Copy heitorrosa/logun-base/sft/checkpoint-1350/trainer_state.json to heitorrosa/logun-base/sft-0.5/checkpoint-1350/trainer_state.json
Browse files
sft-0.5/checkpoint-1350/trainer_state.json
ADDED
|
@@ -0,0 +1,754 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.5307626878366885,
|
| 6 |
+
"eval_steps": 150,
|
| 7 |
+
"global_step": 1350,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.01701162461015027,
|
| 14 |
+
"grad_norm": 116.99359893798828,
|
| 15 |
+
"learning_rate": 4.666666666666667e-06,
|
| 16 |
+
"loss": 39.077762858072916,
|
| 17 |
+
"step": 15
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.03402324922030054,
|
| 21 |
+
"grad_norm": 114.67162322998047,
|
| 22 |
+
"learning_rate": 9.666666666666667e-06,
|
| 23 |
+
"loss": 37.629154459635416,
|
| 24 |
+
"step": 30
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.05103487383045081,
|
| 28 |
+
"grad_norm": 78.71202850341797,
|
| 29 |
+
"learning_rate": 1.4666666666666668e-05,
|
| 30 |
+
"loss": 36.40402018229167,
|
| 31 |
+
"step": 45
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.06804649844060108,
|
| 35 |
+
"grad_norm": 77.94384002685547,
|
| 36 |
+
"learning_rate": 1.9666666666666666e-05,
|
| 37 |
+
"loss": 35.03992513020833,
|
| 38 |
+
"step": 60
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.08505812305075135,
|
| 42 |
+
"grad_norm": 69.6011734008789,
|
| 43 |
+
"learning_rate": 2.466666666666667e-05,
|
| 44 |
+
"loss": 34.81482747395833,
|
| 45 |
+
"step": 75
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.10206974766090161,
|
| 49 |
+
"grad_norm": 74.6609878540039,
|
| 50 |
+
"learning_rate": 2.9666666666666672e-05,
|
| 51 |
+
"loss": 33.83201497395833,
|
| 52 |
+
"step": 90
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.11908137227105188,
|
| 56 |
+
"grad_norm": 73.47967529296875,
|
| 57 |
+
"learning_rate": 3.466666666666667e-05,
|
| 58 |
+
"loss": 34.1611328125,
|
| 59 |
+
"step": 105
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.13609299688120216,
|
| 63 |
+
"grad_norm": 57.3227424621582,
|
| 64 |
+
"learning_rate": 3.966666666666667e-05,
|
| 65 |
+
"loss": 33.17495524088542,
|
| 66 |
+
"step": 120
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.15310462149135243,
|
| 70 |
+
"grad_norm": 41.92803955078125,
|
| 71 |
+
"learning_rate": 4.466666666666667e-05,
|
| 72 |
+
"loss": 33.14850056966146,
|
| 73 |
+
"step": 135
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.1701162461015027,
|
| 77 |
+
"grad_norm": 57.70565414428711,
|
| 78 |
+
"learning_rate": 4.966666666666667e-05,
|
| 79 |
+
"loss": 33.152496337890625,
|
| 80 |
+
"step": 150
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.1701162461015027,
|
| 84 |
+
"eval_accuracy": 0.519298245614035,
|
| 85 |
+
"eval_loss": 1.9665578603744507,
|
| 86 |
+
"eval_macro_f1": 0.43657905964809113,
|
| 87 |
+
"eval_runtime": 7.6538,
|
| 88 |
+
"eval_samples_per_second": 74.472,
|
| 89 |
+
"eval_steps_per_second": 9.407,
|
| 90 |
+
"step": 150
|
| 91 |
+
},
|
| 92 |
+
{
|
| 93 |
+
"epoch": 0.18712787071165296,
|
| 94 |
+
"grad_norm": 63.17024612426758,
|
| 95 |
+
"learning_rate": 4.940323955669225e-05,
|
| 96 |
+
"loss": 31.66871541341146,
|
| 97 |
+
"step": 165
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.20413949532180323,
|
| 101 |
+
"grad_norm": 73.78040313720703,
|
| 102 |
+
"learning_rate": 4.8763853367433935e-05,
|
| 103 |
+
"loss": 31.002229817708333,
|
| 104 |
+
"step": 180
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.2211511199319535,
|
| 108 |
+
"grad_norm": 46.99962615966797,
|
| 109 |
+
"learning_rate": 4.812446717817562e-05,
|
| 110 |
+
"loss": 31.49439900716146,
|
| 111 |
+
"step": 195
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 0.23816274454210376,
|
| 115 |
+
"grad_norm": 59.367454528808594,
|
| 116 |
+
"learning_rate": 4.748508098891731e-05,
|
| 117 |
+
"loss": 30.110245768229166,
|
| 118 |
+
"step": 210
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 0.25517436915225405,
|
| 122 |
+
"grad_norm": 65.87191009521484,
|
| 123 |
+
"learning_rate": 4.6845694799659e-05,
|
| 124 |
+
"loss": 31.120613606770835,
|
| 125 |
+
"step": 225
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 0.2721859937624043,
|
| 129 |
+
"grad_norm": 54.601192474365234,
|
| 130 |
+
"learning_rate": 4.620630861040069e-05,
|
| 131 |
+
"loss": 30.689495849609376,
|
| 132 |
+
"step": 240
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 0.2891976183725546,
|
| 136 |
+
"grad_norm": 85.62554931640625,
|
| 137 |
+
"learning_rate": 4.5566922421142375e-05,
|
| 138 |
+
"loss": 28.9639892578125,
|
| 139 |
+
"step": 255
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 0.30620924298270485,
|
| 143 |
+
"grad_norm": 55.505699157714844,
|
| 144 |
+
"learning_rate": 4.492753623188406e-05,
|
| 145 |
+
"loss": 29.698612467447916,
|
| 146 |
+
"step": 270
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.3232208675928551,
|
| 150 |
+
"grad_norm": 48.24406051635742,
|
| 151 |
+
"learning_rate": 4.428815004262575e-05,
|
| 152 |
+
"loss": 28.051249186197918,
|
| 153 |
+
"step": 285
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.3402324922030054,
|
| 157 |
+
"grad_norm": 83.22586822509766,
|
| 158 |
+
"learning_rate": 4.364876385336744e-05,
|
| 159 |
+
"loss": 27.165751139322918,
|
| 160 |
+
"step": 300
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.3402324922030054,
|
| 164 |
+
"eval_accuracy": 0.6140350877192983,
|
| 165 |
+
"eval_loss": 1.7097147703170776,
|
| 166 |
+
"eval_macro_f1": 0.5762605966740592,
|
| 167 |
+
"eval_runtime": 3.4179,
|
| 168 |
+
"eval_samples_per_second": 166.769,
|
| 169 |
+
"eval_steps_per_second": 21.066,
|
| 170 |
+
"step": 300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.35724411681315565,
|
| 174 |
+
"grad_norm": 60.6512451171875,
|
| 175 |
+
"learning_rate": 4.300937766410913e-05,
|
| 176 |
+
"loss": 27.327327473958334,
|
| 177 |
+
"step": 315
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.3742557414233059,
|
| 181 |
+
"grad_norm": 75.39328002929688,
|
| 182 |
+
"learning_rate": 4.2369991474850815e-05,
|
| 183 |
+
"loss": 28.38462931315104,
|
| 184 |
+
"step": 330
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.3912673660334562,
|
| 188 |
+
"grad_norm": 60.53544235229492,
|
| 189 |
+
"learning_rate": 4.17306052855925e-05,
|
| 190 |
+
"loss": 28.571195475260417,
|
| 191 |
+
"step": 345
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.40827899064360645,
|
| 195 |
+
"grad_norm": 66.51234436035156,
|
| 196 |
+
"learning_rate": 4.109121909633419e-05,
|
| 197 |
+
"loss": 27.315740966796874,
|
| 198 |
+
"step": 360
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.4252906152537567,
|
| 202 |
+
"grad_norm": 71.48542785644531,
|
| 203 |
+
"learning_rate": 4.045183290707588e-05,
|
| 204 |
+
"loss": 26.424430338541665,
|
| 205 |
+
"step": 375
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.442302239863907,
|
| 209 |
+
"grad_norm": 65.96041107177734,
|
| 210 |
+
"learning_rate": 3.981244671781757e-05,
|
| 211 |
+
"loss": 26.818721516927084,
|
| 212 |
+
"step": 390
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.45931386447405725,
|
| 216 |
+
"grad_norm": 96.44197082519531,
|
| 217 |
+
"learning_rate": 3.9173060528559255e-05,
|
| 218 |
+
"loss": 27.456418863932292,
|
| 219 |
+
"step": 405
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.4763254890842075,
|
| 223 |
+
"grad_norm": 74.46075439453125,
|
| 224 |
+
"learning_rate": 3.8533674339300943e-05,
|
| 225 |
+
"loss": 27.342032877604165,
|
| 226 |
+
"step": 420
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.4933371136943578,
|
| 230 |
+
"grad_norm": 98.2132339477539,
|
| 231 |
+
"learning_rate": 3.789428815004263e-05,
|
| 232 |
+
"loss": 27.26170654296875,
|
| 233 |
+
"step": 435
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.5103487383045081,
|
| 237 |
+
"grad_norm": 66.33806610107422,
|
| 238 |
+
"learning_rate": 3.725490196078432e-05,
|
| 239 |
+
"loss": 26.62706298828125,
|
| 240 |
+
"step": 450
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.5103487383045081,
|
| 244 |
+
"eval_accuracy": 0.6473684210526316,
|
| 245 |
+
"eval_loss": 1.612095832824707,
|
| 246 |
+
"eval_macro_f1": 0.6036091764418378,
|
| 247 |
+
"eval_runtime": 3.4226,
|
| 248 |
+
"eval_samples_per_second": 166.541,
|
| 249 |
+
"eval_steps_per_second": 21.037,
|
| 250 |
+
"step": 450
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.5273603629146584,
|
| 254 |
+
"grad_norm": 59.41883087158203,
|
| 255 |
+
"learning_rate": 3.661551577152601e-05,
|
| 256 |
+
"loss": 25.6656494140625,
|
| 257 |
+
"step": 465
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.5443719875248086,
|
| 261 |
+
"grad_norm": 86.57988739013672,
|
| 262 |
+
"learning_rate": 3.5976129582267696e-05,
|
| 263 |
+
"loss": 25.688761393229168,
|
| 264 |
+
"step": 480
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 0.5613836121349589,
|
| 268 |
+
"grad_norm": 87.54814147949219,
|
| 269 |
+
"learning_rate": 3.5336743393009384e-05,
|
| 270 |
+
"loss": 26.806388346354165,
|
| 271 |
+
"step": 495
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.5783952367451092,
|
| 275 |
+
"grad_norm": 65.40018463134766,
|
| 276 |
+
"learning_rate": 3.469735720375107e-05,
|
| 277 |
+
"loss": 25.533500162760415,
|
| 278 |
+
"step": 510
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.5954068613552594,
|
| 282 |
+
"grad_norm": 65.12799072265625,
|
| 283 |
+
"learning_rate": 3.405797101449276e-05,
|
| 284 |
+
"loss": 26.228350830078124,
|
| 285 |
+
"step": 525
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 0.6124184859654097,
|
| 289 |
+
"grad_norm": 65.55587768554688,
|
| 290 |
+
"learning_rate": 3.341858482523445e-05,
|
| 291 |
+
"loss": 26.634810384114584,
|
| 292 |
+
"step": 540
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.62943011057556,
|
| 296 |
+
"grad_norm": 68.18641662597656,
|
| 297 |
+
"learning_rate": 3.2779198635976136e-05,
|
| 298 |
+
"loss": 25.533231608072917,
|
| 299 |
+
"step": 555
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 0.6464417351857102,
|
| 303 |
+
"grad_norm": 80.25837707519531,
|
| 304 |
+
"learning_rate": 3.2139812446717824e-05,
|
| 305 |
+
"loss": 24.925142415364583,
|
| 306 |
+
"step": 570
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.6634533597958605,
|
| 310 |
+
"grad_norm": 58.9185905456543,
|
| 311 |
+
"learning_rate": 3.150042625745951e-05,
|
| 312 |
+
"loss": 24.743680826822917,
|
| 313 |
+
"step": 585
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.6804649844060108,
|
| 317 |
+
"grad_norm": 59.13557815551758,
|
| 318 |
+
"learning_rate": 3.08610400682012e-05,
|
| 319 |
+
"loss": 25.899686686197917,
|
| 320 |
+
"step": 600
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.6804649844060108,
|
| 324 |
+
"eval_accuracy": 0.6421052631578947,
|
| 325 |
+
"eval_loss": 1.5752904415130615,
|
| 326 |
+
"eval_macro_f1": 0.6148540921288014,
|
| 327 |
+
"eval_runtime": 3.4091,
|
| 328 |
+
"eval_samples_per_second": 167.201,
|
| 329 |
+
"eval_steps_per_second": 21.12,
|
| 330 |
+
"step": 600
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 0.697476609016161,
|
| 334 |
+
"grad_norm": 63.18887710571289,
|
| 335 |
+
"learning_rate": 3.0221653878942884e-05,
|
| 336 |
+
"loss": 25.722517903645834,
|
| 337 |
+
"step": 615
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 0.7144882336263113,
|
| 341 |
+
"grad_norm": 81.7510757446289,
|
| 342 |
+
"learning_rate": 2.9582267689684573e-05,
|
| 343 |
+
"loss": 24.722237141927085,
|
| 344 |
+
"step": 630
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.7314998582364616,
|
| 348 |
+
"grad_norm": 59.331199645996094,
|
| 349 |
+
"learning_rate": 2.894288150042626e-05,
|
| 350 |
+
"loss": 25.852589925130207,
|
| 351 |
+
"step": 645
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 0.7485114828466118,
|
| 355 |
+
"grad_norm": 82.33928680419922,
|
| 356 |
+
"learning_rate": 2.830349531116795e-05,
|
| 357 |
+
"loss": 25.581608072916666,
|
| 358 |
+
"step": 660
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 0.7655231074567621,
|
| 362 |
+
"grad_norm": 73.93915557861328,
|
| 363 |
+
"learning_rate": 2.7664109121909637e-05,
|
| 364 |
+
"loss": 24.140205891927085,
|
| 365 |
+
"step": 675
|
| 366 |
+
},
|
| 367 |
+
{
|
| 368 |
+
"epoch": 0.7825347320669124,
|
| 369 |
+
"grad_norm": 88.01911926269531,
|
| 370 |
+
"learning_rate": 2.7024722932651325e-05,
|
| 371 |
+
"loss": 25.460465494791666,
|
| 372 |
+
"step": 690
|
| 373 |
+
},
|
| 374 |
+
{
|
| 375 |
+
"epoch": 0.7995463566770626,
|
| 376 |
+
"grad_norm": 74.82182312011719,
|
| 377 |
+
"learning_rate": 2.6385336743393013e-05,
|
| 378 |
+
"loss": 25.53566691080729,
|
| 379 |
+
"step": 705
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"epoch": 0.8165579812872129,
|
| 383 |
+
"grad_norm": 53.480384826660156,
|
| 384 |
+
"learning_rate": 2.57459505541347e-05,
|
| 385 |
+
"loss": 24.06074422200521,
|
| 386 |
+
"step": 720
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.8335696058973632,
|
| 390 |
+
"grad_norm": 62.765567779541016,
|
| 391 |
+
"learning_rate": 2.510656436487639e-05,
|
| 392 |
+
"loss": 25.66954142252604,
|
| 393 |
+
"step": 735
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.8505812305075134,
|
| 397 |
+
"grad_norm": 61.93996047973633,
|
| 398 |
+
"learning_rate": 2.4467178175618073e-05,
|
| 399 |
+
"loss": 25.64231160481771,
|
| 400 |
+
"step": 750
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 0.8505812305075134,
|
| 404 |
+
"eval_accuracy": 0.6666666666666666,
|
| 405 |
+
"eval_loss": 1.4923547506332397,
|
| 406 |
+
"eval_macro_f1": 0.634289650129149,
|
| 407 |
+
"eval_runtime": 3.4068,
|
| 408 |
+
"eval_samples_per_second": 167.313,
|
| 409 |
+
"eval_steps_per_second": 21.134,
|
| 410 |
+
"step": 750
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.8675928551176637,
|
| 414 |
+
"grad_norm": 88.9131851196289,
|
| 415 |
+
"learning_rate": 2.382779198635976e-05,
|
| 416 |
+
"loss": 25.273152669270832,
|
| 417 |
+
"step": 765
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.884604479727814,
|
| 421 |
+
"grad_norm": 103.49991607666016,
|
| 422 |
+
"learning_rate": 2.318840579710145e-05,
|
| 423 |
+
"loss": 25.633394368489583,
|
| 424 |
+
"step": 780
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 0.9016161043379642,
|
| 428 |
+
"grad_norm": 71.47494506835938,
|
| 429 |
+
"learning_rate": 2.2549019607843138e-05,
|
| 430 |
+
"loss": 24.569466145833335,
|
| 431 |
+
"step": 795
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 0.9186277289481145,
|
| 435 |
+
"grad_norm": 84.61474609375,
|
| 436 |
+
"learning_rate": 2.1909633418584826e-05,
|
| 437 |
+
"loss": 24.982352701822915,
|
| 438 |
+
"step": 810
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 0.9356393535582648,
|
| 442 |
+
"grad_norm": 106.77224731445312,
|
| 443 |
+
"learning_rate": 2.1270247229326514e-05,
|
| 444 |
+
"loss": 23.428892008463542,
|
| 445 |
+
"step": 825
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 0.952650978168415,
|
| 449 |
+
"grad_norm": 64.02266693115234,
|
| 450 |
+
"learning_rate": 2.06308610400682e-05,
|
| 451 |
+
"loss": 23.625968424479165,
|
| 452 |
+
"step": 840
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.9696626027785653,
|
| 456 |
+
"grad_norm": 83.47174835205078,
|
| 457 |
+
"learning_rate": 1.999147485080989e-05,
|
| 458 |
+
"loss": 24.876424153645832,
|
| 459 |
+
"step": 855
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 0.9866742273887156,
|
| 463 |
+
"grad_norm": 68.72810363769531,
|
| 464 |
+
"learning_rate": 1.9352088661551578e-05,
|
| 465 |
+
"loss": 23.115681966145832,
|
| 466 |
+
"step": 870
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.00340232492203,
|
| 470 |
+
"grad_norm": 52.63280487060547,
|
| 471 |
+
"learning_rate": 1.8712702472293266e-05,
|
| 472 |
+
"loss": 23.462247721354167,
|
| 473 |
+
"step": 885
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.0204139495321802,
|
| 477 |
+
"grad_norm": 54.332237243652344,
|
| 478 |
+
"learning_rate": 1.8073316283034954e-05,
|
| 479 |
+
"loss": 25.237931315104166,
|
| 480 |
+
"step": 900
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.0204139495321802,
|
| 484 |
+
"eval_accuracy": 0.6631578947368421,
|
| 485 |
+
"eval_loss": 1.4844019412994385,
|
| 486 |
+
"eval_macro_f1": 0.6347159169539845,
|
| 487 |
+
"eval_runtime": 3.3664,
|
| 488 |
+
"eval_samples_per_second": 169.319,
|
| 489 |
+
"eval_steps_per_second": 21.388,
|
| 490 |
+
"step": 900
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 1.0374255741423306,
|
| 494 |
+
"grad_norm": 76.4985580444336,
|
| 495 |
+
"learning_rate": 1.7433930093776642e-05,
|
| 496 |
+
"loss": 25.129193115234376,
|
| 497 |
+
"step": 915
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 1.0544371987524808,
|
| 501 |
+
"grad_norm": 63.44844436645508,
|
| 502 |
+
"learning_rate": 1.679454390451833e-05,
|
| 503 |
+
"loss": 25.12100626627604,
|
| 504 |
+
"step": 930
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 1.0714488233626311,
|
| 508 |
+
"grad_norm": 85.40037536621094,
|
| 509 |
+
"learning_rate": 1.6155157715260018e-05,
|
| 510 |
+
"loss": 24.0731689453125,
|
| 511 |
+
"step": 945
|
| 512 |
+
},
|
| 513 |
+
{
|
| 514 |
+
"epoch": 1.0884604479727813,
|
| 515 |
+
"grad_norm": 61.12846755981445,
|
| 516 |
+
"learning_rate": 1.5515771526001706e-05,
|
| 517 |
+
"loss": 23.815816243489582,
|
| 518 |
+
"step": 960
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 1.1054720725829317,
|
| 522 |
+
"grad_norm": 88.95614624023438,
|
| 523 |
+
"learning_rate": 1.4876385336743392e-05,
|
| 524 |
+
"loss": 23.149064127604166,
|
| 525 |
+
"step": 975
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 1.1224836971930818,
|
| 529 |
+
"grad_norm": 81.11243438720703,
|
| 530 |
+
"learning_rate": 1.423699914748508e-05,
|
| 531 |
+
"loss": 24.921358235677083,
|
| 532 |
+
"step": 990
|
| 533 |
+
},
|
| 534 |
+
{
|
| 535 |
+
"epoch": 1.1394953218032322,
|
| 536 |
+
"grad_norm": 89.6662826538086,
|
| 537 |
+
"learning_rate": 1.3597612958226768e-05,
|
| 538 |
+
"loss": 24.148970540364584,
|
| 539 |
+
"step": 1005
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 1.1565069464133826,
|
| 543 |
+
"grad_norm": 76.58106231689453,
|
| 544 |
+
"learning_rate": 1.2958226768968456e-05,
|
| 545 |
+
"loss": 23.3828125,
|
| 546 |
+
"step": 1020
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 1.1735185710235327,
|
| 550 |
+
"grad_norm": 57.90241622924805,
|
| 551 |
+
"learning_rate": 1.2318840579710146e-05,
|
| 552 |
+
"loss": 23.80557861328125,
|
| 553 |
+
"step": 1035
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 1.190530195633683,
|
| 557 |
+
"grad_norm": 78.67855072021484,
|
| 558 |
+
"learning_rate": 1.1679454390451834e-05,
|
| 559 |
+
"loss": 25.015262858072916,
|
| 560 |
+
"step": 1050
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 1.190530195633683,
|
| 564 |
+
"eval_accuracy": 0.6771929824561403,
|
| 565 |
+
"eval_loss": 1.4616774320602417,
|
| 566 |
+
"eval_macro_f1": 0.6469672228058898,
|
| 567 |
+
"eval_runtime": 3.3981,
|
| 568 |
+
"eval_samples_per_second": 167.739,
|
| 569 |
+
"eval_steps_per_second": 21.188,
|
| 570 |
+
"step": 1050
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.2075418202438333,
|
| 574 |
+
"grad_norm": NaN,
|
| 575 |
+
"learning_rate": 1.104006820119352e-05,
|
| 576 |
+
"loss": 24.526371256510416,
|
| 577 |
+
"step": 1065
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.2245534448539837,
|
| 581 |
+
"grad_norm": 94.34043884277344,
|
| 582 |
+
"learning_rate": 1.0400682011935209e-05,
|
| 583 |
+
"loss": 24.23281453450521,
|
| 584 |
+
"step": 1080
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.2415650694641338,
|
| 588 |
+
"grad_norm": 84.55120086669922,
|
| 589 |
+
"learning_rate": 9.761295822676897e-06,
|
| 590 |
+
"loss": 23.37388916015625,
|
| 591 |
+
"step": 1095
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.258576694074284,
|
| 595 |
+
"grad_norm": 98.98963928222656,
|
| 596 |
+
"learning_rate": 9.121909633418585e-06,
|
| 597 |
+
"loss": 24.879107666015624,
|
| 598 |
+
"step": 1110
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.2755883186844343,
|
| 602 |
+
"grad_norm": 93.77626037597656,
|
| 603 |
+
"learning_rate": 8.482523444160273e-06,
|
| 604 |
+
"loss": 23.596321614583335,
|
| 605 |
+
"step": 1125
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.2925999432945847,
|
| 609 |
+
"grad_norm": 76.52169036865234,
|
| 610 |
+
"learning_rate": 7.84313725490196e-06,
|
| 611 |
+
"loss": 23.6485595703125,
|
| 612 |
+
"step": 1140
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.3096115679047349,
|
| 616 |
+
"grad_norm": 64.10005950927734,
|
| 617 |
+
"learning_rate": 7.203751065643649e-06,
|
| 618 |
+
"loss": 23.198685709635416,
|
| 619 |
+
"step": 1155
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.3266231925148853,
|
| 623 |
+
"grad_norm": 61.74102783203125,
|
| 624 |
+
"learning_rate": 6.564364876385337e-06,
|
| 625 |
+
"loss": 22.906901041666668,
|
| 626 |
+
"step": 1170
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.3436348171250354,
|
| 630 |
+
"grad_norm": 68.70774841308594,
|
| 631 |
+
"learning_rate": 5.924978687127026e-06,
|
| 632 |
+
"loss": 23.642818196614584,
|
| 633 |
+
"step": 1185
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.3606464417351858,
|
| 637 |
+
"grad_norm": 105.86067199707031,
|
| 638 |
+
"learning_rate": 5.285592497868714e-06,
|
| 639 |
+
"loss": 23.631290690104166,
|
| 640 |
+
"step": 1200
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.3606464417351858,
|
| 644 |
+
"eval_accuracy": 0.6684210526315789,
|
| 645 |
+
"eval_loss": 1.4395803213119507,
|
| 646 |
+
"eval_macro_f1": 0.6398363484455968,
|
| 647 |
+
"eval_runtime": 3.3938,
|
| 648 |
+
"eval_samples_per_second": 167.951,
|
| 649 |
+
"eval_steps_per_second": 21.215,
|
| 650 |
+
"step": 1200
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 1.377658066345336,
|
| 654 |
+
"grad_norm": 74.67951965332031,
|
| 655 |
+
"learning_rate": 1.7038413878562578e-05,
|
| 656 |
+
"loss": 23.69179484049479,
|
| 657 |
+
"step": 1215
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 1.3946696909554863,
|
| 661 |
+
"grad_norm": 66.90929412841797,
|
| 662 |
+
"learning_rate": 1.657372986369269e-05,
|
| 663 |
+
"loss": 24.378173828125,
|
| 664 |
+
"step": 1230
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"epoch": 1.4116813155656365,
|
| 668 |
+
"grad_norm": 72.71094512939453,
|
| 669 |
+
"learning_rate": 1.61090458488228e-05,
|
| 670 |
+
"loss": 25.47337443033854,
|
| 671 |
+
"step": 1245
|
| 672 |
+
},
|
| 673 |
+
{
|
| 674 |
+
"epoch": 1.4286929401757869,
|
| 675 |
+
"grad_norm": 82.03230285644531,
|
| 676 |
+
"learning_rate": 1.5644361833952914e-05,
|
| 677 |
+
"loss": 24.95899658203125,
|
| 678 |
+
"step": 1260
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 1.445704564785937,
|
| 682 |
+
"grad_norm": 86.57196044921875,
|
| 683 |
+
"learning_rate": 1.5179677819083023e-05,
|
| 684 |
+
"loss": 24.54622802734375,
|
| 685 |
+
"step": 1275
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 1.4627161893960874,
|
| 689 |
+
"grad_norm": 74.7414321899414,
|
| 690 |
+
"learning_rate": 1.4714993804213136e-05,
|
| 691 |
+
"loss": 23.085076904296876,
|
| 692 |
+
"step": 1290
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 1.4797278140062375,
|
| 696 |
+
"grad_norm": 94.59770202636719,
|
| 697 |
+
"learning_rate": 1.4250309789343247e-05,
|
| 698 |
+
"loss": 23.9911865234375,
|
| 699 |
+
"step": 1305
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 1.496739438616388,
|
| 703 |
+
"grad_norm": 101.66040802001953,
|
| 704 |
+
"learning_rate": 1.378562577447336e-05,
|
| 705 |
+
"loss": 23.332171630859374,
|
| 706 |
+
"step": 1320
|
| 707 |
+
},
|
| 708 |
+
{
|
| 709 |
+
"epoch": 1.5137510632265383,
|
| 710 |
+
"grad_norm": 62.254207611083984,
|
| 711 |
+
"learning_rate": 1.332094175960347e-05,
|
| 712 |
+
"loss": 23.564457194010416,
|
| 713 |
+
"step": 1335
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 1.5307626878366885,
|
| 717 |
+
"grad_norm": 70.31668090820312,
|
| 718 |
+
"learning_rate": 1.2856257744733583e-05,
|
| 719 |
+
"loss": 23.473478190104167,
|
| 720 |
+
"step": 1350
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 1.5307626878366885,
|
| 724 |
+
"eval_accuracy": 0.6789473684210526,
|
| 725 |
+
"eval_loss": 1.4255586862564087,
|
| 726 |
+
"eval_macro_f1": 0.6559084859040256,
|
| 727 |
+
"eval_runtime": 7.8053,
|
| 728 |
+
"eval_samples_per_second": 73.028,
|
| 729 |
+
"eval_steps_per_second": 9.225,
|
| 730 |
+
"step": 1350
|
| 731 |
+
}
|
| 732 |
+
],
|
| 733 |
+
"logging_steps": 15,
|
| 734 |
+
"max_steps": 1764,
|
| 735 |
+
"num_input_tokens_seen": 0,
|
| 736 |
+
"num_train_epochs": 2,
|
| 737 |
+
"save_steps": 150,
|
| 738 |
+
"stateful_callbacks": {
|
| 739 |
+
"TrainerControl": {
|
| 740 |
+
"args": {
|
| 741 |
+
"should_epoch_stop": false,
|
| 742 |
+
"should_evaluate": false,
|
| 743 |
+
"should_log": false,
|
| 744 |
+
"should_save": true,
|
| 745 |
+
"should_training_stop": false
|
| 746 |
+
},
|
| 747 |
+
"attributes": {}
|
| 748 |
+
}
|
| 749 |
+
},
|
| 750 |
+
"total_flos": 3238560714981376.0,
|
| 751 |
+
"train_batch_size": 2,
|
| 752 |
+
"trial_name": null,
|
| 753 |
+
"trial_params": null
|
| 754 |
+
}
|