Commit
•
1d8e01b
1
Parent(s):
0884c59
End of training
Browse files- all_results.json +15 -0
- eval_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +589 -0
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 30.0,
|
| 3 |
+
"eval_accuracy": 0.8791137178233952,
|
| 4 |
+
"eval_loss": 0.48096638917922974,
|
| 5 |
+
"eval_runtime": 1.2644,
|
| 6 |
+
"eval_samples": 30,
|
| 7 |
+
"eval_samples_per_second": 23.726,
|
| 8 |
+
"eval_steps_per_second": 11.863,
|
| 9 |
+
"perplexity": 1.6176369138835944,
|
| 10 |
+
"train_loss": 0.9465416105698796,
|
| 11 |
+
"train_runtime": 17544.8246,
|
| 12 |
+
"train_samples": 3163,
|
| 13 |
+
"train_samples_per_second": 5.408,
|
| 14 |
+
"train_steps_per_second": 2.705
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 30.0,
|
| 3 |
+
"eval_accuracy": 0.8791137178233952,
|
| 4 |
+
"eval_loss": 0.48096638917922974,
|
| 5 |
+
"eval_runtime": 1.2644,
|
| 6 |
+
"eval_samples": 30,
|
| 7 |
+
"eval_samples_per_second": 23.726,
|
| 8 |
+
"eval_steps_per_second": 11.863,
|
| 9 |
+
"perplexity": 1.6176369138835944
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 30.0,
|
| 3 |
+
"train_loss": 0.9465416105698796,
|
| 4 |
+
"train_runtime": 17544.8246,
|
| 5 |
+
"train_samples": 3163,
|
| 6 |
+
"train_samples_per_second": 5.408,
|
| 7 |
+
"train_steps_per_second": 2.705
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,589 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 30.0,
|
| 5 |
+
"global_step": 47460,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.32,
|
| 12 |
+
"learning_rate": 4.94732406236831e-05,
|
| 13 |
+
"loss": 1.2748,
|
| 14 |
+
"step": 500
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.63,
|
| 18 |
+
"learning_rate": 4.8946481247366205e-05,
|
| 19 |
+
"loss": 1.2878,
|
| 20 |
+
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.95,
|
| 24 |
+
"learning_rate": 4.8419721871049303e-05,
|
| 25 |
+
"loss": 1.2988,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 1.26,
|
| 30 |
+
"learning_rate": 4.789296249473241e-05,
|
| 31 |
+
"loss": 1.2645,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 1.58,
|
| 36 |
+
"learning_rate": 4.7366203118415506e-05,
|
| 37 |
+
"loss": 1.2449,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 1.9,
|
| 42 |
+
"learning_rate": 4.683944374209861e-05,
|
| 43 |
+
"loss": 1.2536,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 2.21,
|
| 48 |
+
"learning_rate": 4.631268436578171e-05,
|
| 49 |
+
"loss": 1.2103,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 2.53,
|
| 54 |
+
"learning_rate": 4.5785924989464814e-05,
|
| 55 |
+
"loss": 1.2159,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 2.84,
|
| 60 |
+
"learning_rate": 4.525916561314791e-05,
|
| 61 |
+
"loss": 1.2245,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 3.16,
|
| 66 |
+
"learning_rate": 4.473240623683102e-05,
|
| 67 |
+
"loss": 1.1803,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 3.48,
|
| 72 |
+
"learning_rate": 4.420564686051412e-05,
|
| 73 |
+
"loss": 1.1739,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 3.79,
|
| 78 |
+
"learning_rate": 4.367888748419722e-05,
|
| 79 |
+
"loss": 1.1882,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 4.11,
|
| 84 |
+
"learning_rate": 4.3152128107880325e-05,
|
| 85 |
+
"loss": 1.169,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 4.42,
|
| 90 |
+
"learning_rate": 4.262536873156342e-05,
|
| 91 |
+
"loss": 1.1294,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 4.74,
|
| 96 |
+
"learning_rate": 4.209860935524653e-05,
|
| 97 |
+
"loss": 1.1658,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 5.06,
|
| 102 |
+
"learning_rate": 4.1571849978929626e-05,
|
| 103 |
+
"loss": 1.1395,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 5.37,
|
| 108 |
+
"learning_rate": 4.104509060261273e-05,
|
| 109 |
+
"loss": 1.0973,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 5.69,
|
| 114 |
+
"learning_rate": 4.051833122629583e-05,
|
| 115 |
+
"loss": 1.1112,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 6.01,
|
| 120 |
+
"learning_rate": 3.9991571849978934e-05,
|
| 121 |
+
"loss": 1.1368,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 6.32,
|
| 126 |
+
"learning_rate": 3.946481247366203e-05,
|
| 127 |
+
"loss": 1.0661,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 6.64,
|
| 132 |
+
"learning_rate": 3.893805309734514e-05,
|
| 133 |
+
"loss": 1.0824,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 6.95,
|
| 138 |
+
"learning_rate": 3.8411293721028235e-05,
|
| 139 |
+
"loss": 1.1049,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 7.27,
|
| 144 |
+
"learning_rate": 3.788453434471134e-05,
|
| 145 |
+
"loss": 1.0503,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 7.59,
|
| 150 |
+
"learning_rate": 3.735777496839444e-05,
|
| 151 |
+
"loss": 1.0634,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 7.9,
|
| 156 |
+
"learning_rate": 3.6831015592077536e-05,
|
| 157 |
+
"loss": 1.0592,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 8.22,
|
| 162 |
+
"learning_rate": 3.630425621576064e-05,
|
| 163 |
+
"loss": 1.0401,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 8.53,
|
| 168 |
+
"learning_rate": 3.577749683944374e-05,
|
| 169 |
+
"loss": 1.0138,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 8.85,
|
| 174 |
+
"learning_rate": 3.5250737463126844e-05,
|
| 175 |
+
"loss": 1.0445,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 9.17,
|
| 180 |
+
"learning_rate": 3.472397808680995e-05,
|
| 181 |
+
"loss": 1.0185,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 9.48,
|
| 186 |
+
"learning_rate": 3.419721871049305e-05,
|
| 187 |
+
"loss": 0.9941,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 9.8,
|
| 192 |
+
"learning_rate": 3.367045933417615e-05,
|
| 193 |
+
"loss": 1.0206,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 10.11,
|
| 198 |
+
"learning_rate": 3.3143699957859256e-05,
|
| 199 |
+
"loss": 1.0188,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 10.43,
|
| 204 |
+
"learning_rate": 3.2616940581542354e-05,
|
| 205 |
+
"loss": 0.972,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 10.75,
|
| 210 |
+
"learning_rate": 3.209018120522546e-05,
|
| 211 |
+
"loss": 0.9909,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 11.06,
|
| 216 |
+
"learning_rate": 3.156342182890856e-05,
|
| 217 |
+
"loss": 0.9792,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 11.38,
|
| 222 |
+
"learning_rate": 3.1036662452591655e-05,
|
| 223 |
+
"loss": 0.9532,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 11.69,
|
| 228 |
+
"learning_rate": 3.050990307627476e-05,
|
| 229 |
+
"loss": 0.9616,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 12.01,
|
| 234 |
+
"learning_rate": 2.998314369995786e-05,
|
| 235 |
+
"loss": 0.9773,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 12.33,
|
| 240 |
+
"learning_rate": 2.9456384323640963e-05,
|
| 241 |
+
"loss": 0.9291,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 12.64,
|
| 246 |
+
"learning_rate": 2.8929624947324065e-05,
|
| 247 |
+
"loss": 0.9526,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 12.96,
|
| 252 |
+
"learning_rate": 2.8402865571007166e-05,
|
| 253 |
+
"loss": 0.9516,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 13.27,
|
| 258 |
+
"learning_rate": 2.7876106194690264e-05,
|
| 259 |
+
"loss": 0.9112,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 13.59,
|
| 264 |
+
"learning_rate": 2.7349346818373366e-05,
|
| 265 |
+
"loss": 0.9198,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 13.91,
|
| 270 |
+
"learning_rate": 2.6822587442056467e-05,
|
| 271 |
+
"loss": 0.9436,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 14.22,
|
| 276 |
+
"learning_rate": 2.629582806573957e-05,
|
| 277 |
+
"loss": 0.9024,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 14.54,
|
| 282 |
+
"learning_rate": 2.576906868942267e-05,
|
| 283 |
+
"loss": 0.9076,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 14.85,
|
| 288 |
+
"learning_rate": 2.524230931310577e-05,
|
| 289 |
+
"loss": 0.9142,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 15.17,
|
| 294 |
+
"learning_rate": 2.4715549936788876e-05,
|
| 295 |
+
"loss": 0.9001,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 15.49,
|
| 300 |
+
"learning_rate": 2.4188790560471978e-05,
|
| 301 |
+
"loss": 0.8884,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 15.8,
|
| 306 |
+
"learning_rate": 2.366203118415508e-05,
|
| 307 |
+
"loss": 0.8909,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 16.12,
|
| 312 |
+
"learning_rate": 2.313527180783818e-05,
|
| 313 |
+
"loss": 0.8959,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 16.43,
|
| 318 |
+
"learning_rate": 2.2608512431521282e-05,
|
| 319 |
+
"loss": 0.8557,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 16.75,
|
| 324 |
+
"learning_rate": 2.2081753055204384e-05,
|
| 325 |
+
"loss": 0.8859,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 17.07,
|
| 330 |
+
"learning_rate": 2.1554993678887485e-05,
|
| 331 |
+
"loss": 0.8828,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 17.38,
|
| 336 |
+
"learning_rate": 2.1028234302570587e-05,
|
| 337 |
+
"loss": 0.8477,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 17.7,
|
| 342 |
+
"learning_rate": 2.0501474926253688e-05,
|
| 343 |
+
"loss": 0.8648,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 18.02,
|
| 348 |
+
"learning_rate": 1.997471554993679e-05,
|
| 349 |
+
"loss": 0.8785,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 18.33,
|
| 354 |
+
"learning_rate": 1.944795617361989e-05,
|
| 355 |
+
"loss": 0.8449,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 18.65,
|
| 360 |
+
"learning_rate": 1.8921196797302992e-05,
|
| 361 |
+
"loss": 0.8559,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 18.96,
|
| 366 |
+
"learning_rate": 1.8394437420986094e-05,
|
| 367 |
+
"loss": 0.8546,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 19.28,
|
| 372 |
+
"learning_rate": 1.7867678044669195e-05,
|
| 373 |
+
"loss": 0.8199,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 19.6,
|
| 378 |
+
"learning_rate": 1.7340918668352297e-05,
|
| 379 |
+
"loss": 0.8308,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 19.91,
|
| 384 |
+
"learning_rate": 1.6814159292035402e-05,
|
| 385 |
+
"loss": 0.8645,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 20.23,
|
| 390 |
+
"learning_rate": 1.62873999157185e-05,
|
| 391 |
+
"loss": 0.8307,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 20.54,
|
| 396 |
+
"learning_rate": 1.57606405394016e-05,
|
| 397 |
+
"loss": 0.8243,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 20.86,
|
| 402 |
+
"learning_rate": 1.5233881163084704e-05,
|
| 403 |
+
"loss": 0.8225,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 21.18,
|
| 408 |
+
"learning_rate": 1.4707121786767806e-05,
|
| 409 |
+
"loss": 0.8214,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 21.49,
|
| 414 |
+
"learning_rate": 1.4180362410450906e-05,
|
| 415 |
+
"loss": 0.8093,
|
| 416 |
+
"step": 34000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 21.81,
|
| 420 |
+
"learning_rate": 1.3653603034134007e-05,
|
| 421 |
+
"loss": 0.817,
|
| 422 |
+
"step": 34500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 22.12,
|
| 426 |
+
"learning_rate": 1.3126843657817109e-05,
|
| 427 |
+
"loss": 0.8373,
|
| 428 |
+
"step": 35000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 22.44,
|
| 432 |
+
"learning_rate": 1.260008428150021e-05,
|
| 433 |
+
"loss": 0.8028,
|
| 434 |
+
"step": 35500
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 22.76,
|
| 438 |
+
"learning_rate": 1.2073324905183313e-05,
|
| 439 |
+
"loss": 0.8156,
|
| 440 |
+
"step": 36000
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 23.07,
|
| 444 |
+
"learning_rate": 1.1546565528866415e-05,
|
| 445 |
+
"loss": 0.8105,
|
| 446 |
+
"step": 36500
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 23.39,
|
| 450 |
+
"learning_rate": 1.1019806152549515e-05,
|
| 451 |
+
"loss": 0.8085,
|
| 452 |
+
"step": 37000
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 23.7,
|
| 456 |
+
"learning_rate": 1.0493046776232618e-05,
|
| 457 |
+
"loss": 0.8071,
|
| 458 |
+
"step": 37500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 24.02,
|
| 462 |
+
"learning_rate": 9.96628739991572e-06,
|
| 463 |
+
"loss": 0.7915,
|
| 464 |
+
"step": 38000
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 24.34,
|
| 468 |
+
"learning_rate": 9.43952802359882e-06,
|
| 469 |
+
"loss": 0.7973,
|
| 470 |
+
"step": 38500
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 24.65,
|
| 474 |
+
"learning_rate": 8.912768647281922e-06,
|
| 475 |
+
"loss": 0.7963,
|
| 476 |
+
"step": 39000
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 24.97,
|
| 480 |
+
"learning_rate": 8.386009270965024e-06,
|
| 481 |
+
"loss": 0.7947,
|
| 482 |
+
"step": 39500
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 25.28,
|
| 486 |
+
"learning_rate": 7.859249894648125e-06,
|
| 487 |
+
"loss": 0.7929,
|
| 488 |
+
"step": 40000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 25.6,
|
| 492 |
+
"learning_rate": 7.3324905183312265e-06,
|
| 493 |
+
"loss": 0.794,
|
| 494 |
+
"step": 40500
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 25.92,
|
| 498 |
+
"learning_rate": 6.805731142014328e-06,
|
| 499 |
+
"loss": 0.7901,
|
| 500 |
+
"step": 41000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 26.23,
|
| 504 |
+
"learning_rate": 6.2789717656974295e-06,
|
| 505 |
+
"loss": 0.7732,
|
| 506 |
+
"step": 41500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 26.55,
|
| 510 |
+
"learning_rate": 5.752212389380531e-06,
|
| 511 |
+
"loss": 0.7862,
|
| 512 |
+
"step": 42000
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 26.86,
|
| 516 |
+
"learning_rate": 5.225453013063633e-06,
|
| 517 |
+
"loss": 0.7973,
|
| 518 |
+
"step": 42500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 27.18,
|
| 522 |
+
"learning_rate": 4.698693636746735e-06,
|
| 523 |
+
"loss": 0.7813,
|
| 524 |
+
"step": 43000
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 27.5,
|
| 528 |
+
"learning_rate": 4.171934260429835e-06,
|
| 529 |
+
"loss": 0.7834,
|
| 530 |
+
"step": 43500
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 27.81,
|
| 534 |
+
"learning_rate": 3.6451748841129377e-06,
|
| 535 |
+
"loss": 0.785,
|
| 536 |
+
"step": 44000
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 28.13,
|
| 540 |
+
"learning_rate": 3.1184155077960387e-06,
|
| 541 |
+
"loss": 0.7904,
|
| 542 |
+
"step": 44500
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 28.45,
|
| 546 |
+
"learning_rate": 2.59165613147914e-06,
|
| 547 |
+
"loss": 0.7809,
|
| 548 |
+
"step": 45000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 28.76,
|
| 552 |
+
"learning_rate": 2.064896755162242e-06,
|
| 553 |
+
"loss": 0.7804,
|
| 554 |
+
"step": 45500
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 29.08,
|
| 558 |
+
"learning_rate": 1.5381373788453435e-06,
|
| 559 |
+
"loss": 0.7755,
|
| 560 |
+
"step": 46000
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 29.39,
|
| 564 |
+
"learning_rate": 1.0113780025284452e-06,
|
| 565 |
+
"loss": 0.7789,
|
| 566 |
+
"step": 46500
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 29.71,
|
| 570 |
+
"learning_rate": 4.846186262115466e-07,
|
| 571 |
+
"loss": 0.7694,
|
| 572 |
+
"step": 47000
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 30.0,
|
| 576 |
+
"step": 47460,
|
| 577 |
+
"total_flos": 4.958800183296e+16,
|
| 578 |
+
"train_loss": 0.9465416105698796,
|
| 579 |
+
"train_runtime": 17544.8246,
|
| 580 |
+
"train_samples_per_second": 5.408,
|
| 581 |
+
"train_steps_per_second": 2.705
|
| 582 |
+
}
|
| 583 |
+
],
|
| 584 |
+
"max_steps": 47460,
|
| 585 |
+
"num_train_epochs": 30,
|
| 586 |
+
"total_flos": 4.958800183296e+16,
|
| 587 |
+
"trial_name": null,
|
| 588 |
+
"trial_params": null
|
| 589 |
+
}
|