mataoxun commited on
Commit ·
e320939
1
Parent(s): e330121
提交
Browse files- optimizer.pt +1 -1
- pytorch_model.bin +1 -1
- rng_state.pth +1 -1
- scaler.pt +1 -1
- scheduler.pt +1 -1
- trainer_state.json +403 -7
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 721685265
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bb68bd928c8754c184cf5882a6e988da39c287f4ab5a5c6d3faee7f2ff7c44ef
|
| 3 |
size 721685265
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 377656855
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aa908b4d2d203d24ea9769f94d1ad6bffd4549d185e32c1a9f1fdaf53f1aa9c2
|
| 3 |
size 377656855
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14567
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:528bc3fc8c38e000630377e0cedd7d8ab65a2802b96dd640fccfa096eb42e4b7
|
| 3 |
size 14567
|
scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 559
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d6cc39bb03982f702acecfed4016e22c45ed660765a892634670fed905013c55
|
| 3 |
size 559
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 623
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4ffec80af644b58242768aec89d26ef0a97999a12561e3f4e5a27ef80389fbca
|
| 3 |
size 623
|
trainer_state.json
CHANGED
|
@@ -1,28 +1,424 @@
|
|
| 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": 0.43,
|
| 12 |
-
"learning_rate":
|
| 13 |
-
"loss":
|
| 14 |
"step": 500
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.85,
|
| 18 |
-
"learning_rate": 9.
|
| 19 |
-
"loss":
|
| 20 |
"step": 1000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 21 |
}
|
| 22 |
],
|
| 23 |
"max_steps": 34046,
|
| 24 |
"num_train_epochs": 29,
|
| 25 |
-
"total_flos":
|
| 26 |
"trial_name": null,
|
| 27 |
"trial_params": null
|
| 28 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 28.960817717206133,
|
| 5 |
+
"global_step": 34000,
|
| 6 |
"is_hyper_param_search": false,
|
| 7 |
"is_local_process_zero": true,
|
| 8 |
"is_world_process_zero": true,
|
| 9 |
"log_history": [
|
| 10 |
{
|
| 11 |
"epoch": 0.43,
|
| 12 |
+
"learning_rate": 9.369363916964232e-05,
|
| 13 |
+
"loss": 0.5598,
|
| 14 |
"step": 500
|
| 15 |
},
|
| 16 |
{
|
| 17 |
"epoch": 0.85,
|
| 18 |
+
"learning_rate": 9.218362282878412e-05,
|
| 19 |
+
"loss": 0.5495,
|
| 20 |
"step": 1000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 1.28,
|
| 24 |
+
"learning_rate": 9.067058040307452e-05,
|
| 25 |
+
"loss": 0.5008,
|
| 26 |
+
"step": 1500
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 1.7,
|
| 30 |
+
"learning_rate": 8.915753797736489e-05,
|
| 31 |
+
"loss": 0.4756,
|
| 32 |
+
"step": 2000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 2.13,
|
| 36 |
+
"learning_rate": 8.764449555165527e-05,
|
| 37 |
+
"loss": 0.4689,
|
| 38 |
+
"step": 2500
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 2.56,
|
| 42 |
+
"learning_rate": 8.613447921079707e-05,
|
| 43 |
+
"loss": 0.4223,
|
| 44 |
+
"step": 3000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 2.98,
|
| 48 |
+
"learning_rate": 8.462143678508746e-05,
|
| 49 |
+
"loss": 0.4444,
|
| 50 |
+
"step": 3500
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 3.41,
|
| 54 |
+
"learning_rate": 8.310839435937784e-05,
|
| 55 |
+
"loss": 0.3734,
|
| 56 |
+
"step": 4000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 3.83,
|
| 60 |
+
"learning_rate": 8.159535193366822e-05,
|
| 61 |
+
"loss": 0.4001,
|
| 62 |
+
"step": 4500
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 4.26,
|
| 66 |
+
"learning_rate": 8.00823095079586e-05,
|
| 67 |
+
"loss": 0.3875,
|
| 68 |
+
"step": 5000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 4.68,
|
| 72 |
+
"learning_rate": 7.856926708224899e-05,
|
| 73 |
+
"loss": 0.3557,
|
| 74 |
+
"step": 5500
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 5.11,
|
| 78 |
+
"learning_rate": 7.705622465653937e-05,
|
| 79 |
+
"loss": 0.34,
|
| 80 |
+
"step": 6000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 5.54,
|
| 84 |
+
"learning_rate": 7.554318223082976e-05,
|
| 85 |
+
"loss": 0.3244,
|
| 86 |
+
"step": 6500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 5.96,
|
| 90 |
+
"learning_rate": 7.403013980512014e-05,
|
| 91 |
+
"loss": 0.3372,
|
| 92 |
+
"step": 7000
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 6.39,
|
| 96 |
+
"learning_rate": 7.251709737941052e-05,
|
| 97 |
+
"loss": 0.3099,
|
| 98 |
+
"step": 7500
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 6.81,
|
| 102 |
+
"learning_rate": 7.10040549537009e-05,
|
| 103 |
+
"loss": 0.3108,
|
| 104 |
+
"step": 8000
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 7.24,
|
| 108 |
+
"learning_rate": 6.949101252799129e-05,
|
| 109 |
+
"loss": 0.2705,
|
| 110 |
+
"step": 8500
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 7.67,
|
| 114 |
+
"learning_rate": 6.797797010228167e-05,
|
| 115 |
+
"loss": 0.2856,
|
| 116 |
+
"step": 9000
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 8.09,
|
| 120 |
+
"learning_rate": 6.646795376142348e-05,
|
| 121 |
+
"loss": 0.2885,
|
| 122 |
+
"step": 9500
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 8.52,
|
| 126 |
+
"learning_rate": 6.495491133571386e-05,
|
| 127 |
+
"loss": 0.2729,
|
| 128 |
+
"step": 10000
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 8.94,
|
| 132 |
+
"learning_rate": 6.344186891000424e-05,
|
| 133 |
+
"loss": 0.274,
|
| 134 |
+
"step": 10500
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 9.37,
|
| 138 |
+
"learning_rate": 6.192882648429462e-05,
|
| 139 |
+
"loss": 0.2465,
|
| 140 |
+
"step": 11000
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 9.8,
|
| 144 |
+
"learning_rate": 6.041881014343642e-05,
|
| 145 |
+
"loss": 0.2517,
|
| 146 |
+
"step": 11500
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 10.22,
|
| 150 |
+
"learning_rate": 5.890576771772681e-05,
|
| 151 |
+
"loss": 0.2512,
|
| 152 |
+
"step": 12000
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 10.65,
|
| 156 |
+
"learning_rate": 5.739272529201719e-05,
|
| 157 |
+
"loss": 0.239,
|
| 158 |
+
"step": 12500
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 11.07,
|
| 162 |
+
"learning_rate": 5.587968286630757e-05,
|
| 163 |
+
"loss": 0.2302,
|
| 164 |
+
"step": 13000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 11.5,
|
| 168 |
+
"learning_rate": 5.436966652544938e-05,
|
| 169 |
+
"loss": 0.2256,
|
| 170 |
+
"step": 13500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 11.93,
|
| 174 |
+
"learning_rate": 5.285662409973976e-05,
|
| 175 |
+
"loss": 0.2259,
|
| 176 |
+
"step": 14000
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 12.35,
|
| 180 |
+
"learning_rate": 5.1343581674030136e-05,
|
| 181 |
+
"loss": 0.206,
|
| 182 |
+
"step": 14500
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 12.78,
|
| 186 |
+
"learning_rate": 4.9830539248320526e-05,
|
| 187 |
+
"loss": 0.2197,
|
| 188 |
+
"step": 15000
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 13.2,
|
| 192 |
+
"learning_rate": 4.831749682261091e-05,
|
| 193 |
+
"loss": 0.2008,
|
| 194 |
+
"step": 15500
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 13.63,
|
| 198 |
+
"learning_rate": 4.680445439690129e-05,
|
| 199 |
+
"loss": 0.1843,
|
| 200 |
+
"step": 16000
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 14.05,
|
| 204 |
+
"learning_rate": 4.5291411971191675e-05,
|
| 205 |
+
"loss": 0.1896,
|
| 206 |
+
"step": 16500
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 14.48,
|
| 210 |
+
"learning_rate": 4.377836954548206e-05,
|
| 211 |
+
"loss": 0.184,
|
| 212 |
+
"step": 17000
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 14.91,
|
| 216 |
+
"learning_rate": 4.226532711977244e-05,
|
| 217 |
+
"loss": 0.1788,
|
| 218 |
+
"step": 17500
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 15.33,
|
| 222 |
+
"learning_rate": 4.0752284694062824e-05,
|
| 223 |
+
"loss": 0.174,
|
| 224 |
+
"step": 18000
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 15.76,
|
| 228 |
+
"learning_rate": 3.923924226835321e-05,
|
| 229 |
+
"loss": 0.177,
|
| 230 |
+
"step": 18500
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 16.18,
|
| 234 |
+
"learning_rate": 3.772619984264359e-05,
|
| 235 |
+
"loss": 0.161,
|
| 236 |
+
"step": 19000
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 16.61,
|
| 240 |
+
"learning_rate": 3.6213157416933974e-05,
|
| 241 |
+
"loss": 0.1548,
|
| 242 |
+
"step": 19500
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 17.04,
|
| 246 |
+
"learning_rate": 3.470616716092719e-05,
|
| 247 |
+
"loss": 0.1719,
|
| 248 |
+
"step": 20000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 17.46,
|
| 252 |
+
"learning_rate": 3.3193124735217576e-05,
|
| 253 |
+
"loss": 0.159,
|
| 254 |
+
"step": 20500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 17.89,
|
| 258 |
+
"learning_rate": 3.168008230950796e-05,
|
| 259 |
+
"loss": 0.1466,
|
| 260 |
+
"step": 21000
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 18.31,
|
| 264 |
+
"learning_rate": 3.0167039883798342e-05,
|
| 265 |
+
"loss": 0.153,
|
| 266 |
+
"step": 21500
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 18.74,
|
| 270 |
+
"learning_rate": 2.865399745808873e-05,
|
| 271 |
+
"loss": 0.1399,
|
| 272 |
+
"step": 22000
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 19.17,
|
| 276 |
+
"learning_rate": 2.7140955032379112e-05,
|
| 277 |
+
"loss": 0.1369,
|
| 278 |
+
"step": 22500
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 19.59,
|
| 282 |
+
"learning_rate": 2.5630938691520913e-05,
|
| 283 |
+
"loss": 0.1397,
|
| 284 |
+
"step": 23000
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 20.02,
|
| 288 |
+
"learning_rate": 2.4117896265811293e-05,
|
| 289 |
+
"loss": 0.1351,
|
| 290 |
+
"step": 23500
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 20.44,
|
| 294 |
+
"learning_rate": 2.260485384010168e-05,
|
| 295 |
+
"loss": 0.1256,
|
| 296 |
+
"step": 24000
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 20.87,
|
| 300 |
+
"learning_rate": 2.109181141439206e-05,
|
| 301 |
+
"loss": 0.128,
|
| 302 |
+
"step": 24500
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 21.29,
|
| 306 |
+
"learning_rate": 1.9581795073533863e-05,
|
| 307 |
+
"loss": 0.1184,
|
| 308 |
+
"step": 25000
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 21.72,
|
| 312 |
+
"learning_rate": 1.8068752647824246e-05,
|
| 313 |
+
"loss": 0.1312,
|
| 314 |
+
"step": 25500
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 22.15,
|
| 318 |
+
"learning_rate": 1.655571022211463e-05,
|
| 319 |
+
"loss": 0.1138,
|
| 320 |
+
"step": 26000
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 22.57,
|
| 324 |
+
"learning_rate": 1.5042667796405011e-05,
|
| 325 |
+
"loss": 0.1123,
|
| 326 |
+
"step": 26500
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 23.0,
|
| 330 |
+
"learning_rate": 1.3529625370695396e-05,
|
| 331 |
+
"loss": 0.1144,
|
| 332 |
+
"step": 27000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 23.42,
|
| 336 |
+
"learning_rate": 1.2019609029837197e-05,
|
| 337 |
+
"loss": 0.1076,
|
| 338 |
+
"step": 27500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 23.85,
|
| 342 |
+
"learning_rate": 1.0509592688979e-05,
|
| 343 |
+
"loss": 0.1054,
|
| 344 |
+
"step": 28000
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 24.28,
|
| 348 |
+
"learning_rate": 8.996550263269383e-06,
|
| 349 |
+
"loss": 0.1033,
|
| 350 |
+
"step": 28500
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 24.7,
|
| 354 |
+
"learning_rate": 7.483507837559765e-06,
|
| 355 |
+
"loss": 0.0993,
|
| 356 |
+
"step": 29000
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 25.13,
|
| 360 |
+
"learning_rate": 5.970465411850148e-06,
|
| 361 |
+
"loss": 0.1,
|
| 362 |
+
"step": 29500
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 25.55,
|
| 366 |
+
"learning_rate": 4.457422986140532e-06,
|
| 367 |
+
"loss": 0.0985,
|
| 368 |
+
"step": 30000
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 25.98,
|
| 372 |
+
"learning_rate": 2.9443805604309144e-06,
|
| 373 |
+
"loss": 0.0928,
|
| 374 |
+
"step": 30500
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 26.41,
|
| 378 |
+
"learning_rate": 1.4313381347212977e-06,
|
| 379 |
+
"loss": 0.0913,
|
| 380 |
+
"step": 31000
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 26.83,
|
| 384 |
+
"learning_rate": 0.0,
|
| 385 |
+
"loss": 0.0947,
|
| 386 |
+
"step": 31500
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 27.26,
|
| 390 |
+
"learning_rate": 0.0,
|
| 391 |
+
"loss": 0.1028,
|
| 392 |
+
"step": 32000
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 27.68,
|
| 396 |
+
"learning_rate": 0.0,
|
| 397 |
+
"loss": 0.0895,
|
| 398 |
+
"step": 32500
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 28.11,
|
| 402 |
+
"learning_rate": 0.0,
|
| 403 |
+
"loss": 0.0894,
|
| 404 |
+
"step": 33000
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 28.53,
|
| 408 |
+
"learning_rate": 0.0,
|
| 409 |
+
"loss": 0.0894,
|
| 410 |
+
"step": 33500
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 28.96,
|
| 414 |
+
"learning_rate": 0.0,
|
| 415 |
+
"loss": 0.0924,
|
| 416 |
+
"step": 34000
|
| 417 |
}
|
| 418 |
],
|
| 419 |
"max_steps": 34046,
|
| 420 |
"num_train_epochs": 29,
|
| 421 |
+
"total_flos": 9.909309665241907e+18,
|
| 422 |
"trial_name": null,
|
| 423 |
"trial_params": null
|
| 424 |
}
|