Upload trainer_state.json with huggingface_hub
Browse files- trainer_state.json +757 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,757 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 3654,
|
| 3 |
+
"best_metric": 0.25175856147050574,
|
| 4 |
+
"best_model_checkpoint": "outputs/textcnn/checkpoint-3654",
|
| 5 |
+
"epoch": 42.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 3654,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 1.0,
|
| 14 |
+
"grad_norm": 5.989197731018066,
|
| 15 |
+
"learning_rate": 3.44e-06,
|
| 16 |
+
"loss": 2.0625,
|
| 17 |
+
"step": 87
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 1.0,
|
| 21 |
+
"eval_accuracy": 0.31486880466472306,
|
| 22 |
+
"eval_loss": 1.9201620817184448,
|
| 23 |
+
"eval_macro_f1": 0.08069587306875443,
|
| 24 |
+
"eval_runtime": 0.0822,
|
| 25 |
+
"eval_samples_per_second": 8343.34,
|
| 26 |
+
"eval_steps_per_second": 133.785,
|
| 27 |
+
"step": 87
|
| 28 |
+
},
|
| 29 |
+
{
|
| 30 |
+
"epoch": 2.0,
|
| 31 |
+
"grad_norm": 4.812881946563721,
|
| 32 |
+
"learning_rate": 6.92e-06,
|
| 33 |
+
"loss": 2.0517,
|
| 34 |
+
"step": 174
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 2.0,
|
| 38 |
+
"eval_accuracy": 0.3163265306122449,
|
| 39 |
+
"eval_loss": 1.867976188659668,
|
| 40 |
+
"eval_macro_f1": 0.08271289631865439,
|
| 41 |
+
"eval_runtime": 0.0798,
|
| 42 |
+
"eval_samples_per_second": 8598.327,
|
| 43 |
+
"eval_steps_per_second": 137.874,
|
| 44 |
+
"step": 174
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 3.0,
|
| 48 |
+
"grad_norm": 4.3668341636657715,
|
| 49 |
+
"learning_rate": 1.04e-05,
|
| 50 |
+
"loss": 1.9731,
|
| 51 |
+
"step": 261
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 3.0,
|
| 55 |
+
"eval_accuracy": 0.31924198250728864,
|
| 56 |
+
"eval_loss": 1.8050793409347534,
|
| 57 |
+
"eval_macro_f1": 0.09255094257382587,
|
| 58 |
+
"eval_runtime": 0.0765,
|
| 59 |
+
"eval_samples_per_second": 8966.769,
|
| 60 |
+
"eval_steps_per_second": 143.782,
|
| 61 |
+
"step": 261
|
| 62 |
+
},
|
| 63 |
+
{
|
| 64 |
+
"epoch": 4.0,
|
| 65 |
+
"grad_norm": 4.521817207336426,
|
| 66 |
+
"learning_rate": 1.3880000000000001e-05,
|
| 67 |
+
"loss": 1.9355,
|
| 68 |
+
"step": 348
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 4.0,
|
| 72 |
+
"eval_accuracy": 0.3206997084548105,
|
| 73 |
+
"eval_loss": 1.760697841644287,
|
| 74 |
+
"eval_macro_f1": 0.09828636171767466,
|
| 75 |
+
"eval_runtime": 0.0776,
|
| 76 |
+
"eval_samples_per_second": 8839.5,
|
| 77 |
+
"eval_steps_per_second": 141.741,
|
| 78 |
+
"step": 348
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"epoch": 5.0,
|
| 82 |
+
"grad_norm": 4.453028678894043,
|
| 83 |
+
"learning_rate": 1.736e-05,
|
| 84 |
+
"loss": 1.9005,
|
| 85 |
+
"step": 435
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 5.0,
|
| 89 |
+
"eval_accuracy": 0.32653061224489793,
|
| 90 |
+
"eval_loss": 1.7365907430648804,
|
| 91 |
+
"eval_macro_f1": 0.11368274326806725,
|
| 92 |
+
"eval_runtime": 0.0784,
|
| 93 |
+
"eval_samples_per_second": 8750.81,
|
| 94 |
+
"eval_steps_per_second": 140.319,
|
| 95 |
+
"step": 435
|
| 96 |
+
},
|
| 97 |
+
{
|
| 98 |
+
"epoch": 6.0,
|
| 99 |
+
"grad_norm": 4.236560344696045,
|
| 100 |
+
"learning_rate": 1.999967634800249e-05,
|
| 101 |
+
"loss": 1.8734,
|
| 102 |
+
"step": 522
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 6.0,
|
| 106 |
+
"eval_accuracy": 0.33527696793002915,
|
| 107 |
+
"eval_loss": 1.7214981317520142,
|
| 108 |
+
"eval_macro_f1": 0.12326578903926995,
|
| 109 |
+
"eval_runtime": 0.0768,
|
| 110 |
+
"eval_samples_per_second": 8932.057,
|
| 111 |
+
"eval_steps_per_second": 143.225,
|
| 112 |
+
"step": 522
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 7.0,
|
| 116 |
+
"grad_norm": 4.54931640625,
|
| 117 |
+
"learning_rate": 1.999144090999249e-05,
|
| 118 |
+
"loss": 1.8459,
|
| 119 |
+
"step": 609
|
| 120 |
+
},
|
| 121 |
+
{
|
| 122 |
+
"epoch": 7.0,
|
| 123 |
+
"eval_accuracy": 0.33819241982507287,
|
| 124 |
+
"eval_loss": 1.7085658311843872,
|
| 125 |
+
"eval_macro_f1": 0.12473908000560015,
|
| 126 |
+
"eval_runtime": 0.0772,
|
| 127 |
+
"eval_samples_per_second": 8891.235,
|
| 128 |
+
"eval_steps_per_second": 142.571,
|
| 129 |
+
"step": 609
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 8.0,
|
| 133 |
+
"grad_norm": 4.285991191864014,
|
| 134 |
+
"learning_rate": 1.9972106098590665e-05,
|
| 135 |
+
"loss": 1.8271,
|
| 136 |
+
"step": 696
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 8.0,
|
| 140 |
+
"eval_accuracy": 0.3469387755102041,
|
| 141 |
+
"eval_loss": 1.6989842653274536,
|
| 142 |
+
"eval_macro_f1": 0.1374280909101405,
|
| 143 |
+
"eval_runtime": 0.0776,
|
| 144 |
+
"eval_samples_per_second": 8844.853,
|
| 145 |
+
"eval_steps_per_second": 141.827,
|
| 146 |
+
"step": 696
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 9.0,
|
| 150 |
+
"grad_norm": 5.04291296005249,
|
| 151 |
+
"learning_rate": 1.994169339261005e-05,
|
| 152 |
+
"loss": 1.8219,
|
| 153 |
+
"step": 783
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 9.0,
|
| 157 |
+
"eval_accuracy": 0.3498542274052478,
|
| 158 |
+
"eval_loss": 1.6909065246582031,
|
| 159 |
+
"eval_macro_f1": 0.14476905523124012,
|
| 160 |
+
"eval_runtime": 0.0766,
|
| 161 |
+
"eval_samples_per_second": 8950.395,
|
| 162 |
+
"eval_steps_per_second": 143.519,
|
| 163 |
+
"step": 783
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 10.0,
|
| 167 |
+
"grad_norm": 3.8076608180999756,
|
| 168 |
+
"learning_rate": 1.990023657716558e-05,
|
| 169 |
+
"loss": 1.8039,
|
| 170 |
+
"step": 870
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 10.0,
|
| 174 |
+
"eval_accuracy": 0.3469387755102041,
|
| 175 |
+
"eval_loss": 1.6828982830047607,
|
| 176 |
+
"eval_macro_f1": 0.14793175460560187,
|
| 177 |
+
"eval_runtime": 0.0764,
|
| 178 |
+
"eval_samples_per_second": 8983.678,
|
| 179 |
+
"eval_steps_per_second": 144.053,
|
| 180 |
+
"step": 870
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 11.0,
|
| 184 |
+
"grad_norm": 3.9515891075134277,
|
| 185 |
+
"learning_rate": 1.9847781706142608e-05,
|
| 186 |
+
"loss": 1.7898,
|
| 187 |
+
"step": 957
|
| 188 |
+
},
|
| 189 |
+
{
|
| 190 |
+
"epoch": 11.0,
|
| 191 |
+
"eval_accuracy": 0.35131195335276966,
|
| 192 |
+
"eval_loss": 1.675271987915039,
|
| 193 |
+
"eval_macro_f1": 0.15154267292502702,
|
| 194 |
+
"eval_runtime": 0.0776,
|
| 195 |
+
"eval_samples_per_second": 8837.816,
|
| 196 |
+
"eval_steps_per_second": 141.714,
|
| 197 |
+
"step": 957
|
| 198 |
+
},
|
| 199 |
+
{
|
| 200 |
+
"epoch": 12.0,
|
| 201 |
+
"grad_norm": 3.946139097213745,
|
| 202 |
+
"learning_rate": 1.978438705103621e-05,
|
| 203 |
+
"loss": 1.7634,
|
| 204 |
+
"step": 1044
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 12.0,
|
| 208 |
+
"eval_accuracy": 0.36151603498542273,
|
| 209 |
+
"eval_loss": 1.6689125299453735,
|
| 210 |
+
"eval_macro_f1": 0.16042602782078802,
|
| 211 |
+
"eval_runtime": 0.0767,
|
| 212 |
+
"eval_samples_per_second": 8948.781,
|
| 213 |
+
"eval_steps_per_second": 143.494,
|
| 214 |
+
"step": 1044
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"epoch": 13.0,
|
| 218 |
+
"grad_norm": 3.8360438346862793,
|
| 219 |
+
"learning_rate": 1.9710123036218044e-05,
|
| 220 |
+
"loss": 1.7572,
|
| 221 |
+
"step": 1131
|
| 222 |
+
},
|
| 223 |
+
{
|
| 224 |
+
"epoch": 13.0,
|
| 225 |
+
"eval_accuracy": 0.37026239067055394,
|
| 226 |
+
"eval_loss": 1.6614633798599243,
|
| 227 |
+
"eval_macro_f1": 0.17028534014340227,
|
| 228 |
+
"eval_runtime": 0.0767,
|
| 229 |
+
"eval_samples_per_second": 8943.635,
|
| 230 |
+
"eval_steps_per_second": 143.411,
|
| 231 |
+
"step": 1131
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 14.0,
|
| 235 |
+
"grad_norm": 3.794384479522705,
|
| 236 |
+
"learning_rate": 1.962507216070276e-05,
|
| 237 |
+
"loss": 1.7411,
|
| 238 |
+
"step": 1218
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 14.0,
|
| 242 |
+
"eval_accuracy": 0.36151603498542273,
|
| 243 |
+
"eval_loss": 1.6555291414260864,
|
| 244 |
+
"eval_macro_f1": 0.17229172694357175,
|
| 245 |
+
"eval_runtime": 0.0764,
|
| 246 |
+
"eval_samples_per_second": 8976.952,
|
| 247 |
+
"eval_steps_per_second": 143.945,
|
| 248 |
+
"step": 1218
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 15.0,
|
| 252 |
+
"grad_norm": 3.697802782058716,
|
| 253 |
+
"learning_rate": 1.9529328906500833e-05,
|
| 254 |
+
"loss": 1.7355,
|
| 255 |
+
"step": 1305
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 15.0,
|
| 259 |
+
"eval_accuracy": 0.36443148688046645,
|
| 260 |
+
"eval_loss": 1.6497727632522583,
|
| 261 |
+
"eval_macro_f1": 0.17077990977186067,
|
| 262 |
+
"eval_runtime": 0.0762,
|
| 263 |
+
"eval_samples_per_second": 8997.528,
|
| 264 |
+
"eval_steps_per_second": 144.275,
|
| 265 |
+
"step": 1305
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 16.0,
|
| 269 |
+
"grad_norm": 4.296336650848389,
|
| 270 |
+
"learning_rate": 1.9422999633659592e-05,
|
| 271 |
+
"loss": 1.7163,
|
| 272 |
+
"step": 1392
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 16.0,
|
| 276 |
+
"eval_accuracy": 0.3717201166180758,
|
| 277 |
+
"eval_loss": 1.6435818672180176,
|
| 278 |
+
"eval_macro_f1": 0.1808240545174343,
|
| 279 |
+
"eval_runtime": 0.0787,
|
| 280 |
+
"eval_samples_per_second": 8721.685,
|
| 281 |
+
"eval_steps_per_second": 139.852,
|
| 282 |
+
"step": 1392
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 17.0,
|
| 286 |
+
"grad_norm": 4.240530490875244,
|
| 287 |
+
"learning_rate": 1.9306202462109128e-05,
|
| 288 |
+
"loss": 1.6979,
|
| 289 |
+
"step": 1479
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 17.0,
|
| 293 |
+
"eval_accuracy": 0.3760932944606414,
|
| 294 |
+
"eval_loss": 1.6384371519088745,
|
| 295 |
+
"eval_macro_f1": 0.18768397854098065,
|
| 296 |
+
"eval_runtime": 0.0781,
|
| 297 |
+
"eval_samples_per_second": 8785.572,
|
| 298 |
+
"eval_steps_per_second": 140.877,
|
| 299 |
+
"step": 1479
|
| 300 |
+
},
|
| 301 |
+
{
|
| 302 |
+
"epoch": 18.0,
|
| 303 |
+
"grad_norm": 4.70124626159668,
|
| 304 |
+
"learning_rate": 1.9179067140444246e-05,
|
| 305 |
+
"loss": 1.7027,
|
| 306 |
+
"step": 1566
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 18.0,
|
| 310 |
+
"eval_accuracy": 0.37317784256559766,
|
| 311 |
+
"eval_loss": 1.6329833269119263,
|
| 312 |
+
"eval_macro_f1": 0.1832569421283258,
|
| 313 |
+
"eval_runtime": 0.0773,
|
| 314 |
+
"eval_samples_per_second": 8871.989,
|
| 315 |
+
"eval_steps_per_second": 142.262,
|
| 316 |
+
"step": 1566
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 19.0,
|
| 320 |
+
"grad_norm": 4.254021167755127,
|
| 321 |
+
"learning_rate": 1.9041734901788285e-05,
|
| 322 |
+
"loss": 1.6776,
|
| 323 |
+
"step": 1653
|
| 324 |
+
},
|
| 325 |
+
{
|
| 326 |
+
"epoch": 19.0,
|
| 327 |
+
"eval_accuracy": 0.3760932944606414,
|
| 328 |
+
"eval_loss": 1.6269856691360474,
|
| 329 |
+
"eval_macro_f1": 0.18890492604023376,
|
| 330 |
+
"eval_runtime": 0.0793,
|
| 331 |
+
"eval_samples_per_second": 8655.516,
|
| 332 |
+
"eval_steps_per_second": 138.791,
|
| 333 |
+
"step": 1653
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 20.0,
|
| 337 |
+
"grad_norm": 3.7426421642303467,
|
| 338 |
+
"learning_rate": 1.8894358306898934e-05,
|
| 339 |
+
"loss": 1.6651,
|
| 340 |
+
"step": 1740
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 20.0,
|
| 344 |
+
"eval_accuracy": 0.37900874635568516,
|
| 345 |
+
"eval_loss": 1.621616005897522,
|
| 346 |
+
"eval_macro_f1": 0.1934068278580951,
|
| 347 |
+
"eval_runtime": 0.0784,
|
| 348 |
+
"eval_samples_per_second": 8746.314,
|
| 349 |
+
"eval_steps_per_second": 140.247,
|
| 350 |
+
"step": 1740
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 21.0,
|
| 354 |
+
"grad_norm": 4.341787338256836,
|
| 355 |
+
"learning_rate": 1.8737101074690274e-05,
|
| 356 |
+
"loss": 1.6694,
|
| 357 |
+
"step": 1827
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 21.0,
|
| 361 |
+
"eval_accuracy": 0.38338192419825073,
|
| 362 |
+
"eval_loss": 1.617226243019104,
|
| 363 |
+
"eval_macro_f1": 0.19340109033111008,
|
| 364 |
+
"eval_runtime": 0.0763,
|
| 365 |
+
"eval_samples_per_second": 8984.969,
|
| 366 |
+
"eval_steps_per_second": 144.074,
|
| 367 |
+
"step": 1827
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 22.0,
|
| 371 |
+
"grad_norm": 4.18576717376709,
|
| 372 |
+
"learning_rate": 1.8570137900359382e-05,
|
| 373 |
+
"loss": 1.6561,
|
| 374 |
+
"step": 1914
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 22.0,
|
| 378 |
+
"eval_accuracy": 0.38338192419825073,
|
| 379 |
+
"eval_loss": 1.6133029460906982,
|
| 380 |
+
"eval_macro_f1": 0.19942474851997433,
|
| 381 |
+
"eval_runtime": 0.0768,
|
| 382 |
+
"eval_samples_per_second": 8932.362,
|
| 383 |
+
"eval_steps_per_second": 143.23,
|
| 384 |
+
"step": 1914
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 23.0,
|
| 388 |
+
"grad_norm": 4.433280944824219,
|
| 389 |
+
"learning_rate": 1.8393654261319504e-05,
|
| 390 |
+
"loss": 1.6456,
|
| 391 |
+
"step": 2001
|
| 392 |
+
},
|
| 393 |
+
{
|
| 394 |
+
"epoch": 23.0,
|
| 395 |
+
"eval_accuracy": 0.3877551020408163,
|
| 396 |
+
"eval_loss": 1.6075658798217773,
|
| 397 |
+
"eval_macro_f1": 0.2021179986320824,
|
| 398 |
+
"eval_runtime": 0.0779,
|
| 399 |
+
"eval_samples_per_second": 8804.876,
|
| 400 |
+
"eval_steps_per_second": 141.186,
|
| 401 |
+
"step": 2001
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 24.0,
|
| 405 |
+
"grad_norm": 3.650712490081787,
|
| 406 |
+
"learning_rate": 1.8207846211155388e-05,
|
| 407 |
+
"loss": 1.6412,
|
| 408 |
+
"step": 2088
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 24.0,
|
| 412 |
+
"eval_accuracy": 0.39212827988338195,
|
| 413 |
+
"eval_loss": 1.6046576499938965,
|
| 414 |
+
"eval_macro_f1": 0.20558065728483735,
|
| 415 |
+
"eval_runtime": 0.0777,
|
| 416 |
+
"eval_samples_per_second": 8832.472,
|
| 417 |
+
"eval_steps_per_second": 141.629,
|
| 418 |
+
"step": 2088
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 25.0,
|
| 422 |
+
"grad_norm": 3.7270474433898926,
|
| 423 |
+
"learning_rate": 1.8012920161829693e-05,
|
| 424 |
+
"loss": 1.6369,
|
| 425 |
+
"step": 2175
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 25.0,
|
| 429 |
+
"eval_accuracy": 0.39504373177842567,
|
| 430 |
+
"eval_loss": 1.6002745628356934,
|
| 431 |
+
"eval_macro_f1": 0.2085329794328549,
|
| 432 |
+
"eval_runtime": 0.0763,
|
| 433 |
+
"eval_samples_per_second": 8986.091,
|
| 434 |
+
"eval_steps_per_second": 144.092,
|
| 435 |
+
"step": 2175
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 26.0,
|
| 439 |
+
"grad_norm": 3.5878360271453857,
|
| 440 |
+
"learning_rate": 1.7809092654382368e-05,
|
| 441 |
+
"loss": 1.6141,
|
| 442 |
+
"step": 2262
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 26.0,
|
| 446 |
+
"eval_accuracy": 0.39941690962099125,
|
| 447 |
+
"eval_loss": 1.5953983068466187,
|
| 448 |
+
"eval_macro_f1": 0.21136043336239665,
|
| 449 |
+
"eval_runtime": 0.0767,
|
| 450 |
+
"eval_samples_per_second": 8941.745,
|
| 451 |
+
"eval_steps_per_second": 143.381,
|
| 452 |
+
"step": 2262
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 27.0,
|
| 456 |
+
"grad_norm": 3.669312000274658,
|
| 457 |
+
"learning_rate": 1.7596590118377787e-05,
|
| 458 |
+
"loss": 1.5989,
|
| 459 |
+
"step": 2349
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 27.0,
|
| 463 |
+
"eval_accuracy": 0.40233236151603496,
|
| 464 |
+
"eval_loss": 1.5911133289337158,
|
| 465 |
+
"eval_macro_f1": 0.21358021621926357,
|
| 466 |
+
"eval_runtime": 0.0772,
|
| 467 |
+
"eval_samples_per_second": 8884.756,
|
| 468 |
+
"eval_steps_per_second": 142.467,
|
| 469 |
+
"step": 2349
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 28.0,
|
| 473 |
+
"grad_norm": 3.686958074569702,
|
| 474 |
+
"learning_rate": 1.7375648620366817e-05,
|
| 475 |
+
"loss": 1.6096,
|
| 476 |
+
"step": 2436
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 28.0,
|
| 480 |
+
"eval_accuracy": 0.40524781341107874,
|
| 481 |
+
"eval_loss": 1.5873298645019531,
|
| 482 |
+
"eval_macro_f1": 0.214485741970254,
|
| 483 |
+
"eval_runtime": 0.08,
|
| 484 |
+
"eval_samples_per_second": 8573.194,
|
| 485 |
+
"eval_steps_per_second": 137.471,
|
| 486 |
+
"step": 2436
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 29.0,
|
| 490 |
+
"grad_norm": 3.553083896636963,
|
| 491 |
+
"learning_rate": 1.7146513601643282e-05,
|
| 492 |
+
"loss": 1.6039,
|
| 493 |
+
"step": 2523
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 29.0,
|
| 497 |
+
"eval_accuracy": 0.4067055393586006,
|
| 498 |
+
"eval_loss": 1.584189534187317,
|
| 499 |
+
"eval_macro_f1": 0.21667857809163207,
|
| 500 |
+
"eval_runtime": 0.0762,
|
| 501 |
+
"eval_samples_per_second": 9000.399,
|
| 502 |
+
"eval_steps_per_second": 144.321,
|
| 503 |
+
"step": 2523
|
| 504 |
+
},
|
| 505 |
+
{
|
| 506 |
+
"epoch": 30.0,
|
| 507 |
+
"grad_norm": 3.9078423976898193,
|
| 508 |
+
"learning_rate": 1.6909439605586156e-05,
|
| 509 |
+
"loss": 1.5928,
|
| 510 |
+
"step": 2610
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 30.0,
|
| 514 |
+
"eval_accuracy": 0.40816326530612246,
|
| 515 |
+
"eval_loss": 1.579264521598816,
|
| 516 |
+
"eval_macro_f1": 0.21831730879606145,
|
| 517 |
+
"eval_runtime": 0.0775,
|
| 518 |
+
"eval_samples_per_second": 8854.652,
|
| 519 |
+
"eval_steps_per_second": 141.984,
|
| 520 |
+
"step": 2610
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 31.0,
|
| 524 |
+
"grad_norm": 3.7723805904388428,
|
| 525 |
+
"learning_rate": 1.6664689994890307e-05,
|
| 526 |
+
"loss": 1.5824,
|
| 527 |
+
"step": 2697
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 31.0,
|
| 531 |
+
"eval_accuracy": 0.40816326530612246,
|
| 532 |
+
"eval_loss": 1.5762993097305298,
|
| 533 |
+
"eval_macro_f1": 0.21682052505544805,
|
| 534 |
+
"eval_runtime": 0.0776,
|
| 535 |
+
"eval_samples_per_second": 8835.265,
|
| 536 |
+
"eval_steps_per_second": 141.673,
|
| 537 |
+
"step": 2697
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 32.0,
|
| 541 |
+
"grad_norm": 4.051678657531738,
|
| 542 |
+
"learning_rate": 1.641253665900002e-05,
|
| 543 |
+
"loss": 1.5877,
|
| 544 |
+
"step": 2784
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 32.0,
|
| 548 |
+
"eval_accuracy": 0.41545189504373176,
|
| 549 |
+
"eval_loss": 1.5732570886611938,
|
| 550 |
+
"eval_macro_f1": 0.2262251950436546,
|
| 551 |
+
"eval_runtime": 0.0769,
|
| 552 |
+
"eval_samples_per_second": 8918.878,
|
| 553 |
+
"eval_steps_per_second": 143.014,
|
| 554 |
+
"step": 2784
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 33.0,
|
| 558 |
+
"grad_norm": 3.396827459335327,
|
| 559 |
+
"learning_rate": 1.6153259712070225e-05,
|
| 560 |
+
"loss": 1.5722,
|
| 561 |
+
"step": 2871
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 33.0,
|
| 565 |
+
"eval_accuracy": 0.4110787172011662,
|
| 566 |
+
"eval_loss": 1.5706168413162231,
|
| 567 |
+
"eval_macro_f1": 0.22060087456248262,
|
| 568 |
+
"eval_runtime": 0.0769,
|
| 569 |
+
"eval_samples_per_second": 8923.138,
|
| 570 |
+
"eval_steps_per_second": 143.082,
|
| 571 |
+
"step": 2871
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 34.0,
|
| 575 |
+
"grad_norm": 3.510072708129883,
|
| 576 |
+
"learning_rate": 1.5887147181791e-05,
|
| 577 |
+
"loss": 1.5649,
|
| 578 |
+
"step": 2958
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 34.0,
|
| 582 |
+
"eval_accuracy": 0.41690962099125367,
|
| 583 |
+
"eval_loss": 1.5673753023147583,
|
| 584 |
+
"eval_macro_f1": 0.2265284337566022,
|
| 585 |
+
"eval_runtime": 0.0781,
|
| 586 |
+
"eval_samples_per_second": 8778.335,
|
| 587 |
+
"eval_steps_per_second": 140.76,
|
| 588 |
+
"step": 2958
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 35.0,
|
| 592 |
+
"grad_norm": 3.531944513320923,
|
| 593 |
+
"learning_rate": 1.5614494689421032e-05,
|
| 594 |
+
"loss": 1.5662,
|
| 595 |
+
"step": 3045
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 35.0,
|
| 599 |
+
"eval_accuracy": 0.4227405247813411,
|
| 600 |
+
"eval_loss": 1.5635616779327393,
|
| 601 |
+
"eval_macro_f1": 0.23237846476317717,
|
| 602 |
+
"eval_runtime": 0.0769,
|
| 603 |
+
"eval_samples_per_second": 8921.948,
|
| 604 |
+
"eval_steps_per_second": 143.063,
|
| 605 |
+
"step": 3045
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 36.0,
|
| 609 |
+
"grad_norm": 3.724010944366455,
|
| 610 |
+
"learning_rate": 1.533560512138543e-05,
|
| 611 |
+
"loss": 1.5545,
|
| 612 |
+
"step": 3132
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 36.0,
|
| 616 |
+
"eval_accuracy": 0.42419825072886297,
|
| 617 |
+
"eval_loss": 1.5617172718048096,
|
| 618 |
+
"eval_macro_f1": 0.23396270153240778,
|
| 619 |
+
"eval_runtime": 0.0766,
|
| 620 |
+
"eval_samples_per_second": 8960.737,
|
| 621 |
+
"eval_steps_per_second": 143.685,
|
| 622 |
+
"step": 3132
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"epoch": 37.0,
|
| 626 |
+
"grad_norm": 3.6395723819732666,
|
| 627 |
+
"learning_rate": 1.5050788292802812e-05,
|
| 628 |
+
"loss": 1.5416,
|
| 629 |
+
"step": 3219
|
| 630 |
+
},
|
| 631 |
+
{
|
| 632 |
+
"epoch": 37.0,
|
| 633 |
+
"eval_accuracy": 0.43440233236151604,
|
| 634 |
+
"eval_loss": 1.5581672191619873,
|
| 635 |
+
"eval_macro_f1": 0.24389742844346657,
|
| 636 |
+
"eval_runtime": 0.0768,
|
| 637 |
+
"eval_samples_per_second": 8935.719,
|
| 638 |
+
"eval_steps_per_second": 143.284,
|
| 639 |
+
"step": 3219
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 38.0,
|
| 643 |
+
"grad_norm": 4.1144866943359375,
|
| 644 |
+
"learning_rate": 1.4760360603315362e-05,
|
| 645 |
+
"loss": 1.5351,
|
| 646 |
+
"step": 3306
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 38.0,
|
| 650 |
+
"eval_accuracy": 0.4329446064139942,
|
| 651 |
+
"eval_loss": 1.55453622341156,
|
| 652 |
+
"eval_macro_f1": 0.23991444298311637,
|
| 653 |
+
"eval_runtime": 0.0769,
|
| 654 |
+
"eval_samples_per_second": 8923.913,
|
| 655 |
+
"eval_steps_per_second": 143.095,
|
| 656 |
+
"step": 3306
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 39.0,
|
| 660 |
+
"grad_norm": 3.656245708465576,
|
| 661 |
+
"learning_rate": 1.4464644685604184e-05,
|
| 662 |
+
"loss": 1.5424,
|
| 663 |
+
"step": 3393
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 39.0,
|
| 667 |
+
"eval_accuracy": 0.4329446064139942,
|
| 668 |
+
"eval_loss": 1.5531222820281982,
|
| 669 |
+
"eval_macro_f1": 0.24107844449857171,
|
| 670 |
+
"eval_runtime": 0.0777,
|
| 671 |
+
"eval_samples_per_second": 8833.123,
|
| 672 |
+
"eval_steps_per_second": 141.639,
|
| 673 |
+
"step": 3393
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 40.0,
|
| 677 |
+
"grad_norm": 3.530606746673584,
|
| 678 |
+
"learning_rate": 1.41639690469805e-05,
|
| 679 |
+
"loss": 1.5232,
|
| 680 |
+
"step": 3480
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 40.0,
|
| 684 |
+
"eval_accuracy": 0.4329446064139942,
|
| 685 |
+
"eval_loss": 1.5503716468811035,
|
| 686 |
+
"eval_macro_f1": 0.24171834592844124,
|
| 687 |
+
"eval_runtime": 0.0767,
|
| 688 |
+
"eval_samples_per_second": 8939.189,
|
| 689 |
+
"eval_steps_per_second": 143.34,
|
| 690 |
+
"step": 3480
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 41.0,
|
| 694 |
+
"grad_norm": 3.4572715759277344,
|
| 695 |
+
"learning_rate": 1.3858667704450763e-05,
|
| 696 |
+
"loss": 1.5277,
|
| 697 |
+
"step": 3567
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 41.0,
|
| 701 |
+
"eval_accuracy": 0.43440233236151604,
|
| 702 |
+
"eval_loss": 1.5470139980316162,
|
| 703 |
+
"eval_macro_f1": 0.24209898836089624,
|
| 704 |
+
"eval_runtime": 0.0769,
|
| 705 |
+
"eval_samples_per_second": 8921.395,
|
| 706 |
+
"eval_steps_per_second": 143.054,
|
| 707 |
+
"step": 3567
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 42.0,
|
| 711 |
+
"grad_norm": 4.217586517333984,
|
| 712 |
+
"learning_rate": 1.3549079813661123e-05,
|
| 713 |
+
"loss": 1.5112,
|
| 714 |
+
"step": 3654
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 42.0,
|
| 718 |
+
"eval_accuracy": 0.4446064139941691,
|
| 719 |
+
"eval_loss": 1.5440438985824585,
|
| 720 |
+
"eval_macro_f1": 0.25175856147050574,
|
| 721 |
+
"eval_runtime": 0.0767,
|
| 722 |
+
"eval_samples_per_second": 8948.113,
|
| 723 |
+
"eval_steps_per_second": 143.483,
|
| 724 |
+
"step": 3654
|
| 725 |
+
}
|
| 726 |
+
],
|
| 727 |
+
"logging_steps": 87,
|
| 728 |
+
"max_steps": 8700,
|
| 729 |
+
"num_input_tokens_seen": 0,
|
| 730 |
+
"num_train_epochs": 100,
|
| 731 |
+
"save_steps": 500,
|
| 732 |
+
"stateful_callbacks": {
|
| 733 |
+
"EarlyStoppingCallback": {
|
| 734 |
+
"args": {
|
| 735 |
+
"early_stopping_patience": 5,
|
| 736 |
+
"early_stopping_threshold": 0.0
|
| 737 |
+
},
|
| 738 |
+
"attributes": {
|
| 739 |
+
"early_stopping_patience_counter": 0
|
| 740 |
+
}
|
| 741 |
+
},
|
| 742 |
+
"TrainerControl": {
|
| 743 |
+
"args": {
|
| 744 |
+
"should_epoch_stop": false,
|
| 745 |
+
"should_evaluate": false,
|
| 746 |
+
"should_log": false,
|
| 747 |
+
"should_save": true,
|
| 748 |
+
"should_training_stop": false
|
| 749 |
+
},
|
| 750 |
+
"attributes": {}
|
| 751 |
+
}
|
| 752 |
+
},
|
| 753 |
+
"total_flos": 0.0,
|
| 754 |
+
"train_batch_size": 64,
|
| 755 |
+
"trial_name": null,
|
| 756 |
+
"trial_params": null
|
| 757 |
+
}
|