End of training
Browse files- all_results.json +15 -0
- eval_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +667 -0
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.99,
|
| 3 |
+
"eval_accuracy": 0.611281497151223,
|
| 4 |
+
"eval_loss": 2.2855756282806396,
|
| 5 |
+
"eval_runtime": 56.4639,
|
| 6 |
+
"eval_samples": 1091,
|
| 7 |
+
"eval_samples_per_second": 19.322,
|
| 8 |
+
"eval_steps_per_second": 19.322,
|
| 9 |
+
"perplexity": 9.831343795057329,
|
| 10 |
+
"train_loss": 1.8919498779271777,
|
| 11 |
+
"train_runtime": 5682.4938,
|
| 12 |
+
"train_samples": 9791,
|
| 13 |
+
"train_samples_per_second": 6.892,
|
| 14 |
+
"train_steps_per_second": 0.053
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.99,
|
| 3 |
+
"eval_accuracy": 0.611281497151223,
|
| 4 |
+
"eval_loss": 2.2855756282806396,
|
| 5 |
+
"eval_runtime": 56.4639,
|
| 6 |
+
"eval_samples": 1091,
|
| 7 |
+
"eval_samples_per_second": 19.322,
|
| 8 |
+
"eval_steps_per_second": 19.322,
|
| 9 |
+
"perplexity": 9.831343795057329
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.99,
|
| 3 |
+
"train_loss": 1.8919498779271777,
|
| 4 |
+
"train_runtime": 5682.4938,
|
| 5 |
+
"train_samples": 9791,
|
| 6 |
+
"train_samples_per_second": 6.892,
|
| 7 |
+
"train_steps_per_second": 0.053
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,667 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.9934640522875817,
|
| 5 |
+
"global_step": 304,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.04,
|
| 12 |
+
"learning_rate": 0.0001125,
|
| 13 |
+
"loss": 3.8539,
|
| 14 |
+
"step": 3
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.08,
|
| 18 |
+
"learning_rate": 0.000225,
|
| 19 |
+
"loss": 3.5808,
|
| 20 |
+
"step": 6
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.12,
|
| 24 |
+
"learning_rate": 0.00033749999999999996,
|
| 25 |
+
"loss": 3.335,
|
| 26 |
+
"step": 9
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.16,
|
| 30 |
+
"learning_rate": 0.00045,
|
| 31 |
+
"loss": 3.2031,
|
| 32 |
+
"step": 12
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.2,
|
| 36 |
+
"learning_rate": 0.0005625,
|
| 37 |
+
"loss": 3.0706,
|
| 38 |
+
"step": 15
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.24,
|
| 42 |
+
"learning_rate": 0.0005999286081239726,
|
| 43 |
+
"loss": 3.084,
|
| 44 |
+
"step": 18
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.27,
|
| 48 |
+
"learning_rate": 0.0005995538936819289,
|
| 49 |
+
"loss": 2.9744,
|
| 50 |
+
"step": 21
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.31,
|
| 54 |
+
"learning_rate": 0.0005988584094275236,
|
| 55 |
+
"loss": 2.8987,
|
| 56 |
+
"step": 24
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.35,
|
| 60 |
+
"learning_rate": 0.0005978429001027164,
|
| 61 |
+
"loss": 2.8293,
|
| 62 |
+
"step": 27
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.39,
|
| 66 |
+
"learning_rate": 0.0005965084531403281,
|
| 67 |
+
"loss": 2.7609,
|
| 68 |
+
"step": 30
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.43,
|
| 72 |
+
"learning_rate": 0.0005948564974995903,
|
| 73 |
+
"loss": 2.7155,
|
| 74 |
+
"step": 33
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.47,
|
| 78 |
+
"learning_rate": 0.00059288880213598,
|
| 79 |
+
"loss": 2.7184,
|
| 80 |
+
"step": 36
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.51,
|
| 84 |
+
"learning_rate": 0.0005906074741069779,
|
| 85 |
+
"loss": 2.6533,
|
| 86 |
+
"step": 39
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.55,
|
| 90 |
+
"learning_rate": 0.0005880149563157786,
|
| 91 |
+
"loss": 2.6509,
|
| 92 |
+
"step": 42
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.59,
|
| 96 |
+
"learning_rate": 0.0005851140248953683,
|
| 97 |
+
"loss": 2.5859,
|
| 98 |
+
"step": 45
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.63,
|
| 102 |
+
"learning_rate": 0.0005819077862357724,
|
| 103 |
+
"loss": 2.5516,
|
| 104 |
+
"step": 48
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.67,
|
| 108 |
+
"learning_rate": 0.0005783996736576553,
|
| 109 |
+
"loss": 2.5855,
|
| 110 |
+
"step": 51
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.71,
|
| 114 |
+
"learning_rate": 0.0005745934437358341,
|
| 115 |
+
"loss": 2.5306,
|
| 116 |
+
"step": 54
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.75,
|
| 120 |
+
"learning_rate": 0.0005704931722766448,
|
| 121 |
+
"loss": 2.5159,
|
| 122 |
+
"step": 57
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.78,
|
| 126 |
+
"learning_rate": 0.0005661032499534664,
|
| 127 |
+
"loss": 2.5304,
|
| 128 |
+
"step": 60
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.82,
|
| 132 |
+
"learning_rate": 0.0005614283776050784,
|
| 133 |
+
"loss": 2.5005,
|
| 134 |
+
"step": 63
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.86,
|
| 138 |
+
"learning_rate": 0.0005564735612018839,
|
| 139 |
+
"loss": 2.4884,
|
| 140 |
+
"step": 66
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.9,
|
| 144 |
+
"learning_rate": 0.0005512441064853923,
|
| 145 |
+
"loss": 2.4729,
|
| 146 |
+
"step": 69
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.94,
|
| 150 |
+
"learning_rate": 0.0005457456132866975,
|
| 151 |
+
"loss": 2.4226,
|
| 152 |
+
"step": 72
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.98,
|
| 156 |
+
"learning_rate": 0.0005399839695300389,
|
| 157 |
+
"loss": 2.412,
|
| 158 |
+
"step": 75
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.99,
|
| 162 |
+
"eval_accuracy": 0.5457690353760842,
|
| 163 |
+
"eval_loss": 2.502744436264038,
|
| 164 |
+
"eval_runtime": 56.9596,
|
| 165 |
+
"eval_samples_per_second": 19.154,
|
| 166 |
+
"eval_steps_per_second": 19.154,
|
| 167 |
+
"step": 76
|
| 168 |
+
},
|
| 169 |
+
{
|
| 170 |
+
"epoch": 1.03,
|
| 171 |
+
"learning_rate": 0.0005339653449278644,
|
| 172 |
+
"loss": 2.6265,
|
| 173 |
+
"step": 78
|
| 174 |
+
},
|
| 175 |
+
{
|
| 176 |
+
"epoch": 1.07,
|
| 177 |
+
"learning_rate": 0.0005276961843741485,
|
| 178 |
+
"loss": 2.148,
|
| 179 |
+
"step": 81
|
| 180 |
+
},
|
| 181 |
+
{
|
| 182 |
+
"epoch": 1.1,
|
| 183 |
+
"learning_rate": 0.0005211832010430372,
|
| 184 |
+
"loss": 2.1056,
|
| 185 |
+
"step": 84
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 1.14,
|
| 189 |
+
"learning_rate": 0.0005144333692002139,
|
| 190 |
+
"loss": 2.1074,
|
| 191 |
+
"step": 87
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 1.18,
|
| 195 |
+
"learning_rate": 0.0005074539167346808,
|
| 196 |
+
"loss": 2.118,
|
| 197 |
+
"step": 90
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 1.22,
|
| 201 |
+
"learning_rate": 0.0005002523174189542,
|
| 202 |
+
"loss": 2.1045,
|
| 203 |
+
"step": 93
|
| 204 |
+
},
|
| 205 |
+
{
|
| 206 |
+
"epoch": 1.26,
|
| 207 |
+
"learning_rate": 0.0004928362829059618,
|
| 208 |
+
"loss": 2.061,
|
| 209 |
+
"step": 96
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 1.3,
|
| 213 |
+
"learning_rate": 0.0004852137544712115,
|
| 214 |
+
"loss": 2.0959,
|
| 215 |
+
"step": 99
|
| 216 |
+
},
|
| 217 |
+
{
|
| 218 |
+
"epoch": 1.34,
|
| 219 |
+
"learning_rate": 0.0004773928945090747,
|
| 220 |
+
"loss": 2.0637,
|
| 221 |
+
"step": 102
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 1.38,
|
| 225 |
+
"learning_rate": 0.0004693820777922901,
|
| 226 |
+
"loss": 2.004,
|
| 227 |
+
"step": 105
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 1.42,
|
| 231 |
+
"learning_rate": 0.00046118988250404714,
|
| 232 |
+
"loss": 2.0516,
|
| 233 |
+
"step": 108
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 1.46,
|
| 237 |
+
"learning_rate": 0.00045282508105225254,
|
| 238 |
+
"loss": 2.1182,
|
| 239 |
+
"step": 111
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 1.5,
|
| 243 |
+
"learning_rate": 0.00044429663067581626,
|
| 244 |
+
"loss": 2.0648,
|
| 245 |
+
"step": 114
|
| 246 |
+
},
|
| 247 |
+
{
|
| 248 |
+
"epoch": 1.54,
|
| 249 |
+
"learning_rate": 0.0004356136638530159,
|
| 250 |
+
"loss": 2.058,
|
| 251 |
+
"step": 117
|
| 252 |
+
},
|
| 253 |
+
{
|
| 254 |
+
"epoch": 1.58,
|
| 255 |
+
"learning_rate": 0.0004267854785222098,
|
| 256 |
+
"loss": 2.012,
|
| 257 |
+
"step": 120
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 1.61,
|
| 261 |
+
"learning_rate": 0.00041782152812537223,
|
| 262 |
+
"loss": 2.0105,
|
| 263 |
+
"step": 123
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 1.65,
|
| 267 |
+
"learning_rate": 0.00040873141148511043,
|
| 268 |
+
"loss": 1.9976,
|
| 269 |
+
"step": 126
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 1.69,
|
| 273 |
+
"learning_rate": 0.00039952486252600565,
|
| 274 |
+
"loss": 2.0034,
|
| 275 |
+
"step": 129
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 1.73,
|
| 279 |
+
"learning_rate": 0.00039021173985128186,
|
| 280 |
+
"loss": 2.0171,
|
| 281 |
+
"step": 132
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 1.77,
|
| 285 |
+
"learning_rate": 0.00038080201618596784,
|
| 286 |
+
"loss": 2.0163,
|
| 287 |
+
"step": 135
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 1.81,
|
| 291 |
+
"learning_rate": 0.0003713057676978519,
|
| 292 |
+
"loss": 1.9683,
|
| 293 |
+
"step": 138
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 1.85,
|
| 297 |
+
"learning_rate": 0.00036173316320767046,
|
| 298 |
+
"loss": 2.0209,
|
| 299 |
+
"step": 141
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 1.89,
|
| 303 |
+
"learning_rate": 0.0003520944533000791,
|
| 304 |
+
"loss": 1.9655,
|
| 305 |
+
"step": 144
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 1.93,
|
| 309 |
+
"learning_rate": 0.0003423999593470703,
|
| 310 |
+
"loss": 2.0127,
|
| 311 |
+
"step": 147
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 1.97,
|
| 315 |
+
"learning_rate": 0.00033266006245558934,
|
| 316 |
+
"loss": 1.9702,
|
| 317 |
+
"step": 150
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 1.99,
|
| 321 |
+
"eval_accuracy": 0.5849799255079998,
|
| 322 |
+
"eval_loss": 2.2756919860839844,
|
| 323 |
+
"eval_runtime": 56.7417,
|
| 324 |
+
"eval_samples_per_second": 19.227,
|
| 325 |
+
"eval_steps_per_second": 19.227,
|
| 326 |
+
"step": 152
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 2.01,
|
| 330 |
+
"learning_rate": 0.00032288519235118573,
|
| 331 |
+
"loss": 2.1744,
|
| 332 |
+
"step": 153
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 2.05,
|
| 336 |
+
"learning_rate": 0.00031308581620960083,
|
| 337 |
+
"loss": 1.6022,
|
| 338 |
+
"step": 156
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 2.09,
|
| 342 |
+
"learning_rate": 0.0003032724274482547,
|
| 343 |
+
"loss": 1.6168,
|
| 344 |
+
"step": 159
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 2.13,
|
| 348 |
+
"learning_rate": 0.0002934555344896317,
|
| 349 |
+
"loss": 1.5467,
|
| 350 |
+
"step": 162
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 2.17,
|
| 354 |
+
"learning_rate": 0.00028364564950859807,
|
| 355 |
+
"loss": 1.6023,
|
| 356 |
+
"step": 165
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 2.21,
|
| 360 |
+
"learning_rate": 0.0002738532771757025,
|
| 361 |
+
"loss": 1.4977,
|
| 362 |
+
"step": 168
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 2.25,
|
| 366 |
+
"learning_rate": 0.0002640889034085113,
|
| 367 |
+
"loss": 1.5448,
|
| 368 |
+
"step": 171
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 2.29,
|
| 372 |
+
"learning_rate": 0.00025436298414302494,
|
| 373 |
+
"loss": 1.5738,
|
| 374 |
+
"step": 174
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 2.33,
|
| 378 |
+
"learning_rate": 0.000244685934137201,
|
| 379 |
+
"loss": 1.559,
|
| 380 |
+
"step": 177
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 2.37,
|
| 384 |
+
"learning_rate": 0.00023506811581856912,
|
| 385 |
+
"loss": 1.5734,
|
| 386 |
+
"step": 180
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 2.41,
|
| 390 |
+
"learning_rate": 0.00022551982818788506,
|
| 391 |
+
"loss": 1.4986,
|
| 392 |
+
"step": 183
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 2.44,
|
| 396 |
+
"learning_rate": 0.00021605129579070238,
|
| 397 |
+
"loss": 1.545,
|
| 398 |
+
"step": 186
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 2.48,
|
| 402 |
+
"learning_rate": 0.00020667265776867276,
|
| 403 |
+
"loss": 1.5496,
|
| 404 |
+
"step": 189
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 2.52,
|
| 408 |
+
"learning_rate": 0.00019739395700229937,
|
| 409 |
+
"loss": 1.5426,
|
| 410 |
+
"step": 192
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.56,
|
| 414 |
+
"learning_rate": 0.0001882251293567691,
|
| 415 |
+
"loss": 1.4687,
|
| 416 |
+
"step": 195
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 2.6,
|
| 420 |
+
"learning_rate": 0.00017917599304237886,
|
| 421 |
+
"loss": 1.5415,
|
| 422 |
+
"step": 198
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.64,
|
| 426 |
+
"learning_rate": 0.0001702562381009501,
|
| 427 |
+
"loss": 1.5289,
|
| 428 |
+
"step": 201
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 2.68,
|
| 432 |
+
"learning_rate": 0.0001614754160294899,
|
| 433 |
+
"loss": 1.5449,
|
| 434 |
+
"step": 204
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 2.72,
|
| 438 |
+
"learning_rate": 0.0001528429295522076,
|
| 439 |
+
"loss": 1.5273,
|
| 440 |
+
"step": 207
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.76,
|
| 444 |
+
"learning_rate": 0.0001443680225518435,
|
| 445 |
+
"loss": 1.5146,
|
| 446 |
+
"step": 210
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 2.8,
|
| 450 |
+
"learning_rate": 0.000136059770171087,
|
| 451 |
+
"loss": 1.5164,
|
| 452 |
+
"step": 213
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.84,
|
| 456 |
+
"learning_rate": 0.00012792706909468623,
|
| 457 |
+
"loss": 1.5239,
|
| 458 |
+
"step": 216
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 2.88,
|
| 462 |
+
"learning_rate": 0.00011997862802265573,
|
| 463 |
+
"loss": 1.4905,
|
| 464 |
+
"step": 219
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.92,
|
| 468 |
+
"learning_rate": 0.00011222295834478227,
|
| 469 |
+
"loss": 1.4968,
|
| 470 |
+
"step": 222
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 2.95,
|
| 474 |
+
"learning_rate": 0.0001046683650264153,
|
| 475 |
+
"loss": 1.4934,
|
| 476 |
+
"step": 225
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 2.99,
|
| 480 |
+
"learning_rate": 9.732293771530192e-05,
|
| 481 |
+
"loss": 1.4628,
|
| 482 |
+
"step": 228
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.99,
|
| 486 |
+
"eval_accuracy": 0.6081706452777681,
|
| 487 |
+
"eval_loss": 2.2162108421325684,
|
| 488 |
+
"eval_runtime": 57.053,
|
| 489 |
+
"eval_samples_per_second": 19.123,
|
| 490 |
+
"eval_steps_per_second": 19.123,
|
| 491 |
+
"step": 228
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 3.04,
|
| 495 |
+
"learning_rate": 9.019454207898983e-05,
|
| 496 |
+
"loss": 1.4385,
|
| 497 |
+
"step": 231
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 3.08,
|
| 501 |
+
"learning_rate": 8.329081138207334e-05,
|
| 502 |
+
"loss": 1.2032,
|
| 503 |
+
"step": 234
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 3.12,
|
| 507 |
+
"learning_rate": 7.661913831230212e-05,
|
| 508 |
+
"loss": 1.1659,
|
| 509 |
+
"step": 237
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 3.16,
|
| 513 |
+
"learning_rate": 7.018666706430662e-05,
|
| 514 |
+
"loss": 1.1521,
|
| 515 |
+
"step": 240
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 3.2,
|
| 519 |
+
"learning_rate": 6.40002856894149e-05,
|
| 520 |
+
"loss": 1.1916,
|
| 521 |
+
"step": 243
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 3.24,
|
| 525 |
+
"learning_rate": 5.8066618719755195e-05,
|
| 526 |
+
"loss": 1.163,
|
| 527 |
+
"step": 246
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 3.27,
|
| 531 |
+
"learning_rate": 5.239202007454086e-05,
|
| 532 |
+
"loss": 1.1624,
|
| 533 |
+
"step": 249
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 3.31,
|
| 537 |
+
"learning_rate": 4.698256625613435e-05,
|
| 538 |
+
"loss": 1.1203,
|
| 539 |
+
"step": 252
|
| 540 |
+
},
|
| 541 |
+
{
|
| 542 |
+
"epoch": 3.35,
|
| 543 |
+
"learning_rate": 4.1844049843176334e-05,
|
| 544 |
+
"loss": 1.1709,
|
| 545 |
+
"step": 255
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 3.39,
|
| 549 |
+
"learning_rate": 3.698197328774769e-05,
|
| 550 |
+
"loss": 1.1263,
|
| 551 |
+
"step": 258
|
| 552 |
+
},
|
| 553 |
+
{
|
| 554 |
+
"epoch": 3.43,
|
| 555 |
+
"learning_rate": 3.2401543023205764e-05,
|
| 556 |
+
"loss": 1.1277,
|
| 557 |
+
"step": 261
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 3.47,
|
| 561 |
+
"learning_rate": 2.8107663889005016e-05,
|
| 562 |
+
"loss": 1.1759,
|
| 563 |
+
"step": 264
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 3.51,
|
| 567 |
+
"learning_rate": 2.410493387847232e-05,
|
| 568 |
+
"loss": 1.157,
|
| 569 |
+
"step": 267
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 3.55,
|
| 573 |
+
"learning_rate": 2.0397639215160466e-05,
|
| 574 |
+
"loss": 1.1721,
|
| 575 |
+
"step": 270
|
| 576 |
+
},
|
| 577 |
+
{
|
| 578 |
+
"epoch": 3.59,
|
| 579 |
+
"learning_rate": 1.698974976305243e-05,
|
| 580 |
+
"loss": 1.1577,
|
| 581 |
+
"step": 273
|
| 582 |
+
},
|
| 583 |
+
{
|
| 584 |
+
"epoch": 3.63,
|
| 585 |
+
"learning_rate": 1.3884914775531952e-05,
|
| 586 |
+
"loss": 1.1546,
|
| 587 |
+
"step": 276
|
| 588 |
+
},
|
| 589 |
+
{
|
| 590 |
+
"epoch": 3.67,
|
| 591 |
+
"learning_rate": 1.1086458987671187e-05,
|
| 592 |
+
"loss": 1.1564,
|
| 593 |
+
"step": 279
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 3.71,
|
| 597 |
+
"learning_rate": 8.59737905602157e-06,
|
| 598 |
+
"loss": 1.1241,
|
| 599 |
+
"step": 282
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 3.75,
|
| 603 |
+
"learning_rate": 6.4203403497185e-06,
|
| 604 |
+
"loss": 1.1605,
|
| 605 |
+
"step": 285
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 3.78,
|
| 609 |
+
"learning_rate": 4.557674096337593e-06,
|
| 610 |
+
"loss": 1.1114,
|
| 611 |
+
"step": 288
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 3.82,
|
| 615 |
+
"learning_rate": 3.011374885557638e-06,
|
| 616 |
+
"loss": 1.1377,
|
| 617 |
+
"step": 291
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 3.86,
|
| 621 |
+
"learning_rate": 1.783098533304106e-06,
|
| 622 |
+
"loss": 1.1541,
|
| 623 |
+
"step": 294
|
| 624 |
+
},
|
| 625 |
+
{
|
| 626 |
+
"epoch": 3.9,
|
| 627 |
+
"learning_rate": 8.741603086600102e-07,
|
| 628 |
+
"loss": 1.174,
|
| 629 |
+
"step": 297
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 3.94,
|
| 633 |
+
"learning_rate": 2.855335254426605e-07,
|
| 634 |
+
"loss": 1.1415,
|
| 635 |
+
"step": 300
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 3.98,
|
| 639 |
+
"learning_rate": 1.7848499955075423e-08,
|
| 640 |
+
"loss": 1.1662,
|
| 641 |
+
"step": 303
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 3.99,
|
| 645 |
+
"eval_accuracy": 0.611281497151223,
|
| 646 |
+
"eval_loss": 2.2855756282806396,
|
| 647 |
+
"eval_runtime": 56.9619,
|
| 648 |
+
"eval_samples_per_second": 19.153,
|
| 649 |
+
"eval_steps_per_second": 19.153,
|
| 650 |
+
"step": 304
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 3.99,
|
| 654 |
+
"step": 304,
|
| 655 |
+
"total_flos": 2.9714731304484864e+16,
|
| 656 |
+
"train_loss": 1.8919498779271777,
|
| 657 |
+
"train_runtime": 5682.4938,
|
| 658 |
+
"train_samples_per_second": 6.892,
|
| 659 |
+
"train_steps_per_second": 0.053
|
| 660 |
+
}
|
| 661 |
+
],
|
| 662 |
+
"max_steps": 304,
|
| 663 |
+
"num_train_epochs": 4,
|
| 664 |
+
"total_flos": 2.9714731304484864e+16,
|
| 665 |
+
"trial_name": null,
|
| 666 |
+
"trial_params": null
|
| 667 |
+
}
|