entropy": 1.7221096634864808,
|
| 17 |
+
"num_tokens": 480339.0,
|
| 18 |
+
"mean_token_accuracy": 0.5810627460479736,
|
| 19 |
+
"epoch": 0.03336113427856547,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"loss": 1.5177,
|
| 24 |
+
"grad_norm": 1.1159552335739136,
|
| 25 |
+
"learning_rate": 5.8e-05,
|
| 26 |
+
"entropy": 1.5653360843658448,
|
| 27 |
+
"num_tokens": 719994.0,
|
| 28 |
+
"mean_token_accuracy": 0.6268167346715927,
|
| 29 |
+
"epoch": 0.05004170141784821,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"loss": 1.0739,
|
| 34 |
+
"grad_norm": 0.4417094886302948,
|
| 35 |
+
"learning_rate": 7.800000000000001e-05,
|
| 36 |
+
"entropy": 1.053076034784317,
|
| 37 |
+
"num_tokens": 957912.0,
|
| 38 |
+
"mean_token_accuracy": 0.7257216989994049,
|
| 39 |
+
"epoch": 0.06672226855713094,
|
| 40 |
+
"step": 40
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"loss": 0.9505,
|
| 44 |
+
"grad_norm": 0.3340432941913605,
|
| 45 |
+
"learning_rate": 9.8e-05,
|
| 46 |
+
"entropy": 0.9432537257671356,
|
| 47 |
+
"num_tokens": 1197740.0,
|
| 48 |
+
"mean_token_accuracy": 0.7573244959115982,
|
| 49 |
+
"epoch": 0.08340283569641367,
|
| 50 |
+
"step": 50
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"loss": 0.9094,
|
| 54 |
+
"grad_norm": 0.3062295615673065,
|
| 55 |
+
"learning_rate": 9.993394529660307e-05,
|
| 56 |
+
"entropy": 0.902097898721695,
|
| 57 |
+
"num_tokens": 1434364.0,
|
| 58 |
+
"mean_token_accuracy": 0.761836588382721,
|
| 59 |
+
"epoch": 0.10008340283569642,
|
| 60 |
+
"step": 60
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"loss": 0.8602,
|
| 64 |
+
"grad_norm": 0.302666038274765,
|
| 65 |
+
"learning_rate": 9.970583211010007e-05,
|
| 66 |
+
"entropy": 0.8552027255296707,
|
| 67 |
+
"num_tokens": 1673721.0,
|
| 68 |
+
"mean_token_accuracy": 0.7724049925804138,
|
| 69 |
+
"epoch": 0.11676396997497915,
|
| 70 |
+
"step": 70
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"loss": 0.8386,
|
| 74 |
+
"grad_norm": 0.35731419920921326,
|
| 75 |
+
"learning_rate": 9.931558883072403e-05,
|
| 76 |
+
"entropy": 0.8345210403203964,
|
| 77 |
+
"num_tokens": 1921434.0,
|
| 78 |
+
"mean_token_accuracy": 0.7766192525625228,
|
| 79 |
+
"epoch": 0.13344453711426188,
|
| 80 |
+
"step": 80
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"loss": 0.8519,
|
| 84 |
+
"grad_norm": 0.3715258538722992,
|
| 85 |
+
"learning_rate": 9.87644883509393e-05,
|
| 86 |
+
"entropy": 0.8489096939563752,
|
| 87 |
+
"num_tokens": 2162407.0,
|
| 88 |
+
"mean_token_accuracy": 0.7723332077264786,
|
| 89 |
+
"epoch": 0.1501251042535446,
|
| 90 |
+
"step": 90
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"loss": 0.766,
|
| 94 |
+
"grad_norm": 0.3652157783508301,
|
| 95 |
+
"learning_rate": 9.805432824596348e-05,
|
| 96 |
+
"entropy": 0.7692611366510391,
|
| 97 |
+
"num_tokens": 2399946.0,
|
| 98 |
+
"mean_token_accuracy": 0.7895216166973114,
|
| 99 |
+
"epoch": 0.16680567139282734,
|
| 100 |
+
"step": 100
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"loss": 0.7927,
|
| 104 |
+
"grad_norm": 0.3489046096801758,
|
| 105 |
+
"learning_rate": 9.71874249104506e-05,
|
| 106 |
+
"entropy": 0.7879800289869309,
|
| 107 |
+
"num_tokens": 2643290.0,
|
| 108 |
+
"mean_token_accuracy": 0.7858752965927124,
|
| 109 |
+
"epoch": 0.1834862385321101,
|
| 110 |
+
"step": 110
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"loss": 0.7939,
|
| 114 |
+
"grad_norm": 0.3456013798713684,
|
| 115 |
+
"learning_rate": 9.61666060028933e-05,
|
| 116 |
+
"entropy": 0.8026239216327667,
|
| 117 |
+
"num_tokens": 2890007.0,
|
| 118 |
+
"mean_token_accuracy": 0.7865915298461914,
|
| 119 |
+
"epoch": 0.20016680567139283,
|
| 120 |
+
"step": 120
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"loss": 0.7367,
|
| 124 |
+
"grad_norm": 0.390689492225647,
|
| 125 |
+
"learning_rate": 9.499520122238845e-05,
|
| 126 |
+
"entropy": 0.7422888517379761,
|
| 127 |
+
"num_tokens": 3134139.0,
|
| 128 |
+
"mean_token_accuracy": 0.7972356855869294,
|
| 129 |
+
"epoch": 0.21684737281067556,
|
| 130 |
+
"step": 130
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"loss": 0.7339,
|
| 134 |
+
"grad_norm": 0.3555961549282074,
|
| 135 |
+
"learning_rate": 9.367703144785096e-05,
|
| 136 |
+
"entropy": 0.740860840678215,
|
| 137 |
+
"num_tokens": 3382098.0,
|
| 138 |
+
"mean_token_accuracy": 0.797610804438591,
|
| 139 |
+
"epoch": 0.2335279399499583,
|
| 140 |
+
"step": 140
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"loss": 0.7405,
|
| 144 |
+
"grad_norm": 0.3681407570838928,
|
| 145 |
+
"learning_rate": 9.221639627510076e-05,
|
| 146 |
+
"entropy": 0.744793301820755,
|
| 147 |
+
"num_tokens": 3622551.0,
|
| 148 |
+
"mean_token_accuracy": 0.7954040169715881,
|
| 149 |
+
"epoch": 0.25020850708924103,
|
| 150 |
+
"step": 150
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"loss": 0.7253,
|
| 154 |
+
"grad_norm": 0.32773861289024353,
|
| 155 |
+
"learning_rate": 9.061805999247504e-05,
|
| 156 |
+
"entropy": 0.7334282338619232,
|
| 157 |
+
"num_tokens": 3865667.0,
|
| 158 |
+
"mean_token_accuracy": 0.799143373966217,
|
| 159 |
+
"epoch": 0.26688907422852376,
|
| 160 |
+
"step": 160
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"loss": 0.711,
|
| 164 |
+
"grad_norm": 0.32971468567848206,
|
| 165 |
+
"learning_rate": 8.88872360407099e-05,
|
| 166 |
+
"entropy": 0.7231753587722778,
|
| 167 |
+
"num_tokens": 4101163.0,
|
| 168 |
+
"mean_token_accuracy": 0.803010705113411,
|
| 169 |
+
"epoch": 0.2835696413678065,
|
| 170 |
+
"step": 170
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"loss": 0.7172,
|
| 174 |
+
"grad_norm": 0.3492438495159149,
|
| 175 |
+
"learning_rate": 8.70295700077803e-05,
|
| 176 |
+
"entropy": 0.7242234885692597,
|
| 177 |
+
"num_tokens": 4344116.0,
|
| 178 |
+
"mean_token_accuracy": 0.8021188855171204,
|
| 179 |
+
"epoch": 0.3002502085070892,
|
| 180 |
+
"step": 180
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"loss": 0.7291,
|
| 184 |
+
"grad_norm": 0.33426719903945923,
|
| 185 |
+
"learning_rate": 8.505112121416555e-05,
|
| 186 |
+
"entropy": 0.7333565503358841,
|
| 187 |
+
"num_tokens": 4581952.0,
|
| 188 |
+
"mean_token_accuracy": 0.7988870143890381,
|
| 189 |
+
"epoch": 0.31693077564637195,
|
| 190 |
+
"step": 190
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"loss": 0.7138,
|
| 194 |
+
"grad_norm": 0.33009985089302063,
|
| 195 |
+
"learning_rate": 8.295834294860535e-05,
|
| 196 |
+
"entropy": 0.7183975011110306,
|
| 197 |
+
"num_tokens": 4829123.0,
|
| 198 |
+
"mean_token_accuracy": 0.8039891421794891,
|
| 199 |
+
"epoch": 0.3336113427856547,
|
| 200 |
+
"step": 200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"loss": 0.685,
|
| 204 |
+
"grad_norm": 0.35757768154144287,
|
| 205 |
+
"learning_rate": 8.075806141881326e-05,
|
| 206 |
+
"entropy": 0.6890431672334671,
|
| 207 |
+
"num_tokens": 5067338.0,
|
| 208 |
+
"mean_token_accuracy": 0.8091971307992936,
|
| 209 |
+
"epoch": 0.3502919099249375,
|
| 210 |
+
"step": 210
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"loss": 0.6954,
|
| 214 |
+
"grad_norm": 0.3246421813964844,
|
| 215 |
+
"learning_rate": 7.845745348580591e-05,
|
| 216 |
+
"entropy": 0.6971507608890534,
|
| 217 |
+
"num_tokens": 5311215.0,
|
| 218 |
+
"mean_token_accuracy": 0.8067805171012878,
|
| 219 |
+
"epoch": 0.3669724770642202,
|
| 220 |
+
"step": 220
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"loss": 0.7073,
|
| 224 |
+
"grad_norm": 0.3495083749294281,
|
| 225 |
+
"learning_rate": 7.60640232544742e-05,
|
| 226 |
+
"entropy": 0.7148334234952927,
|
| 227 |
+
"num_tokens": 5554999.0,
|
| 228 |
+
"mean_token_accuracy": 0.8036890238523483,
|
| 229 |
+
"epoch": 0.38365304420350294,
|
| 230 |
+
"step": 230
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"loss": 0.676,
|
| 234 |
+
"grad_norm": 0.38008829951286316,
|
| 235 |
+
"learning_rate": 7.358557759675284e-05,
|
| 236 |
+
"entropy": 0.6754449099302292,
|
| 237 |
+
"num_tokens": 5802860.0,
|
| 238 |
+
"mean_token_accuracy": 0.8100622564554214,
|
| 239 |
+
"epoch": 0.40033361134278567,
|
| 240 |
+
"step": 240
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"loss": 0.682,
|
| 244 |
+
"grad_norm": 0.3694831430912018,
|
| 245 |
+
"learning_rate": 7.103020068722674e-05,
|
| 246 |
+
"entropy": 0.6828817337751388,
|
| 247 |
+
"num_tokens": 6043669.0,
|
| 248 |
+
"mean_token_accuracy": 0.8099299311637879,
|
| 249 |
+
"epoch": 0.4170141784820684,
|
| 250 |
+
"step": 250
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"loss": 0.6754,
|
| 254 |
+
"grad_norm": 0.34716424345970154,
|
| 255 |
+
"learning_rate": 6.840622763423391e-05,
|
| 256 |
+
"entropy": 0.6824066579341889,
|
| 257 |
+
"num_tokens": 6279154.0,
|
| 258 |
+
"mean_token_accuracy": 0.8119429558515548,
|
| 259 |
+
"epoch": 0.43369474562135113,
|
| 260 |
+
"step": 260
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"loss": 0.6567,
|
| 264 |
+
"grad_norm": 0.3316477835178375,
|
| 265 |
+
"learning_rate": 6.572221729247441e-05,
|
| 266 |
+
"entropy": 0.6584962159395218,
|
| 267 |
+
"num_tokens": 6514535.0,
|
| 268 |
+
"mean_token_accuracy": 0.8146260052919387,
|
| 269 |
+
"epoch": 0.45037531276063386,
|
| 270 |
+
"step": 270
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"loss": 0.6816,
|
| 274 |
+
"grad_norm": 0.36305058002471924,
|
| 275 |
+
"learning_rate": 6.298692434580542e-05,
|
| 276 |
+
"entropy": 0.6815447330474853,
|
| 277 |
+
"num_tokens": 6757908.0,
|
| 278 |
+
"mean_token_accuracy": 0.8089588433504105,
|
| 279 |
+
"epoch": 0.4670558798999166,
|
| 280 |
+
"step": 280
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"loss": 0.6897,
|
| 284 |
+
"grad_norm": 0.33401617407798767,
|
| 285 |
+
"learning_rate": 6.0209270751282165e-05,
|
| 286 |
+
"entropy": 0.696434935927391,
|
| 287 |
+
"num_tokens": 6996464.0,
|
| 288 |
+
"mean_token_accuracy": 0.8082989573478698,
|
| 289 |
+
"epoch": 0.4837364470391993,
|
| 290 |
+
"step": 290
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"loss": 0.6921,
|
| 294 |
+
"grad_norm": 0.34646838903427124,
|
| 295 |
+
"learning_rate": 5.7398316637588336e-05,
|
| 296 |
+
"entropy": 0.6913289308547974,
|
| 297 |
+
"num_tokens": 7234102.0,
|
| 298 |
+
"mean_token_accuracy": 0.8069221824407578,
|
| 299 |
+
"epoch": 0.5004170141784821,
|
| 300 |
+
"step": 300
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"loss": 0.6802,
|
| 304 |
+
"grad_norm": 0.37350520491600037,
|
| 305 |
+
"learning_rate": 5.4563230752779595e-05,
|
| 306 |
+
"entropy": 0.6853398621082306,
|
| 307 |
+
"num_tokens": 7479635.0,
|
| 308 |
+
"mean_token_accuracy": 0.810760036110878,
|
| 309 |
+
"epoch": 0.5170975813177648,
|
| 310 |
+
"step": 310
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"loss": 0.6676,
|
| 314 |
+
"grad_norm": 0.3853772282600403,
|
| 315 |
+
"learning_rate": 5.1713260557733176e-05,
|
| 316 |
+
"entropy": 0.6696277529001236,
|
| 317 |
+
"num_tokens": 7718364.0,
|
| 318 |
+
"mean_token_accuracy": 0.8130882978439331,
|
| 319 |
+
"epoch": 0.5337781484570475,
|
| 320 |
+
"step": 320
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"loss": 0.6537,
|
| 324 |
+
"grad_norm": 0.338049054145813,
|
| 325 |
+
"learning_rate": 4.885770206285252e-05,
|
| 326 |
+
"entropy": 0.6589328497648239,
|
| 327 |
+
"num_tokens": 7958054.0,
|
| 328 |
+
"mean_token_accuracy": 0.8148987740278244,
|
| 329 |
+
"epoch": 0.5504587155963303,
|
| 330 |
+
"step": 330
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"loss": 0.6902,
|
| 334 |
+
"grad_norm": 0.303418904542923,
|
| 335 |
+
"learning_rate": 4.600586950641362e-05,
|
| 336 |
+
"entropy": 0.692239272594452,
|
| 337 |
+
"num_tokens": 8209298.0,
|
| 338 |
+
"mean_token_accuracy": 0.8085194110870362,
|
| 339 |
+
"epoch": 0.567139282735613,
|
| 340 |
+
"step": 340
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"loss": 0.6659,
|
| 344 |
+
"grad_norm": 0.38714390993118286,
|
| 345 |
+
"learning_rate": 4.316706497345572e-05,
|
| 346 |
+
"entropy": 0.6669029265642166,
|
| 347 |
+
"num_tokens": 8455256.0,
|
| 348 |
+
"mean_token_accuracy": 0.8131787478923798,
|
| 349 |
+
"epoch": 0.5838198498748958,
|
| 350 |
+
"step": 350
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"loss": 0.6579,
|
| 354 |
+
"grad_norm": 0.35571643710136414,
|
| 355 |
+
"learning_rate": 4.035054805431334e-05,
|
| 356 |
+
"entropy": 0.6595758140087128,
|
| 357 |
+
"num_tokens": 8695771.0,
|
| 358 |
+
"mean_token_accuracy": 0.8157992452383042,
|
| 359 |
+
"epoch": 0.6005004170141784,
|
| 360 |
+
"step": 360
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"loss": 0.6641,
|
| 364 |
+
"grad_norm": 0.3794392943382263,
|
| 365 |
+
"learning_rate": 3.756550564175727e-05,
|
| 366 |
+
"entropy": 0.6648996561765671,
|
| 367 |
+
"num_tokens": 8944844.0,
|
| 368 |
+
"mean_token_accuracy": 0.8132083296775818,
|
| 369 |
+
"epoch": 0.6171809841534612,
|
| 370 |
+
"step": 370
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"loss": 0.6577,
|
| 374 |
+
"grad_norm": 0.3610455393791199,
|
| 375 |
+
"learning_rate": 3.4821021965259904e-05,
|
| 376 |
+
"entropy": 0.6610071390867234,
|
| 377 |
+
"num_tokens": 9191203.0,
|
| 378 |
+
"mean_token_accuracy": 0.8145314186811448,
|
| 379 |
+
"epoch": 0.6338615512927439,
|
| 380 |
+
"step": 380
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"loss": 0.659,
|
| 384 |
+
"grad_norm": 0.3606182634830475,
|
| 385 |
+
"learning_rate": 3.212604896012679e-05,
|
| 386 |
+
"entropy": 0.660676684975624,
|
| 387 |
+
"num_tokens": 9426243.0,
|
| 388 |
+
"mean_token_accuracy": 0.8144401997327805,
|
| 389 |
+
"epoch": 0.6505421184320267,
|
| 390 |
+
"step": 390
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"loss": 0.6568,
|
| 394 |
+
"grad_norm": 0.3522314727306366,
|
| 395 |
+
"learning_rate": 2.948937706814442e-05,
|
| 396 |
+
"entropy": 0.6625478237867355,
|
| 397 |
+
"num_tokens": 9664795.0,
|
| 398 |
+
"mean_token_accuracy": 0.8151305437088012,
|
| 399 |
+
"epoch": 0.6672226855713094,
|
| 400 |
+
"step": 400
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"loss": 0.6536,
|
| 404 |
+
"grad_norm": 0.3643236756324768,
|
| 405 |
+
"learning_rate": 2.691960656498621e-05,
|
| 406 |
+
"entropy": 0.657148751616478,
|
| 407 |
+
"num_tokens": 9904436.0,
|
| 408 |
+
"mean_token_accuracy": 0.8148669481277466,
|
| 409 |
+
"epoch": 0.6839032527105922,
|
| 410 |
+
"step": 410
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"loss": 0.6704,
|
| 414 |
+
"grad_norm": 0.3429484963417053,
|
| 415 |
+
"learning_rate": 2.4425119507900813e-05,
|
| 416 |
+
"entropy": 0.6674410969018936,
|
| 417 |
+
"num_tokens": 10150234.0,
|
| 418 |
+
"mean_token_accuracy": 0.8119391798973083,
|
| 419 |
+
"epoch": 0.700583819849875,
|
| 420 |
+
"step": 420
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"loss": 0.7034,
|
| 424 |
+
"grad_norm": 0.3803344666957855,
|
| 425 |
+
"learning_rate": 2.2014052395183627e-05,
|
| 426 |
+
"entropy": 0.7082253098487854,
|
| 427 |
+
"num_tokens": 10391148.0,
|
| 428 |
+
"mean_token_accuracy": 0.8065144032239914,
|
| 429 |
+
"epoch": 0.7172643869891576,
|
| 430 |
+
"step": 430
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"loss": 0.6943,
|
| 434 |
+
"grad_norm": 0.30864450335502625,
|
| 435 |
+
"learning_rate": 1.969426962661059e-05,
|
| 436 |
+
"entropy": 0.7002600938081741,
|
| 437 |
+
"num_tokens": 10636950.0,
|
| 438 |
+
"mean_token_accuracy": 0.8064252346754074,
|
| 439 |
+
"epoch": 0.7339449541284404,
|
| 440 |
+
"step": 440
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"loss": 0.6803,
|
| 444 |
+
"grad_norm": 0.35500553250312805,
|
| 445 |
+
"learning_rate": 1.747333785140066e-05,
|
| 446 |
+
"entropy": 0.681396746635437,
|
| 447 |
+
"num_tokens": 10879243.0,
|
| 448 |
+
"mean_token_accuracy": 0.8101641505956649,
|
| 449 |
+
"epoch": 0.7506255212677231,
|
| 450 |
+
"step": 450
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"loss": 0.6421,
|
| 454 |
+
"grad_norm": 0.36364850401878357,
|
| 455 |
+
"learning_rate": 1.535850128737884e-05,
|
| 456 |
+
"entropy": 0.6462778955698013,
|
| 457 |
+
"num_tokens": 11114856.0,
|
| 458 |
+
"mean_token_accuracy": 0.8183332979679108,
|
| 459 |
+
"epoch": 0.7673060884070059,
|
| 460 |
+
"step": 460
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"loss": 0.6766,
|
| 464 |
+
"grad_norm": 0.33400440216064453,
|
| 465 |
+
"learning_rate": 1.335665809184341e-05,
|
| 466 |
+
"entropy": 0.6827076494693756,
|
| 467 |
+
"num_tokens": 11360827.0,
|
| 468 |
+
"mean_token_accuracy": 0.8112966924905777,
|
| 469 |
+
"epoch": 0.7839866555462885,
|
| 470 |
+
"step": 470
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"loss": 0.647,
|
| 474 |
+
"grad_norm": 0.3601098954677582,
|
| 475 |
+
"learning_rate": 1.1474337861210543e-05,
|
| 476 |
+
"entropy": 0.6471425205469131,
|
| 477 |
+
"num_tokens": 11601682.0,
|
| 478 |
+
"mean_token_accuracy": 0.8174957513809205,
|
| 479 |
+
"epoch": 0.8006672226855713,
|
| 480 |
+
"step": 480
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"loss": 0.6936,
|
| 484 |
+
"grad_norm": 0.36329320073127747,
|
| 485 |
+
"learning_rate": 9.717680332828016e-06,
|
| 486 |
+
"entropy": 0.7031383335590362,
|
| 487 |
+
"num_tokens": 11844375.0,
|
| 488 |
+
"mean_token_accuracy": 0.805937397480011,
|
| 489 |
+
"epoch": 0.817347789824854,
|
| 490 |
+
"step": 490
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"loss": 0.6743,
|
| 494 |
+
"grad_norm": 0.3474838137626648,
|
| 495 |
+
"learning_rate": 8.092415358428173e-06,
|
| 496 |
+
"entropy": 0.6719307899475098,
|
| 497 |
+
"num_tokens": 12085115.0,
|
| 498 |
+
"mean_token_accuracy": 0.8131552457809448,
|
| 499 |
+
"epoch": 0.8340283569641368,
|
| 500 |
+
"step": 500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"loss": 0.6344,
|
| 504 |
+
"grad_norm": 0.35991644859313965,
|
| 505 |
+
"learning_rate": 6.603844214542487e-06,
|
| 506 |
+
"entropy": 0.637131878733635,
|
| 507 |
+
"num_tokens": 12332095.0,
|
| 508 |
+
"mean_token_accuracy": 0.8191376119852066,
|
| 509 |
+
"epoch": 0.8507089241034195,
|
| 510 |
+
"step": 510
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"loss": 0.6349,
|
| 514 |
+
"grad_norm": 0.34625497460365295,
|
| 515 |
+
"learning_rate": 5.256822310839405e-06,
|
| 516 |
+
"entropy": 0.6384859681129456,
|
| 517 |
+
"num_tokens": 12575942.0,
|
| 518 |
+
"mean_token_accuracy": 0.8205374091863632,
|
| 519 |
+
"epoch": 0.8673894912427023,
|
| 520 |
+
"step": 520
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"loss": 0.6527,
|
| 524 |
+
"grad_norm": 0.34414827823638916,
|
| 525 |
+
"learning_rate": 4.055743352787267e-06,
|
| 526 |
+
"entropy": 0.6543797552585602,
|
| 527 |
+
"num_tokens": 12817224.0,
|
| 528 |
+
"mean_token_accuracy": 0.8158955574035645,
|
| 529 |
+
"epoch": 0.8840700583819849,
|
| 530 |
+
"step": 530
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"loss": 0.6387,
|
| 534 |
+
"grad_norm": 0.32114923000335693,
|
| 535 |
+
"learning_rate": 3.0045250103002296e-06,
|
| 536 |
+
"entropy": 0.6461419433355331,
|
| 537 |
+
"num_tokens": 13055918.0,
|
| 538 |
+
"mean_token_accuracy": 0.8180903941392899,
|
| 539 |
+
"epoch": 0.9007506255212677,
|
| 540 |
+
"step": 540
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"loss": 0.6749,
|
| 544 |
+
"grad_norm": 0.38731274008750916,
|
| 545 |
+
"learning_rate": 2.1065961391133702e-06,
|
| 546 |
+
"entropy": 0.6730036258697509,
|
| 547 |
+
"num_tokens": 13296781.0,
|
| 548 |
+
"mean_token_accuracy": 0.8106796860694885,
|
| 549 |
+
"epoch": 0.9174311926605505,
|
| 550 |
+
"step": 550
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"loss": 0.63,
|
| 554 |
+
"grad_norm": 0.3419639766216278,
|
| 555 |
+
"learning_rate": 1.3648855965679497e-06,
|
| 556 |
+
"entropy": 0.6305610507726669,
|
| 557 |
+
"num_tokens": 13536924.0,
|
| 558 |
+
"mean_token_accuracy": 0.8212851226329804,
|
| 559 |
+
"epoch": 0.9341117597998332,
|
| 560 |
+
"step": 560
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"loss": 0.641,
|
| 564 |
+
"grad_norm": 0.3530997931957245,
|
| 565 |
+
"learning_rate": 7.818126882875254e-07,
|
| 566 |
+
"entropy": 0.647763478755951,
|
| 567 |
+
"num_tokens": 13773817.0,
|
| 568 |
+
"mean_token_accuracy": 0.8177464485168457,
|
| 569 |
+
"epoch": 0.950792326939116,
|
| 570 |
+
"step": 570
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"loss": 0.6724,
|
| 574 |
+
"grad_norm": 0.3384498953819275,
|
| 575 |
+
"learning_rate": 3.5927927690588283e-07,
|
| 576 |
+
"entropy": 0.675562146306038,
|
| 577 |
+
"num_tokens": 14012569.0,
|
| 578 |
+
"mean_token_accuracy": 0.8116778939962387,
|
| 579 |
+
"epoch": 0.9674728940783986,
|
| 580 |
+
"step": 580
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"loss": 0.6523,
|
| 584 |
+
"grad_norm": 0.3589385747909546,
|
| 585 |
+
"learning_rate": 9.866357858642205e-08,
|
| 586 |
+
"entropy": 0.6570082008838654,
|
| 587 |
+
"num_tokens": 14254389.0,
|
| 588 |
+
"mean_token_accuracy": 0.8173295229673385,
|
| 589 |
+
"epoch": 0.9841534612176814,
|
| 590 |
+
"step": 590
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"loss": 0.6664,
|
| 594 |
+
"grad_norm": 0.5415430665016174,
|
| 595 |
+
"learning_rate": 8.156675674941828e-10,
|
| 596 |
+
"entropy": 0.6749062004842257,
|
| 597 |
+
"num_tokens": 14485000.0,
|
| 598 |
+
"mean_token_accuracy": 0.8125993891766197,
|
| 599 |
+
"epoch": 1.0,
|
| 600 |
+
"step": 600
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"train_runtime": 6029.5713,
|
| 604 |
+
"train_samples_per_second": 3.181,
|
| 605 |
+
"train_steps_per_second": 0.1,
|
| 606 |
+
"total_flos": 6.620616404100956e+18,
|
| 607 |
+
"train_loss": 0.7606559038162232,
|
| 608 |
+
"epoch": 1.0,
|
| 609 |
+
"step": 600
|
| 610 |
+
}
|
| 611 |
+
]
|