End of training
Browse files- all_results.json +15 -0
- eval_results.json +10 -0
- train_results.json +8 -0
- trainer_state.json +721 -0
all_results.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_bleu": 34.2122,
|
| 4 |
+
"eval_gen_len": 65.0263,
|
| 5 |
+
"eval_loss": 0.7057932019233704,
|
| 6 |
+
"eval_runtime": 3541.072,
|
| 7 |
+
"eval_samples": 12422,
|
| 8 |
+
"eval_samples_per_second": 3.508,
|
| 9 |
+
"eval_steps_per_second": 0.219,
|
| 10 |
+
"train_loss": 1.0872868923270456,
|
| 11 |
+
"train_runtime": 30832.4334,
|
| 12 |
+
"train_samples": 62107,
|
| 13 |
+
"train_samples_per_second": 6.043,
|
| 14 |
+
"train_steps_per_second": 0.378
|
| 15 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_bleu": 34.2122,
|
| 4 |
+
"eval_gen_len": 65.0263,
|
| 5 |
+
"eval_loss": 0.7057932019233704,
|
| 6 |
+
"eval_runtime": 3541.072,
|
| 7 |
+
"eval_samples": 12422,
|
| 8 |
+
"eval_samples_per_second": 3.508,
|
| 9 |
+
"eval_steps_per_second": 0.219
|
| 10 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"train_loss": 1.0872868923270456,
|
| 4 |
+
"train_runtime": 30832.4334,
|
| 5 |
+
"train_samples": 62107,
|
| 6 |
+
"train_samples_per_second": 6.043,
|
| 7 |
+
"train_steps_per_second": 0.378
|
| 8 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,721 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"global_step": 11646,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.03,
|
| 12 |
+
"learning_rate": 4.9570668040528936e-05,
|
| 13 |
+
"loss": 2.5131,
|
| 14 |
+
"step": 100
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.05,
|
| 18 |
+
"learning_rate": 4.914133608105788e-05,
|
| 19 |
+
"loss": 2.1951,
|
| 20 |
+
"step": 200
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.08,
|
| 24 |
+
"learning_rate": 4.871200412158681e-05,
|
| 25 |
+
"loss": 1.9936,
|
| 26 |
+
"step": 300
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.1,
|
| 30 |
+
"learning_rate": 4.828267216211575e-05,
|
| 31 |
+
"loss": 1.8599,
|
| 32 |
+
"step": 400
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.13,
|
| 36 |
+
"learning_rate": 4.7853340202644685e-05,
|
| 37 |
+
"loss": 1.7738,
|
| 38 |
+
"step": 500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.15,
|
| 42 |
+
"learning_rate": 4.7424008243173625e-05,
|
| 43 |
+
"loss": 1.7447,
|
| 44 |
+
"step": 600
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.18,
|
| 48 |
+
"learning_rate": 4.699467628370256e-05,
|
| 49 |
+
"loss": 1.664,
|
| 50 |
+
"step": 700
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.21,
|
| 54 |
+
"learning_rate": 4.65653443242315e-05,
|
| 55 |
+
"loss": 1.6198,
|
| 56 |
+
"step": 800
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.23,
|
| 60 |
+
"learning_rate": 4.6136012364760434e-05,
|
| 61 |
+
"loss": 1.5546,
|
| 62 |
+
"step": 900
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.26,
|
| 66 |
+
"learning_rate": 4.570668040528937e-05,
|
| 67 |
+
"loss": 1.5249,
|
| 68 |
+
"step": 1000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.28,
|
| 72 |
+
"learning_rate": 4.527734844581831e-05,
|
| 73 |
+
"loss": 1.4905,
|
| 74 |
+
"step": 1100
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.31,
|
| 78 |
+
"learning_rate": 4.484801648634724e-05,
|
| 79 |
+
"loss": 1.4535,
|
| 80 |
+
"step": 1200
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.33,
|
| 84 |
+
"learning_rate": 4.441868452687618e-05,
|
| 85 |
+
"loss": 1.4336,
|
| 86 |
+
"step": 1300
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.36,
|
| 90 |
+
"learning_rate": 4.3989352567405116e-05,
|
| 91 |
+
"loss": 1.4008,
|
| 92 |
+
"step": 1400
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.39,
|
| 96 |
+
"learning_rate": 4.356002060793406e-05,
|
| 97 |
+
"loss": 1.3552,
|
| 98 |
+
"step": 1500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.41,
|
| 102 |
+
"learning_rate": 4.313068864846299e-05,
|
| 103 |
+
"loss": 1.3854,
|
| 104 |
+
"step": 1600
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.44,
|
| 108 |
+
"learning_rate": 4.270135668899193e-05,
|
| 109 |
+
"loss": 1.3722,
|
| 110 |
+
"step": 1700
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.46,
|
| 114 |
+
"learning_rate": 4.2272024729520865e-05,
|
| 115 |
+
"loss": 1.3153,
|
| 116 |
+
"step": 1800
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.49,
|
| 120 |
+
"learning_rate": 4.1842692770049805e-05,
|
| 121 |
+
"loss": 1.2735,
|
| 122 |
+
"step": 1900
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.52,
|
| 126 |
+
"learning_rate": 4.1413360810578746e-05,
|
| 127 |
+
"loss": 1.293,
|
| 128 |
+
"step": 2000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.54,
|
| 132 |
+
"learning_rate": 4.098402885110768e-05,
|
| 133 |
+
"loss": 1.2661,
|
| 134 |
+
"step": 2100
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.57,
|
| 138 |
+
"learning_rate": 4.055469689163662e-05,
|
| 139 |
+
"loss": 1.2543,
|
| 140 |
+
"step": 2200
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.59,
|
| 144 |
+
"learning_rate": 4.0125364932165554e-05,
|
| 145 |
+
"loss": 1.2635,
|
| 146 |
+
"step": 2300
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.62,
|
| 150 |
+
"learning_rate": 3.969603297269449e-05,
|
| 151 |
+
"loss": 1.2306,
|
| 152 |
+
"step": 2400
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.64,
|
| 156 |
+
"learning_rate": 3.926670101322343e-05,
|
| 157 |
+
"loss": 1.2114,
|
| 158 |
+
"step": 2500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.67,
|
| 162 |
+
"learning_rate": 3.883736905375236e-05,
|
| 163 |
+
"loss": 1.2305,
|
| 164 |
+
"step": 2600
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.7,
|
| 168 |
+
"learning_rate": 3.84080370942813e-05,
|
| 169 |
+
"loss": 1.1961,
|
| 170 |
+
"step": 2700
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.72,
|
| 174 |
+
"learning_rate": 3.797870513481024e-05,
|
| 175 |
+
"loss": 1.1866,
|
| 176 |
+
"step": 2800
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.75,
|
| 180 |
+
"learning_rate": 3.754937317533918e-05,
|
| 181 |
+
"loss": 1.1808,
|
| 182 |
+
"step": 2900
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.77,
|
| 186 |
+
"learning_rate": 3.712004121586811e-05,
|
| 187 |
+
"loss": 1.1883,
|
| 188 |
+
"step": 3000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.8,
|
| 192 |
+
"learning_rate": 3.669070925639705e-05,
|
| 193 |
+
"loss": 1.1482,
|
| 194 |
+
"step": 3100
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.82,
|
| 198 |
+
"learning_rate": 3.6261377296925985e-05,
|
| 199 |
+
"loss": 1.1377,
|
| 200 |
+
"step": 3200
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.85,
|
| 204 |
+
"learning_rate": 3.5832045337454926e-05,
|
| 205 |
+
"loss": 1.1463,
|
| 206 |
+
"step": 3300
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.88,
|
| 210 |
+
"learning_rate": 3.540271337798386e-05,
|
| 211 |
+
"loss": 1.1347,
|
| 212 |
+
"step": 3400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.9,
|
| 216 |
+
"learning_rate": 3.4973381418512793e-05,
|
| 217 |
+
"loss": 1.1217,
|
| 218 |
+
"step": 3500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.93,
|
| 222 |
+
"learning_rate": 3.4544049459041734e-05,
|
| 223 |
+
"loss": 1.1249,
|
| 224 |
+
"step": 3600
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.95,
|
| 228 |
+
"learning_rate": 3.411471749957067e-05,
|
| 229 |
+
"loss": 1.0972,
|
| 230 |
+
"step": 3700
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.98,
|
| 234 |
+
"learning_rate": 3.368538554009961e-05,
|
| 235 |
+
"loss": 1.1206,
|
| 236 |
+
"step": 3800
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.0,
|
| 240 |
+
"learning_rate": 3.325605358062854e-05,
|
| 241 |
+
"loss": 1.0817,
|
| 242 |
+
"step": 3900
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.03,
|
| 246 |
+
"learning_rate": 3.282672162115748e-05,
|
| 247 |
+
"loss": 1.0486,
|
| 248 |
+
"step": 4000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.06,
|
| 252 |
+
"learning_rate": 3.2397389661686417e-05,
|
| 253 |
+
"loss": 1.0737,
|
| 254 |
+
"step": 4100
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.08,
|
| 258 |
+
"learning_rate": 3.196805770221536e-05,
|
| 259 |
+
"loss": 1.0448,
|
| 260 |
+
"step": 4200
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.11,
|
| 264 |
+
"learning_rate": 3.153872574274429e-05,
|
| 265 |
+
"loss": 1.0633,
|
| 266 |
+
"step": 4300
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.13,
|
| 270 |
+
"learning_rate": 3.1109393783273225e-05,
|
| 271 |
+
"loss": 1.0297,
|
| 272 |
+
"step": 4400
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.16,
|
| 276 |
+
"learning_rate": 3.0680061823802165e-05,
|
| 277 |
+
"loss": 1.0405,
|
| 278 |
+
"step": 4500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 1.18,
|
| 282 |
+
"learning_rate": 3.0250729864331102e-05,
|
| 283 |
+
"loss": 1.024,
|
| 284 |
+
"step": 4600
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.21,
|
| 288 |
+
"learning_rate": 2.9821397904860036e-05,
|
| 289 |
+
"loss": 0.9993,
|
| 290 |
+
"step": 4700
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.24,
|
| 294 |
+
"learning_rate": 2.9392065945388973e-05,
|
| 295 |
+
"loss": 1.0126,
|
| 296 |
+
"step": 4800
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.26,
|
| 300 |
+
"learning_rate": 2.896273398591791e-05,
|
| 301 |
+
"loss": 1.0055,
|
| 302 |
+
"step": 4900
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 1.29,
|
| 306 |
+
"learning_rate": 2.8533402026446848e-05,
|
| 307 |
+
"loss": 1.0017,
|
| 308 |
+
"step": 5000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.31,
|
| 312 |
+
"learning_rate": 2.8104070066975785e-05,
|
| 313 |
+
"loss": 1.0052,
|
| 314 |
+
"step": 5100
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 1.34,
|
| 318 |
+
"learning_rate": 2.7674738107504722e-05,
|
| 319 |
+
"loss": 0.995,
|
| 320 |
+
"step": 5200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.37,
|
| 324 |
+
"learning_rate": 2.724540614803366e-05,
|
| 325 |
+
"loss": 1.0071,
|
| 326 |
+
"step": 5300
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.39,
|
| 330 |
+
"learning_rate": 2.6816074188562596e-05,
|
| 331 |
+
"loss": 0.9835,
|
| 332 |
+
"step": 5400
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.42,
|
| 336 |
+
"learning_rate": 2.6386742229091534e-05,
|
| 337 |
+
"loss": 0.9986,
|
| 338 |
+
"step": 5500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.44,
|
| 342 |
+
"learning_rate": 2.5957410269620467e-05,
|
| 343 |
+
"loss": 0.9837,
|
| 344 |
+
"step": 5600
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 1.47,
|
| 348 |
+
"learning_rate": 2.552807831014941e-05,
|
| 349 |
+
"loss": 0.9897,
|
| 350 |
+
"step": 5700
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.49,
|
| 354 |
+
"learning_rate": 2.509874635067835e-05,
|
| 355 |
+
"loss": 0.996,
|
| 356 |
+
"step": 5800
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 1.52,
|
| 360 |
+
"learning_rate": 2.4669414391207282e-05,
|
| 361 |
+
"loss": 0.9831,
|
| 362 |
+
"step": 5900
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.55,
|
| 366 |
+
"learning_rate": 2.424008243173622e-05,
|
| 367 |
+
"loss": 0.9518,
|
| 368 |
+
"step": 6000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.57,
|
| 372 |
+
"learning_rate": 2.3810750472265157e-05,
|
| 373 |
+
"loss": 0.9758,
|
| 374 |
+
"step": 6100
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.6,
|
| 378 |
+
"learning_rate": 2.3381418512794094e-05,
|
| 379 |
+
"loss": 0.9629,
|
| 380 |
+
"step": 6200
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.62,
|
| 384 |
+
"learning_rate": 2.295208655332303e-05,
|
| 385 |
+
"loss": 0.9607,
|
| 386 |
+
"step": 6300
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 1.65,
|
| 390 |
+
"learning_rate": 2.2522754593851965e-05,
|
| 391 |
+
"loss": 0.9619,
|
| 392 |
+
"step": 6400
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.67,
|
| 396 |
+
"learning_rate": 2.2093422634380902e-05,
|
| 397 |
+
"loss": 0.9528,
|
| 398 |
+
"step": 6500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.7,
|
| 402 |
+
"learning_rate": 2.1664090674909843e-05,
|
| 403 |
+
"loss": 0.9504,
|
| 404 |
+
"step": 6600
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.73,
|
| 408 |
+
"learning_rate": 2.123475871543878e-05,
|
| 409 |
+
"loss": 0.9569,
|
| 410 |
+
"step": 6700
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.75,
|
| 414 |
+
"learning_rate": 2.0805426755967717e-05,
|
| 415 |
+
"loss": 0.9721,
|
| 416 |
+
"step": 6800
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.78,
|
| 420 |
+
"learning_rate": 2.0376094796496654e-05,
|
| 421 |
+
"loss": 0.9521,
|
| 422 |
+
"step": 6900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 1.8,
|
| 426 |
+
"learning_rate": 1.994676283702559e-05,
|
| 427 |
+
"loss": 0.9583,
|
| 428 |
+
"step": 7000
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 1.83,
|
| 432 |
+
"learning_rate": 1.9517430877554525e-05,
|
| 433 |
+
"loss": 0.9444,
|
| 434 |
+
"step": 7100
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 1.85,
|
| 438 |
+
"learning_rate": 1.9088098918083462e-05,
|
| 439 |
+
"loss": 0.9116,
|
| 440 |
+
"step": 7200
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 1.88,
|
| 444 |
+
"learning_rate": 1.86587669586124e-05,
|
| 445 |
+
"loss": 0.9321,
|
| 446 |
+
"step": 7300
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 1.91,
|
| 450 |
+
"learning_rate": 1.8229434999141337e-05,
|
| 451 |
+
"loss": 0.9296,
|
| 452 |
+
"step": 7400
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.93,
|
| 456 |
+
"learning_rate": 1.7800103039670274e-05,
|
| 457 |
+
"loss": 0.9124,
|
| 458 |
+
"step": 7500
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.96,
|
| 462 |
+
"learning_rate": 1.737077108019921e-05,
|
| 463 |
+
"loss": 0.9234,
|
| 464 |
+
"step": 7600
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 1.98,
|
| 468 |
+
"learning_rate": 1.6941439120728148e-05,
|
| 469 |
+
"loss": 0.9581,
|
| 470 |
+
"step": 7700
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 2.01,
|
| 474 |
+
"learning_rate": 1.6512107161257085e-05,
|
| 475 |
+
"loss": 0.9333,
|
| 476 |
+
"step": 7800
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 2.04,
|
| 480 |
+
"learning_rate": 1.6082775201786023e-05,
|
| 481 |
+
"loss": 0.8964,
|
| 482 |
+
"step": 7900
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.06,
|
| 486 |
+
"learning_rate": 1.5653443242314956e-05,
|
| 487 |
+
"loss": 0.9115,
|
| 488 |
+
"step": 8000
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 2.09,
|
| 492 |
+
"learning_rate": 1.5224111282843895e-05,
|
| 493 |
+
"loss": 0.9046,
|
| 494 |
+
"step": 8100
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.11,
|
| 498 |
+
"learning_rate": 1.4794779323372832e-05,
|
| 499 |
+
"loss": 0.8765,
|
| 500 |
+
"step": 8200
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 2.14,
|
| 504 |
+
"learning_rate": 1.4365447363901768e-05,
|
| 505 |
+
"loss": 0.8779,
|
| 506 |
+
"step": 8300
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 2.16,
|
| 510 |
+
"learning_rate": 1.3936115404430705e-05,
|
| 511 |
+
"loss": 0.8801,
|
| 512 |
+
"step": 8400
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 2.19,
|
| 516 |
+
"learning_rate": 1.3506783444959644e-05,
|
| 517 |
+
"loss": 0.8843,
|
| 518 |
+
"step": 8500
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 2.22,
|
| 522 |
+
"learning_rate": 1.3077451485488581e-05,
|
| 523 |
+
"loss": 0.8848,
|
| 524 |
+
"step": 8600
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 2.24,
|
| 528 |
+
"learning_rate": 1.2648119526017518e-05,
|
| 529 |
+
"loss": 0.8957,
|
| 530 |
+
"step": 8700
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 2.27,
|
| 534 |
+
"learning_rate": 1.2218787566546454e-05,
|
| 535 |
+
"loss": 0.8997,
|
| 536 |
+
"step": 8800
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 2.29,
|
| 540 |
+
"learning_rate": 1.1789455607075391e-05,
|
| 541 |
+
"loss": 0.8977,
|
| 542 |
+
"step": 8900
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 2.32,
|
| 546 |
+
"learning_rate": 1.1360123647604328e-05,
|
| 547 |
+
"loss": 0.8699,
|
| 548 |
+
"step": 9000
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 2.34,
|
| 552 |
+
"learning_rate": 1.0930791688133265e-05,
|
| 553 |
+
"loss": 0.8895,
|
| 554 |
+
"step": 9100
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 2.37,
|
| 558 |
+
"learning_rate": 1.0501459728662203e-05,
|
| 559 |
+
"loss": 0.8848,
|
| 560 |
+
"step": 9200
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 2.4,
|
| 564 |
+
"learning_rate": 1.007212776919114e-05,
|
| 565 |
+
"loss": 0.8647,
|
| 566 |
+
"step": 9300
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 2.42,
|
| 570 |
+
"learning_rate": 9.642795809720075e-06,
|
| 571 |
+
"loss": 0.8836,
|
| 572 |
+
"step": 9400
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 2.45,
|
| 576 |
+
"learning_rate": 9.213463850249012e-06,
|
| 577 |
+
"loss": 0.8824,
|
| 578 |
+
"step": 9500
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 2.47,
|
| 582 |
+
"learning_rate": 8.78413189077795e-06,
|
| 583 |
+
"loss": 0.8672,
|
| 584 |
+
"step": 9600
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 2.5,
|
| 588 |
+
"learning_rate": 8.354799931306887e-06,
|
| 589 |
+
"loss": 0.881,
|
| 590 |
+
"step": 9700
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 2.52,
|
| 594 |
+
"learning_rate": 7.925467971835824e-06,
|
| 595 |
+
"loss": 0.8505,
|
| 596 |
+
"step": 9800
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 2.55,
|
| 600 |
+
"learning_rate": 7.496136012364761e-06,
|
| 601 |
+
"loss": 0.8925,
|
| 602 |
+
"step": 9900
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 2.58,
|
| 606 |
+
"learning_rate": 7.066804052893698e-06,
|
| 607 |
+
"loss": 0.8876,
|
| 608 |
+
"step": 10000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 2.6,
|
| 612 |
+
"learning_rate": 6.637472093422635e-06,
|
| 613 |
+
"loss": 0.8716,
|
| 614 |
+
"step": 10100
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 2.63,
|
| 618 |
+
"learning_rate": 6.208140133951572e-06,
|
| 619 |
+
"loss": 0.875,
|
| 620 |
+
"step": 10200
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 2.65,
|
| 624 |
+
"learning_rate": 5.778808174480509e-06,
|
| 625 |
+
"loss": 0.8848,
|
| 626 |
+
"step": 10300
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 2.68,
|
| 630 |
+
"learning_rate": 5.349476215009445e-06,
|
| 631 |
+
"loss": 0.8629,
|
| 632 |
+
"step": 10400
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 2.7,
|
| 636 |
+
"learning_rate": 4.9201442555383824e-06,
|
| 637 |
+
"loss": 0.8797,
|
| 638 |
+
"step": 10500
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 2.73,
|
| 642 |
+
"learning_rate": 4.49081229606732e-06,
|
| 643 |
+
"loss": 0.8478,
|
| 644 |
+
"step": 10600
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 2.76,
|
| 648 |
+
"learning_rate": 4.061480336596257e-06,
|
| 649 |
+
"loss": 0.8526,
|
| 650 |
+
"step": 10700
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 2.78,
|
| 654 |
+
"learning_rate": 3.6321483771251936e-06,
|
| 655 |
+
"loss": 0.8619,
|
| 656 |
+
"step": 10800
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 2.81,
|
| 660 |
+
"learning_rate": 3.2028164176541303e-06,
|
| 661 |
+
"loss": 0.8923,
|
| 662 |
+
"step": 10900
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 2.83,
|
| 666 |
+
"learning_rate": 2.7734844581830675e-06,
|
| 667 |
+
"loss": 0.8737,
|
| 668 |
+
"step": 11000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 2.86,
|
| 672 |
+
"learning_rate": 2.3441524987120042e-06,
|
| 673 |
+
"loss": 0.8556,
|
| 674 |
+
"step": 11100
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 2.89,
|
| 678 |
+
"learning_rate": 1.9148205392409414e-06,
|
| 679 |
+
"loss": 0.8767,
|
| 680 |
+
"step": 11200
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 2.91,
|
| 684 |
+
"learning_rate": 1.4854885797698782e-06,
|
| 685 |
+
"loss": 0.886,
|
| 686 |
+
"step": 11300
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 2.94,
|
| 690 |
+
"learning_rate": 1.0561566202988151e-06,
|
| 691 |
+
"loss": 0.8638,
|
| 692 |
+
"step": 11400
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 2.96,
|
| 696 |
+
"learning_rate": 6.26824660827752e-07,
|
| 697 |
+
"loss": 0.8885,
|
| 698 |
+
"step": 11500
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 2.99,
|
| 702 |
+
"learning_rate": 1.97492701356689e-07,
|
| 703 |
+
"loss": 0.881,
|
| 704 |
+
"step": 11600
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 3.0,
|
| 708 |
+
"step": 11646,
|
| 709 |
+
"total_flos": 6.280174860167578e+16,
|
| 710 |
+
"train_loss": 1.0872868923270456,
|
| 711 |
+
"train_runtime": 30832.4334,
|
| 712 |
+
"train_samples_per_second": 6.043,
|
| 713 |
+
"train_steps_per_second": 0.378
|
| 714 |
+
}
|
| 715 |
+
],
|
| 716 |
+
"max_steps": 11646,
|
| 717 |
+
"num_train_epochs": 3,
|
| 718 |
+
"total_flos": 6.280174860167578e+16,
|
| 719 |
+
"trial_name": null,
|
| 720 |
+
"trial_params": null
|
| 721 |
+
}
|