Delete checkpoints-v6.0-discrete
Browse files- checkpoints-v6.0-discrete/checkpoint-21504/manifest.md +0 -10
- checkpoints-v6.0-discrete/checkpoint-21504/model.safetensors +0 -3
- checkpoints-v6.0-discrete/checkpoint-21504/optimizer.pt +0 -3
- checkpoints-v6.0-discrete/checkpoint-21504/rng_state.pth +0 -3
- checkpoints-v6.0-discrete/checkpoint-21504/scaler.pt +0 -3
- checkpoints-v6.0-discrete/checkpoint-21504/scheduler.pt +0 -3
- checkpoints-v6.0-discrete/checkpoint-21504/trainer_state.json +0 -979
- checkpoints-v6.0-discrete/checkpoint-21504/training_args.bin +0 -3
checkpoints-v6.0-discrete/checkpoint-21504/manifest.md
DELETED
|
@@ -1,10 +0,0 @@
|
|
| 1 |
-
## checkpoint-v6.0-discrete (AE - checkpoint-v2.0)
|
| 2 |
-
|
| 3 |
-
# Changes
|
| 4 |
-
```
|
| 5 |
-
# Compute latents
|
| 6 |
-
x1 = self.encoder(input_ids)
|
| 7 |
-
|
| 8 |
-
# Noise
|
| 9 |
-
x0 = torch.randn_like(x1)
|
| 10 |
-
```
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
checkpoints-v6.0-discrete/checkpoint-21504/model.safetensors
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:9501ce0bf8e29d3feb1f4017200188ae444e82eccd472f5e5bfd7f8e48430c4b
|
| 3 |
-
size 24033312
|
|
|
|
|
|
|
|
|
|
|
|
checkpoints-v6.0-discrete/checkpoint-21504/optimizer.pt
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:142bfbb0216417d9bd2230bde64832ba6b0ec65c21fe0f512ee6dc9c80edbebc
|
| 3 |
-
size 1641995
|
|
|
|
|
|
|
|
|
|
|
|
checkpoints-v6.0-discrete/checkpoint-21504/rng_state.pth
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:0411746dbddd96f831447eb37448633a29927da0077d8ee1f63f8f45af4ee83f
|
| 3 |
-
size 14645
|
|
|
|
|
|
|
|
|
|
|
|
checkpoints-v6.0-discrete/checkpoint-21504/scaler.pt
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:7365d240cb075aef4a0a430755ab52030d11068817832fdd479f5b390eb81c10
|
| 3 |
-
size 1383
|
|
|
|
|
|
|
|
|
|
|
|
checkpoints-v6.0-discrete/checkpoint-21504/scheduler.pt
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:c8d0f20f63f45b1b1775e9f97d9c6a1a2904fa04b7d5a41f30f11b9aef0479eb
|
| 3 |
-
size 1465
|
|
|
|
|
|
|
|
|
|
|
|
checkpoints-v6.0-discrete/checkpoint-21504/trainer_state.json
DELETED
|
@@ -1,979 +0,0 @@
|
|
| 1 |
-
{
|
| 2 |
-
"best_global_step": null,
|
| 3 |
-
"best_metric": null,
|
| 4 |
-
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.9932104752667313,
|
| 6 |
-
"eval_steps": 1024,
|
| 7 |
-
"global_step": 21504,
|
| 8 |
-
"is_hyper_param_search": false,
|
| 9 |
-
"is_local_process_zero": true,
|
| 10 |
-
"is_world_process_zero": true,
|
| 11 |
-
"log_history": [
|
| 12 |
-
{
|
| 13 |
-
"epoch": 0.011823934229365849,
|
| 14 |
-
"grad_norm": 0.1075989231467247,
|
| 15 |
-
"learning_rate": 0.000498046875,
|
| 16 |
-
"loss": 2.1448066234588623,
|
| 17 |
-
"step": 256
|
| 18 |
-
},
|
| 19 |
-
{
|
| 20 |
-
"epoch": 0.023647868458731697,
|
| 21 |
-
"grad_norm": 0.44961342215538025,
|
| 22 |
-
"learning_rate": 0.000998046875,
|
| 23 |
-
"loss": 1.5786627531051636,
|
| 24 |
-
"step": 512
|
| 25 |
-
},
|
| 26 |
-
{
|
| 27 |
-
"epoch": 0.03547180268809755,
|
| 28 |
-
"grad_norm": 0.5772096514701843,
|
| 29 |
-
"learning_rate": 0.000999640996023194,
|
| 30 |
-
"loss": 1.41376531124115,
|
| 31 |
-
"step": 768
|
| 32 |
-
},
|
| 33 |
-
{
|
| 34 |
-
"epoch": 0.047295736917463395,
|
| 35 |
-
"grad_norm": 0.7530072927474976,
|
| 36 |
-
"learning_rate": 0.0009985588674043958,
|
| 37 |
-
"loss": 1.3673983812332153,
|
| 38 |
-
"step": 1024
|
| 39 |
-
},
|
| 40 |
-
{
|
| 41 |
-
"epoch": 0.047295736917463395,
|
| 42 |
-
"eval_cos_loss": 0.45404665485092494,
|
| 43 |
-
"eval_loss": 1.3462410258920225,
|
| 44 |
-
"eval_mse_loss": 1.3008363587126883,
|
| 45 |
-
"step": 1024
|
| 46 |
-
},
|
| 47 |
-
{
|
| 48 |
-
"epoch": 0.047295736917463395,
|
| 49 |
-
"eval_cos_loss": 0.45404665485092494,
|
| 50 |
-
"eval_loss": 1.3462410258920225,
|
| 51 |
-
"eval_mse_loss": 1.3008363587126883,
|
| 52 |
-
"eval_runtime": 22.2392,
|
| 53 |
-
"eval_samples_per_second": 1258.724,
|
| 54 |
-
"eval_steps_per_second": 19.695,
|
| 55 |
-
"step": 1024
|
| 56 |
-
},
|
| 57 |
-
{
|
| 58 |
-
"epoch": 0.05911967114682925,
|
| 59 |
-
"grad_norm": 1.1749802827835083,
|
| 60 |
-
"learning_rate": 0.0009967551747861387,
|
| 61 |
-
"loss": 1.3432152271270752,
|
| 62 |
-
"step": 1280
|
| 63 |
-
},
|
| 64 |
-
{
|
| 65 |
-
"epoch": 0.0709436053761951,
|
| 66 |
-
"grad_norm": 1.0016200542449951,
|
| 67 |
-
"learning_rate": 0.000994232528651847,
|
| 68 |
-
"loss": 1.3247753381729126,
|
| 69 |
-
"step": 1536
|
| 70 |
-
},
|
| 71 |
-
{
|
| 72 |
-
"epoch": 0.08276753960556095,
|
| 73 |
-
"grad_norm": 0.7005159258842468,
|
| 74 |
-
"learning_rate": 0.0009909945800260092,
|
| 75 |
-
"loss": 1.3099638223648071,
|
| 76 |
-
"step": 1792
|
| 77 |
-
},
|
| 78 |
-
{
|
| 79 |
-
"epoch": 0.09459147383492679,
|
| 80 |
-
"grad_norm": 1.0100877285003662,
|
| 81 |
-
"learning_rate": 0.0009870460151900522,
|
| 82 |
-
"loss": 1.305666208267212,
|
| 83 |
-
"step": 2048
|
| 84 |
-
},
|
| 85 |
-
{
|
| 86 |
-
"epoch": 0.09459147383492679,
|
| 87 |
-
"eval_cos_loss": 0.4357716066924404,
|
| 88 |
-
"eval_loss": 1.2961699818367283,
|
| 89 |
-
"eval_mse_loss": 1.2525928232767811,
|
| 90 |
-
"step": 2048
|
| 91 |
-
},
|
| 92 |
-
{
|
| 93 |
-
"epoch": 0.09459147383492679,
|
| 94 |
-
"eval_cos_loss": 0.4357716066924404,
|
| 95 |
-
"eval_loss": 1.2961699818367283,
|
| 96 |
-
"eval_mse_loss": 1.2525928232767811,
|
| 97 |
-
"eval_runtime": 22.5703,
|
| 98 |
-
"eval_samples_per_second": 1240.257,
|
| 99 |
-
"eval_steps_per_second": 19.406,
|
| 100 |
-
"step": 2048
|
| 101 |
-
},
|
| 102 |
-
{
|
| 103 |
-
"epoch": 0.10641540806429264,
|
| 104 |
-
"grad_norm": 1.0679256916046143,
|
| 105 |
-
"learning_rate": 0.0009823925488998885,
|
| 106 |
-
"loss": 1.2951494455337524,
|
| 107 |
-
"step": 2304
|
| 108 |
-
},
|
| 109 |
-
{
|
| 110 |
-
"epoch": 0.1182393422936585,
|
| 111 |
-
"grad_norm": 1.8469953536987305,
|
| 112 |
-
"learning_rate": 0.0009770409161149525,
|
| 113 |
-
"loss": 1.2876818180084229,
|
| 114 |
-
"step": 2560
|
| 115 |
-
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.13006327652302435,
|
| 118 |
-
"grad_norm": 0.6149023771286011,
|
| 119 |
-
"learning_rate": 0.0009709988622506973,
|
| 120 |
-
"loss": 1.2823055982589722,
|
| 121 |
-
"step": 2816
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 0.1418872107523902,
|
| 125 |
-
"grad_norm": 0.6905927062034607,
|
| 126 |
-
"learning_rate": 0.000964275131968659,
|
| 127 |
-
"loss": 1.2750741243362427,
|
| 128 |
-
"step": 3072
|
| 129 |
-
},
|
| 130 |
-
{
|
| 131 |
-
"epoch": 0.1418872107523902,
|
| 132 |
-
"eval_cos_loss": 0.42528551592402264,
|
| 133 |
-
"eval_loss": 1.2683077813827828,
|
| 134 |
-
"eval_mse_loss": 1.2257792293753254,
|
| 135 |
-
"step": 3072
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 0.1418872107523902,
|
| 139 |
-
"eval_cos_loss": 0.42528551592402264,
|
| 140 |
-
"eval_loss": 1.2683077813827828,
|
| 141 |
-
"eval_mse_loss": 1.2257792293753254,
|
| 142 |
-
"eval_runtime": 23.6166,
|
| 143 |
-
"eval_samples_per_second": 1185.312,
|
| 144 |
-
"eval_steps_per_second": 18.546,
|
| 145 |
-
"step": 3072
|
| 146 |
-
},
|
| 147 |
-
{
|
| 148 |
-
"epoch": 0.15371114498175603,
|
| 149 |
-
"grad_norm": 0.8839517831802368,
|
| 150 |
-
"learning_rate": 0.0009568794565203123,
|
| 151 |
-
"loss": 1.2709108591079712,
|
| 152 |
-
"step": 3328
|
| 153 |
-
},
|
| 154 |
-
{
|
| 155 |
-
"epoch": 0.1655350792111219,
|
| 156 |
-
"grad_norm": 1.089804768562317,
|
| 157 |
-
"learning_rate": 0.0009488225396630347,
|
| 158 |
-
"loss": 1.2688775062561035,
|
| 159 |
-
"step": 3584
|
| 160 |
-
},
|
| 161 |
-
{
|
| 162 |
-
"epoch": 0.17735901344048774,
|
| 163 |
-
"grad_norm": 0.6663815379142761,
|
| 164 |
-
"learning_rate": 0.0009401160421685646,
|
| 165 |
-
"loss": 1.2621169090270996,
|
| 166 |
-
"step": 3840
|
| 167 |
-
},
|
| 168 |
-
{
|
| 169 |
-
"epoch": 0.18918294766985358,
|
| 170 |
-
"grad_norm": 1.0516070127487183,
|
| 171 |
-
"learning_rate": 0.0009307725649463714,
|
| 172 |
-
"loss": 1.2586530447006226,
|
| 173 |
-
"step": 4096
|
| 174 |
-
},
|
| 175 |
-
{
|
| 176 |
-
"epoch": 0.18918294766985358,
|
| 177 |
-
"eval_cos_loss": 0.4208076707848675,
|
| 178 |
-
"eval_loss": 1.256359337127372,
|
| 179 |
-
"eval_mse_loss": 1.2142785700488852,
|
| 180 |
-
"step": 4096
|
| 181 |
-
},
|
| 182 |
-
{
|
| 183 |
-
"epoch": 0.18918294766985358,
|
| 184 |
-
"eval_cos_loss": 0.4208076707848675,
|
| 185 |
-
"eval_loss": 1.256359337127372,
|
| 186 |
-
"eval_mse_loss": 1.2142785700488852,
|
| 187 |
-
"eval_runtime": 22.2012,
|
| 188 |
-
"eval_samples_per_second": 1260.877,
|
| 189 |
-
"eval_steps_per_second": 19.729,
|
| 190 |
-
"step": 4096
|
| 191 |
-
},
|
| 192 |
-
{
|
| 193 |
-
"epoch": 0.20100688189921945,
|
| 194 |
-
"grad_norm": 1.3545875549316406,
|
| 195 |
-
"learning_rate": 0.0009208056308063659,
|
| 196 |
-
"loss": 1.253294587135315,
|
| 197 |
-
"step": 4352
|
| 198 |
-
},
|
| 199 |
-
{
|
| 200 |
-
"epoch": 0.2128308161285853,
|
| 201 |
-
"grad_norm": 1.4382582902908325,
|
| 202 |
-
"learning_rate": 0.0009102296648873445,
|
| 203 |
-
"loss": 1.254496455192566,
|
| 204 |
-
"step": 4608
|
| 205 |
-
},
|
| 206 |
-
{
|
| 207 |
-
"epoch": 0.22465475035795113,
|
| 208 |
-
"grad_norm": 0.9149227738380432,
|
| 209 |
-
"learning_rate": 0.0008990599737794927,
|
| 210 |
-
"loss": 1.2505712509155273,
|
| 211 |
-
"step": 4864
|
| 212 |
-
},
|
| 213 |
-
{
|
| 214 |
-
"epoch": 0.236478684587317,
|
| 215 |
-
"grad_norm": 1.2436659336090088,
|
| 216 |
-
"learning_rate": 0.0008873127233711644,
|
| 217 |
-
"loss": 1.2458906173706055,
|
| 218 |
-
"step": 5120
|
| 219 |
-
},
|
| 220 |
-
{
|
| 221 |
-
"epoch": 0.236478684587317,
|
| 222 |
-
"eval_cos_loss": 0.41453840518923113,
|
| 223 |
-
"eval_loss": 1.2386129865907642,
|
| 224 |
-
"eval_mse_loss": 1.1971591430711963,
|
| 225 |
-
"step": 5120
|
| 226 |
-
},
|
| 227 |
-
{
|
| 228 |
-
"epoch": 0.236478684587317,
|
| 229 |
-
"eval_cos_loss": 0.41453840518923113,
|
| 230 |
-
"eval_loss": 1.2386129865907642,
|
| 231 |
-
"eval_mse_loss": 1.1971591430711963,
|
| 232 |
-
"eval_runtime": 22.3663,
|
| 233 |
-
"eval_samples_per_second": 1251.571,
|
| 234 |
-
"eval_steps_per_second": 19.583,
|
| 235 |
-
"step": 5120
|
| 236 |
-
},
|
| 237 |
-
{
|
| 238 |
-
"epoch": 0.24830261881668284,
|
| 239 |
-
"grad_norm": 1.1406697034835815,
|
| 240 |
-
"learning_rate": 0.0008750049154520011,
|
| 241 |
-
"loss": 1.2400823831558228,
|
| 242 |
-
"step": 5376
|
| 243 |
-
},
|
| 244 |
-
{
|
| 245 |
-
"epoch": 0.2601265530460487,
|
| 246 |
-
"grad_norm": 1.2730367183685303,
|
| 247 |
-
"learning_rate": 0.0008621543631062487,
|
| 248 |
-
"loss": 1.2361854314804077,
|
| 249 |
-
"step": 5632
|
| 250 |
-
},
|
| 251 |
-
{
|
| 252 |
-
"epoch": 0.27195048727541454,
|
| 253 |
-
"grad_norm": 0.91032475233078,
|
| 254 |
-
"learning_rate": 0.0008487796649318904,
|
| 255 |
-
"loss": 1.2351124286651611,
|
| 256 |
-
"step": 5888
|
| 257 |
-
},
|
| 258 |
-
{
|
| 259 |
-
"epoch": 0.2837744215047804,
|
| 260 |
-
"grad_norm": 1.5020276308059692,
|
| 261 |
-
"learning_rate": 0.0008349001781229053,
|
| 262 |
-
"loss": 1.232508897781372,
|
| 263 |
-
"step": 6144
|
| 264 |
-
},
|
| 265 |
-
{
|
| 266 |
-
"epoch": 0.2837744215047804,
|
| 267 |
-
"eval_cos_loss": 0.41207376571550763,
|
| 268 |
-
"eval_loss": 1.2308254878814906,
|
| 269 |
-
"eval_mse_loss": 1.1896181109289057,
|
| 270 |
-
"step": 6144
|
| 271 |
-
},
|
| 272 |
-
{
|
| 273 |
-
"epoch": 0.2837744215047804,
|
| 274 |
-
"eval_cos_loss": 0.41207376571550763,
|
| 275 |
-
"eval_loss": 1.2308254878814906,
|
| 276 |
-
"eval_mse_loss": 1.1896181109289057,
|
| 277 |
-
"eval_runtime": 22.9215,
|
| 278 |
-
"eval_samples_per_second": 1221.257,
|
| 279 |
-
"eval_steps_per_second": 19.109,
|
| 280 |
-
"step": 6144
|
| 281 |
-
},
|
| 282 |
-
{
|
| 283 |
-
"epoch": 0.2955983557341462,
|
| 284 |
-
"grad_norm": 0.9625284671783447,
|
| 285 |
-
"learning_rate": 0.0008205359904536107,
|
| 286 |
-
"loss": 1.2305837869644165,
|
| 287 |
-
"step": 6400
|
| 288 |
-
},
|
| 289 |
-
{
|
| 290 |
-
"epoch": 0.30742228996351206,
|
| 291 |
-
"grad_norm": 1.1228785514831543,
|
| 292 |
-
"learning_rate": 0.0008057078912056363,
|
| 293 |
-
"loss": 1.226149320602417,
|
| 294 |
-
"step": 6656
|
| 295 |
-
},
|
| 296 |
-
{
|
| 297 |
-
"epoch": 0.3192462241928779,
|
| 298 |
-
"grad_norm": 0.9577246904373169,
|
| 299 |
-
"learning_rate": 0.0007904373410796086,
|
| 300 |
-
"loss": 1.2268048524856567,
|
| 301 |
-
"step": 6912
|
| 302 |
-
},
|
| 303 |
-
{
|
| 304 |
-
"epoch": 0.3310701584222438,
|
| 305 |
-
"grad_norm": 0.956436812877655,
|
| 306 |
-
"learning_rate": 0.0007747464411350876,
|
| 307 |
-
"loss": 1.2231967449188232,
|
| 308 |
-
"step": 7168
|
| 309 |
-
},
|
| 310 |
-
{
|
| 311 |
-
"epoch": 0.3310701584222438,
|
| 312 |
-
"eval_cos_loss": 0.40833413709788563,
|
| 313 |
-
"eval_loss": 1.2216057725692993,
|
| 314 |
-
"eval_mse_loss": 1.180772360057047,
|
| 315 |
-
"step": 7168
|
| 316 |
-
},
|
| 317 |
-
{
|
| 318 |
-
"epoch": 0.3310701584222438,
|
| 319 |
-
"eval_cos_loss": 0.40833413709788563,
|
| 320 |
-
"eval_loss": 1.2216057725692993,
|
| 321 |
-
"eval_mse_loss": 1.180772360057047,
|
| 322 |
-
"eval_runtime": 22.6856,
|
| 323 |
-
"eval_samples_per_second": 1233.956,
|
| 324 |
-
"eval_steps_per_second": 19.307,
|
| 325 |
-
"step": 7168
|
| 326 |
-
},
|
| 327 |
-
{
|
| 328 |
-
"epoch": 0.34289409265160964,
|
| 329 |
-
"grad_norm": 1.1598716974258423,
|
| 330 |
-
"learning_rate": 0.000758657900803716,
|
| 331 |
-
"loss": 1.221030831336975,
|
| 332 |
-
"step": 7424
|
| 333 |
-
},
|
| 334 |
-
{
|
| 335 |
-
"epoch": 0.3547180268809755,
|
| 336 |
-
"grad_norm": 2.4167025089263916,
|
| 337 |
-
"learning_rate": 0.000742195005021869,
|
| 338 |
-
"loss": 1.2203080654144287,
|
| 339 |
-
"step": 7680
|
| 340 |
-
},
|
| 341 |
-
{
|
| 342 |
-
"epoch": 0.3665419611103413,
|
| 343 |
-
"grad_norm": 0.8216366171836853,
|
| 344 |
-
"learning_rate": 0.0007253815805303786,
|
| 345 |
-
"loss": 1.2149429321289062,
|
| 346 |
-
"step": 7936
|
| 347 |
-
},
|
| 348 |
-
{
|
| 349 |
-
"epoch": 0.37836589533970716,
|
| 350 |
-
"grad_norm": 0.9276426434516907,
|
| 351 |
-
"learning_rate": 0.0007082419613901028,
|
| 352 |
-
"loss": 1.2138267755508423,
|
| 353 |
-
"step": 8192
|
| 354 |
-
},
|
| 355 |
-
{
|
| 356 |
-
"epoch": 0.37836589533970716,
|
| 357 |
-
"eval_cos_loss": 0.40511899138694485,
|
| 358 |
-
"eval_loss": 1.212123587523421,
|
| 359 |
-
"eval_mse_loss": 1.171611689813605,
|
| 360 |
-
"step": 8192
|
| 361 |
-
},
|
| 362 |
-
{
|
| 363 |
-
"epoch": 0.37836589533970716,
|
| 364 |
-
"eval_cos_loss": 0.40511899138694485,
|
| 365 |
-
"eval_loss": 1.212123587523421,
|
| 366 |
-
"eval_mse_loss": 1.171611689813605,
|
| 367 |
-
"eval_runtime": 22.7982,
|
| 368 |
-
"eval_samples_per_second": 1227.858,
|
| 369 |
-
"eval_steps_per_second": 19.212,
|
| 370 |
-
"step": 8192
|
| 371 |
-
},
|
| 372 |
-
{
|
| 373 |
-
"epoch": 0.390189829569073,
|
| 374 |
-
"grad_norm": 1.1356655359268188,
|
| 375 |
-
"learning_rate": 0.0006908009537632514,
|
| 376 |
-
"loss": 1.2129006385803223,
|
| 377 |
-
"step": 8448
|
| 378 |
-
},
|
| 379 |
-
{
|
| 380 |
-
"epoch": 0.4020137637984389,
|
| 381 |
-
"grad_norm": 1.165170669555664,
|
| 382 |
-
"learning_rate": 0.0006730838000114403,
|
| 383 |
-
"loss": 1.2100634574890137,
|
| 384 |
-
"step": 8704
|
| 385 |
-
},
|
| 386 |
-
{
|
| 387 |
-
"epoch": 0.41383769802780473,
|
| 388 |
-
"grad_norm": 1.0138978958129883,
|
| 389 |
-
"learning_rate": 0.0006551161421624341,
|
| 390 |
-
"loss": 1.2095133066177368,
|
| 391 |
-
"step": 8960
|
| 392 |
-
},
|
| 393 |
-
{
|
| 394 |
-
"epoch": 0.4256616322571706,
|
| 395 |
-
"grad_norm": 1.826959490776062,
|
| 396 |
-
"learning_rate": 0.0006369239847984517,
|
| 397 |
-
"loss": 1.206162929534912,
|
| 398 |
-
"step": 9216
|
| 399 |
-
},
|
| 400 |
-
{
|
| 401 |
-
"epoch": 0.4256616322571706,
|
| 402 |
-
"eval_cos_loss": 0.4033830788445799,
|
| 403 |
-
"eval_loss": 1.2079350161770164,
|
| 404 |
-
"eval_mse_loss": 1.1675967067888338,
|
| 405 |
-
"step": 9216
|
| 406 |
-
},
|
| 407 |
-
{
|
| 408 |
-
"epoch": 0.4256616322571706,
|
| 409 |
-
"eval_cos_loss": 0.4033830788445799,
|
| 410 |
-
"eval_loss": 1.2079350161770164,
|
| 411 |
-
"eval_mse_loss": 1.1675967067888338,
|
| 412 |
-
"eval_runtime": 22.7338,
|
| 413 |
-
"eval_samples_per_second": 1231.339,
|
| 414 |
-
"eval_steps_per_second": 19.266,
|
| 415 |
-
"step": 9216
|
| 416 |
-
},
|
| 417 |
-
{
|
| 418 |
-
"epoch": 0.4374855664865364,
|
| 419 |
-
"grad_norm": 2.6399316787719727,
|
| 420 |
-
"learning_rate": 0.0006185336574197479,
|
| 421 |
-
"loss": 1.2027764320373535,
|
| 422 |
-
"step": 9472
|
| 423 |
-
},
|
| 424 |
-
{
|
| 425 |
-
"epoch": 0.44930950071590225,
|
| 426 |
-
"grad_norm": 1.1962329149246216,
|
| 427 |
-
"learning_rate": 0.0005999717763379407,
|
| 428 |
-
"loss": 1.2066611051559448,
|
| 429 |
-
"step": 9728
|
| 430 |
-
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 0.4611334349452681,
|
| 433 |
-
"grad_norm": 1.6900142431259155,
|
| 434 |
-
"learning_rate": 0.0005812652061542363,
|
| 435 |
-
"loss": 1.2036194801330566,
|
| 436 |
-
"step": 9984
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 0.472957369174634,
|
| 440 |
-
"grad_norm": 1.740152359008789,
|
| 441 |
-
"learning_rate": 0.0005624410208783071,
|
| 442 |
-
"loss": 1.2026172876358032,
|
| 443 |
-
"step": 10240
|
| 444 |
-
},
|
| 445 |
-
{
|
| 446 |
-
"epoch": 0.472957369174634,
|
| 447 |
-
"eval_cos_loss": 0.4016102112863706,
|
| 448 |
-
"eval_loss": 1.202477820387714,
|
| 449 |
-
"eval_mse_loss": 1.1623167983473164,
|
| 450 |
-
"step": 10240
|
| 451 |
-
},
|
| 452 |
-
{
|
| 453 |
-
"epoch": 0.472957369174634,
|
| 454 |
-
"eval_cos_loss": 0.4016102112863706,
|
| 455 |
-
"eval_loss": 1.202477820387714,
|
| 456 |
-
"eval_mse_loss": 1.1623167983473164,
|
| 457 |
-
"eval_runtime": 23.2606,
|
| 458 |
-
"eval_samples_per_second": 1203.45,
|
| 459 |
-
"eval_steps_per_second": 18.83,
|
| 460 |
-
"step": 10240
|
| 461 |
-
},
|
| 462 |
-
{
|
| 463 |
-
"epoch": 0.48478130340399983,
|
| 464 |
-
"grad_norm": 1.9587527513504028,
|
| 465 |
-
"learning_rate": 0.0005435264647440881,
|
| 466 |
-
"loss": 1.2006222009658813,
|
| 467 |
-
"step": 10496
|
| 468 |
-
},
|
| 469 |
-
{
|
| 470 |
-
"epoch": 0.49660523763336567,
|
| 471 |
-
"grad_norm": 1.8280595541000366,
|
| 472 |
-
"learning_rate": 0.000524548912779213,
|
| 473 |
-
"loss": 1.1970598697662354,
|
| 474 |
-
"step": 10752
|
| 475 |
-
},
|
| 476 |
-
{
|
| 477 |
-
"epoch": 0.5084291718627315,
|
| 478 |
-
"grad_norm": 1.2811309099197388,
|
| 479 |
-
"learning_rate": 0.0005055358311851499,
|
| 480 |
-
"loss": 1.1963506937026978,
|
| 481 |
-
"step": 11008
|
| 482 |
-
},
|
| 483 |
-
{
|
| 484 |
-
"epoch": 0.5202531060920974,
|
| 485 |
-
"grad_norm": 2.4028587341308594,
|
| 486 |
-
"learning_rate": 0.0004865147375853812,
|
| 487 |
-
"loss": 1.1964491605758667,
|
| 488 |
-
"step": 11264
|
| 489 |
-
},
|
| 490 |
-
{
|
| 491 |
-
"epoch": 0.5202531060920974,
|
| 492 |
-
"eval_cos_loss": 0.4001519478619371,
|
| 493 |
-
"eval_loss": 1.199085911387178,
|
| 494 |
-
"eval_mse_loss": 1.1590707160566496,
|
| 495 |
-
"step": 11264
|
| 496 |
-
},
|
| 497 |
-
{
|
| 498 |
-
"epoch": 0.5202531060920974,
|
| 499 |
-
"eval_cos_loss": 0.4001519478619371,
|
| 500 |
-
"eval_loss": 1.199085911387178,
|
| 501 |
-
"eval_mse_loss": 1.1590707160566496,
|
| 502 |
-
"eval_runtime": 22.1469,
|
| 503 |
-
"eval_samples_per_second": 1263.967,
|
| 504 |
-
"eval_steps_per_second": 19.777,
|
| 505 |
-
"step": 11264
|
| 506 |
-
},
|
| 507 |
-
{
|
| 508 |
-
"epoch": 0.5320770403214632,
|
| 509 |
-
"grad_norm": 1.3925809860229492,
|
| 510 |
-
"learning_rate": 0.0004675131611991607,
|
| 511 |
-
"loss": 1.1931047439575195,
|
| 512 |
-
"step": 11520
|
| 513 |
-
},
|
| 514 |
-
{
|
| 515 |
-
"epoch": 0.5439009745508291,
|
| 516 |
-
"grad_norm": 2.021632671356201,
|
| 517 |
-
"learning_rate": 0.0004485586029984899,
|
| 518 |
-
"loss": 1.1920230388641357,
|
| 519 |
-
"step": 11776
|
| 520 |
-
},
|
| 521 |
-
{
|
| 522 |
-
"epoch": 0.5557249087801949,
|
| 523 |
-
"grad_norm": 1.274520754814148,
|
| 524 |
-
"learning_rate": 0.00042967849590597266,
|
| 525 |
-
"loss": 1.1933492422103882,
|
| 526 |
-
"step": 12032
|
| 527 |
-
},
|
| 528 |
-
{
|
| 529 |
-
"epoch": 0.5675488430095608,
|
| 530 |
-
"grad_norm": 2.561692714691162,
|
| 531 |
-
"learning_rate": 0.0004109001650911621,
|
| 532 |
-
"loss": 1.1902129650115967,
|
| 533 |
-
"step": 12288
|
| 534 |
-
},
|
| 535 |
-
{
|
| 536 |
-
"epoch": 0.5675488430095608,
|
| 537 |
-
"eval_cos_loss": 0.39772830930746855,
|
| 538 |
-
"eval_loss": 1.1911358833312988,
|
| 539 |
-
"eval_mse_loss": 1.151363052733957,
|
| 540 |
-
"step": 12288
|
| 541 |
-
},
|
| 542 |
-
{
|
| 543 |
-
"epoch": 0.5675488430095608,
|
| 544 |
-
"eval_cos_loss": 0.39772830930746855,
|
| 545 |
-
"eval_loss": 1.1911358833312988,
|
| 546 |
-
"eval_mse_loss": 1.151363052733957,
|
| 547 |
-
"eval_runtime": 23.0679,
|
| 548 |
-
"eval_samples_per_second": 1213.503,
|
| 549 |
-
"eval_steps_per_second": 18.987,
|
| 550 |
-
"step": 12288
|
| 551 |
-
},
|
| 552 |
-
{
|
| 553 |
-
"epoch": 0.5793727772389267,
|
| 554 |
-
"grad_norm": 1.2727264165878296,
|
| 555 |
-
"learning_rate": 0.0003922507884228551,
|
| 556 |
-
"loss": 1.1917052268981934,
|
| 557 |
-
"step": 12544
|
| 558 |
-
},
|
| 559 |
-
{
|
| 560 |
-
"epoch": 0.5911967114682924,
|
| 561 |
-
"grad_norm": 1.0674834251403809,
|
| 562 |
-
"learning_rate": 0.00037375735713457723,
|
| 563 |
-
"loss": 1.1877367496490479,
|
| 564 |
-
"step": 12800
|
| 565 |
-
},
|
| 566 |
-
{
|
| 567 |
-
"epoch": 0.6030206456976583,
|
| 568 |
-
"grad_norm": 1.2175348997116089,
|
| 569 |
-
"learning_rate": 0.00035544663676018276,
|
| 570 |
-
"loss": 1.189236044883728,
|
| 571 |
-
"step": 13056
|
| 572 |
-
},
|
| 573 |
-
{
|
| 574 |
-
"epoch": 0.6148445799270241,
|
| 575 |
-
"grad_norm": 0.6856772303581238,
|
| 576 |
-
"learning_rate": 0.00033734512839611255,
|
| 577 |
-
"loss": 1.1856091022491455,
|
| 578 |
-
"step": 13312
|
| 579 |
-
},
|
| 580 |
-
{
|
| 581 |
-
"epoch": 0.6148445799270241,
|
| 582 |
-
"eval_cos_loss": 0.39564488399518677,
|
| 583 |
-
"eval_loss": 1.1844980667715204,
|
| 584 |
-
"eval_mse_loss": 1.144933577269724,
|
| 585 |
-
"step": 13312
|
| 586 |
-
},
|
| 587 |
-
{
|
| 588 |
-
"epoch": 0.6148445799270241,
|
| 589 |
-
"eval_cos_loss": 0.39564488399518677,
|
| 590 |
-
"eval_loss": 1.1844980667715204,
|
| 591 |
-
"eval_mse_loss": 1.144933577269724,
|
| 592 |
-
"eval_runtime": 23.3265,
|
| 593 |
-
"eval_samples_per_second": 1200.049,
|
| 594 |
-
"eval_steps_per_second": 18.777,
|
| 595 |
-
"step": 13312
|
| 596 |
-
},
|
| 597 |
-
{
|
| 598 |
-
"epoch": 0.62666851415639,
|
| 599 |
-
"grad_norm": 0.760578989982605,
|
| 600 |
-
"learning_rate": 0.0003194790303463687,
|
| 601 |
-
"loss": 1.1841545104980469,
|
| 602 |
-
"step": 13568
|
| 603 |
-
},
|
| 604 |
-
{
|
| 605 |
-
"epoch": 0.6384924483857558,
|
| 606 |
-
"grad_norm": 1.0841054916381836,
|
| 607 |
-
"learning_rate": 0.00030187420020572406,
|
| 608 |
-
"loss": 1.1827186346054077,
|
| 609 |
-
"step": 13824
|
| 610 |
-
},
|
| 611 |
-
{
|
| 612 |
-
"epoch": 0.6503163826151217,
|
| 613 |
-
"grad_norm": 1.048738956451416,
|
| 614 |
-
"learning_rate": 0.00028455611743603626,
|
| 615 |
-
"loss": 1.1802786588668823,
|
| 616 |
-
"step": 14080
|
| 617 |
-
},
|
| 618 |
-
{
|
| 619 |
-
"epoch": 0.6621403168444876,
|
| 620 |
-
"grad_norm": 1.1549500226974487,
|
| 621 |
-
"learning_rate": 0.0002675498464898373,
|
| 622 |
-
"loss": 1.1840592622756958,
|
| 623 |
-
"step": 14336
|
| 624 |
-
},
|
| 625 |
-
{
|
| 626 |
-
"epoch": 0.6621403168444876,
|
| 627 |
-
"eval_cos_loss": 0.3939972570905947,
|
| 628 |
-
"eval_loss": 1.180122692563218,
|
| 629 |
-
"eval_mse_loss": 1.1407229658131186,
|
| 630 |
-
"step": 14336
|
| 631 |
-
},
|
| 632 |
-
{
|
| 633 |
-
"epoch": 0.6621403168444876,
|
| 634 |
-
"eval_cos_loss": 0.3939972570905947,
|
| 635 |
-
"eval_loss": 1.180122692563218,
|
| 636 |
-
"eval_mse_loss": 1.1407229658131186,
|
| 637 |
-
"eval_runtime": 22.0393,
|
| 638 |
-
"eval_samples_per_second": 1270.139,
|
| 639 |
-
"eval_steps_per_second": 19.874,
|
| 640 |
-
"step": 14336
|
| 641 |
-
},
|
| 642 |
-
{
|
| 643 |
-
"epoch": 0.6739642510738534,
|
| 644 |
-
"grad_norm": 1.9604671001434326,
|
| 645 |
-
"learning_rate": 0.0002508800005345623,
|
| 646 |
-
"loss": 1.1857727766036987,
|
| 647 |
-
"step": 14592
|
| 648 |
-
},
|
| 649 |
-
{
|
| 650 |
-
"epoch": 0.6857881853032193,
|
| 651 |
-
"grad_norm": 0.8758797645568848,
|
| 652 |
-
"learning_rate": 0.00023457070582992562,
|
| 653 |
-
"loss": 1.185521125793457,
|
| 654 |
-
"step": 14848
|
| 655 |
-
},
|
| 656 |
-
{
|
| 657 |
-
"epoch": 0.6976121195325851,
|
| 658 |
-
"grad_norm": 0.44620171189308167,
|
| 659 |
-
"learning_rate": 0.00021864556680999692,
|
| 660 |
-
"loss": 1.1780372858047485,
|
| 661 |
-
"step": 15104
|
| 662 |
-
},
|
| 663 |
-
{
|
| 664 |
-
"epoch": 0.709436053761951,
|
| 665 |
-
"grad_norm": 1.1207635402679443,
|
| 666 |
-
"learning_rate": 0.0002031276319205152,
|
| 667 |
-
"loss": 1.1769145727157593,
|
| 668 |
-
"step": 15360
|
| 669 |
-
},
|
| 670 |
-
{
|
| 671 |
-
"epoch": 0.709436053761951,
|
| 672 |
-
"eval_cos_loss": 0.3943222426388362,
|
| 673 |
-
"eval_loss": 1.1810845996691213,
|
| 674 |
-
"eval_mse_loss": 1.14165237505142,
|
| 675 |
-
"step": 15360
|
| 676 |
-
},
|
| 677 |
-
{
|
| 678 |
-
"epoch": 0.709436053761951,
|
| 679 |
-
"eval_cos_loss": 0.3943222426388362,
|
| 680 |
-
"eval_loss": 1.1810845996691213,
|
| 681 |
-
"eval_mse_loss": 1.14165237505142,
|
| 682 |
-
"eval_runtime": 23.1251,
|
| 683 |
-
"eval_samples_per_second": 1210.504,
|
| 684 |
-
"eval_steps_per_second": 18.94,
|
| 685 |
-
"step": 15360
|
| 686 |
-
},
|
| 687 |
-
{
|
| 688 |
-
"epoch": 0.7212599879913169,
|
| 689 |
-
"grad_norm": 0.919623613357544,
|
| 690 |
-
"learning_rate": 0.00018803936026088542,
|
| 691 |
-
"loss": 1.177735447883606,
|
| 692 |
-
"step": 15616
|
| 693 |
-
},
|
| 694 |
-
{
|
| 695 |
-
"epoch": 0.7330839222206826,
|
| 696 |
-
"grad_norm": 0.9698303937911987,
|
| 697 |
-
"learning_rate": 0.00017340258907913464,
|
| 698 |
-
"loss": 1.181166648864746,
|
| 699 |
-
"step": 15872
|
| 700 |
-
},
|
| 701 |
-
{
|
| 702 |
-
"epoch": 0.7449078564500485,
|
| 703 |
-
"grad_norm": 1.3776527643203735,
|
| 704 |
-
"learning_rate": 0.0001592385021668743,
|
| 705 |
-
"loss": 1.1758204698562622,
|
| 706 |
-
"step": 16128
|
| 707 |
-
},
|
| 708 |
-
{
|
| 709 |
-
"epoch": 0.7567317906794143,
|
| 710 |
-
"grad_norm": 1.5056192874908447,
|
| 711 |
-
"learning_rate": 0.0001455675992000087,
|
| 712 |
-
"loss": 1.1798039674758911,
|
| 713 |
-
"step": 16384
|
| 714 |
-
},
|
| 715 |
-
{
|
| 716 |
-
"epoch": 0.7567317906794143,
|
| 717 |
-
"eval_cos_loss": 0.39351845338736496,
|
| 718 |
-
"eval_loss": 1.1777457766336938,
|
| 719 |
-
"eval_mse_loss": 1.1383939290155558,
|
| 720 |
-
"step": 16384
|
| 721 |
-
},
|
| 722 |
-
{
|
| 723 |
-
"epoch": 0.7567317906794143,
|
| 724 |
-
"eval_cos_loss": 0.39351845338736496,
|
| 725 |
-
"eval_loss": 1.1777457766336938,
|
| 726 |
-
"eval_mse_loss": 1.1383939290155558,
|
| 727 |
-
"eval_runtime": 22.833,
|
| 728 |
-
"eval_samples_per_second": 1225.988,
|
| 729 |
-
"eval_steps_per_second": 19.183,
|
| 730 |
-
"step": 16384
|
| 731 |
-
},
|
| 732 |
-
{
|
| 733 |
-
"epoch": 0.7685557249087802,
|
| 734 |
-
"grad_norm": 0.9075937271118164,
|
| 735 |
-
"learning_rate": 0.000132409666069565,
|
| 736 |
-
"loss": 1.1790531873703003,
|
| 737 |
-
"step": 16640
|
| 738 |
-
},
|
| 739 |
-
{
|
| 740 |
-
"epoch": 0.780379659138146,
|
| 741 |
-
"grad_norm": 1.562941551208496,
|
| 742 |
-
"learning_rate": 0.0001197837462455823,
|
| 743 |
-
"loss": 1.1757910251617432,
|
| 744 |
-
"step": 16896
|
| 745 |
-
},
|
| 746 |
-
{
|
| 747 |
-
"epoch": 0.7922035933675119,
|
| 748 |
-
"grad_norm": 0.8990936875343323,
|
| 749 |
-
"learning_rate": 0.00010770811321550749,
|
| 750 |
-
"loss": 1.1700949668884277,
|
| 751 |
-
"step": 17152
|
| 752 |
-
},
|
| 753 |
-
{
|
| 754 |
-
"epoch": 0.8040275275968778,
|
| 755 |
-
"grad_norm": 0.9770736694335938,
|
| 756 |
-
"learning_rate": 9.620024403698591e-05,
|
| 757 |
-
"loss": 1.173874855041504,
|
| 758 |
-
"step": 17408
|
| 759 |
-
},
|
| 760 |
-
{
|
| 761 |
-
"epoch": 0.8040275275968778,
|
| 762 |
-
"eval_cos_loss": 0.3929577595022715,
|
| 763 |
-
"eval_loss": 1.176220568892074,
|
| 764 |
-
"eval_mse_loss": 1.1369247942754668,
|
| 765 |
-
"step": 17408
|
| 766 |
-
},
|
| 767 |
-
{
|
| 768 |
-
"epoch": 0.8040275275968778,
|
| 769 |
-
"eval_cos_loss": 0.3929577595022715,
|
| 770 |
-
"eval_loss": 1.176220568892074,
|
| 771 |
-
"eval_mse_loss": 1.1369247942754668,
|
| 772 |
-
"eval_runtime": 22.1075,
|
| 773 |
-
"eval_samples_per_second": 1266.22,
|
| 774 |
-
"eval_steps_per_second": 19.812,
|
| 775 |
-
"step": 17408
|
| 776 |
-
},
|
| 777 |
-
{
|
| 778 |
-
"epoch": 0.8158514618262436,
|
| 779 |
-
"grad_norm": 1.4291921854019165,
|
| 780 |
-
"learning_rate": 8.527679404332429e-05,
|
| 781 |
-
"loss": 1.176650881767273,
|
| 782 |
-
"step": 17664
|
| 783 |
-
},
|
| 784 |
-
{
|
| 785 |
-
"epoch": 0.8276753960556095,
|
| 786 |
-
"grad_norm": 0.7356229424476624,
|
| 787 |
-
"learning_rate": 7.495357273823544e-05,
|
| 788 |
-
"loss": 1.1727054119110107,
|
| 789 |
-
"step": 17920
|
| 790 |
-
},
|
| 791 |
-
{
|
| 792 |
-
"epoch": 0.8394993302849753,
|
| 793 |
-
"grad_norm": 1.2436726093292236,
|
| 794 |
-
"learning_rate": 6.524552091475183e-05,
|
| 795 |
-
"loss": 1.1703311204910278,
|
| 796 |
-
"step": 18176
|
| 797 |
-
},
|
| 798 |
-
{
|
| 799 |
-
"epoch": 0.8513232645143411,
|
| 800 |
-
"grad_norm": 3.1195194721221924,
|
| 801 |
-
"learning_rate": 5.6166689031422024e-05,
|
| 802 |
-
"loss": 1.1724638938903809,
|
| 803 |
-
"step": 18432
|
| 804 |
-
},
|
| 805 |
-
{
|
| 806 |
-
"epoch": 0.8513232645143411,
|
| 807 |
-
"eval_cos_loss": 0.39162174615685796,
|
| 808 |
-
"eval_loss": 1.1706689879230168,
|
| 809 |
-
"eval_mse_loss": 1.1315068142599167,
|
| 810 |
-
"step": 18432
|
| 811 |
-
},
|
| 812 |
-
{
|
| 813 |
-
"epoch": 0.8513232645143411,
|
| 814 |
-
"eval_cos_loss": 0.39162174615685796,
|
| 815 |
-
"eval_loss": 1.1706689879230168,
|
| 816 |
-
"eval_mse_loss": 1.1315068142599167,
|
| 817 |
-
"eval_runtime": 23.039,
|
| 818 |
-
"eval_samples_per_second": 1215.025,
|
| 819 |
-
"eval_steps_per_second": 19.011,
|
| 820 |
-
"step": 18432
|
| 821 |
-
},
|
| 822 |
-
{
|
| 823 |
-
"epoch": 0.8631471987437069,
|
| 824 |
-
"grad_norm": 1.4888677597045898,
|
| 825 |
-
"learning_rate": 4.773021687709067e-05,
|
| 826 |
-
"loss": 1.1717946529388428,
|
| 827 |
-
"step": 18688
|
| 828 |
-
},
|
| 829 |
-
{
|
| 830 |
-
"epoch": 0.8749711329730728,
|
| 831 |
-
"grad_norm": 0.46844610571861267,
|
| 832 |
-
"learning_rate": 3.994831455368719e-05,
|
| 833 |
-
"loss": 1.1667778491973877,
|
| 834 |
-
"step": 18944
|
| 835 |
-
},
|
| 836 |
-
{
|
| 837 |
-
"epoch": 0.8867950672024387,
|
| 838 |
-
"grad_norm": 1.4531166553497314,
|
| 839 |
-
"learning_rate": 3.283224480455282e-05,
|
| 840 |
-
"loss": 1.1718573570251465,
|
| 841 |
-
"step": 19200
|
| 842 |
-
},
|
| 843 |
-
{
|
| 844 |
-
"epoch": 0.8986190014318045,
|
| 845 |
-
"grad_norm": 0.9106882810592651,
|
| 846 |
-
"learning_rate": 2.639230671387627e-05,
|
| 847 |
-
"loss": 1.1701186895370483,
|
| 848 |
-
"step": 19456
|
| 849 |
-
},
|
| 850 |
-
{
|
| 851 |
-
"epoch": 0.8986190014318045,
|
| 852 |
-
"eval_cos_loss": 0.3913356080719325,
|
| 853 |
-
"eval_loss": 1.1712680537951048,
|
| 854 |
-
"eval_mse_loss": 1.1321344944440066,
|
| 855 |
-
"step": 19456
|
| 856 |
-
},
|
| 857 |
-
{
|
| 858 |
-
"epoch": 0.8986190014318045,
|
| 859 |
-
"eval_cos_loss": 0.3913356080719325,
|
| 860 |
-
"eval_loss": 1.1712680537951048,
|
| 861 |
-
"eval_mse_loss": 1.1321344944440066,
|
| 862 |
-
"eval_runtime": 22.9377,
|
| 863 |
-
"eval_samples_per_second": 1220.394,
|
| 864 |
-
"eval_steps_per_second": 19.095,
|
| 865 |
-
"step": 19456
|
| 866 |
-
},
|
| 867 |
-
{
|
| 868 |
-
"epoch": 0.9104429356611704,
|
| 869 |
-
"grad_norm": 1.2496660947799683,
|
| 870 |
-
"learning_rate": 2.063782080083576e-05,
|
| 871 |
-
"loss": 1.170915961265564,
|
| 872 |
-
"step": 19712
|
| 873 |
-
},
|
| 874 |
-
{
|
| 875 |
-
"epoch": 0.9222668698905362,
|
| 876 |
-
"grad_norm": 0.9893367886543274,
|
| 877 |
-
"learning_rate": 1.557711553001523e-05,
|
| 878 |
-
"loss": 1.1714890003204346,
|
| 879 |
-
"step": 19968
|
| 880 |
-
},
|
| 881 |
-
{
|
| 882 |
-
"epoch": 0.9340908041199021,
|
| 883 |
-
"grad_norm": 0.43370935320854187,
|
| 884 |
-
"learning_rate": 1.1217515257622269e-05,
|
| 885 |
-
"loss": 1.1684632301330566,
|
| 886 |
-
"step": 20224
|
| 887 |
-
},
|
| 888 |
-
{
|
| 889 |
-
"epoch": 0.945914738349268,
|
| 890 |
-
"grad_norm": 1.3099125623703003,
|
| 891 |
-
"learning_rate": 7.565329630950746e-06,
|
| 892 |
-
"loss": 1.1713871955871582,
|
| 893 |
-
"step": 20480
|
| 894 |
-
},
|
| 895 |
-
{
|
| 896 |
-
"epoch": 0.945914738349268,
|
| 897 |
-
"eval_cos_loss": 0.39161273020587556,
|
| 898 |
-
"eval_loss": 1.1713832663074475,
|
| 899 |
-
"eval_mse_loss": 1.1322219940625369,
|
| 900 |
-
"step": 20480
|
| 901 |
-
},
|
| 902 |
-
{
|
| 903 |
-
"epoch": 0.945914738349268,
|
| 904 |
-
"eval_cos_loss": 0.39161273020587556,
|
| 905 |
-
"eval_loss": 1.1713832663074475,
|
| 906 |
-
"eval_mse_loss": 1.1322219940625369,
|
| 907 |
-
"eval_runtime": 25.5104,
|
| 908 |
-
"eval_samples_per_second": 1097.317,
|
| 909 |
-
"eval_steps_per_second": 17.169,
|
| 910 |
-
"step": 20480
|
| 911 |
-
},
|
| 912 |
-
{
|
| 913 |
-
"epoch": 0.9577386725786338,
|
| 914 |
-
"grad_norm": 0.8409441709518433,
|
| 915 |
-
"learning_rate": 4.62584445643166e-06,
|
| 916 |
-
"loss": 1.1694996356964111,
|
| 917 |
-
"step": 20736
|
| 918 |
-
},
|
| 919 |
-
{
|
| 920 |
-
"epoch": 0.9695626068079997,
|
| 921 |
-
"grad_norm": 0.2852078974246979,
|
| 922 |
-
"learning_rate": 2.40331404948807e-06,
|
| 923 |
-
"loss": 1.1747032403945923,
|
| 924 |
-
"step": 20992
|
| 925 |
-
},
|
| 926 |
-
{
|
| 927 |
-
"epoch": 0.9813865410373654,
|
| 928 |
-
"grad_norm": 1.3213860988616943,
|
| 929 |
-
"learning_rate": 9.009550772663965e-07,
|
| 930 |
-
"loss": 1.171601414680481,
|
| 931 |
-
"step": 21248
|
| 932 |
-
},
|
| 933 |
-
{
|
| 934 |
-
"epoch": 0.9932104752667313,
|
| 935 |
-
"grad_norm": 0.9441902041435242,
|
| 936 |
-
"learning_rate": 1.2094190315575791e-07,
|
| 937 |
-
"loss": 1.171757459640503,
|
| 938 |
-
"step": 21504
|
| 939 |
-
},
|
| 940 |
-
{
|
| 941 |
-
"epoch": 0.9932104752667313,
|
| 942 |
-
"eval_cos_loss": 0.38952490823453967,
|
| 943 |
-
"eval_loss": 1.165849423843976,
|
| 944 |
-
"eval_mse_loss": 1.1268969341500166,
|
| 945 |
-
"step": 21504
|
| 946 |
-
},
|
| 947 |
-
{
|
| 948 |
-
"epoch": 0.9932104752667313,
|
| 949 |
-
"eval_cos_loss": 0.38952490823453967,
|
| 950 |
-
"eval_loss": 1.165849423843976,
|
| 951 |
-
"eval_mse_loss": 1.1268969341500166,
|
| 952 |
-
"eval_runtime": 24.1604,
|
| 953 |
-
"eval_samples_per_second": 1158.629,
|
| 954 |
-
"eval_steps_per_second": 18.129,
|
| 955 |
-
"step": 21504
|
| 956 |
-
}
|
| 957 |
-
],
|
| 958 |
-
"logging_steps": 256,
|
| 959 |
-
"max_steps": 21651,
|
| 960 |
-
"num_input_tokens_seen": 0,
|
| 961 |
-
"num_train_epochs": 1,
|
| 962 |
-
"save_steps": 1024,
|
| 963 |
-
"stateful_callbacks": {
|
| 964 |
-
"TrainerControl": {
|
| 965 |
-
"args": {
|
| 966 |
-
"should_epoch_stop": false,
|
| 967 |
-
"should_evaluate": false,
|
| 968 |
-
"should_log": false,
|
| 969 |
-
"should_save": true,
|
| 970 |
-
"should_training_stop": false
|
| 971 |
-
},
|
| 972 |
-
"attributes": {}
|
| 973 |
-
}
|
| 974 |
-
},
|
| 975 |
-
"total_flos": 0.0,
|
| 976 |
-
"train_batch_size": 64,
|
| 977 |
-
"trial_name": null,
|
| 978 |
-
"trial_params": null
|
| 979 |
-
}
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
checkpoints-v6.0-discrete/checkpoint-21504/training_args.bin
DELETED
|
@@ -1,3 +0,0 @@
|
|
| 1 |
-
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:61a16bb839f687414b8e48611327c4b9cfddeefe38c031ca70808f9a97c476b7
|
| 3 |
-
size 5137
|
|
|
|
|
|
|
|
|
|
|
|