Fathom: upload expert-e8-analyst/training_log.json
Browse files
adapters/expert-e8-analyst/training_log.json
ADDED
|
@@ -0,0 +1,621 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"loss": 1.9181,
|
| 4 |
+
"grad_norm": 0.8695156574249268,
|
| 5 |
+
"learning_rate": 1.8e-05,
|
| 6 |
+
"entropy": 1.457271361351013,
|
| 7 |
+
"num_tokens": 223301.0,
|
| 8 |
+
"mean_token_accuracy": 0.5781058162450791,
|
| 9 |
+
"epoch": 0.016406890894175553,
|
| 10 |
+
"step": 10
|
| 11 |
+
},
|
| 12 |
+
{
|
| 13 |
+
"loss": 1.7798,
|
| 14 |
+
"grad_norm": 0.6769833564758301,
|
| 15 |
+
"learning_rate": 3.8e-05,
|
| 16 |
+
"entropy": 1.6721824526786804,
|
| 17 |
+
"num_tokens": 449106.0,
|
| 18 |
+
"mean_token_accuracy": 0.585400715470314,
|
| 19 |
+
"epoch": 0.03281378178835111,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"loss": 1.5073,
|
| 24 |
+
"grad_norm": 0.903017520904541,
|
| 25 |
+
"learning_rate": 5.8e-05,
|
| 26 |
+
"entropy": 1.568133682012558,
|
| 27 |
+
"num_tokens": 681766.0,
|
| 28 |
+
"mean_token_accuracy": 0.6311331987380981,
|
| 29 |
+
"epoch": 0.04922067268252666,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"loss": 1.1072,
|
| 34 |
+
"grad_norm": 0.4333101511001587,
|
| 35 |
+
"learning_rate": 7.800000000000001e-05,
|
| 36 |
+
"entropy": 1.0811064839363098,
|
| 37 |
+
"num_tokens": 919022.0,
|
| 38 |
+
"mean_token_accuracy": 0.7193854421377182,
|
| 39 |
+
"epoch": 0.06562756357670221,
|
| 40 |
+
"step": 40
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"loss": 0.9973,
|
| 44 |
+
"grad_norm": 0.33554309606552124,
|
| 45 |
+
"learning_rate": 9.8e-05,
|
| 46 |
+
"entropy": 0.9973530560731888,
|
| 47 |
+
"num_tokens": 1148282.0,
|
| 48 |
+
"mean_token_accuracy": 0.7452630162239074,
|
| 49 |
+
"epoch": 0.08203445447087777,
|
| 50 |
+
"step": 50
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"loss": 0.8928,
|
| 54 |
+
"grad_norm": 0.338527649641037,
|
| 55 |
+
"learning_rate": 9.993628283308581e-05,
|
| 56 |
+
"entropy": 0.8918034493923187,
|
| 57 |
+
"num_tokens": 1372211.0,
|
| 58 |
+
"mean_token_accuracy": 0.7673784762620925,
|
| 59 |
+
"epoch": 0.09844134536505332,
|
| 60 |
+
"step": 60
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"loss": 0.8782,
|
| 64 |
+
"grad_norm": 0.32297801971435547,
|
| 65 |
+
"learning_rate": 9.971623444249021e-05,
|
| 66 |
+
"entropy": 0.8743007063865662,
|
| 67 |
+
"num_tokens": 1604226.0,
|
| 68 |
+
"mean_token_accuracy": 0.7686779618263244,
|
| 69 |
+
"epoch": 0.11484823625922888,
|
| 70 |
+
"step": 70
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"loss": 0.8258,
|
| 74 |
+
"grad_norm": 0.33724743127822876,
|
| 75 |
+
"learning_rate": 9.933976038510333e-05,
|
| 76 |
+
"entropy": 0.8295134097337723,
|
| 77 |
+
"num_tokens": 1838081.0,
|
| 78 |
+
"mean_token_accuracy": 0.7803715646266938,
|
| 79 |
+
"epoch": 0.13125512715340443,
|
| 80 |
+
"step": 80
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"loss": 0.7891,
|
| 84 |
+
"grad_norm": 0.3447152078151703,
|
| 85 |
+
"learning_rate": 9.88080451875917e-05,
|
| 86 |
+
"entropy": 0.788675120472908,
|
| 87 |
+
"num_tokens": 2071855.0,
|
| 88 |
+
"mean_token_accuracy": 0.7867679089307785,
|
| 89 |
+
"epoch": 0.14766201804757997,
|
| 90 |
+
"step": 90
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"loss": 0.8272,
|
| 94 |
+
"grad_norm": 0.3010920286178589,
|
| 95 |
+
"learning_rate": 9.812276182268236e-05,
|
| 96 |
+
"entropy": 0.8329822063446045,
|
| 97 |
+
"num_tokens": 2299745.0,
|
| 98 |
+
"mean_token_accuracy": 0.7769433975219726,
|
| 99 |
+
"epoch": 0.16406890894175555,
|
| 100 |
+
"step": 100
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"loss": 0.797,
|
| 104 |
+
"grad_norm": 0.3699570298194885,
|
| 105 |
+
"learning_rate": 9.728606644537178e-05,
|
| 106 |
+
"entropy": 0.8039814531803131,
|
| 107 |
+
"num_tokens": 2528567.0,
|
| 108 |
+
"mean_token_accuracy": 0.784355628490448,
|
| 109 |
+
"epoch": 0.1804757998359311,
|
| 110 |
+
"step": 110
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"loss": 0.7743,
|
| 114 |
+
"grad_norm": 0.3595486581325531,
|
| 115 |
+
"learning_rate": 9.63005916088644e-05,
|
| 116 |
+
"entropy": 0.7768057614564896,
|
| 117 |
+
"num_tokens": 2751275.0,
|
| 118 |
+
"mean_token_accuracy": 0.7889596104621888,
|
| 119 |
+
"epoch": 0.19688269073010664,
|
| 120 |
+
"step": 120
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"loss": 0.7735,
|
| 124 |
+
"grad_norm": 0.3556647002696991,
|
| 125 |
+
"learning_rate": 9.516943798158649e-05,
|
| 126 |
+
"entropy": 0.776089146733284,
|
| 127 |
+
"num_tokens": 2978023.0,
|
| 128 |
+
"mean_token_accuracy": 0.7897650897502899,
|
| 129 |
+
"epoch": 0.2132895816242822,
|
| 130 |
+
"step": 130
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"loss": 0.7622,
|
| 134 |
+
"grad_norm": 0.3334764242172241,
|
| 135 |
+
"learning_rate": 9.389616459133597e-05,
|
| 136 |
+
"entropy": 0.7803491950035095,
|
| 137 |
+
"num_tokens": 3204517.0,
|
| 138 |
+
"mean_token_accuracy": 0.7920954555273056,
|
| 139 |
+
"epoch": 0.22969647251845776,
|
| 140 |
+
"step": 140
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"loss": 0.7576,
|
| 144 |
+
"grad_norm": 0.4009985029697418,
|
| 145 |
+
"learning_rate": 9.248477762726437e-05,
|
| 146 |
+
"entropy": 0.7722930639982224,
|
| 147 |
+
"num_tokens": 3434323.0,
|
| 148 |
+
"mean_token_accuracy": 0.792155721783638,
|
| 149 |
+
"epoch": 0.2461033634126333,
|
| 150 |
+
"step": 150
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"loss": 0.7741,
|
| 154 |
+
"grad_norm": 0.3408016264438629,
|
| 155 |
+
"learning_rate": 9.093971783492355e-05,
|
| 156 |
+
"entropy": 0.7827848076820374,
|
| 157 |
+
"num_tokens": 3661534.0,
|
| 158 |
+
"mean_token_accuracy": 0.7909984678030014,
|
| 159 |
+
"epoch": 0.26251025430680885,
|
| 160 |
+
"step": 160
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"loss": 0.7409,
|
| 164 |
+
"grad_norm": 0.3399813175201416,
|
| 165 |
+
"learning_rate": 8.926584654403724e-05,
|
| 166 |
+
"entropy": 0.7509049952030182,
|
| 167 |
+
"num_tokens": 3890921.0,
|
| 168 |
+
"mean_token_accuracy": 0.7963089287281037,
|
| 169 |
+
"epoch": 0.27891714520098443,
|
| 170 |
+
"step": 170
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"loss": 0.7478,
|
| 174 |
+
"grad_norm": 0.30348479747772217,
|
| 175 |
+
"learning_rate": 8.746843037295937e-05,
|
| 176 |
+
"entropy": 0.7562560260295867,
|
| 177 |
+
"num_tokens": 4119537.0,
|
| 178 |
+
"mean_token_accuracy": 0.794661870598793,
|
| 179 |
+
"epoch": 0.29532403609515995,
|
| 180 |
+
"step": 180
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"loss": 0.737,
|
| 184 |
+
"grad_norm": 0.3403272330760956,
|
| 185 |
+
"learning_rate": 8.555312465794403e-05,
|
| 186 |
+
"entropy": 0.7475169450044632,
|
| 187 |
+
"num_tokens": 4357977.0,
|
| 188 |
+
"mean_token_accuracy": 0.796341797709465,
|
| 189 |
+
"epoch": 0.3117309269893355,
|
| 190 |
+
"step": 190
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"loss": 0.7512,
|
| 194 |
+
"grad_norm": 0.3219321072101593,
|
| 195 |
+
"learning_rate": 8.352595565936554e-05,
|
| 196 |
+
"entropy": 0.7560538798570633,
|
| 197 |
+
"num_tokens": 4581046.0,
|
| 198 |
+
"mean_token_accuracy": 0.7952380329370499,
|
| 199 |
+
"epoch": 0.3281378178835111,
|
| 200 |
+
"step": 200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"loss": 0.7317,
|
| 204 |
+
"grad_norm": 0.3038958013057709,
|
| 205 |
+
"learning_rate": 8.139330160087374e-05,
|
| 206 |
+
"entropy": 0.7322431743144989,
|
| 207 |
+
"num_tokens": 4814227.0,
|
| 208 |
+
"mean_token_accuracy": 0.7974914610385895,
|
| 209 |
+
"epoch": 0.3445447087776866,
|
| 210 |
+
"step": 210
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"loss": 0.7205,
|
| 214 |
+
"grad_norm": 0.32955309748649597,
|
| 215 |
+
"learning_rate": 7.916187260114263e-05,
|
| 216 |
+
"entropy": 0.7275226473808288,
|
| 217 |
+
"num_tokens": 5048157.0,
|
| 218 |
+
"mean_token_accuracy": 0.8014717370271682,
|
| 219 |
+
"epoch": 0.3609515996718622,
|
| 220 |
+
"step": 220
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"loss": 0.7282,
|
| 224 |
+
"grad_norm": 0.4224933683872223,
|
| 225 |
+
"learning_rate": 7.68386895613546e-05,
|
| 226 |
+
"entropy": 0.7310106873512268,
|
| 227 |
+
"num_tokens": 5274071.0,
|
| 228 |
+
"mean_token_accuracy": 0.7992997527122497,
|
| 229 |
+
"epoch": 0.37735849056603776,
|
| 230 |
+
"step": 230
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"loss": 0.6971,
|
| 234 |
+
"grad_norm": 0.33381229639053345,
|
| 235 |
+
"learning_rate": 7.443106207484776e-05,
|
| 236 |
+
"entropy": 0.6995288044214248,
|
| 237 |
+
"num_tokens": 5509360.0,
|
| 238 |
+
"mean_token_accuracy": 0.8060027480125427,
|
| 239 |
+
"epoch": 0.3937653814602133,
|
| 240 |
+
"step": 240
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"loss": 0.711,
|
| 244 |
+
"grad_norm": 0.31851011514663696,
|
| 245 |
+
"learning_rate": 7.194656542843102e-05,
|
| 246 |
+
"entropy": 0.7142476379871369,
|
| 247 |
+
"num_tokens": 5738285.0,
|
| 248 |
+
"mean_token_accuracy": 0.8031993210315704,
|
| 249 |
+
"epoch": 0.41017227235438886,
|
| 250 |
+
"step": 250
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"loss": 0.7237,
|
| 254 |
+
"grad_norm": 0.33157357573509216,
|
| 255 |
+
"learning_rate": 6.939301676772927e-05,
|
| 256 |
+
"entropy": 0.7256091266870499,
|
| 257 |
+
"num_tokens": 5962580.0,
|
| 258 |
+
"mean_token_accuracy": 0.8013624370098114,
|
| 259 |
+
"epoch": 0.4265791632485644,
|
| 260 |
+
"step": 260
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"loss": 0.721,
|
| 264 |
+
"grad_norm": 0.3370811641216278,
|
| 265 |
+
"learning_rate": 6.677845050155107e-05,
|
| 266 |
+
"entropy": 0.7265744864940643,
|
| 267 |
+
"num_tokens": 6196094.0,
|
| 268 |
+
"mean_token_accuracy": 0.7991349190473557,
|
| 269 |
+
"epoch": 0.44298605414273995,
|
| 270 |
+
"step": 270
|
| 271 |
+
},
|
| 272 |
+
{
|
| 273 |
+
"loss": 0.6843,
|
| 274 |
+
"grad_norm": 0.38120874762535095,
|
| 275 |
+
"learning_rate": 6.411109302266616e-05,
|
| 276 |
+
"entropy": 0.6908316820859909,
|
| 277 |
+
"num_tokens": 6420601.0,
|
| 278 |
+
"mean_token_accuracy": 0.8087756901979446,
|
| 279 |
+
"epoch": 0.4593929450369155,
|
| 280 |
+
"step": 280
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"loss": 0.7072,
|
| 284 |
+
"grad_norm": 0.38430851697921753,
|
| 285 |
+
"learning_rate": 6.139933682453036e-05,
|
| 286 |
+
"entropy": 0.7136244118213654,
|
| 287 |
+
"num_tokens": 6655119.0,
|
| 288 |
+
"mean_token_accuracy": 0.8047497570514679,
|
| 289 |
+
"epoch": 0.47579983593109104,
|
| 290 |
+
"step": 290
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"loss": 0.7252,
|
| 294 |
+
"grad_norm": 0.3509667217731476,
|
| 295 |
+
"learning_rate": 5.8651714095396135e-05,
|
| 296 |
+
"entropy": 0.7334865719079972,
|
| 297 |
+
"num_tokens": 6876910.0,
|
| 298 |
+
"mean_token_accuracy": 0.799770200252533,
|
| 299 |
+
"epoch": 0.4922067268252666,
|
| 300 |
+
"step": 300
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"loss": 0.6821,
|
| 304 |
+
"grad_norm": 0.3153151571750641,
|
| 305 |
+
"learning_rate": 5.587686987289189e-05,
|
| 306 |
+
"entropy": 0.6873683601617813,
|
| 307 |
+
"num_tokens": 7112299.0,
|
| 308 |
+
"mean_token_accuracy": 0.8088241666555405,
|
| 309 |
+
"epoch": 0.5086136177194421,
|
| 310 |
+
"step": 310
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"loss": 0.7196,
|
| 314 |
+
"grad_norm": 0.34774187207221985,
|
| 315 |
+
"learning_rate": 5.3083534843535074e-05,
|
| 316 |
+
"entropy": 0.7214434593915939,
|
| 317 |
+
"num_tokens": 7346455.0,
|
| 318 |
+
"mean_token_accuracy": 0.8039845436811447,
|
| 319 |
+
"epoch": 0.5250205086136177,
|
| 320 |
+
"step": 320
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"loss": 0.6638,
|
| 324 |
+
"grad_norm": 0.387768030166626,
|
| 325 |
+
"learning_rate": 5.028049787276249e-05,
|
| 326 |
+
"entropy": 0.6638175457715988,
|
| 327 |
+
"num_tokens": 7571791.0,
|
| 328 |
+
"mean_token_accuracy": 0.812444058060646,
|
| 329 |
+
"epoch": 0.5414273995077933,
|
| 330 |
+
"step": 330
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"loss": 0.6766,
|
| 334 |
+
"grad_norm": 0.3517005741596222,
|
| 335 |
+
"learning_rate": 4.7476578351907954e-05,
|
| 336 |
+
"entropy": 0.6799941658973694,
|
| 337 |
+
"num_tokens": 7801695.0,
|
| 338 |
+
"mean_token_accuracy": 0.811230742931366,
|
| 339 |
+
"epoch": 0.5578342904019689,
|
| 340 |
+
"step": 340
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"loss": 0.6779,
|
| 344 |
+
"grad_norm": 0.32577675580978394,
|
| 345 |
+
"learning_rate": 4.468059844913444e-05,
|
| 346 |
+
"entropy": 0.6814499109983444,
|
| 347 |
+
"num_tokens": 8039821.0,
|
| 348 |
+
"mean_token_accuracy": 0.8104382246732712,
|
| 349 |
+
"epoch": 0.5742411812961444,
|
| 350 |
+
"step": 350
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"loss": 0.6539,
|
| 354 |
+
"grad_norm": 0.35933127999305725,
|
| 355 |
+
"learning_rate": 4.1901355351628945e-05,
|
| 356 |
+
"entropy": 0.6585495263338089,
|
| 357 |
+
"num_tokens": 8273149.0,
|
| 358 |
+
"mean_token_accuracy": 0.8166852772235871,
|
| 359 |
+
"epoch": 0.5906480721903199,
|
| 360 |
+
"step": 360
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"loss": 0.6843,
|
| 364 |
+
"grad_norm": 0.31598055362701416,
|
| 365 |
+
"learning_rate": 3.914759358639719e-05,
|
| 366 |
+
"entropy": 0.6861207246780395,
|
| 367 |
+
"num_tokens": 8503164.0,
|
| 368 |
+
"mean_token_accuracy": 0.8086160510778427,
|
| 369 |
+
"epoch": 0.6070549630844955,
|
| 370 |
+
"step": 370
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"loss": 0.7094,
|
| 374 |
+
"grad_norm": 0.3427006006240845,
|
| 375 |
+
"learning_rate": 3.642797750674629e-05,
|
| 376 |
+
"entropy": 0.7133786290884018,
|
| 377 |
+
"num_tokens": 8726435.0,
|
| 378 |
+
"mean_token_accuracy": 0.8027824640274048,
|
| 379 |
+
"epoch": 0.623461853978671,
|
| 380 |
+
"step": 380
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"loss": 0.6877,
|
| 384 |
+
"grad_norm": 0.34877264499664307,
|
| 385 |
+
"learning_rate": 3.375106403102389e-05,
|
| 386 |
+
"entropy": 0.6881168276071549,
|
| 387 |
+
"num_tokens": 8954291.0,
|
| 388 |
+
"mean_token_accuracy": 0.8073496133089065,
|
| 389 |
+
"epoch": 0.6398687448728466,
|
| 390 |
+
"step": 390
|
| 391 |
+
},
|
| 392 |
+
{
|
| 393 |
+
"loss": 0.6835,
|
| 394 |
+
"grad_norm": 0.3225726783275604,
|
| 395 |
+
"learning_rate": 3.112527571938717e-05,
|
| 396 |
+
"entropy": 0.6862167656421662,
|
| 397 |
+
"num_tokens": 9177163.0,
|
| 398 |
+
"mean_token_accuracy": 0.8089945495128632,
|
| 399 |
+
"epoch": 0.6562756357670222,
|
| 400 |
+
"step": 400
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"loss": 0.7008,
|
| 404 |
+
"grad_norm": 0.329756498336792,
|
| 405 |
+
"learning_rate": 2.8558874273312674e-05,
|
| 406 |
+
"entropy": 0.7071986079216004,
|
| 407 |
+
"num_tokens": 9404151.0,
|
| 408 |
+
"mean_token_accuracy": 0.8044474363327027,
|
| 409 |
+
"epoch": 0.6726825266611977,
|
| 410 |
+
"step": 410
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"loss": 0.6947,
|
| 414 |
+
"grad_norm": 0.3715651035308838,
|
| 415 |
+
"learning_rate": 2.605993454122687e-05,
|
| 416 |
+
"entropy": 0.69432153403759,
|
| 417 |
+
"num_tokens": 9639400.0,
|
| 418 |
+
"mean_token_accuracy": 0.8064981371164321,
|
| 419 |
+
"epoch": 0.6890894175553732,
|
| 420 |
+
"step": 420
|
| 421 |
+
},
|
| 422 |
+
{
|
| 423 |
+
"loss": 0.7066,
|
| 424 |
+
"grad_norm": 0.3599180281162262,
|
| 425 |
+
"learning_rate": 2.3636319112045496e-05,
|
| 426 |
+
"entropy": 0.7111173301935196,
|
| 427 |
+
"num_tokens": 9867668.0,
|
| 428 |
+
"mean_token_accuracy": 0.8044642627239227,
|
| 429 |
+
"epoch": 0.7054963084495488,
|
| 430 |
+
"step": 430
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"loss": 0.7259,
|
| 434 |
+
"grad_norm": 0.2912443280220032,
|
| 435 |
+
"learning_rate": 2.1295653576560163e-05,
|
| 436 |
+
"entropy": 0.7254415988922119,
|
| 437 |
+
"num_tokens": 10100826.0,
|
| 438 |
+
"mean_token_accuracy": 0.8003069430589675,
|
| 439 |
+
"epoch": 0.7219031993437244,
|
| 440 |
+
"step": 440
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"loss": 0.6761,
|
| 444 |
+
"grad_norm": 0.30693626403808594,
|
| 445 |
+
"learning_rate": 1.9045302534508297e-05,
|
| 446 |
+
"entropy": 0.6833124309778214,
|
| 447 |
+
"num_tokens": 10332359.0,
|
| 448 |
+
"mean_token_accuracy": 0.8109049916267395,
|
| 449 |
+
"epoch": 0.7383100902379,
|
| 450 |
+
"step": 450
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"loss": 0.736,
|
| 454 |
+
"grad_norm": 0.3155220150947571,
|
| 455 |
+
"learning_rate": 1.6892346422817946e-05,
|
| 456 |
+
"entropy": 0.736938726902008,
|
| 457 |
+
"num_tokens": 10563841.0,
|
| 458 |
+
"mean_token_accuracy": 0.7979681819677353,
|
| 459 |
+
"epoch": 0.7547169811320755,
|
| 460 |
+
"step": 460
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"loss": 0.6945,
|
| 464 |
+
"grad_norm": 0.3748078942298889,
|
| 465 |
+
"learning_rate": 1.4843559237933473e-05,
|
| 466 |
+
"entropy": 0.7031238079071045,
|
| 467 |
+
"num_tokens": 10788876.0,
|
| 468 |
+
"mean_token_accuracy": 0.8057133972644805,
|
| 469 |
+
"epoch": 0.771123872026251,
|
| 470 |
+
"step": 470
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"loss": 0.6776,
|
| 474 |
+
"grad_norm": 0.3635546565055847,
|
| 475 |
+
"learning_rate": 1.2905387222316822e-05,
|
| 476 |
+
"entropy": 0.6805126667022705,
|
| 477 |
+
"num_tokens": 11015156.0,
|
| 478 |
+
"mean_token_accuracy": 0.8101104766130447,
|
| 479 |
+
"epoch": 0.7875307629204266,
|
| 480 |
+
"step": 480
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"loss": 0.676,
|
| 484 |
+
"grad_norm": 0.3111382722854614,
|
| 485 |
+
"learning_rate": 1.1083928582183711e-05,
|
| 486 |
+
"entropy": 0.6774959295988083,
|
| 487 |
+
"num_tokens": 11245860.0,
|
| 488 |
+
"mean_token_accuracy": 0.8107922226190567,
|
| 489 |
+
"epoch": 0.8039376538146021,
|
| 490 |
+
"step": 490
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"loss": 0.6742,
|
| 494 |
+
"grad_norm": 0.32188844680786133,
|
| 495 |
+
"learning_rate": 9.384914300290748e-06,
|
| 496 |
+
"entropy": 0.6842435419559478,
|
| 497 |
+
"num_tokens": 11476241.0,
|
| 498 |
+
"mean_token_accuracy": 0.8111602008342743,
|
| 499 |
+
"epoch": 0.8203445447087777,
|
| 500 |
+
"step": 500
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"loss": 0.6544,
|
| 504 |
+
"grad_norm": 0.36185422539711,
|
| 505 |
+
"learning_rate": 7.813690104143557e-06,
|
| 506 |
+
"entropy": 0.6514311820268631,
|
| 507 |
+
"num_tokens": 11708112.0,
|
| 508 |
+
"mean_token_accuracy": 0.8149820327758789,
|
| 509 |
+
"epoch": 0.8367514356029533,
|
| 510 |
+
"step": 510
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"loss": 0.6765,
|
| 514 |
+
"grad_norm": 0.3183876574039459,
|
| 515 |
+
"learning_rate": 6.375199646360142e-06,
|
| 516 |
+
"entropy": 0.6856429934501648,
|
| 517 |
+
"num_tokens": 11939337.0,
|
| 518 |
+
"mean_token_accuracy": 0.8090052843093872,
|
| 519 |
+
"epoch": 0.8531583264971287,
|
| 520 |
+
"step": 520
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"loss": 0.6761,
|
| 524 |
+
"grad_norm": 0.3287002742290497,
|
| 525 |
+
"learning_rate": 5.073968950110941e-06,
|
| 526 |
+
"entropy": 0.6834310472011567,
|
| 527 |
+
"num_tokens": 12174723.0,
|
| 528 |
+
"mean_token_accuracy": 0.8104397505521774,
|
| 529 |
+
"epoch": 0.8695652173913043,
|
| 530 |
+
"step": 530
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"loss": 0.6751,
|
| 534 |
+
"grad_norm": 0.35229238867759705,
|
| 535 |
+
"learning_rate": 3.914092168575306e-06,
|
| 536 |
+
"entropy": 0.6824660181999207,
|
| 537 |
+
"num_tokens": 12398555.0,
|
| 538 |
+
"mean_token_accuracy": 0.8104325562715531,
|
| 539 |
+
"epoch": 0.8859721082854799,
|
| 540 |
+
"step": 540
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"loss": 0.6834,
|
| 544 |
+
"grad_norm": 0.38912639021873474,
|
| 545 |
+
"learning_rate": 2.8992187032210518e-06,
|
| 546 |
+
"entropy": 0.682240468263626,
|
| 547 |
+
"num_tokens": 12624846.0,
|
| 548 |
+
"mean_token_accuracy": 0.8091065347194671,
|
| 549 |
+
"epoch": 0.9023789991796555,
|
| 550 |
+
"step": 550
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"loss": 0.696,
|
| 554 |
+
"grad_norm": 0.306355744600296,
|
| 555 |
+
"learning_rate": 2.032541721437209e-06,
|
| 556 |
+
"entropy": 0.7058492481708527,
|
| 557 |
+
"num_tokens": 12859015.0,
|
| 558 |
+
"mean_token_accuracy": 0.8039765357971191,
|
| 559 |
+
"epoch": 0.918785890073831,
|
| 560 |
+
"step": 560
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"loss": 0.6727,
|
| 564 |
+
"grad_norm": 0.38508960604667664,
|
| 565 |
+
"learning_rate": 1.3167881096480372e-06,
|
| 566 |
+
"entropy": 0.681548210978508,
|
| 567 |
+
"num_tokens": 13083551.0,
|
| 568 |
+
"mean_token_accuracy": 0.8100948423147202,
|
| 569 |
+
"epoch": 0.9351927809680065,
|
| 570 |
+
"step": 570
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"loss": 0.7208,
|
| 574 |
+
"grad_norm": 0.33893731236457825,
|
| 575 |
+
"learning_rate": 7.542098935195918e-07,
|
| 576 |
+
"entropy": 0.7220237284898758,
|
| 577 |
+
"num_tokens": 13308857.0,
|
| 578 |
+
"mean_token_accuracy": 0.8005945891141891,
|
| 579 |
+
"epoch": 0.9515996718621821,
|
| 580 |
+
"step": 580
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"loss": 0.6759,
|
| 584 |
+
"grad_norm": 0.3534739911556244,
|
| 585 |
+
"learning_rate": 3.465771522536854e-07,
|
| 586 |
+
"entropy": 0.6739370882511139,
|
| 587 |
+
"num_tokens": 13543857.0,
|
| 588 |
+
"mean_token_accuracy": 0.8097480118274689,
|
| 589 |
+
"epoch": 0.9680065627563577,
|
| 590 |
+
"step": 590
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"loss": 0.6865,
|
| 594 |
+
"grad_norm": 0.3553875982761383,
|
| 595 |
+
"learning_rate": 9.517244926393609e-08,
|
| 596 |
+
"entropy": 0.6908959478139878,
|
| 597 |
+
"num_tokens": 13769574.0,
|
| 598 |
+
"mean_token_accuracy": 0.806584045290947,
|
| 599 |
+
"epoch": 0.9844134536505332,
|
| 600 |
+
"step": 600
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"loss": 0.6525,
|
| 604 |
+
"grad_norm": 0.5078703761100769,
|
| 605 |
+
"learning_rate": 7.867967567354306e-10,
|
| 606 |
+
"entropy": 0.6598060852602908,
|
| 607 |
+
"num_tokens": 13978118.0,
|
| 608 |
+
"mean_token_accuracy": 0.8165042933664823,
|
| 609 |
+
"epoch": 1.0,
|
| 610 |
+
"step": 610
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"train_runtime": 6449.4338,
|
| 614 |
+
"train_samples_per_second": 3.024,
|
| 615 |
+
"train_steps_per_second": 0.095,
|
| 616 |
+
"total_flos": 7.156995496917074e+18,
|
| 617 |
+
"train_loss": 0.7796625786140317,
|
| 618 |
+
"epoch": 1.0,
|
| 619 |
+
"step": 610
|
| 620 |
+
}
|
| 621 |
+
]
|