Upload trainer_state.json
Browse files- trainer_state.json +898 -0
trainer_state.json
ADDED
|
@@ -0,0 +1,898 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 49.01960784313726,
|
| 5 |
+
"global_step": 12500,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 1.0,
|
| 12 |
+
"learning_rate": 4.9e-05,
|
| 13 |
+
"loss": 28.6186,
|
| 14 |
+
"step": 255
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 1.0,
|
| 18 |
+
"eval_accuracy": 0.0,
|
| 19 |
+
"eval_f1_macro": 0.0,
|
| 20 |
+
"eval_gen_len": 1.1552,
|
| 21 |
+
"eval_loss": 7.702219009399414,
|
| 22 |
+
"eval_precision": 0.0,
|
| 23 |
+
"eval_recall": 0.0,
|
| 24 |
+
"eval_runtime": 16.3347,
|
| 25 |
+
"eval_samples_per_second": 298.2,
|
| 26 |
+
"step": 255
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 2.0,
|
| 30 |
+
"learning_rate": 4.8e-05,
|
| 31 |
+
"loss": 5.2711,
|
| 32 |
+
"step": 510
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 2.0,
|
| 36 |
+
"eval_accuracy": 0.2498,
|
| 37 |
+
"eval_f1_macro": 0.0493,
|
| 38 |
+
"eval_gen_len": 2.2652,
|
| 39 |
+
"eval_loss": 1.0767085552215576,
|
| 40 |
+
"eval_precision": 0.0424,
|
| 41 |
+
"eval_recall": 0.0797,
|
| 42 |
+
"eval_runtime": 18.0067,
|
| 43 |
+
"eval_samples_per_second": 270.51,
|
| 44 |
+
"step": 510
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 3.0,
|
| 48 |
+
"learning_rate": 4.7e-05,
|
| 49 |
+
"loss": 1.2251,
|
| 50 |
+
"step": 765
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 3.0,
|
| 54 |
+
"eval_accuracy": 0.3798,
|
| 55 |
+
"eval_f1_macro": 0.1033,
|
| 56 |
+
"eval_gen_len": 2.2634,
|
| 57 |
+
"eval_loss": 0.8948290944099426,
|
| 58 |
+
"eval_precision": 0.0823,
|
| 59 |
+
"eval_recall": 0.1436,
|
| 60 |
+
"eval_runtime": 18.1477,
|
| 61 |
+
"eval_samples_per_second": 268.409,
|
| 62 |
+
"step": 765
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 4.0,
|
| 66 |
+
"learning_rate": 4.600000000000001e-05,
|
| 67 |
+
"loss": 1.0537,
|
| 68 |
+
"step": 1020
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 4.0,
|
| 72 |
+
"eval_accuracy": 0.4003,
|
| 73 |
+
"eval_f1_macro": 0.1197,
|
| 74 |
+
"eval_gen_len": 2.2447,
|
| 75 |
+
"eval_loss": 0.8492642641067505,
|
| 76 |
+
"eval_precision": 0.1675,
|
| 77 |
+
"eval_recall": 0.1574,
|
| 78 |
+
"eval_runtime": 18.1704,
|
| 79 |
+
"eval_samples_per_second": 268.073,
|
| 80 |
+
"step": 1020
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 5.0,
|
| 84 |
+
"learning_rate": 4.5e-05,
|
| 85 |
+
"loss": 0.9861,
|
| 86 |
+
"step": 1275
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 5.0,
|
| 90 |
+
"eval_accuracy": 0.4192,
|
| 91 |
+
"eval_f1_macro": 0.1713,
|
| 92 |
+
"eval_gen_len": 2.2788,
|
| 93 |
+
"eval_loss": 0.8204376101493835,
|
| 94 |
+
"eval_precision": 0.2101,
|
| 95 |
+
"eval_recall": 0.1903,
|
| 96 |
+
"eval_runtime": 18.2564,
|
| 97 |
+
"eval_samples_per_second": 266.811,
|
| 98 |
+
"step": 1275
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 6.0,
|
| 102 |
+
"learning_rate": 4.4000000000000006e-05,
|
| 103 |
+
"loss": 0.9487,
|
| 104 |
+
"step": 1530
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 6.0,
|
| 108 |
+
"eval_accuracy": 0.4256,
|
| 109 |
+
"eval_f1_macro": 0.1852,
|
| 110 |
+
"eval_gen_len": 2.2798,
|
| 111 |
+
"eval_loss": 0.8069906830787659,
|
| 112 |
+
"eval_precision": 0.2583,
|
| 113 |
+
"eval_recall": 0.2033,
|
| 114 |
+
"eval_runtime": 18.2457,
|
| 115 |
+
"eval_samples_per_second": 266.967,
|
| 116 |
+
"step": 1530
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 7.0,
|
| 120 |
+
"learning_rate": 4.3e-05,
|
| 121 |
+
"loss": 0.9149,
|
| 122 |
+
"step": 1785
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 7.0,
|
| 126 |
+
"eval_accuracy": 0.4309,
|
| 127 |
+
"eval_f1_macro": 0.1939,
|
| 128 |
+
"eval_gen_len": 2.3252,
|
| 129 |
+
"eval_loss": 0.7942913770675659,
|
| 130 |
+
"eval_precision": 0.2515,
|
| 131 |
+
"eval_recall": 0.2152,
|
| 132 |
+
"eval_runtime": 18.3033,
|
| 133 |
+
"eval_samples_per_second": 266.127,
|
| 134 |
+
"step": 1785
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 8.0,
|
| 138 |
+
"learning_rate": 4.2e-05,
|
| 139 |
+
"loss": 0.8982,
|
| 140 |
+
"step": 2040
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 8.0,
|
| 144 |
+
"eval_accuracy": 0.4402,
|
| 145 |
+
"eval_f1_macro": 0.2158,
|
| 146 |
+
"eval_gen_len": 2.2735,
|
| 147 |
+
"eval_loss": 0.7909517884254456,
|
| 148 |
+
"eval_precision": 0.3176,
|
| 149 |
+
"eval_recall": 0.2341,
|
| 150 |
+
"eval_runtime": 18.2233,
|
| 151 |
+
"eval_samples_per_second": 267.295,
|
| 152 |
+
"step": 2040
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 9.0,
|
| 156 |
+
"learning_rate": 4.1e-05,
|
| 157 |
+
"loss": 0.8755,
|
| 158 |
+
"step": 2295
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 9.0,
|
| 162 |
+
"eval_accuracy": 0.4488,
|
| 163 |
+
"eval_f1_macro": 0.2325,
|
| 164 |
+
"eval_gen_len": 2.2704,
|
| 165 |
+
"eval_loss": 0.7753563523292542,
|
| 166 |
+
"eval_precision": 0.3133,
|
| 167 |
+
"eval_recall": 0.2408,
|
| 168 |
+
"eval_runtime": 18.3149,
|
| 169 |
+
"eval_samples_per_second": 265.959,
|
| 170 |
+
"step": 2295
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 10.0,
|
| 174 |
+
"learning_rate": 4e-05,
|
| 175 |
+
"loss": 0.861,
|
| 176 |
+
"step": 2550
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 10.0,
|
| 180 |
+
"eval_accuracy": 0.4482,
|
| 181 |
+
"eval_f1_macro": 0.2398,
|
| 182 |
+
"eval_gen_len": 2.2985,
|
| 183 |
+
"eval_loss": 0.7741720676422119,
|
| 184 |
+
"eval_precision": 0.3266,
|
| 185 |
+
"eval_recall": 0.2503,
|
| 186 |
+
"eval_runtime": 18.2556,
|
| 187 |
+
"eval_samples_per_second": 266.822,
|
| 188 |
+
"step": 2550
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 11.0,
|
| 192 |
+
"learning_rate": 3.9000000000000006e-05,
|
| 193 |
+
"loss": 0.8454,
|
| 194 |
+
"step": 2805
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 11.0,
|
| 198 |
+
"eval_accuracy": 0.4527,
|
| 199 |
+
"eval_f1_macro": 0.251,
|
| 200 |
+
"eval_gen_len": 2.2889,
|
| 201 |
+
"eval_loss": 0.7647759318351746,
|
| 202 |
+
"eval_precision": 0.3233,
|
| 203 |
+
"eval_recall": 0.2525,
|
| 204 |
+
"eval_runtime": 18.3159,
|
| 205 |
+
"eval_samples_per_second": 265.944,
|
| 206 |
+
"step": 2805
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 12.0,
|
| 210 |
+
"learning_rate": 3.8e-05,
|
| 211 |
+
"loss": 0.8323,
|
| 212 |
+
"step": 3060
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 12.0,
|
| 216 |
+
"eval_accuracy": 0.4527,
|
| 217 |
+
"eval_f1_macro": 0.2481,
|
| 218 |
+
"eval_gen_len": 2.2753,
|
| 219 |
+
"eval_loss": 0.7750440239906311,
|
| 220 |
+
"eval_precision": 0.3238,
|
| 221 |
+
"eval_recall": 0.2549,
|
| 222 |
+
"eval_runtime": 18.2537,
|
| 223 |
+
"eval_samples_per_second": 266.85,
|
| 224 |
+
"step": 3060
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 13.0,
|
| 228 |
+
"learning_rate": 3.7e-05,
|
| 229 |
+
"loss": 0.8212,
|
| 230 |
+
"step": 3315
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 13.0,
|
| 234 |
+
"eval_accuracy": 0.4588,
|
| 235 |
+
"eval_f1_macro": 0.2551,
|
| 236 |
+
"eval_gen_len": 2.3053,
|
| 237 |
+
"eval_loss": 0.7678483724594116,
|
| 238 |
+
"eval_precision": 0.3446,
|
| 239 |
+
"eval_recall": 0.2561,
|
| 240 |
+
"eval_runtime": 18.3189,
|
| 241 |
+
"eval_samples_per_second": 265.901,
|
| 242 |
+
"step": 3315
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 14.0,
|
| 246 |
+
"learning_rate": 3.6e-05,
|
| 247 |
+
"loss": 0.8167,
|
| 248 |
+
"step": 3570
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 14.0,
|
| 252 |
+
"eval_accuracy": 0.457,
|
| 253 |
+
"eval_f1_macro": 0.2555,
|
| 254 |
+
"eval_gen_len": 2.296,
|
| 255 |
+
"eval_loss": 0.7742500901222229,
|
| 256 |
+
"eval_precision": 0.3303,
|
| 257 |
+
"eval_recall": 0.2597,
|
| 258 |
+
"eval_runtime": 18.2816,
|
| 259 |
+
"eval_samples_per_second": 266.443,
|
| 260 |
+
"step": 3570
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 15.0,
|
| 264 |
+
"learning_rate": 3.5e-05,
|
| 265 |
+
"loss": 0.807,
|
| 266 |
+
"step": 3825
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 15.0,
|
| 270 |
+
"eval_accuracy": 0.4584,
|
| 271 |
+
"eval_f1_macro": 0.2645,
|
| 272 |
+
"eval_gen_len": 2.317,
|
| 273 |
+
"eval_loss": 0.7675516605377197,
|
| 274 |
+
"eval_precision": 0.3309,
|
| 275 |
+
"eval_recall": 0.2686,
|
| 276 |
+
"eval_runtime": 18.3188,
|
| 277 |
+
"eval_samples_per_second": 265.902,
|
| 278 |
+
"step": 3825
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 16.0,
|
| 282 |
+
"learning_rate": 3.4000000000000007e-05,
|
| 283 |
+
"loss": 0.7961,
|
| 284 |
+
"step": 4080
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 16.0,
|
| 288 |
+
"eval_accuracy": 0.4553,
|
| 289 |
+
"eval_f1_macro": 0.2618,
|
| 290 |
+
"eval_gen_len": 2.3398,
|
| 291 |
+
"eval_loss": 0.7658052444458008,
|
| 292 |
+
"eval_precision": 0.3443,
|
| 293 |
+
"eval_recall": 0.2678,
|
| 294 |
+
"eval_runtime": 18.2865,
|
| 295 |
+
"eval_samples_per_second": 266.371,
|
| 296 |
+
"step": 4080
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 17.0,
|
| 300 |
+
"learning_rate": 3.3e-05,
|
| 301 |
+
"loss": 0.793,
|
| 302 |
+
"step": 4335
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 17.0,
|
| 306 |
+
"eval_accuracy": 0.4642,
|
| 307 |
+
"eval_f1_macro": 0.2629,
|
| 308 |
+
"eval_gen_len": 2.3057,
|
| 309 |
+
"eval_loss": 0.760092556476593,
|
| 310 |
+
"eval_precision": 0.3423,
|
| 311 |
+
"eval_recall": 0.2651,
|
| 312 |
+
"eval_runtime": 18.3181,
|
| 313 |
+
"eval_samples_per_second": 265.911,
|
| 314 |
+
"step": 4335
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 18.0,
|
| 318 |
+
"learning_rate": 3.2000000000000005e-05,
|
| 319 |
+
"loss": 0.7792,
|
| 320 |
+
"step": 4590
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 18.0,
|
| 324 |
+
"eval_accuracy": 0.4613,
|
| 325 |
+
"eval_f1_macro": 0.2687,
|
| 326 |
+
"eval_gen_len": 2.3079,
|
| 327 |
+
"eval_loss": 0.7579958438873291,
|
| 328 |
+
"eval_precision": 0.3561,
|
| 329 |
+
"eval_recall": 0.2684,
|
| 330 |
+
"eval_runtime": 18.2367,
|
| 331 |
+
"eval_samples_per_second": 267.099,
|
| 332 |
+
"step": 4590
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 19.0,
|
| 336 |
+
"learning_rate": 3.1e-05,
|
| 337 |
+
"loss": 0.7704,
|
| 338 |
+
"step": 4845
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 19.0,
|
| 342 |
+
"eval_accuracy": 0.4654,
|
| 343 |
+
"eval_f1_macro": 0.2752,
|
| 344 |
+
"eval_gen_len": 2.3051,
|
| 345 |
+
"eval_loss": 0.7610541582107544,
|
| 346 |
+
"eval_precision": 0.375,
|
| 347 |
+
"eval_recall": 0.2723,
|
| 348 |
+
"eval_runtime": 18.3152,
|
| 349 |
+
"eval_samples_per_second": 265.954,
|
| 350 |
+
"step": 4845
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 20.0,
|
| 354 |
+
"learning_rate": 3e-05,
|
| 355 |
+
"loss": 0.7622,
|
| 356 |
+
"step": 5100
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 20.0,
|
| 360 |
+
"eval_accuracy": 0.4631,
|
| 361 |
+
"eval_f1_macro": 0.2759,
|
| 362 |
+
"eval_gen_len": 2.2804,
|
| 363 |
+
"eval_loss": 0.7701175212860107,
|
| 364 |
+
"eval_precision": 0.3679,
|
| 365 |
+
"eval_recall": 0.2734,
|
| 366 |
+
"eval_runtime": 18.233,
|
| 367 |
+
"eval_samples_per_second": 267.153,
|
| 368 |
+
"step": 5100
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 21.0,
|
| 372 |
+
"learning_rate": 2.9e-05,
|
| 373 |
+
"loss": 0.7529,
|
| 374 |
+
"step": 5355
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 21.0,
|
| 378 |
+
"eval_accuracy": 0.4617,
|
| 379 |
+
"eval_f1_macro": 0.279,
|
| 380 |
+
"eval_gen_len": 2.3346,
|
| 381 |
+
"eval_loss": 0.7670521140098572,
|
| 382 |
+
"eval_precision": 0.3509,
|
| 383 |
+
"eval_recall": 0.279,
|
| 384 |
+
"eval_runtime": 18.3163,
|
| 385 |
+
"eval_samples_per_second": 265.937,
|
| 386 |
+
"step": 5355
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 22.0,
|
| 390 |
+
"learning_rate": 2.8000000000000003e-05,
|
| 391 |
+
"loss": 0.747,
|
| 392 |
+
"step": 5610
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 22.0,
|
| 396 |
+
"eval_accuracy": 0.4662,
|
| 397 |
+
"eval_f1_macro": 0.2796,
|
| 398 |
+
"eval_gen_len": 2.3069,
|
| 399 |
+
"eval_loss": 0.7644122242927551,
|
| 400 |
+
"eval_precision": 0.3613,
|
| 401 |
+
"eval_recall": 0.2766,
|
| 402 |
+
"eval_runtime": 18.2901,
|
| 403 |
+
"eval_samples_per_second": 266.318,
|
| 404 |
+
"step": 5610
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 23.0,
|
| 408 |
+
"learning_rate": 2.7000000000000002e-05,
|
| 409 |
+
"loss": 0.7425,
|
| 410 |
+
"step": 5865
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 23.0,
|
| 414 |
+
"eval_accuracy": 0.4648,
|
| 415 |
+
"eval_f1_macro": 0.2723,
|
| 416 |
+
"eval_gen_len": 2.3047,
|
| 417 |
+
"eval_loss": 0.7637659907341003,
|
| 418 |
+
"eval_precision": 0.3654,
|
| 419 |
+
"eval_recall": 0.2699,
|
| 420 |
+
"eval_runtime": 18.318,
|
| 421 |
+
"eval_samples_per_second": 265.913,
|
| 422 |
+
"step": 5865
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 24.0,
|
| 426 |
+
"learning_rate": 2.6000000000000002e-05,
|
| 427 |
+
"loss": 0.7351,
|
| 428 |
+
"step": 6120
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 24.0,
|
| 432 |
+
"eval_accuracy": 0.4689,
|
| 433 |
+
"eval_f1_macro": 0.2763,
|
| 434 |
+
"eval_gen_len": 2.3118,
|
| 435 |
+
"eval_loss": 0.7734237909317017,
|
| 436 |
+
"eval_precision": 0.3711,
|
| 437 |
+
"eval_recall": 0.2742,
|
| 438 |
+
"eval_runtime": 18.3098,
|
| 439 |
+
"eval_samples_per_second": 266.032,
|
| 440 |
+
"step": 6120
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 25.0,
|
| 444 |
+
"learning_rate": 2.5e-05,
|
| 445 |
+
"loss": 0.7279,
|
| 446 |
+
"step": 6375
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 25.0,
|
| 450 |
+
"eval_accuracy": 0.4668,
|
| 451 |
+
"eval_f1_macro": 0.2753,
|
| 452 |
+
"eval_gen_len": 2.3069,
|
| 453 |
+
"eval_loss": 0.7742076516151428,
|
| 454 |
+
"eval_precision": 0.3749,
|
| 455 |
+
"eval_recall": 0.2717,
|
| 456 |
+
"eval_runtime": 18.319,
|
| 457 |
+
"eval_samples_per_second": 265.899,
|
| 458 |
+
"step": 6375
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 26.0,
|
| 462 |
+
"learning_rate": 2.4e-05,
|
| 463 |
+
"loss": 0.7251,
|
| 464 |
+
"step": 6630
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 26.0,
|
| 468 |
+
"eval_accuracy": 0.4675,
|
| 469 |
+
"eval_f1_macro": 0.2824,
|
| 470 |
+
"eval_gen_len": 2.319,
|
| 471 |
+
"eval_loss": 0.777134358882904,
|
| 472 |
+
"eval_precision": 0.3585,
|
| 473 |
+
"eval_recall": 0.2806,
|
| 474 |
+
"eval_runtime": 18.2882,
|
| 475 |
+
"eval_samples_per_second": 266.346,
|
| 476 |
+
"step": 6630
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 27.0,
|
| 480 |
+
"learning_rate": 2.3000000000000003e-05,
|
| 481 |
+
"loss": 0.7149,
|
| 482 |
+
"step": 6885
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 27.0,
|
| 486 |
+
"eval_accuracy": 0.4699,
|
| 487 |
+
"eval_f1_macro": 0.2853,
|
| 488 |
+
"eval_gen_len": 2.316,
|
| 489 |
+
"eval_loss": 0.7695617079734802,
|
| 490 |
+
"eval_precision": 0.3587,
|
| 491 |
+
"eval_recall": 0.2807,
|
| 492 |
+
"eval_runtime": 18.3083,
|
| 493 |
+
"eval_samples_per_second": 266.054,
|
| 494 |
+
"step": 6885
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 28.0,
|
| 498 |
+
"learning_rate": 2.2000000000000003e-05,
|
| 499 |
+
"loss": 0.7122,
|
| 500 |
+
"step": 7140
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 28.0,
|
| 504 |
+
"eval_accuracy": 0.4644,
|
| 505 |
+
"eval_f1_macro": 0.2764,
|
| 506 |
+
"eval_gen_len": 2.3059,
|
| 507 |
+
"eval_loss": 0.7785841822624207,
|
| 508 |
+
"eval_precision": 0.3624,
|
| 509 |
+
"eval_recall": 0.2775,
|
| 510 |
+
"eval_runtime": 18.2576,
|
| 511 |
+
"eval_samples_per_second": 266.793,
|
| 512 |
+
"step": 7140
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 29.0,
|
| 516 |
+
"learning_rate": 2.1e-05,
|
| 517 |
+
"loss": 0.7105,
|
| 518 |
+
"step": 7395
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 29.0,
|
| 522 |
+
"eval_accuracy": 0.4681,
|
| 523 |
+
"eval_f1_macro": 0.287,
|
| 524 |
+
"eval_gen_len": 2.2829,
|
| 525 |
+
"eval_loss": 0.7693188786506653,
|
| 526 |
+
"eval_precision": 0.3864,
|
| 527 |
+
"eval_recall": 0.2772,
|
| 528 |
+
"eval_runtime": 18.3321,
|
| 529 |
+
"eval_samples_per_second": 265.709,
|
| 530 |
+
"step": 7395
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 30.0,
|
| 534 |
+
"learning_rate": 2e-05,
|
| 535 |
+
"loss": 0.7039,
|
| 536 |
+
"step": 7650
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 30.0,
|
| 540 |
+
"eval_accuracy": 0.4666,
|
| 541 |
+
"eval_f1_macro": 0.2828,
|
| 542 |
+
"eval_gen_len": 2.3014,
|
| 543 |
+
"eval_loss": 0.7784000039100647,
|
| 544 |
+
"eval_precision": 0.3699,
|
| 545 |
+
"eval_recall": 0.278,
|
| 546 |
+
"eval_runtime": 18.3242,
|
| 547 |
+
"eval_samples_per_second": 265.823,
|
| 548 |
+
"step": 7650
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 31.0,
|
| 552 |
+
"learning_rate": 1.9e-05,
|
| 553 |
+
"loss": 0.6944,
|
| 554 |
+
"step": 7905
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 31.0,
|
| 558 |
+
"eval_accuracy": 0.4689,
|
| 559 |
+
"eval_f1_macro": 0.2873,
|
| 560 |
+
"eval_gen_len": 2.3211,
|
| 561 |
+
"eval_loss": 0.7755969166755676,
|
| 562 |
+
"eval_precision": 0.3557,
|
| 563 |
+
"eval_recall": 0.2835,
|
| 564 |
+
"eval_runtime": 18.3091,
|
| 565 |
+
"eval_samples_per_second": 266.043,
|
| 566 |
+
"step": 7905
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 32.0,
|
| 570 |
+
"learning_rate": 1.8e-05,
|
| 571 |
+
"loss": 0.6949,
|
| 572 |
+
"step": 8160
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 32.0,
|
| 576 |
+
"eval_accuracy": 0.4689,
|
| 577 |
+
"eval_f1_macro": 0.2854,
|
| 578 |
+
"eval_gen_len": 2.3176,
|
| 579 |
+
"eval_loss": 0.7752977609634399,
|
| 580 |
+
"eval_precision": 0.3604,
|
| 581 |
+
"eval_recall": 0.2814,
|
| 582 |
+
"eval_runtime": 18.2988,
|
| 583 |
+
"eval_samples_per_second": 266.193,
|
| 584 |
+
"step": 8160
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 33.0,
|
| 588 |
+
"learning_rate": 1.7000000000000003e-05,
|
| 589 |
+
"loss": 0.6876,
|
| 590 |
+
"step": 8415
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 33.0,
|
| 594 |
+
"eval_accuracy": 0.4712,
|
| 595 |
+
"eval_f1_macro": 0.2907,
|
| 596 |
+
"eval_gen_len": 2.3196,
|
| 597 |
+
"eval_loss": 0.7808494567871094,
|
| 598 |
+
"eval_precision": 0.3858,
|
| 599 |
+
"eval_recall": 0.2848,
|
| 600 |
+
"eval_runtime": 18.3097,
|
| 601 |
+
"eval_samples_per_second": 266.034,
|
| 602 |
+
"step": 8415
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 34.0,
|
| 606 |
+
"learning_rate": 1.6000000000000003e-05,
|
| 607 |
+
"loss": 0.6872,
|
| 608 |
+
"step": 8670
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 34.0,
|
| 612 |
+
"eval_accuracy": 0.4695,
|
| 613 |
+
"eval_f1_macro": 0.289,
|
| 614 |
+
"eval_gen_len": 2.311,
|
| 615 |
+
"eval_loss": 0.77425217628479,
|
| 616 |
+
"eval_precision": 0.3482,
|
| 617 |
+
"eval_recall": 0.2849,
|
| 618 |
+
"eval_runtime": 18.2885,
|
| 619 |
+
"eval_samples_per_second": 266.342,
|
| 620 |
+
"step": 8670
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 35.0,
|
| 624 |
+
"learning_rate": 1.5e-05,
|
| 625 |
+
"loss": 0.685,
|
| 626 |
+
"step": 8925
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 35.0,
|
| 630 |
+
"eval_accuracy": 0.4677,
|
| 631 |
+
"eval_f1_macro": 0.2876,
|
| 632 |
+
"eval_gen_len": 2.3127,
|
| 633 |
+
"eval_loss": 0.7812894582748413,
|
| 634 |
+
"eval_precision": 0.3644,
|
| 635 |
+
"eval_recall": 0.2814,
|
| 636 |
+
"eval_runtime": 18.3325,
|
| 637 |
+
"eval_samples_per_second": 265.703,
|
| 638 |
+
"step": 8925
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 36.0,
|
| 642 |
+
"learning_rate": 1.4000000000000001e-05,
|
| 643 |
+
"loss": 0.681,
|
| 644 |
+
"step": 9180
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 36.0,
|
| 648 |
+
"eval_accuracy": 0.4636,
|
| 649 |
+
"eval_f1_macro": 0.2814,
|
| 650 |
+
"eval_gen_len": 2.3309,
|
| 651 |
+
"eval_loss": 0.7843227386474609,
|
| 652 |
+
"eval_precision": 0.3367,
|
| 653 |
+
"eval_recall": 0.2784,
|
| 654 |
+
"eval_runtime": 18.3007,
|
| 655 |
+
"eval_samples_per_second": 266.165,
|
| 656 |
+
"step": 9180
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 37.0,
|
| 660 |
+
"learning_rate": 1.3000000000000001e-05,
|
| 661 |
+
"loss": 0.6777,
|
| 662 |
+
"step": 9435
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 37.0,
|
| 666 |
+
"eval_accuracy": 0.4679,
|
| 667 |
+
"eval_f1_macro": 0.2887,
|
| 668 |
+
"eval_gen_len": 2.3244,
|
| 669 |
+
"eval_loss": 0.7880710363388062,
|
| 670 |
+
"eval_precision": 0.347,
|
| 671 |
+
"eval_recall": 0.2834,
|
| 672 |
+
"eval_runtime": 18.3303,
|
| 673 |
+
"eval_samples_per_second": 265.735,
|
| 674 |
+
"step": 9435
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 38.0,
|
| 678 |
+
"learning_rate": 1.2e-05,
|
| 679 |
+
"loss": 0.669,
|
| 680 |
+
"step": 9690
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 38.0,
|
| 684 |
+
"eval_accuracy": 0.4689,
|
| 685 |
+
"eval_f1_macro": 0.2915,
|
| 686 |
+
"eval_gen_len": 2.3153,
|
| 687 |
+
"eval_loss": 0.7872263789176941,
|
| 688 |
+
"eval_precision": 0.3575,
|
| 689 |
+
"eval_recall": 0.2861,
|
| 690 |
+
"eval_runtime": 18.2879,
|
| 691 |
+
"eval_samples_per_second": 266.351,
|
| 692 |
+
"step": 9690
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 39.0,
|
| 696 |
+
"learning_rate": 1.1000000000000001e-05,
|
| 697 |
+
"loss": 0.6724,
|
| 698 |
+
"step": 9945
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 39.0,
|
| 702 |
+
"eval_accuracy": 0.4687,
|
| 703 |
+
"eval_f1_macro": 0.2886,
|
| 704 |
+
"eval_gen_len": 2.3387,
|
| 705 |
+
"eval_loss": 0.7843508124351501,
|
| 706 |
+
"eval_precision": 0.3525,
|
| 707 |
+
"eval_recall": 0.2867,
|
| 708 |
+
"eval_runtime": 18.3152,
|
| 709 |
+
"eval_samples_per_second": 265.954,
|
| 710 |
+
"step": 9945
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 40.0,
|
| 714 |
+
"learning_rate": 1e-05,
|
| 715 |
+
"loss": 0.669,
|
| 716 |
+
"step": 10200
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 40.0,
|
| 720 |
+
"eval_accuracy": 0.4662,
|
| 721 |
+
"eval_f1_macro": 0.2876,
|
| 722 |
+
"eval_gen_len": 2.3285,
|
| 723 |
+
"eval_loss": 0.7868720889091492,
|
| 724 |
+
"eval_precision": 0.348,
|
| 725 |
+
"eval_recall": 0.2836,
|
| 726 |
+
"eval_runtime": 18.2931,
|
| 727 |
+
"eval_samples_per_second": 266.276,
|
| 728 |
+
"step": 10200
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 41.0,
|
| 732 |
+
"learning_rate": 9e-06,
|
| 733 |
+
"loss": 0.6636,
|
| 734 |
+
"step": 10455
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 41.0,
|
| 738 |
+
"eval_accuracy": 0.4683,
|
| 739 |
+
"eval_f1_macro": 0.2857,
|
| 740 |
+
"eval_gen_len": 2.3215,
|
| 741 |
+
"eval_loss": 0.790763795375824,
|
| 742 |
+
"eval_precision": 0.3488,
|
| 743 |
+
"eval_recall": 0.2821,
|
| 744 |
+
"eval_runtime": 18.3162,
|
| 745 |
+
"eval_samples_per_second": 265.94,
|
| 746 |
+
"step": 10455
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 42.0,
|
| 750 |
+
"learning_rate": 8.000000000000001e-06,
|
| 751 |
+
"loss": 0.6654,
|
| 752 |
+
"step": 10710
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 42.0,
|
| 756 |
+
"eval_accuracy": 0.4705,
|
| 757 |
+
"eval_f1_macro": 0.2912,
|
| 758 |
+
"eval_gen_len": 2.3225,
|
| 759 |
+
"eval_loss": 0.78489089012146,
|
| 760 |
+
"eval_precision": 0.3519,
|
| 761 |
+
"eval_recall": 0.2872,
|
| 762 |
+
"eval_runtime": 18.2957,
|
| 763 |
+
"eval_samples_per_second": 266.237,
|
| 764 |
+
"step": 10710
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 43.0,
|
| 768 |
+
"learning_rate": 7.000000000000001e-06,
|
| 769 |
+
"loss": 0.6626,
|
| 770 |
+
"step": 10965
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 43.0,
|
| 774 |
+
"eval_accuracy": 0.4707,
|
| 775 |
+
"eval_f1_macro": 0.2908,
|
| 776 |
+
"eval_gen_len": 2.3242,
|
| 777 |
+
"eval_loss": 0.7898949384689331,
|
| 778 |
+
"eval_precision": 0.3521,
|
| 779 |
+
"eval_recall": 0.2875,
|
| 780 |
+
"eval_runtime": 18.3173,
|
| 781 |
+
"eval_samples_per_second": 265.923,
|
| 782 |
+
"step": 10965
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 44.0,
|
| 786 |
+
"learning_rate": 6e-06,
|
| 787 |
+
"loss": 0.6619,
|
| 788 |
+
"step": 11220
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 44.0,
|
| 792 |
+
"eval_accuracy": 0.467,
|
| 793 |
+
"eval_f1_macro": 0.2874,
|
| 794 |
+
"eval_gen_len": 2.3233,
|
| 795 |
+
"eval_loss": 0.7907114028930664,
|
| 796 |
+
"eval_precision": 0.347,
|
| 797 |
+
"eval_recall": 0.2835,
|
| 798 |
+
"eval_runtime": 18.2912,
|
| 799 |
+
"eval_samples_per_second": 266.303,
|
| 800 |
+
"step": 11220
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 45.0,
|
| 804 |
+
"learning_rate": 5e-06,
|
| 805 |
+
"loss": 0.6593,
|
| 806 |
+
"step": 11475
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 45.0,
|
| 810 |
+
"eval_accuracy": 0.4687,
|
| 811 |
+
"eval_f1_macro": 0.2901,
|
| 812 |
+
"eval_gen_len": 2.3346,
|
| 813 |
+
"eval_loss": 0.7884129285812378,
|
| 814 |
+
"eval_precision": 0.3514,
|
| 815 |
+
"eval_recall": 0.2867,
|
| 816 |
+
"eval_runtime": 18.3091,
|
| 817 |
+
"eval_samples_per_second": 266.042,
|
| 818 |
+
"step": 11475
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 46.0,
|
| 822 |
+
"learning_rate": 4.000000000000001e-06,
|
| 823 |
+
"loss": 0.6562,
|
| 824 |
+
"step": 11730
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 46.0,
|
| 828 |
+
"eval_accuracy": 0.4697,
|
| 829 |
+
"eval_f1_macro": 0.2898,
|
| 830 |
+
"eval_gen_len": 2.3299,
|
| 831 |
+
"eval_loss": 0.7906068563461304,
|
| 832 |
+
"eval_precision": 0.3527,
|
| 833 |
+
"eval_recall": 0.285,
|
| 834 |
+
"eval_runtime": 18.2501,
|
| 835 |
+
"eval_samples_per_second": 266.902,
|
| 836 |
+
"step": 11730
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 47.0,
|
| 840 |
+
"learning_rate": 3e-06,
|
| 841 |
+
"loss": 0.6573,
|
| 842 |
+
"step": 11985
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 47.0,
|
| 846 |
+
"eval_accuracy": 0.4673,
|
| 847 |
+
"eval_f1_macro": 0.2879,
|
| 848 |
+
"eval_gen_len": 2.3285,
|
| 849 |
+
"eval_loss": 0.7910804748535156,
|
| 850 |
+
"eval_precision": 0.344,
|
| 851 |
+
"eval_recall": 0.2839,
|
| 852 |
+
"eval_runtime": 18.3029,
|
| 853 |
+
"eval_samples_per_second": 266.132,
|
| 854 |
+
"step": 11985
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 48.0,
|
| 858 |
+
"learning_rate": 2.0000000000000003e-06,
|
| 859 |
+
"loss": 0.6552,
|
| 860 |
+
"step": 12240
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 48.0,
|
| 864 |
+
"eval_accuracy": 0.4687,
|
| 865 |
+
"eval_f1_macro": 0.2903,
|
| 866 |
+
"eval_gen_len": 2.3324,
|
| 867 |
+
"eval_loss": 0.7917013764381409,
|
| 868 |
+
"eval_precision": 0.344,
|
| 869 |
+
"eval_recall": 0.2863,
|
| 870 |
+
"eval_runtime": 18.2692,
|
| 871 |
+
"eval_samples_per_second": 266.624,
|
| 872 |
+
"step": 12240
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 49.0,
|
| 876 |
+
"learning_rate": 1.0000000000000002e-06,
|
| 877 |
+
"loss": 0.6545,
|
| 878 |
+
"step": 12495
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 49.0,
|
| 882 |
+
"eval_accuracy": 0.4685,
|
| 883 |
+
"eval_f1_macro": 0.2896,
|
| 884 |
+
"eval_gen_len": 2.3322,
|
| 885 |
+
"eval_loss": 0.7909209728240967,
|
| 886 |
+
"eval_precision": 0.3462,
|
| 887 |
+
"eval_recall": 0.2856,
|
| 888 |
+
"eval_runtime": 18.2849,
|
| 889 |
+
"eval_samples_per_second": 266.395,
|
| 890 |
+
"step": 12495
|
| 891 |
+
}
|
| 892 |
+
],
|
| 893 |
+
"max_steps": 12750,
|
| 894 |
+
"num_train_epochs": 50,
|
| 895 |
+
"total_flos": 1.14107158131029e+17,
|
| 896 |
+
"trial_name": null,
|
| 897 |
+
"trial_params": null
|
| 898 |
+
}
|