Upload trainer_state.json
Browse files- trainer_state.json +407 -101
trainer_state.json
CHANGED
|
@@ -1,208 +1,514 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch":
|
| 5 |
-
"global_step":
|
| 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.
|
| 12 |
-
"learning_rate":
|
| 13 |
-
"loss": 2.
|
| 14 |
"step": 3
|
| 15 |
},
|
| 16 |
{
|
| 17 |
-
"epoch": 0.
|
| 18 |
-
"learning_rate":
|
| 19 |
-
"loss": 2.
|
| 20 |
"step": 6
|
| 21 |
},
|
| 22 |
{
|
| 23 |
-
"epoch": 0.
|
| 24 |
-
"learning_rate":
|
| 25 |
-
"loss": 2.
|
| 26 |
"step": 9
|
| 27 |
},
|
| 28 |
{
|
| 29 |
-
"epoch": 0.
|
| 30 |
-
"learning_rate":
|
| 31 |
-
"loss": 2.
|
| 32 |
"step": 12
|
| 33 |
},
|
| 34 |
{
|
| 35 |
-
"epoch": 0.
|
| 36 |
-
"learning_rate":
|
| 37 |
-
"loss": 2.
|
| 38 |
"step": 15
|
| 39 |
},
|
| 40 |
{
|
| 41 |
-
"epoch": 0.
|
| 42 |
-
"learning_rate":
|
| 43 |
-
"loss": 2.
|
| 44 |
"step": 18
|
| 45 |
},
|
| 46 |
{
|
| 47 |
-
"epoch": 0.
|
| 48 |
-
"learning_rate":
|
| 49 |
-
"loss": 2.
|
| 50 |
"step": 21
|
| 51 |
},
|
| 52 |
{
|
| 53 |
-
"epoch": 0.
|
| 54 |
-
"learning_rate":
|
| 55 |
-
"loss": 2.
|
| 56 |
"step": 24
|
| 57 |
},
|
| 58 |
{
|
| 59 |
-
"epoch": 0.
|
| 60 |
-
"learning_rate":
|
| 61 |
-
"loss": 2.
|
| 62 |
"step": 27
|
| 63 |
},
|
| 64 |
{
|
| 65 |
-
"epoch": 0.
|
| 66 |
-
"learning_rate":
|
| 67 |
-
"loss": 2.
|
| 68 |
"step": 30
|
| 69 |
},
|
| 70 |
{
|
| 71 |
-
"epoch": 0.
|
| 72 |
-
"learning_rate":
|
| 73 |
-
"loss": 2.
|
| 74 |
"step": 33
|
| 75 |
},
|
| 76 |
{
|
| 77 |
-
"epoch": 0.
|
| 78 |
-
"learning_rate":
|
| 79 |
-
"loss": 2.
|
| 80 |
"step": 36
|
| 81 |
},
|
| 82 |
{
|
| 83 |
-
"epoch": 0.
|
| 84 |
-
"learning_rate":
|
| 85 |
-
"loss": 2.
|
| 86 |
"step": 39
|
| 87 |
},
|
| 88 |
{
|
| 89 |
-
"epoch": 0.
|
| 90 |
-
"learning_rate":
|
| 91 |
-
"loss": 2.
|
| 92 |
"step": 42
|
| 93 |
},
|
| 94 |
{
|
| 95 |
-
"epoch": 0.
|
| 96 |
-
"learning_rate":
|
| 97 |
-
"loss": 2.
|
| 98 |
"step": 45
|
| 99 |
},
|
| 100 |
{
|
| 101 |
-
"epoch": 0.
|
| 102 |
-
"learning_rate":
|
| 103 |
-
"loss": 2.
|
| 104 |
"step": 48
|
| 105 |
},
|
| 106 |
{
|
| 107 |
-
"epoch":
|
| 108 |
-
"learning_rate":
|
| 109 |
-
"loss":
|
| 110 |
"step": 51
|
| 111 |
},
|
| 112 |
{
|
| 113 |
-
"epoch":
|
| 114 |
-
"learning_rate":
|
| 115 |
-
"loss": 2.
|
| 116 |
"step": 54
|
| 117 |
},
|
| 118 |
{
|
| 119 |
-
"epoch":
|
| 120 |
-
"learning_rate":
|
| 121 |
-
"loss": 2.
|
| 122 |
"step": 57
|
| 123 |
},
|
| 124 |
{
|
| 125 |
-
"epoch":
|
| 126 |
-
"learning_rate":
|
| 127 |
-
"loss": 2.
|
| 128 |
"step": 60
|
| 129 |
},
|
| 130 |
{
|
| 131 |
-
"epoch":
|
| 132 |
-
"learning_rate":
|
| 133 |
-
"loss": 2.
|
| 134 |
"step": 63
|
| 135 |
},
|
| 136 |
{
|
| 137 |
-
"epoch":
|
| 138 |
-
"learning_rate":
|
| 139 |
-
"loss": 2.
|
| 140 |
"step": 66
|
| 141 |
},
|
| 142 |
{
|
| 143 |
-
"epoch":
|
| 144 |
-
"learning_rate":
|
| 145 |
-
"loss": 2.
|
| 146 |
"step": 69
|
| 147 |
},
|
| 148 |
{
|
| 149 |
-
"epoch":
|
| 150 |
-
"learning_rate":
|
| 151 |
-
"loss": 2.
|
| 152 |
"step": 72
|
| 153 |
},
|
| 154 |
{
|
| 155 |
-
"epoch":
|
| 156 |
-
"learning_rate":
|
| 157 |
-
"loss": 2.
|
| 158 |
"step": 75
|
| 159 |
},
|
| 160 |
{
|
| 161 |
-
"epoch":
|
| 162 |
-
"learning_rate":
|
| 163 |
-
"loss": 2.
|
| 164 |
"step": 78
|
| 165 |
},
|
| 166 |
{
|
| 167 |
-
"epoch":
|
| 168 |
-
"learning_rate":
|
| 169 |
-
"loss": 2.
|
| 170 |
"step": 81
|
| 171 |
},
|
| 172 |
{
|
| 173 |
-
"epoch":
|
| 174 |
-
"learning_rate":
|
| 175 |
-
"loss": 2.
|
| 176 |
"step": 84
|
| 177 |
},
|
| 178 |
{
|
| 179 |
-
"epoch":
|
| 180 |
-
"learning_rate":
|
| 181 |
-
"loss": 2.
|
| 182 |
"step": 87
|
| 183 |
},
|
| 184 |
{
|
| 185 |
-
"epoch":
|
| 186 |
-
"learning_rate":
|
| 187 |
-
"loss": 2.
|
| 188 |
"step": 90
|
| 189 |
},
|
| 190 |
{
|
| 191 |
-
"epoch":
|
| 192 |
-
"learning_rate":
|
| 193 |
-
"loss": 2.
|
| 194 |
"step": 93
|
| 195 |
},
|
| 196 |
{
|
| 197 |
-
"epoch":
|
| 198 |
-
"learning_rate":
|
| 199 |
-
"loss": 2.
|
| 200 |
"step": 96
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 201 |
}
|
| 202 |
],
|
| 203 |
-
"max_steps":
|
| 204 |
-
"num_train_epochs":
|
| 205 |
-
"total_flos":
|
| 206 |
"trial_name": null,
|
| 207 |
"trial_params": null
|
| 208 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 2.4508378122036043,
|
| 5 |
+
"global_step": 250,
|
| 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": 1.5000000000000002e-05,
|
| 13 |
+
"loss": 2.8422,
|
| 14 |
"step": 3
|
| 15 |
},
|
| 16 |
{
|
| 17 |
+
"epoch": 0.06,
|
| 18 |
+
"learning_rate": 3.0000000000000004e-05,
|
| 19 |
+
"loss": 2.7419,
|
| 20 |
"step": 6
|
| 21 |
},
|
| 22 |
{
|
| 23 |
+
"epoch": 0.09,
|
| 24 |
+
"learning_rate": 4.5e-05,
|
| 25 |
+
"loss": 2.722,
|
| 26 |
"step": 9
|
| 27 |
},
|
| 28 |
{
|
| 29 |
+
"epoch": 0.12,
|
| 30 |
+
"learning_rate": 6.000000000000001e-05,
|
| 31 |
+
"loss": 2.6536,
|
| 32 |
"step": 12
|
| 33 |
},
|
| 34 |
{
|
| 35 |
+
"epoch": 0.15,
|
| 36 |
+
"learning_rate": 7.500000000000001e-05,
|
| 37 |
+
"loss": 2.795,
|
| 38 |
"step": 15
|
| 39 |
},
|
| 40 |
{
|
| 41 |
+
"epoch": 0.18,
|
| 42 |
+
"learning_rate": 8e-05,
|
| 43 |
+
"loss": 2.7823,
|
| 44 |
"step": 18
|
| 45 |
},
|
| 46 |
{
|
| 47 |
+
"epoch": 0.21,
|
| 48 |
+
"learning_rate": 8e-05,
|
| 49 |
+
"loss": 2.7227,
|
| 50 |
"step": 21
|
| 51 |
},
|
| 52 |
{
|
| 53 |
+
"epoch": 0.24,
|
| 54 |
+
"learning_rate": 8e-05,
|
| 55 |
+
"loss": 2.7894,
|
| 56 |
"step": 24
|
| 57 |
},
|
| 58 |
{
|
| 59 |
+
"epoch": 0.26,
|
| 60 |
+
"learning_rate": 8e-05,
|
| 61 |
+
"loss": 2.8172,
|
| 62 |
"step": 27
|
| 63 |
},
|
| 64 |
{
|
| 65 |
+
"epoch": 0.29,
|
| 66 |
+
"learning_rate": 8e-05,
|
| 67 |
+
"loss": 2.8764,
|
| 68 |
"step": 30
|
| 69 |
},
|
| 70 |
{
|
| 71 |
+
"epoch": 0.32,
|
| 72 |
+
"learning_rate": 8e-05,
|
| 73 |
+
"loss": 2.7575,
|
| 74 |
"step": 33
|
| 75 |
},
|
| 76 |
{
|
| 77 |
+
"epoch": 0.35,
|
| 78 |
+
"learning_rate": 8e-05,
|
| 79 |
+
"loss": 2.7926,
|
| 80 |
"step": 36
|
| 81 |
},
|
| 82 |
{
|
| 83 |
+
"epoch": 0.38,
|
| 84 |
+
"learning_rate": 8e-05,
|
| 85 |
+
"loss": 2.8379,
|
| 86 |
"step": 39
|
| 87 |
},
|
| 88 |
{
|
| 89 |
+
"epoch": 0.41,
|
| 90 |
+
"learning_rate": 8e-05,
|
| 91 |
+
"loss": 2.8733,
|
| 92 |
"step": 42
|
| 93 |
},
|
| 94 |
{
|
| 95 |
+
"epoch": 0.44,
|
| 96 |
+
"learning_rate": 8e-05,
|
| 97 |
+
"loss": 2.7793,
|
| 98 |
"step": 45
|
| 99 |
},
|
| 100 |
{
|
| 101 |
+
"epoch": 0.47,
|
| 102 |
+
"learning_rate": 8e-05,
|
| 103 |
+
"loss": 2.8022,
|
| 104 |
"step": 48
|
| 105 |
},
|
| 106 |
{
|
| 107 |
+
"epoch": 0.5,
|
| 108 |
+
"learning_rate": 8e-05,
|
| 109 |
+
"loss": 2.8124,
|
| 110 |
"step": 51
|
| 111 |
},
|
| 112 |
{
|
| 113 |
+
"epoch": 0.53,
|
| 114 |
+
"learning_rate": 8e-05,
|
| 115 |
+
"loss": 2.8472,
|
| 116 |
"step": 54
|
| 117 |
},
|
| 118 |
{
|
| 119 |
+
"epoch": 0.56,
|
| 120 |
+
"learning_rate": 8e-05,
|
| 121 |
+
"loss": 2.7553,
|
| 122 |
"step": 57
|
| 123 |
},
|
| 124 |
{
|
| 125 |
+
"epoch": 0.59,
|
| 126 |
+
"learning_rate": 8e-05,
|
| 127 |
+
"loss": 2.8276,
|
| 128 |
"step": 60
|
| 129 |
},
|
| 130 |
{
|
| 131 |
+
"epoch": 0.62,
|
| 132 |
+
"learning_rate": 8e-05,
|
| 133 |
+
"loss": 2.7441,
|
| 134 |
"step": 63
|
| 135 |
},
|
| 136 |
{
|
| 137 |
+
"epoch": 0.65,
|
| 138 |
+
"learning_rate": 8e-05,
|
| 139 |
+
"loss": 2.7039,
|
| 140 |
"step": 66
|
| 141 |
},
|
| 142 |
{
|
| 143 |
+
"epoch": 0.68,
|
| 144 |
+
"learning_rate": 8e-05,
|
| 145 |
+
"loss": 2.7785,
|
| 146 |
"step": 69
|
| 147 |
},
|
| 148 |
{
|
| 149 |
+
"epoch": 0.71,
|
| 150 |
+
"learning_rate": 8e-05,
|
| 151 |
+
"loss": 2.8397,
|
| 152 |
"step": 72
|
| 153 |
},
|
| 154 |
{
|
| 155 |
+
"epoch": 0.74,
|
| 156 |
+
"learning_rate": 8e-05,
|
| 157 |
+
"loss": 2.9083,
|
| 158 |
"step": 75
|
| 159 |
},
|
| 160 |
{
|
| 161 |
+
"epoch": 0.76,
|
| 162 |
+
"learning_rate": 8e-05,
|
| 163 |
+
"loss": 2.8106,
|
| 164 |
"step": 78
|
| 165 |
},
|
| 166 |
{
|
| 167 |
+
"epoch": 0.79,
|
| 168 |
+
"learning_rate": 8e-05,
|
| 169 |
+
"loss": 2.7675,
|
| 170 |
"step": 81
|
| 171 |
},
|
| 172 |
{
|
| 173 |
+
"epoch": 0.82,
|
| 174 |
+
"learning_rate": 8e-05,
|
| 175 |
+
"loss": 2.6857,
|
| 176 |
"step": 84
|
| 177 |
},
|
| 178 |
{
|
| 179 |
+
"epoch": 0.85,
|
| 180 |
+
"learning_rate": 8e-05,
|
| 181 |
+
"loss": 2.8444,
|
| 182 |
"step": 87
|
| 183 |
},
|
| 184 |
{
|
| 185 |
+
"epoch": 0.88,
|
| 186 |
+
"learning_rate": 8e-05,
|
| 187 |
+
"loss": 2.7138,
|
| 188 |
"step": 90
|
| 189 |
},
|
| 190 |
{
|
| 191 |
+
"epoch": 0.91,
|
| 192 |
+
"learning_rate": 8e-05,
|
| 193 |
+
"loss": 2.7543,
|
| 194 |
"step": 93
|
| 195 |
},
|
| 196 |
{
|
| 197 |
+
"epoch": 0.94,
|
| 198 |
+
"learning_rate": 8e-05,
|
| 199 |
+
"loss": 2.722,
|
| 200 |
"step": 96
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.97,
|
| 204 |
+
"learning_rate": 8e-05,
|
| 205 |
+
"loss": 2.749,
|
| 206 |
+
"step": 99
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 1.0,
|
| 210 |
+
"learning_rate": 8e-05,
|
| 211 |
+
"loss": 2.7536,
|
| 212 |
+
"step": 102
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 1.03,
|
| 216 |
+
"learning_rate": 8e-05,
|
| 217 |
+
"loss": 2.7219,
|
| 218 |
+
"step": 105
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 1.06,
|
| 222 |
+
"learning_rate": 8e-05,
|
| 223 |
+
"loss": 2.7833,
|
| 224 |
+
"step": 108
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 1.09,
|
| 228 |
+
"learning_rate": 8e-05,
|
| 229 |
+
"loss": 2.7061,
|
| 230 |
+
"step": 111
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 1.12,
|
| 234 |
+
"learning_rate": 8e-05,
|
| 235 |
+
"loss": 2.7059,
|
| 236 |
+
"step": 114
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 1.15,
|
| 240 |
+
"learning_rate": 8e-05,
|
| 241 |
+
"loss": 2.7589,
|
| 242 |
+
"step": 117
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 1.18,
|
| 246 |
+
"learning_rate": 8e-05,
|
| 247 |
+
"loss": 2.7999,
|
| 248 |
+
"step": 120
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 1.21,
|
| 252 |
+
"learning_rate": 8e-05,
|
| 253 |
+
"loss": 2.6974,
|
| 254 |
+
"step": 123
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 1.24,
|
| 258 |
+
"learning_rate": 8e-05,
|
| 259 |
+
"loss": 2.6833,
|
| 260 |
+
"step": 126
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 1.26,
|
| 264 |
+
"learning_rate": 8e-05,
|
| 265 |
+
"loss": 2.8006,
|
| 266 |
+
"step": 129
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 1.29,
|
| 270 |
+
"learning_rate": 8e-05,
|
| 271 |
+
"loss": 2.7804,
|
| 272 |
+
"step": 132
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 1.32,
|
| 276 |
+
"learning_rate": 8e-05,
|
| 277 |
+
"loss": 2.6436,
|
| 278 |
+
"step": 135
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 1.35,
|
| 282 |
+
"learning_rate": 8e-05,
|
| 283 |
+
"loss": 2.8129,
|
| 284 |
+
"step": 138
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 1.38,
|
| 288 |
+
"learning_rate": 8e-05,
|
| 289 |
+
"loss": 2.7529,
|
| 290 |
+
"step": 141
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 1.41,
|
| 294 |
+
"learning_rate": 8e-05,
|
| 295 |
+
"loss": 2.7318,
|
| 296 |
+
"step": 144
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 1.44,
|
| 300 |
+
"learning_rate": 8e-05,
|
| 301 |
+
"loss": 2.7619,
|
| 302 |
+
"step": 147
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 1.47,
|
| 306 |
+
"learning_rate": 8e-05,
|
| 307 |
+
"loss": 2.7427,
|
| 308 |
+
"step": 150
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 1.5,
|
| 312 |
+
"learning_rate": 8e-05,
|
| 313 |
+
"loss": 2.7136,
|
| 314 |
+
"step": 153
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 1.53,
|
| 318 |
+
"learning_rate": 8e-05,
|
| 319 |
+
"loss": 2.6851,
|
| 320 |
+
"step": 156
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 1.56,
|
| 324 |
+
"learning_rate": 8e-05,
|
| 325 |
+
"loss": 2.8152,
|
| 326 |
+
"step": 159
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.59,
|
| 330 |
+
"learning_rate": 8e-05,
|
| 331 |
+
"loss": 2.8048,
|
| 332 |
+
"step": 162
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 1.62,
|
| 336 |
+
"learning_rate": 8e-05,
|
| 337 |
+
"loss": 2.6799,
|
| 338 |
+
"step": 165
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 1.65,
|
| 342 |
+
"learning_rate": 8e-05,
|
| 343 |
+
"loss": 2.6929,
|
| 344 |
+
"step": 168
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 1.68,
|
| 348 |
+
"learning_rate": 8e-05,
|
| 349 |
+
"loss": 2.797,
|
| 350 |
+
"step": 171
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 1.71,
|
| 354 |
+
"learning_rate": 8e-05,
|
| 355 |
+
"loss": 2.7406,
|
| 356 |
+
"step": 174
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 1.74,
|
| 360 |
+
"learning_rate": 8e-05,
|
| 361 |
+
"loss": 2.7053,
|
| 362 |
+
"step": 177
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 1.76,
|
| 366 |
+
"learning_rate": 8e-05,
|
| 367 |
+
"loss": 2.7513,
|
| 368 |
+
"step": 180
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.79,
|
| 372 |
+
"learning_rate": 8e-05,
|
| 373 |
+
"loss": 2.7608,
|
| 374 |
+
"step": 183
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.82,
|
| 378 |
+
"learning_rate": 8e-05,
|
| 379 |
+
"loss": 2.6898,
|
| 380 |
+
"step": 186
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 1.85,
|
| 384 |
+
"learning_rate": 8e-05,
|
| 385 |
+
"loss": 2.7014,
|
| 386 |
+
"step": 189
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 1.88,
|
| 390 |
+
"learning_rate": 8e-05,
|
| 391 |
+
"loss": 2.7365,
|
| 392 |
+
"step": 192
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 1.91,
|
| 396 |
+
"learning_rate": 8e-05,
|
| 397 |
+
"loss": 2.7513,
|
| 398 |
+
"step": 195
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 1.94,
|
| 402 |
+
"learning_rate": 8e-05,
|
| 403 |
+
"loss": 2.7584,
|
| 404 |
+
"step": 198
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 1.97,
|
| 408 |
+
"learning_rate": 8e-05,
|
| 409 |
+
"loss": 2.78,
|
| 410 |
+
"step": 201
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 2.0,
|
| 414 |
+
"learning_rate": 8e-05,
|
| 415 |
+
"loss": 2.7855,
|
| 416 |
+
"step": 204
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 2.03,
|
| 420 |
+
"learning_rate": 8e-05,
|
| 421 |
+
"loss": 2.7826,
|
| 422 |
+
"step": 207
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 2.06,
|
| 426 |
+
"learning_rate": 8e-05,
|
| 427 |
+
"loss": 2.7339,
|
| 428 |
+
"step": 210
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 2.09,
|
| 432 |
+
"learning_rate": 8e-05,
|
| 433 |
+
"loss": 2.7663,
|
| 434 |
+
"step": 213
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 2.12,
|
| 438 |
+
"learning_rate": 8e-05,
|
| 439 |
+
"loss": 2.775,
|
| 440 |
+
"step": 216
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 2.15,
|
| 444 |
+
"learning_rate": 8e-05,
|
| 445 |
+
"loss": 2.7496,
|
| 446 |
+
"step": 219
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 2.18,
|
| 450 |
+
"learning_rate": 8e-05,
|
| 451 |
+
"loss": 2.6277,
|
| 452 |
+
"step": 222
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 2.21,
|
| 456 |
+
"learning_rate": 8e-05,
|
| 457 |
+
"loss": 2.6725,
|
| 458 |
+
"step": 225
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 2.24,
|
| 462 |
+
"learning_rate": 8e-05,
|
| 463 |
+
"loss": 2.6692,
|
| 464 |
+
"step": 228
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 2.26,
|
| 468 |
+
"learning_rate": 8e-05,
|
| 469 |
+
"loss": 2.7542,
|
| 470 |
+
"step": 231
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 2.29,
|
| 474 |
+
"learning_rate": 8e-05,
|
| 475 |
+
"loss": 2.7232,
|
| 476 |
+
"step": 234
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 2.32,
|
| 480 |
+
"learning_rate": 8e-05,
|
| 481 |
+
"loss": 2.6342,
|
| 482 |
+
"step": 237
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 2.35,
|
| 486 |
+
"learning_rate": 8e-05,
|
| 487 |
+
"loss": 2.7458,
|
| 488 |
+
"step": 240
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 2.38,
|
| 492 |
+
"learning_rate": 8e-05,
|
| 493 |
+
"loss": 2.5471,
|
| 494 |
+
"step": 243
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 2.41,
|
| 498 |
+
"learning_rate": 8e-05,
|
| 499 |
+
"loss": 2.7454,
|
| 500 |
+
"step": 246
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 2.44,
|
| 504 |
+
"learning_rate": 8e-05,
|
| 505 |
+
"loss": 2.6822,
|
| 506 |
+
"step": 249
|
| 507 |
}
|
| 508 |
],
|
| 509 |
+
"max_steps": 510,
|
| 510 |
+
"num_train_epochs": 5,
|
| 511 |
+
"total_flos": 2.8472026748839526e+17,
|
| 512 |
"trial_name": null,
|
| 513 |
"trial_params": null
|
| 514 |
}
|