End of training
Browse files- all_results.json +15 -0
- eval_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +214 -0
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_accuracy": 0.5416644525230297,
|
| 4 |
+
"eval_loss": 2.2877631187438965,
|
| 5 |
+
"eval_runtime": 4.5703,
|
| 6 |
+
"eval_samples": 5038,
|
| 7 |
+
"eval_samples_per_second": 1102.339,
|
| 8 |
+
"eval_steps_per_second": 2.188,
|
| 9 |
+
"perplexity": 9.852873305063827,
|
| 10 |
+
"train_loss": 2.5528629182903297,
|
| 11 |
+
"train_runtime": 747.1228,
|
| 12 |
+
"train_samples": 52812,
|
| 13 |
+
"train_samples_per_second": 706.872,
|
| 14 |
+
"train_steps_per_second": 11.056
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_accuracy": 0.5416644525230297,
|
| 4 |
+
"eval_loss": 2.2877631187438965,
|
| 5 |
+
"eval_runtime": 4.5703,
|
| 6 |
+
"eval_samples": 5038,
|
| 7 |
+
"eval_samples_per_second": 1102.339,
|
| 8 |
+
"eval_steps_per_second": 2.188,
|
| 9 |
+
"perplexity": 9.852873305063827
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"train_loss": 2.5528629182903297,
|
| 4 |
+
"train_runtime": 747.1228,
|
| 5 |
+
"train_samples": 52812,
|
| 6 |
+
"train_samples_per_second": 706.872,
|
| 7 |
+
"train_steps_per_second": 11.056
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,214 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 8260,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.61,
|
| 13 |
+
"learning_rate": 6.25e-05,
|
| 14 |
+
"loss": 5.8796,
|
| 15 |
+
"step": 500
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 1.0,
|
| 19 |
+
"eval_accuracy": 0.4610925470362255,
|
| 20 |
+
"eval_loss": 3.108325958251953,
|
| 21 |
+
"eval_runtime": 4.3682,
|
| 22 |
+
"eval_samples_per_second": 1153.323,
|
| 23 |
+
"eval_steps_per_second": 2.289,
|
| 24 |
+
"step": 826
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 1.21,
|
| 28 |
+
"learning_rate": 0.000125,
|
| 29 |
+
"loss": 3.1165,
|
| 30 |
+
"step": 1000
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 1.82,
|
| 34 |
+
"learning_rate": 0.0001875,
|
| 35 |
+
"loss": 2.802,
|
| 36 |
+
"step": 1500
|
| 37 |
+
},
|
| 38 |
+
{
|
| 39 |
+
"epoch": 2.0,
|
| 40 |
+
"eval_accuracy": 0.49653186960204804,
|
| 41 |
+
"eval_loss": 2.7454917430877686,
|
| 42 |
+
"eval_runtime": 4.161,
|
| 43 |
+
"eval_samples_per_second": 1210.754,
|
| 44 |
+
"eval_steps_per_second": 2.403,
|
| 45 |
+
"step": 1652
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 2.42,
|
| 49 |
+
"learning_rate": 0.00025,
|
| 50 |
+
"loss": 2.6268,
|
| 51 |
+
"step": 2000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 3.0,
|
| 55 |
+
"eval_accuracy": 0.5115797107338558,
|
| 56 |
+
"eval_loss": 2.573380947113037,
|
| 57 |
+
"eval_runtime": 4.3721,
|
| 58 |
+
"eval_samples_per_second": 1152.294,
|
| 59 |
+
"eval_steps_per_second": 2.287,
|
| 60 |
+
"step": 2478
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 3.03,
|
| 64 |
+
"learning_rate": 0.0003125,
|
| 65 |
+
"loss": 2.5157,
|
| 66 |
+
"step": 2500
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 3.63,
|
| 70 |
+
"learning_rate": 0.000375,
|
| 71 |
+
"loss": 2.4165,
|
| 72 |
+
"step": 3000
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 4.0,
|
| 76 |
+
"eval_accuracy": 0.5211369966209564,
|
| 77 |
+
"eval_loss": 2.4666714668273926,
|
| 78 |
+
"eval_runtime": 4.4473,
|
| 79 |
+
"eval_samples_per_second": 1132.814,
|
| 80 |
+
"eval_steps_per_second": 2.249,
|
| 81 |
+
"step": 3304
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 4.24,
|
| 85 |
+
"learning_rate": 0.00043750000000000006,
|
| 86 |
+
"loss": 2.3502,
|
| 87 |
+
"step": 3500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 4.84,
|
| 91 |
+
"learning_rate": 0.0005,
|
| 92 |
+
"loss": 2.2892,
|
| 93 |
+
"step": 4000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 5.0,
|
| 97 |
+
"eval_accuracy": 0.5287937658050783,
|
| 98 |
+
"eval_loss": 2.394850969314575,
|
| 99 |
+
"eval_runtime": 4.5706,
|
| 100 |
+
"eval_samples_per_second": 1102.258,
|
| 101 |
+
"eval_steps_per_second": 2.188,
|
| 102 |
+
"step": 4130
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 5.45,
|
| 106 |
+
"learning_rate": 0.0005625000000000001,
|
| 107 |
+
"loss": 2.2315,
|
| 108 |
+
"step": 4500
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 6.0,
|
| 112 |
+
"eval_accuracy": 0.5337701187510354,
|
| 113 |
+
"eval_loss": 2.344557523727417,
|
| 114 |
+
"eval_runtime": 4.6212,
|
| 115 |
+
"eval_samples_per_second": 1090.182,
|
| 116 |
+
"eval_steps_per_second": 2.164,
|
| 117 |
+
"step": 4956
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 6.05,
|
| 121 |
+
"learning_rate": 0.000625,
|
| 122 |
+
"loss": 2.2096,
|
| 123 |
+
"step": 5000
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 6.66,
|
| 127 |
+
"learning_rate": 0.0006875,
|
| 128 |
+
"loss": 2.1587,
|
| 129 |
+
"step": 5500
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 7.0,
|
| 133 |
+
"eval_accuracy": 0.5373570314429236,
|
| 134 |
+
"eval_loss": 2.3208389282226562,
|
| 135 |
+
"eval_runtime": 4.2475,
|
| 136 |
+
"eval_samples_per_second": 1186.103,
|
| 137 |
+
"eval_steps_per_second": 2.354,
|
| 138 |
+
"step": 5782
|
| 139 |
+
},
|
| 140 |
+
{
|
| 141 |
+
"epoch": 7.26,
|
| 142 |
+
"learning_rate": 0.00075,
|
| 143 |
+
"loss": 2.139,
|
| 144 |
+
"step": 6000
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 7.87,
|
| 148 |
+
"learning_rate": 0.0008125,
|
| 149 |
+
"loss": 2.1253,
|
| 150 |
+
"step": 6500
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 8.0,
|
| 154 |
+
"eval_accuracy": 0.5394279069622053,
|
| 155 |
+
"eval_loss": 2.3043758869171143,
|
| 156 |
+
"eval_runtime": 4.4837,
|
| 157 |
+
"eval_samples_per_second": 1123.621,
|
| 158 |
+
"eval_steps_per_second": 2.23,
|
| 159 |
+
"step": 6608
|
| 160 |
+
},
|
| 161 |
+
{
|
| 162 |
+
"epoch": 8.47,
|
| 163 |
+
"learning_rate": 0.0008750000000000001,
|
| 164 |
+
"loss": 2.0858,
|
| 165 |
+
"step": 7000
|
| 166 |
+
},
|
| 167 |
+
{
|
| 168 |
+
"epoch": 9.0,
|
| 169 |
+
"eval_accuracy": 0.5403687877641734,
|
| 170 |
+
"eval_loss": 2.2939975261688232,
|
| 171 |
+
"eval_runtime": 4.172,
|
| 172 |
+
"eval_samples_per_second": 1207.562,
|
| 173 |
+
"eval_steps_per_second": 2.397,
|
| 174 |
+
"step": 7434
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 9.08,
|
| 178 |
+
"learning_rate": 0.0009375,
|
| 179 |
+
"loss": 2.0892,
|
| 180 |
+
"step": 7500
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 9.69,
|
| 184 |
+
"learning_rate": 0.001,
|
| 185 |
+
"loss": 2.0556,
|
| 186 |
+
"step": 8000
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 10.0,
|
| 190 |
+
"eval_accuracy": 0.5416644525230297,
|
| 191 |
+
"eval_loss": 2.2877631187438965,
|
| 192 |
+
"eval_runtime": 4.4277,
|
| 193 |
+
"eval_samples_per_second": 1137.831,
|
| 194 |
+
"eval_steps_per_second": 2.258,
|
| 195 |
+
"step": 8260
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 10.0,
|
| 199 |
+
"step": 8260,
|
| 200 |
+
"total_flos": 2562796651806720.0,
|
| 201 |
+
"train_loss": 2.5528629182903297,
|
| 202 |
+
"train_runtime": 747.1228,
|
| 203 |
+
"train_samples_per_second": 706.872,
|
| 204 |
+
"train_steps_per_second": 11.056
|
| 205 |
+
}
|
| 206 |
+
],
|
| 207 |
+
"logging_steps": 500,
|
| 208 |
+
"max_steps": 8260,
|
| 209 |
+
"num_train_epochs": 10,
|
| 210 |
+
"save_steps": 2000,
|
| 211 |
+
"total_flos": 2562796651806720.0,
|
| 212 |
+
"trial_name": null,
|
| 213 |
+
"trial_params": null
|
| 214 |
+
}
|