Upload 7 files
Browse files- pytorch_model.bin +3 -0
- rng_state_0.pth +3 -0
- rng_state_1.pth +3 -0
- rng_state_2.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +256 -0
- training_args.bin +3 -0
pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:73ad8f79a630af8c080504accae2319ebdad84507b20d9d7445ac57f5c2ba1b5
|
| 3 |
+
size 711493997
|
rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8938ef67075cb4e3f800591efb4ebcbbae812b52006cd01daad84b386e575bbb
|
| 3 |
+
size 14503
|
rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8938ef67075cb4e3f800591efb4ebcbbae812b52006cd01daad84b386e575bbb
|
| 3 |
+
size 14503
|
rng_state_2.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8938ef67075cb4e3f800591efb4ebcbbae812b52006cd01daad84b386e575bbb
|
| 3 |
+
size 14503
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f3a6f3b73f0fabe66db6f6604c5ac99d8166a37ed0b1e9bdd2d6de950875fed5
|
| 3 |
+
size 623
|
trainer_state.json
ADDED
|
@@ -0,0 +1,256 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 3.0,
|
| 5 |
+
"global_step": 36816,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.08,
|
| 12 |
+
"learning_rate": 5.431830526887561e-06,
|
| 13 |
+
"loss": 1.0008,
|
| 14 |
+
"step": 1000
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.16,
|
| 18 |
+
"learning_rate": 1.0863661053775122e-05,
|
| 19 |
+
"loss": 0.7988,
|
| 20 |
+
"step": 2000
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.24,
|
| 24 |
+
"learning_rate": 1.6295491580662687e-05,
|
| 25 |
+
"loss": 0.7246,
|
| 26 |
+
"step": 3000
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.33,
|
| 30 |
+
"learning_rate": 1.9808052151868172e-05,
|
| 31 |
+
"loss": 0.6939,
|
| 32 |
+
"step": 4000
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.41,
|
| 36 |
+
"learning_rate": 1.9204442566547958e-05,
|
| 37 |
+
"loss": 0.6612,
|
| 38 |
+
"step": 5000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.49,
|
| 42 |
+
"learning_rate": 1.8600832981227744e-05,
|
| 43 |
+
"loss": 0.6373,
|
| 44 |
+
"step": 6000
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.57,
|
| 48 |
+
"learning_rate": 1.7997223395907527e-05,
|
| 49 |
+
"loss": 0.6202,
|
| 50 |
+
"step": 7000
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.65,
|
| 54 |
+
"learning_rate": 1.7393613810587313e-05,
|
| 55 |
+
"loss": 0.6067,
|
| 56 |
+
"step": 8000
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.73,
|
| 60 |
+
"learning_rate": 1.67900042252671e-05,
|
| 61 |
+
"loss": 0.5993,
|
| 62 |
+
"step": 9000
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.81,
|
| 66 |
+
"learning_rate": 1.618639463994688e-05,
|
| 67 |
+
"loss": 0.5762,
|
| 68 |
+
"step": 10000
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.9,
|
| 72 |
+
"learning_rate": 1.5582785054626667e-05,
|
| 73 |
+
"loss": 0.5743,
|
| 74 |
+
"step": 11000
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.98,
|
| 78 |
+
"learning_rate": 1.4979175469306453e-05,
|
| 79 |
+
"loss": 0.5693,
|
| 80 |
+
"step": 12000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 1.0,
|
| 84 |
+
"eval_loss": 0.5552455186843872,
|
| 85 |
+
"eval_runtime": 2.5925,
|
| 86 |
+
"eval_samples_per_second": 960.462,
|
| 87 |
+
"eval_steps_per_second": 30.087,
|
| 88 |
+
"step": 12272
|
| 89 |
+
},
|
| 90 |
+
{
|
| 91 |
+
"epoch": 1.06,
|
| 92 |
+
"learning_rate": 1.437556588398624e-05,
|
| 93 |
+
"loss": 0.5115,
|
| 94 |
+
"step": 13000
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 1.14,
|
| 98 |
+
"learning_rate": 1.3771956298666025e-05,
|
| 99 |
+
"loss": 0.4904,
|
| 100 |
+
"step": 14000
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 1.22,
|
| 104 |
+
"learning_rate": 1.316834671334581e-05,
|
| 105 |
+
"loss": 0.4926,
|
| 106 |
+
"step": 15000
|
| 107 |
+
},
|
| 108 |
+
{
|
| 109 |
+
"epoch": 1.3,
|
| 110 |
+
"learning_rate": 1.2564737128025594e-05,
|
| 111 |
+
"loss": 0.4882,
|
| 112 |
+
"step": 16000
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 1.39,
|
| 116 |
+
"learning_rate": 1.196112754270538e-05,
|
| 117 |
+
"loss": 0.4862,
|
| 118 |
+
"step": 17000
|
| 119 |
+
},
|
| 120 |
+
{
|
| 121 |
+
"epoch": 1.47,
|
| 122 |
+
"learning_rate": 1.1357517957385164e-05,
|
| 123 |
+
"loss": 0.4829,
|
| 124 |
+
"step": 18000
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 1.55,
|
| 128 |
+
"learning_rate": 1.0753908372064949e-05,
|
| 129 |
+
"loss": 0.4797,
|
| 130 |
+
"step": 19000
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 1.63,
|
| 134 |
+
"learning_rate": 1.0150298786744735e-05,
|
| 135 |
+
"loss": 0.4749,
|
| 136 |
+
"step": 20000
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 1.71,
|
| 140 |
+
"learning_rate": 9.546689201424519e-06,
|
| 141 |
+
"loss": 0.4649,
|
| 142 |
+
"step": 21000
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.79,
|
| 146 |
+
"learning_rate": 8.943079616104305e-06,
|
| 147 |
+
"loss": 0.4638,
|
| 148 |
+
"step": 22000
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 1.87,
|
| 152 |
+
"learning_rate": 8.33947003078409e-06,
|
| 153 |
+
"loss": 0.4629,
|
| 154 |
+
"step": 23000
|
| 155 |
+
},
|
| 156 |
+
{
|
| 157 |
+
"epoch": 1.96,
|
| 158 |
+
"learning_rate": 7.735860445463874e-06,
|
| 159 |
+
"loss": 0.4666,
|
| 160 |
+
"step": 24000
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 2.0,
|
| 164 |
+
"eval_loss": 0.4878082573413849,
|
| 165 |
+
"eval_runtime": 2.5611,
|
| 166 |
+
"eval_samples_per_second": 972.256,
|
| 167 |
+
"eval_steps_per_second": 30.456,
|
| 168 |
+
"step": 24544
|
| 169 |
+
},
|
| 170 |
+
{
|
| 171 |
+
"epoch": 2.04,
|
| 172 |
+
"learning_rate": 7.1322508601436605e-06,
|
| 173 |
+
"loss": 0.4189,
|
| 174 |
+
"step": 25000
|
| 175 |
+
},
|
| 176 |
+
{
|
| 177 |
+
"epoch": 2.12,
|
| 178 |
+
"learning_rate": 6.528641274823445e-06,
|
| 179 |
+
"loss": 0.3709,
|
| 180 |
+
"step": 26000
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 2.2,
|
| 184 |
+
"learning_rate": 5.92503168950323e-06,
|
| 185 |
+
"loss": 0.3749,
|
| 186 |
+
"step": 27000
|
| 187 |
+
},
|
| 188 |
+
{
|
| 189 |
+
"epoch": 2.28,
|
| 190 |
+
"learning_rate": 5.321422104183015e-06,
|
| 191 |
+
"loss": 0.3729,
|
| 192 |
+
"step": 28000
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 2.36,
|
| 196 |
+
"learning_rate": 4.7178125188628e-06,
|
| 197 |
+
"loss": 0.3657,
|
| 198 |
+
"step": 29000
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 2.44,
|
| 202 |
+
"learning_rate": 4.114202933542585e-06,
|
| 203 |
+
"loss": 0.3694,
|
| 204 |
+
"step": 30000
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 2.53,
|
| 208 |
+
"learning_rate": 3.51059334822237e-06,
|
| 209 |
+
"loss": 0.3715,
|
| 210 |
+
"step": 31000
|
| 211 |
+
},
|
| 212 |
+
{
|
| 213 |
+
"epoch": 2.61,
|
| 214 |
+
"learning_rate": 2.9069837629021554e-06,
|
| 215 |
+
"loss": 0.3666,
|
| 216 |
+
"step": 32000
|
| 217 |
+
},
|
| 218 |
+
{
|
| 219 |
+
"epoch": 2.69,
|
| 220 |
+
"learning_rate": 2.30337417758194e-06,
|
| 221 |
+
"loss": 0.3672,
|
| 222 |
+
"step": 33000
|
| 223 |
+
},
|
| 224 |
+
{
|
| 225 |
+
"epoch": 2.77,
|
| 226 |
+
"learning_rate": 1.6997645922617255e-06,
|
| 227 |
+
"loss": 0.3717,
|
| 228 |
+
"step": 34000
|
| 229 |
+
},
|
| 230 |
+
{
|
| 231 |
+
"epoch": 2.85,
|
| 232 |
+
"learning_rate": 1.0961550069415102e-06,
|
| 233 |
+
"loss": 0.3555,
|
| 234 |
+
"step": 35000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 2.93,
|
| 238 |
+
"learning_rate": 4.925454216212955e-07,
|
| 239 |
+
"loss": 0.355,
|
| 240 |
+
"step": 36000
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 3.0,
|
| 244 |
+
"eval_loss": 0.5086637735366821,
|
| 245 |
+
"eval_runtime": 2.5595,
|
| 246 |
+
"eval_samples_per_second": 972.842,
|
| 247 |
+
"eval_steps_per_second": 30.475,
|
| 248 |
+
"step": 36816
|
| 249 |
+
}
|
| 250 |
+
],
|
| 251 |
+
"max_steps": 36816,
|
| 252 |
+
"num_train_epochs": 3,
|
| 253 |
+
"total_flos": 7.749426953899213e+16,
|
| 254 |
+
"trial_name": null,
|
| 255 |
+
"trial_params": null
|
| 256 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c69ac141c8f34349d8fec9ea05bd9623303d6c3905fd5e465d5298dde1cfd70f
|
| 3 |
+
size 3439
|