|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 1.0, |
|
"global_step": 201, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.0, |
|
"learning_rate": 0.0, |
|
"loss": 2.5469, |
|
"step": 1 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"learning_rate": 7.124143742160444e-06, |
|
"loss": 3.1875, |
|
"step": 2 |
|
}, |
|
{ |
|
"epoch": 0.01, |
|
"learning_rate": 1.1291500681071595e-05, |
|
"loss": 2.7812, |
|
"step": 3 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"learning_rate": 1.4248287484320888e-05, |
|
"loss": 0.0, |
|
"step": 4 |
|
}, |
|
{ |
|
"epoch": 0.02, |
|
"learning_rate": 1.6541749506938325e-05, |
|
"loss": 3.75, |
|
"step": 5 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"learning_rate": 1.8415644423232037e-05, |
|
"loss": 0.0, |
|
"step": 6 |
|
}, |
|
{ |
|
"epoch": 0.03, |
|
"learning_rate": 2e-05, |
|
"loss": 2.7969, |
|
"step": 7 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"learning_rate": 2e-05, |
|
"loss": 2.6719, |
|
"step": 8 |
|
}, |
|
{ |
|
"epoch": 0.04, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 9 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 10 |
|
}, |
|
{ |
|
"epoch": 0.05, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 11 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"learning_rate": 2e-05, |
|
"loss": 2.7031, |
|
"step": 12 |
|
}, |
|
{ |
|
"epoch": 0.06, |
|
"learning_rate": 2e-05, |
|
"loss": 2.3906, |
|
"step": 13 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"learning_rate": 2e-05, |
|
"loss": 2.1406, |
|
"step": 14 |
|
}, |
|
{ |
|
"epoch": 0.07, |
|
"learning_rate": 2e-05, |
|
"loss": 4.875, |
|
"step": 15 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 16 |
|
}, |
|
{ |
|
"epoch": 0.08, |
|
"learning_rate": 2e-05, |
|
"loss": 2.625, |
|
"step": 17 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 18 |
|
}, |
|
{ |
|
"epoch": 0.09, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 19 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"learning_rate": 2e-05, |
|
"loss": 2.5938, |
|
"step": 20 |
|
}, |
|
{ |
|
"epoch": 0.1, |
|
"learning_rate": 2e-05, |
|
"loss": 2.625, |
|
"step": 21 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"learning_rate": 2e-05, |
|
"loss": 2.6875, |
|
"step": 22 |
|
}, |
|
{ |
|
"epoch": 0.11, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 23 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 24 |
|
}, |
|
{ |
|
"epoch": 0.12, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 25 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"learning_rate": 2e-05, |
|
"loss": 2.7344, |
|
"step": 26 |
|
}, |
|
{ |
|
"epoch": 0.13, |
|
"learning_rate": 2e-05, |
|
"loss": 5.125, |
|
"step": 27 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 28 |
|
}, |
|
{ |
|
"epoch": 0.14, |
|
"learning_rate": 2e-05, |
|
"loss": 3.7969, |
|
"step": 29 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 30 |
|
}, |
|
{ |
|
"epoch": 0.15, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 31 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"learning_rate": 2e-05, |
|
"loss": 3.0312, |
|
"step": 32 |
|
}, |
|
{ |
|
"epoch": 0.16, |
|
"learning_rate": 2e-05, |
|
"loss": 3.1094, |
|
"step": 33 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"learning_rate": 2e-05, |
|
"loss": 2.3438, |
|
"step": 34 |
|
}, |
|
{ |
|
"epoch": 0.17, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 35 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"learning_rate": 2e-05, |
|
"loss": 4.0625, |
|
"step": 36 |
|
}, |
|
{ |
|
"epoch": 0.18, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 37 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 38 |
|
}, |
|
{ |
|
"epoch": 0.19, |
|
"learning_rate": 2e-05, |
|
"loss": 2.4844, |
|
"step": 39 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 40 |
|
}, |
|
{ |
|
"epoch": 0.2, |
|
"learning_rate": 2e-05, |
|
"loss": 3.375, |
|
"step": 41 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"learning_rate": 2e-05, |
|
"loss": 2.9219, |
|
"step": 42 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"learning_rate": 2e-05, |
|
"loss": 4.4062, |
|
"step": 43 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 44 |
|
}, |
|
{ |
|
"epoch": 0.22, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 45 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 46 |
|
}, |
|
{ |
|
"epoch": 0.23, |
|
"learning_rate": 2e-05, |
|
"loss": 2.75, |
|
"step": 47 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 48 |
|
}, |
|
{ |
|
"epoch": 0.24, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 49 |
|
}, |
|
{ |
|
"epoch": 0.25, |
|
"learning_rate": 2e-05, |
|
"loss": 3.7344, |
|
"step": 50 |
|
}, |
|
{ |
|
"epoch": 0.25, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 51 |
|
}, |
|
{ |
|
"epoch": 0.26, |
|
"learning_rate": 2e-05, |
|
"loss": 2.6406, |
|
"step": 52 |
|
}, |
|
{ |
|
"epoch": 0.26, |
|
"learning_rate": 2e-05, |
|
"loss": 2.4531, |
|
"step": 53 |
|
}, |
|
{ |
|
"epoch": 0.27, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 54 |
|
}, |
|
{ |
|
"epoch": 0.27, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 55 |
|
}, |
|
{ |
|
"epoch": 0.28, |
|
"learning_rate": 2e-05, |
|
"loss": 2.7188, |
|
"step": 56 |
|
}, |
|
{ |
|
"epoch": 0.28, |
|
"learning_rate": 2e-05, |
|
"loss": 3.3438, |
|
"step": 57 |
|
}, |
|
{ |
|
"epoch": 0.29, |
|
"learning_rate": 2e-05, |
|
"loss": 3.3594, |
|
"step": 58 |
|
}, |
|
{ |
|
"epoch": 0.29, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 59 |
|
}, |
|
{ |
|
"epoch": 0.3, |
|
"learning_rate": 2e-05, |
|
"loss": 3.0938, |
|
"step": 60 |
|
}, |
|
{ |
|
"epoch": 0.3, |
|
"learning_rate": 2e-05, |
|
"loss": 2.9531, |
|
"step": 61 |
|
}, |
|
{ |
|
"epoch": 0.31, |
|
"learning_rate": 2e-05, |
|
"loss": 3.75, |
|
"step": 62 |
|
}, |
|
{ |
|
"epoch": 0.31, |
|
"learning_rate": 2e-05, |
|
"loss": 3.5, |
|
"step": 63 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"learning_rate": 2e-05, |
|
"loss": 2.5, |
|
"step": 64 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"learning_rate": 2e-05, |
|
"loss": 2.2969, |
|
"step": 65 |
|
}, |
|
{ |
|
"epoch": 0.33, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 66 |
|
}, |
|
{ |
|
"epoch": 0.33, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 67 |
|
}, |
|
{ |
|
"epoch": 0.34, |
|
"learning_rate": 2e-05, |
|
"loss": 3.1562, |
|
"step": 68 |
|
}, |
|
{ |
|
"epoch": 0.34, |
|
"learning_rate": 2e-05, |
|
"loss": 3.3281, |
|
"step": 69 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 70 |
|
}, |
|
{ |
|
"epoch": 0.35, |
|
"learning_rate": 2e-05, |
|
"loss": 2.5, |
|
"step": 71 |
|
}, |
|
{ |
|
"epoch": 0.36, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 72 |
|
}, |
|
{ |
|
"epoch": 0.36, |
|
"learning_rate": 2e-05, |
|
"loss": 3.9688, |
|
"step": 73 |
|
}, |
|
{ |
|
"epoch": 0.37, |
|
"learning_rate": 2e-05, |
|
"loss": 2.5156, |
|
"step": 74 |
|
}, |
|
{ |
|
"epoch": 0.37, |
|
"learning_rate": 2e-05, |
|
"loss": 3.5312, |
|
"step": 75 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 76 |
|
}, |
|
{ |
|
"epoch": 0.38, |
|
"learning_rate": 2e-05, |
|
"loss": 2.4375, |
|
"step": 77 |
|
}, |
|
{ |
|
"epoch": 0.39, |
|
"learning_rate": 2e-05, |
|
"loss": 3.3594, |
|
"step": 78 |
|
}, |
|
{ |
|
"epoch": 0.39, |
|
"learning_rate": 2e-05, |
|
"loss": 2.9844, |
|
"step": 79 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"learning_rate": 2e-05, |
|
"loss": 3.0, |
|
"step": 80 |
|
}, |
|
{ |
|
"epoch": 0.4, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 81 |
|
}, |
|
{ |
|
"epoch": 0.41, |
|
"learning_rate": 2e-05, |
|
"loss": 2.875, |
|
"step": 82 |
|
}, |
|
{ |
|
"epoch": 0.41, |
|
"learning_rate": 2e-05, |
|
"loss": 2.4844, |
|
"step": 83 |
|
}, |
|
{ |
|
"epoch": 0.42, |
|
"learning_rate": 2e-05, |
|
"loss": 2.4688, |
|
"step": 84 |
|
}, |
|
{ |
|
"epoch": 0.42, |
|
"learning_rate": 2e-05, |
|
"loss": 2.7344, |
|
"step": 85 |
|
}, |
|
{ |
|
"epoch": 0.43, |
|
"learning_rate": 2e-05, |
|
"loss": 2.6875, |
|
"step": 86 |
|
}, |
|
{ |
|
"epoch": 0.43, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 87 |
|
}, |
|
{ |
|
"epoch": 0.44, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 88 |
|
}, |
|
{ |
|
"epoch": 0.44, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 89 |
|
}, |
|
{ |
|
"epoch": 0.45, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 90 |
|
}, |
|
{ |
|
"epoch": 0.45, |
|
"learning_rate": 2e-05, |
|
"loss": 2.7188, |
|
"step": 91 |
|
}, |
|
{ |
|
"epoch": 0.46, |
|
"learning_rate": 2e-05, |
|
"loss": 4.625, |
|
"step": 92 |
|
}, |
|
{ |
|
"epoch": 0.46, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 93 |
|
}, |
|
{ |
|
"epoch": 0.47, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 94 |
|
}, |
|
{ |
|
"epoch": 0.47, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 95 |
|
}, |
|
{ |
|
"epoch": 0.48, |
|
"learning_rate": 2e-05, |
|
"loss": 3.8906, |
|
"step": 96 |
|
}, |
|
{ |
|
"epoch": 0.48, |
|
"learning_rate": 2e-05, |
|
"loss": 5.4375, |
|
"step": 97 |
|
}, |
|
{ |
|
"epoch": 0.49, |
|
"learning_rate": 2e-05, |
|
"loss": 3.2031, |
|
"step": 98 |
|
}, |
|
{ |
|
"epoch": 0.49, |
|
"learning_rate": 2e-05, |
|
"loss": 2.7812, |
|
"step": 99 |
|
}, |
|
{ |
|
"epoch": 0.5, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 100 |
|
}, |
|
{ |
|
"epoch": 0.5, |
|
"learning_rate": 2e-05, |
|
"loss": 2.8594, |
|
"step": 101 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"learning_rate": 2e-05, |
|
"loss": 3.0625, |
|
"step": 102 |
|
}, |
|
{ |
|
"epoch": 0.51, |
|
"learning_rate": 2e-05, |
|
"loss": 2.4844, |
|
"step": 103 |
|
}, |
|
{ |
|
"epoch": 0.52, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 104 |
|
}, |
|
{ |
|
"epoch": 0.52, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 105 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 106 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 107 |
|
}, |
|
{ |
|
"epoch": 0.54, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 108 |
|
}, |
|
{ |
|
"epoch": 0.54, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 109 |
|
}, |
|
{ |
|
"epoch": 0.55, |
|
"learning_rate": 2e-05, |
|
"loss": 2.75, |
|
"step": 110 |
|
}, |
|
{ |
|
"epoch": 0.55, |
|
"learning_rate": 2e-05, |
|
"loss": 2.6094, |
|
"step": 111 |
|
}, |
|
{ |
|
"epoch": 0.56, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 112 |
|
}, |
|
{ |
|
"epoch": 0.56, |
|
"learning_rate": 2e-05, |
|
"loss": 2.5625, |
|
"step": 113 |
|
}, |
|
{ |
|
"epoch": 0.57, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 114 |
|
}, |
|
{ |
|
"epoch": 0.57, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 115 |
|
}, |
|
{ |
|
"epoch": 0.58, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 116 |
|
}, |
|
{ |
|
"epoch": 0.58, |
|
"learning_rate": 2e-05, |
|
"loss": 2.4062, |
|
"step": 117 |
|
}, |
|
{ |
|
"epoch": 0.59, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 118 |
|
}, |
|
{ |
|
"epoch": 0.59, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 119 |
|
}, |
|
{ |
|
"epoch": 0.6, |
|
"learning_rate": 2e-05, |
|
"loss": 2.8906, |
|
"step": 120 |
|
}, |
|
{ |
|
"epoch": 0.6, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 121 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 122 |
|
}, |
|
{ |
|
"epoch": 0.61, |
|
"learning_rate": 2e-05, |
|
"loss": 2.9375, |
|
"step": 123 |
|
}, |
|
{ |
|
"epoch": 0.62, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 124 |
|
}, |
|
{ |
|
"epoch": 0.62, |
|
"learning_rate": 2e-05, |
|
"loss": 3.9062, |
|
"step": 125 |
|
}, |
|
{ |
|
"epoch": 0.63, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 126 |
|
}, |
|
{ |
|
"epoch": 0.63, |
|
"learning_rate": 2e-05, |
|
"loss": 2.2812, |
|
"step": 127 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"learning_rate": 2e-05, |
|
"loss": 3.3125, |
|
"step": 128 |
|
}, |
|
{ |
|
"epoch": 0.64, |
|
"learning_rate": 2e-05, |
|
"loss": 2.4844, |
|
"step": 129 |
|
}, |
|
{ |
|
"epoch": 0.65, |
|
"learning_rate": 2e-05, |
|
"loss": 2.0625, |
|
"step": 130 |
|
}, |
|
{ |
|
"epoch": 0.65, |
|
"learning_rate": 2e-05, |
|
"loss": 2.875, |
|
"step": 131 |
|
}, |
|
{ |
|
"epoch": 0.66, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 132 |
|
}, |
|
{ |
|
"epoch": 0.66, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 133 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"learning_rate": 2e-05, |
|
"loss": 1.9453, |
|
"step": 134 |
|
}, |
|
{ |
|
"epoch": 0.67, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 135 |
|
}, |
|
{ |
|
"epoch": 0.68, |
|
"learning_rate": 2e-05, |
|
"loss": 2.8594, |
|
"step": 136 |
|
}, |
|
{ |
|
"epoch": 0.68, |
|
"learning_rate": 2e-05, |
|
"loss": 2.5625, |
|
"step": 137 |
|
}, |
|
{ |
|
"epoch": 0.69, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 138 |
|
}, |
|
{ |
|
"epoch": 0.69, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 139 |
|
}, |
|
{ |
|
"epoch": 0.7, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 140 |
|
}, |
|
{ |
|
"epoch": 0.7, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 141 |
|
}, |
|
{ |
|
"epoch": 0.71, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 142 |
|
}, |
|
{ |
|
"epoch": 0.71, |
|
"learning_rate": 2e-05, |
|
"loss": 2.4531, |
|
"step": 143 |
|
}, |
|
{ |
|
"epoch": 0.72, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 144 |
|
}, |
|
{ |
|
"epoch": 0.72, |
|
"learning_rate": 2e-05, |
|
"loss": 2.625, |
|
"step": 145 |
|
}, |
|
{ |
|
"epoch": 0.73, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 146 |
|
}, |
|
{ |
|
"epoch": 0.73, |
|
"learning_rate": 2e-05, |
|
"loss": 2.9688, |
|
"step": 147 |
|
}, |
|
{ |
|
"epoch": 0.74, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 148 |
|
}, |
|
{ |
|
"epoch": 0.74, |
|
"learning_rate": 2e-05, |
|
"loss": 1.875, |
|
"step": 149 |
|
}, |
|
{ |
|
"epoch": 0.75, |
|
"learning_rate": 2e-05, |
|
"loss": 2.5938, |
|
"step": 150 |
|
}, |
|
{ |
|
"epoch": 0.75, |
|
"learning_rate": 2e-05, |
|
"loss": 3.875, |
|
"step": 151 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"learning_rate": 2e-05, |
|
"loss": 2.4062, |
|
"step": 152 |
|
}, |
|
{ |
|
"epoch": 0.76, |
|
"learning_rate": 2e-05, |
|
"loss": 2.6406, |
|
"step": 153 |
|
}, |
|
{ |
|
"epoch": 0.77, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 154 |
|
}, |
|
{ |
|
"epoch": 0.77, |
|
"learning_rate": 2e-05, |
|
"loss": 4.5312, |
|
"step": 155 |
|
}, |
|
{ |
|
"epoch": 0.78, |
|
"learning_rate": 2e-05, |
|
"loss": 2.2812, |
|
"step": 156 |
|
}, |
|
{ |
|
"epoch": 0.78, |
|
"learning_rate": 2e-05, |
|
"loss": 2.5469, |
|
"step": 157 |
|
}, |
|
{ |
|
"epoch": 0.79, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 158 |
|
}, |
|
{ |
|
"epoch": 0.79, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 159 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"learning_rate": 2e-05, |
|
"loss": 2.8906, |
|
"step": 160 |
|
}, |
|
{ |
|
"epoch": 0.8, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 161 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"learning_rate": 2e-05, |
|
"loss": 2.5625, |
|
"step": 162 |
|
}, |
|
{ |
|
"epoch": 0.81, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 163 |
|
}, |
|
{ |
|
"epoch": 0.82, |
|
"learning_rate": 2e-05, |
|
"loss": 2.2812, |
|
"step": 164 |
|
}, |
|
{ |
|
"epoch": 0.82, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 165 |
|
}, |
|
{ |
|
"epoch": 0.83, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 166 |
|
}, |
|
{ |
|
"epoch": 0.83, |
|
"learning_rate": 2e-05, |
|
"loss": 2.8125, |
|
"step": 167 |
|
}, |
|
{ |
|
"epoch": 0.84, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 168 |
|
}, |
|
{ |
|
"epoch": 0.84, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 169 |
|
}, |
|
{ |
|
"epoch": 0.85, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 170 |
|
}, |
|
{ |
|
"epoch": 0.85, |
|
"learning_rate": 2e-05, |
|
"loss": 2.6562, |
|
"step": 171 |
|
}, |
|
{ |
|
"epoch": 0.86, |
|
"learning_rate": 2e-05, |
|
"loss": 2.8281, |
|
"step": 172 |
|
}, |
|
{ |
|
"epoch": 0.86, |
|
"learning_rate": 2e-05, |
|
"loss": 4.125, |
|
"step": 173 |
|
}, |
|
{ |
|
"epoch": 0.87, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 174 |
|
}, |
|
{ |
|
"epoch": 0.87, |
|
"learning_rate": 2e-05, |
|
"loss": 2.7656, |
|
"step": 175 |
|
}, |
|
{ |
|
"epoch": 0.88, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 176 |
|
}, |
|
{ |
|
"epoch": 0.88, |
|
"learning_rate": 2e-05, |
|
"loss": 2.4688, |
|
"step": 177 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 178 |
|
}, |
|
{ |
|
"epoch": 0.89, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 179 |
|
}, |
|
{ |
|
"epoch": 0.9, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 180 |
|
}, |
|
{ |
|
"epoch": 0.9, |
|
"learning_rate": 2e-05, |
|
"loss": 2.0938, |
|
"step": 181 |
|
}, |
|
{ |
|
"epoch": 0.91, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 182 |
|
}, |
|
{ |
|
"epoch": 0.91, |
|
"learning_rate": 2e-05, |
|
"loss": 2.6562, |
|
"step": 183 |
|
}, |
|
{ |
|
"epoch": 0.92, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 184 |
|
}, |
|
{ |
|
"epoch": 0.92, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 185 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"learning_rate": 2e-05, |
|
"loss": 2.9688, |
|
"step": 186 |
|
}, |
|
{ |
|
"epoch": 0.93, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 187 |
|
}, |
|
{ |
|
"epoch": 0.94, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 188 |
|
}, |
|
{ |
|
"epoch": 0.94, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 189 |
|
}, |
|
{ |
|
"epoch": 0.95, |
|
"learning_rate": 2e-05, |
|
"loss": 2.5781, |
|
"step": 190 |
|
}, |
|
{ |
|
"epoch": 0.95, |
|
"learning_rate": 2e-05, |
|
"loss": 2.4219, |
|
"step": 191 |
|
}, |
|
{ |
|
"epoch": 0.96, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 192 |
|
}, |
|
{ |
|
"epoch": 0.96, |
|
"learning_rate": 2e-05, |
|
"loss": 2.7969, |
|
"step": 193 |
|
}, |
|
{ |
|
"epoch": 0.97, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 194 |
|
}, |
|
{ |
|
"epoch": 0.97, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 195 |
|
}, |
|
{ |
|
"epoch": 0.98, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 196 |
|
}, |
|
{ |
|
"epoch": 0.98, |
|
"learning_rate": 2e-05, |
|
"loss": 2.2031, |
|
"step": 197 |
|
}, |
|
{ |
|
"epoch": 0.99, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 198 |
|
}, |
|
{ |
|
"epoch": 0.99, |
|
"learning_rate": 2e-05, |
|
"loss": 0.0, |
|
"step": 199 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"learning_rate": 2e-05, |
|
"loss": 2.375, |
|
"step": 200 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"learning_rate": 2e-05, |
|
"loss": 4.0938, |
|
"step": 201 |
|
}, |
|
{ |
|
"epoch": 1.0, |
|
"step": 201, |
|
"total_flos": 1.632975817015296e+16, |
|
"train_loss": 1.5560867537313432, |
|
"train_runtime": 88.8069, |
|
"train_samples_per_second": 2.263, |
|
"train_steps_per_second": 2.263 |
|
} |
|
], |
|
"max_steps": 201, |
|
"num_train_epochs": 1, |
|
"total_flos": 1.632975817015296e+16, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|