Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +847 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,847 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 19.982758620689655,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 1159,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.1724137931034483,
|
| 13 |
+
"grad_norm": 4.288447380065918,
|
| 14 |
+
"learning_rate": 3.4482758620689657e-05,
|
| 15 |
+
"loss": 1.24,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.3448275862068966,
|
| 20 |
+
"grad_norm": 3.9249799251556396,
|
| 21 |
+
"learning_rate": 6.896551724137931e-05,
|
| 22 |
+
"loss": 0.5159,
|
| 23 |
+
"step": 20
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.5172413793103449,
|
| 27 |
+
"grad_norm": 1.7602132558822632,
|
| 28 |
+
"learning_rate": 0.00010344827586206898,
|
| 29 |
+
"loss": 0.3011,
|
| 30 |
+
"step": 30
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.6896551724137931,
|
| 34 |
+
"grad_norm": 0.8162740468978882,
|
| 35 |
+
"learning_rate": 0.00013793103448275863,
|
| 36 |
+
"loss": 0.1873,
|
| 37 |
+
"step": 40
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.8620689655172413,
|
| 41 |
+
"grad_norm": 0.7871344685554504,
|
| 42 |
+
"learning_rate": 0.00017241379310344826,
|
| 43 |
+
"loss": 0.1371,
|
| 44 |
+
"step": 50
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.0344827586206897,
|
| 48 |
+
"grad_norm": 1.0209136009216309,
|
| 49 |
+
"learning_rate": 0.00019999837162688113,
|
| 50 |
+
"loss": 0.1282,
|
| 51 |
+
"step": 60
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 1.206896551724138,
|
| 55 |
+
"grad_norm": 0.5767734050750732,
|
| 56 |
+
"learning_rate": 0.00019994138413588491,
|
| 57 |
+
"loss": 0.117,
|
| 58 |
+
"step": 70
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 1.3793103448275863,
|
| 62 |
+
"grad_norm": 0.9807557463645935,
|
| 63 |
+
"learning_rate": 0.00019980303101316228,
|
| 64 |
+
"loss": 0.1053,
|
| 65 |
+
"step": 80
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 1.5517241379310345,
|
| 69 |
+
"grad_norm": 0.8148991465568542,
|
| 70 |
+
"learning_rate": 0.00019958342489662917,
|
| 71 |
+
"loss": 0.095,
|
| 72 |
+
"step": 90
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 1.7241379310344827,
|
| 76 |
+
"grad_norm": 0.7483856678009033,
|
| 77 |
+
"learning_rate": 0.00019928274457498818,
|
| 78 |
+
"loss": 0.0823,
|
| 79 |
+
"step": 100
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.896551724137931,
|
| 83 |
+
"grad_norm": 0.38755494356155396,
|
| 84 |
+
"learning_rate": 0.00019890123484217056,
|
| 85 |
+
"loss": 0.0852,
|
| 86 |
+
"step": 110
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 2.0689655172413794,
|
| 90 |
+
"grad_norm": 0.40888020396232605,
|
| 91 |
+
"learning_rate": 0.0001984392062980413,
|
| 92 |
+
"loss": 0.0776,
|
| 93 |
+
"step": 120
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 2.2413793103448274,
|
| 97 |
+
"grad_norm": 0.4752354919910431,
|
| 98 |
+
"learning_rate": 0.00019789703509552945,
|
| 99 |
+
"loss": 0.065,
|
| 100 |
+
"step": 130
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 2.413793103448276,
|
| 104 |
+
"grad_norm": 0.2943930923938751,
|
| 105 |
+
"learning_rate": 0.00019727516263438918,
|
| 106 |
+
"loss": 0.0703,
|
| 107 |
+
"step": 140
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 2.586206896551724,
|
| 111 |
+
"grad_norm": 0.3636096119880676,
|
| 112 |
+
"learning_rate": 0.0001965740952018417,
|
| 113 |
+
"loss": 0.0734,
|
| 114 |
+
"step": 150
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 2.7586206896551726,
|
| 118 |
+
"grad_norm": 0.49514421820640564,
|
| 119 |
+
"learning_rate": 0.00019579440356038967,
|
| 120 |
+
"loss": 0.0685,
|
| 121 |
+
"step": 160
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 2.9310344827586206,
|
| 125 |
+
"grad_norm": 0.7530002593994141,
|
| 126 |
+
"learning_rate": 0.00019493672248314044,
|
| 127 |
+
"loss": 0.0653,
|
| 128 |
+
"step": 170
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 3.103448275862069,
|
| 132 |
+
"grad_norm": 0.4943099021911621,
|
| 133 |
+
"learning_rate": 0.00019400175023701586,
|
| 134 |
+
"loss": 0.0569,
|
| 135 |
+
"step": 180
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 3.2758620689655173,
|
| 139 |
+
"grad_norm": 0.5510227084159851,
|
| 140 |
+
"learning_rate": 0.00019299024801426994,
|
| 141 |
+
"loss": 0.0612,
|
| 142 |
+
"step": 190
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 3.4482758620689653,
|
| 146 |
+
"grad_norm": 0.4278947114944458,
|
| 147 |
+
"learning_rate": 0.0001919030393127765,
|
| 148 |
+
"loss": 0.0554,
|
| 149 |
+
"step": 200
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 3.6206896551724137,
|
| 153 |
+
"grad_norm": 0.3584755063056946,
|
| 154 |
+
"learning_rate": 0.0001907410092655921,
|
| 155 |
+
"loss": 0.0599,
|
| 156 |
+
"step": 210
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 3.793103448275862,
|
| 160 |
+
"grad_norm": 0.5437096953392029,
|
| 161 |
+
"learning_rate": 0.00018950510392033945,
|
| 162 |
+
"loss": 0.0597,
|
| 163 |
+
"step": 220
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 3.9655172413793105,
|
| 167 |
+
"grad_norm": 0.3559582531452179,
|
| 168 |
+
"learning_rate": 0.0001881963294689984,
|
| 169 |
+
"loss": 0.0569,
|
| 170 |
+
"step": 230
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 4.137931034482759,
|
| 174 |
+
"grad_norm": 0.4884796440601349,
|
| 175 |
+
"learning_rate": 0.00018681575142873126,
|
| 176 |
+
"loss": 0.0578,
|
| 177 |
+
"step": 240
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 4.310344827586207,
|
| 181 |
+
"grad_norm": 0.4382438659667969,
|
| 182 |
+
"learning_rate": 0.0001853644937744095,
|
| 183 |
+
"loss": 0.051,
|
| 184 |
+
"step": 250
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 4.482758620689655,
|
| 188 |
+
"grad_norm": 0.34379124641418457,
|
| 189 |
+
"learning_rate": 0.0001838437380235483,
|
| 190 |
+
"loss": 0.0529,
|
| 191 |
+
"step": 260
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 4.655172413793103,
|
| 195 |
+
"grad_norm": 0.5280494093894958,
|
| 196 |
+
"learning_rate": 0.0001822547222743933,
|
| 197 |
+
"loss": 0.0503,
|
| 198 |
+
"step": 270
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 4.827586206896552,
|
| 202 |
+
"grad_norm": 0.3714671730995178,
|
| 203 |
+
"learning_rate": 0.00018059874019794351,
|
| 204 |
+
"loss": 0.0474,
|
| 205 |
+
"step": 280
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 5.0,
|
| 209 |
+
"grad_norm": 0.2870706617832184,
|
| 210 |
+
"learning_rate": 0.00017887713998473022,
|
| 211 |
+
"loss": 0.0473,
|
| 212 |
+
"step": 290
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 5.172413793103448,
|
| 216 |
+
"grad_norm": 0.35148724913597107,
|
| 217 |
+
"learning_rate": 0.00017709132324720997,
|
| 218 |
+
"loss": 0.0492,
|
| 219 |
+
"step": 300
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 5.344827586206897,
|
| 223 |
+
"grad_norm": 0.2515222728252411,
|
| 224 |
+
"learning_rate": 0.00017524274387866484,
|
| 225 |
+
"loss": 0.0462,
|
| 226 |
+
"step": 310
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 5.517241379310345,
|
| 230 |
+
"grad_norm": 0.22452038526535034,
|
| 231 |
+
"learning_rate": 0.00017333290686953915,
|
| 232 |
+
"loss": 0.0414,
|
| 233 |
+
"step": 320
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 5.689655172413794,
|
| 237 |
+
"grad_norm": 0.3675800561904907,
|
| 238 |
+
"learning_rate": 0.00017136336708217632,
|
| 239 |
+
"loss": 0.0483,
|
| 240 |
+
"step": 330
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 5.862068965517241,
|
| 244 |
+
"grad_norm": 0.3029170334339142,
|
| 245 |
+
"learning_rate": 0.00016933572798495328,
|
| 246 |
+
"loss": 0.0553,
|
| 247 |
+
"step": 340
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 6.0344827586206895,
|
| 251 |
+
"grad_norm": 0.2198949158191681,
|
| 252 |
+
"learning_rate": 0.00016725164034684317,
|
| 253 |
+
"loss": 0.0466,
|
| 254 |
+
"step": 350
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 6.206896551724138,
|
| 258 |
+
"grad_norm": 0.1851377636194229,
|
| 259 |
+
"learning_rate": 0.00016511280089346876,
|
| 260 |
+
"loss": 0.0462,
|
| 261 |
+
"step": 360
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 6.379310344827586,
|
| 265 |
+
"grad_norm": 0.3081173300743103,
|
| 266 |
+
"learning_rate": 0.00016292095092574154,
|
| 267 |
+
"loss": 0.0433,
|
| 268 |
+
"step": 370
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 6.551724137931035,
|
| 272 |
+
"grad_norm": 0.2842998206615448,
|
| 273 |
+
"learning_rate": 0.00016067787490221008,
|
| 274 |
+
"loss": 0.0389,
|
| 275 |
+
"step": 380
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 6.724137931034483,
|
| 279 |
+
"grad_norm": 0.31966182589530945,
|
| 280 |
+
"learning_rate": 0.00015838539898627242,
|
| 281 |
+
"loss": 0.04,
|
| 282 |
+
"step": 390
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 6.896551724137931,
|
| 286 |
+
"grad_norm": 0.23094028234481812,
|
| 287 |
+
"learning_rate": 0.0001560453895594354,
|
| 288 |
+
"loss": 0.0395,
|
| 289 |
+
"step": 400
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 7.068965517241379,
|
| 293 |
+
"grad_norm": 0.35125771164894104,
|
| 294 |
+
"learning_rate": 0.00015365975170183085,
|
| 295 |
+
"loss": 0.0445,
|
| 296 |
+
"step": 410
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 7.241379310344827,
|
| 300 |
+
"grad_norm": 0.18835288286209106,
|
| 301 |
+
"learning_rate": 0.00015123042764122584,
|
| 302 |
+
"loss": 0.0391,
|
| 303 |
+
"step": 420
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 7.413793103448276,
|
| 307 |
+
"grad_norm": 0.26209092140197754,
|
| 308 |
+
"learning_rate": 0.00014875939517179016,
|
| 309 |
+
"loss": 0.0433,
|
| 310 |
+
"step": 430
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 7.586206896551724,
|
| 314 |
+
"grad_norm": 0.23597943782806396,
|
| 315 |
+
"learning_rate": 0.00014624866604390758,
|
| 316 |
+
"loss": 0.0383,
|
| 317 |
+
"step": 440
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 7.758620689655173,
|
| 321 |
+
"grad_norm": 0.32963162660598755,
|
| 322 |
+
"learning_rate": 0.00014370028432634252,
|
| 323 |
+
"loss": 0.0438,
|
| 324 |
+
"step": 450
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 7.931034482758621,
|
| 328 |
+
"grad_norm": 0.2865336537361145,
|
| 329 |
+
"learning_rate": 0.00014111632474209505,
|
| 330 |
+
"loss": 0.0367,
|
| 331 |
+
"step": 460
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 8.10344827586207,
|
| 335 |
+
"grad_norm": 0.22742541134357452,
|
| 336 |
+
"learning_rate": 0.0001384988909792995,
|
| 337 |
+
"loss": 0.0366,
|
| 338 |
+
"step": 470
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 8.275862068965518,
|
| 342 |
+
"grad_norm": 0.23098573088645935,
|
| 343 |
+
"learning_rate": 0.00013585011397854117,
|
| 344 |
+
"loss": 0.0418,
|
| 345 |
+
"step": 480
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 8.448275862068966,
|
| 349 |
+
"grad_norm": 0.3277537524700165,
|
| 350 |
+
"learning_rate": 0.00013317215019798638,
|
| 351 |
+
"loss": 0.0391,
|
| 352 |
+
"step": 490
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 8.620689655172415,
|
| 356 |
+
"grad_norm": 0.33878064155578613,
|
| 357 |
+
"learning_rate": 0.00013046717985773745,
|
| 358 |
+
"loss": 0.0401,
|
| 359 |
+
"step": 500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 8.793103448275861,
|
| 363 |
+
"grad_norm": 0.4036126434803009,
|
| 364 |
+
"learning_rate": 0.00012773740516484235,
|
| 365 |
+
"loss": 0.0369,
|
| 366 |
+
"step": 510
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 8.96551724137931,
|
| 370 |
+
"grad_norm": 0.2406553477048874,
|
| 371 |
+
"learning_rate": 0.00012498504852040434,
|
| 372 |
+
"loss": 0.0357,
|
| 373 |
+
"step": 520
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 9.137931034482758,
|
| 377 |
+
"grad_norm": 0.33557119965553284,
|
| 378 |
+
"learning_rate": 0.00012221235071025046,
|
| 379 |
+
"loss": 0.0366,
|
| 380 |
+
"step": 530
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 9.310344827586206,
|
| 384 |
+
"grad_norm": 0.2251608818769455,
|
| 385 |
+
"learning_rate": 0.00011942156908063284,
|
| 386 |
+
"loss": 0.0387,
|
| 387 |
+
"step": 540
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 9.482758620689655,
|
| 391 |
+
"grad_norm": 0.331117182970047,
|
| 392 |
+
"learning_rate": 0.00011661497570044738,
|
| 393 |
+
"loss": 0.0306,
|
| 394 |
+
"step": 550
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 9.655172413793103,
|
| 398 |
+
"grad_norm": 0.1580682098865509,
|
| 399 |
+
"learning_rate": 0.00011379485551146621,
|
| 400 |
+
"loss": 0.0332,
|
| 401 |
+
"step": 560
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 9.827586206896552,
|
| 405 |
+
"grad_norm": 0.21949926018714905,
|
| 406 |
+
"learning_rate": 0.00011096350446808998,
|
| 407 |
+
"loss": 0.0378,
|
| 408 |
+
"step": 570
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 10.0,
|
| 412 |
+
"grad_norm": 0.300108402967453,
|
| 413 |
+
"learning_rate": 0.00010812322766813461,
|
| 414 |
+
"loss": 0.0354,
|
| 415 |
+
"step": 580
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 10.172413793103448,
|
| 419 |
+
"grad_norm": 0.2582547664642334,
|
| 420 |
+
"learning_rate": 0.00010527633747617362,
|
| 421 |
+
"loss": 0.0348,
|
| 422 |
+
"step": 590
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 10.344827586206897,
|
| 426 |
+
"grad_norm": 0.3702482283115387,
|
| 427 |
+
"learning_rate": 0.00010242515164096471,
|
| 428 |
+
"loss": 0.0334,
|
| 429 |
+
"step": 600
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 10.517241379310345,
|
| 433 |
+
"grad_norm": 0.2247171700000763,
|
| 434 |
+
"learning_rate": 9.957199140849278e-05,
|
| 435 |
+
"loss": 0.0313,
|
| 436 |
+
"step": 610
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 10.689655172413794,
|
| 440 |
+
"grad_norm": 0.26956915855407715,
|
| 441 |
+
"learning_rate": 9.671917963216574e-05,
|
| 442 |
+
"loss": 0.0336,
|
| 443 |
+
"step": 620
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 10.862068965517242,
|
| 447 |
+
"grad_norm": 0.19159068167209625,
|
| 448 |
+
"learning_rate": 9.386903888170172e-05,
|
| 449 |
+
"loss": 0.0311,
|
| 450 |
+
"step": 630
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 11.03448275862069,
|
| 454 |
+
"grad_norm": 0.244057759642601,
|
| 455 |
+
"learning_rate": 9.102388955224703e-05,
|
| 456 |
+
"loss": 0.033,
|
| 457 |
+
"step": 640
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 11.206896551724139,
|
| 461 |
+
"grad_norm": 0.24532181024551392,
|
| 462 |
+
"learning_rate": 8.818604797526493e-05,
|
| 463 |
+
"loss": 0.0365,
|
| 464 |
+
"step": 650
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 11.379310344827585,
|
| 468 |
+
"grad_norm": 0.3451012670993805,
|
| 469 |
+
"learning_rate": 8.535782453273217e-05,
|
| 470 |
+
"loss": 0.0297,
|
| 471 |
+
"step": 660
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 11.551724137931034,
|
| 475 |
+
"grad_norm": 0.28239578008651733,
|
| 476 |
+
"learning_rate": 8.254152177618e-05,
|
| 477 |
+
"loss": 0.0285,
|
| 478 |
+
"step": 670
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 11.724137931034482,
|
| 482 |
+
"grad_norm": 0.2965984046459198,
|
| 483 |
+
"learning_rate": 7.97394325521093e-05,
|
| 484 |
+
"loss": 0.0303,
|
| 485 |
+
"step": 680
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 11.89655172413793,
|
| 489 |
+
"grad_norm": 0.25484567880630493,
|
| 490 |
+
"learning_rate": 7.695383813530776e-05,
|
| 491 |
+
"loss": 0.0283,
|
| 492 |
+
"step": 690
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 12.068965517241379,
|
| 496 |
+
"grad_norm": 0.15973417460918427,
|
| 497 |
+
"learning_rate": 7.418700637158742e-05,
|
| 498 |
+
"loss": 0.0324,
|
| 499 |
+
"step": 700
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 12.241379310344827,
|
| 503 |
+
"grad_norm": 0.16871625185012817,
|
| 504 |
+
"learning_rate": 7.144118983145521e-05,
|
| 505 |
+
"loss": 0.029,
|
| 506 |
+
"step": 710
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 12.413793103448276,
|
| 510 |
+
"grad_norm": 0.19246120750904083,
|
| 511 |
+
"learning_rate": 6.871862397622005e-05,
|
| 512 |
+
"loss": 0.0309,
|
| 513 |
+
"step": 720
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 12.586206896551724,
|
| 517 |
+
"grad_norm": 0.28560781478881836,
|
| 518 |
+
"learning_rate": 6.60215253380287e-05,
|
| 519 |
+
"loss": 0.03,
|
| 520 |
+
"step": 730
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 12.758620689655173,
|
| 524 |
+
"grad_norm": 0.2544420659542084,
|
| 525 |
+
"learning_rate": 6.335208971531265e-05,
|
| 526 |
+
"loss": 0.0338,
|
| 527 |
+
"step": 740
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 12.931034482758621,
|
| 531 |
+
"grad_norm": 0.16777703166007996,
|
| 532 |
+
"learning_rate": 6.071249038511497e-05,
|
| 533 |
+
"loss": 0.03,
|
| 534 |
+
"step": 750
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 13.10344827586207,
|
| 538 |
+
"grad_norm": 0.13849933445453644,
|
| 539 |
+
"learning_rate": 5.810487633375261e-05,
|
| 540 |
+
"loss": 0.03,
|
| 541 |
+
"step": 760
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 13.275862068965518,
|
| 545 |
+
"grad_norm": 0.2304876148700714,
|
| 546 |
+
"learning_rate": 5.553137050725459e-05,
|
| 547 |
+
"loss": 0.032,
|
| 548 |
+
"step": 770
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 13.448275862068966,
|
| 552 |
+
"grad_norm": 0.17199701070785522,
|
| 553 |
+
"learning_rate": 5.2994068083000495e-05,
|
| 554 |
+
"loss": 0.0324,
|
| 555 |
+
"step": 780
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 13.620689655172415,
|
| 559 |
+
"grad_norm": 0.15231479704380035,
|
| 560 |
+
"learning_rate": 5.049503476396627e-05,
|
| 561 |
+
"loss": 0.0269,
|
| 562 |
+
"step": 790
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 13.793103448275861,
|
| 566 |
+
"grad_norm": 0.20847296714782715,
|
| 567 |
+
"learning_rate": 4.803630509696627e-05,
|
| 568 |
+
"loss": 0.0276,
|
| 569 |
+
"step": 800
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 13.96551724137931,
|
| 573 |
+
"grad_norm": 0.2147344946861267,
|
| 574 |
+
"learning_rate": 4.561988081626023e-05,
|
| 575 |
+
"loss": 0.0252,
|
| 576 |
+
"step": 810
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 14.137931034482758,
|
| 580 |
+
"grad_norm": 0.1602960079908371,
|
| 581 |
+
"learning_rate": 4.32477292138746e-05,
|
| 582 |
+
"loss": 0.025,
|
| 583 |
+
"step": 820
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 14.310344827586206,
|
| 587 |
+
"grad_norm": 0.19987209141254425,
|
| 588 |
+
"learning_rate": 4.092178153796401e-05,
|
| 589 |
+
"loss": 0.0268,
|
| 590 |
+
"step": 830
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 14.482758620689655,
|
| 594 |
+
"grad_norm": 0.18080690503120422,
|
| 595 |
+
"learning_rate": 3.8643931420517255e-05,
|
| 596 |
+
"loss": 0.0279,
|
| 597 |
+
"step": 840
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 14.655172413793103,
|
| 601 |
+
"grad_norm": 0.1728052943944931,
|
| 602 |
+
"learning_rate": 3.641603333568831e-05,
|
| 603 |
+
"loss": 0.0226,
|
| 604 |
+
"step": 850
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 14.827586206896552,
|
| 608 |
+
"grad_norm": 0.19159851968288422,
|
| 609 |
+
"learning_rate": 3.423990109000641e-05,
|
| 610 |
+
"loss": 0.0281,
|
| 611 |
+
"step": 860
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 15.0,
|
| 615 |
+
"grad_norm": 0.17135493457317352,
|
| 616 |
+
"learning_rate": 3.211730634569563e-05,
|
| 617 |
+
"loss": 0.0291,
|
| 618 |
+
"step": 870
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 15.172413793103448,
|
| 622 |
+
"grad_norm": 0.24832935631275177,
|
| 623 |
+
"learning_rate": 3.0049977178305076e-05,
|
| 624 |
+
"loss": 0.0277,
|
| 625 |
+
"step": 880
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 15.344827586206897,
|
| 629 |
+
"grad_norm": 0.22584789991378784,
|
| 630 |
+
"learning_rate": 2.803959666982474e-05,
|
| 631 |
+
"loss": 0.025,
|
| 632 |
+
"step": 890
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 15.517241379310345,
|
| 636 |
+
"grad_norm": 0.17386794090270996,
|
| 637 |
+
"learning_rate": 2.60878015384319e-05,
|
| 638 |
+
"loss": 0.0289,
|
| 639 |
+
"step": 900
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 15.689655172413794,
|
| 643 |
+
"grad_norm": 0.16346997022628784,
|
| 644 |
+
"learning_rate": 2.419618080598417e-05,
|
| 645 |
+
"loss": 0.0207,
|
| 646 |
+
"step": 910
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 15.862068965517242,
|
| 650 |
+
"grad_norm": 0.11085375398397446,
|
| 651 |
+
"learning_rate": 2.23662745043433e-05,
|
| 652 |
+
"loss": 0.0227,
|
| 653 |
+
"step": 920
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 16.03448275862069,
|
| 657 |
+
"grad_norm": 0.1747630387544632,
|
| 658 |
+
"learning_rate": 2.05995724215838e-05,
|
| 659 |
+
"loss": 0.0257,
|
| 660 |
+
"step": 930
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 16.20689655172414,
|
| 664 |
+
"grad_norm": 0.15860582888126373,
|
| 665 |
+
"learning_rate": 1.889751288910645e-05,
|
| 666 |
+
"loss": 0.0254,
|
| 667 |
+
"step": 940
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 16.379310344827587,
|
| 671 |
+
"grad_norm": 0.15299658477306366,
|
| 672 |
+
"learning_rate": 1.726148161064437e-05,
|
| 673 |
+
"loss": 0.0251,
|
| 674 |
+
"step": 950
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 16.551724137931036,
|
| 678 |
+
"grad_norm": 0.1972336769104004,
|
| 679 |
+
"learning_rate": 1.569281053411532e-05,
|
| 680 |
+
"loss": 0.0256,
|
| 681 |
+
"step": 960
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 16.724137931034484,
|
| 685 |
+
"grad_norm": 0.13377782702445984,
|
| 686 |
+
"learning_rate": 1.4192776767238158e-05,
|
| 687 |
+
"loss": 0.0249,
|
| 688 |
+
"step": 970
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 16.896551724137932,
|
| 692 |
+
"grad_norm": 0.17352934181690216,
|
| 693 |
+
"learning_rate": 1.276260153779667e-05,
|
| 694 |
+
"loss": 0.0231,
|
| 695 |
+
"step": 980
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 17.06896551724138,
|
| 699 |
+
"grad_norm": 0.21800649166107178,
|
| 700 |
+
"learning_rate": 1.1403449199396966e-05,
|
| 701 |
+
"loss": 0.0279,
|
| 702 |
+
"step": 990
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 17.24137931034483,
|
| 706 |
+
"grad_norm": 0.18025921285152435,
|
| 707 |
+
"learning_rate": 1.0116426283528302e-05,
|
| 708 |
+
"loss": 0.0227,
|
| 709 |
+
"step": 1000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 17.413793103448278,
|
| 713 |
+
"grad_norm": 0.17031840980052948,
|
| 714 |
+
"learning_rate": 8.902580598698663e-06,
|
| 715 |
+
"loss": 0.0208,
|
| 716 |
+
"step": 1010
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 17.586206896551722,
|
| 720 |
+
"grad_norm": 0.12430210411548615,
|
| 721 |
+
"learning_rate": 7.7629003773787e-06,
|
| 722 |
+
"loss": 0.0269,
|
| 723 |
+
"step": 1020
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 17.75862068965517,
|
| 727 |
+
"grad_norm": 0.17070575058460236,
|
| 728 |
+
"learning_rate": 6.698313471448547e-06,
|
| 729 |
+
"loss": 0.0209,
|
| 730 |
+
"step": 1030
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 17.93103448275862,
|
| 734 |
+
"grad_norm": 0.17201951146125793,
|
| 735 |
+
"learning_rate": 5.709686596802521e-06,
|
| 736 |
+
"loss": 0.0218,
|
| 737 |
+
"step": 1040
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 18.103448275862068,
|
| 741 |
+
"grad_norm": 0.10487387329339981,
|
| 742 |
+
"learning_rate": 4.797824627726577e-06,
|
| 743 |
+
"loss": 0.0201,
|
| 744 |
+
"step": 1050
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 18.275862068965516,
|
| 748 |
+
"grad_norm": 0.14748798310756683,
|
| 749 |
+
"learning_rate": 3.963469941623288e-06,
|
| 750 |
+
"loss": 0.0202,
|
| 751 |
+
"step": 1060
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 18.448275862068964,
|
| 755 |
+
"grad_norm": 0.1550891101360321,
|
| 756 |
+
"learning_rate": 3.2073018146173805e-06,
|
| 757 |
+
"loss": 0.0202,
|
| 758 |
+
"step": 1070
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 18.620689655172413,
|
| 762 |
+
"grad_norm": 0.14703543484210968,
|
| 763 |
+
"learning_rate": 2.529935868534372e-06,
|
| 764 |
+
"loss": 0.0243,
|
| 765 |
+
"step": 1080
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 18.79310344827586,
|
| 769 |
+
"grad_norm": 0.1516178995370865,
|
| 770 |
+
"learning_rate": 1.9319235697021763e-06,
|
| 771 |
+
"loss": 0.0207,
|
| 772 |
+
"step": 1090
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 18.96551724137931,
|
| 776 |
+
"grad_norm": 0.12201597541570663,
|
| 777 |
+
"learning_rate": 1.4137517799837852e-06,
|
| 778 |
+
"loss": 0.0225,
|
| 779 |
+
"step": 1100
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 19.137931034482758,
|
| 783 |
+
"grad_norm": 0.15553522109985352,
|
| 784 |
+
"learning_rate": 9.758423604068312e-07,
|
| 785 |
+
"loss": 0.0188,
|
| 786 |
+
"step": 1110
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 19.310344827586206,
|
| 790 |
+
"grad_norm": 0.13390740752220154,
|
| 791 |
+
"learning_rate": 6.185518277123214e-07,
|
| 792 |
+
"loss": 0.0214,
|
| 793 |
+
"step": 1120
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 19.482758620689655,
|
| 797 |
+
"grad_norm": 0.17719490826129913,
|
| 798 |
+
"learning_rate": 3.421710641023923e-07,
|
| 799 |
+
"loss": 0.0245,
|
| 800 |
+
"step": 1130
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 19.655172413793103,
|
| 804 |
+
"grad_norm": 0.206907719373703,
|
| 805 |
+
"learning_rate": 1.4692508042337682e-07,
|
| 806 |
+
"loss": 0.0216,
|
| 807 |
+
"step": 1140
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 19.82758620689655,
|
| 811 |
+
"grad_norm": 0.08690398931503296,
|
| 812 |
+
"learning_rate": 3.2972832976918554e-08,
|
| 813 |
+
"loss": 0.0197,
|
| 814 |
+
"step": 1150
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 19.982758620689655,
|
| 818 |
+
"step": 1159,
|
| 819 |
+
"total_flos": 8.16714982017792e+16,
|
| 820 |
+
"train_loss": 0.05862578803655708,
|
| 821 |
+
"train_runtime": 806.3047,
|
| 822 |
+
"train_samples_per_second": 91.995,
|
| 823 |
+
"train_steps_per_second": 1.437
|
| 824 |
+
}
|
| 825 |
+
],
|
| 826 |
+
"logging_steps": 10,
|
| 827 |
+
"max_steps": 1159,
|
| 828 |
+
"num_input_tokens_seen": 0,
|
| 829 |
+
"num_train_epochs": 20,
|
| 830 |
+
"save_steps": 10000,
|
| 831 |
+
"stateful_callbacks": {
|
| 832 |
+
"TrainerControl": {
|
| 833 |
+
"args": {
|
| 834 |
+
"should_epoch_stop": false,
|
| 835 |
+
"should_evaluate": false,
|
| 836 |
+
"should_log": false,
|
| 837 |
+
"should_save": true,
|
| 838 |
+
"should_training_stop": true
|
| 839 |
+
},
|
| 840 |
+
"attributes": {}
|
| 841 |
+
}
|
| 842 |
+
},
|
| 843 |
+
"total_flos": 8.16714982017792e+16,
|
| 844 |
+
"train_batch_size": 64,
|
| 845 |
+
"trial_name": null,
|
| 846 |
+
"trial_params": null
|
| 847 |
+
}
|