defect
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"eval_accuracy": 1.0,
|
| 4 |
+
"eval_loss": 0.0012111830292269588,
|
| 5 |
+
"eval_runtime": 4.3953,
|
| 6 |
+
"eval_samples_per_second": 1.82,
|
| 7 |
+
"eval_steps_per_second": 0.228,
|
| 8 |
+
"total_flos": 2.789711626125312e+17,
|
| 9 |
+
"train_loss": 0.008226247422862798,
|
| 10 |
+
"train_runtime": 8928.1619,
|
| 11 |
+
"train_samples_per_second": 0.403,
|
| 12 |
+
"train_steps_per_second": 0.045
|
| 13 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"eval_accuracy": 1.0,
|
| 4 |
+
"eval_loss": 0.0012111830292269588,
|
| 5 |
+
"eval_runtime": 4.3953,
|
| 6 |
+
"eval_samples_per_second": 1.82,
|
| 7 |
+
"eval_steps_per_second": 0.228
|
| 8 |
+
}
|
runs/May24_14-33-43_ddd353ae11b5/events.out.tfevents.1684947836.ddd353ae11b5.383.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c785097dbdd6dbc0bead8dc76aabae87c6ba3e31ad63a4ec4183a4637f60c34e
|
| 3 |
+
size 411
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 50.0,
|
| 3 |
+
"total_flos": 2.789711626125312e+17,
|
| 4 |
+
"train_loss": 0.008226247422862798,
|
| 5 |
+
"train_runtime": 8928.1619,
|
| 6 |
+
"train_samples_per_second": 0.403,
|
| 7 |
+
"train_steps_per_second": 0.045
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,625 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.0012111830292269588,
|
| 3 |
+
"best_model_checkpoint": "/leaves/checkpoint-400",
|
| 4 |
+
"epoch": 50.0,
|
| 5 |
+
"global_step": 400,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 1.25,
|
| 12 |
+
"learning_rate": 0.000195,
|
| 13 |
+
"loss": 0.2249,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.25,
|
| 18 |
+
"eval_accuracy": 1.0,
|
| 19 |
+
"eval_loss": 0.03227296099066734,
|
| 20 |
+
"eval_runtime": 6.7134,
|
| 21 |
+
"eval_samples_per_second": 1.192,
|
| 22 |
+
"eval_steps_per_second": 0.149,
|
| 23 |
+
"step": 10
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 2.5,
|
| 27 |
+
"learning_rate": 0.00019,
|
| 28 |
+
"loss": 0.0177,
|
| 29 |
+
"step": 20
|
| 30 |
+
},
|
| 31 |
+
{
|
| 32 |
+
"epoch": 2.5,
|
| 33 |
+
"eval_accuracy": 1.0,
|
| 34 |
+
"eval_loss": 0.011206778697669506,
|
| 35 |
+
"eval_runtime": 4.4578,
|
| 36 |
+
"eval_samples_per_second": 1.795,
|
| 37 |
+
"eval_steps_per_second": 0.224,
|
| 38 |
+
"step": 20
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 3.75,
|
| 42 |
+
"learning_rate": 0.00018500000000000002,
|
| 43 |
+
"loss": 0.0086,
|
| 44 |
+
"step": 30
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 3.75,
|
| 48 |
+
"eval_accuracy": 1.0,
|
| 49 |
+
"eval_loss": 0.00745726190507412,
|
| 50 |
+
"eval_runtime": 6.1312,
|
| 51 |
+
"eval_samples_per_second": 1.305,
|
| 52 |
+
"eval_steps_per_second": 0.163,
|
| 53 |
+
"step": 30
|
| 54 |
+
},
|
| 55 |
+
{
|
| 56 |
+
"epoch": 5.0,
|
| 57 |
+
"learning_rate": 0.00018,
|
| 58 |
+
"loss": 0.0063,
|
| 59 |
+
"step": 40
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 5.0,
|
| 63 |
+
"eval_accuracy": 1.0,
|
| 64 |
+
"eval_loss": 0.0058769905008375645,
|
| 65 |
+
"eval_runtime": 4.3511,
|
| 66 |
+
"eval_samples_per_second": 1.839,
|
| 67 |
+
"eval_steps_per_second": 0.23,
|
| 68 |
+
"step": 40
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 6.25,
|
| 72 |
+
"learning_rate": 0.000175,
|
| 73 |
+
"loss": 0.0051,
|
| 74 |
+
"step": 50
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 6.25,
|
| 78 |
+
"eval_accuracy": 1.0,
|
| 79 |
+
"eval_loss": 0.005005857907235622,
|
| 80 |
+
"eval_runtime": 6.0951,
|
| 81 |
+
"eval_samples_per_second": 1.313,
|
| 82 |
+
"eval_steps_per_second": 0.164,
|
| 83 |
+
"step": 50
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 7.5,
|
| 87 |
+
"learning_rate": 0.00017,
|
| 88 |
+
"loss": 0.0045,
|
| 89 |
+
"step": 60
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 7.5,
|
| 93 |
+
"eval_accuracy": 1.0,
|
| 94 |
+
"eval_loss": 0.004425144288688898,
|
| 95 |
+
"eval_runtime": 4.3832,
|
| 96 |
+
"eval_samples_per_second": 1.825,
|
| 97 |
+
"eval_steps_per_second": 0.228,
|
| 98 |
+
"step": 60
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 8.75,
|
| 102 |
+
"learning_rate": 0.000165,
|
| 103 |
+
"loss": 0.004,
|
| 104 |
+
"step": 70
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 8.75,
|
| 108 |
+
"eval_accuracy": 1.0,
|
| 109 |
+
"eval_loss": 0.0039963009767234325,
|
| 110 |
+
"eval_runtime": 4.3367,
|
| 111 |
+
"eval_samples_per_second": 1.845,
|
| 112 |
+
"eval_steps_per_second": 0.231,
|
| 113 |
+
"step": 70
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 10.0,
|
| 117 |
+
"learning_rate": 0.00016,
|
| 118 |
+
"loss": 0.0036,
|
| 119 |
+
"step": 80
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 10.0,
|
| 123 |
+
"eval_accuracy": 1.0,
|
| 124 |
+
"eval_loss": 0.003639573697000742,
|
| 125 |
+
"eval_runtime": 4.3969,
|
| 126 |
+
"eval_samples_per_second": 1.819,
|
| 127 |
+
"eval_steps_per_second": 0.227,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 11.25,
|
| 132 |
+
"learning_rate": 0.000155,
|
| 133 |
+
"loss": 0.0033,
|
| 134 |
+
"step": 90
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 11.25,
|
| 138 |
+
"eval_accuracy": 1.0,
|
| 139 |
+
"eval_loss": 0.0033505873288959265,
|
| 140 |
+
"eval_runtime": 4.3672,
|
| 141 |
+
"eval_samples_per_second": 1.832,
|
| 142 |
+
"eval_steps_per_second": 0.229,
|
| 143 |
+
"step": 90
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 12.5,
|
| 147 |
+
"learning_rate": 0.00015000000000000001,
|
| 148 |
+
"loss": 0.0031,
|
| 149 |
+
"step": 100
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 12.5,
|
| 153 |
+
"eval_accuracy": 1.0,
|
| 154 |
+
"eval_loss": 0.003097910899668932,
|
| 155 |
+
"eval_runtime": 4.4215,
|
| 156 |
+
"eval_samples_per_second": 1.809,
|
| 157 |
+
"eval_steps_per_second": 0.226,
|
| 158 |
+
"step": 100
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 13.75,
|
| 162 |
+
"learning_rate": 0.000145,
|
| 163 |
+
"loss": 0.0028,
|
| 164 |
+
"step": 110
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 13.75,
|
| 168 |
+
"eval_accuracy": 1.0,
|
| 169 |
+
"eval_loss": 0.0028831621166318655,
|
| 170 |
+
"eval_runtime": 4.4829,
|
| 171 |
+
"eval_samples_per_second": 1.785,
|
| 172 |
+
"eval_steps_per_second": 0.223,
|
| 173 |
+
"step": 110
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 15.0,
|
| 177 |
+
"learning_rate": 0.00014,
|
| 178 |
+
"loss": 0.0026,
|
| 179 |
+
"step": 120
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 15.0,
|
| 183 |
+
"eval_accuracy": 1.0,
|
| 184 |
+
"eval_loss": 0.002691818866878748,
|
| 185 |
+
"eval_runtime": 6.1147,
|
| 186 |
+
"eval_samples_per_second": 1.308,
|
| 187 |
+
"eval_steps_per_second": 0.164,
|
| 188 |
+
"step": 120
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 16.25,
|
| 192 |
+
"learning_rate": 0.00013500000000000003,
|
| 193 |
+
"loss": 0.0025,
|
| 194 |
+
"step": 130
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 16.25,
|
| 198 |
+
"eval_accuracy": 1.0,
|
| 199 |
+
"eval_loss": 0.002529613673686981,
|
| 200 |
+
"eval_runtime": 4.923,
|
| 201 |
+
"eval_samples_per_second": 1.625,
|
| 202 |
+
"eval_steps_per_second": 0.203,
|
| 203 |
+
"step": 130
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 17.5,
|
| 207 |
+
"learning_rate": 0.00013000000000000002,
|
| 208 |
+
"loss": 0.0023,
|
| 209 |
+
"step": 140
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 17.5,
|
| 213 |
+
"eval_accuracy": 1.0,
|
| 214 |
+
"eval_loss": 0.00238333223387599,
|
| 215 |
+
"eval_runtime": 4.4274,
|
| 216 |
+
"eval_samples_per_second": 1.807,
|
| 217 |
+
"eval_steps_per_second": 0.226,
|
| 218 |
+
"step": 140
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 18.75,
|
| 222 |
+
"learning_rate": 0.000125,
|
| 223 |
+
"loss": 0.0022,
|
| 224 |
+
"step": 150
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 18.75,
|
| 228 |
+
"eval_accuracy": 1.0,
|
| 229 |
+
"eval_loss": 0.002253755461424589,
|
| 230 |
+
"eval_runtime": 5.8958,
|
| 231 |
+
"eval_samples_per_second": 1.357,
|
| 232 |
+
"eval_steps_per_second": 0.17,
|
| 233 |
+
"step": 150
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 20.0,
|
| 237 |
+
"learning_rate": 0.00012,
|
| 238 |
+
"loss": 0.0021,
|
| 239 |
+
"step": 160
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 20.0,
|
| 243 |
+
"eval_accuracy": 1.0,
|
| 244 |
+
"eval_loss": 0.002138556446880102,
|
| 245 |
+
"eval_runtime": 4.3757,
|
| 246 |
+
"eval_samples_per_second": 1.828,
|
| 247 |
+
"eval_steps_per_second": 0.229,
|
| 248 |
+
"step": 160
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 21.25,
|
| 252 |
+
"learning_rate": 0.00011499999999999999,
|
| 253 |
+
"loss": 0.002,
|
| 254 |
+
"step": 170
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 21.25,
|
| 258 |
+
"eval_accuracy": 1.0,
|
| 259 |
+
"eval_loss": 0.002034764038398862,
|
| 260 |
+
"eval_runtime": 6.0111,
|
| 261 |
+
"eval_samples_per_second": 1.331,
|
| 262 |
+
"eval_steps_per_second": 0.166,
|
| 263 |
+
"step": 170
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 22.5,
|
| 267 |
+
"learning_rate": 0.00011000000000000002,
|
| 268 |
+
"loss": 0.0019,
|
| 269 |
+
"step": 180
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 22.5,
|
| 273 |
+
"eval_accuracy": 1.0,
|
| 274 |
+
"eval_loss": 0.0019434543792158365,
|
| 275 |
+
"eval_runtime": 8.0684,
|
| 276 |
+
"eval_samples_per_second": 0.992,
|
| 277 |
+
"eval_steps_per_second": 0.124,
|
| 278 |
+
"step": 180
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 23.75,
|
| 282 |
+
"learning_rate": 0.000105,
|
| 283 |
+
"loss": 0.0018,
|
| 284 |
+
"step": 190
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 23.75,
|
| 288 |
+
"eval_accuracy": 1.0,
|
| 289 |
+
"eval_loss": 0.0018578022718429565,
|
| 290 |
+
"eval_runtime": 4.356,
|
| 291 |
+
"eval_samples_per_second": 1.837,
|
| 292 |
+
"eval_steps_per_second": 0.23,
|
| 293 |
+
"step": 190
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 25.0,
|
| 297 |
+
"learning_rate": 0.0001,
|
| 298 |
+
"loss": 0.0017,
|
| 299 |
+
"step": 200
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 25.0,
|
| 303 |
+
"eval_accuracy": 1.0,
|
| 304 |
+
"eval_loss": 0.0017840128857642412,
|
| 305 |
+
"eval_runtime": 4.4024,
|
| 306 |
+
"eval_samples_per_second": 1.817,
|
| 307 |
+
"eval_steps_per_second": 0.227,
|
| 308 |
+
"step": 200
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 26.25,
|
| 312 |
+
"learning_rate": 9.5e-05,
|
| 313 |
+
"loss": 0.0016,
|
| 314 |
+
"step": 210
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 26.25,
|
| 318 |
+
"eval_accuracy": 1.0,
|
| 319 |
+
"eval_loss": 0.0017146510072052479,
|
| 320 |
+
"eval_runtime": 6.2173,
|
| 321 |
+
"eval_samples_per_second": 1.287,
|
| 322 |
+
"eval_steps_per_second": 0.161,
|
| 323 |
+
"step": 210
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 27.5,
|
| 327 |
+
"learning_rate": 9e-05,
|
| 328 |
+
"loss": 0.0016,
|
| 329 |
+
"step": 220
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 27.5,
|
| 333 |
+
"eval_accuracy": 1.0,
|
| 334 |
+
"eval_loss": 0.0016539423959329724,
|
| 335 |
+
"eval_runtime": 4.4503,
|
| 336 |
+
"eval_samples_per_second": 1.798,
|
| 337 |
+
"eval_steps_per_second": 0.225,
|
| 338 |
+
"step": 220
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 28.75,
|
| 342 |
+
"learning_rate": 8.5e-05,
|
| 343 |
+
"loss": 0.0015,
|
| 344 |
+
"step": 230
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 28.75,
|
| 348 |
+
"eval_accuracy": 1.0,
|
| 349 |
+
"eval_loss": 0.0015984964556992054,
|
| 350 |
+
"eval_runtime": 5.2504,
|
| 351 |
+
"eval_samples_per_second": 1.524,
|
| 352 |
+
"eval_steps_per_second": 0.19,
|
| 353 |
+
"step": 230
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 30.0,
|
| 357 |
+
"learning_rate": 8e-05,
|
| 358 |
+
"loss": 0.0015,
|
| 359 |
+
"step": 240
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 30.0,
|
| 363 |
+
"eval_accuracy": 1.0,
|
| 364 |
+
"eval_loss": 0.0015484185423702002,
|
| 365 |
+
"eval_runtime": 4.3678,
|
| 366 |
+
"eval_samples_per_second": 1.832,
|
| 367 |
+
"eval_steps_per_second": 0.229,
|
| 368 |
+
"step": 240
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 31.25,
|
| 372 |
+
"learning_rate": 7.500000000000001e-05,
|
| 373 |
+
"loss": 0.0014,
|
| 374 |
+
"step": 250
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 31.25,
|
| 378 |
+
"eval_accuracy": 1.0,
|
| 379 |
+
"eval_loss": 0.0015031290240585804,
|
| 380 |
+
"eval_runtime": 4.7164,
|
| 381 |
+
"eval_samples_per_second": 1.696,
|
| 382 |
+
"eval_steps_per_second": 0.212,
|
| 383 |
+
"step": 250
|
| 384 |
+
},
|
| 385 |
+
{
|
| 386 |
+
"epoch": 32.5,
|
| 387 |
+
"learning_rate": 7e-05,
|
| 388 |
+
"loss": 0.0014,
|
| 389 |
+
"step": 260
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 32.5,
|
| 393 |
+
"eval_accuracy": 1.0,
|
| 394 |
+
"eval_loss": 0.0014625097392126918,
|
| 395 |
+
"eval_runtime": 5.6004,
|
| 396 |
+
"eval_samples_per_second": 1.428,
|
| 397 |
+
"eval_steps_per_second": 0.179,
|
| 398 |
+
"step": 260
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 33.75,
|
| 402 |
+
"learning_rate": 6.500000000000001e-05,
|
| 403 |
+
"loss": 0.0013,
|
| 404 |
+
"step": 270
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 33.75,
|
| 408 |
+
"eval_accuracy": 1.0,
|
| 409 |
+
"eval_loss": 0.0014250578824430704,
|
| 410 |
+
"eval_runtime": 4.4145,
|
| 411 |
+
"eval_samples_per_second": 1.812,
|
| 412 |
+
"eval_steps_per_second": 0.227,
|
| 413 |
+
"step": 270
|
| 414 |
+
},
|
| 415 |
+
{
|
| 416 |
+
"epoch": 35.0,
|
| 417 |
+
"learning_rate": 6e-05,
|
| 418 |
+
"loss": 0.0013,
|
| 419 |
+
"step": 280
|
| 420 |
+
},
|
| 421 |
+
{
|
| 422 |
+
"epoch": 35.0,
|
| 423 |
+
"eval_accuracy": 1.0,
|
| 424 |
+
"eval_loss": 0.0013918605400249362,
|
| 425 |
+
"eval_runtime": 6.1306,
|
| 426 |
+
"eval_samples_per_second": 1.305,
|
| 427 |
+
"eval_steps_per_second": 0.163,
|
| 428 |
+
"step": 280
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 36.25,
|
| 432 |
+
"learning_rate": 5.500000000000001e-05,
|
| 433 |
+
"loss": 0.0013,
|
| 434 |
+
"step": 290
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 36.25,
|
| 438 |
+
"eval_accuracy": 1.0,
|
| 439 |
+
"eval_loss": 0.0013616828946396708,
|
| 440 |
+
"eval_runtime": 4.4127,
|
| 441 |
+
"eval_samples_per_second": 1.813,
|
| 442 |
+
"eval_steps_per_second": 0.227,
|
| 443 |
+
"step": 290
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 37.5,
|
| 447 |
+
"learning_rate": 5e-05,
|
| 448 |
+
"loss": 0.0013,
|
| 449 |
+
"step": 300
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 37.5,
|
| 453 |
+
"eval_accuracy": 1.0,
|
| 454 |
+
"eval_loss": 0.0013343169121071696,
|
| 455 |
+
"eval_runtime": 4.3912,
|
| 456 |
+
"eval_samples_per_second": 1.822,
|
| 457 |
+
"eval_steps_per_second": 0.228,
|
| 458 |
+
"step": 300
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 38.75,
|
| 462 |
+
"learning_rate": 4.5e-05,
|
| 463 |
+
"loss": 0.0012,
|
| 464 |
+
"step": 310
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 38.75,
|
| 468 |
+
"eval_accuracy": 1.0,
|
| 469 |
+
"eval_loss": 0.0013108787825331092,
|
| 470 |
+
"eval_runtime": 4.8701,
|
| 471 |
+
"eval_samples_per_second": 1.643,
|
| 472 |
+
"eval_steps_per_second": 0.205,
|
| 473 |
+
"step": 310
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 40.0,
|
| 477 |
+
"learning_rate": 4e-05,
|
| 478 |
+
"loss": 0.0012,
|
| 479 |
+
"step": 320
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 40.0,
|
| 483 |
+
"eval_accuracy": 1.0,
|
| 484 |
+
"eval_loss": 0.001289389911107719,
|
| 485 |
+
"eval_runtime": 5.9706,
|
| 486 |
+
"eval_samples_per_second": 1.34,
|
| 487 |
+
"eval_steps_per_second": 0.167,
|
| 488 |
+
"step": 320
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 41.25,
|
| 492 |
+
"learning_rate": 3.5e-05,
|
| 493 |
+
"loss": 0.0012,
|
| 494 |
+
"step": 330
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 41.25,
|
| 498 |
+
"eval_accuracy": 1.0,
|
| 499 |
+
"eval_loss": 0.0012705642729997635,
|
| 500 |
+
"eval_runtime": 4.9479,
|
| 501 |
+
"eval_samples_per_second": 1.617,
|
| 502 |
+
"eval_steps_per_second": 0.202,
|
| 503 |
+
"step": 330
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 42.5,
|
| 507 |
+
"learning_rate": 3e-05,
|
| 508 |
+
"loss": 0.0012,
|
| 509 |
+
"step": 340
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 42.5,
|
| 513 |
+
"eval_accuracy": 1.0,
|
| 514 |
+
"eval_loss": 0.00125490827485919,
|
| 515 |
+
"eval_runtime": 4.9186,
|
| 516 |
+
"eval_samples_per_second": 1.626,
|
| 517 |
+
"eval_steps_per_second": 0.203,
|
| 518 |
+
"step": 340
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 43.75,
|
| 522 |
+
"learning_rate": 2.5e-05,
|
| 523 |
+
"loss": 0.0012,
|
| 524 |
+
"step": 350
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 43.75,
|
| 528 |
+
"eval_accuracy": 1.0,
|
| 529 |
+
"eval_loss": 0.0012415139935910702,
|
| 530 |
+
"eval_runtime": 4.3622,
|
| 531 |
+
"eval_samples_per_second": 1.834,
|
| 532 |
+
"eval_steps_per_second": 0.229,
|
| 533 |
+
"step": 350
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 45.0,
|
| 537 |
+
"learning_rate": 2e-05,
|
| 538 |
+
"loss": 0.0012,
|
| 539 |
+
"step": 360
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 45.0,
|
| 543 |
+
"eval_accuracy": 1.0,
|
| 544 |
+
"eval_loss": 0.0012306944699957967,
|
| 545 |
+
"eval_runtime": 6.2742,
|
| 546 |
+
"eval_samples_per_second": 1.275,
|
| 547 |
+
"eval_steps_per_second": 0.159,
|
| 548 |
+
"step": 360
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 46.25,
|
| 552 |
+
"learning_rate": 1.5e-05,
|
| 553 |
+
"loss": 0.0011,
|
| 554 |
+
"step": 370
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 46.25,
|
| 558 |
+
"eval_accuracy": 1.0,
|
| 559 |
+
"eval_loss": 0.0012223005760461092,
|
| 560 |
+
"eval_runtime": 5.7273,
|
| 561 |
+
"eval_samples_per_second": 1.397,
|
| 562 |
+
"eval_steps_per_second": 0.175,
|
| 563 |
+
"step": 370
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 47.5,
|
| 567 |
+
"learning_rate": 1e-05,
|
| 568 |
+
"loss": 0.0012,
|
| 569 |
+
"step": 380
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 47.5,
|
| 573 |
+
"eval_accuracy": 1.0,
|
| 574 |
+
"eval_loss": 0.0012162283528596163,
|
| 575 |
+
"eval_runtime": 4.7043,
|
| 576 |
+
"eval_samples_per_second": 1.701,
|
| 577 |
+
"eval_steps_per_second": 0.213,
|
| 578 |
+
"step": 380
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 48.75,
|
| 582 |
+
"learning_rate": 5e-06,
|
| 583 |
+
"loss": 0.0011,
|
| 584 |
+
"step": 390
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 48.75,
|
| 588 |
+
"eval_accuracy": 1.0,
|
| 589 |
+
"eval_loss": 0.0012125077191740274,
|
| 590 |
+
"eval_runtime": 4.4338,
|
| 591 |
+
"eval_samples_per_second": 1.804,
|
| 592 |
+
"eval_steps_per_second": 0.226,
|
| 593 |
+
"step": 390
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 50.0,
|
| 597 |
+
"learning_rate": 0.0,
|
| 598 |
+
"loss": 0.0011,
|
| 599 |
+
"step": 400
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 50.0,
|
| 603 |
+
"eval_accuracy": 1.0,
|
| 604 |
+
"eval_loss": 0.0012111830292269588,
|
| 605 |
+
"eval_runtime": 5.2993,
|
| 606 |
+
"eval_samples_per_second": 1.51,
|
| 607 |
+
"eval_steps_per_second": 0.189,
|
| 608 |
+
"step": 400
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 50.0,
|
| 612 |
+
"step": 400,
|
| 613 |
+
"total_flos": 2.789711626125312e+17,
|
| 614 |
+
"train_loss": 0.008226247422862798,
|
| 615 |
+
"train_runtime": 8928.1619,
|
| 616 |
+
"train_samples_per_second": 0.403,
|
| 617 |
+
"train_steps_per_second": 0.045
|
| 618 |
+
}
|
| 619 |
+
],
|
| 620 |
+
"max_steps": 400,
|
| 621 |
+
"num_train_epochs": 50,
|
| 622 |
+
"total_flos": 2.789711626125312e+17,
|
| 623 |
+
"trial_name": null,
|
| 624 |
+
"trial_params": null
|
| 625 |
+
}
|