Training in progress, step 500
Browse files- all_results.json +4 -4
- pytorch_model.bin +1 -1
- train_results.json +4 -4
- trainer_state.json +586 -88
- training_args.bin +1 -1
all_results.json
CHANGED
|
@@ -4,8 +4,8 @@
|
|
| 4 |
"eval_runtime": 205.5006,
|
| 5 |
"eval_samples_per_second": 121.722,
|
| 6 |
"eval_steps_per_second": 0.954,
|
| 7 |
-
"train_loss": 0.
|
| 8 |
-
"train_runtime":
|
| 9 |
-
"train_samples_per_second":
|
| 10 |
-
"train_steps_per_second": 0.
|
| 11 |
}
|
|
|
|
| 4 |
"eval_runtime": 205.5006,
|
| 5 |
"eval_samples_per_second": 121.722,
|
| 6 |
"eval_steps_per_second": 0.954,
|
| 7 |
+
"train_loss": 0.3350037185057538,
|
| 8 |
+
"train_runtime": 233521.0358,
|
| 9 |
+
"train_samples_per_second": 7.602,
|
| 10 |
+
"train_steps_per_second": 0.238
|
| 11 |
}
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 851693457
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:26415fe1a4799ab4e959825baec1082c82cfdbc9532b97a76dcb16ce7416950f
|
| 3 |
size 851693457
|
train_results.json
CHANGED
|
@@ -1,7 +1,7 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 3.0,
|
| 3 |
-
"train_loss": 0.
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples_per_second":
|
| 6 |
-
"train_steps_per_second": 0.
|
| 7 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 3.0,
|
| 3 |
+
"train_loss": 0.3350037185057538,
|
| 4 |
+
"train_runtime": 233521.0358,
|
| 5 |
+
"train_samples_per_second": 7.602,
|
| 6 |
+
"train_steps_per_second": 0.238
|
| 7 |
}
|
trainer_state.json
CHANGED
|
@@ -3,185 +3,683 @@
|
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
"epoch": 3.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"learning_rate": 4.
|
| 14 |
-
"loss": 1.
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
{
|
| 18 |
-
"epoch": 0.
|
| 19 |
-
"learning_rate": 4.
|
| 20 |
-
"loss":
|
| 21 |
"step": 1000
|
| 22 |
},
|
| 23 |
{
|
| 24 |
-
"epoch": 0.
|
| 25 |
-
"learning_rate": 4.
|
| 26 |
-
"loss": 0.
|
| 27 |
"step": 1500
|
| 28 |
},
|
| 29 |
{
|
| 30 |
-
"epoch": 0.
|
| 31 |
-
"learning_rate": 4.
|
| 32 |
-
"loss": 0.
|
| 33 |
"step": 2000
|
| 34 |
},
|
| 35 |
{
|
| 36 |
-
"epoch": 0.
|
| 37 |
-
"learning_rate": 4.
|
| 38 |
-
"loss": 0.
|
| 39 |
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
-
"epoch": 0.
|
| 43 |
-
"learning_rate":
|
| 44 |
-
"loss": 0.
|
| 45 |
"step": 3000
|
| 46 |
},
|
| 47 |
{
|
| 48 |
-
"epoch": 0.
|
| 49 |
-
"learning_rate":
|
| 50 |
-
"loss": 0.
|
| 51 |
"step": 3500
|
| 52 |
},
|
| 53 |
{
|
| 54 |
-
"epoch": 0.
|
| 55 |
-
"learning_rate":
|
| 56 |
-
"loss": 0.
|
| 57 |
"step": 4000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
-
"epoch": 0.
|
| 61 |
-
"learning_rate":
|
| 62 |
-
"loss": 0.
|
| 63 |
"step": 4500
|
| 64 |
},
|
| 65 |
{
|
| 66 |
-
"epoch":
|
| 67 |
-
"learning_rate":
|
| 68 |
-
"loss": 0.
|
| 69 |
"step": 5000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
-
"epoch":
|
| 73 |
-
"learning_rate":
|
| 74 |
-
"loss": 0.
|
| 75 |
"step": 5500
|
| 76 |
},
|
| 77 |
{
|
| 78 |
-
"epoch":
|
| 79 |
-
"learning_rate":
|
| 80 |
-
"loss": 0.
|
| 81 |
"step": 6000
|
| 82 |
},
|
| 83 |
{
|
| 84 |
-
"epoch":
|
| 85 |
-
"learning_rate":
|
| 86 |
-
"loss": 0.
|
| 87 |
"step": 6500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
-
"epoch":
|
| 91 |
-
"learning_rate":
|
| 92 |
-
"loss": 0.
|
| 93 |
"step": 7000
|
| 94 |
},
|
| 95 |
{
|
| 96 |
-
"epoch":
|
| 97 |
-
"learning_rate":
|
| 98 |
-
"loss": 0.
|
| 99 |
"step": 7500
|
| 100 |
},
|
| 101 |
{
|
| 102 |
-
"epoch":
|
| 103 |
-
"learning_rate":
|
| 104 |
-
"loss": 0.
|
| 105 |
"step": 8000
|
| 106 |
},
|
| 107 |
{
|
| 108 |
-
"epoch":
|
| 109 |
-
"learning_rate":
|
| 110 |
-
"loss": 0.
|
| 111 |
"step": 8500
|
| 112 |
},
|
| 113 |
{
|
| 114 |
-
"epoch":
|
| 115 |
-
"learning_rate":
|
| 116 |
-
"loss": 0.
|
| 117 |
"step": 9000
|
| 118 |
},
|
| 119 |
{
|
| 120 |
-
"epoch":
|
| 121 |
-
"learning_rate":
|
| 122 |
-
"loss": 0.
|
| 123 |
"step": 9500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
-
"epoch":
|
| 127 |
-
"learning_rate":
|
| 128 |
-
"loss": 0.
|
| 129 |
"step": 10000
|
| 130 |
},
|
| 131 |
{
|
| 132 |
-
"epoch":
|
| 133 |
-
"learning_rate":
|
| 134 |
-
"loss": 0.
|
| 135 |
"step": 10500
|
| 136 |
},
|
| 137 |
{
|
| 138 |
-
"epoch":
|
| 139 |
-
"learning_rate":
|
| 140 |
-
"loss": 0.
|
| 141 |
"step": 11000
|
| 142 |
},
|
| 143 |
{
|
| 144 |
-
"epoch":
|
| 145 |
-
"learning_rate":
|
| 146 |
-
"loss": 0.
|
| 147 |
"step": 11500
|
| 148 |
},
|
| 149 |
{
|
| 150 |
-
"epoch":
|
| 151 |
-
"learning_rate":
|
| 152 |
-
"loss": 0.
|
| 153 |
"step": 12000
|
| 154 |
},
|
| 155 |
{
|
| 156 |
-
"epoch":
|
| 157 |
-
"learning_rate":
|
| 158 |
-
"loss": 0.
|
| 159 |
"step": 12500
|
| 160 |
},
|
| 161 |
{
|
| 162 |
-
"epoch":
|
| 163 |
-
"learning_rate": 3.
|
| 164 |
-
"loss": 0.
|
| 165 |
"step": 13000
|
| 166 |
},
|
| 167 |
{
|
| 168 |
-
"epoch":
|
| 169 |
-
"learning_rate":
|
| 170 |
-
"loss": 0.
|
| 171 |
"step": 13500
|
| 172 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 173 |
{
|
| 174 |
"epoch": 3.0,
|
| 175 |
-
"step":
|
| 176 |
"total_flos": 2.3702775297552e+17,
|
| 177 |
-
"train_loss": 0.
|
| 178 |
-
"train_runtime":
|
| 179 |
-
"train_samples_per_second":
|
| 180 |
-
"train_steps_per_second": 0.
|
| 181 |
}
|
| 182 |
],
|
| 183 |
"logging_steps": 500,
|
| 184 |
-
"max_steps":
|
| 185 |
"num_train_epochs": 3,
|
| 186 |
"save_steps": 500,
|
| 187 |
"total_flos": 2.3702775297552e+17,
|
|
|
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
"epoch": 3.0,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 55479,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.03,
|
| 13 |
+
"learning_rate": 4.954937904432308e-05,
|
| 14 |
+
"loss": 1.8047,
|
| 15 |
"step": 500
|
| 16 |
},
|
| 17 |
{
|
| 18 |
+
"epoch": 0.05,
|
| 19 |
+
"learning_rate": 4.9098758088646155e-05,
|
| 20 |
+
"loss": 1.0161,
|
| 21 |
"step": 1000
|
| 22 |
},
|
| 23 |
{
|
| 24 |
+
"epoch": 0.08,
|
| 25 |
+
"learning_rate": 4.864813713296923e-05,
|
| 26 |
+
"loss": 0.9184,
|
| 27 |
"step": 1500
|
| 28 |
},
|
| 29 |
{
|
| 30 |
+
"epoch": 0.11,
|
| 31 |
+
"learning_rate": 4.8197516177292315e-05,
|
| 32 |
+
"loss": 0.8381,
|
| 33 |
"step": 2000
|
| 34 |
},
|
| 35 |
{
|
| 36 |
+
"epoch": 0.14,
|
| 37 |
+
"learning_rate": 4.774689522161539e-05,
|
| 38 |
+
"loss": 0.8042,
|
| 39 |
"step": 2500
|
| 40 |
},
|
| 41 |
{
|
| 42 |
+
"epoch": 0.16,
|
| 43 |
+
"learning_rate": 4.729627426593847e-05,
|
| 44 |
+
"loss": 0.7784,
|
| 45 |
"step": 3000
|
| 46 |
},
|
| 47 |
{
|
| 48 |
+
"epoch": 0.19,
|
| 49 |
+
"learning_rate": 4.6845653310261544e-05,
|
| 50 |
+
"loss": 0.7531,
|
| 51 |
"step": 3500
|
| 52 |
},
|
| 53 |
{
|
| 54 |
+
"epoch": 0.22,
|
| 55 |
+
"learning_rate": 4.639503235458462e-05,
|
| 56 |
+
"loss": 0.7263,
|
| 57 |
"step": 4000
|
| 58 |
},
|
| 59 |
{
|
| 60 |
+
"epoch": 0.24,
|
| 61 |
+
"learning_rate": 4.59444113989077e-05,
|
| 62 |
+
"loss": 0.7171,
|
| 63 |
"step": 4500
|
| 64 |
},
|
| 65 |
{
|
| 66 |
+
"epoch": 0.27,
|
| 67 |
+
"learning_rate": 4.549379044323077e-05,
|
| 68 |
+
"loss": 0.6689,
|
| 69 |
"step": 5000
|
| 70 |
},
|
| 71 |
{
|
| 72 |
+
"epoch": 0.3,
|
| 73 |
+
"learning_rate": 4.504316948755385e-05,
|
| 74 |
+
"loss": 0.6679,
|
| 75 |
"step": 5500
|
| 76 |
},
|
| 77 |
{
|
| 78 |
+
"epoch": 0.32,
|
| 79 |
+
"learning_rate": 4.4592548531876926e-05,
|
| 80 |
+
"loss": 0.6308,
|
| 81 |
"step": 6000
|
| 82 |
},
|
| 83 |
{
|
| 84 |
+
"epoch": 0.35,
|
| 85 |
+
"learning_rate": 4.41419275762e-05,
|
| 86 |
+
"loss": 0.6113,
|
| 87 |
"step": 6500
|
| 88 |
},
|
| 89 |
{
|
| 90 |
+
"epoch": 0.38,
|
| 91 |
+
"learning_rate": 4.369130662052308e-05,
|
| 92 |
+
"loss": 0.6029,
|
| 93 |
"step": 7000
|
| 94 |
},
|
| 95 |
{
|
| 96 |
+
"epoch": 0.41,
|
| 97 |
+
"learning_rate": 4.324068566484616e-05,
|
| 98 |
+
"loss": 0.6022,
|
| 99 |
"step": 7500
|
| 100 |
},
|
| 101 |
{
|
| 102 |
+
"epoch": 0.43,
|
| 103 |
+
"learning_rate": 4.279006470916924e-05,
|
| 104 |
+
"loss": 0.5795,
|
| 105 |
"step": 8000
|
| 106 |
},
|
| 107 |
{
|
| 108 |
+
"epoch": 0.46,
|
| 109 |
+
"learning_rate": 4.2339443753492315e-05,
|
| 110 |
+
"loss": 0.5509,
|
| 111 |
"step": 8500
|
| 112 |
},
|
| 113 |
{
|
| 114 |
+
"epoch": 0.49,
|
| 115 |
+
"learning_rate": 4.188882279781539e-05,
|
| 116 |
+
"loss": 0.5502,
|
| 117 |
"step": 9000
|
| 118 |
},
|
| 119 |
{
|
| 120 |
+
"epoch": 0.51,
|
| 121 |
+
"learning_rate": 4.143820184213847e-05,
|
| 122 |
+
"loss": 0.5475,
|
| 123 |
"step": 9500
|
| 124 |
},
|
| 125 |
{
|
| 126 |
+
"epoch": 0.54,
|
| 127 |
+
"learning_rate": 4.0987580886461544e-05,
|
| 128 |
+
"loss": 0.5176,
|
| 129 |
"step": 10000
|
| 130 |
},
|
| 131 |
{
|
| 132 |
+
"epoch": 0.57,
|
| 133 |
+
"learning_rate": 4.053695993078463e-05,
|
| 134 |
+
"loss": 0.524,
|
| 135 |
"step": 10500
|
| 136 |
},
|
| 137 |
{
|
| 138 |
+
"epoch": 0.59,
|
| 139 |
+
"learning_rate": 4.0086338975107703e-05,
|
| 140 |
+
"loss": 0.4914,
|
| 141 |
"step": 11000
|
| 142 |
},
|
| 143 |
{
|
| 144 |
+
"epoch": 0.62,
|
| 145 |
+
"learning_rate": 3.963571801943078e-05,
|
| 146 |
+
"loss": 0.4989,
|
| 147 |
"step": 11500
|
| 148 |
},
|
| 149 |
{
|
| 150 |
+
"epoch": 0.65,
|
| 151 |
+
"learning_rate": 3.9185097063753856e-05,
|
| 152 |
+
"loss": 0.4852,
|
| 153 |
"step": 12000
|
| 154 |
},
|
| 155 |
{
|
| 156 |
+
"epoch": 0.68,
|
| 157 |
+
"learning_rate": 3.873447610807693e-05,
|
| 158 |
+
"loss": 0.4751,
|
| 159 |
"step": 12500
|
| 160 |
},
|
| 161 |
{
|
| 162 |
+
"epoch": 0.7,
|
| 163 |
+
"learning_rate": 3.828385515240001e-05,
|
| 164 |
+
"loss": 0.4766,
|
| 165 |
"step": 13000
|
| 166 |
},
|
| 167 |
{
|
| 168 |
+
"epoch": 0.73,
|
| 169 |
+
"learning_rate": 3.7833234196723085e-05,
|
| 170 |
+
"loss": 0.4558,
|
| 171 |
"step": 13500
|
| 172 |
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.76,
|
| 175 |
+
"learning_rate": 3.738261324104617e-05,
|
| 176 |
+
"loss": 0.46,
|
| 177 |
+
"step": 14000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.78,
|
| 181 |
+
"learning_rate": 3.693199228536924e-05,
|
| 182 |
+
"loss": 0.4379,
|
| 183 |
+
"step": 14500
|
| 184 |
+
},
|
| 185 |
+
{
|
| 186 |
+
"epoch": 0.81,
|
| 187 |
+
"learning_rate": 3.6481371329692315e-05,
|
| 188 |
+
"loss": 0.4219,
|
| 189 |
+
"step": 15000
|
| 190 |
+
},
|
| 191 |
+
{
|
| 192 |
+
"epoch": 0.84,
|
| 193 |
+
"learning_rate": 3.603075037401539e-05,
|
| 194 |
+
"loss": 0.4261,
|
| 195 |
+
"step": 15500
|
| 196 |
+
},
|
| 197 |
+
{
|
| 198 |
+
"epoch": 0.87,
|
| 199 |
+
"learning_rate": 3.558012941833847e-05,
|
| 200 |
+
"loss": 0.4075,
|
| 201 |
+
"step": 16000
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.89,
|
| 205 |
+
"learning_rate": 3.512950846266155e-05,
|
| 206 |
+
"loss": 0.4027,
|
| 207 |
+
"step": 16500
|
| 208 |
+
},
|
| 209 |
+
{
|
| 210 |
+
"epoch": 0.92,
|
| 211 |
+
"learning_rate": 3.467888750698463e-05,
|
| 212 |
+
"loss": 0.3952,
|
| 213 |
+
"step": 17000
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.95,
|
| 217 |
+
"learning_rate": 3.4228266551307703e-05,
|
| 218 |
+
"loss": 0.3859,
|
| 219 |
+
"step": 17500
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.97,
|
| 223 |
+
"learning_rate": 3.377764559563078e-05,
|
| 224 |
+
"loss": 0.3845,
|
| 225 |
+
"step": 18000
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 1.0,
|
| 229 |
+
"learning_rate": 3.3327024639953856e-05,
|
| 230 |
+
"loss": 0.3848,
|
| 231 |
+
"step": 18500
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 1.03,
|
| 235 |
+
"learning_rate": 3.287640368427693e-05,
|
| 236 |
+
"loss": 0.3174,
|
| 237 |
+
"step": 19000
|
| 238 |
+
},
|
| 239 |
+
{
|
| 240 |
+
"epoch": 1.05,
|
| 241 |
+
"learning_rate": 3.2425782728600016e-05,
|
| 242 |
+
"loss": 0.3251,
|
| 243 |
+
"step": 19500
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"epoch": 1.08,
|
| 247 |
+
"learning_rate": 3.197516177292309e-05,
|
| 248 |
+
"loss": 0.3244,
|
| 249 |
+
"step": 20000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 1.11,
|
| 253 |
+
"learning_rate": 3.152454081724617e-05,
|
| 254 |
+
"loss": 0.3186,
|
| 255 |
+
"step": 20500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 1.14,
|
| 259 |
+
"learning_rate": 3.1073919861569245e-05,
|
| 260 |
+
"loss": 0.3151,
|
| 261 |
+
"step": 21000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 1.16,
|
| 265 |
+
"learning_rate": 3.062329890589232e-05,
|
| 266 |
+
"loss": 0.2993,
|
| 267 |
+
"step": 21500
|
| 268 |
+
},
|
| 269 |
+
{
|
| 270 |
+
"epoch": 1.19,
|
| 271 |
+
"learning_rate": 3.01726779502154e-05,
|
| 272 |
+
"loss": 0.2997,
|
| 273 |
+
"step": 22000
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 1.22,
|
| 277 |
+
"learning_rate": 2.9722056994538478e-05,
|
| 278 |
+
"loss": 0.3078,
|
| 279 |
+
"step": 22500
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 1.24,
|
| 283 |
+
"learning_rate": 2.9271436038861554e-05,
|
| 284 |
+
"loss": 0.2988,
|
| 285 |
+
"step": 23000
|
| 286 |
+
},
|
| 287 |
+
{
|
| 288 |
+
"epoch": 1.27,
|
| 289 |
+
"learning_rate": 2.8820815083184634e-05,
|
| 290 |
+
"loss": 0.2912,
|
| 291 |
+
"step": 23500
|
| 292 |
+
},
|
| 293 |
+
{
|
| 294 |
+
"epoch": 1.3,
|
| 295 |
+
"learning_rate": 2.8370194127507703e-05,
|
| 296 |
+
"loss": 0.288,
|
| 297 |
+
"step": 24000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 1.32,
|
| 301 |
+
"learning_rate": 2.7919573171830783e-05,
|
| 302 |
+
"loss": 0.277,
|
| 303 |
+
"step": 24500
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 1.35,
|
| 307 |
+
"learning_rate": 2.746895221615386e-05,
|
| 308 |
+
"loss": 0.2886,
|
| 309 |
+
"step": 25000
|
| 310 |
+
},
|
| 311 |
+
{
|
| 312 |
+
"epoch": 1.38,
|
| 313 |
+
"learning_rate": 2.7018331260476936e-05,
|
| 314 |
+
"loss": 0.2659,
|
| 315 |
+
"step": 25500
|
| 316 |
+
},
|
| 317 |
+
{
|
| 318 |
+
"epoch": 1.41,
|
| 319 |
+
"learning_rate": 2.6567710304800016e-05,
|
| 320 |
+
"loss": 0.2713,
|
| 321 |
+
"step": 26000
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 1.43,
|
| 325 |
+
"learning_rate": 2.6117089349123092e-05,
|
| 326 |
+
"loss": 0.2755,
|
| 327 |
+
"step": 26500
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 1.46,
|
| 331 |
+
"learning_rate": 2.566646839344617e-05,
|
| 332 |
+
"loss": 0.2723,
|
| 333 |
+
"step": 27000
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.49,
|
| 337 |
+
"learning_rate": 2.5215847437769248e-05,
|
| 338 |
+
"loss": 0.261,
|
| 339 |
+
"step": 27500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 1.51,
|
| 343 |
+
"learning_rate": 2.4765226482092325e-05,
|
| 344 |
+
"loss": 0.2628,
|
| 345 |
+
"step": 28000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 1.54,
|
| 349 |
+
"learning_rate": 2.43146055264154e-05,
|
| 350 |
+
"loss": 0.2538,
|
| 351 |
+
"step": 28500
|
| 352 |
+
},
|
| 353 |
+
{
|
| 354 |
+
"epoch": 1.57,
|
| 355 |
+
"learning_rate": 2.3863984570738477e-05,
|
| 356 |
+
"loss": 0.2526,
|
| 357 |
+
"step": 29000
|
| 358 |
+
},
|
| 359 |
+
{
|
| 360 |
+
"epoch": 1.6,
|
| 361 |
+
"learning_rate": 2.3413363615061557e-05,
|
| 362 |
+
"loss": 0.2595,
|
| 363 |
+
"step": 29500
|
| 364 |
+
},
|
| 365 |
+
{
|
| 366 |
+
"epoch": 1.62,
|
| 367 |
+
"learning_rate": 2.2962742659384634e-05,
|
| 368 |
+
"loss": 0.2523,
|
| 369 |
+
"step": 30000
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 1.65,
|
| 373 |
+
"learning_rate": 2.251212170370771e-05,
|
| 374 |
+
"loss": 0.2431,
|
| 375 |
+
"step": 30500
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.68,
|
| 379 |
+
"learning_rate": 2.2061500748030786e-05,
|
| 380 |
+
"loss": 0.2351,
|
| 381 |
+
"step": 31000
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.7,
|
| 385 |
+
"learning_rate": 2.1610879792353863e-05,
|
| 386 |
+
"loss": 0.2409,
|
| 387 |
+
"step": 31500
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 1.73,
|
| 391 |
+
"learning_rate": 2.1160258836676943e-05,
|
| 392 |
+
"loss": 0.23,
|
| 393 |
+
"step": 32000
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 1.76,
|
| 397 |
+
"learning_rate": 2.070963788100002e-05,
|
| 398 |
+
"loss": 0.223,
|
| 399 |
+
"step": 32500
|
| 400 |
+
},
|
| 401 |
+
{
|
| 402 |
+
"epoch": 1.78,
|
| 403 |
+
"learning_rate": 2.0259016925323095e-05,
|
| 404 |
+
"loss": 0.2247,
|
| 405 |
+
"step": 33000
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"epoch": 1.81,
|
| 409 |
+
"learning_rate": 1.9808395969646172e-05,
|
| 410 |
+
"loss": 0.221,
|
| 411 |
+
"step": 33500
|
| 412 |
+
},
|
| 413 |
+
{
|
| 414 |
+
"epoch": 1.84,
|
| 415 |
+
"learning_rate": 1.935777501396925e-05,
|
| 416 |
+
"loss": 0.222,
|
| 417 |
+
"step": 34000
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.87,
|
| 421 |
+
"learning_rate": 1.8907154058292328e-05,
|
| 422 |
+
"loss": 0.2262,
|
| 423 |
+
"step": 34500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.89,
|
| 427 |
+
"learning_rate": 1.8456533102615404e-05,
|
| 428 |
+
"loss": 0.2137,
|
| 429 |
+
"step": 35000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 1.92,
|
| 433 |
+
"learning_rate": 1.8005912146938484e-05,
|
| 434 |
+
"loss": 0.2143,
|
| 435 |
+
"step": 35500
|
| 436 |
+
},
|
| 437 |
+
{
|
| 438 |
+
"epoch": 1.95,
|
| 439 |
+
"learning_rate": 1.7555291191261557e-05,
|
| 440 |
+
"loss": 0.21,
|
| 441 |
+
"step": 36000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 1.97,
|
| 445 |
+
"learning_rate": 1.7104670235584637e-05,
|
| 446 |
+
"loss": 0.2025,
|
| 447 |
+
"step": 36500
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"epoch": 2.0,
|
| 451 |
+
"learning_rate": 1.6654049279907713e-05,
|
| 452 |
+
"loss": 0.2038,
|
| 453 |
+
"step": 37000
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 2.03,
|
| 457 |
+
"learning_rate": 1.620342832423079e-05,
|
| 458 |
+
"loss": 0.1669,
|
| 459 |
+
"step": 37500
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 2.05,
|
| 463 |
+
"learning_rate": 1.575280736855387e-05,
|
| 464 |
+
"loss": 0.1703,
|
| 465 |
+
"step": 38000
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 2.08,
|
| 469 |
+
"learning_rate": 1.5302186412876946e-05,
|
| 470 |
+
"loss": 0.1641,
|
| 471 |
+
"step": 38500
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 2.11,
|
| 475 |
+
"learning_rate": 1.4851565457200022e-05,
|
| 476 |
+
"loss": 0.1578,
|
| 477 |
+
"step": 39000
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 2.14,
|
| 481 |
+
"learning_rate": 1.44009445015231e-05,
|
| 482 |
+
"loss": 0.1523,
|
| 483 |
+
"step": 39500
|
| 484 |
+
},
|
| 485 |
+
{
|
| 486 |
+
"epoch": 2.16,
|
| 487 |
+
"learning_rate": 1.3950323545846179e-05,
|
| 488 |
+
"loss": 0.1566,
|
| 489 |
+
"step": 40000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 2.19,
|
| 493 |
+
"learning_rate": 1.3499702590169253e-05,
|
| 494 |
+
"loss": 0.1602,
|
| 495 |
+
"step": 40500
|
| 496 |
+
},
|
| 497 |
+
{
|
| 498 |
+
"epoch": 2.22,
|
| 499 |
+
"learning_rate": 1.304908163449233e-05,
|
| 500 |
+
"loss": 0.1523,
|
| 501 |
+
"step": 41000
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 2.24,
|
| 505 |
+
"learning_rate": 1.2598460678815408e-05,
|
| 506 |
+
"loss": 0.1551,
|
| 507 |
+
"step": 41500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 2.27,
|
| 511 |
+
"learning_rate": 1.2147839723138484e-05,
|
| 512 |
+
"loss": 0.1533,
|
| 513 |
+
"step": 42000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 2.3,
|
| 517 |
+
"learning_rate": 1.1697218767461562e-05,
|
| 518 |
+
"loss": 0.1517,
|
| 519 |
+
"step": 42500
|
| 520 |
+
},
|
| 521 |
+
{
|
| 522 |
+
"epoch": 2.33,
|
| 523 |
+
"learning_rate": 1.124659781178464e-05,
|
| 524 |
+
"loss": 0.1394,
|
| 525 |
+
"step": 43000
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 2.35,
|
| 529 |
+
"learning_rate": 1.0795976856107717e-05,
|
| 530 |
+
"loss": 0.1471,
|
| 531 |
+
"step": 43500
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 2.38,
|
| 535 |
+
"learning_rate": 1.0345355900430793e-05,
|
| 536 |
+
"loss": 0.1484,
|
| 537 |
+
"step": 44000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 2.41,
|
| 541 |
+
"learning_rate": 9.894734944753871e-06,
|
| 542 |
+
"loss": 0.1454,
|
| 543 |
+
"step": 44500
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 2.43,
|
| 547 |
+
"learning_rate": 9.444113989076948e-06,
|
| 548 |
+
"loss": 0.1368,
|
| 549 |
+
"step": 45000
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 2.46,
|
| 553 |
+
"learning_rate": 8.993493033400026e-06,
|
| 554 |
+
"loss": 0.1404,
|
| 555 |
+
"step": 45500
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 2.49,
|
| 559 |
+
"learning_rate": 8.542872077723104e-06,
|
| 560 |
+
"loss": 0.1362,
|
| 561 |
+
"step": 46000
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 2.51,
|
| 565 |
+
"learning_rate": 8.09225112204618e-06,
|
| 566 |
+
"loss": 0.135,
|
| 567 |
+
"step": 46500
|
| 568 |
+
},
|
| 569 |
+
{
|
| 570 |
+
"epoch": 2.54,
|
| 571 |
+
"learning_rate": 7.641630166369257e-06,
|
| 572 |
+
"loss": 0.1329,
|
| 573 |
+
"step": 47000
|
| 574 |
+
},
|
| 575 |
+
{
|
| 576 |
+
"epoch": 2.57,
|
| 577 |
+
"learning_rate": 7.191009210692335e-06,
|
| 578 |
+
"loss": 0.1313,
|
| 579 |
+
"step": 47500
|
| 580 |
+
},
|
| 581 |
+
{
|
| 582 |
+
"epoch": 2.6,
|
| 583 |
+
"learning_rate": 6.740388255015412e-06,
|
| 584 |
+
"loss": 0.1327,
|
| 585 |
+
"step": 48000
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 2.62,
|
| 589 |
+
"learning_rate": 6.289767299338489e-06,
|
| 590 |
+
"loss": 0.1242,
|
| 591 |
+
"step": 48500
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 2.65,
|
| 595 |
+
"learning_rate": 5.839146343661566e-06,
|
| 596 |
+
"loss": 0.1232,
|
| 597 |
+
"step": 49000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 2.68,
|
| 601 |
+
"learning_rate": 5.388525387984643e-06,
|
| 602 |
+
"loss": 0.1294,
|
| 603 |
+
"step": 49500
|
| 604 |
+
},
|
| 605 |
+
{
|
| 606 |
+
"epoch": 2.7,
|
| 607 |
+
"learning_rate": 4.93790443230772e-06,
|
| 608 |
+
"loss": 0.1275,
|
| 609 |
+
"step": 50000
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 2.73,
|
| 613 |
+
"learning_rate": 4.487283476630798e-06,
|
| 614 |
+
"loss": 0.1261,
|
| 615 |
+
"step": 50500
|
| 616 |
+
},
|
| 617 |
+
{
|
| 618 |
+
"epoch": 2.76,
|
| 619 |
+
"learning_rate": 4.0366625209538745e-06,
|
| 620 |
+
"loss": 0.1282,
|
| 621 |
+
"step": 51000
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 2.78,
|
| 625 |
+
"learning_rate": 3.586041565276952e-06,
|
| 626 |
+
"loss": 0.1245,
|
| 627 |
+
"step": 51500
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 2.81,
|
| 631 |
+
"learning_rate": 3.135420609600029e-06,
|
| 632 |
+
"loss": 0.1247,
|
| 633 |
+
"step": 52000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 2.84,
|
| 637 |
+
"learning_rate": 2.6847996539231063e-06,
|
| 638 |
+
"loss": 0.121,
|
| 639 |
+
"step": 52500
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 2.87,
|
| 643 |
+
"learning_rate": 2.2341786982461835e-06,
|
| 644 |
+
"loss": 0.1104,
|
| 645 |
+
"step": 53000
|
| 646 |
+
},
|
| 647 |
+
{
|
| 648 |
+
"epoch": 2.89,
|
| 649 |
+
"learning_rate": 1.7835577425692605e-06,
|
| 650 |
+
"loss": 0.1234,
|
| 651 |
+
"step": 53500
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 2.92,
|
| 655 |
+
"learning_rate": 1.3329367868923376e-06,
|
| 656 |
+
"loss": 0.1146,
|
| 657 |
+
"step": 54000
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 2.95,
|
| 661 |
+
"learning_rate": 8.823158312154149e-07,
|
| 662 |
+
"loss": 0.1197,
|
| 663 |
+
"step": 54500
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 2.97,
|
| 667 |
+
"learning_rate": 4.3169487553849206e-07,
|
| 668 |
+
"loss": 0.1116,
|
| 669 |
+
"step": 55000
|
| 670 |
+
},
|
| 671 |
{
|
| 672 |
"epoch": 3.0,
|
| 673 |
+
"step": 55479,
|
| 674 |
"total_flos": 2.3702775297552e+17,
|
| 675 |
+
"train_loss": 0.3350037185057538,
|
| 676 |
+
"train_runtime": 233521.0358,
|
| 677 |
+
"train_samples_per_second": 7.602,
|
| 678 |
+
"train_steps_per_second": 0.238
|
| 679 |
}
|
| 680 |
],
|
| 681 |
"logging_steps": 500,
|
| 682 |
+
"max_steps": 55479,
|
| 683 |
"num_train_epochs": 3,
|
| 684 |
"save_steps": 500,
|
| 685 |
"total_flos": 2.3702775297552e+17,
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4027
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d3a0286c3fcc4229a47d81ceff75544e3ab75dd419177f05c245554c91e4044a
|
| 3 |
size 4027
|