Fathom: upload expert-e3-network/training_log.json
Browse files
adapters/expert-e3-network/training_log.json
ADDED
|
@@ -0,0 +1,634 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"loss": 1.6756,
|
| 4 |
+
"grad_norm": 0.9338335990905762,
|
| 5 |
+
"learning_rate": 1.8e-05,
|
| 6 |
+
"entropy": 1.2653020560741424,
|
| 7 |
+
"num_tokens": 211745.0,
|
| 8 |
+
"mean_token_accuracy": 0.626320019364357,
|
| 9 |
+
"epoch": 0.016,
|
| 10 |
+
"step": 10
|
| 11 |
+
},
|
| 12 |
+
{
|
| 13 |
+
"loss": 1.6675,
|
| 14 |
+
"grad_norm": 0.4600104093551636,
|
| 15 |
+
"learning_rate": 3.8e-05,
|
| 16 |
+
"entropy": 1.5156079709529877,
|
| 17 |
+
"num_tokens": 413783.0,
|
| 18 |
+
"mean_token_accuracy": 0.6156355381011963,
|
| 19 |
+
"epoch": 0.032,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"loss": 1.4313,
|
| 24 |
+
"grad_norm": 0.5679482221603394,
|
| 25 |
+
"learning_rate": 5.8e-05,
|
| 26 |
+
"entropy": 1.511820811033249,
|
| 27 |
+
"num_tokens": 618078.0,
|
| 28 |
+
"mean_token_accuracy": 0.6563311725854873,
|
| 29 |
+
"epoch": 0.048,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"loss": 1.1182,
|
| 34 |
+
"grad_norm": 0.5529033541679382,
|
| 35 |
+
"learning_rate": 7.800000000000001e-05,
|
| 36 |
+
"entropy": 1.1426804274320603,
|
| 37 |
+
"num_tokens": 824579.0,
|
| 38 |
+
"mean_token_accuracy": 0.7136697113513947,
|
| 39 |
+
"epoch": 0.064,
|
| 40 |
+
"step": 40
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"loss": 0.9574,
|
| 44 |
+
"grad_norm": 0.4290783405303955,
|
| 45 |
+
"learning_rate": 9.8e-05,
|
| 46 |
+
"entropy": 0.9684652209281921,
|
| 47 |
+
"num_tokens": 1022359.0,
|
| 48 |
+
"mean_token_accuracy": 0.7489378124475479,
|
| 49 |
+
"epoch": 0.08,
|
| 50 |
+
"step": 50
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"loss": 0.8724,
|
| 54 |
+
"grad_norm": 0.34012866020202637,
|
| 55 |
+
"learning_rate": 9.993956318446873e-05,
|
| 56 |
+
"entropy": 0.8785539418458939,
|
| 57 |
+
"num_tokens": 1237697.0,
|
| 58 |
+
"mean_token_accuracy": 0.7662121266126632,
|
| 59 |
+
"epoch": 0.096,
|
| 60 |
+
"step": 60
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"loss": 0.8526,
|
| 64 |
+
"grad_norm": 0.4122697710990906,
|
| 65 |
+
"learning_rate": 9.973083336646172e-05,
|
| 66 |
+
"entropy": 0.8518908679485321,
|
| 67 |
+
"num_tokens": 1438438.0,
|
| 68 |
+
"mean_token_accuracy": 0.7753641337156296,
|
| 69 |
+
"epoch": 0.112,
|
| 70 |
+
"step": 70
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"loss": 0.7437,
|
| 74 |
+
"grad_norm": 0.3228837251663208,
|
| 75 |
+
"learning_rate": 9.937368719576374e-05,
|
| 76 |
+
"entropy": 0.7489307045936584,
|
| 77 |
+
"num_tokens": 1631589.0,
|
| 78 |
+
"mean_token_accuracy": 0.7941762745380402,
|
| 79 |
+
"epoch": 0.128,
|
| 80 |
+
"step": 80
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"loss": 0.7492,
|
| 84 |
+
"grad_norm": 0.33696669340133667,
|
| 85 |
+
"learning_rate": 9.886919053691883e-05,
|
| 86 |
+
"entropy": 0.7552473366260528,
|
| 87 |
+
"num_tokens": 1836258.0,
|
| 88 |
+
"mean_token_accuracy": 0.7946388304233551,
|
| 89 |
+
"epoch": 0.144,
|
| 90 |
+
"step": 90
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"loss": 0.7372,
|
| 94 |
+
"grad_norm": 0.2882266640663147,
|
| 95 |
+
"learning_rate": 9.82188490062415e-05,
|
| 96 |
+
"entropy": 0.7464645385742188,
|
| 97 |
+
"num_tokens": 2046902.0,
|
| 98 |
+
"mean_token_accuracy": 0.7958032578229904,
|
| 99 |
+
"epoch": 0.16,
|
| 100 |
+
"step": 100
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"loss": 0.7528,
|
| 104 |
+
"grad_norm": 0.34871771931648254,
|
| 105 |
+
"learning_rate": 9.742460347846587e-05,
|
| 106 |
+
"entropy": 0.7529336363077164,
|
| 107 |
+
"num_tokens": 2254427.0,
|
| 108 |
+
"mean_token_accuracy": 0.7927632719278336,
|
| 109 |
+
"epoch": 0.176,
|
| 110 |
+
"step": 110
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"loss": 0.6854,
|
| 114 |
+
"grad_norm": 0.344471275806427,
|
| 115 |
+
"learning_rate": 9.648882429441257e-05,
|
| 116 |
+
"entropy": 0.6884129703044891,
|
| 117 |
+
"num_tokens": 2452097.0,
|
| 118 |
+
"mean_token_accuracy": 0.8083581000566482,
|
| 119 |
+
"epoch": 0.192,
|
| 120 |
+
"step": 120
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"loss": 0.7035,
|
| 124 |
+
"grad_norm": 0.2790994942188263,
|
| 125 |
+
"learning_rate": 9.541430418696018e-05,
|
| 126 |
+
"entropy": 0.712130931019783,
|
| 127 |
+
"num_tokens": 2647959.0,
|
| 128 |
+
"mean_token_accuracy": 0.8047784239053726,
|
| 129 |
+
"epoch": 0.208,
|
| 130 |
+
"step": 130
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"loss": 0.7175,
|
| 134 |
+
"grad_norm": 0.3251378536224365,
|
| 135 |
+
"learning_rate": 9.420424994643266e-05,
|
| 136 |
+
"entropy": 0.7225965172052383,
|
| 137 |
+
"num_tokens": 2851644.0,
|
| 138 |
+
"mean_token_accuracy": 0.8003872334957123,
|
| 139 |
+
"epoch": 0.224,
|
| 140 |
+
"step": 140
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"loss": 0.6633,
|
| 144 |
+
"grad_norm": 0.3178017735481262,
|
| 145 |
+
"learning_rate": 9.28622728502766e-05,
|
| 146 |
+
"entropy": 0.6729940563440323,
|
| 147 |
+
"num_tokens": 3053179.0,
|
| 148 |
+
"mean_token_accuracy": 0.8133321523666381,
|
| 149 |
+
"epoch": 0.24,
|
| 150 |
+
"step": 150
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"loss": 0.7078,
|
| 154 |
+
"grad_norm": 0.32298043370246887,
|
| 155 |
+
"learning_rate": 9.139237788559e-05,
|
| 156 |
+
"entropy": 0.716013514995575,
|
| 157 |
+
"num_tokens": 3244569.0,
|
| 158 |
+
"mean_token_accuracy": 0.8037430495023727,
|
| 159 |
+
"epoch": 0.256,
|
| 160 |
+
"step": 160
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"loss": 0.7011,
|
| 164 |
+
"grad_norm": 0.3123107850551605,
|
| 165 |
+
"learning_rate": 8.979895179666673e-05,
|
| 166 |
+
"entropy": 0.6998003959655762,
|
| 167 |
+
"num_tokens": 3448693.0,
|
| 168 |
+
"mean_token_accuracy": 0.8040744125843048,
|
| 169 |
+
"epoch": 0.272,
|
| 170 |
+
"step": 170
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"loss": 0.6887,
|
| 174 |
+
"grad_norm": 0.3589050769805908,
|
| 175 |
+
"learning_rate": 8.808674999322728e-05,
|
| 176 |
+
"entropy": 0.6893662005662918,
|
| 177 |
+
"num_tokens": 3650299.0,
|
| 178 |
+
"mean_token_accuracy": 0.8075590372085572,
|
| 179 |
+
"epoch": 0.288,
|
| 180 |
+
"step": 180
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"loss": 0.7397,
|
| 184 |
+
"grad_norm": 0.3437199890613556,
|
| 185 |
+
"learning_rate": 8.626088235840726e-05,
|
| 186 |
+
"entropy": 0.7490889757871628,
|
| 187 |
+
"num_tokens": 3860706.0,
|
| 188 |
+
"mean_token_accuracy": 0.7960624009370804,
|
| 189 |
+
"epoch": 0.304,
|
| 190 |
+
"step": 190
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"loss": 0.7104,
|
| 194 |
+
"grad_norm": 0.31088024377822876,
|
| 195 |
+
"learning_rate": 8.43267979988576e-05,
|
| 196 |
+
"entropy": 0.7115123689174652,
|
| 197 |
+
"num_tokens": 4057498.0,
|
| 198 |
+
"mean_token_accuracy": 0.8006591260433197,
|
| 199 |
+
"epoch": 0.32,
|
| 200 |
+
"step": 200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"loss": 0.6628,
|
| 204 |
+
"grad_norm": 0.3589341640472412,
|
| 205 |
+
"learning_rate": 8.229026898246885e-05,
|
| 206 |
+
"entropy": 0.6689569860696792,
|
| 207 |
+
"num_tokens": 4251824.0,
|
| 208 |
+
"mean_token_accuracy": 0.812098690867424,
|
| 209 |
+
"epoch": 0.336,
|
| 210 |
+
"step": 210
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"loss": 0.676,
|
| 214 |
+
"grad_norm": 0.31112033128738403,
|
| 215 |
+
"learning_rate": 8.015737311225172e-05,
|
| 216 |
+
"entropy": 0.6726008355617523,
|
| 217 |
+
"num_tokens": 4465389.0,
|
| 218 |
+
"mean_token_accuracy": 0.8103758335113526,
|
| 219 |
+
"epoch": 0.352,
|
| 220 |
+
"step": 220
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"loss": 0.6668,
|
| 224 |
+
"grad_norm": 0.33138927817344666,
|
| 225 |
+
"learning_rate": 7.793447578778427e-05,
|
| 226 |
+
"entropy": 0.6735879182815552,
|
| 227 |
+
"num_tokens": 4685518.0,
|
| 228 |
+
"mean_token_accuracy": 0.8112154036760331,
|
| 229 |
+
"epoch": 0.368,
|
| 230 |
+
"step": 230
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"loss": 0.6414,
|
| 234 |
+
"grad_norm": 0.31951209902763367,
|
| 235 |
+
"learning_rate": 7.56282110083577e-05,
|
| 236 |
+
"entropy": 0.6442272961139679,
|
| 237 |
+
"num_tokens": 4886383.0,
|
| 238 |
+
"mean_token_accuracy": 0.8160840034484863,
|
| 239 |
+
"epoch": 0.384,
|
| 240 |
+
"step": 240
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"loss": 0.6817,
|
| 244 |
+
"grad_norm": 0.3405151665210724,
|
| 245 |
+
"learning_rate": 7.324546157451508e-05,
|
| 246 |
+
"entropy": 0.6818776607513428,
|
| 247 |
+
"num_tokens": 5086529.0,
|
| 248 |
+
"mean_token_accuracy": 0.8091985791921615,
|
| 249 |
+
"epoch": 0.4,
|
| 250 |
+
"step": 250
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"loss": 0.6843,
|
| 254 |
+
"grad_norm": 0.30915334820747375,
|
| 255 |
+
"learning_rate": 7.079333854706938e-05,
|
| 256 |
+
"entropy": 0.6939111232757569,
|
| 257 |
+
"num_tokens": 5298416.0,
|
| 258 |
+
"mean_token_accuracy": 0.8071506917476654,
|
| 259 |
+
"epoch": 0.416,
|
| 260 |
+
"step": 260
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"loss": 0.6805,
|
| 264 |
+
"grad_norm": 0.3196154832839966,
|
| 265 |
+
"learning_rate": 6.827916002490304e-05,
|
| 266 |
+
"entropy": 0.6853002905845642,
|
| 267 |
+
"num_tokens": 5502693.0,
|
| 268 |
+
"mean_token_accuracy": 0.8068861842155457,
|
| 269 |
+
"epoch": 0.432,
|
| 270 |
+
"step": 270
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"loss": 0.6497,
|
| 274 |
+
"grad_norm": 0.3364115059375763,
|
| 275 |
+
"learning_rate": 6.571042930488474e-05,
|
| 276 |
+
"entropy": 0.6549761325120926,
|
| 277 |
+
"num_tokens": 5697863.0,
|
| 278 |
+
"mean_token_accuracy": 0.81393081843853,
|
| 279 |
+
"epoch": 0.448,
|
| 280 |
+
"step": 280
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"loss": 0.6697,
|
| 284 |
+
"grad_norm": 0.33196505904197693,
|
| 285 |
+
"learning_rate": 6.309481248908235e-05,
|
| 286 |
+
"entropy": 0.6747944802045822,
|
| 287 |
+
"num_tokens": 5910487.0,
|
| 288 |
+
"mean_token_accuracy": 0.8128439605236053,
|
| 289 |
+
"epoch": 0.464,
|
| 290 |
+
"step": 290
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"loss": 0.6535,
|
| 294 |
+
"grad_norm": 0.30423638224601746,
|
| 295 |
+
"learning_rate": 6.044011560610153e-05,
|
| 296 |
+
"entropy": 0.6566255927085877,
|
| 297 |
+
"num_tokens": 6127477.0,
|
| 298 |
+
"mean_token_accuracy": 0.8121924549341202,
|
| 299 |
+
"epoch": 0.48,
|
| 300 |
+
"step": 300
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"loss": 0.6518,
|
| 304 |
+
"grad_norm": 0.3487173914909363,
|
| 305 |
+
"learning_rate": 5.775426131482811e-05,
|
| 306 |
+
"entropy": 0.6593502879142761,
|
| 307 |
+
"num_tokens": 6328197.0,
|
| 308 |
+
"mean_token_accuracy": 0.814925542473793,
|
| 309 |
+
"epoch": 0.496,
|
| 310 |
+
"step": 310
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"loss": 0.6931,
|
| 314 |
+
"grad_norm": 0.3142063021659851,
|
| 315 |
+
"learning_rate": 5.5045265260100234e-05,
|
| 316 |
+
"entropy": 0.700766509771347,
|
| 317 |
+
"num_tokens": 6536365.0,
|
| 318 |
+
"mean_token_accuracy": 0.8070295184850693,
|
| 319 |
+
"epoch": 0.512,
|
| 320 |
+
"step": 320
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"loss": 0.6785,
|
| 324 |
+
"grad_norm": 0.3409480154514313,
|
| 325 |
+
"learning_rate": 5.232121215087369e-05,
|
| 326 |
+
"entropy": 0.6776748090982437,
|
| 327 |
+
"num_tokens": 6742650.0,
|
| 328 |
+
"mean_token_accuracy": 0.809091717004776,
|
| 329 |
+
"epoch": 0.528,
|
| 330 |
+
"step": 330
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"loss": 0.6438,
|
| 334 |
+
"grad_norm": 0.3378758132457733,
|
| 335 |
+
"learning_rate": 4.959023163227284e-05,
|
| 336 |
+
"entropy": 0.6490969121456146,
|
| 337 |
+
"num_tokens": 6946868.0,
|
| 338 |
+
"mean_token_accuracy": 0.8178076684474945,
|
| 339 |
+
"epoch": 0.544,
|
| 340 |
+
"step": 340
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"loss": 0.6481,
|
| 344 |
+
"grad_norm": 0.3258441388607025,
|
| 345 |
+
"learning_rate": 4.6860474023534335e-05,
|
| 346 |
+
"entropy": 0.6563442766666412,
|
| 347 |
+
"num_tokens": 7138705.0,
|
| 348 |
+
"mean_token_accuracy": 0.817442661523819,
|
| 349 |
+
"epoch": 0.56,
|
| 350 |
+
"step": 350
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"loss": 0.6622,
|
| 354 |
+
"grad_norm": 0.32307952642440796,
|
| 355 |
+
"learning_rate": 4.4140085994251136e-05,
|
| 356 |
+
"entropy": 0.6632782518863678,
|
| 357 |
+
"num_tokens": 7348846.0,
|
| 358 |
+
"mean_token_accuracy": 0.812923377752304,
|
| 359 |
+
"epoch": 0.576,
|
| 360 |
+
"step": 360
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"loss": 0.6436,
|
| 364 |
+
"grad_norm": 0.35711175203323364,
|
| 365 |
+
"learning_rate": 4.143718625150854e-05,
|
| 366 |
+
"entropy": 0.6440205127000809,
|
| 367 |
+
"num_tokens": 7556952.0,
|
| 368 |
+
"mean_token_accuracy": 0.8172848135232925,
|
| 369 |
+
"epoch": 0.592,
|
| 370 |
+
"step": 370
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"loss": 0.6299,
|
| 374 |
+
"grad_norm": 0.3042103052139282,
|
| 375 |
+
"learning_rate": 3.875984131047135e-05,
|
| 376 |
+
"entropy": 0.6329021096229553,
|
| 377 |
+
"num_tokens": 7758339.0,
|
| 378 |
+
"mean_token_accuracy": 0.8194414287805557,
|
| 379 |
+
"epoch": 0.608,
|
| 380 |
+
"step": 380
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"loss": 0.6559,
|
| 384 |
+
"grad_norm": 0.35361942648887634,
|
| 385 |
+
"learning_rate": 3.6116041420732103e-05,
|
| 386 |
+
"entropy": 0.6580730825662613,
|
| 387 |
+
"num_tokens": 7961705.0,
|
| 388 |
+
"mean_token_accuracy": 0.8145245909690857,
|
| 389 |
+
"epoch": 0.624,
|
| 390 |
+
"step": 390
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"loss": 0.6394,
|
| 394 |
+
"grad_norm": 0.35071322321891785,
|
| 395 |
+
"learning_rate": 3.3513676720265934e-05,
|
| 396 |
+
"entropy": 0.6357898592948914,
|
| 397 |
+
"num_tokens": 8160317.0,
|
| 398 |
+
"mean_token_accuracy": 0.8177996665239334,
|
| 399 |
+
"epoch": 0.64,
|
| 400 |
+
"step": 400
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"loss": 0.6545,
|
| 404 |
+
"grad_norm": 0.31700801849365234,
|
| 405 |
+
"learning_rate": 3.0960513688157754e-05,
|
| 406 |
+
"entropy": 0.649863663315773,
|
| 407 |
+
"num_tokens": 8372896.0,
|
| 408 |
+
"mean_token_accuracy": 0.8160782635211945,
|
| 409 |
+
"epoch": 0.656,
|
| 410 |
+
"step": 410
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"loss": 0.656,
|
| 414 |
+
"grad_norm": 0.3056493401527405,
|
| 415 |
+
"learning_rate": 2.846417196637613e-05,
|
| 416 |
+
"entropy": 0.6630118876695633,
|
| 417 |
+
"num_tokens": 8582698.0,
|
| 418 |
+
"mean_token_accuracy": 0.8151620656251908,
|
| 419 |
+
"epoch": 0.672,
|
| 420 |
+
"step": 420
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"loss": 0.6869,
|
| 424 |
+
"grad_norm": 0.3303000330924988,
|
| 425 |
+
"learning_rate": 2.603210161976687e-05,
|
| 426 |
+
"entropy": 0.6842911779880524,
|
| 427 |
+
"num_tokens": 8794332.0,
|
| 428 |
+
"mean_token_accuracy": 0.805791437625885,
|
| 429 |
+
"epoch": 0.688,
|
| 430 |
+
"step": 430
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"loss": 0.6602,
|
| 434 |
+
"grad_norm": 0.3385516405105591,
|
| 435 |
+
"learning_rate": 2.3671560902131447e-05,
|
| 436 |
+
"entropy": 0.6720612704753876,
|
| 437 |
+
"num_tokens": 9008268.0,
|
| 438 |
+
"mean_token_accuracy": 0.811782768368721,
|
| 439 |
+
"epoch": 0.704,
|
| 440 |
+
"step": 440
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"loss": 0.6425,
|
| 444 |
+
"grad_norm": 0.3051205277442932,
|
| 445 |
+
"learning_rate": 2.1389594594744864e-05,
|
| 446 |
+
"entropy": 0.6423384606838226,
|
| 447 |
+
"num_tokens": 9216146.0,
|
| 448 |
+
"mean_token_accuracy": 0.81755151450634,
|
| 449 |
+
"epoch": 0.72,
|
| 450 |
+
"step": 450
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"loss": 0.6262,
|
| 454 |
+
"grad_norm": 0.3302038013935089,
|
| 455 |
+
"learning_rate": 1.9193012981959475e-05,
|
| 456 |
+
"entropy": 0.6281730026006699,
|
| 457 |
+
"num_tokens": 9418629.0,
|
| 458 |
+
"mean_token_accuracy": 0.820314571261406,
|
| 459 |
+
"epoch": 0.736,
|
| 460 |
+
"step": 460
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"loss": 0.6204,
|
| 464 |
+
"grad_norm": 0.2998092472553253,
|
| 465 |
+
"learning_rate": 1.7088371526639853e-05,
|
| 466 |
+
"entropy": 0.6200591921806335,
|
| 467 |
+
"num_tokens": 9624430.0,
|
| 468 |
+
"mean_token_accuracy": 0.8230848222970962,
|
| 469 |
+
"epoch": 0.752,
|
| 470 |
+
"step": 470
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"loss": 0.6133,
|
| 474 |
+
"grad_norm": 0.3248184025287628,
|
| 475 |
+
"learning_rate": 1.508195130608504e-05,
|
| 476 |
+
"entropy": 0.6175663053989411,
|
| 477 |
+
"num_tokens": 9824038.0,
|
| 478 |
+
"mean_token_accuracy": 0.8229408711194992,
|
| 479 |
+
"epoch": 0.768,
|
| 480 |
+
"step": 480
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"loss": 0.6471,
|
| 484 |
+
"grad_norm": 0.3508250117301941,
|
| 485 |
+
"learning_rate": 1.3179740266825253e-05,
|
| 486 |
+
"entropy": 0.6427986294031143,
|
| 487 |
+
"num_tokens": 10029090.0,
|
| 488 |
+
"mean_token_accuracy": 0.8153195798397064,
|
| 489 |
+
"epoch": 0.784,
|
| 490 |
+
"step": 490
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"loss": 0.6494,
|
| 494 |
+
"grad_norm": 0.3246026933193207,
|
| 495 |
+
"learning_rate": 1.1387415354235887e-05,
|
| 496 |
+
"entropy": 0.6497834831476211,
|
| 497 |
+
"num_tokens": 10233813.0,
|
| 498 |
+
"mean_token_accuracy": 0.8152256548404694,
|
| 499 |
+
"epoch": 0.8,
|
| 500 |
+
"step": 500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"loss": 0.621,
|
| 504 |
+
"grad_norm": 0.3171588182449341,
|
| 505 |
+
"learning_rate": 9.710325570301348e-06,
|
| 506 |
+
"entropy": 0.6244612663984299,
|
| 507 |
+
"num_tokens": 10435960.0,
|
| 508 |
+
"mean_token_accuracy": 0.8210752457380295,
|
| 509 |
+
"epoch": 0.816,
|
| 510 |
+
"step": 510
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"loss": 0.5986,
|
| 514 |
+
"grad_norm": 0.32061895728111267,
|
| 515 |
+
"learning_rate": 8.153476010090788e-06,
|
| 516 |
+
"entropy": 0.5975286513566971,
|
| 517 |
+
"num_tokens": 10636030.0,
|
| 518 |
+
"mean_token_accuracy": 0.827234199643135,
|
| 519 |
+
"epoch": 0.832,
|
| 520 |
+
"step": 520
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"loss": 0.6372,
|
| 524 |
+
"grad_norm": 0.3552604615688324,
|
| 525 |
+
"learning_rate": 6.721512924587242e-06,
|
| 526 |
+
"entropy": 0.6376388996839524,
|
| 527 |
+
"num_tokens": 10838906.0,
|
| 528 |
+
"mean_token_accuracy": 0.818820309638977,
|
| 529 |
+
"epoch": 0.848,
|
| 530 |
+
"step": 530
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"loss": 0.6304,
|
| 534 |
+
"grad_norm": 0.33599594235420227,
|
| 535 |
+
"learning_rate": 5.418709854448489e-06,
|
| 536 |
+
"entropy": 0.6320249021053315,
|
| 537 |
+
"num_tokens": 11026457.0,
|
| 538 |
+
"mean_token_accuracy": 0.8198111385107041,
|
| 539 |
+
"epoch": 0.864,
|
| 540 |
+
"step": 540
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"loss": 0.6235,
|
| 544 |
+
"grad_norm": 0.41409239172935486,
|
| 545 |
+
"learning_rate": 4.248954876082195e-06,
|
| 546 |
+
"entropy": 0.6273589804768562,
|
| 547 |
+
"num_tokens": 11211997.0,
|
| 548 |
+
"mean_token_accuracy": 0.8219926416873932,
|
| 549 |
+
"epoch": 0.88,
|
| 550 |
+
"step": 550
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"loss": 0.6511,
|
| 554 |
+
"grad_norm": 0.337541401386261,
|
| 555 |
+
"learning_rate": 3.2157389980976783e-06,
|
| 556 |
+
"entropy": 0.6515613377094269,
|
| 557 |
+
"num_tokens": 11421486.0,
|
| 558 |
+
"mean_token_accuracy": 0.815014871954918,
|
| 559 |
+
"epoch": 0.896,
|
| 560 |
+
"step": 560
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"loss": 0.6209,
|
| 564 |
+
"grad_norm": 0.3296094536781311,
|
| 565 |
+
"learning_rate": 2.322145742764181e-06,
|
| 566 |
+
"entropy": 0.6218553513288498,
|
| 567 |
+
"num_tokens": 11621269.0,
|
| 568 |
+
"mean_token_accuracy": 0.8230124861001968,
|
| 569 |
+
"epoch": 0.912,
|
| 570 |
+
"step": 570
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"loss": 0.634,
|
| 574 |
+
"grad_norm": 0.34123408794403076,
|
| 575 |
+
"learning_rate": 1.5708419435684462e-06,
|
| 576 |
+
"entropy": 0.6355304539203643,
|
| 577 |
+
"num_tokens": 11830087.0,
|
| 578 |
+
"mean_token_accuracy": 0.818019899725914,
|
| 579 |
+
"epoch": 0.928,
|
| 580 |
+
"step": 580
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"loss": 0.6276,
|
| 584 |
+
"grad_norm": 0.2921477258205414,
|
| 585 |
+
"learning_rate": 9.640697863354698e-07,
|
| 586 |
+
"entropy": 0.6308085203170777,
|
| 587 |
+
"num_tokens": 12030300.0,
|
| 588 |
+
"mean_token_accuracy": 0.8203556030988693,
|
| 589 |
+
"epoch": 0.944,
|
| 590 |
+
"step": 590
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"loss": 0.623,
|
| 594 |
+
"grad_norm": 0.3348216712474823,
|
| 595 |
+
"learning_rate": 5.036401176647332e-07,
|
| 596 |
+
"entropy": 0.6214668571949005,
|
| 597 |
+
"num_tokens": 12226666.0,
|
| 598 |
+
"mean_token_accuracy": 0.8228818953037262,
|
| 599 |
+
"epoch": 0.96,
|
| 600 |
+
"step": 600
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"loss": 0.6305,
|
| 604 |
+
"grad_norm": 0.3399432301521301,
|
| 605 |
+
"learning_rate": 1.9092704065222788e-07,
|
| 606 |
+
"entropy": 0.6307568103075027,
|
| 607 |
+
"num_tokens": 12433050.0,
|
| 608 |
+
"mean_token_accuracy": 0.8195924639701844,
|
| 609 |
+
"epoch": 0.976,
|
| 610 |
+
"step": 610
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"loss": 0.6685,
|
| 614 |
+
"grad_norm": 0.3575892746448517,
|
| 615 |
+
"learning_rate": 2.6863814026734102e-08,
|
| 616 |
+
"entropy": 0.6648932933807373,
|
| 617 |
+
"num_tokens": 12635860.0,
|
| 618 |
+
"mean_token_accuracy": 0.8116440147161483,
|
| 619 |
+
"epoch": 0.992,
|
| 620 |
+
"step": 620
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"train_runtime": 6933.2338,
|
| 624 |
+
"train_samples_per_second": 2.883,
|
| 625 |
+
"train_steps_per_second": 0.09,
|
| 626 |
+
"total_flos": 7.776202046473175e+18,
|
| 627 |
+
"train_loss": 0.7265998420715332,
|
| 628 |
+
"entropy": 0.6022609889507293,
|
| 629 |
+
"num_tokens": 12730954.0,
|
| 630 |
+
"mean_token_accuracy": 0.8287682414054871,
|
| 631 |
+
"epoch": 1.0,
|
| 632 |
+
"step": 625
|
| 633 |
+
}
|
| 634 |
+
]
|