End of training
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_accuracy": 0.4996186598919624,
|
| 4 |
+
"eval_loss": 2.4745402336120605,
|
| 5 |
+
"eval_runtime": 3.375,
|
| 6 |
+
"eval_samples": 4491,
|
| 7 |
+
"eval_samples_per_second": 1330.683,
|
| 8 |
+
"eval_steps_per_second": 10.667,
|
| 9 |
+
"perplexity": 11.876245559153128,
|
| 10 |
+
"train_loss": 2.566726820065024,
|
| 11 |
+
"train_runtime": 751.1564,
|
| 12 |
+
"train_samples": 46845,
|
| 13 |
+
"train_samples_per_second": 623.638,
|
| 14 |
+
"train_steps_per_second": 38.98
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"eval_accuracy": 0.4996186598919624,
|
| 4 |
+
"eval_loss": 2.4745402336120605,
|
| 5 |
+
"eval_runtime": 3.375,
|
| 6 |
+
"eval_samples": 4491,
|
| 7 |
+
"eval_samples_per_second": 1330.683,
|
| 8 |
+
"eval_steps_per_second": 10.667,
|
| 9 |
+
"perplexity": 11.876245559153128
|
| 10 |
+
}
|
runs/Mar13_12-34-37_phyl-ling-p01.la.utexas.edu/events.out.tfevents.1710352038.phyl-ling-p01.la.utexas.edu.51695.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:995c956f1ddb2d96615673024f2a6e74b05c5586f3ab69ba76bab73971fe240a
|
| 3 |
+
size 417
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 10.0,
|
| 3 |
+
"train_loss": 2.566726820065024,
|
| 4 |
+
"train_runtime": 751.1564,
|
| 5 |
+
"train_samples": 46845,
|
| 6 |
+
"train_samples_per_second": 623.638,
|
| 7 |
+
"train_steps_per_second": 38.98
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,526 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 10.0,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 29280,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.17,
|
| 13 |
+
"grad_norm": 1.187337040901184,
|
| 14 |
+
"learning_rate": 6.25e-05,
|
| 15 |
+
"loss": 6.2023,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.34,
|
| 20 |
+
"grad_norm": 1.0917108058929443,
|
| 21 |
+
"learning_rate": 0.000125,
|
| 22 |
+
"loss": 3.6263,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.51,
|
| 27 |
+
"grad_norm": 1.0571826696395874,
|
| 28 |
+
"learning_rate": 0.0001875,
|
| 29 |
+
"loss": 3.3348,
|
| 30 |
+
"step": 1500
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.68,
|
| 34 |
+
"grad_norm": 0.9771557450294495,
|
| 35 |
+
"learning_rate": 0.00025,
|
| 36 |
+
"loss": 3.1685,
|
| 37 |
+
"step": 2000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.85,
|
| 41 |
+
"grad_norm": 0.9500265717506409,
|
| 42 |
+
"learning_rate": 0.0003125,
|
| 43 |
+
"loss": 3.0457,
|
| 44 |
+
"step": 2500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 1.0,
|
| 48 |
+
"eval_accuracy": 0.4381255950220651,
|
| 49 |
+
"eval_loss": 3.016178607940674,
|
| 50 |
+
"eval_runtime": 3.353,
|
| 51 |
+
"eval_samples_per_second": 1339.396,
|
| 52 |
+
"eval_steps_per_second": 10.737,
|
| 53 |
+
"step": 2928
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 1.02,
|
| 57 |
+
"grad_norm": 0.9907193779945374,
|
| 58 |
+
"learning_rate": 0.000375,
|
| 59 |
+
"loss": 2.9546,
|
| 60 |
+
"step": 3000
|
| 61 |
+
},
|
| 62 |
+
{
|
| 63 |
+
"epoch": 1.2,
|
| 64 |
+
"grad_norm": 0.9292730093002319,
|
| 65 |
+
"learning_rate": 0.00043750000000000006,
|
| 66 |
+
"loss": 2.8611,
|
| 67 |
+
"step": 3500
|
| 68 |
+
},
|
| 69 |
+
{
|
| 70 |
+
"epoch": 1.37,
|
| 71 |
+
"grad_norm": 0.8982778787612915,
|
| 72 |
+
"learning_rate": 0.0005,
|
| 73 |
+
"loss": 2.8115,
|
| 74 |
+
"step": 4000
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 1.54,
|
| 78 |
+
"grad_norm": 0.899147093296051,
|
| 79 |
+
"learning_rate": 0.0005625000000000001,
|
| 80 |
+
"loss": 2.7697,
|
| 81 |
+
"step": 4500
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 1.71,
|
| 85 |
+
"grad_norm": 0.889224648475647,
|
| 86 |
+
"learning_rate": 0.000625,
|
| 87 |
+
"loss": 2.7455,
|
| 88 |
+
"step": 5000
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.88,
|
| 92 |
+
"grad_norm": 0.8336099982261658,
|
| 93 |
+
"learning_rate": 0.0006875,
|
| 94 |
+
"loss": 2.7215,
|
| 95 |
+
"step": 5500
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 2.0,
|
| 99 |
+
"eval_accuracy": 0.4596735027360057,
|
| 100 |
+
"eval_loss": 2.7846643924713135,
|
| 101 |
+
"eval_runtime": 3.384,
|
| 102 |
+
"eval_samples_per_second": 1327.111,
|
| 103 |
+
"eval_steps_per_second": 10.638,
|
| 104 |
+
"step": 5856
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 2.05,
|
| 108 |
+
"grad_norm": 0.8637300729751587,
|
| 109 |
+
"learning_rate": 0.00075,
|
| 110 |
+
"loss": 2.6694,
|
| 111 |
+
"step": 6000
|
| 112 |
+
},
|
| 113 |
+
{
|
| 114 |
+
"epoch": 2.22,
|
| 115 |
+
"grad_norm": 0.7789953351020813,
|
| 116 |
+
"learning_rate": 0.0008125,
|
| 117 |
+
"loss": 2.6306,
|
| 118 |
+
"step": 6500
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 2.39,
|
| 122 |
+
"grad_norm": 0.7371299862861633,
|
| 123 |
+
"learning_rate": 0.0008750000000000001,
|
| 124 |
+
"loss": 2.6088,
|
| 125 |
+
"step": 7000
|
| 126 |
+
},
|
| 127 |
+
{
|
| 128 |
+
"epoch": 2.56,
|
| 129 |
+
"grad_norm": 0.6360913515090942,
|
| 130 |
+
"learning_rate": 0.0009375,
|
| 131 |
+
"loss": 2.6008,
|
| 132 |
+
"step": 7500
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 2.73,
|
| 136 |
+
"grad_norm": 0.6919972896575928,
|
| 137 |
+
"learning_rate": 0.001,
|
| 138 |
+
"loss": 2.5956,
|
| 139 |
+
"step": 8000
|
| 140 |
+
},
|
| 141 |
+
{
|
| 142 |
+
"epoch": 2.9,
|
| 143 |
+
"grad_norm": 0.653573751449585,
|
| 144 |
+
"learning_rate": 0.0010625,
|
| 145 |
+
"loss": 2.5825,
|
| 146 |
+
"step": 8500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 3.0,
|
| 150 |
+
"eval_accuracy": 0.47033349288252796,
|
| 151 |
+
"eval_loss": 2.6924686431884766,
|
| 152 |
+
"eval_runtime": 3.3904,
|
| 153 |
+
"eval_samples_per_second": 1324.616,
|
| 154 |
+
"eval_steps_per_second": 10.618,
|
| 155 |
+
"step": 8784
|
| 156 |
+
},
|
| 157 |
+
{
|
| 158 |
+
"epoch": 3.07,
|
| 159 |
+
"grad_norm": 0.621619701385498,
|
| 160 |
+
"learning_rate": 0.0011250000000000001,
|
| 161 |
+
"loss": 2.5387,
|
| 162 |
+
"step": 9000
|
| 163 |
+
},
|
| 164 |
+
{
|
| 165 |
+
"epoch": 3.24,
|
| 166 |
+
"grad_norm": 0.6471651792526245,
|
| 167 |
+
"learning_rate": 0.0011875,
|
| 168 |
+
"loss": 2.5135,
|
| 169 |
+
"step": 9500
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 3.42,
|
| 173 |
+
"grad_norm": 0.5801098942756653,
|
| 174 |
+
"learning_rate": 0.00125,
|
| 175 |
+
"loss": 2.5157,
|
| 176 |
+
"step": 10000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 3.59,
|
| 180 |
+
"grad_norm": 0.5154350399971008,
|
| 181 |
+
"learning_rate": 0.0013125,
|
| 182 |
+
"loss": 2.5178,
|
| 183 |
+
"step": 10500
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 3.76,
|
| 187 |
+
"grad_norm": 0.4794241189956665,
|
| 188 |
+
"learning_rate": 0.001375,
|
| 189 |
+
"loss": 2.5135,
|
| 190 |
+
"step": 11000
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 3.93,
|
| 194 |
+
"grad_norm": 0.4597519040107727,
|
| 195 |
+
"learning_rate": 0.0014375000000000002,
|
| 196 |
+
"loss": 2.5059,
|
| 197 |
+
"step": 11500
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 4.0,
|
| 201 |
+
"eval_accuracy": 0.47582479043826936,
|
| 202 |
+
"eval_loss": 2.6392264366149902,
|
| 203 |
+
"eval_runtime": 3.4286,
|
| 204 |
+
"eval_samples_per_second": 1309.863,
|
| 205 |
+
"eval_steps_per_second": 10.5,
|
| 206 |
+
"step": 11712
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 4.1,
|
| 210 |
+
"grad_norm": 0.4645042419433594,
|
| 211 |
+
"learning_rate": 0.0015,
|
| 212 |
+
"loss": 2.4568,
|
| 213 |
+
"step": 12000
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 4.27,
|
| 217 |
+
"grad_norm": 0.46030470728874207,
|
| 218 |
+
"learning_rate": 0.0015625,
|
| 219 |
+
"loss": 2.4486,
|
| 220 |
+
"step": 12500
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 4.44,
|
| 224 |
+
"grad_norm": 0.44470083713531494,
|
| 225 |
+
"learning_rate": 0.001625,
|
| 226 |
+
"loss": 2.4657,
|
| 227 |
+
"step": 13000
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 4.61,
|
| 231 |
+
"grad_norm": 0.4096473157405853,
|
| 232 |
+
"learning_rate": 0.0016875,
|
| 233 |
+
"loss": 2.4596,
|
| 234 |
+
"step": 13500
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 4.78,
|
| 238 |
+
"grad_norm": 0.3900606334209442,
|
| 239 |
+
"learning_rate": 0.0017500000000000003,
|
| 240 |
+
"loss": 2.4585,
|
| 241 |
+
"step": 14000
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 4.95,
|
| 245 |
+
"grad_norm": 0.40039312839508057,
|
| 246 |
+
"learning_rate": 0.0018124999999999999,
|
| 247 |
+
"loss": 2.4667,
|
| 248 |
+
"step": 14500
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 5.0,
|
| 252 |
+
"eval_accuracy": 0.480080020057613,
|
| 253 |
+
"eval_loss": 2.6032490730285645,
|
| 254 |
+
"eval_runtime": 3.3841,
|
| 255 |
+
"eval_samples_per_second": 1327.1,
|
| 256 |
+
"eval_steps_per_second": 10.638,
|
| 257 |
+
"step": 14640
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 5.12,
|
| 261 |
+
"grad_norm": 0.3943752348423004,
|
| 262 |
+
"learning_rate": 0.001875,
|
| 263 |
+
"loss": 2.4054,
|
| 264 |
+
"step": 15000
|
| 265 |
+
},
|
| 266 |
+
{
|
| 267 |
+
"epoch": 5.29,
|
| 268 |
+
"grad_norm": 0.4194723069667816,
|
| 269 |
+
"learning_rate": 0.0019375000000000002,
|
| 270 |
+
"loss": 2.4006,
|
| 271 |
+
"step": 15500
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 5.46,
|
| 275 |
+
"grad_norm": 0.3874049484729767,
|
| 276 |
+
"learning_rate": 0.002,
|
| 277 |
+
"loss": 2.4215,
|
| 278 |
+
"step": 16000
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 5.64,
|
| 282 |
+
"grad_norm": 0.4187428057193756,
|
| 283 |
+
"learning_rate": 0.0020625,
|
| 284 |
+
"loss": 2.409,
|
| 285 |
+
"step": 16500
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 5.81,
|
| 289 |
+
"grad_norm": 0.3771882951259613,
|
| 290 |
+
"learning_rate": 0.002125,
|
| 291 |
+
"loss": 2.4208,
|
| 292 |
+
"step": 17000
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 5.98,
|
| 296 |
+
"grad_norm": 0.34806564450263977,
|
| 297 |
+
"learning_rate": 0.0021874999999999998,
|
| 298 |
+
"loss": 2.4261,
|
| 299 |
+
"step": 17500
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 6.0,
|
| 303 |
+
"eval_accuracy": 0.4827099518371827,
|
| 304 |
+
"eval_loss": 2.585068941116333,
|
| 305 |
+
"eval_runtime": 3.4455,
|
| 306 |
+
"eval_samples_per_second": 1303.458,
|
| 307 |
+
"eval_steps_per_second": 10.449,
|
| 308 |
+
"step": 17568
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 6.15,
|
| 312 |
+
"grad_norm": 0.38157713413238525,
|
| 313 |
+
"learning_rate": 0.0022500000000000003,
|
| 314 |
+
"loss": 2.3536,
|
| 315 |
+
"step": 18000
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 6.32,
|
| 319 |
+
"grad_norm": 0.35192397236824036,
|
| 320 |
+
"learning_rate": 0.0023125000000000003,
|
| 321 |
+
"loss": 2.3635,
|
| 322 |
+
"step": 18500
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 6.49,
|
| 326 |
+
"grad_norm": 0.36528274416923523,
|
| 327 |
+
"learning_rate": 0.002375,
|
| 328 |
+
"loss": 2.3799,
|
| 329 |
+
"step": 19000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 6.66,
|
| 333 |
+
"grad_norm": 0.3123536705970764,
|
| 334 |
+
"learning_rate": 0.0024375,
|
| 335 |
+
"loss": 2.3808,
|
| 336 |
+
"step": 19500
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 6.83,
|
| 340 |
+
"grad_norm": 0.32904407382011414,
|
| 341 |
+
"learning_rate": 0.0025,
|
| 342 |
+
"loss": 2.3888,
|
| 343 |
+
"step": 20000
|
| 344 |
+
},
|
| 345 |
+
{
|
| 346 |
+
"epoch": 7.0,
|
| 347 |
+
"eval_accuracy": 0.484838443290781,
|
| 348 |
+
"eval_loss": 2.566373348236084,
|
| 349 |
+
"eval_runtime": 3.3915,
|
| 350 |
+
"eval_samples_per_second": 1324.208,
|
| 351 |
+
"eval_steps_per_second": 10.615,
|
| 352 |
+
"step": 20496
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 7.0,
|
| 356 |
+
"grad_norm": 0.316089928150177,
|
| 357 |
+
"learning_rate": 0.0025625,
|
| 358 |
+
"loss": 2.397,
|
| 359 |
+
"step": 20500
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 7.17,
|
| 363 |
+
"grad_norm": 0.36722826957702637,
|
| 364 |
+
"learning_rate": 0.002625,
|
| 365 |
+
"loss": 2.3088,
|
| 366 |
+
"step": 21000
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 7.34,
|
| 370 |
+
"grad_norm": 0.3584025800228119,
|
| 371 |
+
"learning_rate": 0.0026875000000000002,
|
| 372 |
+
"loss": 2.3392,
|
| 373 |
+
"step": 21500
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 7.51,
|
| 377 |
+
"grad_norm": 0.39147356152534485,
|
| 378 |
+
"learning_rate": 0.00275,
|
| 379 |
+
"loss": 2.3481,
|
| 380 |
+
"step": 22000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 7.68,
|
| 384 |
+
"grad_norm": 0.30665042996406555,
|
| 385 |
+
"learning_rate": 0.0028125,
|
| 386 |
+
"loss": 2.3524,
|
| 387 |
+
"step": 22500
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 7.86,
|
| 391 |
+
"grad_norm": 0.34735408425331116,
|
| 392 |
+
"learning_rate": 0.0028750000000000004,
|
| 393 |
+
"loss": 2.3623,
|
| 394 |
+
"step": 23000
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 8.0,
|
| 398 |
+
"eval_accuracy": 0.4860008731373508,
|
| 399 |
+
"eval_loss": 2.5573480129241943,
|
| 400 |
+
"eval_runtime": 3.38,
|
| 401 |
+
"eval_samples_per_second": 1328.716,
|
| 402 |
+
"eval_steps_per_second": 10.651,
|
| 403 |
+
"step": 23424
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 8.03,
|
| 407 |
+
"grad_norm": 0.3066130578517914,
|
| 408 |
+
"learning_rate": 0.0029375,
|
| 409 |
+
"loss": 2.359,
|
| 410 |
+
"step": 23500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 8.2,
|
| 414 |
+
"grad_norm": 0.4078517556190491,
|
| 415 |
+
"learning_rate": 0.003,
|
| 416 |
+
"loss": 2.285,
|
| 417 |
+
"step": 24000
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 8.37,
|
| 421 |
+
"grad_norm": 0.2971705198287964,
|
| 422 |
+
"learning_rate": 0.002715909090909091,
|
| 423 |
+
"loss": 2.3045,
|
| 424 |
+
"step": 24500
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 8.54,
|
| 428 |
+
"grad_norm": 0.30169880390167236,
|
| 429 |
+
"learning_rate": 0.0024318181818181817,
|
| 430 |
+
"loss": 2.3081,
|
| 431 |
+
"step": 25000
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 8.71,
|
| 435 |
+
"grad_norm": 0.30579808354377747,
|
| 436 |
+
"learning_rate": 0.002147727272727273,
|
| 437 |
+
"loss": 2.2906,
|
| 438 |
+
"step": 25500
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 8.88,
|
| 442 |
+
"grad_norm": 0.3257807791233063,
|
| 443 |
+
"learning_rate": 0.0018636363636363638,
|
| 444 |
+
"loss": 2.2903,
|
| 445 |
+
"step": 26000
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 9.0,
|
| 449 |
+
"eval_accuracy": 0.4940852834277479,
|
| 450 |
+
"eval_loss": 2.502375841140747,
|
| 451 |
+
"eval_runtime": 3.3782,
|
| 452 |
+
"eval_samples_per_second": 1329.388,
|
| 453 |
+
"eval_steps_per_second": 10.656,
|
| 454 |
+
"step": 26352
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 9.05,
|
| 458 |
+
"grad_norm": 0.27855226397514343,
|
| 459 |
+
"learning_rate": 0.0015795454545454546,
|
| 460 |
+
"loss": 2.2424,
|
| 461 |
+
"step": 26500
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 9.22,
|
| 465 |
+
"grad_norm": 0.2956109941005707,
|
| 466 |
+
"learning_rate": 0.0012954545454545456,
|
| 467 |
+
"loss": 2.1635,
|
| 468 |
+
"step": 27000
|
| 469 |
+
},
|
| 470 |
+
{
|
| 471 |
+
"epoch": 9.39,
|
| 472 |
+
"grad_norm": 0.3226345479488373,
|
| 473 |
+
"learning_rate": 0.0010113636363636364,
|
| 474 |
+
"loss": 2.1661,
|
| 475 |
+
"step": 27500
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 9.56,
|
| 479 |
+
"grad_norm": 0.3127693831920624,
|
| 480 |
+
"learning_rate": 0.0007272727272727273,
|
| 481 |
+
"loss": 2.1605,
|
| 482 |
+
"step": 28000
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 9.73,
|
| 486 |
+
"grad_norm": 0.3304063379764557,
|
| 487 |
+
"learning_rate": 0.0004431818181818182,
|
| 488 |
+
"loss": 2.1557,
|
| 489 |
+
"step": 28500
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 9.9,
|
| 493 |
+
"grad_norm": 0.3265271782875061,
|
| 494 |
+
"learning_rate": 0.0001590909090909091,
|
| 495 |
+
"loss": 2.135,
|
| 496 |
+
"step": 29000
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 10.0,
|
| 500 |
+
"eval_accuracy": 0.4996186598919624,
|
| 501 |
+
"eval_loss": 2.4745402336120605,
|
| 502 |
+
"eval_runtime": 3.376,
|
| 503 |
+
"eval_samples_per_second": 1330.26,
|
| 504 |
+
"eval_steps_per_second": 10.663,
|
| 505 |
+
"step": 29280
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 10.0,
|
| 509 |
+
"step": 29280,
|
| 510 |
+
"total_flos": 2273237316403200.0,
|
| 511 |
+
"train_loss": 2.566726820065024,
|
| 512 |
+
"train_runtime": 751.1564,
|
| 513 |
+
"train_samples_per_second": 623.638,
|
| 514 |
+
"train_steps_per_second": 38.98
|
| 515 |
+
}
|
| 516 |
+
],
|
| 517 |
+
"logging_steps": 500,
|
| 518 |
+
"max_steps": 29280,
|
| 519 |
+
"num_input_tokens_seen": 0,
|
| 520 |
+
"num_train_epochs": 10,
|
| 521 |
+
"save_steps": 2000,
|
| 522 |
+
"total_flos": 2273237316403200.0,
|
| 523 |
+
"train_batch_size": 16,
|
| 524 |
+
"trial_name": null,
|
| 525 |
+
"trial_params": null
|
| 526 |
+
}
|