Upload folder using huggingface_hub
Browse files- all_results.json +3 -3
- checkpoint-1480/global_step1479/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +1 -1
- checkpoint-1480/global_step1479/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt +1 -1
- checkpoint-1480/global_step1479/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt +1 -1
- checkpoint-1480/global_step1479/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt +1 -1
- checkpoint-1480/model-00001-of-00004.safetensors +1 -1
- checkpoint-1480/model-00002-of-00004.safetensors +1 -1
- checkpoint-1480/model-00003-of-00004.safetensors +1 -1
- checkpoint-1480/model-00004-of-00004.safetensors +1 -1
- checkpoint-1480/tokenizer.json +2 -2
- checkpoint-1480/trainer_state.json +556 -556
- checkpoint-1480/training_args.bin +1 -1
- model-00001-of-00004.safetensors +1 -1
- model-00002-of-00004.safetensors +1 -1
- model-00003-of-00004.safetensors +1 -1
- model-00004-of-00004.safetensors +1 -1
- tokenizer.json +2 -2
- train_results.json +3 -3
- trainer_state.json +559 -559
- training_args.bin +1 -1
all_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.9980866629150253,
|
| 3 |
"total_flos": 3.6719988787579453e+18,
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 35538,
|
| 7 |
-
"train_samples_per_second": 5.
|
| 8 |
"train_steps_per_second": 0.112
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.9980866629150253,
|
| 3 |
"total_flos": 3.6719988787579453e+18,
|
| 4 |
+
"train_loss": 0.37467364362768224,
|
| 5 |
+
"train_runtime": 13260.7751,
|
| 6 |
"train_samples": 35538,
|
| 7 |
+
"train_samples_per_second": 5.36,
|
| 8 |
"train_steps_per_second": 0.112
|
| 9 |
}
|
checkpoint-1480/global_step1479/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 24090788620
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3576ce12151079b052c59095e6430fb0f28b9bd15decdec4def7759afc7fcf9d
|
| 3 |
size 24090788620
|
checkpoint-1480/global_step1479/bf16_zero_pp_rank_1_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 24090788620
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6cc6afb0420ab474c145073970fe7e49bdb21b3fe19074b08220bdfe578a4965
|
| 3 |
size 24090788620
|
checkpoint-1480/global_step1479/bf16_zero_pp_rank_2_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 24090788620
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1d87cc66ce69345659854a64c98e6a4434a70632da3ec8cb6f127f026c70d41b
|
| 3 |
size 24090788620
|
checkpoint-1480/global_step1479/bf16_zero_pp_rank_3_mp_rank_00_optim_states.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 24090788620
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cbe08f990693517a2757f0cfa829dec8ce46c9d602ba9e94eaafacd0820cb756
|
| 3 |
size 24090788620
|
checkpoint-1480/model-00001-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4976698672
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bd78dcad6a6e52066ad8f69de370e285ef65b53be6a6f5b003ec73161153f1a1
|
| 3 |
size 4976698672
|
checkpoint-1480/model-00002-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4999802720
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eaf863d03a19d67f3b5bd124a1af51716f33a11900ec798135fa04c015b7780c
|
| 3 |
size 4999802720
|
checkpoint-1480/model-00003-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4915916176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0701dd793589970557efd0a6b3fa5a9e330186f1276390178e9fd5ee51010030
|
| 3 |
size 4915916176
|
checkpoint-1480/model-00004-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1168138808
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1652d220d7874225cd65e0501667096059583d52796ad6ed749081c58c12b3c7
|
| 3 |
size 1168138808
|
checkpoint-1480/tokenizer.json
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c5cf44023714fb39b05e71e425f8d7b92805ff73f7988b083b8c87f0bf87393
|
| 3 |
+
size 17209961
|
checkpoint-1480/trainer_state.json
CHANGED
|
@@ -10,2097 +10,2097 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0013505908835115363,
|
| 13 |
-
"grad_norm": 2.
|
| 14 |
"learning_rate": 6.756756756756757e-08,
|
| 15 |
"loss": 2.1725,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.006752954417557681,
|
| 20 |
-
"grad_norm": 2.
|
| 21 |
"learning_rate": 3.378378378378379e-07,
|
| 22 |
-
"loss": 2.
|
| 23 |
"step": 5
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.013505908835115363,
|
| 27 |
-
"grad_norm": 1.
|
| 28 |
"learning_rate": 6.756756756756758e-07,
|
| 29 |
-
"loss": 2.
|
| 30 |
"step": 10
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.020258863252673044,
|
| 34 |
-
"grad_norm": 0.
|
| 35 |
"learning_rate": 1.0135135135135136e-06,
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 15
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.027011817670230726,
|
| 41 |
-
"grad_norm": 0.
|
| 42 |
"learning_rate": 1.3513513513513515e-06,
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 20
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.03376477208778841,
|
| 48 |
-
"grad_norm": 0.
|
| 49 |
"learning_rate": 1.6891891891891894e-06,
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 25
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.04051772650534609,
|
| 55 |
-
"grad_norm": 0.
|
| 56 |
"learning_rate": 2.0270270270270273e-06,
|
| 57 |
-
"loss": 0.
|
| 58 |
"step": 30
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.04727068092290377,
|
| 62 |
-
"grad_norm": 0.
|
| 63 |
"learning_rate": 2.364864864864865e-06,
|
| 64 |
-
"loss": 0.
|
| 65 |
"step": 35
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.05402363534046145,
|
| 69 |
-
"grad_norm": 0.
|
| 70 |
"learning_rate": 2.702702702702703e-06,
|
| 71 |
-
"loss": 0.
|
| 72 |
"step": 40
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.06077658975801913,
|
| 76 |
-
"grad_norm": 0.
|
| 77 |
"learning_rate": 3.040540540540541e-06,
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 45
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.06752954417557681,
|
| 83 |
-
"grad_norm": 0.
|
| 84 |
"learning_rate": 3.3783783783783788e-06,
|
| 85 |
-
"loss": 0.
|
| 86 |
"step": 50
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.0742824985931345,
|
| 90 |
-
"grad_norm": 0.
|
| 91 |
"learning_rate": 3.7162162162162162e-06,
|
| 92 |
"loss": 0.5621,
|
| 93 |
"step": 55
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.08103545301069218,
|
| 97 |
-
"grad_norm": 0.
|
| 98 |
"learning_rate": 4.0540540540540545e-06,
|
| 99 |
-
"loss": 0.
|
| 100 |
"step": 60
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.08778840742824986,
|
| 104 |
-
"grad_norm": 0.
|
| 105 |
"learning_rate": 4.391891891891892e-06,
|
| 106 |
-
"loss": 0.
|
| 107 |
"step": 65
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.09454136184580754,
|
| 111 |
-
"grad_norm": 0.
|
| 112 |
"learning_rate": 4.72972972972973e-06,
|
| 113 |
-
"loss": 0.
|
| 114 |
"step": 70
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.10129431626336523,
|
| 118 |
-
"grad_norm": 0.
|
| 119 |
"learning_rate": 5.067567567567568e-06,
|
| 120 |
-
"loss": 0.
|
| 121 |
"step": 75
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.1080472706809229,
|
| 125 |
-
"grad_norm": 0.
|
| 126 |
"learning_rate": 5.405405405405406e-06,
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 80
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.11480022509848059,
|
| 132 |
-
"grad_norm": 0.
|
| 133 |
"learning_rate": 5.743243243243244e-06,
|
| 134 |
"loss": 0.471,
|
| 135 |
"step": 85
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.12155317951603826,
|
| 139 |
-
"grad_norm": 0.
|
| 140 |
"learning_rate": 6.081081081081082e-06,
|
| 141 |
-
"loss": 0.
|
| 142 |
"step": 90
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.12830613393359594,
|
| 146 |
-
"grad_norm": 0.
|
| 147 |
"learning_rate": 6.41891891891892e-06,
|
| 148 |
-
"loss": 0.
|
| 149 |
"step": 95
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.13505908835115363,
|
| 153 |
-
"grad_norm": 0.
|
| 154 |
"learning_rate": 6.7567567567567575e-06,
|
| 155 |
"loss": 0.4398,
|
| 156 |
"step": 100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.14181204276871132,
|
| 160 |
-
"grad_norm": 0.
|
| 161 |
"learning_rate": 7.0945945945945946e-06,
|
| 162 |
"loss": 0.3899,
|
| 163 |
"step": 105
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.148564997186269,
|
| 167 |
-
"grad_norm": 0.
|
| 168 |
"learning_rate": 7.4324324324324324e-06,
|
| 169 |
-
"loss": 0.
|
| 170 |
"step": 110
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.15531795160382666,
|
| 174 |
-
"grad_norm": 0.
|
| 175 |
"learning_rate": 7.77027027027027e-06,
|
| 176 |
-
"loss": 0.
|
| 177 |
"step": 115
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.16207090602138435,
|
| 181 |
-
"grad_norm": 0.
|
| 182 |
"learning_rate": 8.108108108108109e-06,
|
| 183 |
-
"loss": 0.
|
| 184 |
"step": 120
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.16882386043894204,
|
| 188 |
-
"grad_norm": 0.
|
| 189 |
"learning_rate": 8.445945945945948e-06,
|
| 190 |
-
"loss": 0.
|
| 191 |
"step": 125
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.17557681485649973,
|
| 195 |
-
"grad_norm": 0.
|
| 196 |
"learning_rate": 8.783783783783785e-06,
|
| 197 |
-
"loss": 0.
|
| 198 |
"step": 130
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.1823297692740574,
|
| 202 |
-
"grad_norm": 0.
|
| 203 |
"learning_rate": 9.121621621621622e-06,
|
| 204 |
"loss": 0.3895,
|
| 205 |
"step": 135
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.18908272369161508,
|
| 209 |
-
"grad_norm": 0.
|
| 210 |
"learning_rate": 9.45945945945946e-06,
|
| 211 |
"loss": 0.4029,
|
| 212 |
"step": 140
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.19583567810917277,
|
| 216 |
-
"grad_norm": 0.
|
| 217 |
"learning_rate": 9.797297297297298e-06,
|
| 218 |
-
"loss": 0.
|
| 219 |
"step": 145
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.20258863252673046,
|
| 223 |
-
"grad_norm": 0.
|
| 224 |
"learning_rate": 9.999944372378571e-06,
|
| 225 |
-
"loss": 0.
|
| 226 |
"step": 150
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.20934158694428812,
|
| 230 |
-
"grad_norm": 0.
|
| 231 |
"learning_rate": 9.999318575852451e-06,
|
| 232 |
-
"loss": 0.
|
| 233 |
"step": 155
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.2160945413618458,
|
| 237 |
-
"grad_norm": 0.
|
| 238 |
"learning_rate": 9.99799753559161e-06,
|
| 239 |
-
"loss": 0.
|
| 240 |
"step": 160
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.2228474957794035,
|
| 244 |
-
"grad_norm": 0.
|
| 245 |
"learning_rate": 9.995981435310078e-06,
|
| 246 |
-
"loss": 0.
|
| 247 |
"step": 165
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.22960045019696118,
|
| 251 |
-
"grad_norm": 0.
|
| 252 |
"learning_rate": 9.993270555382283e-06,
|
| 253 |
"loss": 0.4058,
|
| 254 |
"step": 170
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.23635340461451884,
|
| 258 |
-
"grad_norm": 0.
|
| 259 |
"learning_rate": 9.989865272804064e-06,
|
| 260 |
-
"loss": 0.
|
| 261 |
"step": 175
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.24310635903207653,
|
| 265 |
-
"grad_norm": 0.
|
| 266 |
"learning_rate": 9.985766061140233e-06,
|
| 267 |
-
"loss": 0.
|
| 268 |
"step": 180
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.24985931344963422,
|
| 272 |
-
"grad_norm": 0.
|
| 273 |
"learning_rate": 9.980973490458728e-06,
|
| 274 |
-
"loss": 0.
|
| 275 |
"step": 185
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.2566122678671919,
|
| 279 |
-
"grad_norm": 0.
|
| 280 |
"learning_rate": 9.97548822725133e-06,
|
| 281 |
-
"loss": 0.
|
| 282 |
"step": 190
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.26336522228474957,
|
| 286 |
-
"grad_norm": 0.
|
| 287 |
"learning_rate": 9.969311034340977e-06,
|
| 288 |
-
"loss": 0.
|
| 289 |
"step": 195
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.27011817670230726,
|
| 293 |
-
"grad_norm": 0.
|
| 294 |
"learning_rate": 9.962442770775675e-06,
|
| 295 |
-
"loss": 0.
|
| 296 |
"step": 200
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.27687113111986494,
|
| 300 |
-
"grad_norm": 0.
|
| 301 |
"learning_rate": 9.954884391709043e-06,
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 205
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.28362408553742263,
|
| 307 |
-
"grad_norm": 0.
|
| 308 |
"learning_rate": 9.946636948267468e-06,
|
| 309 |
"loss": 0.349,
|
| 310 |
"step": 210
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.2903770399549803,
|
| 314 |
-
"grad_norm": 0.
|
| 315 |
"learning_rate": 9.937701587403941e-06,
|
| 316 |
-
"loss": 0.
|
| 317 |
"step": 215
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.297129994372538,
|
| 321 |
-
"grad_norm": 0.
|
| 322 |
"learning_rate": 9.928079551738542e-06,
|
| 323 |
-
"loss": 0.
|
| 324 |
"step": 220
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.30388294879009564,
|
| 328 |
-
"grad_norm": 0.
|
| 329 |
"learning_rate": 9.91777217938564e-06,
|
| 330 |
-
"loss": 0.
|
| 331 |
"step": 225
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.31063590320765333,
|
| 335 |
-
"grad_norm": 0.
|
| 336 |
"learning_rate": 9.906780903767799e-06,
|
| 337 |
-
"loss": 0.
|
| 338 |
"step": 230
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.317388857625211,
|
| 342 |
-
"grad_norm": 0.
|
| 343 |
"learning_rate": 9.895107253416434e-06,
|
| 344 |
-
"loss": 0.
|
| 345 |
"step": 235
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.3241418120427687,
|
| 349 |
-
"grad_norm": 0.
|
| 350 |
"learning_rate": 9.882752851759247e-06,
|
| 351 |
-
"loss": 0.
|
| 352 |
"step": 240
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.3308947664603264,
|
| 356 |
-
"grad_norm": 0.
|
| 357 |
"learning_rate": 9.869719416894462e-06,
|
| 358 |
-
"loss": 0.
|
| 359 |
"step": 245
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.3376477208778841,
|
| 363 |
-
"grad_norm": 0.
|
| 364 |
"learning_rate": 9.856008761351882e-06,
|
| 365 |
-
"loss": 0.
|
| 366 |
"step": 250
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 0.34440067529544177,
|
| 370 |
-
"grad_norm": 0.
|
| 371 |
"learning_rate": 9.841622791840839e-06,
|
| 372 |
-
"loss": 0.
|
| 373 |
"step": 255
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.35115362971299946,
|
| 377 |
-
"grad_norm": 0.
|
| 378 |
"learning_rate": 9.826563508985017e-06,
|
| 379 |
-
"loss": 0.
|
| 380 |
"step": 260
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.3579065841305571,
|
| 384 |
-
"grad_norm": 0.
|
| 385 |
"learning_rate": 9.810833007044247e-06,
|
| 386 |
"loss": 0.3925,
|
| 387 |
"step": 265
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.3646595385481148,
|
| 391 |
-
"grad_norm": 0.
|
| 392 |
"learning_rate": 9.794433473623249e-06,
|
| 393 |
-
"loss": 0.
|
| 394 |
"step": 270
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 0.37141249296567247,
|
| 398 |
-
"grad_norm": 0.
|
| 399 |
"learning_rate": 9.777367189367412e-06,
|
| 400 |
"loss": 0.3779,
|
| 401 |
"step": 275
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.37816544738323016,
|
| 405 |
-
"grad_norm": 0.
|
| 406 |
"learning_rate": 9.759636527645633e-06,
|
| 407 |
-
"loss": 0.
|
| 408 |
"step": 280
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.38491840180078785,
|
| 412 |
-
"grad_norm": 0.
|
| 413 |
"learning_rate": 9.74124395422025e-06,
|
| 414 |
"loss": 0.3725,
|
| 415 |
"step": 285
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.39167135621834553,
|
| 419 |
-
"grad_norm": 0.
|
| 420 |
"learning_rate": 9.722192026904145e-06,
|
| 421 |
-
"loss": 0.
|
| 422 |
"step": 290
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.3984243106359032,
|
| 426 |
-
"grad_norm": 0.
|
| 427 |
"learning_rate": 9.702483395205023e-06,
|
| 428 |
-
"loss": 0.
|
| 429 |
"step": 295
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.4051772650534609,
|
| 433 |
-
"grad_norm": 0.
|
| 434 |
"learning_rate": 9.682120799956961e-06,
|
| 435 |
-
"loss": 0.
|
| 436 |
"step": 300
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 0.41193021947101854,
|
| 440 |
-
"grad_norm": 0.
|
| 441 |
"learning_rate": 9.661107072939244e-06,
|
| 442 |
-
"loss": 0.
|
| 443 |
"step": 305
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.41868317388857623,
|
| 447 |
-
"grad_norm": 0.
|
| 448 |
"learning_rate": 9.639445136482549e-06,
|
| 449 |
-
"loss": 0.
|
| 450 |
"step": 310
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.4254361283061339,
|
| 454 |
-
"grad_norm": 0.
|
| 455 |
"learning_rate": 9.61713800306255e-06,
|
| 456 |
-
"loss": 0.
|
| 457 |
"step": 315
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.4321890827236916,
|
| 461 |
-
"grad_norm": 0.
|
| 462 |
"learning_rate": 9.594188774880981e-06,
|
| 463 |
-
"loss": 0.
|
| 464 |
"step": 320
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.4389420371412493,
|
| 468 |
-
"grad_norm": 0.
|
| 469 |
"learning_rate": 9.570600643434217e-06,
|
| 470 |
"loss": 0.3371,
|
| 471 |
"step": 325
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.445694991558807,
|
| 475 |
-
"grad_norm": 0.
|
| 476 |
"learning_rate": 9.546376889069442e-06,
|
| 477 |
-
"loss": 0.
|
| 478 |
"step": 330
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 0.4524479459763647,
|
| 482 |
-
"grad_norm": 0.
|
| 483 |
"learning_rate": 9.521520880528453e-06,
|
| 484 |
-
"loss": 0.
|
| 485 |
"step": 335
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.45920090039392236,
|
| 489 |
-
"grad_norm": 0.
|
| 490 |
"learning_rate": 9.496036074479184e-06,
|
| 491 |
"loss": 0.3604,
|
| 492 |
"step": 340
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 0.46595385481148,
|
| 496 |
-
"grad_norm": 0.
|
| 497 |
"learning_rate": 9.46992601503499e-06,
|
| 498 |
-
"loss": 0.
|
| 499 |
"step": 345
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.4727068092290377,
|
| 503 |
-
"grad_norm": 0.
|
| 504 |
"learning_rate": 9.44319433326178e-06,
|
| 505 |
-
"loss": 0.
|
| 506 |
"step": 350
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.47945976364659537,
|
| 510 |
-
"grad_norm": 0.
|
| 511 |
"learning_rate": 9.415844746673047e-06,
|
| 512 |
-
"loss": 0.
|
| 513 |
"step": 355
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.48621271806415306,
|
| 517 |
-
"grad_norm": 0.
|
| 518 |
"learning_rate": 9.387881058712888e-06,
|
| 519 |
-
"loss": 0.
|
| 520 |
"step": 360
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 0.49296567248171075,
|
| 524 |
-
"grad_norm": 0.
|
| 525 |
"learning_rate": 9.359307158227067e-06,
|
| 526 |
-
"loss": 0.
|
| 527 |
"step": 365
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.49971862689926844,
|
| 531 |
-
"grad_norm": 0.
|
| 532 |
"learning_rate": 9.330127018922195e-06,
|
| 533 |
-
"loss": 0.
|
| 534 |
"step": 370
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.5064715813168261,
|
| 538 |
-
"grad_norm": 0.
|
| 539 |
"learning_rate": 9.300344698813124e-06,
|
| 540 |
-
"loss": 0.
|
| 541 |
"step": 375
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.5132245357343838,
|
| 545 |
-
"grad_norm": 0.
|
| 546 |
"learning_rate": 9.269964339658605e-06,
|
| 547 |
-
"loss": 0.
|
| 548 |
"step": 380
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.5199774901519415,
|
| 552 |
-
"grad_norm": 0.
|
| 553 |
"learning_rate": 9.238990166385304e-06,
|
| 554 |
-
"loss": 0.
|
| 555 |
"step": 385
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.5267304445694991,
|
| 559 |
-
"grad_norm": 0.
|
| 560 |
"learning_rate": 9.207426486500252e-06,
|
| 561 |
"loss": 0.3587,
|
| 562 |
"step": 390
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 0.5334833989870569,
|
| 566 |
-
"grad_norm": 0.
|
| 567 |
"learning_rate": 9.175277689491804e-06,
|
| 568 |
-
"loss": 0.
|
| 569 |
"step": 395
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.5402363534046145,
|
| 573 |
-
"grad_norm": 0.
|
| 574 |
"learning_rate": 9.142548246219212e-06,
|
| 575 |
-
"loss": 0.
|
| 576 |
"step": 400
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 0.5469893078221723,
|
| 580 |
-
"grad_norm": 0.
|
| 581 |
"learning_rate": 9.109242708290864e-06,
|
| 582 |
"loss": 0.3427,
|
| 583 |
"step": 405
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.5537422622397299,
|
| 587 |
-
"grad_norm": 0.
|
| 588 |
"learning_rate": 9.075365707431311e-06,
|
| 589 |
-
"loss": 0.
|
| 590 |
"step": 410
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.5604952166572875,
|
| 594 |
-
"grad_norm": 0.
|
| 595 |
"learning_rate": 9.040921954837139e-06,
|
| 596 |
-
"loss": 0.
|
| 597 |
"step": 415
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.5672481710748453,
|
| 601 |
-
"grad_norm": 0.
|
| 602 |
"learning_rate": 9.005916240521788e-06,
|
| 603 |
-
"loss": 0.
|
| 604 |
"step": 420
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 0.5740011254924029,
|
| 608 |
-
"grad_norm": 0.
|
| 609 |
"learning_rate": 8.97035343264943e-06,
|
| 610 |
-
"loss": 0.
|
| 611 |
"step": 425
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.5807540799099606,
|
| 615 |
-
"grad_norm": 0.
|
| 616 |
"learning_rate": 8.93423847685795e-06,
|
| 617 |
-
"loss": 0.
|
| 618 |
"step": 430
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 0.5875070343275183,
|
| 622 |
-
"grad_norm": 0.
|
| 623 |
"learning_rate": 8.89757639557118e-06,
|
| 624 |
-
"loss": 0.
|
| 625 |
"step": 435
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.594259988745076,
|
| 629 |
-
"grad_norm": 0.
|
| 630 |
"learning_rate": 8.860372287300432e-06,
|
| 631 |
"loss": 0.3703,
|
| 632 |
"step": 440
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 0.6010129431626337,
|
| 636 |
-
"grad_norm": 0.
|
| 637 |
"learning_rate": 8.822631325935463e-06,
|
| 638 |
-
"loss": 0.
|
| 639 |
"step": 445
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.6077658975801913,
|
| 643 |
-
"grad_norm": 0.
|
| 644 |
"learning_rate": 8.78435876002496e-06,
|
| 645 |
-
"loss": 0.
|
| 646 |
"step": 450
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 0.614518851997749,
|
| 650 |
-
"grad_norm": 0.
|
| 651 |
"learning_rate": 8.745559912046625e-06,
|
| 652 |
-
"loss": 0.
|
| 653 |
"step": 455
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.6212718064153067,
|
| 657 |
-
"grad_norm": 0.
|
| 658 |
"learning_rate": 8.706240177667003e-06,
|
| 659 |
"loss": 0.3476,
|
| 660 |
"step": 460
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 0.6280247608328644,
|
| 664 |
-
"grad_norm": 0.
|
| 665 |
"learning_rate": 8.666405024991105e-06,
|
| 666 |
-
"loss": 0.
|
| 667 |
"step": 465
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.634777715250422,
|
| 671 |
-
"grad_norm": 0.
|
| 672 |
"learning_rate": 8.626059993801988e-06,
|
| 673 |
-
"loss": 0.
|
| 674 |
"step": 470
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.6415306696679798,
|
| 678 |
-
"grad_norm": 0.
|
| 679 |
"learning_rate": 8.585210694790333e-06,
|
| 680 |
-
"loss": 0.
|
| 681 |
"step": 475
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.6482836240855374,
|
| 685 |
-
"grad_norm": 0.
|
| 686 |
"learning_rate": 8.543862808774193e-06,
|
| 687 |
-
"loss": 0.
|
| 688 |
"step": 480
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 0.6550365785030952,
|
| 692 |
-
"grad_norm": 0.
|
| 693 |
"learning_rate": 8.502022085908963e-06,
|
| 694 |
"loss": 0.3471,
|
| 695 |
"step": 485
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.6617895329206528,
|
| 699 |
-
"grad_norm": 0.
|
| 700 |
"learning_rate": 8.459694344887732e-06,
|
| 701 |
-
"loss": 0.
|
| 702 |
"step": 490
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 0.6685424873382104,
|
| 706 |
-
"grad_norm": 0.
|
| 707 |
"learning_rate": 8.416885472132077e-06,
|
| 708 |
-
"loss": 0.
|
| 709 |
"step": 495
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.6752954417557682,
|
| 713 |
-
"grad_norm": 0.
|
| 714 |
"learning_rate": 8.373601420973464e-06,
|
| 715 |
-
"loss": 0.
|
| 716 |
"step": 500
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 0.6820483961733258,
|
| 720 |
-
"grad_norm": 0.
|
| 721 |
"learning_rate": 8.329848210825322e-06,
|
| 722 |
-
"loss": 0.
|
| 723 |
"step": 505
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.6888013505908835,
|
| 727 |
-
"grad_norm": 0.
|
| 728 |
"learning_rate": 8.285631926345943e-06,
|
| 729 |
-
"loss": 0.
|
| 730 |
"step": 510
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 0.6955543050084412,
|
| 734 |
-
"grad_norm": 0.
|
| 735 |
"learning_rate": 8.240958716592304e-06,
|
| 736 |
-
"loss": 0.
|
| 737 |
"step": 515
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.7023072594259989,
|
| 741 |
-
"grad_norm": 0.
|
| 742 |
"learning_rate": 8.195834794164925e-06,
|
| 743 |
"loss": 0.3501,
|
| 744 |
"step": 520
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 0.7090602138435566,
|
| 748 |
-
"grad_norm": 0.
|
| 749 |
"learning_rate": 8.150266434343904e-06,
|
| 750 |
"loss": 0.3296,
|
| 751 |
"step": 525
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.7158131682611142,
|
| 755 |
-
"grad_norm": 0.
|
| 756 |
"learning_rate": 8.104259974216219e-06,
|
| 757 |
-
"loss": 0.
|
| 758 |
"step": 530
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.7225661226786719,
|
| 762 |
-
"grad_norm": 0.
|
| 763 |
"learning_rate": 8.057821811794457e-06,
|
| 764 |
-
"loss": 0.
|
| 765 |
"step": 535
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.7293190770962296,
|
| 769 |
-
"grad_norm": 0.
|
| 770 |
"learning_rate": 8.010958405127048e-06,
|
| 771 |
-
"loss": 0.
|
| 772 |
"step": 540
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 0.7360720315137873,
|
| 776 |
-
"grad_norm": 0.
|
| 777 |
"learning_rate": 7.963676271400158e-06,
|
| 778 |
-
"loss": 0.
|
| 779 |
"step": 545
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.7428249859313449,
|
| 783 |
-
"grad_norm": 0.
|
| 784 |
"learning_rate": 7.915981986031367e-06,
|
| 785 |
-
"loss": 0.
|
| 786 |
"step": 550
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 0.7495779403489027,
|
| 790 |
-
"grad_norm": 0.
|
| 791 |
"learning_rate": 7.86788218175523e-06,
|
| 792 |
-
"loss": 0.
|
| 793 |
"step": 555
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.7563308947664603,
|
| 797 |
-
"grad_norm": 0.
|
| 798 |
"learning_rate": 7.819383547700889e-06,
|
| 799 |
"loss": 0.3354,
|
| 800 |
"step": 560
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 0.7630838491840181,
|
| 804 |
-
"grad_norm": 0.
|
| 805 |
"learning_rate": 7.770492828461824e-06,
|
| 806 |
-
"loss": 0.
|
| 807 |
"step": 565
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.7698368036015757,
|
| 811 |
-
"grad_norm": 0.
|
| 812 |
"learning_rate": 7.721216823157896e-06,
|
| 813 |
-
"loss": 0.
|
| 814 |
"step": 570
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 0.7765897580191333,
|
| 818 |
-
"grad_norm": 0.
|
| 819 |
"learning_rate": 7.671562384489819e-06,
|
| 820 |
-
"loss": 0.
|
| 821 |
"step": 575
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.7833427124366911,
|
| 825 |
-
"grad_norm": 0.
|
| 826 |
"learning_rate": 7.621536417786159e-06,
|
| 827 |
-
"loss": 0.
|
| 828 |
"step": 580
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 0.7900956668542487,
|
| 832 |
-
"grad_norm": 0.
|
| 833 |
"learning_rate": 7.571145880043036e-06,
|
| 834 |
-
"loss": 0.
|
| 835 |
"step": 585
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.7968486212718064,
|
| 839 |
-
"grad_norm": 0.
|
| 840 |
"learning_rate": 7.520397778956623e-06,
|
| 841 |
-
"loss": 0.
|
| 842 |
"step": 590
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.8036015756893641,
|
| 846 |
-
"grad_norm": 0.
|
| 847 |
"learning_rate": 7.469299171948608e-06,
|
| 848 |
-
"loss": 0.
|
| 849 |
"step": 595
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.8103545301069218,
|
| 853 |
-
"grad_norm": 0.
|
| 854 |
"learning_rate": 7.417857165184723e-06,
|
| 855 |
"loss": 0.3474,
|
| 856 |
"step": 600
|
| 857 |
},
|
| 858 |
{
|
| 859 |
"epoch": 0.8171074845244795,
|
| 860 |
-
"grad_norm": 0.
|
| 861 |
"learning_rate": 7.366078912586523e-06,
|
| 862 |
-
"loss": 0.
|
| 863 |
"step": 605
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 0.8238604389420371,
|
| 867 |
-
"grad_norm": 0.
|
| 868 |
"learning_rate": 7.313971614836496e-06,
|
| 869 |
-
"loss": 0.
|
| 870 |
"step": 610
|
| 871 |
},
|
| 872 |
{
|
| 873 |
"epoch": 0.8306133933595948,
|
| 874 |
-
"grad_norm": 0.
|
| 875 |
"learning_rate": 7.261542518376677e-06,
|
| 876 |
-
"loss": 0.
|
| 877 |
"step": 615
|
| 878 |
},
|
| 879 |
{
|
| 880 |
"epoch": 0.8373663477771525,
|
| 881 |
-
"grad_norm": 0.
|
| 882 |
"learning_rate": 7.208798914400916e-06,
|
| 883 |
-
"loss": 0.
|
| 884 |
"step": 620
|
| 885 |
},
|
| 886 |
{
|
| 887 |
"epoch": 0.8441193021947102,
|
| 888 |
-
"grad_norm": 0.
|
| 889 |
"learning_rate": 7.155748137840892e-06,
|
| 890 |
-
"loss": 0.
|
| 891 |
"step": 625
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 0.8508722566122678,
|
| 895 |
-
"grad_norm": 0.
|
| 896 |
"learning_rate": 7.102397566346073e-06,
|
| 897 |
-
"loss": 0.
|
| 898 |
"step": 630
|
| 899 |
},
|
| 900 |
{
|
| 901 |
"epoch": 0.8576252110298256,
|
| 902 |
-
"grad_norm": 0.
|
| 903 |
"learning_rate": 7.048754619257716e-06,
|
| 904 |
-
"loss": 0.
|
| 905 |
"step": 635
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.8643781654473832,
|
| 909 |
-
"grad_norm": 0.
|
| 910 |
"learning_rate": 6.994826756577082e-06,
|
| 911 |
-
"loss": 0.
|
| 912 |
"step": 640
|
| 913 |
},
|
| 914 |
{
|
| 915 |
"epoch": 0.871131119864941,
|
| 916 |
-
"grad_norm": 0.
|
| 917 |
"learning_rate": 6.940621477927988e-06,
|
| 918 |
-
"loss": 0.
|
| 919 |
"step": 645
|
| 920 |
},
|
| 921 |
{
|
| 922 |
"epoch": 0.8778840742824986,
|
| 923 |
-
"grad_norm": 0.
|
| 924 |
"learning_rate": 6.88614632151385e-06,
|
| 925 |
-
"loss": 0.
|
| 926 |
"step": 650
|
| 927 |
},
|
| 928 |
{
|
| 929 |
"epoch": 0.8846370287000562,
|
| 930 |
-
"grad_norm": 0.
|
| 931 |
"learning_rate": 6.831408863069364e-06,
|
| 932 |
-
"loss": 0.
|
| 933 |
"step": 655
|
| 934 |
},
|
| 935 |
{
|
| 936 |
"epoch": 0.891389983117614,
|
| 937 |
-
"grad_norm": 0.
|
| 938 |
"learning_rate": 6.7764167148069695e-06,
|
| 939 |
"loss": 0.3251,
|
| 940 |
"step": 660
|
| 941 |
},
|
| 942 |
{
|
| 943 |
"epoch": 0.8981429375351716,
|
| 944 |
-
"grad_norm": 0.
|
| 945 |
"learning_rate": 6.721177524358226e-06,
|
| 946 |
-
"loss": 0.
|
| 947 |
"step": 665
|
| 948 |
},
|
| 949 |
{
|
| 950 |
"epoch": 0.9048958919527293,
|
| 951 |
-
"grad_norm": 0.
|
| 952 |
"learning_rate": 6.665698973710289e-06,
|
| 953 |
-
"loss": 0.
|
| 954 |
"step": 670
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 0.911648846370287,
|
| 958 |
-
"grad_norm": 0.
|
| 959 |
"learning_rate": 6.609988778137582e-06,
|
| 960 |
-
"loss": 0.
|
| 961 |
"step": 675
|
| 962 |
},
|
| 963 |
{
|
| 964 |
"epoch": 0.9184018007878447,
|
| 965 |
-
"grad_norm": 0.
|
| 966 |
"learning_rate": 6.554054685128857e-06,
|
| 967 |
-
"loss": 0.
|
| 968 |
"step": 680
|
| 969 |
},
|
| 970 |
{
|
| 971 |
"epoch": 0.9251547552054024,
|
| 972 |
-
"grad_norm": 0.
|
| 973 |
"learning_rate": 6.497904473309766e-06,
|
| 974 |
"loss": 0.3519,
|
| 975 |
"step": 685
|
| 976 |
},
|
| 977 |
{
|
| 978 |
"epoch": 0.93190770962296,
|
| 979 |
-
"grad_norm": 0.
|
| 980 |
"learning_rate": 6.44154595136111e-06,
|
| 981 |
-
"loss": 0.
|
| 982 |
"step": 690
|
| 983 |
},
|
| 984 |
{
|
| 985 |
"epoch": 0.9386606640405177,
|
| 986 |
-
"grad_norm": 0.
|
| 987 |
"learning_rate": 6.384986956932897e-06,
|
| 988 |
"loss": 0.3339,
|
| 989 |
"step": 695
|
| 990 |
},
|
| 991 |
{
|
| 992 |
"epoch": 0.9454136184580754,
|
| 993 |
-
"grad_norm": 0.
|
| 994 |
"learning_rate": 6.328235355554382e-06,
|
| 995 |
-
"loss": 0.
|
| 996 |
"step": 700
|
| 997 |
},
|
| 998 |
{
|
| 999 |
"epoch": 0.9521665728756331,
|
| 1000 |
-
"grad_norm": 0.
|
| 1001 |
"learning_rate": 6.271299039540228e-06,
|
| 1002 |
"loss": 0.3645,
|
| 1003 |
"step": 705
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
"epoch": 0.9589195272931907,
|
| 1007 |
-
"grad_norm": 0.
|
| 1008 |
"learning_rate": 6.214185926892936e-06,
|
| 1009 |
-
"loss": 0.
|
| 1010 |
"step": 710
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
"epoch": 0.9656724817107485,
|
| 1014 |
-
"grad_norm": 0.
|
| 1015 |
"learning_rate": 6.156903960201709e-06,
|
| 1016 |
-
"loss": 0.
|
| 1017 |
"step": 715
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.9724254361283061,
|
| 1021 |
-
"grad_norm": 0.
|
| 1022 |
"learning_rate": 6.099461105537889e-06,
|
| 1023 |
-
"loss": 0.
|
| 1024 |
"step": 720
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
"epoch": 0.9791783905458639,
|
| 1028 |
-
"grad_norm": 0.
|
| 1029 |
"learning_rate": 6.041865351347146e-06,
|
| 1030 |
"loss": 0.3421,
|
| 1031 |
"step": 725
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
"epoch": 0.9859313449634215,
|
| 1035 |
-
"grad_norm": 0.
|
| 1036 |
"learning_rate": 5.984124707338528e-06,
|
| 1037 |
-
"loss": 0.
|
| 1038 |
"step": 730
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
"epoch": 0.9926842993809791,
|
| 1042 |
-
"grad_norm": 0.
|
| 1043 |
"learning_rate": 5.926247203370583e-06,
|
| 1044 |
-
"loss": 0.
|
| 1045 |
"step": 735
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 0.9994372537985369,
|
| 1049 |
-
"grad_norm": 0.
|
| 1050 |
"learning_rate": 5.8682408883346535e-06,
|
| 1051 |
"loss": 0.2909,
|
| 1052 |
"step": 740
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 1.0,
|
| 1056 |
-
"eval_loss": 0.
|
| 1057 |
-
"eval_runtime": 18.
|
| 1058 |
-
"eval_samples_per_second": 19.
|
| 1059 |
-
"eval_steps_per_second": 4.
|
| 1060 |
"step": 741
|
| 1061 |
},
|
| 1062 |
{
|
| 1063 |
"epoch": 1.005402363534046,
|
| 1064 |
-
"grad_norm": 0.
|
| 1065 |
"learning_rate": 5.810113829035544e-06,
|
| 1066 |
-
"loss": 0.
|
| 1067 |
"step": 745
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
"epoch": 1.0121553179516039,
|
| 1071 |
-
"grad_norm": 0.
|
| 1072 |
"learning_rate": 5.751874109069685e-06,
|
| 1073 |
"loss": 0.3611,
|
| 1074 |
"step": 750
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 1.0189082723691616,
|
| 1078 |
-
"grad_norm": 0.
|
| 1079 |
"learning_rate": 5.693529827700967e-06,
|
| 1080 |
-
"loss": 0.
|
| 1081 |
"step": 755
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
"epoch": 1.0256612267867191,
|
| 1085 |
-
"grad_norm": 0.
|
| 1086 |
"learning_rate": 5.635089098734394e-06,
|
| 1087 |
-
"loss": 0.
|
| 1088 |
"step": 760
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
"epoch": 1.0324141812042769,
|
| 1092 |
-
"grad_norm": 0.
|
| 1093 |
"learning_rate": 5.576560049387717e-06,
|
| 1094 |
"loss": 0.339,
|
| 1095 |
"step": 765
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
"epoch": 1.0391671356218346,
|
| 1099 |
-
"grad_norm": 0.
|
| 1100 |
"learning_rate": 5.517950819161196e-06,
|
| 1101 |
"loss": 0.328,
|
| 1102 |
"step": 770
|
| 1103 |
},
|
| 1104 |
{
|
| 1105 |
"epoch": 1.0459200900393921,
|
| 1106 |
-
"grad_norm": 0.
|
| 1107 |
"learning_rate": 5.459269558705667e-06,
|
| 1108 |
-
"loss": 0.
|
| 1109 |
"step": 775
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
"epoch": 1.0526730444569499,
|
| 1113 |
-
"grad_norm": 0.
|
| 1114 |
"learning_rate": 5.400524428689035e-06,
|
| 1115 |
-
"loss": 0.
|
| 1116 |
"step": 780
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
"epoch": 1.0594259988745076,
|
| 1120 |
-
"grad_norm": 0.
|
| 1121 |
"learning_rate": 5.341723598661409e-06,
|
| 1122 |
"loss": 0.3267,
|
| 1123 |
"step": 785
|
| 1124 |
},
|
| 1125 |
{
|
| 1126 |
"epoch": 1.0661789532920654,
|
| 1127 |
-
"grad_norm": 0.
|
| 1128 |
"learning_rate": 5.282875245918963e-06,
|
| 1129 |
-
"loss": 0.
|
| 1130 |
"step": 790
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
"epoch": 1.0729319077096229,
|
| 1134 |
-
"grad_norm": 0.
|
| 1135 |
"learning_rate": 5.2239875543667465e-06,
|
| 1136 |
-
"loss": 0.
|
| 1137 |
"step": 795
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
"epoch": 1.0796848621271806,
|
| 1141 |
-
"grad_norm": 0.
|
| 1142 |
"learning_rate": 5.165068713380568e-06,
|
| 1143 |
-
"loss": 0.
|
| 1144 |
"step": 800
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
"epoch": 1.0864378165447384,
|
| 1148 |
-
"grad_norm": 0.
|
| 1149 |
"learning_rate": 5.106126916668118e-06,
|
| 1150 |
"loss": 0.3192,
|
| 1151 |
"step": 805
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
"epoch": 1.093190770962296,
|
| 1155 |
-
"grad_norm": 0.
|
| 1156 |
"learning_rate": 5.047170361129484e-06,
|
| 1157 |
-
"loss": 0.
|
| 1158 |
"step": 810
|
| 1159 |
},
|
| 1160 |
{
|
| 1161 |
"epoch": 1.0999437253798536,
|
| 1162 |
-
"grad_norm": 0.
|
| 1163 |
"learning_rate": 4.988207245717232e-06,
|
| 1164 |
"loss": 0.3558,
|
| 1165 |
"step": 815
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
"epoch": 1.1066966797974114,
|
| 1169 |
-
"grad_norm": 0.
|
| 1170 |
"learning_rate": 4.929245770296191e-06,
|
| 1171 |
-
"loss": 0.
|
| 1172 |
"step": 820
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
"epoch": 1.1134496342149691,
|
| 1176 |
-
"grad_norm": 0.
|
| 1177 |
"learning_rate": 4.870294134503123e-06,
|
| 1178 |
-
"loss": 0.
|
| 1179 |
"step": 825
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 1.1202025886325266,
|
| 1183 |
-
"grad_norm": 0.
|
| 1184 |
"learning_rate": 4.811360536606416e-06,
|
| 1185 |
-
"loss": 0.
|
| 1186 |
"step": 830
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
"epoch": 1.1269555430500844,
|
| 1190 |
-
"grad_norm": 0.
|
| 1191 |
"learning_rate": 4.752453172365966e-06,
|
| 1192 |
-
"loss": 0.
|
| 1193 |
"step": 835
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 1.1337084974676421,
|
| 1197 |
-
"grad_norm": 0.
|
| 1198 |
"learning_rate": 4.69358023389342e-06,
|
| 1199 |
-
"loss": 0.
|
| 1200 |
"step": 840
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 1.1404614518851997,
|
| 1204 |
-
"grad_norm": 0.
|
| 1205 |
"learning_rate": 4.634749908512907e-06,
|
| 1206 |
-
"loss": 0.
|
| 1207 |
"step": 845
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
"epoch": 1.1472144063027574,
|
| 1211 |
-
"grad_norm": 0.
|
| 1212 |
"learning_rate": 4.575970377622456e-06,
|
| 1213 |
-
"loss": 0.
|
| 1214 |
"step": 850
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
"epoch": 1.1539673607203151,
|
| 1218 |
-
"grad_norm": 0.
|
| 1219 |
"learning_rate": 4.517249815556219e-06,
|
| 1220 |
-
"loss": 0.
|
| 1221 |
"step": 855
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
"epoch": 1.1607203151378729,
|
| 1225 |
-
"grad_norm": 0.
|
| 1226 |
"learning_rate": 4.458596388447691e-06,
|
| 1227 |
-
"loss": 0.
|
| 1228 |
"step": 860
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
"epoch": 1.1674732695554304,
|
| 1232 |
-
"grad_norm": 0.
|
| 1233 |
"learning_rate": 4.400018253094065e-06,
|
| 1234 |
-
"loss": 0.
|
| 1235 |
"step": 865
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
"epoch": 1.1742262239729881,
|
| 1239 |
-
"grad_norm": 0.
|
| 1240 |
"learning_rate": 4.341523555821882e-06,
|
| 1241 |
-
"loss": 0.
|
| 1242 |
"step": 870
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 1.180979178390546,
|
| 1246 |
-
"grad_norm": 0.
|
| 1247 |
"learning_rate": 4.283120431354137e-06,
|
| 1248 |
-
"loss": 0.
|
| 1249 |
"step": 875
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 1.1877321328081036,
|
| 1253 |
-
"grad_norm": 0.
|
| 1254 |
"learning_rate": 4.224817001679011e-06,
|
| 1255 |
-
"loss": 0.
|
| 1256 |
"step": 880
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
"epoch": 1.1944850872256612,
|
| 1260 |
-
"grad_norm": 0.
|
| 1261 |
"learning_rate": 4.1666213749203545e-06,
|
| 1262 |
"loss": 0.3237,
|
| 1263 |
"step": 885
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
"epoch": 1.201238041643219,
|
| 1267 |
-
"grad_norm": 0.
|
| 1268 |
"learning_rate": 4.1085416442101205e-06,
|
| 1269 |
-
"loss": 0.
|
| 1270 |
"step": 890
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 1.2079909960607766,
|
| 1274 |
-
"grad_norm": 0.
|
| 1275 |
"learning_rate": 4.050585886562858e-06,
|
| 1276 |
-
"loss": 0.
|
| 1277 |
"step": 895
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
"epoch": 1.2147439504783342,
|
| 1281 |
-
"grad_norm": 0.
|
| 1282 |
"learning_rate": 3.992762161752474e-06,
|
| 1283 |
-
"loss": 0.
|
| 1284 |
"step": 900
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 1.221496904895892,
|
| 1288 |
-
"grad_norm": 0.
|
| 1289 |
"learning_rate": 3.935078511191368e-06,
|
| 1290 |
-
"loss": 0.
|
| 1291 |
"step": 905
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
"epoch": 1.2282498593134497,
|
| 1295 |
-
"grad_norm": 0.
|
| 1296 |
"learning_rate": 3.877542956812137e-06,
|
| 1297 |
"loss": 0.3221,
|
| 1298 |
"step": 910
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
"epoch": 1.2350028137310074,
|
| 1302 |
-
"grad_norm": 0.
|
| 1303 |
"learning_rate": 3.820163499951984e-06,
|
| 1304 |
"loss": 0.3579,
|
| 1305 |
"step": 915
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 1.241755768148565,
|
| 1309 |
-
"grad_norm": 0.
|
| 1310 |
"learning_rate": 3.7629481202399886e-06,
|
| 1311 |
-
"loss": 0.
|
| 1312 |
"step": 920
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 1.2485087225661227,
|
| 1316 |
-
"grad_norm": 0.
|
| 1317 |
"learning_rate": 3.705904774487396e-06,
|
| 1318 |
-
"loss": 0.
|
| 1319 |
"step": 925
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
"epoch": 1.2552616769836804,
|
| 1323 |
-
"grad_norm": 0.
|
| 1324 |
"learning_rate": 3.64904139558109e-06,
|
| 1325 |
-
"loss": 0.
|
| 1326 |
"step": 930
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
"epoch": 1.2620146314012382,
|
| 1330 |
-
"grad_norm": 0.
|
| 1331 |
"learning_rate": 3.5923658913803726e-06,
|
| 1332 |
-
"loss": 0.
|
| 1333 |
"step": 935
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
"epoch": 1.2687675858187957,
|
| 1337 |
-
"grad_norm": 0.
|
| 1338 |
"learning_rate": 3.5358861436172487e-06,
|
| 1339 |
-
"loss": 0.
|
| 1340 |
"step": 940
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 1.2755205402363534,
|
| 1344 |
-
"grad_norm": 0.
|
| 1345 |
"learning_rate": 3.47961000680032e-06,
|
| 1346 |
-
"loss": 0.
|
| 1347 |
"step": 945
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
"epoch": 1.2822734946539112,
|
| 1351 |
-
"grad_norm": 0.
|
| 1352 |
"learning_rate": 3.4235453071224882e-06,
|
| 1353 |
-
"loss": 0.
|
| 1354 |
"step": 950
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
"epoch": 1.2890264490714687,
|
| 1358 |
-
"grad_norm": 0.
|
| 1359 |
"learning_rate": 3.3676998413725726e-06,
|
| 1360 |
-
"loss": 0.
|
| 1361 |
"step": 955
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
"epoch": 1.2957794034890264,
|
| 1365 |
-
"grad_norm": 0.
|
| 1366 |
"learning_rate": 3.3120813758510385e-06,
|
| 1367 |
-
"loss": 0.
|
| 1368 |
"step": 960
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
"epoch": 1.3025323579065842,
|
| 1372 |
-
"grad_norm": 0.
|
| 1373 |
"learning_rate": 3.2566976452899507e-06,
|
| 1374 |
-
"loss": 0.
|
| 1375 |
"step": 965
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
"epoch": 1.3092853123241417,
|
| 1379 |
-
"grad_norm": 0.
|
| 1380 |
"learning_rate": 3.2015563517773214e-06,
|
| 1381 |
-
"loss": 0.
|
| 1382 |
"step": 970
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
"epoch": 1.3160382667416994,
|
| 1386 |
-
"grad_norm": 0.
|
| 1387 |
"learning_rate": 3.1466651636860025e-06,
|
| 1388 |
"loss": 0.3325,
|
| 1389 |
"step": 975
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
"epoch": 1.3227912211592572,
|
| 1393 |
-
"grad_norm": 0.
|
| 1394 |
"learning_rate": 3.0920317146072577e-06,
|
| 1395 |
-
"loss": 0.
|
| 1396 |
"step": 980
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
"epoch": 1.329544175576815,
|
| 1400 |
-
"grad_norm": 0.
|
| 1401 |
"learning_rate": 3.0376636022891813e-06,
|
| 1402 |
-
"loss": 0.
|
| 1403 |
"step": 985
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
"epoch": 1.3362971299943727,
|
| 1407 |
-
"grad_norm": 0.
|
| 1408 |
"learning_rate": 2.983568387580094e-06,
|
| 1409 |
-
"loss": 0.
|
| 1410 |
"step": 990
|
| 1411 |
},
|
| 1412 |
{
|
| 1413 |
"epoch": 1.3430500844119302,
|
| 1414 |
-
"grad_norm": 0.
|
| 1415 |
"learning_rate": 2.9297535933770732e-06,
|
| 1416 |
-
"loss": 0.
|
| 1417 |
"step": 995
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
"epoch": 1.349803038829488,
|
| 1421 |
-
"grad_norm": 0.
|
| 1422 |
"learning_rate": 2.8762267035797607e-06,
|
| 1423 |
-
"loss": 0.
|
| 1424 |
"step": 1000
|
| 1425 |
},
|
| 1426 |
{
|
| 1427 |
"epoch": 1.3565559932470457,
|
| 1428 |
-
"grad_norm": 0.
|
| 1429 |
"learning_rate": 2.822995162049599e-06,
|
| 1430 |
-
"loss": 0.
|
| 1431 |
"step": 1005
|
| 1432 |
},
|
| 1433 |
{
|
| 1434 |
"epoch": 1.3633089476646032,
|
| 1435 |
-
"grad_norm": 0.
|
| 1436 |
"learning_rate": 2.7700663715746213e-06,
|
| 1437 |
-
"loss": 0.
|
| 1438 |
"step": 1010
|
| 1439 |
},
|
| 1440 |
{
|
| 1441 |
"epoch": 1.370061902082161,
|
| 1442 |
-
"grad_norm": 0.
|
| 1443 |
"learning_rate": 2.7174476928399685e-06,
|
| 1444 |
-
"loss": 0.
|
| 1445 |
"step": 1015
|
| 1446 |
},
|
| 1447 |
{
|
| 1448 |
"epoch": 1.3768148564997187,
|
| 1449 |
-
"grad_norm": 0.
|
| 1450 |
"learning_rate": 2.66514644340426e-06,
|
| 1451 |
"loss": 0.3425,
|
| 1452 |
"step": 1020
|
| 1453 |
},
|
| 1454 |
{
|
| 1455 |
"epoch": 1.3835678109172762,
|
| 1456 |
-
"grad_norm": 0.
|
| 1457 |
"learning_rate": 2.613169896681949e-06,
|
| 1458 |
-
"loss": 0.
|
| 1459 |
"step": 1025
|
| 1460 |
},
|
| 1461 |
{
|
| 1462 |
"epoch": 1.390320765334834,
|
| 1463 |
-
"grad_norm": 0.
|
| 1464 |
"learning_rate": 2.5615252809318287e-06,
|
| 1465 |
-
"loss": 0.
|
| 1466 |
"step": 1030
|
| 1467 |
},
|
| 1468 |
{
|
| 1469 |
"epoch": 1.3970737197523917,
|
| 1470 |
-
"grad_norm": 0.
|
| 1471 |
"learning_rate": 2.5102197782518145e-06,
|
| 1472 |
-
"loss": 0.
|
| 1473 |
"step": 1035
|
| 1474 |
},
|
| 1475 |
{
|
| 1476 |
"epoch": 1.4038266741699492,
|
| 1477 |
-
"grad_norm": 0.
|
| 1478 |
"learning_rate": 2.4592605235801544e-06,
|
| 1479 |
-
"loss": 0.
|
| 1480 |
"step": 1040
|
| 1481 |
},
|
| 1482 |
{
|
| 1483 |
"epoch": 1.410579628587507,
|
| 1484 |
-
"grad_norm": 0.
|
| 1485 |
"learning_rate": 2.4086546037031734e-06,
|
| 1486 |
-
"loss": 0.
|
| 1487 |
"step": 1045
|
| 1488 |
},
|
| 1489 |
{
|
| 1490 |
"epoch": 1.4173325830050647,
|
| 1491 |
-
"grad_norm": 0.
|
| 1492 |
"learning_rate": 2.3584090562697427e-06,
|
| 1493 |
-
"loss": 0.
|
| 1494 |
"step": 1050
|
| 1495 |
},
|
| 1496 |
{
|
| 1497 |
"epoch": 1.4240855374226224,
|
| 1498 |
-
"grad_norm": 0.
|
| 1499 |
"learning_rate": 2.3085308688125695e-06,
|
| 1500 |
-
"loss": 0.
|
| 1501 |
"step": 1055
|
| 1502 |
},
|
| 1503 |
{
|
| 1504 |
"epoch": 1.4308384918401802,
|
| 1505 |
-
"grad_norm": 0.
|
| 1506 |
"learning_rate": 2.2590269777764516e-06,
|
| 1507 |
-
"loss": 0.
|
| 1508 |
"step": 1060
|
| 1509 |
},
|
| 1510 |
{
|
| 1511 |
"epoch": 1.4375914462577377,
|
| 1512 |
-
"grad_norm": 0.
|
| 1513 |
"learning_rate": 2.2099042675536437e-06,
|
| 1514 |
-
"loss": 0.
|
| 1515 |
"step": 1065
|
| 1516 |
},
|
| 1517 |
{
|
| 1518 |
"epoch": 1.4443444006752955,
|
| 1519 |
-
"grad_norm": 0.
|
| 1520 |
"learning_rate": 2.161169569526461e-06,
|
| 1521 |
-
"loss": 0.
|
| 1522 |
"step": 1070
|
| 1523 |
},
|
| 1524 |
{
|
| 1525 |
"epoch": 1.4510973550928532,
|
| 1526 |
-
"grad_norm": 0.
|
| 1527 |
"learning_rate": 2.1128296611172593e-06,
|
| 1528 |
-
"loss": 0.
|
| 1529 |
"step": 1075
|
| 1530 |
},
|
| 1531 |
{
|
| 1532 |
"epoch": 1.4578503095104107,
|
| 1533 |
-
"grad_norm": 0.
|
| 1534 |
"learning_rate": 2.0648912648459072e-06,
|
| 1535 |
"loss": 0.3447,
|
| 1536 |
"step": 1080
|
| 1537 |
},
|
| 1538 |
{
|
| 1539 |
"epoch": 1.4646032639279685,
|
| 1540 |
-
"grad_norm": 0.
|
| 1541 |
"learning_rate": 2.0173610473949048e-06,
|
| 1542 |
-
"loss": 0.
|
| 1543 |
"step": 1085
|
| 1544 |
},
|
| 1545 |
{
|
| 1546 |
"epoch": 1.4713562183455262,
|
| 1547 |
-
"grad_norm": 0.
|
| 1548 |
"learning_rate": 1.9702456186822595e-06,
|
| 1549 |
-
"loss": 0.
|
| 1550 |
"step": 1090
|
| 1551 |
},
|
| 1552 |
{
|
| 1553 |
"epoch": 1.4781091727630837,
|
| 1554 |
-
"grad_norm": 0.
|
| 1555 |
"learning_rate": 1.9235515309422685e-06,
|
| 1556 |
-
"loss": 0.
|
| 1557 |
"step": 1095
|
| 1558 |
},
|
| 1559 |
{
|
| 1560 |
"epoch": 1.4848621271806415,
|
| 1561 |
-
"grad_norm": 0.
|
| 1562 |
"learning_rate": 1.8772852778143064e-06,
|
| 1563 |
-
"loss": 0.
|
| 1564 |
"step": 1100
|
| 1565 |
},
|
| 1566 |
{
|
| 1567 |
"epoch": 1.4916150815981992,
|
| 1568 |
-
"grad_norm": 0.
|
| 1569 |
"learning_rate": 1.831453293439771e-06,
|
| 1570 |
-
"loss": 0.
|
| 1571 |
"step": 1105
|
| 1572 |
},
|
| 1573 |
{
|
| 1574 |
"epoch": 1.498368036015757,
|
| 1575 |
-
"grad_norm": 0.
|
| 1576 |
"learning_rate": 1.7860619515673034e-06,
|
| 1577 |
-
"loss": 0.
|
| 1578 |
"step": 1110
|
| 1579 |
},
|
| 1580 |
{
|
| 1581 |
"epoch": 1.5051209904333147,
|
| 1582 |
-
"grad_norm": 0.
|
| 1583 |
"learning_rate": 1.7411175646664103e-06,
|
| 1584 |
-
"loss": 0.
|
| 1585 |
"step": 1115
|
| 1586 |
},
|
| 1587 |
{
|
| 1588 |
"epoch": 1.5118739448508722,
|
| 1589 |
-
"grad_norm": 0.
|
| 1590 |
"learning_rate": 1.6966263830495939e-06,
|
| 1591 |
-
"loss": 0.
|
| 1592 |
"step": 1120
|
| 1593 |
},
|
| 1594 |
{
|
| 1595 |
"epoch": 1.51862689926843,
|
| 1596 |
-
"grad_norm": 0.
|
| 1597 |
"learning_rate": 1.6525945940031407e-06,
|
| 1598 |
"loss": 0.3692,
|
| 1599 |
"step": 1125
|
| 1600 |
},
|
| 1601 |
{
|
| 1602 |
"epoch": 1.5253798536859877,
|
| 1603 |
-
"grad_norm": 0.
|
| 1604 |
"learning_rate": 1.6090283209266682e-06,
|
| 1605 |
-
"loss": 0.
|
| 1606 |
"step": 1130
|
| 1607 |
},
|
| 1608 |
{
|
| 1609 |
"epoch": 1.5321328081035452,
|
| 1610 |
-
"grad_norm": 0.
|
| 1611 |
"learning_rate": 1.5659336224815642e-06,
|
| 1612 |
-
"loss": 0.
|
| 1613 |
"step": 1135
|
| 1614 |
},
|
| 1615 |
{
|
| 1616 |
"epoch": 1.538885762521103,
|
| 1617 |
-
"grad_norm": 0.
|
| 1618 |
"learning_rate": 1.5233164917484117e-06,
|
| 1619 |
-
"loss": 0.
|
| 1620 |
"step": 1140
|
| 1621 |
},
|
| 1622 |
{
|
| 1623 |
"epoch": 1.5456387169386607,
|
| 1624 |
-
"grad_norm": 0.
|
| 1625 |
"learning_rate": 1.4811828553935498e-06,
|
| 1626 |
-
"loss": 0.
|
| 1627 |
"step": 1145
|
| 1628 |
},
|
| 1629 |
{
|
| 1630 |
"epoch": 1.5523916713562183,
|
| 1631 |
-
"grad_norm": 0.
|
| 1632 |
"learning_rate": 1.439538572844873e-06,
|
| 1633 |
-
"loss": 0.
|
| 1634 |
"step": 1150
|
| 1635 |
},
|
| 1636 |
{
|
| 1637 |
"epoch": 1.559144625773776,
|
| 1638 |
-
"grad_norm": 0.
|
| 1639 |
"learning_rate": 1.3983894354769616e-06,
|
| 1640 |
-
"loss": 0.
|
| 1641 |
"step": 1155
|
| 1642 |
},
|
| 1643 |
{
|
| 1644 |
"epoch": 1.5658975801913337,
|
| 1645 |
-
"grad_norm": 0.
|
| 1646 |
"learning_rate": 1.3577411658056965e-06,
|
| 1647 |
-
"loss": 0.
|
| 1648 |
"step": 1160
|
| 1649 |
},
|
| 1650 |
{
|
| 1651 |
"epoch": 1.5726505346088913,
|
| 1652 |
-
"grad_norm": 0.
|
| 1653 |
"learning_rate": 1.3175994166924394e-06,
|
| 1654 |
-
"loss": 0.
|
| 1655 |
"step": 1165
|
| 1656 |
},
|
| 1657 |
{
|
| 1658 |
"epoch": 1.5794034890264492,
|
| 1659 |
-
"grad_norm": 0.
|
| 1660 |
"learning_rate": 1.2779697705579058e-06,
|
| 1661 |
-
"loss": 0.
|
| 1662 |
"step": 1170
|
| 1663 |
},
|
| 1664 |
{
|
| 1665 |
"epoch": 1.5861564434440067,
|
| 1666 |
-
"grad_norm": 0.
|
| 1667 |
"learning_rate": 1.2388577386058248e-06,
|
| 1668 |
-
"loss": 0.
|
| 1669 |
"step": 1175
|
| 1670 |
},
|
| 1671 |
{
|
| 1672 |
"epoch": 1.5929093978615643,
|
| 1673 |
-
"grad_norm": 0.
|
| 1674 |
"learning_rate": 1.2002687600565138e-06,
|
| 1675 |
-
"loss": 0.
|
| 1676 |
"step": 1180
|
| 1677 |
},
|
| 1678 |
{
|
| 1679 |
"epoch": 1.5996623522791222,
|
| 1680 |
-
"grad_norm": 0.
|
| 1681 |
"learning_rate": 1.1622082013904535e-06,
|
| 1682 |
-
"loss": 0.
|
| 1683 |
"step": 1185
|
| 1684 |
},
|
| 1685 |
{
|
| 1686 |
"epoch": 1.6064153066966798,
|
| 1687 |
-
"grad_norm": 0.
|
| 1688 |
"learning_rate": 1.1246813556019925e-06,
|
| 1689 |
-
"loss": 0.
|
| 1690 |
"step": 1190
|
| 1691 |
},
|
| 1692 |
{
|
| 1693 |
"epoch": 1.6131682611142375,
|
| 1694 |
-
"grad_norm": 0.
|
| 1695 |
"learning_rate": 1.0876934414632523e-06,
|
| 1696 |
-
"loss": 0.
|
| 1697 |
"step": 1195
|
| 1698 |
},
|
| 1699 |
{
|
| 1700 |
"epoch": 1.6199212155317952,
|
| 1701 |
-
"grad_norm": 0.
|
| 1702 |
"learning_rate": 1.0512496027983715e-06,
|
| 1703 |
-
"loss": 0.
|
| 1704 |
"step": 1200
|
| 1705 |
},
|
| 1706 |
{
|
| 1707 |
"epoch": 1.6266741699493528,
|
| 1708 |
-
"grad_norm": 0.
|
| 1709 |
"learning_rate": 1.0153549077681617e-06,
|
| 1710 |
-
"loss": 0.
|
| 1711 |
"step": 1205
|
| 1712 |
},
|
| 1713 |
{
|
| 1714 |
"epoch": 1.6334271243669105,
|
| 1715 |
-
"grad_norm": 0.
|
| 1716 |
"learning_rate": 9.80014348165298e-07,
|
| 1717 |
-
"loss": 0.
|
| 1718 |
"step": 1210
|
| 1719 |
},
|
| 1720 |
{
|
| 1721 |
"epoch": 1.6401800787844683,
|
| 1722 |
-
"grad_norm": 0.
|
| 1723 |
"learning_rate": 9.452328387201104e-07,
|
| 1724 |
-
"loss": 0.
|
| 1725 |
"step": 1215
|
| 1726 |
},
|
| 1727 |
{
|
| 1728 |
"epoch": 1.6469330332020258,
|
| 1729 |
-
"grad_norm": 0.
|
| 1730 |
"learning_rate": 9.110152164171127e-07,
|
| 1731 |
-
"loss": 0.
|
| 1732 |
"step": 1220
|
| 1733 |
},
|
| 1734 |
{
|
| 1735 |
"epoch": 1.6536859876195835,
|
| 1736 |
-
"grad_norm": 0.
|
| 1737 |
"learning_rate": 8.773662398223276e-07,
|
| 1738 |
-
"loss": 0.
|
| 1739 |
"step": 1225
|
| 1740 |
},
|
| 1741 |
{
|
| 1742 |
"epoch": 1.6604389420371413,
|
| 1743 |
-
"grad_norm": 0.
|
| 1744 |
"learning_rate": 8.44290588421533e-07,
|
| 1745 |
-
"loss": 0.
|
| 1746 |
"step": 1230
|
| 1747 |
},
|
| 1748 |
{
|
| 1749 |
"epoch": 1.6671918964546988,
|
| 1750 |
-
"grad_norm": 0.
|
| 1751 |
"learning_rate": 8.117928619694848e-07,
|
| 1752 |
-
"loss": 0.
|
| 1753 |
"step": 1235
|
| 1754 |
},
|
| 1755 |
{
|
| 1756 |
"epoch": 1.6739448508722568,
|
| 1757 |
-
"grad_norm": 0.
|
| 1758 |
"learning_rate": 7.798775798502484e-07,
|
| 1759 |
-
"loss": 0.
|
| 1760 |
"step": 1240
|
| 1761 |
},
|
| 1762 |
{
|
| 1763 |
"epoch": 1.6806978052898143,
|
| 1764 |
-
"grad_norm": 0.
|
| 1765 |
"learning_rate": 7.485491804486972e-07,
|
| 1766 |
"loss": 0.2951,
|
| 1767 |
"step": 1245
|
| 1768 |
},
|
| 1769 |
{
|
| 1770 |
"epoch": 1.687450759707372,
|
| 1771 |
-
"grad_norm": 0.
|
| 1772 |
"learning_rate": 7.178120205332717e-07,
|
| 1773 |
-
"loss": 0.
|
| 1774 |
"step": 1250
|
| 1775 |
},
|
| 1776 |
{
|
| 1777 |
"epoch": 1.6942037141249298,
|
| 1778 |
-
"grad_norm": 0.
|
| 1779 |
"learning_rate": 6.876703746500984e-07,
|
| 1780 |
-
"loss": 0.
|
| 1781 |
"step": 1255
|
| 1782 |
},
|
| 1783 |
{
|
| 1784 |
"epoch": 1.7009566685424873,
|
| 1785 |
-
"grad_norm": 0.
|
| 1786 |
"learning_rate": 6.581284345285371e-07,
|
| 1787 |
-
"loss": 0.
|
| 1788 |
"step": 1260
|
| 1789 |
},
|
| 1790 |
{
|
| 1791 |
"epoch": 1.707709622960045,
|
| 1792 |
-
"grad_norm": 0.
|
| 1793 |
"learning_rate": 6.291903084982481e-07,
|
| 1794 |
-
"loss": 0.
|
| 1795 |
"step": 1265
|
| 1796 |
},
|
| 1797 |
{
|
| 1798 |
"epoch": 1.7144625773776028,
|
| 1799 |
-
"grad_norm": 0.
|
| 1800 |
"learning_rate": 6.008600209178539e-07,
|
| 1801 |
-
"loss": 0.
|
| 1802 |
"step": 1270
|
| 1803 |
},
|
| 1804 |
{
|
| 1805 |
"epoch": 1.7212155317951603,
|
| 1806 |
-
"grad_norm": 0.
|
| 1807 |
"learning_rate": 5.7314151161528e-07,
|
| 1808 |
-
"loss": 0.
|
| 1809 |
"step": 1275
|
| 1810 |
},
|
| 1811 |
{
|
| 1812 |
"epoch": 1.727968486212718,
|
| 1813 |
-
"grad_norm": 0.
|
| 1814 |
"learning_rate": 5.460386353398583e-07,
|
| 1815 |
-
"loss": 0.
|
| 1816 |
"step": 1280
|
| 1817 |
},
|
| 1818 |
{
|
| 1819 |
"epoch": 1.7347214406302758,
|
| 1820 |
-
"grad_norm": 0.
|
| 1821 |
"learning_rate": 5.195551612262478e-07,
|
| 1822 |
-
"loss": 0.
|
| 1823 |
"step": 1285
|
| 1824 |
},
|
| 1825 |
{
|
| 1826 |
"epoch": 1.7414743950478333,
|
| 1827 |
-
"grad_norm": 0.
|
| 1828 |
"learning_rate": 4.936947722702762e-07,
|
| 1829 |
-
"loss": 0.
|
| 1830 |
"step": 1290
|
| 1831 |
},
|
| 1832 |
{
|
| 1833 |
"epoch": 1.7482273494653913,
|
| 1834 |
-
"grad_norm": 0.
|
| 1835 |
"learning_rate": 4.6846106481675035e-07,
|
| 1836 |
-
"loss": 0.
|
| 1837 |
"step": 1295
|
| 1838 |
},
|
| 1839 |
{
|
| 1840 |
"epoch": 1.7549803038829488,
|
| 1841 |
-
"grad_norm": 0.
|
| 1842 |
"learning_rate": 4.43857548059321e-07,
|
| 1843 |
-
"loss": 0.
|
| 1844 |
"step": 1300
|
| 1845 |
},
|
| 1846 |
{
|
| 1847 |
"epoch": 1.7617332583005063,
|
| 1848 |
-
"grad_norm": 0.
|
| 1849 |
"learning_rate": 4.198876435524718e-07,
|
| 1850 |
-
"loss": 0.
|
| 1851 |
"step": 1305
|
| 1852 |
},
|
| 1853 |
{
|
| 1854 |
"epoch": 1.7684862127180643,
|
| 1855 |
-
"grad_norm": 0.
|
| 1856 |
"learning_rate": 3.9655468473568436e-07,
|
| 1857 |
-
"loss": 0.
|
| 1858 |
"step": 1310
|
| 1859 |
},
|
| 1860 |
{
|
| 1861 |
"epoch": 1.7752391671356218,
|
| 1862 |
-
"grad_norm": 0.
|
| 1863 |
"learning_rate": 3.7386191646987094e-07,
|
| 1864 |
-
"loss": 0.
|
| 1865 |
"step": 1315
|
| 1866 |
},
|
| 1867 |
{
|
| 1868 |
"epoch": 1.7819921215531795,
|
| 1869 |
-
"grad_norm": 0.
|
| 1870 |
"learning_rate": 3.51812494586114e-07,
|
| 1871 |
-
"loss": 0.
|
| 1872 |
"step": 1320
|
| 1873 |
},
|
| 1874 |
{
|
| 1875 |
"epoch": 1.7887450759707373,
|
| 1876 |
-
"grad_norm": 0.
|
| 1877 |
"learning_rate": 3.3040948544679817e-07,
|
| 1878 |
-
"loss": 0.
|
| 1879 |
"step": 1325
|
| 1880 |
},
|
| 1881 |
{
|
| 1882 |
"epoch": 1.7954980303882948,
|
| 1883 |
-
"grad_norm": 0.
|
| 1884 |
"learning_rate": 3.096558655191706e-07,
|
| 1885 |
-
"loss": 0.
|
| 1886 |
"step": 1330
|
| 1887 |
},
|
| 1888 |
{
|
| 1889 |
"epoch": 1.8022509848058526,
|
| 1890 |
-
"grad_norm": 0.
|
| 1891 |
"learning_rate": 2.895545209614176e-07,
|
| 1892 |
-
"loss": 0.
|
| 1893 |
"step": 1335
|
| 1894 |
},
|
| 1895 |
{
|
| 1896 |
"epoch": 1.8090039392234103,
|
| 1897 |
-
"grad_norm": 0.
|
| 1898 |
"learning_rate": 2.701082472212879e-07,
|
| 1899 |
-
"loss": 0.
|
| 1900 |
"step": 1340
|
| 1901 |
},
|
| 1902 |
{
|
| 1903 |
"epoch": 1.8157568936409678,
|
| 1904 |
-
"grad_norm": 0.
|
| 1905 |
"learning_rate": 2.5131974864734063e-07,
|
| 1906 |
"loss": 0.3425,
|
| 1907 |
"step": 1345
|
| 1908 |
},
|
| 1909 |
{
|
| 1910 |
"epoch": 1.8225098480585256,
|
| 1911 |
-
"grad_norm": 0.
|
| 1912 |
"learning_rate": 2.331916381128535e-07,
|
| 1913 |
-
"loss": 0.
|
| 1914 |
"step": 1350
|
| 1915 |
},
|
| 1916 |
{
|
| 1917 |
"epoch": 1.8292628024760833,
|
| 1918 |
-
"grad_norm": 0.
|
| 1919 |
"learning_rate": 2.157264366524603e-07,
|
| 1920 |
-
"loss": 0.
|
| 1921 |
"step": 1355
|
| 1922 |
},
|
| 1923 |
{
|
| 1924 |
"epoch": 1.8360157568936408,
|
| 1925 |
-
"grad_norm": 0.
|
| 1926 |
"learning_rate": 1.989265731115525e-07,
|
| 1927 |
-
"loss": 0.
|
| 1928 |
"step": 1360
|
| 1929 |
},
|
| 1930 |
{
|
| 1931 |
"epoch": 1.8427687113111988,
|
| 1932 |
-
"grad_norm": 0.
|
| 1933 |
"learning_rate": 1.827943838085111e-07,
|
| 1934 |
"loss": 0.306,
|
| 1935 |
"step": 1365
|
| 1936 |
},
|
| 1937 |
{
|
| 1938 |
"epoch": 1.8495216657287563,
|
| 1939 |
-
"grad_norm": 0.
|
| 1940 |
"learning_rate": 1.6733211220979316e-07,
|
| 1941 |
-
"loss": 0.
|
| 1942 |
"step": 1370
|
| 1943 |
},
|
| 1944 |
{
|
| 1945 |
"epoch": 1.856274620146314,
|
| 1946 |
-
"grad_norm": 0.
|
| 1947 |
"learning_rate": 1.5254190861794415e-07,
|
| 1948 |
-
"loss": 0.
|
| 1949 |
"step": 1375
|
| 1950 |
},
|
| 1951 |
{
|
| 1952 |
"epoch": 1.8630275745638718,
|
| 1953 |
-
"grad_norm": 0.
|
| 1954 |
"learning_rate": 1.3842582987255494e-07,
|
| 1955 |
-
"loss": 0.
|
| 1956 |
"step": 1380
|
| 1957 |
},
|
| 1958 |
{
|
| 1959 |
"epoch": 1.8697805289814293,
|
| 1960 |
-
"grad_norm": 0.
|
| 1961 |
"learning_rate": 1.2498583906422779e-07,
|
| 1962 |
-
"loss": 0.
|
| 1963 |
"step": 1385
|
| 1964 |
},
|
| 1965 |
{
|
| 1966 |
"epoch": 1.876533483398987,
|
| 1967 |
-
"grad_norm": 0.
|
| 1968 |
"learning_rate": 1.1222380526156929e-07,
|
| 1969 |
-
"loss": 0.
|
| 1970 |
"step": 1390
|
| 1971 |
},
|
| 1972 |
{
|
| 1973 |
"epoch": 1.8832864378165448,
|
| 1974 |
-
"grad_norm": 0.
|
| 1975 |
"learning_rate": 1.0014150325126315e-07,
|
| 1976 |
-
"loss": 0.
|
| 1977 |
"step": 1395
|
| 1978 |
},
|
| 1979 |
{
|
| 1980 |
"epoch": 1.8900393922341023,
|
| 1981 |
-
"grad_norm": 0.
|
| 1982 |
"learning_rate": 8.874061329125939e-08,
|
| 1983 |
-
"loss": 0.
|
| 1984 |
"step": 1400
|
| 1985 |
},
|
| 1986 |
{
|
| 1987 |
"epoch": 1.89679234665166,
|
| 1988 |
-
"grad_norm": 0.
|
| 1989 |
"learning_rate": 7.802272087709951e-08,
|
| 1990 |
-
"loss": 0.
|
| 1991 |
"step": 1405
|
| 1992 |
},
|
| 1993 |
{
|
| 1994 |
"epoch": 1.9035453010692178,
|
| 1995 |
-
"grad_norm": 0.
|
| 1996 |
"learning_rate": 6.798931652142738e-08,
|
| 1997 |
-
"loss": 0.
|
| 1998 |
"step": 1410
|
| 1999 |
},
|
| 2000 |
{
|
| 2001 |
"epoch": 1.9102982554867753,
|
| 2002 |
-
"grad_norm": 0.
|
| 2003 |
"learning_rate": 5.864179554670724e-08,
|
| 2004 |
-
"loss": 0.
|
| 2005 |
"step": 1415
|
| 2006 |
},
|
| 2007 |
{
|
| 2008 |
"epoch": 1.9170512099043333,
|
| 2009 |
-
"grad_norm": 0.
|
| 2010 |
"learning_rate": 4.998145789118114e-08,
|
| 2011 |
-
"loss": 0.
|
| 2012 |
"step": 1420
|
| 2013 |
},
|
| 2014 |
{
|
| 2015 |
"epoch": 1.9238041643218908,
|
| 2016 |
-
"grad_norm": 0.
|
| 2017 |
"learning_rate": 4.2009507928084116e-08,
|
| 2018 |
-
"loss": 0.
|
| 2019 |
"step": 1425
|
| 2020 |
},
|
| 2021 |
{
|
| 2022 |
"epoch": 1.9305571187394484,
|
| 2023 |
-
"grad_norm": 0.
|
| 2024 |
"learning_rate": 3.4727054298161475e-08,
|
| 2025 |
-
"loss": 0.
|
| 2026 |
"step": 1430
|
| 2027 |
},
|
| 2028 |
{
|
| 2029 |
"epoch": 1.9373100731570063,
|
| 2030 |
-
"grad_norm": 0.
|
| 2031 |
"learning_rate": 2.8135109755487723e-08,
|
| 2032 |
-
"loss": 0.
|
| 2033 |
"step": 1435
|
| 2034 |
},
|
| 2035 |
{
|
| 2036 |
"epoch": 1.9440630275745638,
|
| 2037 |
-
"grad_norm": 0.
|
| 2038 |
"learning_rate": 2.223459102662695e-08,
|
| 2039 |
-
"loss": 0.
|
| 2040 |
"step": 1440
|
| 2041 |
},
|
| 2042 |
{
|
| 2043 |
"epoch": 1.9508159819921216,
|
| 2044 |
-
"grad_norm": 0.
|
| 2045 |
"learning_rate": 1.7026318683147082e-08,
|
| 2046 |
-
"loss": 0.
|
| 2047 |
"step": 1445
|
| 2048 |
},
|
| 2049 |
{
|
| 2050 |
"epoch": 1.9575689364096793,
|
| 2051 |
-
"grad_norm": 0.
|
| 2052 |
"learning_rate": 1.2511017027501682e-08,
|
| 2053 |
"loss": 0.3032,
|
| 2054 |
"step": 1450
|
| 2055 |
},
|
| 2056 |
{
|
| 2057 |
"epoch": 1.9643218908272368,
|
| 2058 |
-
"grad_norm": 0.
|
| 2059 |
"learning_rate": 8.689313992306104e-09,
|
| 2060 |
-
"loss": 0.
|
| 2061 |
"step": 1455
|
| 2062 |
},
|
| 2063 |
{
|
| 2064 |
"epoch": 1.9710748452447946,
|
| 2065 |
-
"grad_norm": 0.
|
| 2066 |
"learning_rate": 5.561741053010661e-09,
|
| 2067 |
-
"loss": 0.
|
| 2068 |
"step": 1460
|
| 2069 |
},
|
| 2070 |
{
|
| 2071 |
"epoch": 1.9778277996623523,
|
| 2072 |
-
"grad_norm": 0.
|
| 2073 |
"learning_rate": 3.1287331539903155e-09,
|
| 2074 |
-
"loss": 0.
|
| 2075 |
"step": 1465
|
| 2076 |
},
|
| 2077 |
{
|
| 2078 |
"epoch": 1.9845807540799099,
|
| 2079 |
-
"grad_norm": 0.
|
| 2080 |
"learning_rate": 1.3906286480563913e-09,
|
| 2081 |
-
"loss": 0.
|
| 2082 |
"step": 1470
|
| 2083 |
},
|
| 2084 |
{
|
| 2085 |
"epoch": 1.9913337084974676,
|
| 2086 |
-
"grad_norm": 0.
|
| 2087 |
"learning_rate": 3.4766924940476954e-10,
|
| 2088 |
-
"loss": 0.
|
| 2089 |
"step": 1475
|
| 2090 |
},
|
| 2091 |
{
|
| 2092 |
"epoch": 1.9980866629150253,
|
| 2093 |
-
"grad_norm": 0.
|
| 2094 |
"learning_rate": 0.0,
|
| 2095 |
-
"loss": 0.
|
| 2096 |
"step": 1480
|
| 2097 |
},
|
| 2098 |
{
|
| 2099 |
"epoch": 1.9980866629150253,
|
| 2100 |
-
"eval_loss": 0.
|
| 2101 |
-
"eval_runtime": 17.
|
| 2102 |
-
"eval_samples_per_second": 20.
|
| 2103 |
-
"eval_steps_per_second": 5.
|
| 2104 |
"step": 1480
|
| 2105 |
}
|
| 2106 |
],
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0013505908835115363,
|
| 13 |
+
"grad_norm": 2.148791476511835,
|
| 14 |
"learning_rate": 6.756756756756757e-08,
|
| 15 |
"loss": 2.1725,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.006752954417557681,
|
| 20 |
+
"grad_norm": 2.0693435263055955,
|
| 21 |
"learning_rate": 3.378378378378379e-07,
|
| 22 |
+
"loss": 2.1697,
|
| 23 |
"step": 5
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.013505908835115363,
|
| 27 |
+
"grad_norm": 1.698535954485606,
|
| 28 |
"learning_rate": 6.756756756756758e-07,
|
| 29 |
+
"loss": 2.0659,
|
| 30 |
"step": 10
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.020258863252673044,
|
| 34 |
+
"grad_norm": 0.9452543971000898,
|
| 35 |
"learning_rate": 1.0135135135135136e-06,
|
| 36 |
+
"loss": 1.8928,
|
| 37 |
"step": 15
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.027011817670230726,
|
| 41 |
+
"grad_norm": 0.8038988537572226,
|
| 42 |
"learning_rate": 1.3513513513513515e-06,
|
| 43 |
+
"loss": 1.5709,
|
| 44 |
"step": 20
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.03376477208778841,
|
| 48 |
+
"grad_norm": 0.471508019359059,
|
| 49 |
"learning_rate": 1.6891891891891894e-06,
|
| 50 |
+
"loss": 1.1256,
|
| 51 |
"step": 25
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.04051772650534609,
|
| 55 |
+
"grad_norm": 0.252352654342694,
|
| 56 |
"learning_rate": 2.0270270270270273e-06,
|
| 57 |
+
"loss": 0.9097,
|
| 58 |
"step": 30
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.04727068092290377,
|
| 62 |
+
"grad_norm": 0.24974204597161465,
|
| 63 |
"learning_rate": 2.364864864864865e-06,
|
| 64 |
+
"loss": 0.7799,
|
| 65 |
"step": 35
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.05402363534046145,
|
| 69 |
+
"grad_norm": 0.20795816160607852,
|
| 70 |
"learning_rate": 2.702702702702703e-06,
|
| 71 |
+
"loss": 0.7137,
|
| 72 |
"step": 40
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.06077658975801913,
|
| 76 |
+
"grad_norm": 0.17844205306863006,
|
| 77 |
"learning_rate": 3.040540540540541e-06,
|
| 78 |
+
"loss": 0.5982,
|
| 79 |
"step": 45
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.06752954417557681,
|
| 83 |
+
"grad_norm": 0.14449057063114276,
|
| 84 |
"learning_rate": 3.3783783783783788e-06,
|
| 85 |
+
"loss": 0.5388,
|
| 86 |
"step": 50
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.0742824985931345,
|
| 90 |
+
"grad_norm": 0.12180944441742095,
|
| 91 |
"learning_rate": 3.7162162162162162e-06,
|
| 92 |
"loss": 0.5621,
|
| 93 |
"step": 55
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.08103545301069218,
|
| 97 |
+
"grad_norm": 0.11672624417203623,
|
| 98 |
"learning_rate": 4.0540540540540545e-06,
|
| 99 |
+
"loss": 0.4885,
|
| 100 |
"step": 60
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.08778840742824986,
|
| 104 |
+
"grad_norm": 0.12365779514235553,
|
| 105 |
"learning_rate": 4.391891891891892e-06,
|
| 106 |
+
"loss": 0.4725,
|
| 107 |
"step": 65
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.09454136184580754,
|
| 111 |
+
"grad_norm": 0.13848942929552646,
|
| 112 |
"learning_rate": 4.72972972972973e-06,
|
| 113 |
+
"loss": 0.4737,
|
| 114 |
"step": 70
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.10129431626336523,
|
| 118 |
+
"grad_norm": 0.13947287893715526,
|
| 119 |
"learning_rate": 5.067567567567568e-06,
|
| 120 |
+
"loss": 0.4686,
|
| 121 |
"step": 75
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.1080472706809229,
|
| 125 |
+
"grad_norm": 0.10498499488363304,
|
| 126 |
"learning_rate": 5.405405405405406e-06,
|
| 127 |
+
"loss": 0.4607,
|
| 128 |
"step": 80
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.11480022509848059,
|
| 132 |
+
"grad_norm": 0.1263112338839764,
|
| 133 |
"learning_rate": 5.743243243243244e-06,
|
| 134 |
"loss": 0.471,
|
| 135 |
"step": 85
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.12155317951603826,
|
| 139 |
+
"grad_norm": 0.1082136353357591,
|
| 140 |
"learning_rate": 6.081081081081082e-06,
|
| 141 |
+
"loss": 0.432,
|
| 142 |
"step": 90
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.12830613393359594,
|
| 146 |
+
"grad_norm": 0.08923160665434038,
|
| 147 |
"learning_rate": 6.41891891891892e-06,
|
| 148 |
+
"loss": 0.4433,
|
| 149 |
"step": 95
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.13505908835115363,
|
| 153 |
+
"grad_norm": 0.11807106476588053,
|
| 154 |
"learning_rate": 6.7567567567567575e-06,
|
| 155 |
"loss": 0.4398,
|
| 156 |
"step": 100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.14181204276871132,
|
| 160 |
+
"grad_norm": 0.09395401671145927,
|
| 161 |
"learning_rate": 7.0945945945945946e-06,
|
| 162 |
"loss": 0.3899,
|
| 163 |
"step": 105
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.148564997186269,
|
| 167 |
+
"grad_norm": 0.09428015614908754,
|
| 168 |
"learning_rate": 7.4324324324324324e-06,
|
| 169 |
+
"loss": 0.3983,
|
| 170 |
"step": 110
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.15531795160382666,
|
| 174 |
+
"grad_norm": 0.08605902096557952,
|
| 175 |
"learning_rate": 7.77027027027027e-06,
|
| 176 |
+
"loss": 0.415,
|
| 177 |
"step": 115
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.16207090602138435,
|
| 181 |
+
"grad_norm": 0.08713323888522238,
|
| 182 |
"learning_rate": 8.108108108108109e-06,
|
| 183 |
+
"loss": 0.4075,
|
| 184 |
"step": 120
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.16882386043894204,
|
| 188 |
+
"grad_norm": 0.07858143188282206,
|
| 189 |
"learning_rate": 8.445945945945948e-06,
|
| 190 |
+
"loss": 0.4089,
|
| 191 |
"step": 125
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.17557681485649973,
|
| 195 |
+
"grad_norm": 0.07427563021554226,
|
| 196 |
"learning_rate": 8.783783783783785e-06,
|
| 197 |
+
"loss": 0.4291,
|
| 198 |
"step": 130
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.1823297692740574,
|
| 202 |
+
"grad_norm": 0.09144835191366023,
|
| 203 |
"learning_rate": 9.121621621621622e-06,
|
| 204 |
"loss": 0.3895,
|
| 205 |
"step": 135
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.18908272369161508,
|
| 209 |
+
"grad_norm": 0.07052586919720526,
|
| 210 |
"learning_rate": 9.45945945945946e-06,
|
| 211 |
"loss": 0.4029,
|
| 212 |
"step": 140
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.19583567810917277,
|
| 216 |
+
"grad_norm": 0.0769561636008079,
|
| 217 |
"learning_rate": 9.797297297297298e-06,
|
| 218 |
+
"loss": 0.4148,
|
| 219 |
"step": 145
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.20258863252673046,
|
| 223 |
+
"grad_norm": 0.07216832297932871,
|
| 224 |
"learning_rate": 9.999944372378571e-06,
|
| 225 |
+
"loss": 0.3801,
|
| 226 |
"step": 150
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.20934158694428812,
|
| 230 |
+
"grad_norm": 0.06883498336890326,
|
| 231 |
"learning_rate": 9.999318575852451e-06,
|
| 232 |
+
"loss": 0.3728,
|
| 233 |
"step": 155
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.2160945413618458,
|
| 237 |
+
"grad_norm": 0.07007134671747867,
|
| 238 |
"learning_rate": 9.99799753559161e-06,
|
| 239 |
+
"loss": 0.3831,
|
| 240 |
"step": 160
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.2228474957794035,
|
| 244 |
+
"grad_norm": 0.0638944841323381,
|
| 245 |
"learning_rate": 9.995981435310078e-06,
|
| 246 |
+
"loss": 0.3852,
|
| 247 |
"step": 165
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.22960045019696118,
|
| 251 |
+
"grad_norm": 0.0732472641559478,
|
| 252 |
"learning_rate": 9.993270555382283e-06,
|
| 253 |
"loss": 0.4058,
|
| 254 |
"step": 170
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.23635340461451884,
|
| 258 |
+
"grad_norm": 0.06688107743191332,
|
| 259 |
"learning_rate": 9.989865272804064e-06,
|
| 260 |
+
"loss": 0.3981,
|
| 261 |
"step": 175
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.24310635903207653,
|
| 265 |
+
"grad_norm": 0.0750542374720311,
|
| 266 |
"learning_rate": 9.985766061140233e-06,
|
| 267 |
+
"loss": 0.4034,
|
| 268 |
"step": 180
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.24985931344963422,
|
| 272 |
+
"grad_norm": 0.06261496896671846,
|
| 273 |
"learning_rate": 9.980973490458728e-06,
|
| 274 |
+
"loss": 0.3611,
|
| 275 |
"step": 185
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.2566122678671919,
|
| 279 |
+
"grad_norm": 0.07059993646852766,
|
| 280 |
"learning_rate": 9.97548822725133e-06,
|
| 281 |
+
"loss": 0.3603,
|
| 282 |
"step": 190
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.26336522228474957,
|
| 286 |
+
"grad_norm": 0.06560938196431165,
|
| 287 |
"learning_rate": 9.969311034340977e-06,
|
| 288 |
+
"loss": 0.3853,
|
| 289 |
"step": 195
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.27011817670230726,
|
| 293 |
+
"grad_norm": 0.05548643801780881,
|
| 294 |
"learning_rate": 9.962442770775675e-06,
|
| 295 |
+
"loss": 0.3659,
|
| 296 |
"step": 200
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.27687113111986494,
|
| 300 |
+
"grad_norm": 0.06003482909108544,
|
| 301 |
"learning_rate": 9.954884391709043e-06,
|
| 302 |
+
"loss": 0.3907,
|
| 303 |
"step": 205
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.28362408553742263,
|
| 307 |
+
"grad_norm": 0.04978856362198666,
|
| 308 |
"learning_rate": 9.946636948267468e-06,
|
| 309 |
"loss": 0.349,
|
| 310 |
"step": 210
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.2903770399549803,
|
| 314 |
+
"grad_norm": 0.056553641270721766,
|
| 315 |
"learning_rate": 9.937701587403941e-06,
|
| 316 |
+
"loss": 0.4041,
|
| 317 |
"step": 215
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.297129994372538,
|
| 321 |
+
"grad_norm": 0.06048923977199324,
|
| 322 |
"learning_rate": 9.928079551738542e-06,
|
| 323 |
+
"loss": 0.3704,
|
| 324 |
"step": 220
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.30388294879009564,
|
| 328 |
+
"grad_norm": 0.055403005803656744,
|
| 329 |
"learning_rate": 9.91777217938564e-06,
|
| 330 |
+
"loss": 0.3765,
|
| 331 |
"step": 225
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.31063590320765333,
|
| 335 |
+
"grad_norm": 0.05878900335636033,
|
| 336 |
"learning_rate": 9.906780903767799e-06,
|
| 337 |
+
"loss": 0.3927,
|
| 338 |
"step": 230
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.317388857625211,
|
| 342 |
+
"grad_norm": 0.05009538620005328,
|
| 343 |
"learning_rate": 9.895107253416434e-06,
|
| 344 |
+
"loss": 0.3826,
|
| 345 |
"step": 235
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.3241418120427687,
|
| 349 |
+
"grad_norm": 0.06193069860962348,
|
| 350 |
"learning_rate": 9.882752851759247e-06,
|
| 351 |
+
"loss": 0.3694,
|
| 352 |
"step": 240
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.3308947664603264,
|
| 356 |
+
"grad_norm": 0.05658978100208689,
|
| 357 |
"learning_rate": 9.869719416894462e-06,
|
| 358 |
+
"loss": 0.3729,
|
| 359 |
"step": 245
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.3376477208778841,
|
| 363 |
+
"grad_norm": 0.060376802387617515,
|
| 364 |
"learning_rate": 9.856008761351882e-06,
|
| 365 |
+
"loss": 0.373,
|
| 366 |
"step": 250
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 0.34440067529544177,
|
| 370 |
+
"grad_norm": 0.047418407201082834,
|
| 371 |
"learning_rate": 9.841622791840839e-06,
|
| 372 |
+
"loss": 0.3343,
|
| 373 |
"step": 255
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.35115362971299946,
|
| 377 |
+
"grad_norm": 0.062039985277230605,
|
| 378 |
"learning_rate": 9.826563508985017e-06,
|
| 379 |
+
"loss": 0.3897,
|
| 380 |
"step": 260
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.3579065841305571,
|
| 384 |
+
"grad_norm": 0.04965955489714088,
|
| 385 |
"learning_rate": 9.810833007044247e-06,
|
| 386 |
"loss": 0.3925,
|
| 387 |
"step": 265
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.3646595385481148,
|
| 391 |
+
"grad_norm": 0.05316606675382464,
|
| 392 |
"learning_rate": 9.794433473623249e-06,
|
| 393 |
+
"loss": 0.377,
|
| 394 |
"step": 270
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 0.37141249296567247,
|
| 398 |
+
"grad_norm": 0.0467476855297103,
|
| 399 |
"learning_rate": 9.777367189367412e-06,
|
| 400 |
"loss": 0.3779,
|
| 401 |
"step": 275
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.37816544738323016,
|
| 405 |
+
"grad_norm": 0.04978059836724344,
|
| 406 |
"learning_rate": 9.759636527645633e-06,
|
| 407 |
+
"loss": 0.3521,
|
| 408 |
"step": 280
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.38491840180078785,
|
| 412 |
+
"grad_norm": 0.05999211649054394,
|
| 413 |
"learning_rate": 9.74124395422025e-06,
|
| 414 |
"loss": 0.3725,
|
| 415 |
"step": 285
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.39167135621834553,
|
| 419 |
+
"grad_norm": 0.04365453179095536,
|
| 420 |
"learning_rate": 9.722192026904145e-06,
|
| 421 |
+
"loss": 0.3402,
|
| 422 |
"step": 290
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.3984243106359032,
|
| 426 |
+
"grad_norm": 0.04616864773915026,
|
| 427 |
"learning_rate": 9.702483395205023e-06,
|
| 428 |
+
"loss": 0.3655,
|
| 429 |
"step": 295
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.4051772650534609,
|
| 433 |
+
"grad_norm": 0.04782080283525874,
|
| 434 |
"learning_rate": 9.682120799956961e-06,
|
| 435 |
+
"loss": 0.3263,
|
| 436 |
"step": 300
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 0.41193021947101854,
|
| 440 |
+
"grad_norm": 0.0498656978185221,
|
| 441 |
"learning_rate": 9.661107072939244e-06,
|
| 442 |
+
"loss": 0.3387,
|
| 443 |
"step": 305
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.41868317388857623,
|
| 447 |
+
"grad_norm": 0.04142545069347494,
|
| 448 |
"learning_rate": 9.639445136482549e-06,
|
| 449 |
+
"loss": 0.3388,
|
| 450 |
"step": 310
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.4254361283061339,
|
| 454 |
+
"grad_norm": 0.04183998822986712,
|
| 455 |
"learning_rate": 9.61713800306255e-06,
|
| 456 |
+
"loss": 0.3615,
|
| 457 |
"step": 315
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.4321890827236916,
|
| 461 |
+
"grad_norm": 0.045933316046675596,
|
| 462 |
"learning_rate": 9.594188774880981e-06,
|
| 463 |
+
"loss": 0.3607,
|
| 464 |
"step": 320
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.4389420371412493,
|
| 468 |
+
"grad_norm": 0.04805726955974,
|
| 469 |
"learning_rate": 9.570600643434217e-06,
|
| 470 |
"loss": 0.3371,
|
| 471 |
"step": 325
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.445694991558807,
|
| 475 |
+
"grad_norm": 0.04569635226134762,
|
| 476 |
"learning_rate": 9.546376889069442e-06,
|
| 477 |
+
"loss": 0.3582,
|
| 478 |
"step": 330
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 0.4524479459763647,
|
| 482 |
+
"grad_norm": 0.050347567122321094,
|
| 483 |
"learning_rate": 9.521520880528453e-06,
|
| 484 |
+
"loss": 0.3831,
|
| 485 |
"step": 335
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.45920090039392236,
|
| 489 |
+
"grad_norm": 0.051704382901289696,
|
| 490 |
"learning_rate": 9.496036074479184e-06,
|
| 491 |
"loss": 0.3604,
|
| 492 |
"step": 340
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 0.46595385481148,
|
| 496 |
+
"grad_norm": 0.04717995332220621,
|
| 497 |
"learning_rate": 9.46992601503499e-06,
|
| 498 |
+
"loss": 0.3498,
|
| 499 |
"step": 345
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.4727068092290377,
|
| 503 |
+
"grad_norm": 0.04394756827213909,
|
| 504 |
"learning_rate": 9.44319433326178e-06,
|
| 505 |
+
"loss": 0.3711,
|
| 506 |
"step": 350
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.47945976364659537,
|
| 510 |
+
"grad_norm": 0.05455789140928886,
|
| 511 |
"learning_rate": 9.415844746673047e-06,
|
| 512 |
+
"loss": 0.3395,
|
| 513 |
"step": 355
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.48621271806415306,
|
| 517 |
+
"grad_norm": 0.0486039405429815,
|
| 518 |
"learning_rate": 9.387881058712888e-06,
|
| 519 |
+
"loss": 0.3596,
|
| 520 |
"step": 360
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 0.49296567248171075,
|
| 524 |
+
"grad_norm": 0.049127971484592976,
|
| 525 |
"learning_rate": 9.359307158227067e-06,
|
| 526 |
+
"loss": 0.3759,
|
| 527 |
"step": 365
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.49971862689926844,
|
| 531 |
+
"grad_norm": 0.05804448606366045,
|
| 532 |
"learning_rate": 9.330127018922195e-06,
|
| 533 |
+
"loss": 0.3741,
|
| 534 |
"step": 370
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.5064715813168261,
|
| 538 |
+
"grad_norm": 0.05121055195790688,
|
| 539 |
"learning_rate": 9.300344698813124e-06,
|
| 540 |
+
"loss": 0.3978,
|
| 541 |
"step": 375
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.5132245357343838,
|
| 545 |
+
"grad_norm": 0.05072015885221285,
|
| 546 |
"learning_rate": 9.269964339658605e-06,
|
| 547 |
+
"loss": 0.349,
|
| 548 |
"step": 380
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.5199774901519415,
|
| 552 |
+
"grad_norm": 0.0446955462817659,
|
| 553 |
"learning_rate": 9.238990166385304e-06,
|
| 554 |
+
"loss": 0.3783,
|
| 555 |
"step": 385
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.5267304445694991,
|
| 559 |
+
"grad_norm": 0.047869039864268624,
|
| 560 |
"learning_rate": 9.207426486500252e-06,
|
| 561 |
"loss": 0.3587,
|
| 562 |
"step": 390
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 0.5334833989870569,
|
| 566 |
+
"grad_norm": 0.04183937339521655,
|
| 567 |
"learning_rate": 9.175277689491804e-06,
|
| 568 |
+
"loss": 0.3385,
|
| 569 |
"step": 395
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.5402363534046145,
|
| 573 |
+
"grad_norm": 0.04804994911136886,
|
| 574 |
"learning_rate": 9.142548246219212e-06,
|
| 575 |
+
"loss": 0.3696,
|
| 576 |
"step": 400
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 0.5469893078221723,
|
| 580 |
+
"grad_norm": 0.056116289553118946,
|
| 581 |
"learning_rate": 9.109242708290864e-06,
|
| 582 |
"loss": 0.3427,
|
| 583 |
"step": 405
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.5537422622397299,
|
| 587 |
+
"grad_norm": 0.04610990276019498,
|
| 588 |
"learning_rate": 9.075365707431311e-06,
|
| 589 |
+
"loss": 0.3684,
|
| 590 |
"step": 410
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.5604952166572875,
|
| 594 |
+
"grad_norm": 0.04711791241060553,
|
| 595 |
"learning_rate": 9.040921954837139e-06,
|
| 596 |
+
"loss": 0.3262,
|
| 597 |
"step": 415
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.5672481710748453,
|
| 601 |
+
"grad_norm": 0.0436305028584867,
|
| 602 |
"learning_rate": 9.005916240521788e-06,
|
| 603 |
+
"loss": 0.3466,
|
| 604 |
"step": 420
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 0.5740011254924029,
|
| 608 |
+
"grad_norm": 0.043193879987331325,
|
| 609 |
"learning_rate": 8.97035343264943e-06,
|
| 610 |
+
"loss": 0.3213,
|
| 611 |
"step": 425
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.5807540799099606,
|
| 615 |
+
"grad_norm": 0.04807145092887773,
|
| 616 |
"learning_rate": 8.93423847685795e-06,
|
| 617 |
+
"loss": 0.3553,
|
| 618 |
"step": 430
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 0.5875070343275183,
|
| 622 |
+
"grad_norm": 0.04707322735993806,
|
| 623 |
"learning_rate": 8.89757639557118e-06,
|
| 624 |
+
"loss": 0.3611,
|
| 625 |
"step": 435
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.594259988745076,
|
| 629 |
+
"grad_norm": 0.04694876715909799,
|
| 630 |
"learning_rate": 8.860372287300432e-06,
|
| 631 |
"loss": 0.3703,
|
| 632 |
"step": 440
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 0.6010129431626337,
|
| 636 |
+
"grad_norm": 0.051713244981409834,
|
| 637 |
"learning_rate": 8.822631325935463e-06,
|
| 638 |
+
"loss": 0.3763,
|
| 639 |
"step": 445
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.6077658975801913,
|
| 643 |
+
"grad_norm": 0.07337197692235135,
|
| 644 |
"learning_rate": 8.78435876002496e-06,
|
| 645 |
+
"loss": 0.3867,
|
| 646 |
"step": 450
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 0.614518851997749,
|
| 650 |
+
"grad_norm": 0.03984879157392941,
|
| 651 |
"learning_rate": 8.745559912046625e-06,
|
| 652 |
+
"loss": 0.3292,
|
| 653 |
"step": 455
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.6212718064153067,
|
| 657 |
+
"grad_norm": 0.042340122733838116,
|
| 658 |
"learning_rate": 8.706240177667003e-06,
|
| 659 |
"loss": 0.3476,
|
| 660 |
"step": 460
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 0.6280247608328644,
|
| 664 |
+
"grad_norm": 0.04565714030234626,
|
| 665 |
"learning_rate": 8.666405024991105e-06,
|
| 666 |
+
"loss": 0.3361,
|
| 667 |
"step": 465
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.634777715250422,
|
| 671 |
+
"grad_norm": 0.04169016093573778,
|
| 672 |
"learning_rate": 8.626059993801988e-06,
|
| 673 |
+
"loss": 0.339,
|
| 674 |
"step": 470
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.6415306696679798,
|
| 678 |
+
"grad_norm": 0.04325992397568874,
|
| 679 |
"learning_rate": 8.585210694790333e-06,
|
| 680 |
+
"loss": 0.3366,
|
| 681 |
"step": 475
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.6482836240855374,
|
| 685 |
+
"grad_norm": 0.0415410087146755,
|
| 686 |
"learning_rate": 8.543862808774193e-06,
|
| 687 |
+
"loss": 0.3489,
|
| 688 |
"step": 480
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 0.6550365785030952,
|
| 692 |
+
"grad_norm": 0.04286421536314813,
|
| 693 |
"learning_rate": 8.502022085908963e-06,
|
| 694 |
"loss": 0.3471,
|
| 695 |
"step": 485
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.6617895329206528,
|
| 699 |
+
"grad_norm": 0.04783002216381401,
|
| 700 |
"learning_rate": 8.459694344887732e-06,
|
| 701 |
+
"loss": 0.4098,
|
| 702 |
"step": 490
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 0.6685424873382104,
|
| 706 |
+
"grad_norm": 0.04187887777828664,
|
| 707 |
"learning_rate": 8.416885472132077e-06,
|
| 708 |
+
"loss": 0.3462,
|
| 709 |
"step": 495
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.6752954417557682,
|
| 713 |
+
"grad_norm": 0.05373723455097553,
|
| 714 |
"learning_rate": 8.373601420973464e-06,
|
| 715 |
+
"loss": 0.344,
|
| 716 |
"step": 500
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 0.6820483961733258,
|
| 720 |
+
"grad_norm": 0.04300150710853549,
|
| 721 |
"learning_rate": 8.329848210825322e-06,
|
| 722 |
+
"loss": 0.3647,
|
| 723 |
"step": 505
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.6888013505908835,
|
| 727 |
+
"grad_norm": 0.043097229603458374,
|
| 728 |
"learning_rate": 8.285631926345943e-06,
|
| 729 |
+
"loss": 0.3693,
|
| 730 |
"step": 510
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 0.6955543050084412,
|
| 734 |
+
"grad_norm": 0.040768739194901454,
|
| 735 |
"learning_rate": 8.240958716592304e-06,
|
| 736 |
+
"loss": 0.3645,
|
| 737 |
"step": 515
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.7023072594259989,
|
| 741 |
+
"grad_norm": 0.046324281574150986,
|
| 742 |
"learning_rate": 8.195834794164925e-06,
|
| 743 |
"loss": 0.3501,
|
| 744 |
"step": 520
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 0.7090602138435566,
|
| 748 |
+
"grad_norm": 0.04048770768699556,
|
| 749 |
"learning_rate": 8.150266434343904e-06,
|
| 750 |
"loss": 0.3296,
|
| 751 |
"step": 525
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.7158131682611142,
|
| 755 |
+
"grad_norm": 0.04249502701153697,
|
| 756 |
"learning_rate": 8.104259974216219e-06,
|
| 757 |
+
"loss": 0.3489,
|
| 758 |
"step": 530
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.7225661226786719,
|
| 762 |
+
"grad_norm": 0.03877822496803113,
|
| 763 |
"learning_rate": 8.057821811794457e-06,
|
| 764 |
+
"loss": 0.3673,
|
| 765 |
"step": 535
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.7293190770962296,
|
| 769 |
+
"grad_norm": 0.048769655901584476,
|
| 770 |
"learning_rate": 8.010958405127048e-06,
|
| 771 |
+
"loss": 0.3313,
|
| 772 |
"step": 540
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 0.7360720315137873,
|
| 776 |
+
"grad_norm": 0.07133171500587031,
|
| 777 |
"learning_rate": 7.963676271400158e-06,
|
| 778 |
+
"loss": 0.3472,
|
| 779 |
"step": 545
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.7428249859313449,
|
| 783 |
+
"grad_norm": 0.04064256549350507,
|
| 784 |
"learning_rate": 7.915981986031367e-06,
|
| 785 |
+
"loss": 0.3478,
|
| 786 |
"step": 550
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 0.7495779403489027,
|
| 790 |
+
"grad_norm": 0.04549548056733051,
|
| 791 |
"learning_rate": 7.86788218175523e-06,
|
| 792 |
+
"loss": 0.3342,
|
| 793 |
"step": 555
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.7563308947664603,
|
| 797 |
+
"grad_norm": 0.03599813075890018,
|
| 798 |
"learning_rate": 7.819383547700889e-06,
|
| 799 |
"loss": 0.3354,
|
| 800 |
"step": 560
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 0.7630838491840181,
|
| 804 |
+
"grad_norm": 0.04316680817634258,
|
| 805 |
"learning_rate": 7.770492828461824e-06,
|
| 806 |
+
"loss": 0.3622,
|
| 807 |
"step": 565
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.7698368036015757,
|
| 811 |
+
"grad_norm": 0.04908046724152037,
|
| 812 |
"learning_rate": 7.721216823157896e-06,
|
| 813 |
+
"loss": 0.3559,
|
| 814 |
"step": 570
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 0.7765897580191333,
|
| 818 |
+
"grad_norm": 0.040354310023923715,
|
| 819 |
"learning_rate": 7.671562384489819e-06,
|
| 820 |
+
"loss": 0.3629,
|
| 821 |
"step": 575
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.7833427124366911,
|
| 825 |
+
"grad_norm": 0.043579845331084145,
|
| 826 |
"learning_rate": 7.621536417786159e-06,
|
| 827 |
+
"loss": 0.3404,
|
| 828 |
"step": 580
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 0.7900956668542487,
|
| 832 |
+
"grad_norm": 0.04040670293206675,
|
| 833 |
"learning_rate": 7.571145880043036e-06,
|
| 834 |
+
"loss": 0.3672,
|
| 835 |
"step": 585
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.7968486212718064,
|
| 839 |
+
"grad_norm": 0.04244690517322693,
|
| 840 |
"learning_rate": 7.520397778956623e-06,
|
| 841 |
+
"loss": 0.3499,
|
| 842 |
"step": 590
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.8036015756893641,
|
| 846 |
+
"grad_norm": 0.05433806018620802,
|
| 847 |
"learning_rate": 7.469299171948608e-06,
|
| 848 |
+
"loss": 0.3704,
|
| 849 |
"step": 595
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.8103545301069218,
|
| 853 |
+
"grad_norm": 0.04428365952002865,
|
| 854 |
"learning_rate": 7.417857165184723e-06,
|
| 855 |
"loss": 0.3474,
|
| 856 |
"step": 600
|
| 857 |
},
|
| 858 |
{
|
| 859 |
"epoch": 0.8171074845244795,
|
| 860 |
+
"grad_norm": 0.03948151196873183,
|
| 861 |
"learning_rate": 7.366078912586523e-06,
|
| 862 |
+
"loss": 0.3426,
|
| 863 |
"step": 605
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 0.8238604389420371,
|
| 867 |
+
"grad_norm": 0.03804562874118621,
|
| 868 |
"learning_rate": 7.313971614836496e-06,
|
| 869 |
+
"loss": 0.3418,
|
| 870 |
"step": 610
|
| 871 |
},
|
| 872 |
{
|
| 873 |
"epoch": 0.8306133933595948,
|
| 874 |
+
"grad_norm": 0.04327583528687119,
|
| 875 |
"learning_rate": 7.261542518376677e-06,
|
| 876 |
+
"loss": 0.3408,
|
| 877 |
"step": 615
|
| 878 |
},
|
| 879 |
{
|
| 880 |
"epoch": 0.8373663477771525,
|
| 881 |
+
"grad_norm": 0.038581581579252,
|
| 882 |
"learning_rate": 7.208798914400916e-06,
|
| 883 |
+
"loss": 0.3429,
|
| 884 |
"step": 620
|
| 885 |
},
|
| 886 |
{
|
| 887 |
"epoch": 0.8441193021947102,
|
| 888 |
+
"grad_norm": 0.040274824329860945,
|
| 889 |
"learning_rate": 7.155748137840892e-06,
|
| 890 |
+
"loss": 0.385,
|
| 891 |
"step": 625
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 0.8508722566122678,
|
| 895 |
+
"grad_norm": 0.043944187715932226,
|
| 896 |
"learning_rate": 7.102397566346073e-06,
|
| 897 |
+
"loss": 0.3366,
|
| 898 |
"step": 630
|
| 899 |
},
|
| 900 |
{
|
| 901 |
"epoch": 0.8576252110298256,
|
| 902 |
+
"grad_norm": 0.04145675205661903,
|
| 903 |
"learning_rate": 7.048754619257716e-06,
|
| 904 |
+
"loss": 0.3755,
|
| 905 |
"step": 635
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.8643781654473832,
|
| 909 |
+
"grad_norm": 0.041045163518307656,
|
| 910 |
"learning_rate": 6.994826756577082e-06,
|
| 911 |
+
"loss": 0.3722,
|
| 912 |
"step": 640
|
| 913 |
},
|
| 914 |
{
|
| 915 |
"epoch": 0.871131119864941,
|
| 916 |
+
"grad_norm": 0.05061812503051895,
|
| 917 |
"learning_rate": 6.940621477927988e-06,
|
| 918 |
+
"loss": 0.3687,
|
| 919 |
"step": 645
|
| 920 |
},
|
| 921 |
{
|
| 922 |
"epoch": 0.8778840742824986,
|
| 923 |
+
"grad_norm": 0.03703919699696153,
|
| 924 |
"learning_rate": 6.88614632151385e-06,
|
| 925 |
+
"loss": 0.3454,
|
| 926 |
"step": 650
|
| 927 |
},
|
| 928 |
{
|
| 929 |
"epoch": 0.8846370287000562,
|
| 930 |
+
"grad_norm": 0.039635652764389524,
|
| 931 |
"learning_rate": 6.831408863069364e-06,
|
| 932 |
+
"loss": 0.365,
|
| 933 |
"step": 655
|
| 934 |
},
|
| 935 |
{
|
| 936 |
"epoch": 0.891389983117614,
|
| 937 |
+
"grad_norm": 0.0428338033091146,
|
| 938 |
"learning_rate": 6.7764167148069695e-06,
|
| 939 |
"loss": 0.3251,
|
| 940 |
"step": 660
|
| 941 |
},
|
| 942 |
{
|
| 943 |
"epoch": 0.8981429375351716,
|
| 944 |
+
"grad_norm": 0.0393255629299852,
|
| 945 |
"learning_rate": 6.721177524358226e-06,
|
| 946 |
+
"loss": 0.3519,
|
| 947 |
"step": 665
|
| 948 |
},
|
| 949 |
{
|
| 950 |
"epoch": 0.9048958919527293,
|
| 951 |
+
"grad_norm": 0.03605877433184958,
|
| 952 |
"learning_rate": 6.665698973710289e-06,
|
| 953 |
+
"loss": 0.3586,
|
| 954 |
"step": 670
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 0.911648846370287,
|
| 958 |
+
"grad_norm": 0.0418094941760349,
|
| 959 |
"learning_rate": 6.609988778137582e-06,
|
| 960 |
+
"loss": 0.333,
|
| 961 |
"step": 675
|
| 962 |
},
|
| 963 |
{
|
| 964 |
"epoch": 0.9184018007878447,
|
| 965 |
+
"grad_norm": 0.04338138841178037,
|
| 966 |
"learning_rate": 6.554054685128857e-06,
|
| 967 |
+
"loss": 0.334,
|
| 968 |
"step": 680
|
| 969 |
},
|
| 970 |
{
|
| 971 |
"epoch": 0.9251547552054024,
|
| 972 |
+
"grad_norm": 0.04225445092421321,
|
| 973 |
"learning_rate": 6.497904473309766e-06,
|
| 974 |
"loss": 0.3519,
|
| 975 |
"step": 685
|
| 976 |
},
|
| 977 |
{
|
| 978 |
"epoch": 0.93190770962296,
|
| 979 |
+
"grad_norm": 0.04826057110998316,
|
| 980 |
"learning_rate": 6.44154595136111e-06,
|
| 981 |
+
"loss": 0.3598,
|
| 982 |
"step": 690
|
| 983 |
},
|
| 984 |
{
|
| 985 |
"epoch": 0.9386606640405177,
|
| 986 |
+
"grad_norm": 0.04381442124287324,
|
| 987 |
"learning_rate": 6.384986956932897e-06,
|
| 988 |
"loss": 0.3339,
|
| 989 |
"step": 695
|
| 990 |
},
|
| 991 |
{
|
| 992 |
"epoch": 0.9454136184580754,
|
| 993 |
+
"grad_norm": 0.0414974847898415,
|
| 994 |
"learning_rate": 6.328235355554382e-06,
|
| 995 |
+
"loss": 0.3645,
|
| 996 |
"step": 700
|
| 997 |
},
|
| 998 |
{
|
| 999 |
"epoch": 0.9521665728756331,
|
| 1000 |
+
"grad_norm": 0.04276356905517851,
|
| 1001 |
"learning_rate": 6.271299039540228e-06,
|
| 1002 |
"loss": 0.3645,
|
| 1003 |
"step": 705
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
"epoch": 0.9589195272931907,
|
| 1007 |
+
"grad_norm": 0.046676300677672276,
|
| 1008 |
"learning_rate": 6.214185926892936e-06,
|
| 1009 |
+
"loss": 0.3551,
|
| 1010 |
"step": 710
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
"epoch": 0.9656724817107485,
|
| 1014 |
+
"grad_norm": 0.04006191125365821,
|
| 1015 |
"learning_rate": 6.156903960201709e-06,
|
| 1016 |
+
"loss": 0.3395,
|
| 1017 |
"step": 715
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.9724254361283061,
|
| 1021 |
+
"grad_norm": 0.047147919741739285,
|
| 1022 |
"learning_rate": 6.099461105537889e-06,
|
| 1023 |
+
"loss": 0.3321,
|
| 1024 |
"step": 720
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
"epoch": 0.9791783905458639,
|
| 1028 |
+
"grad_norm": 0.041770572631853936,
|
| 1029 |
"learning_rate": 6.041865351347146e-06,
|
| 1030 |
"loss": 0.3421,
|
| 1031 |
"step": 725
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
"epoch": 0.9859313449634215,
|
| 1035 |
+
"grad_norm": 0.04239922397171486,
|
| 1036 |
"learning_rate": 5.984124707338528e-06,
|
| 1037 |
+
"loss": 0.3373,
|
| 1038 |
"step": 730
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
"epoch": 0.9926842993809791,
|
| 1042 |
+
"grad_norm": 0.046194589474773985,
|
| 1043 |
"learning_rate": 5.926247203370583e-06,
|
| 1044 |
+
"loss": 0.3405,
|
| 1045 |
"step": 735
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 0.9994372537985369,
|
| 1049 |
+
"grad_norm": 0.03944962664459983,
|
| 1050 |
"learning_rate": 5.8682408883346535e-06,
|
| 1051 |
"loss": 0.2909,
|
| 1052 |
"step": 740
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 1.0,
|
| 1056 |
+
"eval_loss": 0.3324998915195465,
|
| 1057 |
+
"eval_runtime": 18.1309,
|
| 1058 |
+
"eval_samples_per_second": 19.8,
|
| 1059 |
+
"eval_steps_per_second": 4.964,
|
| 1060 |
"step": 741
|
| 1061 |
},
|
| 1062 |
{
|
| 1063 |
"epoch": 1.005402363534046,
|
| 1064 |
+
"grad_norm": 0.040121304388969325,
|
| 1065 |
"learning_rate": 5.810113829035544e-06,
|
| 1066 |
+
"loss": 0.355,
|
| 1067 |
"step": 745
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
"epoch": 1.0121553179516039,
|
| 1071 |
+
"grad_norm": 0.047689669773125845,
|
| 1072 |
"learning_rate": 5.751874109069685e-06,
|
| 1073 |
"loss": 0.3611,
|
| 1074 |
"step": 750
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 1.0189082723691616,
|
| 1078 |
+
"grad_norm": 0.03912739831058518,
|
| 1079 |
"learning_rate": 5.693529827700967e-06,
|
| 1080 |
+
"loss": 0.2899,
|
| 1081 |
"step": 755
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
"epoch": 1.0256612267867191,
|
| 1085 |
+
"grad_norm": 0.04377909856345864,
|
| 1086 |
"learning_rate": 5.635089098734394e-06,
|
| 1087 |
+
"loss": 0.3197,
|
| 1088 |
"step": 760
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
"epoch": 1.0324141812042769,
|
| 1092 |
+
"grad_norm": 0.037393171944725745,
|
| 1093 |
"learning_rate": 5.576560049387717e-06,
|
| 1094 |
"loss": 0.339,
|
| 1095 |
"step": 765
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
"epoch": 1.0391671356218346,
|
| 1099 |
+
"grad_norm": 0.042256918616508245,
|
| 1100 |
"learning_rate": 5.517950819161196e-06,
|
| 1101 |
"loss": 0.328,
|
| 1102 |
"step": 770
|
| 1103 |
},
|
| 1104 |
{
|
| 1105 |
"epoch": 1.0459200900393921,
|
| 1106 |
+
"grad_norm": 0.040529502097932234,
|
| 1107 |
"learning_rate": 5.459269558705667e-06,
|
| 1108 |
+
"loss": 0.3259,
|
| 1109 |
"step": 775
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
"epoch": 1.0526730444569499,
|
| 1113 |
+
"grad_norm": 0.04379220562588465,
|
| 1114 |
"learning_rate": 5.400524428689035e-06,
|
| 1115 |
+
"loss": 0.3464,
|
| 1116 |
"step": 780
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
"epoch": 1.0594259988745076,
|
| 1120 |
+
"grad_norm": 0.04725991662965075,
|
| 1121 |
"learning_rate": 5.341723598661409e-06,
|
| 1122 |
"loss": 0.3267,
|
| 1123 |
"step": 785
|
| 1124 |
},
|
| 1125 |
{
|
| 1126 |
"epoch": 1.0661789532920654,
|
| 1127 |
+
"grad_norm": 0.036832603953267457,
|
| 1128 |
"learning_rate": 5.282875245918963e-06,
|
| 1129 |
+
"loss": 0.3261,
|
| 1130 |
"step": 790
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
"epoch": 1.0729319077096229,
|
| 1134 |
+
"grad_norm": 0.04175828917214628,
|
| 1135 |
"learning_rate": 5.2239875543667465e-06,
|
| 1136 |
+
"loss": 0.3346,
|
| 1137 |
"step": 795
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
"epoch": 1.0796848621271806,
|
| 1141 |
+
"grad_norm": 0.03618010757790516,
|
| 1142 |
"learning_rate": 5.165068713380568e-06,
|
| 1143 |
+
"loss": 0.3182,
|
| 1144 |
"step": 800
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
"epoch": 1.0864378165447384,
|
| 1148 |
+
"grad_norm": 0.04097386666815235,
|
| 1149 |
"learning_rate": 5.106126916668118e-06,
|
| 1150 |
"loss": 0.3192,
|
| 1151 |
"step": 805
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
"epoch": 1.093190770962296,
|
| 1155 |
+
"grad_norm": 0.040300039979862916,
|
| 1156 |
"learning_rate": 5.047170361129484e-06,
|
| 1157 |
+
"loss": 0.3138,
|
| 1158 |
"step": 810
|
| 1159 |
},
|
| 1160 |
{
|
| 1161 |
"epoch": 1.0999437253798536,
|
| 1162 |
+
"grad_norm": 0.0413703774570022,
|
| 1163 |
"learning_rate": 4.988207245717232e-06,
|
| 1164 |
"loss": 0.3558,
|
| 1165 |
"step": 815
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
"epoch": 1.1066966797974114,
|
| 1169 |
+
"grad_norm": 0.03938201981503368,
|
| 1170 |
"learning_rate": 4.929245770296191e-06,
|
| 1171 |
+
"loss": 0.3436,
|
| 1172 |
"step": 820
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
"epoch": 1.1134496342149691,
|
| 1176 |
+
"grad_norm": 0.043903225983181984,
|
| 1177 |
"learning_rate": 4.870294134503123e-06,
|
| 1178 |
+
"loss": 0.3301,
|
| 1179 |
"step": 825
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 1.1202025886325266,
|
| 1183 |
+
"grad_norm": 0.04208713272188743,
|
| 1184 |
"learning_rate": 4.811360536606416e-06,
|
| 1185 |
+
"loss": 0.3108,
|
| 1186 |
"step": 830
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
"epoch": 1.1269555430500844,
|
| 1190 |
+
"grad_norm": 0.03660393810864938,
|
| 1191 |
"learning_rate": 4.752453172365966e-06,
|
| 1192 |
+
"loss": 0.333,
|
| 1193 |
"step": 835
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 1.1337084974676421,
|
| 1197 |
+
"grad_norm": 0.04407131833701405,
|
| 1198 |
"learning_rate": 4.69358023389342e-06,
|
| 1199 |
+
"loss": 0.3605,
|
| 1200 |
"step": 840
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 1.1404614518851997,
|
| 1204 |
+
"grad_norm": 0.04616999836259198,
|
| 1205 |
"learning_rate": 4.634749908512907e-06,
|
| 1206 |
+
"loss": 0.3439,
|
| 1207 |
"step": 845
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
"epoch": 1.1472144063027574,
|
| 1211 |
+
"grad_norm": 0.04325365689860756,
|
| 1212 |
"learning_rate": 4.575970377622456e-06,
|
| 1213 |
+
"loss": 0.3049,
|
| 1214 |
"step": 850
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
"epoch": 1.1539673607203151,
|
| 1218 |
+
"grad_norm": 0.03846831221970018,
|
| 1219 |
"learning_rate": 4.517249815556219e-06,
|
| 1220 |
+
"loss": 0.3511,
|
| 1221 |
"step": 855
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
"epoch": 1.1607203151378729,
|
| 1225 |
+
"grad_norm": 0.04569268985609919,
|
| 1226 |
"learning_rate": 4.458596388447691e-06,
|
| 1227 |
+
"loss": 0.3209,
|
| 1228 |
"step": 860
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
"epoch": 1.1674732695554304,
|
| 1232 |
+
"grad_norm": 0.0409252349150715,
|
| 1233 |
"learning_rate": 4.400018253094065e-06,
|
| 1234 |
+
"loss": 0.308,
|
| 1235 |
"step": 865
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
"epoch": 1.1742262239729881,
|
| 1239 |
+
"grad_norm": 0.04133167625196089,
|
| 1240 |
"learning_rate": 4.341523555821882e-06,
|
| 1241 |
+
"loss": 0.3377,
|
| 1242 |
"step": 870
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 1.180979178390546,
|
| 1246 |
+
"grad_norm": 0.043097502447341854,
|
| 1247 |
"learning_rate": 4.283120431354137e-06,
|
| 1248 |
+
"loss": 0.3117,
|
| 1249 |
"step": 875
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 1.1877321328081036,
|
| 1253 |
+
"grad_norm": 0.042543275375795706,
|
| 1254 |
"learning_rate": 4.224817001679011e-06,
|
| 1255 |
+
"loss": 0.3344,
|
| 1256 |
"step": 880
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
"epoch": 1.1944850872256612,
|
| 1260 |
+
"grad_norm": 0.04304997898343135,
|
| 1261 |
"learning_rate": 4.1666213749203545e-06,
|
| 1262 |
"loss": 0.3237,
|
| 1263 |
"step": 885
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
"epoch": 1.201238041643219,
|
| 1267 |
+
"grad_norm": 0.04100222680465783,
|
| 1268 |
"learning_rate": 4.1085416442101205e-06,
|
| 1269 |
+
"loss": 0.3641,
|
| 1270 |
"step": 890
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 1.2079909960607766,
|
| 1274 |
+
"grad_norm": 0.04523396323757555,
|
| 1275 |
"learning_rate": 4.050585886562858e-06,
|
| 1276 |
+
"loss": 0.3494,
|
| 1277 |
"step": 895
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
"epoch": 1.2147439504783342,
|
| 1281 |
+
"grad_norm": 0.042513327919157136,
|
| 1282 |
"learning_rate": 3.992762161752474e-06,
|
| 1283 |
+
"loss": 0.343,
|
| 1284 |
"step": 900
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 1.221496904895892,
|
| 1288 |
+
"grad_norm": 0.046594930027348866,
|
| 1289 |
"learning_rate": 3.935078511191368e-06,
|
| 1290 |
+
"loss": 0.3235,
|
| 1291 |
"step": 905
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
"epoch": 1.2282498593134497,
|
| 1295 |
+
"grad_norm": 0.03906041128591151,
|
| 1296 |
"learning_rate": 3.877542956812137e-06,
|
| 1297 |
"loss": 0.3221,
|
| 1298 |
"step": 910
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
"epoch": 1.2350028137310074,
|
| 1302 |
+
"grad_norm": 0.044515080037505254,
|
| 1303 |
"learning_rate": 3.820163499951984e-06,
|
| 1304 |
"loss": 0.3579,
|
| 1305 |
"step": 915
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 1.241755768148565,
|
| 1309 |
+
"grad_norm": 0.04205841651552162,
|
| 1310 |
"learning_rate": 3.7629481202399886e-06,
|
| 1311 |
+
"loss": 0.3375,
|
| 1312 |
"step": 920
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 1.2485087225661227,
|
| 1316 |
+
"grad_norm": 0.04152789361719496,
|
| 1317 |
"learning_rate": 3.705904774487396e-06,
|
| 1318 |
+
"loss": 0.2971,
|
| 1319 |
"step": 925
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
"epoch": 1.2552616769836804,
|
| 1323 |
+
"grad_norm": 0.05137094607363216,
|
| 1324 |
"learning_rate": 3.64904139558109e-06,
|
| 1325 |
+
"loss": 0.3343,
|
| 1326 |
"step": 930
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
"epoch": 1.2620146314012382,
|
| 1330 |
+
"grad_norm": 0.04499346319735512,
|
| 1331 |
"learning_rate": 3.5923658913803726e-06,
|
| 1332 |
+
"loss": 0.3274,
|
| 1333 |
"step": 935
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
"epoch": 1.2687675858187957,
|
| 1337 |
+
"grad_norm": 0.04194049880241552,
|
| 1338 |
"learning_rate": 3.5358861436172487e-06,
|
| 1339 |
+
"loss": 0.3436,
|
| 1340 |
"step": 940
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 1.2755205402363534,
|
| 1344 |
+
"grad_norm": 0.04739038652189242,
|
| 1345 |
"learning_rate": 3.47961000680032e-06,
|
| 1346 |
+
"loss": 0.3395,
|
| 1347 |
"step": 945
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
"epoch": 1.2822734946539112,
|
| 1351 |
+
"grad_norm": 0.04407332378962217,
|
| 1352 |
"learning_rate": 3.4235453071224882e-06,
|
| 1353 |
+
"loss": 0.3283,
|
| 1354 |
"step": 950
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
"epoch": 1.2890264490714687,
|
| 1358 |
+
"grad_norm": 0.046768041779961654,
|
| 1359 |
"learning_rate": 3.3676998413725726e-06,
|
| 1360 |
+
"loss": 0.3457,
|
| 1361 |
"step": 955
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
"epoch": 1.2957794034890264,
|
| 1365 |
+
"grad_norm": 0.04405892619650361,
|
| 1366 |
"learning_rate": 3.3120813758510385e-06,
|
| 1367 |
+
"loss": 0.3446,
|
| 1368 |
"step": 960
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
"epoch": 1.3025323579065842,
|
| 1372 |
+
"grad_norm": 0.04917585118916133,
|
| 1373 |
"learning_rate": 3.2566976452899507e-06,
|
| 1374 |
+
"loss": 0.3267,
|
| 1375 |
"step": 965
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
"epoch": 1.3092853123241417,
|
| 1379 |
+
"grad_norm": 0.04327329891699119,
|
| 1380 |
"learning_rate": 3.2015563517773214e-06,
|
| 1381 |
+
"loss": 0.3382,
|
| 1382 |
"step": 970
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
"epoch": 1.3160382667416994,
|
| 1386 |
+
"grad_norm": 0.04751650625794217,
|
| 1387 |
"learning_rate": 3.1466651636860025e-06,
|
| 1388 |
"loss": 0.3325,
|
| 1389 |
"step": 975
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
"epoch": 1.3227912211592572,
|
| 1393 |
+
"grad_norm": 0.05680333767716324,
|
| 1394 |
"learning_rate": 3.0920317146072577e-06,
|
| 1395 |
+
"loss": 0.3207,
|
| 1396 |
"step": 980
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
"epoch": 1.329544175576815,
|
| 1400 |
+
"grad_norm": 0.0497656718123479,
|
| 1401 |
"learning_rate": 3.0376636022891813e-06,
|
| 1402 |
+
"loss": 0.329,
|
| 1403 |
"step": 985
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
"epoch": 1.3362971299943727,
|
| 1407 |
+
"grad_norm": 0.048529124206900136,
|
| 1408 |
"learning_rate": 2.983568387580094e-06,
|
| 1409 |
+
"loss": 0.3322,
|
| 1410 |
"step": 990
|
| 1411 |
},
|
| 1412 |
{
|
| 1413 |
"epoch": 1.3430500844119302,
|
| 1414 |
+
"grad_norm": 0.04475425551771912,
|
| 1415 |
"learning_rate": 2.9297535933770732e-06,
|
| 1416 |
+
"loss": 0.3278,
|
| 1417 |
"step": 995
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
"epoch": 1.349803038829488,
|
| 1421 |
+
"grad_norm": 0.043952297509005415,
|
| 1422 |
"learning_rate": 2.8762267035797607e-06,
|
| 1423 |
+
"loss": 0.2961,
|
| 1424 |
"step": 1000
|
| 1425 |
},
|
| 1426 |
{
|
| 1427 |
"epoch": 1.3565559932470457,
|
| 1428 |
+
"grad_norm": 0.04672219098205047,
|
| 1429 |
"learning_rate": 2.822995162049599e-06,
|
| 1430 |
+
"loss": 0.3618,
|
| 1431 |
"step": 1005
|
| 1432 |
},
|
| 1433 |
{
|
| 1434 |
"epoch": 1.3633089476646032,
|
| 1435 |
+
"grad_norm": 0.04260321337685024,
|
| 1436 |
"learning_rate": 2.7700663715746213e-06,
|
| 1437 |
+
"loss": 0.3382,
|
| 1438 |
"step": 1010
|
| 1439 |
},
|
| 1440 |
{
|
| 1441 |
"epoch": 1.370061902082161,
|
| 1442 |
+
"grad_norm": 0.048008697667558865,
|
| 1443 |
"learning_rate": 2.7174476928399685e-06,
|
| 1444 |
+
"loss": 0.3466,
|
| 1445 |
"step": 1015
|
| 1446 |
},
|
| 1447 |
{
|
| 1448 |
"epoch": 1.3768148564997187,
|
| 1449 |
+
"grad_norm": 0.04892067375449484,
|
| 1450 |
"learning_rate": 2.66514644340426e-06,
|
| 1451 |
"loss": 0.3425,
|
| 1452 |
"step": 1020
|
| 1453 |
},
|
| 1454 |
{
|
| 1455 |
"epoch": 1.3835678109172762,
|
| 1456 |
+
"grad_norm": 0.04689882121063575,
|
| 1457 |
"learning_rate": 2.613169896681949e-06,
|
| 1458 |
+
"loss": 0.2877,
|
| 1459 |
"step": 1025
|
| 1460 |
},
|
| 1461 |
{
|
| 1462 |
"epoch": 1.390320765334834,
|
| 1463 |
+
"grad_norm": 0.054157177491435074,
|
| 1464 |
"learning_rate": 2.5615252809318287e-06,
|
| 1465 |
+
"loss": 0.3306,
|
| 1466 |
"step": 1030
|
| 1467 |
},
|
| 1468 |
{
|
| 1469 |
"epoch": 1.3970737197523917,
|
| 1470 |
+
"grad_norm": 0.048672772164928886,
|
| 1471 |
"learning_rate": 2.5102197782518145e-06,
|
| 1472 |
+
"loss": 0.3216,
|
| 1473 |
"step": 1035
|
| 1474 |
},
|
| 1475 |
{
|
| 1476 |
"epoch": 1.4038266741699492,
|
| 1477 |
+
"grad_norm": 0.06752970884761282,
|
| 1478 |
"learning_rate": 2.4592605235801544e-06,
|
| 1479 |
+
"loss": 0.3276,
|
| 1480 |
"step": 1040
|
| 1481 |
},
|
| 1482 |
{
|
| 1483 |
"epoch": 1.410579628587507,
|
| 1484 |
+
"grad_norm": 0.04710574868399225,
|
| 1485 |
"learning_rate": 2.4086546037031734e-06,
|
| 1486 |
+
"loss": 0.3173,
|
| 1487 |
"step": 1045
|
| 1488 |
},
|
| 1489 |
{
|
| 1490 |
"epoch": 1.4173325830050647,
|
| 1491 |
+
"grad_norm": 0.045184769590470804,
|
| 1492 |
"learning_rate": 2.3584090562697427e-06,
|
| 1493 |
+
"loss": 0.3164,
|
| 1494 |
"step": 1050
|
| 1495 |
},
|
| 1496 |
{
|
| 1497 |
"epoch": 1.4240855374226224,
|
| 1498 |
+
"grad_norm": 0.043485542556231396,
|
| 1499 |
"learning_rate": 2.3085308688125695e-06,
|
| 1500 |
+
"loss": 0.3027,
|
| 1501 |
"step": 1055
|
| 1502 |
},
|
| 1503 |
{
|
| 1504 |
"epoch": 1.4308384918401802,
|
| 1505 |
+
"grad_norm": 0.05260603355819636,
|
| 1506 |
"learning_rate": 2.2590269777764516e-06,
|
| 1507 |
+
"loss": 0.3004,
|
| 1508 |
"step": 1060
|
| 1509 |
},
|
| 1510 |
{
|
| 1511 |
"epoch": 1.4375914462577377,
|
| 1512 |
+
"grad_norm": 0.051099932583330926,
|
| 1513 |
"learning_rate": 2.2099042675536437e-06,
|
| 1514 |
+
"loss": 0.3368,
|
| 1515 |
"step": 1065
|
| 1516 |
},
|
| 1517 |
{
|
| 1518 |
"epoch": 1.4443444006752955,
|
| 1519 |
+
"grad_norm": 0.045328406677686464,
|
| 1520 |
"learning_rate": 2.161169569526461e-06,
|
| 1521 |
+
"loss": 0.3056,
|
| 1522 |
"step": 1070
|
| 1523 |
},
|
| 1524 |
{
|
| 1525 |
"epoch": 1.4510973550928532,
|
| 1526 |
+
"grad_norm": 0.05676490778567245,
|
| 1527 |
"learning_rate": 2.1128296611172593e-06,
|
| 1528 |
+
"loss": 0.3052,
|
| 1529 |
"step": 1075
|
| 1530 |
},
|
| 1531 |
{
|
| 1532 |
"epoch": 1.4578503095104107,
|
| 1533 |
+
"grad_norm": 0.044797358674842205,
|
| 1534 |
"learning_rate": 2.0648912648459072e-06,
|
| 1535 |
"loss": 0.3447,
|
| 1536 |
"step": 1080
|
| 1537 |
},
|
| 1538 |
{
|
| 1539 |
"epoch": 1.4646032639279685,
|
| 1540 |
+
"grad_norm": 0.049289994897252776,
|
| 1541 |
"learning_rate": 2.0173610473949048e-06,
|
| 1542 |
+
"loss": 0.3375,
|
| 1543 |
"step": 1085
|
| 1544 |
},
|
| 1545 |
{
|
| 1546 |
"epoch": 1.4713562183455262,
|
| 1547 |
+
"grad_norm": 0.04706434397429944,
|
| 1548 |
"learning_rate": 1.9702456186822595e-06,
|
| 1549 |
+
"loss": 0.3464,
|
| 1550 |
"step": 1090
|
| 1551 |
},
|
| 1552 |
{
|
| 1553 |
"epoch": 1.4781091727630837,
|
| 1554 |
+
"grad_norm": 0.05030965485900401,
|
| 1555 |
"learning_rate": 1.9235515309422685e-06,
|
| 1556 |
+
"loss": 0.3119,
|
| 1557 |
"step": 1095
|
| 1558 |
},
|
| 1559 |
{
|
| 1560 |
"epoch": 1.4848621271806415,
|
| 1561 |
+
"grad_norm": 0.04413503044032754,
|
| 1562 |
"learning_rate": 1.8772852778143064e-06,
|
| 1563 |
+
"loss": 0.3371,
|
| 1564 |
"step": 1100
|
| 1565 |
},
|
| 1566 |
{
|
| 1567 |
"epoch": 1.4916150815981992,
|
| 1568 |
+
"grad_norm": 0.06233556882470439,
|
| 1569 |
"learning_rate": 1.831453293439771e-06,
|
| 1570 |
+
"loss": 0.3253,
|
| 1571 |
"step": 1105
|
| 1572 |
},
|
| 1573 |
{
|
| 1574 |
"epoch": 1.498368036015757,
|
| 1575 |
+
"grad_norm": 0.05282578176334435,
|
| 1576 |
"learning_rate": 1.7860619515673034e-06,
|
| 1577 |
+
"loss": 0.3032,
|
| 1578 |
"step": 1110
|
| 1579 |
},
|
| 1580 |
{
|
| 1581 |
"epoch": 1.5051209904333147,
|
| 1582 |
+
"grad_norm": 0.05220647521633855,
|
| 1583 |
"learning_rate": 1.7411175646664103e-06,
|
| 1584 |
+
"loss": 0.3161,
|
| 1585 |
"step": 1115
|
| 1586 |
},
|
| 1587 |
{
|
| 1588 |
"epoch": 1.5118739448508722,
|
| 1589 |
+
"grad_norm": 0.047487304887968124,
|
| 1590 |
"learning_rate": 1.6966263830495939e-06,
|
| 1591 |
+
"loss": 0.3104,
|
| 1592 |
"step": 1120
|
| 1593 |
},
|
| 1594 |
{
|
| 1595 |
"epoch": 1.51862689926843,
|
| 1596 |
+
"grad_norm": 0.05098969364978786,
|
| 1597 |
"learning_rate": 1.6525945940031407e-06,
|
| 1598 |
"loss": 0.3692,
|
| 1599 |
"step": 1125
|
| 1600 |
},
|
| 1601 |
{
|
| 1602 |
"epoch": 1.5253798536859877,
|
| 1603 |
+
"grad_norm": 0.04810126110395747,
|
| 1604 |
"learning_rate": 1.6090283209266682e-06,
|
| 1605 |
+
"loss": 0.3383,
|
| 1606 |
"step": 1130
|
| 1607 |
},
|
| 1608 |
{
|
| 1609 |
"epoch": 1.5321328081035452,
|
| 1610 |
+
"grad_norm": 0.06656533903224827,
|
| 1611 |
"learning_rate": 1.5659336224815642e-06,
|
| 1612 |
+
"loss": 0.3413,
|
| 1613 |
"step": 1135
|
| 1614 |
},
|
| 1615 |
{
|
| 1616 |
"epoch": 1.538885762521103,
|
| 1617 |
+
"grad_norm": 0.057091145465475585,
|
| 1618 |
"learning_rate": 1.5233164917484117e-06,
|
| 1619 |
+
"loss": 0.3247,
|
| 1620 |
"step": 1140
|
| 1621 |
},
|
| 1622 |
{
|
| 1623 |
"epoch": 1.5456387169386607,
|
| 1624 |
+
"grad_norm": 0.05098329100385167,
|
| 1625 |
"learning_rate": 1.4811828553935498e-06,
|
| 1626 |
+
"loss": 0.3284,
|
| 1627 |
"step": 1145
|
| 1628 |
},
|
| 1629 |
{
|
| 1630 |
"epoch": 1.5523916713562183,
|
| 1631 |
+
"grad_norm": 0.04852633539753427,
|
| 1632 |
"learning_rate": 1.439538572844873e-06,
|
| 1633 |
+
"loss": 0.307,
|
| 1634 |
"step": 1150
|
| 1635 |
},
|
| 1636 |
{
|
| 1637 |
"epoch": 1.559144625773776,
|
| 1638 |
+
"grad_norm": 0.05935932625720792,
|
| 1639 |
"learning_rate": 1.3983894354769616e-06,
|
| 1640 |
+
"loss": 0.3247,
|
| 1641 |
"step": 1155
|
| 1642 |
},
|
| 1643 |
{
|
| 1644 |
"epoch": 1.5658975801913337,
|
| 1645 |
+
"grad_norm": 0.06306295104638487,
|
| 1646 |
"learning_rate": 1.3577411658056965e-06,
|
| 1647 |
+
"loss": 0.3286,
|
| 1648 |
"step": 1160
|
| 1649 |
},
|
| 1650 |
{
|
| 1651 |
"epoch": 1.5726505346088913,
|
| 1652 |
+
"grad_norm": 0.06348638606124773,
|
| 1653 |
"learning_rate": 1.3175994166924394e-06,
|
| 1654 |
+
"loss": 0.3247,
|
| 1655 |
"step": 1165
|
| 1656 |
},
|
| 1657 |
{
|
| 1658 |
"epoch": 1.5794034890264492,
|
| 1659 |
+
"grad_norm": 0.051811719215914646,
|
| 1660 |
"learning_rate": 1.2779697705579058e-06,
|
| 1661 |
+
"loss": 0.3246,
|
| 1662 |
"step": 1170
|
| 1663 |
},
|
| 1664 |
{
|
| 1665 |
"epoch": 1.5861564434440067,
|
| 1666 |
+
"grad_norm": 0.06830322751813911,
|
| 1667 |
"learning_rate": 1.2388577386058248e-06,
|
| 1668 |
+
"loss": 0.3326,
|
| 1669 |
"step": 1175
|
| 1670 |
},
|
| 1671 |
{
|
| 1672 |
"epoch": 1.5929093978615643,
|
| 1673 |
+
"grad_norm": 0.04922857249304829,
|
| 1674 |
"learning_rate": 1.2002687600565138e-06,
|
| 1675 |
+
"loss": 0.3061,
|
| 1676 |
"step": 1180
|
| 1677 |
},
|
| 1678 |
{
|
| 1679 |
"epoch": 1.5996623522791222,
|
| 1680 |
+
"grad_norm": 0.047108708806732846,
|
| 1681 |
"learning_rate": 1.1622082013904535e-06,
|
| 1682 |
+
"loss": 0.3118,
|
| 1683 |
"step": 1185
|
| 1684 |
},
|
| 1685 |
{
|
| 1686 |
"epoch": 1.6064153066966798,
|
| 1687 |
+
"grad_norm": 0.059687932246599115,
|
| 1688 |
"learning_rate": 1.1246813556019925e-06,
|
| 1689 |
+
"loss": 0.3372,
|
| 1690 |
"step": 1190
|
| 1691 |
},
|
| 1692 |
{
|
| 1693 |
"epoch": 1.6131682611142375,
|
| 1694 |
+
"grad_norm": 0.05955615759336213,
|
| 1695 |
"learning_rate": 1.0876934414632523e-06,
|
| 1696 |
+
"loss": 0.2956,
|
| 1697 |
"step": 1195
|
| 1698 |
},
|
| 1699 |
{
|
| 1700 |
"epoch": 1.6199212155317952,
|
| 1701 |
+
"grad_norm": 0.06542544159973171,
|
| 1702 |
"learning_rate": 1.0512496027983715e-06,
|
| 1703 |
+
"loss": 0.3118,
|
| 1704 |
"step": 1200
|
| 1705 |
},
|
| 1706 |
{
|
| 1707 |
"epoch": 1.6266741699493528,
|
| 1708 |
+
"grad_norm": 0.05198203822638065,
|
| 1709 |
"learning_rate": 1.0153549077681617e-06,
|
| 1710 |
+
"loss": 0.346,
|
| 1711 |
"step": 1205
|
| 1712 |
},
|
| 1713 |
{
|
| 1714 |
"epoch": 1.6334271243669105,
|
| 1715 |
+
"grad_norm": 0.06101598768362396,
|
| 1716 |
"learning_rate": 9.80014348165298e-07,
|
| 1717 |
+
"loss": 0.2952,
|
| 1718 |
"step": 1210
|
| 1719 |
},
|
| 1720 |
{
|
| 1721 |
"epoch": 1.6401800787844683,
|
| 1722 |
+
"grad_norm": 0.05552589226769097,
|
| 1723 |
"learning_rate": 9.452328387201104e-07,
|
| 1724 |
+
"loss": 0.2903,
|
| 1725 |
"step": 1215
|
| 1726 |
},
|
| 1727 |
{
|
| 1728 |
"epoch": 1.6469330332020258,
|
| 1729 |
+
"grad_norm": 0.06472026824610987,
|
| 1730 |
"learning_rate": 9.110152164171127e-07,
|
| 1731 |
+
"loss": 0.3081,
|
| 1732 |
"step": 1220
|
| 1733 |
},
|
| 1734 |
{
|
| 1735 |
"epoch": 1.6536859876195835,
|
| 1736 |
+
"grad_norm": 0.07431179400574185,
|
| 1737 |
"learning_rate": 8.773662398223276e-07,
|
| 1738 |
+
"loss": 0.3095,
|
| 1739 |
"step": 1225
|
| 1740 |
},
|
| 1741 |
{
|
| 1742 |
"epoch": 1.6604389420371413,
|
| 1743 |
+
"grad_norm": 0.06752269915658678,
|
| 1744 |
"learning_rate": 8.44290588421533e-07,
|
| 1745 |
+
"loss": 0.3187,
|
| 1746 |
"step": 1230
|
| 1747 |
},
|
| 1748 |
{
|
| 1749 |
"epoch": 1.6671918964546988,
|
| 1750 |
+
"grad_norm": 0.06578494925550707,
|
| 1751 |
"learning_rate": 8.117928619694848e-07,
|
| 1752 |
+
"loss": 0.3036,
|
| 1753 |
"step": 1235
|
| 1754 |
},
|
| 1755 |
{
|
| 1756 |
"epoch": 1.6739448508722568,
|
| 1757 |
+
"grad_norm": 0.06180051968265024,
|
| 1758 |
"learning_rate": 7.798775798502484e-07,
|
| 1759 |
+
"loss": 0.3023,
|
| 1760 |
"step": 1240
|
| 1761 |
},
|
| 1762 |
{
|
| 1763 |
"epoch": 1.6806978052898143,
|
| 1764 |
+
"grad_norm": 0.07560845712778133,
|
| 1765 |
"learning_rate": 7.485491804486972e-07,
|
| 1766 |
"loss": 0.2951,
|
| 1767 |
"step": 1245
|
| 1768 |
},
|
| 1769 |
{
|
| 1770 |
"epoch": 1.687450759707372,
|
| 1771 |
+
"grad_norm": 0.07194010404747912,
|
| 1772 |
"learning_rate": 7.178120205332717e-07,
|
| 1773 |
+
"loss": 0.3372,
|
| 1774 |
"step": 1250
|
| 1775 |
},
|
| 1776 |
{
|
| 1777 |
"epoch": 1.6942037141249298,
|
| 1778 |
+
"grad_norm": 0.06976609157173444,
|
| 1779 |
"learning_rate": 6.876703746500984e-07,
|
| 1780 |
+
"loss": 0.2973,
|
| 1781 |
"step": 1255
|
| 1782 |
},
|
| 1783 |
{
|
| 1784 |
"epoch": 1.7009566685424873,
|
| 1785 |
+
"grad_norm": 0.07659553993307627,
|
| 1786 |
"learning_rate": 6.581284345285371e-07,
|
| 1787 |
+
"loss": 0.284,
|
| 1788 |
"step": 1260
|
| 1789 |
},
|
| 1790 |
{
|
| 1791 |
"epoch": 1.707709622960045,
|
| 1792 |
+
"grad_norm": 0.05961865699755965,
|
| 1793 |
"learning_rate": 6.291903084982481e-07,
|
| 1794 |
+
"loss": 0.3153,
|
| 1795 |
"step": 1265
|
| 1796 |
},
|
| 1797 |
{
|
| 1798 |
"epoch": 1.7144625773776028,
|
| 1799 |
+
"grad_norm": 0.0812482207068001,
|
| 1800 |
"learning_rate": 6.008600209178539e-07,
|
| 1801 |
+
"loss": 0.342,
|
| 1802 |
"step": 1270
|
| 1803 |
},
|
| 1804 |
{
|
| 1805 |
"epoch": 1.7212155317951603,
|
| 1806 |
+
"grad_norm": 0.06939464438812673,
|
| 1807 |
"learning_rate": 5.7314151161528e-07,
|
| 1808 |
+
"loss": 0.3287,
|
| 1809 |
"step": 1275
|
| 1810 |
},
|
| 1811 |
{
|
| 1812 |
"epoch": 1.727968486212718,
|
| 1813 |
+
"grad_norm": 0.06615067862128174,
|
| 1814 |
"learning_rate": 5.460386353398583e-07,
|
| 1815 |
+
"loss": 0.3036,
|
| 1816 |
"step": 1280
|
| 1817 |
},
|
| 1818 |
{
|
| 1819 |
"epoch": 1.7347214406302758,
|
| 1820 |
+
"grad_norm": 0.08487351817752944,
|
| 1821 |
"learning_rate": 5.195551612262478e-07,
|
| 1822 |
+
"loss": 0.3059,
|
| 1823 |
"step": 1285
|
| 1824 |
},
|
| 1825 |
{
|
| 1826 |
"epoch": 1.7414743950478333,
|
| 1827 |
+
"grad_norm": 0.0660084040820942,
|
| 1828 |
"learning_rate": 4.936947722702762e-07,
|
| 1829 |
+
"loss": 0.3004,
|
| 1830 |
"step": 1290
|
| 1831 |
},
|
| 1832 |
{
|
| 1833 |
"epoch": 1.7482273494653913,
|
| 1834 |
+
"grad_norm": 0.07500490253176417,
|
| 1835 |
"learning_rate": 4.6846106481675035e-07,
|
| 1836 |
+
"loss": 0.3163,
|
| 1837 |
"step": 1295
|
| 1838 |
},
|
| 1839 |
{
|
| 1840 |
"epoch": 1.7549803038829488,
|
| 1841 |
+
"grad_norm": 0.07341914462851476,
|
| 1842 |
"learning_rate": 4.43857548059321e-07,
|
| 1843 |
+
"loss": 0.3028,
|
| 1844 |
"step": 1300
|
| 1845 |
},
|
| 1846 |
{
|
| 1847 |
"epoch": 1.7617332583005063,
|
| 1848 |
+
"grad_norm": 0.05936547346570651,
|
| 1849 |
"learning_rate": 4.198876435524718e-07,
|
| 1850 |
+
"loss": 0.3071,
|
| 1851 |
"step": 1305
|
| 1852 |
},
|
| 1853 |
{
|
| 1854 |
"epoch": 1.7684862127180643,
|
| 1855 |
+
"grad_norm": 0.06361743408706552,
|
| 1856 |
"learning_rate": 3.9655468473568436e-07,
|
| 1857 |
+
"loss": 0.3151,
|
| 1858 |
"step": 1310
|
| 1859 |
},
|
| 1860 |
{
|
| 1861 |
"epoch": 1.7752391671356218,
|
| 1862 |
+
"grad_norm": 0.07650389023350329,
|
| 1863 |
"learning_rate": 3.7386191646987094e-07,
|
| 1864 |
+
"loss": 0.333,
|
| 1865 |
"step": 1315
|
| 1866 |
},
|
| 1867 |
{
|
| 1868 |
"epoch": 1.7819921215531795,
|
| 1869 |
+
"grad_norm": 0.06885824668728105,
|
| 1870 |
"learning_rate": 3.51812494586114e-07,
|
| 1871 |
+
"loss": 0.2898,
|
| 1872 |
"step": 1320
|
| 1873 |
},
|
| 1874 |
{
|
| 1875 |
"epoch": 1.7887450759707373,
|
| 1876 |
+
"grad_norm": 0.0645999509118337,
|
| 1877 |
"learning_rate": 3.3040948544679817e-07,
|
| 1878 |
+
"loss": 0.3091,
|
| 1879 |
"step": 1325
|
| 1880 |
},
|
| 1881 |
{
|
| 1882 |
"epoch": 1.7954980303882948,
|
| 1883 |
+
"grad_norm": 0.07732866321608588,
|
| 1884 |
"learning_rate": 3.096558655191706e-07,
|
| 1885 |
+
"loss": 0.3148,
|
| 1886 |
"step": 1330
|
| 1887 |
},
|
| 1888 |
{
|
| 1889 |
"epoch": 1.8022509848058526,
|
| 1890 |
+
"grad_norm": 0.0733331982280887,
|
| 1891 |
"learning_rate": 2.895545209614176e-07,
|
| 1892 |
+
"loss": 0.3101,
|
| 1893 |
"step": 1335
|
| 1894 |
},
|
| 1895 |
{
|
| 1896 |
"epoch": 1.8090039392234103,
|
| 1897 |
+
"grad_norm": 0.06997792694817997,
|
| 1898 |
"learning_rate": 2.701082472212879e-07,
|
| 1899 |
+
"loss": 0.3302,
|
| 1900 |
"step": 1340
|
| 1901 |
},
|
| 1902 |
{
|
| 1903 |
"epoch": 1.8157568936409678,
|
| 1904 |
+
"grad_norm": 0.07837574138979145,
|
| 1905 |
"learning_rate": 2.5131974864734063e-07,
|
| 1906 |
"loss": 0.3425,
|
| 1907 |
"step": 1345
|
| 1908 |
},
|
| 1909 |
{
|
| 1910 |
"epoch": 1.8225098480585256,
|
| 1911 |
+
"grad_norm": 0.07239388554199513,
|
| 1912 |
"learning_rate": 2.331916381128535e-07,
|
| 1913 |
+
"loss": 0.3054,
|
| 1914 |
"step": 1350
|
| 1915 |
},
|
| 1916 |
{
|
| 1917 |
"epoch": 1.8292628024760833,
|
| 1918 |
+
"grad_norm": 0.0674988666661417,
|
| 1919 |
"learning_rate": 2.157264366524603e-07,
|
| 1920 |
+
"loss": 0.28,
|
| 1921 |
"step": 1355
|
| 1922 |
},
|
| 1923 |
{
|
| 1924 |
"epoch": 1.8360157568936408,
|
| 1925 |
+
"grad_norm": 0.07181164697688498,
|
| 1926 |
"learning_rate": 1.989265731115525e-07,
|
| 1927 |
+
"loss": 0.3307,
|
| 1928 |
"step": 1360
|
| 1929 |
},
|
| 1930 |
{
|
| 1931 |
"epoch": 1.8427687113111988,
|
| 1932 |
+
"grad_norm": 0.07548202922443094,
|
| 1933 |
"learning_rate": 1.827943838085111e-07,
|
| 1934 |
"loss": 0.306,
|
| 1935 |
"step": 1365
|
| 1936 |
},
|
| 1937 |
{
|
| 1938 |
"epoch": 1.8495216657287563,
|
| 1939 |
+
"grad_norm": 0.05982075375718766,
|
| 1940 |
"learning_rate": 1.6733211220979316e-07,
|
| 1941 |
+
"loss": 0.2981,
|
| 1942 |
"step": 1370
|
| 1943 |
},
|
| 1944 |
{
|
| 1945 |
"epoch": 1.856274620146314,
|
| 1946 |
+
"grad_norm": 0.08355451454560252,
|
| 1947 |
"learning_rate": 1.5254190861794415e-07,
|
| 1948 |
+
"loss": 0.3083,
|
| 1949 |
"step": 1375
|
| 1950 |
},
|
| 1951 |
{
|
| 1952 |
"epoch": 1.8630275745638718,
|
| 1953 |
+
"grad_norm": 0.07033662219869818,
|
| 1954 |
"learning_rate": 1.3842582987255494e-07,
|
| 1955 |
+
"loss": 0.293,
|
| 1956 |
"step": 1380
|
| 1957 |
},
|
| 1958 |
{
|
| 1959 |
"epoch": 1.8697805289814293,
|
| 1960 |
+
"grad_norm": 0.06615537508166416,
|
| 1961 |
"learning_rate": 1.2498583906422779e-07,
|
| 1962 |
+
"loss": 0.3072,
|
| 1963 |
"step": 1385
|
| 1964 |
},
|
| 1965 |
{
|
| 1966 |
"epoch": 1.876533483398987,
|
| 1967 |
+
"grad_norm": 0.059237664909188886,
|
| 1968 |
"learning_rate": 1.1222380526156929e-07,
|
| 1969 |
+
"loss": 0.313,
|
| 1970 |
"step": 1390
|
| 1971 |
},
|
| 1972 |
{
|
| 1973 |
"epoch": 1.8832864378165448,
|
| 1974 |
+
"grad_norm": 0.08185957993390121,
|
| 1975 |
"learning_rate": 1.0014150325126315e-07,
|
| 1976 |
+
"loss": 0.3071,
|
| 1977 |
"step": 1395
|
| 1978 |
},
|
| 1979 |
{
|
| 1980 |
"epoch": 1.8900393922341023,
|
| 1981 |
+
"grad_norm": 0.06113218790264024,
|
| 1982 |
"learning_rate": 8.874061329125939e-08,
|
| 1983 |
+
"loss": 0.3116,
|
| 1984 |
"step": 1400
|
| 1985 |
},
|
| 1986 |
{
|
| 1987 |
"epoch": 1.89679234665166,
|
| 1988 |
+
"grad_norm": 0.08593985138096456,
|
| 1989 |
"learning_rate": 7.802272087709951e-08,
|
| 1990 |
+
"loss": 0.3106,
|
| 1991 |
"step": 1405
|
| 1992 |
},
|
| 1993 |
{
|
| 1994 |
"epoch": 1.9035453010692178,
|
| 1995 |
+
"grad_norm": 0.0753339474248732,
|
| 1996 |
"learning_rate": 6.798931652142738e-08,
|
| 1997 |
+
"loss": 0.3118,
|
| 1998 |
"step": 1410
|
| 1999 |
},
|
| 2000 |
{
|
| 2001 |
"epoch": 1.9102982554867753,
|
| 2002 |
+
"grad_norm": 0.06406340322576629,
|
| 2003 |
"learning_rate": 5.864179554670724e-08,
|
| 2004 |
+
"loss": 0.3384,
|
| 2005 |
"step": 1415
|
| 2006 |
},
|
| 2007 |
{
|
| 2008 |
"epoch": 1.9170512099043333,
|
| 2009 |
+
"grad_norm": 0.06739406393286214,
|
| 2010 |
"learning_rate": 4.998145789118114e-08,
|
| 2011 |
+
"loss": 0.2755,
|
| 2012 |
"step": 1420
|
| 2013 |
},
|
| 2014 |
{
|
| 2015 |
"epoch": 1.9238041643218908,
|
| 2016 |
+
"grad_norm": 0.07101405306951455,
|
| 2017 |
"learning_rate": 4.2009507928084116e-08,
|
| 2018 |
+
"loss": 0.2802,
|
| 2019 |
"step": 1425
|
| 2020 |
},
|
| 2021 |
{
|
| 2022 |
"epoch": 1.9305571187394484,
|
| 2023 |
+
"grad_norm": 0.07509273619426268,
|
| 2024 |
"learning_rate": 3.4727054298161475e-08,
|
| 2025 |
+
"loss": 0.2778,
|
| 2026 |
"step": 1430
|
| 2027 |
},
|
| 2028 |
{
|
| 2029 |
"epoch": 1.9373100731570063,
|
| 2030 |
+
"grad_norm": 0.0719074660592959,
|
| 2031 |
"learning_rate": 2.8135109755487723e-08,
|
| 2032 |
+
"loss": 0.3238,
|
| 2033 |
"step": 1435
|
| 2034 |
},
|
| 2035 |
{
|
| 2036 |
"epoch": 1.9440630275745638,
|
| 2037 |
+
"grad_norm": 0.07042162359075328,
|
| 2038 |
"learning_rate": 2.223459102662695e-08,
|
| 2039 |
+
"loss": 0.3235,
|
| 2040 |
"step": 1440
|
| 2041 |
},
|
| 2042 |
{
|
| 2043 |
"epoch": 1.9508159819921216,
|
| 2044 |
+
"grad_norm": 0.07906944364159181,
|
| 2045 |
"learning_rate": 1.7026318683147082e-08,
|
| 2046 |
+
"loss": 0.3115,
|
| 2047 |
"step": 1445
|
| 2048 |
},
|
| 2049 |
{
|
| 2050 |
"epoch": 1.9575689364096793,
|
| 2051 |
+
"grad_norm": 0.08613249297162148,
|
| 2052 |
"learning_rate": 1.2511017027501682e-08,
|
| 2053 |
"loss": 0.3032,
|
| 2054 |
"step": 1450
|
| 2055 |
},
|
| 2056 |
{
|
| 2057 |
"epoch": 1.9643218908272368,
|
| 2058 |
+
"grad_norm": 0.07379471104805907,
|
| 2059 |
"learning_rate": 8.689313992306104e-09,
|
| 2060 |
+
"loss": 0.303,
|
| 2061 |
"step": 1455
|
| 2062 |
},
|
| 2063 |
{
|
| 2064 |
"epoch": 1.9710748452447946,
|
| 2065 |
+
"grad_norm": 0.07085802534414035,
|
| 2066 |
"learning_rate": 5.561741053010661e-09,
|
| 2067 |
+
"loss": 0.3059,
|
| 2068 |
"step": 1460
|
| 2069 |
},
|
| 2070 |
{
|
| 2071 |
"epoch": 1.9778277996623523,
|
| 2072 |
+
"grad_norm": 0.07455102650353135,
|
| 2073 |
"learning_rate": 3.1287331539903155e-09,
|
| 2074 |
+
"loss": 0.3165,
|
| 2075 |
"step": 1465
|
| 2076 |
},
|
| 2077 |
{
|
| 2078 |
"epoch": 1.9845807540799099,
|
| 2079 |
+
"grad_norm": 0.07699115693937987,
|
| 2080 |
"learning_rate": 1.3906286480563913e-09,
|
| 2081 |
+
"loss": 0.3108,
|
| 2082 |
"step": 1470
|
| 2083 |
},
|
| 2084 |
{
|
| 2085 |
"epoch": 1.9913337084974676,
|
| 2086 |
+
"grad_norm": 0.06672257472875329,
|
| 2087 |
"learning_rate": 3.4766924940476954e-10,
|
| 2088 |
+
"loss": 0.3134,
|
| 2089 |
"step": 1475
|
| 2090 |
},
|
| 2091 |
{
|
| 2092 |
"epoch": 1.9980866629150253,
|
| 2093 |
+
"grad_norm": 0.06196469052553269,
|
| 2094 |
"learning_rate": 0.0,
|
| 2095 |
+
"loss": 0.3054,
|
| 2096 |
"step": 1480
|
| 2097 |
},
|
| 2098 |
{
|
| 2099 |
"epoch": 1.9980866629150253,
|
| 2100 |
+
"eval_loss": 0.3086739182472229,
|
| 2101 |
+
"eval_runtime": 17.7246,
|
| 2102 |
+
"eval_samples_per_second": 20.254,
|
| 2103 |
+
"eval_steps_per_second": 5.078,
|
| 2104 |
"step": 1480
|
| 2105 |
}
|
| 2106 |
],
|
checkpoint-1480/training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 7288
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6f6a7799de28c93d654b3137724722a60a4634ceedd183783e265a4c550d87c3
|
| 3 |
size 7288
|
model-00001-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4976698672
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:bd78dcad6a6e52066ad8f69de370e285ef65b53be6a6f5b003ec73161153f1a1
|
| 3 |
size 4976698672
|
model-00002-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4999802720
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eaf863d03a19d67f3b5bd124a1af51716f33a11900ec798135fa04c015b7780c
|
| 3 |
size 4999802720
|
model-00003-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4915916176
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0701dd793589970557efd0a6b3fa5a9e330186f1276390178e9fd5ee51010030
|
| 3 |
size 4915916176
|
model-00004-of-00004.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1168138808
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1652d220d7874225cd65e0501667096059583d52796ad6ed749081c58c12b3c7
|
| 3 |
size 1168138808
|
tokenizer.json
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c5cf44023714fb39b05e71e425f8d7b92805ff73f7988b083b8c87f0bf87393
|
| 3 |
+
size 17209961
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"epoch": 1.9980866629150253,
|
| 3 |
"total_flos": 3.6719988787579453e+18,
|
| 4 |
-
"train_loss": 0.
|
| 5 |
-
"train_runtime":
|
| 6 |
"train_samples": 35538,
|
| 7 |
-
"train_samples_per_second": 5.
|
| 8 |
"train_steps_per_second": 0.112
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"epoch": 1.9980866629150253,
|
| 3 |
"total_flos": 3.6719988787579453e+18,
|
| 4 |
+
"train_loss": 0.37467364362768224,
|
| 5 |
+
"train_runtime": 13260.7751,
|
| 6 |
"train_samples": 35538,
|
| 7 |
+
"train_samples_per_second": 5.36,
|
| 8 |
"train_steps_per_second": 0.112
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -10,2106 +10,2106 @@
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0013505908835115363,
|
| 13 |
-
"grad_norm": 2.
|
| 14 |
"learning_rate": 6.756756756756757e-08,
|
| 15 |
"loss": 2.1725,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.006752954417557681,
|
| 20 |
-
"grad_norm": 2.
|
| 21 |
"learning_rate": 3.378378378378379e-07,
|
| 22 |
-
"loss": 2.
|
| 23 |
"step": 5
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.013505908835115363,
|
| 27 |
-
"grad_norm": 1.
|
| 28 |
"learning_rate": 6.756756756756758e-07,
|
| 29 |
-
"loss": 2.
|
| 30 |
"step": 10
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.020258863252673044,
|
| 34 |
-
"grad_norm": 0.
|
| 35 |
"learning_rate": 1.0135135135135136e-06,
|
| 36 |
-
"loss": 1.
|
| 37 |
"step": 15
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.027011817670230726,
|
| 41 |
-
"grad_norm": 0.
|
| 42 |
"learning_rate": 1.3513513513513515e-06,
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 20
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.03376477208778841,
|
| 48 |
-
"grad_norm": 0.
|
| 49 |
"learning_rate": 1.6891891891891894e-06,
|
| 50 |
-
"loss": 1.
|
| 51 |
"step": 25
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.04051772650534609,
|
| 55 |
-
"grad_norm": 0.
|
| 56 |
"learning_rate": 2.0270270270270273e-06,
|
| 57 |
-
"loss": 0.
|
| 58 |
"step": 30
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.04727068092290377,
|
| 62 |
-
"grad_norm": 0.
|
| 63 |
"learning_rate": 2.364864864864865e-06,
|
| 64 |
-
"loss": 0.
|
| 65 |
"step": 35
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.05402363534046145,
|
| 69 |
-
"grad_norm": 0.
|
| 70 |
"learning_rate": 2.702702702702703e-06,
|
| 71 |
-
"loss": 0.
|
| 72 |
"step": 40
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.06077658975801913,
|
| 76 |
-
"grad_norm": 0.
|
| 77 |
"learning_rate": 3.040540540540541e-06,
|
| 78 |
-
"loss": 0.
|
| 79 |
"step": 45
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.06752954417557681,
|
| 83 |
-
"grad_norm": 0.
|
| 84 |
"learning_rate": 3.3783783783783788e-06,
|
| 85 |
-
"loss": 0.
|
| 86 |
"step": 50
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.0742824985931345,
|
| 90 |
-
"grad_norm": 0.
|
| 91 |
"learning_rate": 3.7162162162162162e-06,
|
| 92 |
"loss": 0.5621,
|
| 93 |
"step": 55
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.08103545301069218,
|
| 97 |
-
"grad_norm": 0.
|
| 98 |
"learning_rate": 4.0540540540540545e-06,
|
| 99 |
-
"loss": 0.
|
| 100 |
"step": 60
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.08778840742824986,
|
| 104 |
-
"grad_norm": 0.
|
| 105 |
"learning_rate": 4.391891891891892e-06,
|
| 106 |
-
"loss": 0.
|
| 107 |
"step": 65
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.09454136184580754,
|
| 111 |
-
"grad_norm": 0.
|
| 112 |
"learning_rate": 4.72972972972973e-06,
|
| 113 |
-
"loss": 0.
|
| 114 |
"step": 70
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.10129431626336523,
|
| 118 |
-
"grad_norm": 0.
|
| 119 |
"learning_rate": 5.067567567567568e-06,
|
| 120 |
-
"loss": 0.
|
| 121 |
"step": 75
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.1080472706809229,
|
| 125 |
-
"grad_norm": 0.
|
| 126 |
"learning_rate": 5.405405405405406e-06,
|
| 127 |
-
"loss": 0.
|
| 128 |
"step": 80
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.11480022509848059,
|
| 132 |
-
"grad_norm": 0.
|
| 133 |
"learning_rate": 5.743243243243244e-06,
|
| 134 |
"loss": 0.471,
|
| 135 |
"step": 85
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.12155317951603826,
|
| 139 |
-
"grad_norm": 0.
|
| 140 |
"learning_rate": 6.081081081081082e-06,
|
| 141 |
-
"loss": 0.
|
| 142 |
"step": 90
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.12830613393359594,
|
| 146 |
-
"grad_norm": 0.
|
| 147 |
"learning_rate": 6.41891891891892e-06,
|
| 148 |
-
"loss": 0.
|
| 149 |
"step": 95
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.13505908835115363,
|
| 153 |
-
"grad_norm": 0.
|
| 154 |
"learning_rate": 6.7567567567567575e-06,
|
| 155 |
"loss": 0.4398,
|
| 156 |
"step": 100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.14181204276871132,
|
| 160 |
-
"grad_norm": 0.
|
| 161 |
"learning_rate": 7.0945945945945946e-06,
|
| 162 |
"loss": 0.3899,
|
| 163 |
"step": 105
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.148564997186269,
|
| 167 |
-
"grad_norm": 0.
|
| 168 |
"learning_rate": 7.4324324324324324e-06,
|
| 169 |
-
"loss": 0.
|
| 170 |
"step": 110
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.15531795160382666,
|
| 174 |
-
"grad_norm": 0.
|
| 175 |
"learning_rate": 7.77027027027027e-06,
|
| 176 |
-
"loss": 0.
|
| 177 |
"step": 115
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.16207090602138435,
|
| 181 |
-
"grad_norm": 0.
|
| 182 |
"learning_rate": 8.108108108108109e-06,
|
| 183 |
-
"loss": 0.
|
| 184 |
"step": 120
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.16882386043894204,
|
| 188 |
-
"grad_norm": 0.
|
| 189 |
"learning_rate": 8.445945945945948e-06,
|
| 190 |
-
"loss": 0.
|
| 191 |
"step": 125
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.17557681485649973,
|
| 195 |
-
"grad_norm": 0.
|
| 196 |
"learning_rate": 8.783783783783785e-06,
|
| 197 |
-
"loss": 0.
|
| 198 |
"step": 130
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.1823297692740574,
|
| 202 |
-
"grad_norm": 0.
|
| 203 |
"learning_rate": 9.121621621621622e-06,
|
| 204 |
"loss": 0.3895,
|
| 205 |
"step": 135
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.18908272369161508,
|
| 209 |
-
"grad_norm": 0.
|
| 210 |
"learning_rate": 9.45945945945946e-06,
|
| 211 |
"loss": 0.4029,
|
| 212 |
"step": 140
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.19583567810917277,
|
| 216 |
-
"grad_norm": 0.
|
| 217 |
"learning_rate": 9.797297297297298e-06,
|
| 218 |
-
"loss": 0.
|
| 219 |
"step": 145
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.20258863252673046,
|
| 223 |
-
"grad_norm": 0.
|
| 224 |
"learning_rate": 9.999944372378571e-06,
|
| 225 |
-
"loss": 0.
|
| 226 |
"step": 150
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.20934158694428812,
|
| 230 |
-
"grad_norm": 0.
|
| 231 |
"learning_rate": 9.999318575852451e-06,
|
| 232 |
-
"loss": 0.
|
| 233 |
"step": 155
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.2160945413618458,
|
| 237 |
-
"grad_norm": 0.
|
| 238 |
"learning_rate": 9.99799753559161e-06,
|
| 239 |
-
"loss": 0.
|
| 240 |
"step": 160
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.2228474957794035,
|
| 244 |
-
"grad_norm": 0.
|
| 245 |
"learning_rate": 9.995981435310078e-06,
|
| 246 |
-
"loss": 0.
|
| 247 |
"step": 165
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.22960045019696118,
|
| 251 |
-
"grad_norm": 0.
|
| 252 |
"learning_rate": 9.993270555382283e-06,
|
| 253 |
"loss": 0.4058,
|
| 254 |
"step": 170
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.23635340461451884,
|
| 258 |
-
"grad_norm": 0.
|
| 259 |
"learning_rate": 9.989865272804064e-06,
|
| 260 |
-
"loss": 0.
|
| 261 |
"step": 175
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.24310635903207653,
|
| 265 |
-
"grad_norm": 0.
|
| 266 |
"learning_rate": 9.985766061140233e-06,
|
| 267 |
-
"loss": 0.
|
| 268 |
"step": 180
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.24985931344963422,
|
| 272 |
-
"grad_norm": 0.
|
| 273 |
"learning_rate": 9.980973490458728e-06,
|
| 274 |
-
"loss": 0.
|
| 275 |
"step": 185
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.2566122678671919,
|
| 279 |
-
"grad_norm": 0.
|
| 280 |
"learning_rate": 9.97548822725133e-06,
|
| 281 |
-
"loss": 0.
|
| 282 |
"step": 190
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.26336522228474957,
|
| 286 |
-
"grad_norm": 0.
|
| 287 |
"learning_rate": 9.969311034340977e-06,
|
| 288 |
-
"loss": 0.
|
| 289 |
"step": 195
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.27011817670230726,
|
| 293 |
-
"grad_norm": 0.
|
| 294 |
"learning_rate": 9.962442770775675e-06,
|
| 295 |
-
"loss": 0.
|
| 296 |
"step": 200
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.27687113111986494,
|
| 300 |
-
"grad_norm": 0.
|
| 301 |
"learning_rate": 9.954884391709043e-06,
|
| 302 |
-
"loss": 0.
|
| 303 |
"step": 205
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.28362408553742263,
|
| 307 |
-
"grad_norm": 0.
|
| 308 |
"learning_rate": 9.946636948267468e-06,
|
| 309 |
"loss": 0.349,
|
| 310 |
"step": 210
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.2903770399549803,
|
| 314 |
-
"grad_norm": 0.
|
| 315 |
"learning_rate": 9.937701587403941e-06,
|
| 316 |
-
"loss": 0.
|
| 317 |
"step": 215
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.297129994372538,
|
| 321 |
-
"grad_norm": 0.
|
| 322 |
"learning_rate": 9.928079551738542e-06,
|
| 323 |
-
"loss": 0.
|
| 324 |
"step": 220
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.30388294879009564,
|
| 328 |
-
"grad_norm": 0.
|
| 329 |
"learning_rate": 9.91777217938564e-06,
|
| 330 |
-
"loss": 0.
|
| 331 |
"step": 225
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.31063590320765333,
|
| 335 |
-
"grad_norm": 0.
|
| 336 |
"learning_rate": 9.906780903767799e-06,
|
| 337 |
-
"loss": 0.
|
| 338 |
"step": 230
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.317388857625211,
|
| 342 |
-
"grad_norm": 0.
|
| 343 |
"learning_rate": 9.895107253416434e-06,
|
| 344 |
-
"loss": 0.
|
| 345 |
"step": 235
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.3241418120427687,
|
| 349 |
-
"grad_norm": 0.
|
| 350 |
"learning_rate": 9.882752851759247e-06,
|
| 351 |
-
"loss": 0.
|
| 352 |
"step": 240
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.3308947664603264,
|
| 356 |
-
"grad_norm": 0.
|
| 357 |
"learning_rate": 9.869719416894462e-06,
|
| 358 |
-
"loss": 0.
|
| 359 |
"step": 245
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.3376477208778841,
|
| 363 |
-
"grad_norm": 0.
|
| 364 |
"learning_rate": 9.856008761351882e-06,
|
| 365 |
-
"loss": 0.
|
| 366 |
"step": 250
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 0.34440067529544177,
|
| 370 |
-
"grad_norm": 0.
|
| 371 |
"learning_rate": 9.841622791840839e-06,
|
| 372 |
-
"loss": 0.
|
| 373 |
"step": 255
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.35115362971299946,
|
| 377 |
-
"grad_norm": 0.
|
| 378 |
"learning_rate": 9.826563508985017e-06,
|
| 379 |
-
"loss": 0.
|
| 380 |
"step": 260
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.3579065841305571,
|
| 384 |
-
"grad_norm": 0.
|
| 385 |
"learning_rate": 9.810833007044247e-06,
|
| 386 |
"loss": 0.3925,
|
| 387 |
"step": 265
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.3646595385481148,
|
| 391 |
-
"grad_norm": 0.
|
| 392 |
"learning_rate": 9.794433473623249e-06,
|
| 393 |
-
"loss": 0.
|
| 394 |
"step": 270
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 0.37141249296567247,
|
| 398 |
-
"grad_norm": 0.
|
| 399 |
"learning_rate": 9.777367189367412e-06,
|
| 400 |
"loss": 0.3779,
|
| 401 |
"step": 275
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.37816544738323016,
|
| 405 |
-
"grad_norm": 0.
|
| 406 |
"learning_rate": 9.759636527645633e-06,
|
| 407 |
-
"loss": 0.
|
| 408 |
"step": 280
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.38491840180078785,
|
| 412 |
-
"grad_norm": 0.
|
| 413 |
"learning_rate": 9.74124395422025e-06,
|
| 414 |
"loss": 0.3725,
|
| 415 |
"step": 285
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.39167135621834553,
|
| 419 |
-
"grad_norm": 0.
|
| 420 |
"learning_rate": 9.722192026904145e-06,
|
| 421 |
-
"loss": 0.
|
| 422 |
"step": 290
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.3984243106359032,
|
| 426 |
-
"grad_norm": 0.
|
| 427 |
"learning_rate": 9.702483395205023e-06,
|
| 428 |
-
"loss": 0.
|
| 429 |
"step": 295
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.4051772650534609,
|
| 433 |
-
"grad_norm": 0.
|
| 434 |
"learning_rate": 9.682120799956961e-06,
|
| 435 |
-
"loss": 0.
|
| 436 |
"step": 300
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 0.41193021947101854,
|
| 440 |
-
"grad_norm": 0.
|
| 441 |
"learning_rate": 9.661107072939244e-06,
|
| 442 |
-
"loss": 0.
|
| 443 |
"step": 305
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.41868317388857623,
|
| 447 |
-
"grad_norm": 0.
|
| 448 |
"learning_rate": 9.639445136482549e-06,
|
| 449 |
-
"loss": 0.
|
| 450 |
"step": 310
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.4254361283061339,
|
| 454 |
-
"grad_norm": 0.
|
| 455 |
"learning_rate": 9.61713800306255e-06,
|
| 456 |
-
"loss": 0.
|
| 457 |
"step": 315
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.4321890827236916,
|
| 461 |
-
"grad_norm": 0.
|
| 462 |
"learning_rate": 9.594188774880981e-06,
|
| 463 |
-
"loss": 0.
|
| 464 |
"step": 320
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.4389420371412493,
|
| 468 |
-
"grad_norm": 0.
|
| 469 |
"learning_rate": 9.570600643434217e-06,
|
| 470 |
"loss": 0.3371,
|
| 471 |
"step": 325
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.445694991558807,
|
| 475 |
-
"grad_norm": 0.
|
| 476 |
"learning_rate": 9.546376889069442e-06,
|
| 477 |
-
"loss": 0.
|
| 478 |
"step": 330
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 0.4524479459763647,
|
| 482 |
-
"grad_norm": 0.
|
| 483 |
"learning_rate": 9.521520880528453e-06,
|
| 484 |
-
"loss": 0.
|
| 485 |
"step": 335
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.45920090039392236,
|
| 489 |
-
"grad_norm": 0.
|
| 490 |
"learning_rate": 9.496036074479184e-06,
|
| 491 |
"loss": 0.3604,
|
| 492 |
"step": 340
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 0.46595385481148,
|
| 496 |
-
"grad_norm": 0.
|
| 497 |
"learning_rate": 9.46992601503499e-06,
|
| 498 |
-
"loss": 0.
|
| 499 |
"step": 345
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.4727068092290377,
|
| 503 |
-
"grad_norm": 0.
|
| 504 |
"learning_rate": 9.44319433326178e-06,
|
| 505 |
-
"loss": 0.
|
| 506 |
"step": 350
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.47945976364659537,
|
| 510 |
-
"grad_norm": 0.
|
| 511 |
"learning_rate": 9.415844746673047e-06,
|
| 512 |
-
"loss": 0.
|
| 513 |
"step": 355
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.48621271806415306,
|
| 517 |
-
"grad_norm": 0.
|
| 518 |
"learning_rate": 9.387881058712888e-06,
|
| 519 |
-
"loss": 0.
|
| 520 |
"step": 360
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 0.49296567248171075,
|
| 524 |
-
"grad_norm": 0.
|
| 525 |
"learning_rate": 9.359307158227067e-06,
|
| 526 |
-
"loss": 0.
|
| 527 |
"step": 365
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.49971862689926844,
|
| 531 |
-
"grad_norm": 0.
|
| 532 |
"learning_rate": 9.330127018922195e-06,
|
| 533 |
-
"loss": 0.
|
| 534 |
"step": 370
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.5064715813168261,
|
| 538 |
-
"grad_norm": 0.
|
| 539 |
"learning_rate": 9.300344698813124e-06,
|
| 540 |
-
"loss": 0.
|
| 541 |
"step": 375
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.5132245357343838,
|
| 545 |
-
"grad_norm": 0.
|
| 546 |
"learning_rate": 9.269964339658605e-06,
|
| 547 |
-
"loss": 0.
|
| 548 |
"step": 380
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.5199774901519415,
|
| 552 |
-
"grad_norm": 0.
|
| 553 |
"learning_rate": 9.238990166385304e-06,
|
| 554 |
-
"loss": 0.
|
| 555 |
"step": 385
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.5267304445694991,
|
| 559 |
-
"grad_norm": 0.
|
| 560 |
"learning_rate": 9.207426486500252e-06,
|
| 561 |
"loss": 0.3587,
|
| 562 |
"step": 390
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 0.5334833989870569,
|
| 566 |
-
"grad_norm": 0.
|
| 567 |
"learning_rate": 9.175277689491804e-06,
|
| 568 |
-
"loss": 0.
|
| 569 |
"step": 395
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.5402363534046145,
|
| 573 |
-
"grad_norm": 0.
|
| 574 |
"learning_rate": 9.142548246219212e-06,
|
| 575 |
-
"loss": 0.
|
| 576 |
"step": 400
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 0.5469893078221723,
|
| 580 |
-
"grad_norm": 0.
|
| 581 |
"learning_rate": 9.109242708290864e-06,
|
| 582 |
"loss": 0.3427,
|
| 583 |
"step": 405
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.5537422622397299,
|
| 587 |
-
"grad_norm": 0.
|
| 588 |
"learning_rate": 9.075365707431311e-06,
|
| 589 |
-
"loss": 0.
|
| 590 |
"step": 410
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.5604952166572875,
|
| 594 |
-
"grad_norm": 0.
|
| 595 |
"learning_rate": 9.040921954837139e-06,
|
| 596 |
-
"loss": 0.
|
| 597 |
"step": 415
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.5672481710748453,
|
| 601 |
-
"grad_norm": 0.
|
| 602 |
"learning_rate": 9.005916240521788e-06,
|
| 603 |
-
"loss": 0.
|
| 604 |
"step": 420
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 0.5740011254924029,
|
| 608 |
-
"grad_norm": 0.
|
| 609 |
"learning_rate": 8.97035343264943e-06,
|
| 610 |
-
"loss": 0.
|
| 611 |
"step": 425
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.5807540799099606,
|
| 615 |
-
"grad_norm": 0.
|
| 616 |
"learning_rate": 8.93423847685795e-06,
|
| 617 |
-
"loss": 0.
|
| 618 |
"step": 430
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 0.5875070343275183,
|
| 622 |
-
"grad_norm": 0.
|
| 623 |
"learning_rate": 8.89757639557118e-06,
|
| 624 |
-
"loss": 0.
|
| 625 |
"step": 435
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.594259988745076,
|
| 629 |
-
"grad_norm": 0.
|
| 630 |
"learning_rate": 8.860372287300432e-06,
|
| 631 |
"loss": 0.3703,
|
| 632 |
"step": 440
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 0.6010129431626337,
|
| 636 |
-
"grad_norm": 0.
|
| 637 |
"learning_rate": 8.822631325935463e-06,
|
| 638 |
-
"loss": 0.
|
| 639 |
"step": 445
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.6077658975801913,
|
| 643 |
-
"grad_norm": 0.
|
| 644 |
"learning_rate": 8.78435876002496e-06,
|
| 645 |
-
"loss": 0.
|
| 646 |
"step": 450
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 0.614518851997749,
|
| 650 |
-
"grad_norm": 0.
|
| 651 |
"learning_rate": 8.745559912046625e-06,
|
| 652 |
-
"loss": 0.
|
| 653 |
"step": 455
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.6212718064153067,
|
| 657 |
-
"grad_norm": 0.
|
| 658 |
"learning_rate": 8.706240177667003e-06,
|
| 659 |
"loss": 0.3476,
|
| 660 |
"step": 460
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 0.6280247608328644,
|
| 664 |
-
"grad_norm": 0.
|
| 665 |
"learning_rate": 8.666405024991105e-06,
|
| 666 |
-
"loss": 0.
|
| 667 |
"step": 465
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.634777715250422,
|
| 671 |
-
"grad_norm": 0.
|
| 672 |
"learning_rate": 8.626059993801988e-06,
|
| 673 |
-
"loss": 0.
|
| 674 |
"step": 470
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.6415306696679798,
|
| 678 |
-
"grad_norm": 0.
|
| 679 |
"learning_rate": 8.585210694790333e-06,
|
| 680 |
-
"loss": 0.
|
| 681 |
"step": 475
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.6482836240855374,
|
| 685 |
-
"grad_norm": 0.
|
| 686 |
"learning_rate": 8.543862808774193e-06,
|
| 687 |
-
"loss": 0.
|
| 688 |
"step": 480
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 0.6550365785030952,
|
| 692 |
-
"grad_norm": 0.
|
| 693 |
"learning_rate": 8.502022085908963e-06,
|
| 694 |
"loss": 0.3471,
|
| 695 |
"step": 485
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.6617895329206528,
|
| 699 |
-
"grad_norm": 0.
|
| 700 |
"learning_rate": 8.459694344887732e-06,
|
| 701 |
-
"loss": 0.
|
| 702 |
"step": 490
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 0.6685424873382104,
|
| 706 |
-
"grad_norm": 0.
|
| 707 |
"learning_rate": 8.416885472132077e-06,
|
| 708 |
-
"loss": 0.
|
| 709 |
"step": 495
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.6752954417557682,
|
| 713 |
-
"grad_norm": 0.
|
| 714 |
"learning_rate": 8.373601420973464e-06,
|
| 715 |
-
"loss": 0.
|
| 716 |
"step": 500
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 0.6820483961733258,
|
| 720 |
-
"grad_norm": 0.
|
| 721 |
"learning_rate": 8.329848210825322e-06,
|
| 722 |
-
"loss": 0.
|
| 723 |
"step": 505
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.6888013505908835,
|
| 727 |
-
"grad_norm": 0.
|
| 728 |
"learning_rate": 8.285631926345943e-06,
|
| 729 |
-
"loss": 0.
|
| 730 |
"step": 510
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 0.6955543050084412,
|
| 734 |
-
"grad_norm": 0.
|
| 735 |
"learning_rate": 8.240958716592304e-06,
|
| 736 |
-
"loss": 0.
|
| 737 |
"step": 515
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.7023072594259989,
|
| 741 |
-
"grad_norm": 0.
|
| 742 |
"learning_rate": 8.195834794164925e-06,
|
| 743 |
"loss": 0.3501,
|
| 744 |
"step": 520
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 0.7090602138435566,
|
| 748 |
-
"grad_norm": 0.
|
| 749 |
"learning_rate": 8.150266434343904e-06,
|
| 750 |
"loss": 0.3296,
|
| 751 |
"step": 525
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.7158131682611142,
|
| 755 |
-
"grad_norm": 0.
|
| 756 |
"learning_rate": 8.104259974216219e-06,
|
| 757 |
-
"loss": 0.
|
| 758 |
"step": 530
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.7225661226786719,
|
| 762 |
-
"grad_norm": 0.
|
| 763 |
"learning_rate": 8.057821811794457e-06,
|
| 764 |
-
"loss": 0.
|
| 765 |
"step": 535
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.7293190770962296,
|
| 769 |
-
"grad_norm": 0.
|
| 770 |
"learning_rate": 8.010958405127048e-06,
|
| 771 |
-
"loss": 0.
|
| 772 |
"step": 540
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 0.7360720315137873,
|
| 776 |
-
"grad_norm": 0.
|
| 777 |
"learning_rate": 7.963676271400158e-06,
|
| 778 |
-
"loss": 0.
|
| 779 |
"step": 545
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.7428249859313449,
|
| 783 |
-
"grad_norm": 0.
|
| 784 |
"learning_rate": 7.915981986031367e-06,
|
| 785 |
-
"loss": 0.
|
| 786 |
"step": 550
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 0.7495779403489027,
|
| 790 |
-
"grad_norm": 0.
|
| 791 |
"learning_rate": 7.86788218175523e-06,
|
| 792 |
-
"loss": 0.
|
| 793 |
"step": 555
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.7563308947664603,
|
| 797 |
-
"grad_norm": 0.
|
| 798 |
"learning_rate": 7.819383547700889e-06,
|
| 799 |
"loss": 0.3354,
|
| 800 |
"step": 560
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 0.7630838491840181,
|
| 804 |
-
"grad_norm": 0.
|
| 805 |
"learning_rate": 7.770492828461824e-06,
|
| 806 |
-
"loss": 0.
|
| 807 |
"step": 565
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.7698368036015757,
|
| 811 |
-
"grad_norm": 0.
|
| 812 |
"learning_rate": 7.721216823157896e-06,
|
| 813 |
-
"loss": 0.
|
| 814 |
"step": 570
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 0.7765897580191333,
|
| 818 |
-
"grad_norm": 0.
|
| 819 |
"learning_rate": 7.671562384489819e-06,
|
| 820 |
-
"loss": 0.
|
| 821 |
"step": 575
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.7833427124366911,
|
| 825 |
-
"grad_norm": 0.
|
| 826 |
"learning_rate": 7.621536417786159e-06,
|
| 827 |
-
"loss": 0.
|
| 828 |
"step": 580
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 0.7900956668542487,
|
| 832 |
-
"grad_norm": 0.
|
| 833 |
"learning_rate": 7.571145880043036e-06,
|
| 834 |
-
"loss": 0.
|
| 835 |
"step": 585
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.7968486212718064,
|
| 839 |
-
"grad_norm": 0.
|
| 840 |
"learning_rate": 7.520397778956623e-06,
|
| 841 |
-
"loss": 0.
|
| 842 |
"step": 590
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.8036015756893641,
|
| 846 |
-
"grad_norm": 0.
|
| 847 |
"learning_rate": 7.469299171948608e-06,
|
| 848 |
-
"loss": 0.
|
| 849 |
"step": 595
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.8103545301069218,
|
| 853 |
-
"grad_norm": 0.
|
| 854 |
"learning_rate": 7.417857165184723e-06,
|
| 855 |
"loss": 0.3474,
|
| 856 |
"step": 600
|
| 857 |
},
|
| 858 |
{
|
| 859 |
"epoch": 0.8171074845244795,
|
| 860 |
-
"grad_norm": 0.
|
| 861 |
"learning_rate": 7.366078912586523e-06,
|
| 862 |
-
"loss": 0.
|
| 863 |
"step": 605
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 0.8238604389420371,
|
| 867 |
-
"grad_norm": 0.
|
| 868 |
"learning_rate": 7.313971614836496e-06,
|
| 869 |
-
"loss": 0.
|
| 870 |
"step": 610
|
| 871 |
},
|
| 872 |
{
|
| 873 |
"epoch": 0.8306133933595948,
|
| 874 |
-
"grad_norm": 0.
|
| 875 |
"learning_rate": 7.261542518376677e-06,
|
| 876 |
-
"loss": 0.
|
| 877 |
"step": 615
|
| 878 |
},
|
| 879 |
{
|
| 880 |
"epoch": 0.8373663477771525,
|
| 881 |
-
"grad_norm": 0.
|
| 882 |
"learning_rate": 7.208798914400916e-06,
|
| 883 |
-
"loss": 0.
|
| 884 |
"step": 620
|
| 885 |
},
|
| 886 |
{
|
| 887 |
"epoch": 0.8441193021947102,
|
| 888 |
-
"grad_norm": 0.
|
| 889 |
"learning_rate": 7.155748137840892e-06,
|
| 890 |
-
"loss": 0.
|
| 891 |
"step": 625
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 0.8508722566122678,
|
| 895 |
-
"grad_norm": 0.
|
| 896 |
"learning_rate": 7.102397566346073e-06,
|
| 897 |
-
"loss": 0.
|
| 898 |
"step": 630
|
| 899 |
},
|
| 900 |
{
|
| 901 |
"epoch": 0.8576252110298256,
|
| 902 |
-
"grad_norm": 0.
|
| 903 |
"learning_rate": 7.048754619257716e-06,
|
| 904 |
-
"loss": 0.
|
| 905 |
"step": 635
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.8643781654473832,
|
| 909 |
-
"grad_norm": 0.
|
| 910 |
"learning_rate": 6.994826756577082e-06,
|
| 911 |
-
"loss": 0.
|
| 912 |
"step": 640
|
| 913 |
},
|
| 914 |
{
|
| 915 |
"epoch": 0.871131119864941,
|
| 916 |
-
"grad_norm": 0.
|
| 917 |
"learning_rate": 6.940621477927988e-06,
|
| 918 |
-
"loss": 0.
|
| 919 |
"step": 645
|
| 920 |
},
|
| 921 |
{
|
| 922 |
"epoch": 0.8778840742824986,
|
| 923 |
-
"grad_norm": 0.
|
| 924 |
"learning_rate": 6.88614632151385e-06,
|
| 925 |
-
"loss": 0.
|
| 926 |
"step": 650
|
| 927 |
},
|
| 928 |
{
|
| 929 |
"epoch": 0.8846370287000562,
|
| 930 |
-
"grad_norm": 0.
|
| 931 |
"learning_rate": 6.831408863069364e-06,
|
| 932 |
-
"loss": 0.
|
| 933 |
"step": 655
|
| 934 |
},
|
| 935 |
{
|
| 936 |
"epoch": 0.891389983117614,
|
| 937 |
-
"grad_norm": 0.
|
| 938 |
"learning_rate": 6.7764167148069695e-06,
|
| 939 |
"loss": 0.3251,
|
| 940 |
"step": 660
|
| 941 |
},
|
| 942 |
{
|
| 943 |
"epoch": 0.8981429375351716,
|
| 944 |
-
"grad_norm": 0.
|
| 945 |
"learning_rate": 6.721177524358226e-06,
|
| 946 |
-
"loss": 0.
|
| 947 |
"step": 665
|
| 948 |
},
|
| 949 |
{
|
| 950 |
"epoch": 0.9048958919527293,
|
| 951 |
-
"grad_norm": 0.
|
| 952 |
"learning_rate": 6.665698973710289e-06,
|
| 953 |
-
"loss": 0.
|
| 954 |
"step": 670
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 0.911648846370287,
|
| 958 |
-
"grad_norm": 0.
|
| 959 |
"learning_rate": 6.609988778137582e-06,
|
| 960 |
-
"loss": 0.
|
| 961 |
"step": 675
|
| 962 |
},
|
| 963 |
{
|
| 964 |
"epoch": 0.9184018007878447,
|
| 965 |
-
"grad_norm": 0.
|
| 966 |
"learning_rate": 6.554054685128857e-06,
|
| 967 |
-
"loss": 0.
|
| 968 |
"step": 680
|
| 969 |
},
|
| 970 |
{
|
| 971 |
"epoch": 0.9251547552054024,
|
| 972 |
-
"grad_norm": 0.
|
| 973 |
"learning_rate": 6.497904473309766e-06,
|
| 974 |
"loss": 0.3519,
|
| 975 |
"step": 685
|
| 976 |
},
|
| 977 |
{
|
| 978 |
"epoch": 0.93190770962296,
|
| 979 |
-
"grad_norm": 0.
|
| 980 |
"learning_rate": 6.44154595136111e-06,
|
| 981 |
-
"loss": 0.
|
| 982 |
"step": 690
|
| 983 |
},
|
| 984 |
{
|
| 985 |
"epoch": 0.9386606640405177,
|
| 986 |
-
"grad_norm": 0.
|
| 987 |
"learning_rate": 6.384986956932897e-06,
|
| 988 |
"loss": 0.3339,
|
| 989 |
"step": 695
|
| 990 |
},
|
| 991 |
{
|
| 992 |
"epoch": 0.9454136184580754,
|
| 993 |
-
"grad_norm": 0.
|
| 994 |
"learning_rate": 6.328235355554382e-06,
|
| 995 |
-
"loss": 0.
|
| 996 |
"step": 700
|
| 997 |
},
|
| 998 |
{
|
| 999 |
"epoch": 0.9521665728756331,
|
| 1000 |
-
"grad_norm": 0.
|
| 1001 |
"learning_rate": 6.271299039540228e-06,
|
| 1002 |
"loss": 0.3645,
|
| 1003 |
"step": 705
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
"epoch": 0.9589195272931907,
|
| 1007 |
-
"grad_norm": 0.
|
| 1008 |
"learning_rate": 6.214185926892936e-06,
|
| 1009 |
-
"loss": 0.
|
| 1010 |
"step": 710
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
"epoch": 0.9656724817107485,
|
| 1014 |
-
"grad_norm": 0.
|
| 1015 |
"learning_rate": 6.156903960201709e-06,
|
| 1016 |
-
"loss": 0.
|
| 1017 |
"step": 715
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.9724254361283061,
|
| 1021 |
-
"grad_norm": 0.
|
| 1022 |
"learning_rate": 6.099461105537889e-06,
|
| 1023 |
-
"loss": 0.
|
| 1024 |
"step": 720
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
"epoch": 0.9791783905458639,
|
| 1028 |
-
"grad_norm": 0.
|
| 1029 |
"learning_rate": 6.041865351347146e-06,
|
| 1030 |
"loss": 0.3421,
|
| 1031 |
"step": 725
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
"epoch": 0.9859313449634215,
|
| 1035 |
-
"grad_norm": 0.
|
| 1036 |
"learning_rate": 5.984124707338528e-06,
|
| 1037 |
-
"loss": 0.
|
| 1038 |
"step": 730
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
"epoch": 0.9926842993809791,
|
| 1042 |
-
"grad_norm": 0.
|
| 1043 |
"learning_rate": 5.926247203370583e-06,
|
| 1044 |
-
"loss": 0.
|
| 1045 |
"step": 735
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 0.9994372537985369,
|
| 1049 |
-
"grad_norm": 0.
|
| 1050 |
"learning_rate": 5.8682408883346535e-06,
|
| 1051 |
"loss": 0.2909,
|
| 1052 |
"step": 740
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 1.0,
|
| 1056 |
-
"eval_loss": 0.
|
| 1057 |
-
"eval_runtime": 18.
|
| 1058 |
-
"eval_samples_per_second": 19.
|
| 1059 |
-
"eval_steps_per_second": 4.
|
| 1060 |
"step": 741
|
| 1061 |
},
|
| 1062 |
{
|
| 1063 |
"epoch": 1.005402363534046,
|
| 1064 |
-
"grad_norm": 0.
|
| 1065 |
"learning_rate": 5.810113829035544e-06,
|
| 1066 |
-
"loss": 0.
|
| 1067 |
"step": 745
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
"epoch": 1.0121553179516039,
|
| 1071 |
-
"grad_norm": 0.
|
| 1072 |
"learning_rate": 5.751874109069685e-06,
|
| 1073 |
"loss": 0.3611,
|
| 1074 |
"step": 750
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 1.0189082723691616,
|
| 1078 |
-
"grad_norm": 0.
|
| 1079 |
"learning_rate": 5.693529827700967e-06,
|
| 1080 |
-
"loss": 0.
|
| 1081 |
"step": 755
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
"epoch": 1.0256612267867191,
|
| 1085 |
-
"grad_norm": 0.
|
| 1086 |
"learning_rate": 5.635089098734394e-06,
|
| 1087 |
-
"loss": 0.
|
| 1088 |
"step": 760
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
"epoch": 1.0324141812042769,
|
| 1092 |
-
"grad_norm": 0.
|
| 1093 |
"learning_rate": 5.576560049387717e-06,
|
| 1094 |
"loss": 0.339,
|
| 1095 |
"step": 765
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
"epoch": 1.0391671356218346,
|
| 1099 |
-
"grad_norm": 0.
|
| 1100 |
"learning_rate": 5.517950819161196e-06,
|
| 1101 |
"loss": 0.328,
|
| 1102 |
"step": 770
|
| 1103 |
},
|
| 1104 |
{
|
| 1105 |
"epoch": 1.0459200900393921,
|
| 1106 |
-
"grad_norm": 0.
|
| 1107 |
"learning_rate": 5.459269558705667e-06,
|
| 1108 |
-
"loss": 0.
|
| 1109 |
"step": 775
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
"epoch": 1.0526730444569499,
|
| 1113 |
-
"grad_norm": 0.
|
| 1114 |
"learning_rate": 5.400524428689035e-06,
|
| 1115 |
-
"loss": 0.
|
| 1116 |
"step": 780
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
"epoch": 1.0594259988745076,
|
| 1120 |
-
"grad_norm": 0.
|
| 1121 |
"learning_rate": 5.341723598661409e-06,
|
| 1122 |
"loss": 0.3267,
|
| 1123 |
"step": 785
|
| 1124 |
},
|
| 1125 |
{
|
| 1126 |
"epoch": 1.0661789532920654,
|
| 1127 |
-
"grad_norm": 0.
|
| 1128 |
"learning_rate": 5.282875245918963e-06,
|
| 1129 |
-
"loss": 0.
|
| 1130 |
"step": 790
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
"epoch": 1.0729319077096229,
|
| 1134 |
-
"grad_norm": 0.
|
| 1135 |
"learning_rate": 5.2239875543667465e-06,
|
| 1136 |
-
"loss": 0.
|
| 1137 |
"step": 795
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
"epoch": 1.0796848621271806,
|
| 1141 |
-
"grad_norm": 0.
|
| 1142 |
"learning_rate": 5.165068713380568e-06,
|
| 1143 |
-
"loss": 0.
|
| 1144 |
"step": 800
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
"epoch": 1.0864378165447384,
|
| 1148 |
-
"grad_norm": 0.
|
| 1149 |
"learning_rate": 5.106126916668118e-06,
|
| 1150 |
"loss": 0.3192,
|
| 1151 |
"step": 805
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
"epoch": 1.093190770962296,
|
| 1155 |
-
"grad_norm": 0.
|
| 1156 |
"learning_rate": 5.047170361129484e-06,
|
| 1157 |
-
"loss": 0.
|
| 1158 |
"step": 810
|
| 1159 |
},
|
| 1160 |
{
|
| 1161 |
"epoch": 1.0999437253798536,
|
| 1162 |
-
"grad_norm": 0.
|
| 1163 |
"learning_rate": 4.988207245717232e-06,
|
| 1164 |
"loss": 0.3558,
|
| 1165 |
"step": 815
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
"epoch": 1.1066966797974114,
|
| 1169 |
-
"grad_norm": 0.
|
| 1170 |
"learning_rate": 4.929245770296191e-06,
|
| 1171 |
-
"loss": 0.
|
| 1172 |
"step": 820
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
"epoch": 1.1134496342149691,
|
| 1176 |
-
"grad_norm": 0.
|
| 1177 |
"learning_rate": 4.870294134503123e-06,
|
| 1178 |
-
"loss": 0.
|
| 1179 |
"step": 825
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 1.1202025886325266,
|
| 1183 |
-
"grad_norm": 0.
|
| 1184 |
"learning_rate": 4.811360536606416e-06,
|
| 1185 |
-
"loss": 0.
|
| 1186 |
"step": 830
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
"epoch": 1.1269555430500844,
|
| 1190 |
-
"grad_norm": 0.
|
| 1191 |
"learning_rate": 4.752453172365966e-06,
|
| 1192 |
-
"loss": 0.
|
| 1193 |
"step": 835
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 1.1337084974676421,
|
| 1197 |
-
"grad_norm": 0.
|
| 1198 |
"learning_rate": 4.69358023389342e-06,
|
| 1199 |
-
"loss": 0.
|
| 1200 |
"step": 840
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 1.1404614518851997,
|
| 1204 |
-
"grad_norm": 0.
|
| 1205 |
"learning_rate": 4.634749908512907e-06,
|
| 1206 |
-
"loss": 0.
|
| 1207 |
"step": 845
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
"epoch": 1.1472144063027574,
|
| 1211 |
-
"grad_norm": 0.
|
| 1212 |
"learning_rate": 4.575970377622456e-06,
|
| 1213 |
-
"loss": 0.
|
| 1214 |
"step": 850
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
"epoch": 1.1539673607203151,
|
| 1218 |
-
"grad_norm": 0.
|
| 1219 |
"learning_rate": 4.517249815556219e-06,
|
| 1220 |
-
"loss": 0.
|
| 1221 |
"step": 855
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
"epoch": 1.1607203151378729,
|
| 1225 |
-
"grad_norm": 0.
|
| 1226 |
"learning_rate": 4.458596388447691e-06,
|
| 1227 |
-
"loss": 0.
|
| 1228 |
"step": 860
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
"epoch": 1.1674732695554304,
|
| 1232 |
-
"grad_norm": 0.
|
| 1233 |
"learning_rate": 4.400018253094065e-06,
|
| 1234 |
-
"loss": 0.
|
| 1235 |
"step": 865
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
"epoch": 1.1742262239729881,
|
| 1239 |
-
"grad_norm": 0.
|
| 1240 |
"learning_rate": 4.341523555821882e-06,
|
| 1241 |
-
"loss": 0.
|
| 1242 |
"step": 870
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 1.180979178390546,
|
| 1246 |
-
"grad_norm": 0.
|
| 1247 |
"learning_rate": 4.283120431354137e-06,
|
| 1248 |
-
"loss": 0.
|
| 1249 |
"step": 875
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 1.1877321328081036,
|
| 1253 |
-
"grad_norm": 0.
|
| 1254 |
"learning_rate": 4.224817001679011e-06,
|
| 1255 |
-
"loss": 0.
|
| 1256 |
"step": 880
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
"epoch": 1.1944850872256612,
|
| 1260 |
-
"grad_norm": 0.
|
| 1261 |
"learning_rate": 4.1666213749203545e-06,
|
| 1262 |
"loss": 0.3237,
|
| 1263 |
"step": 885
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
"epoch": 1.201238041643219,
|
| 1267 |
-
"grad_norm": 0.
|
| 1268 |
"learning_rate": 4.1085416442101205e-06,
|
| 1269 |
-
"loss": 0.
|
| 1270 |
"step": 890
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 1.2079909960607766,
|
| 1274 |
-
"grad_norm": 0.
|
| 1275 |
"learning_rate": 4.050585886562858e-06,
|
| 1276 |
-
"loss": 0.
|
| 1277 |
"step": 895
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
"epoch": 1.2147439504783342,
|
| 1281 |
-
"grad_norm": 0.
|
| 1282 |
"learning_rate": 3.992762161752474e-06,
|
| 1283 |
-
"loss": 0.
|
| 1284 |
"step": 900
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 1.221496904895892,
|
| 1288 |
-
"grad_norm": 0.
|
| 1289 |
"learning_rate": 3.935078511191368e-06,
|
| 1290 |
-
"loss": 0.
|
| 1291 |
"step": 905
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
"epoch": 1.2282498593134497,
|
| 1295 |
-
"grad_norm": 0.
|
| 1296 |
"learning_rate": 3.877542956812137e-06,
|
| 1297 |
"loss": 0.3221,
|
| 1298 |
"step": 910
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
"epoch": 1.2350028137310074,
|
| 1302 |
-
"grad_norm": 0.
|
| 1303 |
"learning_rate": 3.820163499951984e-06,
|
| 1304 |
"loss": 0.3579,
|
| 1305 |
"step": 915
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 1.241755768148565,
|
| 1309 |
-
"grad_norm": 0.
|
| 1310 |
"learning_rate": 3.7629481202399886e-06,
|
| 1311 |
-
"loss": 0.
|
| 1312 |
"step": 920
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 1.2485087225661227,
|
| 1316 |
-
"grad_norm": 0.
|
| 1317 |
"learning_rate": 3.705904774487396e-06,
|
| 1318 |
-
"loss": 0.
|
| 1319 |
"step": 925
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
"epoch": 1.2552616769836804,
|
| 1323 |
-
"grad_norm": 0.
|
| 1324 |
"learning_rate": 3.64904139558109e-06,
|
| 1325 |
-
"loss": 0.
|
| 1326 |
"step": 930
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
"epoch": 1.2620146314012382,
|
| 1330 |
-
"grad_norm": 0.
|
| 1331 |
"learning_rate": 3.5923658913803726e-06,
|
| 1332 |
-
"loss": 0.
|
| 1333 |
"step": 935
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
"epoch": 1.2687675858187957,
|
| 1337 |
-
"grad_norm": 0.
|
| 1338 |
"learning_rate": 3.5358861436172487e-06,
|
| 1339 |
-
"loss": 0.
|
| 1340 |
"step": 940
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 1.2755205402363534,
|
| 1344 |
-
"grad_norm": 0.
|
| 1345 |
"learning_rate": 3.47961000680032e-06,
|
| 1346 |
-
"loss": 0.
|
| 1347 |
"step": 945
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
"epoch": 1.2822734946539112,
|
| 1351 |
-
"grad_norm": 0.
|
| 1352 |
"learning_rate": 3.4235453071224882e-06,
|
| 1353 |
-
"loss": 0.
|
| 1354 |
"step": 950
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
"epoch": 1.2890264490714687,
|
| 1358 |
-
"grad_norm": 0.
|
| 1359 |
"learning_rate": 3.3676998413725726e-06,
|
| 1360 |
-
"loss": 0.
|
| 1361 |
"step": 955
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
"epoch": 1.2957794034890264,
|
| 1365 |
-
"grad_norm": 0.
|
| 1366 |
"learning_rate": 3.3120813758510385e-06,
|
| 1367 |
-
"loss": 0.
|
| 1368 |
"step": 960
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
"epoch": 1.3025323579065842,
|
| 1372 |
-
"grad_norm": 0.
|
| 1373 |
"learning_rate": 3.2566976452899507e-06,
|
| 1374 |
-
"loss": 0.
|
| 1375 |
"step": 965
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
"epoch": 1.3092853123241417,
|
| 1379 |
-
"grad_norm": 0.
|
| 1380 |
"learning_rate": 3.2015563517773214e-06,
|
| 1381 |
-
"loss": 0.
|
| 1382 |
"step": 970
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
"epoch": 1.3160382667416994,
|
| 1386 |
-
"grad_norm": 0.
|
| 1387 |
"learning_rate": 3.1466651636860025e-06,
|
| 1388 |
"loss": 0.3325,
|
| 1389 |
"step": 975
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
"epoch": 1.3227912211592572,
|
| 1393 |
-
"grad_norm": 0.
|
| 1394 |
"learning_rate": 3.0920317146072577e-06,
|
| 1395 |
-
"loss": 0.
|
| 1396 |
"step": 980
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
"epoch": 1.329544175576815,
|
| 1400 |
-
"grad_norm": 0.
|
| 1401 |
"learning_rate": 3.0376636022891813e-06,
|
| 1402 |
-
"loss": 0.
|
| 1403 |
"step": 985
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
"epoch": 1.3362971299943727,
|
| 1407 |
-
"grad_norm": 0.
|
| 1408 |
"learning_rate": 2.983568387580094e-06,
|
| 1409 |
-
"loss": 0.
|
| 1410 |
"step": 990
|
| 1411 |
},
|
| 1412 |
{
|
| 1413 |
"epoch": 1.3430500844119302,
|
| 1414 |
-
"grad_norm": 0.
|
| 1415 |
"learning_rate": 2.9297535933770732e-06,
|
| 1416 |
-
"loss": 0.
|
| 1417 |
"step": 995
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
"epoch": 1.349803038829488,
|
| 1421 |
-
"grad_norm": 0.
|
| 1422 |
"learning_rate": 2.8762267035797607e-06,
|
| 1423 |
-
"loss": 0.
|
| 1424 |
"step": 1000
|
| 1425 |
},
|
| 1426 |
{
|
| 1427 |
"epoch": 1.3565559932470457,
|
| 1428 |
-
"grad_norm": 0.
|
| 1429 |
"learning_rate": 2.822995162049599e-06,
|
| 1430 |
-
"loss": 0.
|
| 1431 |
"step": 1005
|
| 1432 |
},
|
| 1433 |
{
|
| 1434 |
"epoch": 1.3633089476646032,
|
| 1435 |
-
"grad_norm": 0.
|
| 1436 |
"learning_rate": 2.7700663715746213e-06,
|
| 1437 |
-
"loss": 0.
|
| 1438 |
"step": 1010
|
| 1439 |
},
|
| 1440 |
{
|
| 1441 |
"epoch": 1.370061902082161,
|
| 1442 |
-
"grad_norm": 0.
|
| 1443 |
"learning_rate": 2.7174476928399685e-06,
|
| 1444 |
-
"loss": 0.
|
| 1445 |
"step": 1015
|
| 1446 |
},
|
| 1447 |
{
|
| 1448 |
"epoch": 1.3768148564997187,
|
| 1449 |
-
"grad_norm": 0.
|
| 1450 |
"learning_rate": 2.66514644340426e-06,
|
| 1451 |
"loss": 0.3425,
|
| 1452 |
"step": 1020
|
| 1453 |
},
|
| 1454 |
{
|
| 1455 |
"epoch": 1.3835678109172762,
|
| 1456 |
-
"grad_norm": 0.
|
| 1457 |
"learning_rate": 2.613169896681949e-06,
|
| 1458 |
-
"loss": 0.
|
| 1459 |
"step": 1025
|
| 1460 |
},
|
| 1461 |
{
|
| 1462 |
"epoch": 1.390320765334834,
|
| 1463 |
-
"grad_norm": 0.
|
| 1464 |
"learning_rate": 2.5615252809318287e-06,
|
| 1465 |
-
"loss": 0.
|
| 1466 |
"step": 1030
|
| 1467 |
},
|
| 1468 |
{
|
| 1469 |
"epoch": 1.3970737197523917,
|
| 1470 |
-
"grad_norm": 0.
|
| 1471 |
"learning_rate": 2.5102197782518145e-06,
|
| 1472 |
-
"loss": 0.
|
| 1473 |
"step": 1035
|
| 1474 |
},
|
| 1475 |
{
|
| 1476 |
"epoch": 1.4038266741699492,
|
| 1477 |
-
"grad_norm": 0.
|
| 1478 |
"learning_rate": 2.4592605235801544e-06,
|
| 1479 |
-
"loss": 0.
|
| 1480 |
"step": 1040
|
| 1481 |
},
|
| 1482 |
{
|
| 1483 |
"epoch": 1.410579628587507,
|
| 1484 |
-
"grad_norm": 0.
|
| 1485 |
"learning_rate": 2.4086546037031734e-06,
|
| 1486 |
-
"loss": 0.
|
| 1487 |
"step": 1045
|
| 1488 |
},
|
| 1489 |
{
|
| 1490 |
"epoch": 1.4173325830050647,
|
| 1491 |
-
"grad_norm": 0.
|
| 1492 |
"learning_rate": 2.3584090562697427e-06,
|
| 1493 |
-
"loss": 0.
|
| 1494 |
"step": 1050
|
| 1495 |
},
|
| 1496 |
{
|
| 1497 |
"epoch": 1.4240855374226224,
|
| 1498 |
-
"grad_norm": 0.
|
| 1499 |
"learning_rate": 2.3085308688125695e-06,
|
| 1500 |
-
"loss": 0.
|
| 1501 |
"step": 1055
|
| 1502 |
},
|
| 1503 |
{
|
| 1504 |
"epoch": 1.4308384918401802,
|
| 1505 |
-
"grad_norm": 0.
|
| 1506 |
"learning_rate": 2.2590269777764516e-06,
|
| 1507 |
-
"loss": 0.
|
| 1508 |
"step": 1060
|
| 1509 |
},
|
| 1510 |
{
|
| 1511 |
"epoch": 1.4375914462577377,
|
| 1512 |
-
"grad_norm": 0.
|
| 1513 |
"learning_rate": 2.2099042675536437e-06,
|
| 1514 |
-
"loss": 0.
|
| 1515 |
"step": 1065
|
| 1516 |
},
|
| 1517 |
{
|
| 1518 |
"epoch": 1.4443444006752955,
|
| 1519 |
-
"grad_norm": 0.
|
| 1520 |
"learning_rate": 2.161169569526461e-06,
|
| 1521 |
-
"loss": 0.
|
| 1522 |
"step": 1070
|
| 1523 |
},
|
| 1524 |
{
|
| 1525 |
"epoch": 1.4510973550928532,
|
| 1526 |
-
"grad_norm": 0.
|
| 1527 |
"learning_rate": 2.1128296611172593e-06,
|
| 1528 |
-
"loss": 0.
|
| 1529 |
"step": 1075
|
| 1530 |
},
|
| 1531 |
{
|
| 1532 |
"epoch": 1.4578503095104107,
|
| 1533 |
-
"grad_norm": 0.
|
| 1534 |
"learning_rate": 2.0648912648459072e-06,
|
| 1535 |
"loss": 0.3447,
|
| 1536 |
"step": 1080
|
| 1537 |
},
|
| 1538 |
{
|
| 1539 |
"epoch": 1.4646032639279685,
|
| 1540 |
-
"grad_norm": 0.
|
| 1541 |
"learning_rate": 2.0173610473949048e-06,
|
| 1542 |
-
"loss": 0.
|
| 1543 |
"step": 1085
|
| 1544 |
},
|
| 1545 |
{
|
| 1546 |
"epoch": 1.4713562183455262,
|
| 1547 |
-
"grad_norm": 0.
|
| 1548 |
"learning_rate": 1.9702456186822595e-06,
|
| 1549 |
-
"loss": 0.
|
| 1550 |
"step": 1090
|
| 1551 |
},
|
| 1552 |
{
|
| 1553 |
"epoch": 1.4781091727630837,
|
| 1554 |
-
"grad_norm": 0.
|
| 1555 |
"learning_rate": 1.9235515309422685e-06,
|
| 1556 |
-
"loss": 0.
|
| 1557 |
"step": 1095
|
| 1558 |
},
|
| 1559 |
{
|
| 1560 |
"epoch": 1.4848621271806415,
|
| 1561 |
-
"grad_norm": 0.
|
| 1562 |
"learning_rate": 1.8772852778143064e-06,
|
| 1563 |
-
"loss": 0.
|
| 1564 |
"step": 1100
|
| 1565 |
},
|
| 1566 |
{
|
| 1567 |
"epoch": 1.4916150815981992,
|
| 1568 |
-
"grad_norm": 0.
|
| 1569 |
"learning_rate": 1.831453293439771e-06,
|
| 1570 |
-
"loss": 0.
|
| 1571 |
"step": 1105
|
| 1572 |
},
|
| 1573 |
{
|
| 1574 |
"epoch": 1.498368036015757,
|
| 1575 |
-
"grad_norm": 0.
|
| 1576 |
"learning_rate": 1.7860619515673034e-06,
|
| 1577 |
-
"loss": 0.
|
| 1578 |
"step": 1110
|
| 1579 |
},
|
| 1580 |
{
|
| 1581 |
"epoch": 1.5051209904333147,
|
| 1582 |
-
"grad_norm": 0.
|
| 1583 |
"learning_rate": 1.7411175646664103e-06,
|
| 1584 |
-
"loss": 0.
|
| 1585 |
"step": 1115
|
| 1586 |
},
|
| 1587 |
{
|
| 1588 |
"epoch": 1.5118739448508722,
|
| 1589 |
-
"grad_norm": 0.
|
| 1590 |
"learning_rate": 1.6966263830495939e-06,
|
| 1591 |
-
"loss": 0.
|
| 1592 |
"step": 1120
|
| 1593 |
},
|
| 1594 |
{
|
| 1595 |
"epoch": 1.51862689926843,
|
| 1596 |
-
"grad_norm": 0.
|
| 1597 |
"learning_rate": 1.6525945940031407e-06,
|
| 1598 |
"loss": 0.3692,
|
| 1599 |
"step": 1125
|
| 1600 |
},
|
| 1601 |
{
|
| 1602 |
"epoch": 1.5253798536859877,
|
| 1603 |
-
"grad_norm": 0.
|
| 1604 |
"learning_rate": 1.6090283209266682e-06,
|
| 1605 |
-
"loss": 0.
|
| 1606 |
"step": 1130
|
| 1607 |
},
|
| 1608 |
{
|
| 1609 |
"epoch": 1.5321328081035452,
|
| 1610 |
-
"grad_norm": 0.
|
| 1611 |
"learning_rate": 1.5659336224815642e-06,
|
| 1612 |
-
"loss": 0.
|
| 1613 |
"step": 1135
|
| 1614 |
},
|
| 1615 |
{
|
| 1616 |
"epoch": 1.538885762521103,
|
| 1617 |
-
"grad_norm": 0.
|
| 1618 |
"learning_rate": 1.5233164917484117e-06,
|
| 1619 |
-
"loss": 0.
|
| 1620 |
"step": 1140
|
| 1621 |
},
|
| 1622 |
{
|
| 1623 |
"epoch": 1.5456387169386607,
|
| 1624 |
-
"grad_norm": 0.
|
| 1625 |
"learning_rate": 1.4811828553935498e-06,
|
| 1626 |
-
"loss": 0.
|
| 1627 |
"step": 1145
|
| 1628 |
},
|
| 1629 |
{
|
| 1630 |
"epoch": 1.5523916713562183,
|
| 1631 |
-
"grad_norm": 0.
|
| 1632 |
"learning_rate": 1.439538572844873e-06,
|
| 1633 |
-
"loss": 0.
|
| 1634 |
"step": 1150
|
| 1635 |
},
|
| 1636 |
{
|
| 1637 |
"epoch": 1.559144625773776,
|
| 1638 |
-
"grad_norm": 0.
|
| 1639 |
"learning_rate": 1.3983894354769616e-06,
|
| 1640 |
-
"loss": 0.
|
| 1641 |
"step": 1155
|
| 1642 |
},
|
| 1643 |
{
|
| 1644 |
"epoch": 1.5658975801913337,
|
| 1645 |
-
"grad_norm": 0.
|
| 1646 |
"learning_rate": 1.3577411658056965e-06,
|
| 1647 |
-
"loss": 0.
|
| 1648 |
"step": 1160
|
| 1649 |
},
|
| 1650 |
{
|
| 1651 |
"epoch": 1.5726505346088913,
|
| 1652 |
-
"grad_norm": 0.
|
| 1653 |
"learning_rate": 1.3175994166924394e-06,
|
| 1654 |
-
"loss": 0.
|
| 1655 |
"step": 1165
|
| 1656 |
},
|
| 1657 |
{
|
| 1658 |
"epoch": 1.5794034890264492,
|
| 1659 |
-
"grad_norm": 0.
|
| 1660 |
"learning_rate": 1.2779697705579058e-06,
|
| 1661 |
-
"loss": 0.
|
| 1662 |
"step": 1170
|
| 1663 |
},
|
| 1664 |
{
|
| 1665 |
"epoch": 1.5861564434440067,
|
| 1666 |
-
"grad_norm": 0.
|
| 1667 |
"learning_rate": 1.2388577386058248e-06,
|
| 1668 |
-
"loss": 0.
|
| 1669 |
"step": 1175
|
| 1670 |
},
|
| 1671 |
{
|
| 1672 |
"epoch": 1.5929093978615643,
|
| 1673 |
-
"grad_norm": 0.
|
| 1674 |
"learning_rate": 1.2002687600565138e-06,
|
| 1675 |
-
"loss": 0.
|
| 1676 |
"step": 1180
|
| 1677 |
},
|
| 1678 |
{
|
| 1679 |
"epoch": 1.5996623522791222,
|
| 1680 |
-
"grad_norm": 0.
|
| 1681 |
"learning_rate": 1.1622082013904535e-06,
|
| 1682 |
-
"loss": 0.
|
| 1683 |
"step": 1185
|
| 1684 |
},
|
| 1685 |
{
|
| 1686 |
"epoch": 1.6064153066966798,
|
| 1687 |
-
"grad_norm": 0.
|
| 1688 |
"learning_rate": 1.1246813556019925e-06,
|
| 1689 |
-
"loss": 0.
|
| 1690 |
"step": 1190
|
| 1691 |
},
|
| 1692 |
{
|
| 1693 |
"epoch": 1.6131682611142375,
|
| 1694 |
-
"grad_norm": 0.
|
| 1695 |
"learning_rate": 1.0876934414632523e-06,
|
| 1696 |
-
"loss": 0.
|
| 1697 |
"step": 1195
|
| 1698 |
},
|
| 1699 |
{
|
| 1700 |
"epoch": 1.6199212155317952,
|
| 1701 |
-
"grad_norm": 0.
|
| 1702 |
"learning_rate": 1.0512496027983715e-06,
|
| 1703 |
-
"loss": 0.
|
| 1704 |
"step": 1200
|
| 1705 |
},
|
| 1706 |
{
|
| 1707 |
"epoch": 1.6266741699493528,
|
| 1708 |
-
"grad_norm": 0.
|
| 1709 |
"learning_rate": 1.0153549077681617e-06,
|
| 1710 |
-
"loss": 0.
|
| 1711 |
"step": 1205
|
| 1712 |
},
|
| 1713 |
{
|
| 1714 |
"epoch": 1.6334271243669105,
|
| 1715 |
-
"grad_norm": 0.
|
| 1716 |
"learning_rate": 9.80014348165298e-07,
|
| 1717 |
-
"loss": 0.
|
| 1718 |
"step": 1210
|
| 1719 |
},
|
| 1720 |
{
|
| 1721 |
"epoch": 1.6401800787844683,
|
| 1722 |
-
"grad_norm": 0.
|
| 1723 |
"learning_rate": 9.452328387201104e-07,
|
| 1724 |
-
"loss": 0.
|
| 1725 |
"step": 1215
|
| 1726 |
},
|
| 1727 |
{
|
| 1728 |
"epoch": 1.6469330332020258,
|
| 1729 |
-
"grad_norm": 0.
|
| 1730 |
"learning_rate": 9.110152164171127e-07,
|
| 1731 |
-
"loss": 0.
|
| 1732 |
"step": 1220
|
| 1733 |
},
|
| 1734 |
{
|
| 1735 |
"epoch": 1.6536859876195835,
|
| 1736 |
-
"grad_norm": 0.
|
| 1737 |
"learning_rate": 8.773662398223276e-07,
|
| 1738 |
-
"loss": 0.
|
| 1739 |
"step": 1225
|
| 1740 |
},
|
| 1741 |
{
|
| 1742 |
"epoch": 1.6604389420371413,
|
| 1743 |
-
"grad_norm": 0.
|
| 1744 |
"learning_rate": 8.44290588421533e-07,
|
| 1745 |
-
"loss": 0.
|
| 1746 |
"step": 1230
|
| 1747 |
},
|
| 1748 |
{
|
| 1749 |
"epoch": 1.6671918964546988,
|
| 1750 |
-
"grad_norm": 0.
|
| 1751 |
"learning_rate": 8.117928619694848e-07,
|
| 1752 |
-
"loss": 0.
|
| 1753 |
"step": 1235
|
| 1754 |
},
|
| 1755 |
{
|
| 1756 |
"epoch": 1.6739448508722568,
|
| 1757 |
-
"grad_norm": 0.
|
| 1758 |
"learning_rate": 7.798775798502484e-07,
|
| 1759 |
-
"loss": 0.
|
| 1760 |
"step": 1240
|
| 1761 |
},
|
| 1762 |
{
|
| 1763 |
"epoch": 1.6806978052898143,
|
| 1764 |
-
"grad_norm": 0.
|
| 1765 |
"learning_rate": 7.485491804486972e-07,
|
| 1766 |
"loss": 0.2951,
|
| 1767 |
"step": 1245
|
| 1768 |
},
|
| 1769 |
{
|
| 1770 |
"epoch": 1.687450759707372,
|
| 1771 |
-
"grad_norm": 0.
|
| 1772 |
"learning_rate": 7.178120205332717e-07,
|
| 1773 |
-
"loss": 0.
|
| 1774 |
"step": 1250
|
| 1775 |
},
|
| 1776 |
{
|
| 1777 |
"epoch": 1.6942037141249298,
|
| 1778 |
-
"grad_norm": 0.
|
| 1779 |
"learning_rate": 6.876703746500984e-07,
|
| 1780 |
-
"loss": 0.
|
| 1781 |
"step": 1255
|
| 1782 |
},
|
| 1783 |
{
|
| 1784 |
"epoch": 1.7009566685424873,
|
| 1785 |
-
"grad_norm": 0.
|
| 1786 |
"learning_rate": 6.581284345285371e-07,
|
| 1787 |
-
"loss": 0.
|
| 1788 |
"step": 1260
|
| 1789 |
},
|
| 1790 |
{
|
| 1791 |
"epoch": 1.707709622960045,
|
| 1792 |
-
"grad_norm": 0.
|
| 1793 |
"learning_rate": 6.291903084982481e-07,
|
| 1794 |
-
"loss": 0.
|
| 1795 |
"step": 1265
|
| 1796 |
},
|
| 1797 |
{
|
| 1798 |
"epoch": 1.7144625773776028,
|
| 1799 |
-
"grad_norm": 0.
|
| 1800 |
"learning_rate": 6.008600209178539e-07,
|
| 1801 |
-
"loss": 0.
|
| 1802 |
"step": 1270
|
| 1803 |
},
|
| 1804 |
{
|
| 1805 |
"epoch": 1.7212155317951603,
|
| 1806 |
-
"grad_norm": 0.
|
| 1807 |
"learning_rate": 5.7314151161528e-07,
|
| 1808 |
-
"loss": 0.
|
| 1809 |
"step": 1275
|
| 1810 |
},
|
| 1811 |
{
|
| 1812 |
"epoch": 1.727968486212718,
|
| 1813 |
-
"grad_norm": 0.
|
| 1814 |
"learning_rate": 5.460386353398583e-07,
|
| 1815 |
-
"loss": 0.
|
| 1816 |
"step": 1280
|
| 1817 |
},
|
| 1818 |
{
|
| 1819 |
"epoch": 1.7347214406302758,
|
| 1820 |
-
"grad_norm": 0.
|
| 1821 |
"learning_rate": 5.195551612262478e-07,
|
| 1822 |
-
"loss": 0.
|
| 1823 |
"step": 1285
|
| 1824 |
},
|
| 1825 |
{
|
| 1826 |
"epoch": 1.7414743950478333,
|
| 1827 |
-
"grad_norm": 0.
|
| 1828 |
"learning_rate": 4.936947722702762e-07,
|
| 1829 |
-
"loss": 0.
|
| 1830 |
"step": 1290
|
| 1831 |
},
|
| 1832 |
{
|
| 1833 |
"epoch": 1.7482273494653913,
|
| 1834 |
-
"grad_norm": 0.
|
| 1835 |
"learning_rate": 4.6846106481675035e-07,
|
| 1836 |
-
"loss": 0.
|
| 1837 |
"step": 1295
|
| 1838 |
},
|
| 1839 |
{
|
| 1840 |
"epoch": 1.7549803038829488,
|
| 1841 |
-
"grad_norm": 0.
|
| 1842 |
"learning_rate": 4.43857548059321e-07,
|
| 1843 |
-
"loss": 0.
|
| 1844 |
"step": 1300
|
| 1845 |
},
|
| 1846 |
{
|
| 1847 |
"epoch": 1.7617332583005063,
|
| 1848 |
-
"grad_norm": 0.
|
| 1849 |
"learning_rate": 4.198876435524718e-07,
|
| 1850 |
-
"loss": 0.
|
| 1851 |
"step": 1305
|
| 1852 |
},
|
| 1853 |
{
|
| 1854 |
"epoch": 1.7684862127180643,
|
| 1855 |
-
"grad_norm": 0.
|
| 1856 |
"learning_rate": 3.9655468473568436e-07,
|
| 1857 |
-
"loss": 0.
|
| 1858 |
"step": 1310
|
| 1859 |
},
|
| 1860 |
{
|
| 1861 |
"epoch": 1.7752391671356218,
|
| 1862 |
-
"grad_norm": 0.
|
| 1863 |
"learning_rate": 3.7386191646987094e-07,
|
| 1864 |
-
"loss": 0.
|
| 1865 |
"step": 1315
|
| 1866 |
},
|
| 1867 |
{
|
| 1868 |
"epoch": 1.7819921215531795,
|
| 1869 |
-
"grad_norm": 0.
|
| 1870 |
"learning_rate": 3.51812494586114e-07,
|
| 1871 |
-
"loss": 0.
|
| 1872 |
"step": 1320
|
| 1873 |
},
|
| 1874 |
{
|
| 1875 |
"epoch": 1.7887450759707373,
|
| 1876 |
-
"grad_norm": 0.
|
| 1877 |
"learning_rate": 3.3040948544679817e-07,
|
| 1878 |
-
"loss": 0.
|
| 1879 |
"step": 1325
|
| 1880 |
},
|
| 1881 |
{
|
| 1882 |
"epoch": 1.7954980303882948,
|
| 1883 |
-
"grad_norm": 0.
|
| 1884 |
"learning_rate": 3.096558655191706e-07,
|
| 1885 |
-
"loss": 0.
|
| 1886 |
"step": 1330
|
| 1887 |
},
|
| 1888 |
{
|
| 1889 |
"epoch": 1.8022509848058526,
|
| 1890 |
-
"grad_norm": 0.
|
| 1891 |
"learning_rate": 2.895545209614176e-07,
|
| 1892 |
-
"loss": 0.
|
| 1893 |
"step": 1335
|
| 1894 |
},
|
| 1895 |
{
|
| 1896 |
"epoch": 1.8090039392234103,
|
| 1897 |
-
"grad_norm": 0.
|
| 1898 |
"learning_rate": 2.701082472212879e-07,
|
| 1899 |
-
"loss": 0.
|
| 1900 |
"step": 1340
|
| 1901 |
},
|
| 1902 |
{
|
| 1903 |
"epoch": 1.8157568936409678,
|
| 1904 |
-
"grad_norm": 0.
|
| 1905 |
"learning_rate": 2.5131974864734063e-07,
|
| 1906 |
"loss": 0.3425,
|
| 1907 |
"step": 1345
|
| 1908 |
},
|
| 1909 |
{
|
| 1910 |
"epoch": 1.8225098480585256,
|
| 1911 |
-
"grad_norm": 0.
|
| 1912 |
"learning_rate": 2.331916381128535e-07,
|
| 1913 |
-
"loss": 0.
|
| 1914 |
"step": 1350
|
| 1915 |
},
|
| 1916 |
{
|
| 1917 |
"epoch": 1.8292628024760833,
|
| 1918 |
-
"grad_norm": 0.
|
| 1919 |
"learning_rate": 2.157264366524603e-07,
|
| 1920 |
-
"loss": 0.
|
| 1921 |
"step": 1355
|
| 1922 |
},
|
| 1923 |
{
|
| 1924 |
"epoch": 1.8360157568936408,
|
| 1925 |
-
"grad_norm": 0.
|
| 1926 |
"learning_rate": 1.989265731115525e-07,
|
| 1927 |
-
"loss": 0.
|
| 1928 |
"step": 1360
|
| 1929 |
},
|
| 1930 |
{
|
| 1931 |
"epoch": 1.8427687113111988,
|
| 1932 |
-
"grad_norm": 0.
|
| 1933 |
"learning_rate": 1.827943838085111e-07,
|
| 1934 |
"loss": 0.306,
|
| 1935 |
"step": 1365
|
| 1936 |
},
|
| 1937 |
{
|
| 1938 |
"epoch": 1.8495216657287563,
|
| 1939 |
-
"grad_norm": 0.
|
| 1940 |
"learning_rate": 1.6733211220979316e-07,
|
| 1941 |
-
"loss": 0.
|
| 1942 |
"step": 1370
|
| 1943 |
},
|
| 1944 |
{
|
| 1945 |
"epoch": 1.856274620146314,
|
| 1946 |
-
"grad_norm": 0.
|
| 1947 |
"learning_rate": 1.5254190861794415e-07,
|
| 1948 |
-
"loss": 0.
|
| 1949 |
"step": 1375
|
| 1950 |
},
|
| 1951 |
{
|
| 1952 |
"epoch": 1.8630275745638718,
|
| 1953 |
-
"grad_norm": 0.
|
| 1954 |
"learning_rate": 1.3842582987255494e-07,
|
| 1955 |
-
"loss": 0.
|
| 1956 |
"step": 1380
|
| 1957 |
},
|
| 1958 |
{
|
| 1959 |
"epoch": 1.8697805289814293,
|
| 1960 |
-
"grad_norm": 0.
|
| 1961 |
"learning_rate": 1.2498583906422779e-07,
|
| 1962 |
-
"loss": 0.
|
| 1963 |
"step": 1385
|
| 1964 |
},
|
| 1965 |
{
|
| 1966 |
"epoch": 1.876533483398987,
|
| 1967 |
-
"grad_norm": 0.
|
| 1968 |
"learning_rate": 1.1222380526156929e-07,
|
| 1969 |
-
"loss": 0.
|
| 1970 |
"step": 1390
|
| 1971 |
},
|
| 1972 |
{
|
| 1973 |
"epoch": 1.8832864378165448,
|
| 1974 |
-
"grad_norm": 0.
|
| 1975 |
"learning_rate": 1.0014150325126315e-07,
|
| 1976 |
-
"loss": 0.
|
| 1977 |
"step": 1395
|
| 1978 |
},
|
| 1979 |
{
|
| 1980 |
"epoch": 1.8900393922341023,
|
| 1981 |
-
"grad_norm": 0.
|
| 1982 |
"learning_rate": 8.874061329125939e-08,
|
| 1983 |
-
"loss": 0.
|
| 1984 |
"step": 1400
|
| 1985 |
},
|
| 1986 |
{
|
| 1987 |
"epoch": 1.89679234665166,
|
| 1988 |
-
"grad_norm": 0.
|
| 1989 |
"learning_rate": 7.802272087709951e-08,
|
| 1990 |
-
"loss": 0.
|
| 1991 |
"step": 1405
|
| 1992 |
},
|
| 1993 |
{
|
| 1994 |
"epoch": 1.9035453010692178,
|
| 1995 |
-
"grad_norm": 0.
|
| 1996 |
"learning_rate": 6.798931652142738e-08,
|
| 1997 |
-
"loss": 0.
|
| 1998 |
"step": 1410
|
| 1999 |
},
|
| 2000 |
{
|
| 2001 |
"epoch": 1.9102982554867753,
|
| 2002 |
-
"grad_norm": 0.
|
| 2003 |
"learning_rate": 5.864179554670724e-08,
|
| 2004 |
-
"loss": 0.
|
| 2005 |
"step": 1415
|
| 2006 |
},
|
| 2007 |
{
|
| 2008 |
"epoch": 1.9170512099043333,
|
| 2009 |
-
"grad_norm": 0.
|
| 2010 |
"learning_rate": 4.998145789118114e-08,
|
| 2011 |
-
"loss": 0.
|
| 2012 |
"step": 1420
|
| 2013 |
},
|
| 2014 |
{
|
| 2015 |
"epoch": 1.9238041643218908,
|
| 2016 |
-
"grad_norm": 0.
|
| 2017 |
"learning_rate": 4.2009507928084116e-08,
|
| 2018 |
-
"loss": 0.
|
| 2019 |
"step": 1425
|
| 2020 |
},
|
| 2021 |
{
|
| 2022 |
"epoch": 1.9305571187394484,
|
| 2023 |
-
"grad_norm": 0.
|
| 2024 |
"learning_rate": 3.4727054298161475e-08,
|
| 2025 |
-
"loss": 0.
|
| 2026 |
"step": 1430
|
| 2027 |
},
|
| 2028 |
{
|
| 2029 |
"epoch": 1.9373100731570063,
|
| 2030 |
-
"grad_norm": 0.
|
| 2031 |
"learning_rate": 2.8135109755487723e-08,
|
| 2032 |
-
"loss": 0.
|
| 2033 |
"step": 1435
|
| 2034 |
},
|
| 2035 |
{
|
| 2036 |
"epoch": 1.9440630275745638,
|
| 2037 |
-
"grad_norm": 0.
|
| 2038 |
"learning_rate": 2.223459102662695e-08,
|
| 2039 |
-
"loss": 0.
|
| 2040 |
"step": 1440
|
| 2041 |
},
|
| 2042 |
{
|
| 2043 |
"epoch": 1.9508159819921216,
|
| 2044 |
-
"grad_norm": 0.
|
| 2045 |
"learning_rate": 1.7026318683147082e-08,
|
| 2046 |
-
"loss": 0.
|
| 2047 |
"step": 1445
|
| 2048 |
},
|
| 2049 |
{
|
| 2050 |
"epoch": 1.9575689364096793,
|
| 2051 |
-
"grad_norm": 0.
|
| 2052 |
"learning_rate": 1.2511017027501682e-08,
|
| 2053 |
"loss": 0.3032,
|
| 2054 |
"step": 1450
|
| 2055 |
},
|
| 2056 |
{
|
| 2057 |
"epoch": 1.9643218908272368,
|
| 2058 |
-
"grad_norm": 0.
|
| 2059 |
"learning_rate": 8.689313992306104e-09,
|
| 2060 |
-
"loss": 0.
|
| 2061 |
"step": 1455
|
| 2062 |
},
|
| 2063 |
{
|
| 2064 |
"epoch": 1.9710748452447946,
|
| 2065 |
-
"grad_norm": 0.
|
| 2066 |
"learning_rate": 5.561741053010661e-09,
|
| 2067 |
-
"loss": 0.
|
| 2068 |
"step": 1460
|
| 2069 |
},
|
| 2070 |
{
|
| 2071 |
"epoch": 1.9778277996623523,
|
| 2072 |
-
"grad_norm": 0.
|
| 2073 |
"learning_rate": 3.1287331539903155e-09,
|
| 2074 |
-
"loss": 0.
|
| 2075 |
"step": 1465
|
| 2076 |
},
|
| 2077 |
{
|
| 2078 |
"epoch": 1.9845807540799099,
|
| 2079 |
-
"grad_norm": 0.
|
| 2080 |
"learning_rate": 1.3906286480563913e-09,
|
| 2081 |
-
"loss": 0.
|
| 2082 |
"step": 1470
|
| 2083 |
},
|
| 2084 |
{
|
| 2085 |
"epoch": 1.9913337084974676,
|
| 2086 |
-
"grad_norm": 0.
|
| 2087 |
"learning_rate": 3.4766924940476954e-10,
|
| 2088 |
-
"loss": 0.
|
| 2089 |
"step": 1475
|
| 2090 |
},
|
| 2091 |
{
|
| 2092 |
"epoch": 1.9980866629150253,
|
| 2093 |
-
"grad_norm": 0.
|
| 2094 |
"learning_rate": 0.0,
|
| 2095 |
-
"loss": 0.
|
| 2096 |
"step": 1480
|
| 2097 |
},
|
| 2098 |
{
|
| 2099 |
"epoch": 1.9980866629150253,
|
| 2100 |
-
"eval_loss": 0.
|
| 2101 |
-
"eval_runtime": 17.
|
| 2102 |
-
"eval_samples_per_second": 20.
|
| 2103 |
-
"eval_steps_per_second": 5.
|
| 2104 |
"step": 1480
|
| 2105 |
},
|
| 2106 |
{
|
| 2107 |
"epoch": 1.9980866629150253,
|
| 2108 |
"step": 1480,
|
| 2109 |
"total_flos": 3.6719988787579453e+18,
|
| 2110 |
-
"train_loss": 0.
|
| 2111 |
-
"train_runtime":
|
| 2112 |
-
"train_samples_per_second": 5.
|
| 2113 |
"train_steps_per_second": 0.112
|
| 2114 |
}
|
| 2115 |
],
|
|
|
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
"epoch": 0.0013505908835115363,
|
| 13 |
+
"grad_norm": 2.148791476511835,
|
| 14 |
"learning_rate": 6.756756756756757e-08,
|
| 15 |
"loss": 2.1725,
|
| 16 |
"step": 1
|
| 17 |
},
|
| 18 |
{
|
| 19 |
"epoch": 0.006752954417557681,
|
| 20 |
+
"grad_norm": 2.0693435263055955,
|
| 21 |
"learning_rate": 3.378378378378379e-07,
|
| 22 |
+
"loss": 2.1697,
|
| 23 |
"step": 5
|
| 24 |
},
|
| 25 |
{
|
| 26 |
"epoch": 0.013505908835115363,
|
| 27 |
+
"grad_norm": 1.698535954485606,
|
| 28 |
"learning_rate": 6.756756756756758e-07,
|
| 29 |
+
"loss": 2.0659,
|
| 30 |
"step": 10
|
| 31 |
},
|
| 32 |
{
|
| 33 |
"epoch": 0.020258863252673044,
|
| 34 |
+
"grad_norm": 0.9452543971000898,
|
| 35 |
"learning_rate": 1.0135135135135136e-06,
|
| 36 |
+
"loss": 1.8928,
|
| 37 |
"step": 15
|
| 38 |
},
|
| 39 |
{
|
| 40 |
"epoch": 0.027011817670230726,
|
| 41 |
+
"grad_norm": 0.8038988537572226,
|
| 42 |
"learning_rate": 1.3513513513513515e-06,
|
| 43 |
+
"loss": 1.5709,
|
| 44 |
"step": 20
|
| 45 |
},
|
| 46 |
{
|
| 47 |
"epoch": 0.03376477208778841,
|
| 48 |
+
"grad_norm": 0.471508019359059,
|
| 49 |
"learning_rate": 1.6891891891891894e-06,
|
| 50 |
+
"loss": 1.1256,
|
| 51 |
"step": 25
|
| 52 |
},
|
| 53 |
{
|
| 54 |
"epoch": 0.04051772650534609,
|
| 55 |
+
"grad_norm": 0.252352654342694,
|
| 56 |
"learning_rate": 2.0270270270270273e-06,
|
| 57 |
+
"loss": 0.9097,
|
| 58 |
"step": 30
|
| 59 |
},
|
| 60 |
{
|
| 61 |
"epoch": 0.04727068092290377,
|
| 62 |
+
"grad_norm": 0.24974204597161465,
|
| 63 |
"learning_rate": 2.364864864864865e-06,
|
| 64 |
+
"loss": 0.7799,
|
| 65 |
"step": 35
|
| 66 |
},
|
| 67 |
{
|
| 68 |
"epoch": 0.05402363534046145,
|
| 69 |
+
"grad_norm": 0.20795816160607852,
|
| 70 |
"learning_rate": 2.702702702702703e-06,
|
| 71 |
+
"loss": 0.7137,
|
| 72 |
"step": 40
|
| 73 |
},
|
| 74 |
{
|
| 75 |
"epoch": 0.06077658975801913,
|
| 76 |
+
"grad_norm": 0.17844205306863006,
|
| 77 |
"learning_rate": 3.040540540540541e-06,
|
| 78 |
+
"loss": 0.5982,
|
| 79 |
"step": 45
|
| 80 |
},
|
| 81 |
{
|
| 82 |
"epoch": 0.06752954417557681,
|
| 83 |
+
"grad_norm": 0.14449057063114276,
|
| 84 |
"learning_rate": 3.3783783783783788e-06,
|
| 85 |
+
"loss": 0.5388,
|
| 86 |
"step": 50
|
| 87 |
},
|
| 88 |
{
|
| 89 |
"epoch": 0.0742824985931345,
|
| 90 |
+
"grad_norm": 0.12180944441742095,
|
| 91 |
"learning_rate": 3.7162162162162162e-06,
|
| 92 |
"loss": 0.5621,
|
| 93 |
"step": 55
|
| 94 |
},
|
| 95 |
{
|
| 96 |
"epoch": 0.08103545301069218,
|
| 97 |
+
"grad_norm": 0.11672624417203623,
|
| 98 |
"learning_rate": 4.0540540540540545e-06,
|
| 99 |
+
"loss": 0.4885,
|
| 100 |
"step": 60
|
| 101 |
},
|
| 102 |
{
|
| 103 |
"epoch": 0.08778840742824986,
|
| 104 |
+
"grad_norm": 0.12365779514235553,
|
| 105 |
"learning_rate": 4.391891891891892e-06,
|
| 106 |
+
"loss": 0.4725,
|
| 107 |
"step": 65
|
| 108 |
},
|
| 109 |
{
|
| 110 |
"epoch": 0.09454136184580754,
|
| 111 |
+
"grad_norm": 0.13848942929552646,
|
| 112 |
"learning_rate": 4.72972972972973e-06,
|
| 113 |
+
"loss": 0.4737,
|
| 114 |
"step": 70
|
| 115 |
},
|
| 116 |
{
|
| 117 |
"epoch": 0.10129431626336523,
|
| 118 |
+
"grad_norm": 0.13947287893715526,
|
| 119 |
"learning_rate": 5.067567567567568e-06,
|
| 120 |
+
"loss": 0.4686,
|
| 121 |
"step": 75
|
| 122 |
},
|
| 123 |
{
|
| 124 |
"epoch": 0.1080472706809229,
|
| 125 |
+
"grad_norm": 0.10498499488363304,
|
| 126 |
"learning_rate": 5.405405405405406e-06,
|
| 127 |
+
"loss": 0.4607,
|
| 128 |
"step": 80
|
| 129 |
},
|
| 130 |
{
|
| 131 |
"epoch": 0.11480022509848059,
|
| 132 |
+
"grad_norm": 0.1263112338839764,
|
| 133 |
"learning_rate": 5.743243243243244e-06,
|
| 134 |
"loss": 0.471,
|
| 135 |
"step": 85
|
| 136 |
},
|
| 137 |
{
|
| 138 |
"epoch": 0.12155317951603826,
|
| 139 |
+
"grad_norm": 0.1082136353357591,
|
| 140 |
"learning_rate": 6.081081081081082e-06,
|
| 141 |
+
"loss": 0.432,
|
| 142 |
"step": 90
|
| 143 |
},
|
| 144 |
{
|
| 145 |
"epoch": 0.12830613393359594,
|
| 146 |
+
"grad_norm": 0.08923160665434038,
|
| 147 |
"learning_rate": 6.41891891891892e-06,
|
| 148 |
+
"loss": 0.4433,
|
| 149 |
"step": 95
|
| 150 |
},
|
| 151 |
{
|
| 152 |
"epoch": 0.13505908835115363,
|
| 153 |
+
"grad_norm": 0.11807106476588053,
|
| 154 |
"learning_rate": 6.7567567567567575e-06,
|
| 155 |
"loss": 0.4398,
|
| 156 |
"step": 100
|
| 157 |
},
|
| 158 |
{
|
| 159 |
"epoch": 0.14181204276871132,
|
| 160 |
+
"grad_norm": 0.09395401671145927,
|
| 161 |
"learning_rate": 7.0945945945945946e-06,
|
| 162 |
"loss": 0.3899,
|
| 163 |
"step": 105
|
| 164 |
},
|
| 165 |
{
|
| 166 |
"epoch": 0.148564997186269,
|
| 167 |
+
"grad_norm": 0.09428015614908754,
|
| 168 |
"learning_rate": 7.4324324324324324e-06,
|
| 169 |
+
"loss": 0.3983,
|
| 170 |
"step": 110
|
| 171 |
},
|
| 172 |
{
|
| 173 |
"epoch": 0.15531795160382666,
|
| 174 |
+
"grad_norm": 0.08605902096557952,
|
| 175 |
"learning_rate": 7.77027027027027e-06,
|
| 176 |
+
"loss": 0.415,
|
| 177 |
"step": 115
|
| 178 |
},
|
| 179 |
{
|
| 180 |
"epoch": 0.16207090602138435,
|
| 181 |
+
"grad_norm": 0.08713323888522238,
|
| 182 |
"learning_rate": 8.108108108108109e-06,
|
| 183 |
+
"loss": 0.4075,
|
| 184 |
"step": 120
|
| 185 |
},
|
| 186 |
{
|
| 187 |
"epoch": 0.16882386043894204,
|
| 188 |
+
"grad_norm": 0.07858143188282206,
|
| 189 |
"learning_rate": 8.445945945945948e-06,
|
| 190 |
+
"loss": 0.4089,
|
| 191 |
"step": 125
|
| 192 |
},
|
| 193 |
{
|
| 194 |
"epoch": 0.17557681485649973,
|
| 195 |
+
"grad_norm": 0.07427563021554226,
|
| 196 |
"learning_rate": 8.783783783783785e-06,
|
| 197 |
+
"loss": 0.4291,
|
| 198 |
"step": 130
|
| 199 |
},
|
| 200 |
{
|
| 201 |
"epoch": 0.1823297692740574,
|
| 202 |
+
"grad_norm": 0.09144835191366023,
|
| 203 |
"learning_rate": 9.121621621621622e-06,
|
| 204 |
"loss": 0.3895,
|
| 205 |
"step": 135
|
| 206 |
},
|
| 207 |
{
|
| 208 |
"epoch": 0.18908272369161508,
|
| 209 |
+
"grad_norm": 0.07052586919720526,
|
| 210 |
"learning_rate": 9.45945945945946e-06,
|
| 211 |
"loss": 0.4029,
|
| 212 |
"step": 140
|
| 213 |
},
|
| 214 |
{
|
| 215 |
"epoch": 0.19583567810917277,
|
| 216 |
+
"grad_norm": 0.0769561636008079,
|
| 217 |
"learning_rate": 9.797297297297298e-06,
|
| 218 |
+
"loss": 0.4148,
|
| 219 |
"step": 145
|
| 220 |
},
|
| 221 |
{
|
| 222 |
"epoch": 0.20258863252673046,
|
| 223 |
+
"grad_norm": 0.07216832297932871,
|
| 224 |
"learning_rate": 9.999944372378571e-06,
|
| 225 |
+
"loss": 0.3801,
|
| 226 |
"step": 150
|
| 227 |
},
|
| 228 |
{
|
| 229 |
"epoch": 0.20934158694428812,
|
| 230 |
+
"grad_norm": 0.06883498336890326,
|
| 231 |
"learning_rate": 9.999318575852451e-06,
|
| 232 |
+
"loss": 0.3728,
|
| 233 |
"step": 155
|
| 234 |
},
|
| 235 |
{
|
| 236 |
"epoch": 0.2160945413618458,
|
| 237 |
+
"grad_norm": 0.07007134671747867,
|
| 238 |
"learning_rate": 9.99799753559161e-06,
|
| 239 |
+
"loss": 0.3831,
|
| 240 |
"step": 160
|
| 241 |
},
|
| 242 |
{
|
| 243 |
"epoch": 0.2228474957794035,
|
| 244 |
+
"grad_norm": 0.0638944841323381,
|
| 245 |
"learning_rate": 9.995981435310078e-06,
|
| 246 |
+
"loss": 0.3852,
|
| 247 |
"step": 165
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 0.22960045019696118,
|
| 251 |
+
"grad_norm": 0.0732472641559478,
|
| 252 |
"learning_rate": 9.993270555382283e-06,
|
| 253 |
"loss": 0.4058,
|
| 254 |
"step": 170
|
| 255 |
},
|
| 256 |
{
|
| 257 |
"epoch": 0.23635340461451884,
|
| 258 |
+
"grad_norm": 0.06688107743191332,
|
| 259 |
"learning_rate": 9.989865272804064e-06,
|
| 260 |
+
"loss": 0.3981,
|
| 261 |
"step": 175
|
| 262 |
},
|
| 263 |
{
|
| 264 |
"epoch": 0.24310635903207653,
|
| 265 |
+
"grad_norm": 0.0750542374720311,
|
| 266 |
"learning_rate": 9.985766061140233e-06,
|
| 267 |
+
"loss": 0.4034,
|
| 268 |
"step": 180
|
| 269 |
},
|
| 270 |
{
|
| 271 |
"epoch": 0.24985931344963422,
|
| 272 |
+
"grad_norm": 0.06261496896671846,
|
| 273 |
"learning_rate": 9.980973490458728e-06,
|
| 274 |
+
"loss": 0.3611,
|
| 275 |
"step": 185
|
| 276 |
},
|
| 277 |
{
|
| 278 |
"epoch": 0.2566122678671919,
|
| 279 |
+
"grad_norm": 0.07059993646852766,
|
| 280 |
"learning_rate": 9.97548822725133e-06,
|
| 281 |
+
"loss": 0.3603,
|
| 282 |
"step": 190
|
| 283 |
},
|
| 284 |
{
|
| 285 |
"epoch": 0.26336522228474957,
|
| 286 |
+
"grad_norm": 0.06560938196431165,
|
| 287 |
"learning_rate": 9.969311034340977e-06,
|
| 288 |
+
"loss": 0.3853,
|
| 289 |
"step": 195
|
| 290 |
},
|
| 291 |
{
|
| 292 |
"epoch": 0.27011817670230726,
|
| 293 |
+
"grad_norm": 0.05548643801780881,
|
| 294 |
"learning_rate": 9.962442770775675e-06,
|
| 295 |
+
"loss": 0.3659,
|
| 296 |
"step": 200
|
| 297 |
},
|
| 298 |
{
|
| 299 |
"epoch": 0.27687113111986494,
|
| 300 |
+
"grad_norm": 0.06003482909108544,
|
| 301 |
"learning_rate": 9.954884391709043e-06,
|
| 302 |
+
"loss": 0.3907,
|
| 303 |
"step": 205
|
| 304 |
},
|
| 305 |
{
|
| 306 |
"epoch": 0.28362408553742263,
|
| 307 |
+
"grad_norm": 0.04978856362198666,
|
| 308 |
"learning_rate": 9.946636948267468e-06,
|
| 309 |
"loss": 0.349,
|
| 310 |
"step": 210
|
| 311 |
},
|
| 312 |
{
|
| 313 |
"epoch": 0.2903770399549803,
|
| 314 |
+
"grad_norm": 0.056553641270721766,
|
| 315 |
"learning_rate": 9.937701587403941e-06,
|
| 316 |
+
"loss": 0.4041,
|
| 317 |
"step": 215
|
| 318 |
},
|
| 319 |
{
|
| 320 |
"epoch": 0.297129994372538,
|
| 321 |
+
"grad_norm": 0.06048923977199324,
|
| 322 |
"learning_rate": 9.928079551738542e-06,
|
| 323 |
+
"loss": 0.3704,
|
| 324 |
"step": 220
|
| 325 |
},
|
| 326 |
{
|
| 327 |
"epoch": 0.30388294879009564,
|
| 328 |
+
"grad_norm": 0.055403005803656744,
|
| 329 |
"learning_rate": 9.91777217938564e-06,
|
| 330 |
+
"loss": 0.3765,
|
| 331 |
"step": 225
|
| 332 |
},
|
| 333 |
{
|
| 334 |
"epoch": 0.31063590320765333,
|
| 335 |
+
"grad_norm": 0.05878900335636033,
|
| 336 |
"learning_rate": 9.906780903767799e-06,
|
| 337 |
+
"loss": 0.3927,
|
| 338 |
"step": 230
|
| 339 |
},
|
| 340 |
{
|
| 341 |
"epoch": 0.317388857625211,
|
| 342 |
+
"grad_norm": 0.05009538620005328,
|
| 343 |
"learning_rate": 9.895107253416434e-06,
|
| 344 |
+
"loss": 0.3826,
|
| 345 |
"step": 235
|
| 346 |
},
|
| 347 |
{
|
| 348 |
"epoch": 0.3241418120427687,
|
| 349 |
+
"grad_norm": 0.06193069860962348,
|
| 350 |
"learning_rate": 9.882752851759247e-06,
|
| 351 |
+
"loss": 0.3694,
|
| 352 |
"step": 240
|
| 353 |
},
|
| 354 |
{
|
| 355 |
"epoch": 0.3308947664603264,
|
| 356 |
+
"grad_norm": 0.05658978100208689,
|
| 357 |
"learning_rate": 9.869719416894462e-06,
|
| 358 |
+
"loss": 0.3729,
|
| 359 |
"step": 245
|
| 360 |
},
|
| 361 |
{
|
| 362 |
"epoch": 0.3376477208778841,
|
| 363 |
+
"grad_norm": 0.060376802387617515,
|
| 364 |
"learning_rate": 9.856008761351882e-06,
|
| 365 |
+
"loss": 0.373,
|
| 366 |
"step": 250
|
| 367 |
},
|
| 368 |
{
|
| 369 |
"epoch": 0.34440067529544177,
|
| 370 |
+
"grad_norm": 0.047418407201082834,
|
| 371 |
"learning_rate": 9.841622791840839e-06,
|
| 372 |
+
"loss": 0.3343,
|
| 373 |
"step": 255
|
| 374 |
},
|
| 375 |
{
|
| 376 |
"epoch": 0.35115362971299946,
|
| 377 |
+
"grad_norm": 0.062039985277230605,
|
| 378 |
"learning_rate": 9.826563508985017e-06,
|
| 379 |
+
"loss": 0.3897,
|
| 380 |
"step": 260
|
| 381 |
},
|
| 382 |
{
|
| 383 |
"epoch": 0.3579065841305571,
|
| 384 |
+
"grad_norm": 0.04965955489714088,
|
| 385 |
"learning_rate": 9.810833007044247e-06,
|
| 386 |
"loss": 0.3925,
|
| 387 |
"step": 265
|
| 388 |
},
|
| 389 |
{
|
| 390 |
"epoch": 0.3646595385481148,
|
| 391 |
+
"grad_norm": 0.05316606675382464,
|
| 392 |
"learning_rate": 9.794433473623249e-06,
|
| 393 |
+
"loss": 0.377,
|
| 394 |
"step": 270
|
| 395 |
},
|
| 396 |
{
|
| 397 |
"epoch": 0.37141249296567247,
|
| 398 |
+
"grad_norm": 0.0467476855297103,
|
| 399 |
"learning_rate": 9.777367189367412e-06,
|
| 400 |
"loss": 0.3779,
|
| 401 |
"step": 275
|
| 402 |
},
|
| 403 |
{
|
| 404 |
"epoch": 0.37816544738323016,
|
| 405 |
+
"grad_norm": 0.04978059836724344,
|
| 406 |
"learning_rate": 9.759636527645633e-06,
|
| 407 |
+
"loss": 0.3521,
|
| 408 |
"step": 280
|
| 409 |
},
|
| 410 |
{
|
| 411 |
"epoch": 0.38491840180078785,
|
| 412 |
+
"grad_norm": 0.05999211649054394,
|
| 413 |
"learning_rate": 9.74124395422025e-06,
|
| 414 |
"loss": 0.3725,
|
| 415 |
"step": 285
|
| 416 |
},
|
| 417 |
{
|
| 418 |
"epoch": 0.39167135621834553,
|
| 419 |
+
"grad_norm": 0.04365453179095536,
|
| 420 |
"learning_rate": 9.722192026904145e-06,
|
| 421 |
+
"loss": 0.3402,
|
| 422 |
"step": 290
|
| 423 |
},
|
| 424 |
{
|
| 425 |
"epoch": 0.3984243106359032,
|
| 426 |
+
"grad_norm": 0.04616864773915026,
|
| 427 |
"learning_rate": 9.702483395205023e-06,
|
| 428 |
+
"loss": 0.3655,
|
| 429 |
"step": 295
|
| 430 |
},
|
| 431 |
{
|
| 432 |
"epoch": 0.4051772650534609,
|
| 433 |
+
"grad_norm": 0.04782080283525874,
|
| 434 |
"learning_rate": 9.682120799956961e-06,
|
| 435 |
+
"loss": 0.3263,
|
| 436 |
"step": 300
|
| 437 |
},
|
| 438 |
{
|
| 439 |
"epoch": 0.41193021947101854,
|
| 440 |
+
"grad_norm": 0.0498656978185221,
|
| 441 |
"learning_rate": 9.661107072939244e-06,
|
| 442 |
+
"loss": 0.3387,
|
| 443 |
"step": 305
|
| 444 |
},
|
| 445 |
{
|
| 446 |
"epoch": 0.41868317388857623,
|
| 447 |
+
"grad_norm": 0.04142545069347494,
|
| 448 |
"learning_rate": 9.639445136482549e-06,
|
| 449 |
+
"loss": 0.3388,
|
| 450 |
"step": 310
|
| 451 |
},
|
| 452 |
{
|
| 453 |
"epoch": 0.4254361283061339,
|
| 454 |
+
"grad_norm": 0.04183998822986712,
|
| 455 |
"learning_rate": 9.61713800306255e-06,
|
| 456 |
+
"loss": 0.3615,
|
| 457 |
"step": 315
|
| 458 |
},
|
| 459 |
{
|
| 460 |
"epoch": 0.4321890827236916,
|
| 461 |
+
"grad_norm": 0.045933316046675596,
|
| 462 |
"learning_rate": 9.594188774880981e-06,
|
| 463 |
+
"loss": 0.3607,
|
| 464 |
"step": 320
|
| 465 |
},
|
| 466 |
{
|
| 467 |
"epoch": 0.4389420371412493,
|
| 468 |
+
"grad_norm": 0.04805726955974,
|
| 469 |
"learning_rate": 9.570600643434217e-06,
|
| 470 |
"loss": 0.3371,
|
| 471 |
"step": 325
|
| 472 |
},
|
| 473 |
{
|
| 474 |
"epoch": 0.445694991558807,
|
| 475 |
+
"grad_norm": 0.04569635226134762,
|
| 476 |
"learning_rate": 9.546376889069442e-06,
|
| 477 |
+
"loss": 0.3582,
|
| 478 |
"step": 330
|
| 479 |
},
|
| 480 |
{
|
| 481 |
"epoch": 0.4524479459763647,
|
| 482 |
+
"grad_norm": 0.050347567122321094,
|
| 483 |
"learning_rate": 9.521520880528453e-06,
|
| 484 |
+
"loss": 0.3831,
|
| 485 |
"step": 335
|
| 486 |
},
|
| 487 |
{
|
| 488 |
"epoch": 0.45920090039392236,
|
| 489 |
+
"grad_norm": 0.051704382901289696,
|
| 490 |
"learning_rate": 9.496036074479184e-06,
|
| 491 |
"loss": 0.3604,
|
| 492 |
"step": 340
|
| 493 |
},
|
| 494 |
{
|
| 495 |
"epoch": 0.46595385481148,
|
| 496 |
+
"grad_norm": 0.04717995332220621,
|
| 497 |
"learning_rate": 9.46992601503499e-06,
|
| 498 |
+
"loss": 0.3498,
|
| 499 |
"step": 345
|
| 500 |
},
|
| 501 |
{
|
| 502 |
"epoch": 0.4727068092290377,
|
| 503 |
+
"grad_norm": 0.04394756827213909,
|
| 504 |
"learning_rate": 9.44319433326178e-06,
|
| 505 |
+
"loss": 0.3711,
|
| 506 |
"step": 350
|
| 507 |
},
|
| 508 |
{
|
| 509 |
"epoch": 0.47945976364659537,
|
| 510 |
+
"grad_norm": 0.05455789140928886,
|
| 511 |
"learning_rate": 9.415844746673047e-06,
|
| 512 |
+
"loss": 0.3395,
|
| 513 |
"step": 355
|
| 514 |
},
|
| 515 |
{
|
| 516 |
"epoch": 0.48621271806415306,
|
| 517 |
+
"grad_norm": 0.0486039405429815,
|
| 518 |
"learning_rate": 9.387881058712888e-06,
|
| 519 |
+
"loss": 0.3596,
|
| 520 |
"step": 360
|
| 521 |
},
|
| 522 |
{
|
| 523 |
"epoch": 0.49296567248171075,
|
| 524 |
+
"grad_norm": 0.049127971484592976,
|
| 525 |
"learning_rate": 9.359307158227067e-06,
|
| 526 |
+
"loss": 0.3759,
|
| 527 |
"step": 365
|
| 528 |
},
|
| 529 |
{
|
| 530 |
"epoch": 0.49971862689926844,
|
| 531 |
+
"grad_norm": 0.05804448606366045,
|
| 532 |
"learning_rate": 9.330127018922195e-06,
|
| 533 |
+
"loss": 0.3741,
|
| 534 |
"step": 370
|
| 535 |
},
|
| 536 |
{
|
| 537 |
"epoch": 0.5064715813168261,
|
| 538 |
+
"grad_norm": 0.05121055195790688,
|
| 539 |
"learning_rate": 9.300344698813124e-06,
|
| 540 |
+
"loss": 0.3978,
|
| 541 |
"step": 375
|
| 542 |
},
|
| 543 |
{
|
| 544 |
"epoch": 0.5132245357343838,
|
| 545 |
+
"grad_norm": 0.05072015885221285,
|
| 546 |
"learning_rate": 9.269964339658605e-06,
|
| 547 |
+
"loss": 0.349,
|
| 548 |
"step": 380
|
| 549 |
},
|
| 550 |
{
|
| 551 |
"epoch": 0.5199774901519415,
|
| 552 |
+
"grad_norm": 0.0446955462817659,
|
| 553 |
"learning_rate": 9.238990166385304e-06,
|
| 554 |
+
"loss": 0.3783,
|
| 555 |
"step": 385
|
| 556 |
},
|
| 557 |
{
|
| 558 |
"epoch": 0.5267304445694991,
|
| 559 |
+
"grad_norm": 0.047869039864268624,
|
| 560 |
"learning_rate": 9.207426486500252e-06,
|
| 561 |
"loss": 0.3587,
|
| 562 |
"step": 390
|
| 563 |
},
|
| 564 |
{
|
| 565 |
"epoch": 0.5334833989870569,
|
| 566 |
+
"grad_norm": 0.04183937339521655,
|
| 567 |
"learning_rate": 9.175277689491804e-06,
|
| 568 |
+
"loss": 0.3385,
|
| 569 |
"step": 395
|
| 570 |
},
|
| 571 |
{
|
| 572 |
"epoch": 0.5402363534046145,
|
| 573 |
+
"grad_norm": 0.04804994911136886,
|
| 574 |
"learning_rate": 9.142548246219212e-06,
|
| 575 |
+
"loss": 0.3696,
|
| 576 |
"step": 400
|
| 577 |
},
|
| 578 |
{
|
| 579 |
"epoch": 0.5469893078221723,
|
| 580 |
+
"grad_norm": 0.056116289553118946,
|
| 581 |
"learning_rate": 9.109242708290864e-06,
|
| 582 |
"loss": 0.3427,
|
| 583 |
"step": 405
|
| 584 |
},
|
| 585 |
{
|
| 586 |
"epoch": 0.5537422622397299,
|
| 587 |
+
"grad_norm": 0.04610990276019498,
|
| 588 |
"learning_rate": 9.075365707431311e-06,
|
| 589 |
+
"loss": 0.3684,
|
| 590 |
"step": 410
|
| 591 |
},
|
| 592 |
{
|
| 593 |
"epoch": 0.5604952166572875,
|
| 594 |
+
"grad_norm": 0.04711791241060553,
|
| 595 |
"learning_rate": 9.040921954837139e-06,
|
| 596 |
+
"loss": 0.3262,
|
| 597 |
"step": 415
|
| 598 |
},
|
| 599 |
{
|
| 600 |
"epoch": 0.5672481710748453,
|
| 601 |
+
"grad_norm": 0.0436305028584867,
|
| 602 |
"learning_rate": 9.005916240521788e-06,
|
| 603 |
+
"loss": 0.3466,
|
| 604 |
"step": 420
|
| 605 |
},
|
| 606 |
{
|
| 607 |
"epoch": 0.5740011254924029,
|
| 608 |
+
"grad_norm": 0.043193879987331325,
|
| 609 |
"learning_rate": 8.97035343264943e-06,
|
| 610 |
+
"loss": 0.3213,
|
| 611 |
"step": 425
|
| 612 |
},
|
| 613 |
{
|
| 614 |
"epoch": 0.5807540799099606,
|
| 615 |
+
"grad_norm": 0.04807145092887773,
|
| 616 |
"learning_rate": 8.93423847685795e-06,
|
| 617 |
+
"loss": 0.3553,
|
| 618 |
"step": 430
|
| 619 |
},
|
| 620 |
{
|
| 621 |
"epoch": 0.5875070343275183,
|
| 622 |
+
"grad_norm": 0.04707322735993806,
|
| 623 |
"learning_rate": 8.89757639557118e-06,
|
| 624 |
+
"loss": 0.3611,
|
| 625 |
"step": 435
|
| 626 |
},
|
| 627 |
{
|
| 628 |
"epoch": 0.594259988745076,
|
| 629 |
+
"grad_norm": 0.04694876715909799,
|
| 630 |
"learning_rate": 8.860372287300432e-06,
|
| 631 |
"loss": 0.3703,
|
| 632 |
"step": 440
|
| 633 |
},
|
| 634 |
{
|
| 635 |
"epoch": 0.6010129431626337,
|
| 636 |
+
"grad_norm": 0.051713244981409834,
|
| 637 |
"learning_rate": 8.822631325935463e-06,
|
| 638 |
+
"loss": 0.3763,
|
| 639 |
"step": 445
|
| 640 |
},
|
| 641 |
{
|
| 642 |
"epoch": 0.6077658975801913,
|
| 643 |
+
"grad_norm": 0.07337197692235135,
|
| 644 |
"learning_rate": 8.78435876002496e-06,
|
| 645 |
+
"loss": 0.3867,
|
| 646 |
"step": 450
|
| 647 |
},
|
| 648 |
{
|
| 649 |
"epoch": 0.614518851997749,
|
| 650 |
+
"grad_norm": 0.03984879157392941,
|
| 651 |
"learning_rate": 8.745559912046625e-06,
|
| 652 |
+
"loss": 0.3292,
|
| 653 |
"step": 455
|
| 654 |
},
|
| 655 |
{
|
| 656 |
"epoch": 0.6212718064153067,
|
| 657 |
+
"grad_norm": 0.042340122733838116,
|
| 658 |
"learning_rate": 8.706240177667003e-06,
|
| 659 |
"loss": 0.3476,
|
| 660 |
"step": 460
|
| 661 |
},
|
| 662 |
{
|
| 663 |
"epoch": 0.6280247608328644,
|
| 664 |
+
"grad_norm": 0.04565714030234626,
|
| 665 |
"learning_rate": 8.666405024991105e-06,
|
| 666 |
+
"loss": 0.3361,
|
| 667 |
"step": 465
|
| 668 |
},
|
| 669 |
{
|
| 670 |
"epoch": 0.634777715250422,
|
| 671 |
+
"grad_norm": 0.04169016093573778,
|
| 672 |
"learning_rate": 8.626059993801988e-06,
|
| 673 |
+
"loss": 0.339,
|
| 674 |
"step": 470
|
| 675 |
},
|
| 676 |
{
|
| 677 |
"epoch": 0.6415306696679798,
|
| 678 |
+
"grad_norm": 0.04325992397568874,
|
| 679 |
"learning_rate": 8.585210694790333e-06,
|
| 680 |
+
"loss": 0.3366,
|
| 681 |
"step": 475
|
| 682 |
},
|
| 683 |
{
|
| 684 |
"epoch": 0.6482836240855374,
|
| 685 |
+
"grad_norm": 0.0415410087146755,
|
| 686 |
"learning_rate": 8.543862808774193e-06,
|
| 687 |
+
"loss": 0.3489,
|
| 688 |
"step": 480
|
| 689 |
},
|
| 690 |
{
|
| 691 |
"epoch": 0.6550365785030952,
|
| 692 |
+
"grad_norm": 0.04286421536314813,
|
| 693 |
"learning_rate": 8.502022085908963e-06,
|
| 694 |
"loss": 0.3471,
|
| 695 |
"step": 485
|
| 696 |
},
|
| 697 |
{
|
| 698 |
"epoch": 0.6617895329206528,
|
| 699 |
+
"grad_norm": 0.04783002216381401,
|
| 700 |
"learning_rate": 8.459694344887732e-06,
|
| 701 |
+
"loss": 0.4098,
|
| 702 |
"step": 490
|
| 703 |
},
|
| 704 |
{
|
| 705 |
"epoch": 0.6685424873382104,
|
| 706 |
+
"grad_norm": 0.04187887777828664,
|
| 707 |
"learning_rate": 8.416885472132077e-06,
|
| 708 |
+
"loss": 0.3462,
|
| 709 |
"step": 495
|
| 710 |
},
|
| 711 |
{
|
| 712 |
"epoch": 0.6752954417557682,
|
| 713 |
+
"grad_norm": 0.05373723455097553,
|
| 714 |
"learning_rate": 8.373601420973464e-06,
|
| 715 |
+
"loss": 0.344,
|
| 716 |
"step": 500
|
| 717 |
},
|
| 718 |
{
|
| 719 |
"epoch": 0.6820483961733258,
|
| 720 |
+
"grad_norm": 0.04300150710853549,
|
| 721 |
"learning_rate": 8.329848210825322e-06,
|
| 722 |
+
"loss": 0.3647,
|
| 723 |
"step": 505
|
| 724 |
},
|
| 725 |
{
|
| 726 |
"epoch": 0.6888013505908835,
|
| 727 |
+
"grad_norm": 0.043097229603458374,
|
| 728 |
"learning_rate": 8.285631926345943e-06,
|
| 729 |
+
"loss": 0.3693,
|
| 730 |
"step": 510
|
| 731 |
},
|
| 732 |
{
|
| 733 |
"epoch": 0.6955543050084412,
|
| 734 |
+
"grad_norm": 0.040768739194901454,
|
| 735 |
"learning_rate": 8.240958716592304e-06,
|
| 736 |
+
"loss": 0.3645,
|
| 737 |
"step": 515
|
| 738 |
},
|
| 739 |
{
|
| 740 |
"epoch": 0.7023072594259989,
|
| 741 |
+
"grad_norm": 0.046324281574150986,
|
| 742 |
"learning_rate": 8.195834794164925e-06,
|
| 743 |
"loss": 0.3501,
|
| 744 |
"step": 520
|
| 745 |
},
|
| 746 |
{
|
| 747 |
"epoch": 0.7090602138435566,
|
| 748 |
+
"grad_norm": 0.04048770768699556,
|
| 749 |
"learning_rate": 8.150266434343904e-06,
|
| 750 |
"loss": 0.3296,
|
| 751 |
"step": 525
|
| 752 |
},
|
| 753 |
{
|
| 754 |
"epoch": 0.7158131682611142,
|
| 755 |
+
"grad_norm": 0.04249502701153697,
|
| 756 |
"learning_rate": 8.104259974216219e-06,
|
| 757 |
+
"loss": 0.3489,
|
| 758 |
"step": 530
|
| 759 |
},
|
| 760 |
{
|
| 761 |
"epoch": 0.7225661226786719,
|
| 762 |
+
"grad_norm": 0.03877822496803113,
|
| 763 |
"learning_rate": 8.057821811794457e-06,
|
| 764 |
+
"loss": 0.3673,
|
| 765 |
"step": 535
|
| 766 |
},
|
| 767 |
{
|
| 768 |
"epoch": 0.7293190770962296,
|
| 769 |
+
"grad_norm": 0.048769655901584476,
|
| 770 |
"learning_rate": 8.010958405127048e-06,
|
| 771 |
+
"loss": 0.3313,
|
| 772 |
"step": 540
|
| 773 |
},
|
| 774 |
{
|
| 775 |
"epoch": 0.7360720315137873,
|
| 776 |
+
"grad_norm": 0.07133171500587031,
|
| 777 |
"learning_rate": 7.963676271400158e-06,
|
| 778 |
+
"loss": 0.3472,
|
| 779 |
"step": 545
|
| 780 |
},
|
| 781 |
{
|
| 782 |
"epoch": 0.7428249859313449,
|
| 783 |
+
"grad_norm": 0.04064256549350507,
|
| 784 |
"learning_rate": 7.915981986031367e-06,
|
| 785 |
+
"loss": 0.3478,
|
| 786 |
"step": 550
|
| 787 |
},
|
| 788 |
{
|
| 789 |
"epoch": 0.7495779403489027,
|
| 790 |
+
"grad_norm": 0.04549548056733051,
|
| 791 |
"learning_rate": 7.86788218175523e-06,
|
| 792 |
+
"loss": 0.3342,
|
| 793 |
"step": 555
|
| 794 |
},
|
| 795 |
{
|
| 796 |
"epoch": 0.7563308947664603,
|
| 797 |
+
"grad_norm": 0.03599813075890018,
|
| 798 |
"learning_rate": 7.819383547700889e-06,
|
| 799 |
"loss": 0.3354,
|
| 800 |
"step": 560
|
| 801 |
},
|
| 802 |
{
|
| 803 |
"epoch": 0.7630838491840181,
|
| 804 |
+
"grad_norm": 0.04316680817634258,
|
| 805 |
"learning_rate": 7.770492828461824e-06,
|
| 806 |
+
"loss": 0.3622,
|
| 807 |
"step": 565
|
| 808 |
},
|
| 809 |
{
|
| 810 |
"epoch": 0.7698368036015757,
|
| 811 |
+
"grad_norm": 0.04908046724152037,
|
| 812 |
"learning_rate": 7.721216823157896e-06,
|
| 813 |
+
"loss": 0.3559,
|
| 814 |
"step": 570
|
| 815 |
},
|
| 816 |
{
|
| 817 |
"epoch": 0.7765897580191333,
|
| 818 |
+
"grad_norm": 0.040354310023923715,
|
| 819 |
"learning_rate": 7.671562384489819e-06,
|
| 820 |
+
"loss": 0.3629,
|
| 821 |
"step": 575
|
| 822 |
},
|
| 823 |
{
|
| 824 |
"epoch": 0.7833427124366911,
|
| 825 |
+
"grad_norm": 0.043579845331084145,
|
| 826 |
"learning_rate": 7.621536417786159e-06,
|
| 827 |
+
"loss": 0.3404,
|
| 828 |
"step": 580
|
| 829 |
},
|
| 830 |
{
|
| 831 |
"epoch": 0.7900956668542487,
|
| 832 |
+
"grad_norm": 0.04040670293206675,
|
| 833 |
"learning_rate": 7.571145880043036e-06,
|
| 834 |
+
"loss": 0.3672,
|
| 835 |
"step": 585
|
| 836 |
},
|
| 837 |
{
|
| 838 |
"epoch": 0.7968486212718064,
|
| 839 |
+
"grad_norm": 0.04244690517322693,
|
| 840 |
"learning_rate": 7.520397778956623e-06,
|
| 841 |
+
"loss": 0.3499,
|
| 842 |
"step": 590
|
| 843 |
},
|
| 844 |
{
|
| 845 |
"epoch": 0.8036015756893641,
|
| 846 |
+
"grad_norm": 0.05433806018620802,
|
| 847 |
"learning_rate": 7.469299171948608e-06,
|
| 848 |
+
"loss": 0.3704,
|
| 849 |
"step": 595
|
| 850 |
},
|
| 851 |
{
|
| 852 |
"epoch": 0.8103545301069218,
|
| 853 |
+
"grad_norm": 0.04428365952002865,
|
| 854 |
"learning_rate": 7.417857165184723e-06,
|
| 855 |
"loss": 0.3474,
|
| 856 |
"step": 600
|
| 857 |
},
|
| 858 |
{
|
| 859 |
"epoch": 0.8171074845244795,
|
| 860 |
+
"grad_norm": 0.03948151196873183,
|
| 861 |
"learning_rate": 7.366078912586523e-06,
|
| 862 |
+
"loss": 0.3426,
|
| 863 |
"step": 605
|
| 864 |
},
|
| 865 |
{
|
| 866 |
"epoch": 0.8238604389420371,
|
| 867 |
+
"grad_norm": 0.03804562874118621,
|
| 868 |
"learning_rate": 7.313971614836496e-06,
|
| 869 |
+
"loss": 0.3418,
|
| 870 |
"step": 610
|
| 871 |
},
|
| 872 |
{
|
| 873 |
"epoch": 0.8306133933595948,
|
| 874 |
+
"grad_norm": 0.04327583528687119,
|
| 875 |
"learning_rate": 7.261542518376677e-06,
|
| 876 |
+
"loss": 0.3408,
|
| 877 |
"step": 615
|
| 878 |
},
|
| 879 |
{
|
| 880 |
"epoch": 0.8373663477771525,
|
| 881 |
+
"grad_norm": 0.038581581579252,
|
| 882 |
"learning_rate": 7.208798914400916e-06,
|
| 883 |
+
"loss": 0.3429,
|
| 884 |
"step": 620
|
| 885 |
},
|
| 886 |
{
|
| 887 |
"epoch": 0.8441193021947102,
|
| 888 |
+
"grad_norm": 0.040274824329860945,
|
| 889 |
"learning_rate": 7.155748137840892e-06,
|
| 890 |
+
"loss": 0.385,
|
| 891 |
"step": 625
|
| 892 |
},
|
| 893 |
{
|
| 894 |
"epoch": 0.8508722566122678,
|
| 895 |
+
"grad_norm": 0.043944187715932226,
|
| 896 |
"learning_rate": 7.102397566346073e-06,
|
| 897 |
+
"loss": 0.3366,
|
| 898 |
"step": 630
|
| 899 |
},
|
| 900 |
{
|
| 901 |
"epoch": 0.8576252110298256,
|
| 902 |
+
"grad_norm": 0.04145675205661903,
|
| 903 |
"learning_rate": 7.048754619257716e-06,
|
| 904 |
+
"loss": 0.3755,
|
| 905 |
"step": 635
|
| 906 |
},
|
| 907 |
{
|
| 908 |
"epoch": 0.8643781654473832,
|
| 909 |
+
"grad_norm": 0.041045163518307656,
|
| 910 |
"learning_rate": 6.994826756577082e-06,
|
| 911 |
+
"loss": 0.3722,
|
| 912 |
"step": 640
|
| 913 |
},
|
| 914 |
{
|
| 915 |
"epoch": 0.871131119864941,
|
| 916 |
+
"grad_norm": 0.05061812503051895,
|
| 917 |
"learning_rate": 6.940621477927988e-06,
|
| 918 |
+
"loss": 0.3687,
|
| 919 |
"step": 645
|
| 920 |
},
|
| 921 |
{
|
| 922 |
"epoch": 0.8778840742824986,
|
| 923 |
+
"grad_norm": 0.03703919699696153,
|
| 924 |
"learning_rate": 6.88614632151385e-06,
|
| 925 |
+
"loss": 0.3454,
|
| 926 |
"step": 650
|
| 927 |
},
|
| 928 |
{
|
| 929 |
"epoch": 0.8846370287000562,
|
| 930 |
+
"grad_norm": 0.039635652764389524,
|
| 931 |
"learning_rate": 6.831408863069364e-06,
|
| 932 |
+
"loss": 0.365,
|
| 933 |
"step": 655
|
| 934 |
},
|
| 935 |
{
|
| 936 |
"epoch": 0.891389983117614,
|
| 937 |
+
"grad_norm": 0.0428338033091146,
|
| 938 |
"learning_rate": 6.7764167148069695e-06,
|
| 939 |
"loss": 0.3251,
|
| 940 |
"step": 660
|
| 941 |
},
|
| 942 |
{
|
| 943 |
"epoch": 0.8981429375351716,
|
| 944 |
+
"grad_norm": 0.0393255629299852,
|
| 945 |
"learning_rate": 6.721177524358226e-06,
|
| 946 |
+
"loss": 0.3519,
|
| 947 |
"step": 665
|
| 948 |
},
|
| 949 |
{
|
| 950 |
"epoch": 0.9048958919527293,
|
| 951 |
+
"grad_norm": 0.03605877433184958,
|
| 952 |
"learning_rate": 6.665698973710289e-06,
|
| 953 |
+
"loss": 0.3586,
|
| 954 |
"step": 670
|
| 955 |
},
|
| 956 |
{
|
| 957 |
"epoch": 0.911648846370287,
|
| 958 |
+
"grad_norm": 0.0418094941760349,
|
| 959 |
"learning_rate": 6.609988778137582e-06,
|
| 960 |
+
"loss": 0.333,
|
| 961 |
"step": 675
|
| 962 |
},
|
| 963 |
{
|
| 964 |
"epoch": 0.9184018007878447,
|
| 965 |
+
"grad_norm": 0.04338138841178037,
|
| 966 |
"learning_rate": 6.554054685128857e-06,
|
| 967 |
+
"loss": 0.334,
|
| 968 |
"step": 680
|
| 969 |
},
|
| 970 |
{
|
| 971 |
"epoch": 0.9251547552054024,
|
| 972 |
+
"grad_norm": 0.04225445092421321,
|
| 973 |
"learning_rate": 6.497904473309766e-06,
|
| 974 |
"loss": 0.3519,
|
| 975 |
"step": 685
|
| 976 |
},
|
| 977 |
{
|
| 978 |
"epoch": 0.93190770962296,
|
| 979 |
+
"grad_norm": 0.04826057110998316,
|
| 980 |
"learning_rate": 6.44154595136111e-06,
|
| 981 |
+
"loss": 0.3598,
|
| 982 |
"step": 690
|
| 983 |
},
|
| 984 |
{
|
| 985 |
"epoch": 0.9386606640405177,
|
| 986 |
+
"grad_norm": 0.04381442124287324,
|
| 987 |
"learning_rate": 6.384986956932897e-06,
|
| 988 |
"loss": 0.3339,
|
| 989 |
"step": 695
|
| 990 |
},
|
| 991 |
{
|
| 992 |
"epoch": 0.9454136184580754,
|
| 993 |
+
"grad_norm": 0.0414974847898415,
|
| 994 |
"learning_rate": 6.328235355554382e-06,
|
| 995 |
+
"loss": 0.3645,
|
| 996 |
"step": 700
|
| 997 |
},
|
| 998 |
{
|
| 999 |
"epoch": 0.9521665728756331,
|
| 1000 |
+
"grad_norm": 0.04276356905517851,
|
| 1001 |
"learning_rate": 6.271299039540228e-06,
|
| 1002 |
"loss": 0.3645,
|
| 1003 |
"step": 705
|
| 1004 |
},
|
| 1005 |
{
|
| 1006 |
"epoch": 0.9589195272931907,
|
| 1007 |
+
"grad_norm": 0.046676300677672276,
|
| 1008 |
"learning_rate": 6.214185926892936e-06,
|
| 1009 |
+
"loss": 0.3551,
|
| 1010 |
"step": 710
|
| 1011 |
},
|
| 1012 |
{
|
| 1013 |
"epoch": 0.9656724817107485,
|
| 1014 |
+
"grad_norm": 0.04006191125365821,
|
| 1015 |
"learning_rate": 6.156903960201709e-06,
|
| 1016 |
+
"loss": 0.3395,
|
| 1017 |
"step": 715
|
| 1018 |
},
|
| 1019 |
{
|
| 1020 |
"epoch": 0.9724254361283061,
|
| 1021 |
+
"grad_norm": 0.047147919741739285,
|
| 1022 |
"learning_rate": 6.099461105537889e-06,
|
| 1023 |
+
"loss": 0.3321,
|
| 1024 |
"step": 720
|
| 1025 |
},
|
| 1026 |
{
|
| 1027 |
"epoch": 0.9791783905458639,
|
| 1028 |
+
"grad_norm": 0.041770572631853936,
|
| 1029 |
"learning_rate": 6.041865351347146e-06,
|
| 1030 |
"loss": 0.3421,
|
| 1031 |
"step": 725
|
| 1032 |
},
|
| 1033 |
{
|
| 1034 |
"epoch": 0.9859313449634215,
|
| 1035 |
+
"grad_norm": 0.04239922397171486,
|
| 1036 |
"learning_rate": 5.984124707338528e-06,
|
| 1037 |
+
"loss": 0.3373,
|
| 1038 |
"step": 730
|
| 1039 |
},
|
| 1040 |
{
|
| 1041 |
"epoch": 0.9926842993809791,
|
| 1042 |
+
"grad_norm": 0.046194589474773985,
|
| 1043 |
"learning_rate": 5.926247203370583e-06,
|
| 1044 |
+
"loss": 0.3405,
|
| 1045 |
"step": 735
|
| 1046 |
},
|
| 1047 |
{
|
| 1048 |
"epoch": 0.9994372537985369,
|
| 1049 |
+
"grad_norm": 0.03944962664459983,
|
| 1050 |
"learning_rate": 5.8682408883346535e-06,
|
| 1051 |
"loss": 0.2909,
|
| 1052 |
"step": 740
|
| 1053 |
},
|
| 1054 |
{
|
| 1055 |
"epoch": 1.0,
|
| 1056 |
+
"eval_loss": 0.3324998915195465,
|
| 1057 |
+
"eval_runtime": 18.1309,
|
| 1058 |
+
"eval_samples_per_second": 19.8,
|
| 1059 |
+
"eval_steps_per_second": 4.964,
|
| 1060 |
"step": 741
|
| 1061 |
},
|
| 1062 |
{
|
| 1063 |
"epoch": 1.005402363534046,
|
| 1064 |
+
"grad_norm": 0.040121304388969325,
|
| 1065 |
"learning_rate": 5.810113829035544e-06,
|
| 1066 |
+
"loss": 0.355,
|
| 1067 |
"step": 745
|
| 1068 |
},
|
| 1069 |
{
|
| 1070 |
"epoch": 1.0121553179516039,
|
| 1071 |
+
"grad_norm": 0.047689669773125845,
|
| 1072 |
"learning_rate": 5.751874109069685e-06,
|
| 1073 |
"loss": 0.3611,
|
| 1074 |
"step": 750
|
| 1075 |
},
|
| 1076 |
{
|
| 1077 |
"epoch": 1.0189082723691616,
|
| 1078 |
+
"grad_norm": 0.03912739831058518,
|
| 1079 |
"learning_rate": 5.693529827700967e-06,
|
| 1080 |
+
"loss": 0.2899,
|
| 1081 |
"step": 755
|
| 1082 |
},
|
| 1083 |
{
|
| 1084 |
"epoch": 1.0256612267867191,
|
| 1085 |
+
"grad_norm": 0.04377909856345864,
|
| 1086 |
"learning_rate": 5.635089098734394e-06,
|
| 1087 |
+
"loss": 0.3197,
|
| 1088 |
"step": 760
|
| 1089 |
},
|
| 1090 |
{
|
| 1091 |
"epoch": 1.0324141812042769,
|
| 1092 |
+
"grad_norm": 0.037393171944725745,
|
| 1093 |
"learning_rate": 5.576560049387717e-06,
|
| 1094 |
"loss": 0.339,
|
| 1095 |
"step": 765
|
| 1096 |
},
|
| 1097 |
{
|
| 1098 |
"epoch": 1.0391671356218346,
|
| 1099 |
+
"grad_norm": 0.042256918616508245,
|
| 1100 |
"learning_rate": 5.517950819161196e-06,
|
| 1101 |
"loss": 0.328,
|
| 1102 |
"step": 770
|
| 1103 |
},
|
| 1104 |
{
|
| 1105 |
"epoch": 1.0459200900393921,
|
| 1106 |
+
"grad_norm": 0.040529502097932234,
|
| 1107 |
"learning_rate": 5.459269558705667e-06,
|
| 1108 |
+
"loss": 0.3259,
|
| 1109 |
"step": 775
|
| 1110 |
},
|
| 1111 |
{
|
| 1112 |
"epoch": 1.0526730444569499,
|
| 1113 |
+
"grad_norm": 0.04379220562588465,
|
| 1114 |
"learning_rate": 5.400524428689035e-06,
|
| 1115 |
+
"loss": 0.3464,
|
| 1116 |
"step": 780
|
| 1117 |
},
|
| 1118 |
{
|
| 1119 |
"epoch": 1.0594259988745076,
|
| 1120 |
+
"grad_norm": 0.04725991662965075,
|
| 1121 |
"learning_rate": 5.341723598661409e-06,
|
| 1122 |
"loss": 0.3267,
|
| 1123 |
"step": 785
|
| 1124 |
},
|
| 1125 |
{
|
| 1126 |
"epoch": 1.0661789532920654,
|
| 1127 |
+
"grad_norm": 0.036832603953267457,
|
| 1128 |
"learning_rate": 5.282875245918963e-06,
|
| 1129 |
+
"loss": 0.3261,
|
| 1130 |
"step": 790
|
| 1131 |
},
|
| 1132 |
{
|
| 1133 |
"epoch": 1.0729319077096229,
|
| 1134 |
+
"grad_norm": 0.04175828917214628,
|
| 1135 |
"learning_rate": 5.2239875543667465e-06,
|
| 1136 |
+
"loss": 0.3346,
|
| 1137 |
"step": 795
|
| 1138 |
},
|
| 1139 |
{
|
| 1140 |
"epoch": 1.0796848621271806,
|
| 1141 |
+
"grad_norm": 0.03618010757790516,
|
| 1142 |
"learning_rate": 5.165068713380568e-06,
|
| 1143 |
+
"loss": 0.3182,
|
| 1144 |
"step": 800
|
| 1145 |
},
|
| 1146 |
{
|
| 1147 |
"epoch": 1.0864378165447384,
|
| 1148 |
+
"grad_norm": 0.04097386666815235,
|
| 1149 |
"learning_rate": 5.106126916668118e-06,
|
| 1150 |
"loss": 0.3192,
|
| 1151 |
"step": 805
|
| 1152 |
},
|
| 1153 |
{
|
| 1154 |
"epoch": 1.093190770962296,
|
| 1155 |
+
"grad_norm": 0.040300039979862916,
|
| 1156 |
"learning_rate": 5.047170361129484e-06,
|
| 1157 |
+
"loss": 0.3138,
|
| 1158 |
"step": 810
|
| 1159 |
},
|
| 1160 |
{
|
| 1161 |
"epoch": 1.0999437253798536,
|
| 1162 |
+
"grad_norm": 0.0413703774570022,
|
| 1163 |
"learning_rate": 4.988207245717232e-06,
|
| 1164 |
"loss": 0.3558,
|
| 1165 |
"step": 815
|
| 1166 |
},
|
| 1167 |
{
|
| 1168 |
"epoch": 1.1066966797974114,
|
| 1169 |
+
"grad_norm": 0.03938201981503368,
|
| 1170 |
"learning_rate": 4.929245770296191e-06,
|
| 1171 |
+
"loss": 0.3436,
|
| 1172 |
"step": 820
|
| 1173 |
},
|
| 1174 |
{
|
| 1175 |
"epoch": 1.1134496342149691,
|
| 1176 |
+
"grad_norm": 0.043903225983181984,
|
| 1177 |
"learning_rate": 4.870294134503123e-06,
|
| 1178 |
+
"loss": 0.3301,
|
| 1179 |
"step": 825
|
| 1180 |
},
|
| 1181 |
{
|
| 1182 |
"epoch": 1.1202025886325266,
|
| 1183 |
+
"grad_norm": 0.04208713272188743,
|
| 1184 |
"learning_rate": 4.811360536606416e-06,
|
| 1185 |
+
"loss": 0.3108,
|
| 1186 |
"step": 830
|
| 1187 |
},
|
| 1188 |
{
|
| 1189 |
"epoch": 1.1269555430500844,
|
| 1190 |
+
"grad_norm": 0.03660393810864938,
|
| 1191 |
"learning_rate": 4.752453172365966e-06,
|
| 1192 |
+
"loss": 0.333,
|
| 1193 |
"step": 835
|
| 1194 |
},
|
| 1195 |
{
|
| 1196 |
"epoch": 1.1337084974676421,
|
| 1197 |
+
"grad_norm": 0.04407131833701405,
|
| 1198 |
"learning_rate": 4.69358023389342e-06,
|
| 1199 |
+
"loss": 0.3605,
|
| 1200 |
"step": 840
|
| 1201 |
},
|
| 1202 |
{
|
| 1203 |
"epoch": 1.1404614518851997,
|
| 1204 |
+
"grad_norm": 0.04616999836259198,
|
| 1205 |
"learning_rate": 4.634749908512907e-06,
|
| 1206 |
+
"loss": 0.3439,
|
| 1207 |
"step": 845
|
| 1208 |
},
|
| 1209 |
{
|
| 1210 |
"epoch": 1.1472144063027574,
|
| 1211 |
+
"grad_norm": 0.04325365689860756,
|
| 1212 |
"learning_rate": 4.575970377622456e-06,
|
| 1213 |
+
"loss": 0.3049,
|
| 1214 |
"step": 850
|
| 1215 |
},
|
| 1216 |
{
|
| 1217 |
"epoch": 1.1539673607203151,
|
| 1218 |
+
"grad_norm": 0.03846831221970018,
|
| 1219 |
"learning_rate": 4.517249815556219e-06,
|
| 1220 |
+
"loss": 0.3511,
|
| 1221 |
"step": 855
|
| 1222 |
},
|
| 1223 |
{
|
| 1224 |
"epoch": 1.1607203151378729,
|
| 1225 |
+
"grad_norm": 0.04569268985609919,
|
| 1226 |
"learning_rate": 4.458596388447691e-06,
|
| 1227 |
+
"loss": 0.3209,
|
| 1228 |
"step": 860
|
| 1229 |
},
|
| 1230 |
{
|
| 1231 |
"epoch": 1.1674732695554304,
|
| 1232 |
+
"grad_norm": 0.0409252349150715,
|
| 1233 |
"learning_rate": 4.400018253094065e-06,
|
| 1234 |
+
"loss": 0.308,
|
| 1235 |
"step": 865
|
| 1236 |
},
|
| 1237 |
{
|
| 1238 |
"epoch": 1.1742262239729881,
|
| 1239 |
+
"grad_norm": 0.04133167625196089,
|
| 1240 |
"learning_rate": 4.341523555821882e-06,
|
| 1241 |
+
"loss": 0.3377,
|
| 1242 |
"step": 870
|
| 1243 |
},
|
| 1244 |
{
|
| 1245 |
"epoch": 1.180979178390546,
|
| 1246 |
+
"grad_norm": 0.043097502447341854,
|
| 1247 |
"learning_rate": 4.283120431354137e-06,
|
| 1248 |
+
"loss": 0.3117,
|
| 1249 |
"step": 875
|
| 1250 |
},
|
| 1251 |
{
|
| 1252 |
"epoch": 1.1877321328081036,
|
| 1253 |
+
"grad_norm": 0.042543275375795706,
|
| 1254 |
"learning_rate": 4.224817001679011e-06,
|
| 1255 |
+
"loss": 0.3344,
|
| 1256 |
"step": 880
|
| 1257 |
},
|
| 1258 |
{
|
| 1259 |
"epoch": 1.1944850872256612,
|
| 1260 |
+
"grad_norm": 0.04304997898343135,
|
| 1261 |
"learning_rate": 4.1666213749203545e-06,
|
| 1262 |
"loss": 0.3237,
|
| 1263 |
"step": 885
|
| 1264 |
},
|
| 1265 |
{
|
| 1266 |
"epoch": 1.201238041643219,
|
| 1267 |
+
"grad_norm": 0.04100222680465783,
|
| 1268 |
"learning_rate": 4.1085416442101205e-06,
|
| 1269 |
+
"loss": 0.3641,
|
| 1270 |
"step": 890
|
| 1271 |
},
|
| 1272 |
{
|
| 1273 |
"epoch": 1.2079909960607766,
|
| 1274 |
+
"grad_norm": 0.04523396323757555,
|
| 1275 |
"learning_rate": 4.050585886562858e-06,
|
| 1276 |
+
"loss": 0.3494,
|
| 1277 |
"step": 895
|
| 1278 |
},
|
| 1279 |
{
|
| 1280 |
"epoch": 1.2147439504783342,
|
| 1281 |
+
"grad_norm": 0.042513327919157136,
|
| 1282 |
"learning_rate": 3.992762161752474e-06,
|
| 1283 |
+
"loss": 0.343,
|
| 1284 |
"step": 900
|
| 1285 |
},
|
| 1286 |
{
|
| 1287 |
"epoch": 1.221496904895892,
|
| 1288 |
+
"grad_norm": 0.046594930027348866,
|
| 1289 |
"learning_rate": 3.935078511191368e-06,
|
| 1290 |
+
"loss": 0.3235,
|
| 1291 |
"step": 905
|
| 1292 |
},
|
| 1293 |
{
|
| 1294 |
"epoch": 1.2282498593134497,
|
| 1295 |
+
"grad_norm": 0.03906041128591151,
|
| 1296 |
"learning_rate": 3.877542956812137e-06,
|
| 1297 |
"loss": 0.3221,
|
| 1298 |
"step": 910
|
| 1299 |
},
|
| 1300 |
{
|
| 1301 |
"epoch": 1.2350028137310074,
|
| 1302 |
+
"grad_norm": 0.044515080037505254,
|
| 1303 |
"learning_rate": 3.820163499951984e-06,
|
| 1304 |
"loss": 0.3579,
|
| 1305 |
"step": 915
|
| 1306 |
},
|
| 1307 |
{
|
| 1308 |
"epoch": 1.241755768148565,
|
| 1309 |
+
"grad_norm": 0.04205841651552162,
|
| 1310 |
"learning_rate": 3.7629481202399886e-06,
|
| 1311 |
+
"loss": 0.3375,
|
| 1312 |
"step": 920
|
| 1313 |
},
|
| 1314 |
{
|
| 1315 |
"epoch": 1.2485087225661227,
|
| 1316 |
+
"grad_norm": 0.04152789361719496,
|
| 1317 |
"learning_rate": 3.705904774487396e-06,
|
| 1318 |
+
"loss": 0.2971,
|
| 1319 |
"step": 925
|
| 1320 |
},
|
| 1321 |
{
|
| 1322 |
"epoch": 1.2552616769836804,
|
| 1323 |
+
"grad_norm": 0.05137094607363216,
|
| 1324 |
"learning_rate": 3.64904139558109e-06,
|
| 1325 |
+
"loss": 0.3343,
|
| 1326 |
"step": 930
|
| 1327 |
},
|
| 1328 |
{
|
| 1329 |
"epoch": 1.2620146314012382,
|
| 1330 |
+
"grad_norm": 0.04499346319735512,
|
| 1331 |
"learning_rate": 3.5923658913803726e-06,
|
| 1332 |
+
"loss": 0.3274,
|
| 1333 |
"step": 935
|
| 1334 |
},
|
| 1335 |
{
|
| 1336 |
"epoch": 1.2687675858187957,
|
| 1337 |
+
"grad_norm": 0.04194049880241552,
|
| 1338 |
"learning_rate": 3.5358861436172487e-06,
|
| 1339 |
+
"loss": 0.3436,
|
| 1340 |
"step": 940
|
| 1341 |
},
|
| 1342 |
{
|
| 1343 |
"epoch": 1.2755205402363534,
|
| 1344 |
+
"grad_norm": 0.04739038652189242,
|
| 1345 |
"learning_rate": 3.47961000680032e-06,
|
| 1346 |
+
"loss": 0.3395,
|
| 1347 |
"step": 945
|
| 1348 |
},
|
| 1349 |
{
|
| 1350 |
"epoch": 1.2822734946539112,
|
| 1351 |
+
"grad_norm": 0.04407332378962217,
|
| 1352 |
"learning_rate": 3.4235453071224882e-06,
|
| 1353 |
+
"loss": 0.3283,
|
| 1354 |
"step": 950
|
| 1355 |
},
|
| 1356 |
{
|
| 1357 |
"epoch": 1.2890264490714687,
|
| 1358 |
+
"grad_norm": 0.046768041779961654,
|
| 1359 |
"learning_rate": 3.3676998413725726e-06,
|
| 1360 |
+
"loss": 0.3457,
|
| 1361 |
"step": 955
|
| 1362 |
},
|
| 1363 |
{
|
| 1364 |
"epoch": 1.2957794034890264,
|
| 1365 |
+
"grad_norm": 0.04405892619650361,
|
| 1366 |
"learning_rate": 3.3120813758510385e-06,
|
| 1367 |
+
"loss": 0.3446,
|
| 1368 |
"step": 960
|
| 1369 |
},
|
| 1370 |
{
|
| 1371 |
"epoch": 1.3025323579065842,
|
| 1372 |
+
"grad_norm": 0.04917585118916133,
|
| 1373 |
"learning_rate": 3.2566976452899507e-06,
|
| 1374 |
+
"loss": 0.3267,
|
| 1375 |
"step": 965
|
| 1376 |
},
|
| 1377 |
{
|
| 1378 |
"epoch": 1.3092853123241417,
|
| 1379 |
+
"grad_norm": 0.04327329891699119,
|
| 1380 |
"learning_rate": 3.2015563517773214e-06,
|
| 1381 |
+
"loss": 0.3382,
|
| 1382 |
"step": 970
|
| 1383 |
},
|
| 1384 |
{
|
| 1385 |
"epoch": 1.3160382667416994,
|
| 1386 |
+
"grad_norm": 0.04751650625794217,
|
| 1387 |
"learning_rate": 3.1466651636860025e-06,
|
| 1388 |
"loss": 0.3325,
|
| 1389 |
"step": 975
|
| 1390 |
},
|
| 1391 |
{
|
| 1392 |
"epoch": 1.3227912211592572,
|
| 1393 |
+
"grad_norm": 0.05680333767716324,
|
| 1394 |
"learning_rate": 3.0920317146072577e-06,
|
| 1395 |
+
"loss": 0.3207,
|
| 1396 |
"step": 980
|
| 1397 |
},
|
| 1398 |
{
|
| 1399 |
"epoch": 1.329544175576815,
|
| 1400 |
+
"grad_norm": 0.0497656718123479,
|
| 1401 |
"learning_rate": 3.0376636022891813e-06,
|
| 1402 |
+
"loss": 0.329,
|
| 1403 |
"step": 985
|
| 1404 |
},
|
| 1405 |
{
|
| 1406 |
"epoch": 1.3362971299943727,
|
| 1407 |
+
"grad_norm": 0.048529124206900136,
|
| 1408 |
"learning_rate": 2.983568387580094e-06,
|
| 1409 |
+
"loss": 0.3322,
|
| 1410 |
"step": 990
|
| 1411 |
},
|
| 1412 |
{
|
| 1413 |
"epoch": 1.3430500844119302,
|
| 1414 |
+
"grad_norm": 0.04475425551771912,
|
| 1415 |
"learning_rate": 2.9297535933770732e-06,
|
| 1416 |
+
"loss": 0.3278,
|
| 1417 |
"step": 995
|
| 1418 |
},
|
| 1419 |
{
|
| 1420 |
"epoch": 1.349803038829488,
|
| 1421 |
+
"grad_norm": 0.043952297509005415,
|
| 1422 |
"learning_rate": 2.8762267035797607e-06,
|
| 1423 |
+
"loss": 0.2961,
|
| 1424 |
"step": 1000
|
| 1425 |
},
|
| 1426 |
{
|
| 1427 |
"epoch": 1.3565559932470457,
|
| 1428 |
+
"grad_norm": 0.04672219098205047,
|
| 1429 |
"learning_rate": 2.822995162049599e-06,
|
| 1430 |
+
"loss": 0.3618,
|
| 1431 |
"step": 1005
|
| 1432 |
},
|
| 1433 |
{
|
| 1434 |
"epoch": 1.3633089476646032,
|
| 1435 |
+
"grad_norm": 0.04260321337685024,
|
| 1436 |
"learning_rate": 2.7700663715746213e-06,
|
| 1437 |
+
"loss": 0.3382,
|
| 1438 |
"step": 1010
|
| 1439 |
},
|
| 1440 |
{
|
| 1441 |
"epoch": 1.370061902082161,
|
| 1442 |
+
"grad_norm": 0.048008697667558865,
|
| 1443 |
"learning_rate": 2.7174476928399685e-06,
|
| 1444 |
+
"loss": 0.3466,
|
| 1445 |
"step": 1015
|
| 1446 |
},
|
| 1447 |
{
|
| 1448 |
"epoch": 1.3768148564997187,
|
| 1449 |
+
"grad_norm": 0.04892067375449484,
|
| 1450 |
"learning_rate": 2.66514644340426e-06,
|
| 1451 |
"loss": 0.3425,
|
| 1452 |
"step": 1020
|
| 1453 |
},
|
| 1454 |
{
|
| 1455 |
"epoch": 1.3835678109172762,
|
| 1456 |
+
"grad_norm": 0.04689882121063575,
|
| 1457 |
"learning_rate": 2.613169896681949e-06,
|
| 1458 |
+
"loss": 0.2877,
|
| 1459 |
"step": 1025
|
| 1460 |
},
|
| 1461 |
{
|
| 1462 |
"epoch": 1.390320765334834,
|
| 1463 |
+
"grad_norm": 0.054157177491435074,
|
| 1464 |
"learning_rate": 2.5615252809318287e-06,
|
| 1465 |
+
"loss": 0.3306,
|
| 1466 |
"step": 1030
|
| 1467 |
},
|
| 1468 |
{
|
| 1469 |
"epoch": 1.3970737197523917,
|
| 1470 |
+
"grad_norm": 0.048672772164928886,
|
| 1471 |
"learning_rate": 2.5102197782518145e-06,
|
| 1472 |
+
"loss": 0.3216,
|
| 1473 |
"step": 1035
|
| 1474 |
},
|
| 1475 |
{
|
| 1476 |
"epoch": 1.4038266741699492,
|
| 1477 |
+
"grad_norm": 0.06752970884761282,
|
| 1478 |
"learning_rate": 2.4592605235801544e-06,
|
| 1479 |
+
"loss": 0.3276,
|
| 1480 |
"step": 1040
|
| 1481 |
},
|
| 1482 |
{
|
| 1483 |
"epoch": 1.410579628587507,
|
| 1484 |
+
"grad_norm": 0.04710574868399225,
|
| 1485 |
"learning_rate": 2.4086546037031734e-06,
|
| 1486 |
+
"loss": 0.3173,
|
| 1487 |
"step": 1045
|
| 1488 |
},
|
| 1489 |
{
|
| 1490 |
"epoch": 1.4173325830050647,
|
| 1491 |
+
"grad_norm": 0.045184769590470804,
|
| 1492 |
"learning_rate": 2.3584090562697427e-06,
|
| 1493 |
+
"loss": 0.3164,
|
| 1494 |
"step": 1050
|
| 1495 |
},
|
| 1496 |
{
|
| 1497 |
"epoch": 1.4240855374226224,
|
| 1498 |
+
"grad_norm": 0.043485542556231396,
|
| 1499 |
"learning_rate": 2.3085308688125695e-06,
|
| 1500 |
+
"loss": 0.3027,
|
| 1501 |
"step": 1055
|
| 1502 |
},
|
| 1503 |
{
|
| 1504 |
"epoch": 1.4308384918401802,
|
| 1505 |
+
"grad_norm": 0.05260603355819636,
|
| 1506 |
"learning_rate": 2.2590269777764516e-06,
|
| 1507 |
+
"loss": 0.3004,
|
| 1508 |
"step": 1060
|
| 1509 |
},
|
| 1510 |
{
|
| 1511 |
"epoch": 1.4375914462577377,
|
| 1512 |
+
"grad_norm": 0.051099932583330926,
|
| 1513 |
"learning_rate": 2.2099042675536437e-06,
|
| 1514 |
+
"loss": 0.3368,
|
| 1515 |
"step": 1065
|
| 1516 |
},
|
| 1517 |
{
|
| 1518 |
"epoch": 1.4443444006752955,
|
| 1519 |
+
"grad_norm": 0.045328406677686464,
|
| 1520 |
"learning_rate": 2.161169569526461e-06,
|
| 1521 |
+
"loss": 0.3056,
|
| 1522 |
"step": 1070
|
| 1523 |
},
|
| 1524 |
{
|
| 1525 |
"epoch": 1.4510973550928532,
|
| 1526 |
+
"grad_norm": 0.05676490778567245,
|
| 1527 |
"learning_rate": 2.1128296611172593e-06,
|
| 1528 |
+
"loss": 0.3052,
|
| 1529 |
"step": 1075
|
| 1530 |
},
|
| 1531 |
{
|
| 1532 |
"epoch": 1.4578503095104107,
|
| 1533 |
+
"grad_norm": 0.044797358674842205,
|
| 1534 |
"learning_rate": 2.0648912648459072e-06,
|
| 1535 |
"loss": 0.3447,
|
| 1536 |
"step": 1080
|
| 1537 |
},
|
| 1538 |
{
|
| 1539 |
"epoch": 1.4646032639279685,
|
| 1540 |
+
"grad_norm": 0.049289994897252776,
|
| 1541 |
"learning_rate": 2.0173610473949048e-06,
|
| 1542 |
+
"loss": 0.3375,
|
| 1543 |
"step": 1085
|
| 1544 |
},
|
| 1545 |
{
|
| 1546 |
"epoch": 1.4713562183455262,
|
| 1547 |
+
"grad_norm": 0.04706434397429944,
|
| 1548 |
"learning_rate": 1.9702456186822595e-06,
|
| 1549 |
+
"loss": 0.3464,
|
| 1550 |
"step": 1090
|
| 1551 |
},
|
| 1552 |
{
|
| 1553 |
"epoch": 1.4781091727630837,
|
| 1554 |
+
"grad_norm": 0.05030965485900401,
|
| 1555 |
"learning_rate": 1.9235515309422685e-06,
|
| 1556 |
+
"loss": 0.3119,
|
| 1557 |
"step": 1095
|
| 1558 |
},
|
| 1559 |
{
|
| 1560 |
"epoch": 1.4848621271806415,
|
| 1561 |
+
"grad_norm": 0.04413503044032754,
|
| 1562 |
"learning_rate": 1.8772852778143064e-06,
|
| 1563 |
+
"loss": 0.3371,
|
| 1564 |
"step": 1100
|
| 1565 |
},
|
| 1566 |
{
|
| 1567 |
"epoch": 1.4916150815981992,
|
| 1568 |
+
"grad_norm": 0.06233556882470439,
|
| 1569 |
"learning_rate": 1.831453293439771e-06,
|
| 1570 |
+
"loss": 0.3253,
|
| 1571 |
"step": 1105
|
| 1572 |
},
|
| 1573 |
{
|
| 1574 |
"epoch": 1.498368036015757,
|
| 1575 |
+
"grad_norm": 0.05282578176334435,
|
| 1576 |
"learning_rate": 1.7860619515673034e-06,
|
| 1577 |
+
"loss": 0.3032,
|
| 1578 |
"step": 1110
|
| 1579 |
},
|
| 1580 |
{
|
| 1581 |
"epoch": 1.5051209904333147,
|
| 1582 |
+
"grad_norm": 0.05220647521633855,
|
| 1583 |
"learning_rate": 1.7411175646664103e-06,
|
| 1584 |
+
"loss": 0.3161,
|
| 1585 |
"step": 1115
|
| 1586 |
},
|
| 1587 |
{
|
| 1588 |
"epoch": 1.5118739448508722,
|
| 1589 |
+
"grad_norm": 0.047487304887968124,
|
| 1590 |
"learning_rate": 1.6966263830495939e-06,
|
| 1591 |
+
"loss": 0.3104,
|
| 1592 |
"step": 1120
|
| 1593 |
},
|
| 1594 |
{
|
| 1595 |
"epoch": 1.51862689926843,
|
| 1596 |
+
"grad_norm": 0.05098969364978786,
|
| 1597 |
"learning_rate": 1.6525945940031407e-06,
|
| 1598 |
"loss": 0.3692,
|
| 1599 |
"step": 1125
|
| 1600 |
},
|
| 1601 |
{
|
| 1602 |
"epoch": 1.5253798536859877,
|
| 1603 |
+
"grad_norm": 0.04810126110395747,
|
| 1604 |
"learning_rate": 1.6090283209266682e-06,
|
| 1605 |
+
"loss": 0.3383,
|
| 1606 |
"step": 1130
|
| 1607 |
},
|
| 1608 |
{
|
| 1609 |
"epoch": 1.5321328081035452,
|
| 1610 |
+
"grad_norm": 0.06656533903224827,
|
| 1611 |
"learning_rate": 1.5659336224815642e-06,
|
| 1612 |
+
"loss": 0.3413,
|
| 1613 |
"step": 1135
|
| 1614 |
},
|
| 1615 |
{
|
| 1616 |
"epoch": 1.538885762521103,
|
| 1617 |
+
"grad_norm": 0.057091145465475585,
|
| 1618 |
"learning_rate": 1.5233164917484117e-06,
|
| 1619 |
+
"loss": 0.3247,
|
| 1620 |
"step": 1140
|
| 1621 |
},
|
| 1622 |
{
|
| 1623 |
"epoch": 1.5456387169386607,
|
| 1624 |
+
"grad_norm": 0.05098329100385167,
|
| 1625 |
"learning_rate": 1.4811828553935498e-06,
|
| 1626 |
+
"loss": 0.3284,
|
| 1627 |
"step": 1145
|
| 1628 |
},
|
| 1629 |
{
|
| 1630 |
"epoch": 1.5523916713562183,
|
| 1631 |
+
"grad_norm": 0.04852633539753427,
|
| 1632 |
"learning_rate": 1.439538572844873e-06,
|
| 1633 |
+
"loss": 0.307,
|
| 1634 |
"step": 1150
|
| 1635 |
},
|
| 1636 |
{
|
| 1637 |
"epoch": 1.559144625773776,
|
| 1638 |
+
"grad_norm": 0.05935932625720792,
|
| 1639 |
"learning_rate": 1.3983894354769616e-06,
|
| 1640 |
+
"loss": 0.3247,
|
| 1641 |
"step": 1155
|
| 1642 |
},
|
| 1643 |
{
|
| 1644 |
"epoch": 1.5658975801913337,
|
| 1645 |
+
"grad_norm": 0.06306295104638487,
|
| 1646 |
"learning_rate": 1.3577411658056965e-06,
|
| 1647 |
+
"loss": 0.3286,
|
| 1648 |
"step": 1160
|
| 1649 |
},
|
| 1650 |
{
|
| 1651 |
"epoch": 1.5726505346088913,
|
| 1652 |
+
"grad_norm": 0.06348638606124773,
|
| 1653 |
"learning_rate": 1.3175994166924394e-06,
|
| 1654 |
+
"loss": 0.3247,
|
| 1655 |
"step": 1165
|
| 1656 |
},
|
| 1657 |
{
|
| 1658 |
"epoch": 1.5794034890264492,
|
| 1659 |
+
"grad_norm": 0.051811719215914646,
|
| 1660 |
"learning_rate": 1.2779697705579058e-06,
|
| 1661 |
+
"loss": 0.3246,
|
| 1662 |
"step": 1170
|
| 1663 |
},
|
| 1664 |
{
|
| 1665 |
"epoch": 1.5861564434440067,
|
| 1666 |
+
"grad_norm": 0.06830322751813911,
|
| 1667 |
"learning_rate": 1.2388577386058248e-06,
|
| 1668 |
+
"loss": 0.3326,
|
| 1669 |
"step": 1175
|
| 1670 |
},
|
| 1671 |
{
|
| 1672 |
"epoch": 1.5929093978615643,
|
| 1673 |
+
"grad_norm": 0.04922857249304829,
|
| 1674 |
"learning_rate": 1.2002687600565138e-06,
|
| 1675 |
+
"loss": 0.3061,
|
| 1676 |
"step": 1180
|
| 1677 |
},
|
| 1678 |
{
|
| 1679 |
"epoch": 1.5996623522791222,
|
| 1680 |
+
"grad_norm": 0.047108708806732846,
|
| 1681 |
"learning_rate": 1.1622082013904535e-06,
|
| 1682 |
+
"loss": 0.3118,
|
| 1683 |
"step": 1185
|
| 1684 |
},
|
| 1685 |
{
|
| 1686 |
"epoch": 1.6064153066966798,
|
| 1687 |
+
"grad_norm": 0.059687932246599115,
|
| 1688 |
"learning_rate": 1.1246813556019925e-06,
|
| 1689 |
+
"loss": 0.3372,
|
| 1690 |
"step": 1190
|
| 1691 |
},
|
| 1692 |
{
|
| 1693 |
"epoch": 1.6131682611142375,
|
| 1694 |
+
"grad_norm": 0.05955615759336213,
|
| 1695 |
"learning_rate": 1.0876934414632523e-06,
|
| 1696 |
+
"loss": 0.2956,
|
| 1697 |
"step": 1195
|
| 1698 |
},
|
| 1699 |
{
|
| 1700 |
"epoch": 1.6199212155317952,
|
| 1701 |
+
"grad_norm": 0.06542544159973171,
|
| 1702 |
"learning_rate": 1.0512496027983715e-06,
|
| 1703 |
+
"loss": 0.3118,
|
| 1704 |
"step": 1200
|
| 1705 |
},
|
| 1706 |
{
|
| 1707 |
"epoch": 1.6266741699493528,
|
| 1708 |
+
"grad_norm": 0.05198203822638065,
|
| 1709 |
"learning_rate": 1.0153549077681617e-06,
|
| 1710 |
+
"loss": 0.346,
|
| 1711 |
"step": 1205
|
| 1712 |
},
|
| 1713 |
{
|
| 1714 |
"epoch": 1.6334271243669105,
|
| 1715 |
+
"grad_norm": 0.06101598768362396,
|
| 1716 |
"learning_rate": 9.80014348165298e-07,
|
| 1717 |
+
"loss": 0.2952,
|
| 1718 |
"step": 1210
|
| 1719 |
},
|
| 1720 |
{
|
| 1721 |
"epoch": 1.6401800787844683,
|
| 1722 |
+
"grad_norm": 0.05552589226769097,
|
| 1723 |
"learning_rate": 9.452328387201104e-07,
|
| 1724 |
+
"loss": 0.2903,
|
| 1725 |
"step": 1215
|
| 1726 |
},
|
| 1727 |
{
|
| 1728 |
"epoch": 1.6469330332020258,
|
| 1729 |
+
"grad_norm": 0.06472026824610987,
|
| 1730 |
"learning_rate": 9.110152164171127e-07,
|
| 1731 |
+
"loss": 0.3081,
|
| 1732 |
"step": 1220
|
| 1733 |
},
|
| 1734 |
{
|
| 1735 |
"epoch": 1.6536859876195835,
|
| 1736 |
+
"grad_norm": 0.07431179400574185,
|
| 1737 |
"learning_rate": 8.773662398223276e-07,
|
| 1738 |
+
"loss": 0.3095,
|
| 1739 |
"step": 1225
|
| 1740 |
},
|
| 1741 |
{
|
| 1742 |
"epoch": 1.6604389420371413,
|
| 1743 |
+
"grad_norm": 0.06752269915658678,
|
| 1744 |
"learning_rate": 8.44290588421533e-07,
|
| 1745 |
+
"loss": 0.3187,
|
| 1746 |
"step": 1230
|
| 1747 |
},
|
| 1748 |
{
|
| 1749 |
"epoch": 1.6671918964546988,
|
| 1750 |
+
"grad_norm": 0.06578494925550707,
|
| 1751 |
"learning_rate": 8.117928619694848e-07,
|
| 1752 |
+
"loss": 0.3036,
|
| 1753 |
"step": 1235
|
| 1754 |
},
|
| 1755 |
{
|
| 1756 |
"epoch": 1.6739448508722568,
|
| 1757 |
+
"grad_norm": 0.06180051968265024,
|
| 1758 |
"learning_rate": 7.798775798502484e-07,
|
| 1759 |
+
"loss": 0.3023,
|
| 1760 |
"step": 1240
|
| 1761 |
},
|
| 1762 |
{
|
| 1763 |
"epoch": 1.6806978052898143,
|
| 1764 |
+
"grad_norm": 0.07560845712778133,
|
| 1765 |
"learning_rate": 7.485491804486972e-07,
|
| 1766 |
"loss": 0.2951,
|
| 1767 |
"step": 1245
|
| 1768 |
},
|
| 1769 |
{
|
| 1770 |
"epoch": 1.687450759707372,
|
| 1771 |
+
"grad_norm": 0.07194010404747912,
|
| 1772 |
"learning_rate": 7.178120205332717e-07,
|
| 1773 |
+
"loss": 0.3372,
|
| 1774 |
"step": 1250
|
| 1775 |
},
|
| 1776 |
{
|
| 1777 |
"epoch": 1.6942037141249298,
|
| 1778 |
+
"grad_norm": 0.06976609157173444,
|
| 1779 |
"learning_rate": 6.876703746500984e-07,
|
| 1780 |
+
"loss": 0.2973,
|
| 1781 |
"step": 1255
|
| 1782 |
},
|
| 1783 |
{
|
| 1784 |
"epoch": 1.7009566685424873,
|
| 1785 |
+
"grad_norm": 0.07659553993307627,
|
| 1786 |
"learning_rate": 6.581284345285371e-07,
|
| 1787 |
+
"loss": 0.284,
|
| 1788 |
"step": 1260
|
| 1789 |
},
|
| 1790 |
{
|
| 1791 |
"epoch": 1.707709622960045,
|
| 1792 |
+
"grad_norm": 0.05961865699755965,
|
| 1793 |
"learning_rate": 6.291903084982481e-07,
|
| 1794 |
+
"loss": 0.3153,
|
| 1795 |
"step": 1265
|
| 1796 |
},
|
| 1797 |
{
|
| 1798 |
"epoch": 1.7144625773776028,
|
| 1799 |
+
"grad_norm": 0.0812482207068001,
|
| 1800 |
"learning_rate": 6.008600209178539e-07,
|
| 1801 |
+
"loss": 0.342,
|
| 1802 |
"step": 1270
|
| 1803 |
},
|
| 1804 |
{
|
| 1805 |
"epoch": 1.7212155317951603,
|
| 1806 |
+
"grad_norm": 0.06939464438812673,
|
| 1807 |
"learning_rate": 5.7314151161528e-07,
|
| 1808 |
+
"loss": 0.3287,
|
| 1809 |
"step": 1275
|
| 1810 |
},
|
| 1811 |
{
|
| 1812 |
"epoch": 1.727968486212718,
|
| 1813 |
+
"grad_norm": 0.06615067862128174,
|
| 1814 |
"learning_rate": 5.460386353398583e-07,
|
| 1815 |
+
"loss": 0.3036,
|
| 1816 |
"step": 1280
|
| 1817 |
},
|
| 1818 |
{
|
| 1819 |
"epoch": 1.7347214406302758,
|
| 1820 |
+
"grad_norm": 0.08487351817752944,
|
| 1821 |
"learning_rate": 5.195551612262478e-07,
|
| 1822 |
+
"loss": 0.3059,
|
| 1823 |
"step": 1285
|
| 1824 |
},
|
| 1825 |
{
|
| 1826 |
"epoch": 1.7414743950478333,
|
| 1827 |
+
"grad_norm": 0.0660084040820942,
|
| 1828 |
"learning_rate": 4.936947722702762e-07,
|
| 1829 |
+
"loss": 0.3004,
|
| 1830 |
"step": 1290
|
| 1831 |
},
|
| 1832 |
{
|
| 1833 |
"epoch": 1.7482273494653913,
|
| 1834 |
+
"grad_norm": 0.07500490253176417,
|
| 1835 |
"learning_rate": 4.6846106481675035e-07,
|
| 1836 |
+
"loss": 0.3163,
|
| 1837 |
"step": 1295
|
| 1838 |
},
|
| 1839 |
{
|
| 1840 |
"epoch": 1.7549803038829488,
|
| 1841 |
+
"grad_norm": 0.07341914462851476,
|
| 1842 |
"learning_rate": 4.43857548059321e-07,
|
| 1843 |
+
"loss": 0.3028,
|
| 1844 |
"step": 1300
|
| 1845 |
},
|
| 1846 |
{
|
| 1847 |
"epoch": 1.7617332583005063,
|
| 1848 |
+
"grad_norm": 0.05936547346570651,
|
| 1849 |
"learning_rate": 4.198876435524718e-07,
|
| 1850 |
+
"loss": 0.3071,
|
| 1851 |
"step": 1305
|
| 1852 |
},
|
| 1853 |
{
|
| 1854 |
"epoch": 1.7684862127180643,
|
| 1855 |
+
"grad_norm": 0.06361743408706552,
|
| 1856 |
"learning_rate": 3.9655468473568436e-07,
|
| 1857 |
+
"loss": 0.3151,
|
| 1858 |
"step": 1310
|
| 1859 |
},
|
| 1860 |
{
|
| 1861 |
"epoch": 1.7752391671356218,
|
| 1862 |
+
"grad_norm": 0.07650389023350329,
|
| 1863 |
"learning_rate": 3.7386191646987094e-07,
|
| 1864 |
+
"loss": 0.333,
|
| 1865 |
"step": 1315
|
| 1866 |
},
|
| 1867 |
{
|
| 1868 |
"epoch": 1.7819921215531795,
|
| 1869 |
+
"grad_norm": 0.06885824668728105,
|
| 1870 |
"learning_rate": 3.51812494586114e-07,
|
| 1871 |
+
"loss": 0.2898,
|
| 1872 |
"step": 1320
|
| 1873 |
},
|
| 1874 |
{
|
| 1875 |
"epoch": 1.7887450759707373,
|
| 1876 |
+
"grad_norm": 0.0645999509118337,
|
| 1877 |
"learning_rate": 3.3040948544679817e-07,
|
| 1878 |
+
"loss": 0.3091,
|
| 1879 |
"step": 1325
|
| 1880 |
},
|
| 1881 |
{
|
| 1882 |
"epoch": 1.7954980303882948,
|
| 1883 |
+
"grad_norm": 0.07732866321608588,
|
| 1884 |
"learning_rate": 3.096558655191706e-07,
|
| 1885 |
+
"loss": 0.3148,
|
| 1886 |
"step": 1330
|
| 1887 |
},
|
| 1888 |
{
|
| 1889 |
"epoch": 1.8022509848058526,
|
| 1890 |
+
"grad_norm": 0.0733331982280887,
|
| 1891 |
"learning_rate": 2.895545209614176e-07,
|
| 1892 |
+
"loss": 0.3101,
|
| 1893 |
"step": 1335
|
| 1894 |
},
|
| 1895 |
{
|
| 1896 |
"epoch": 1.8090039392234103,
|
| 1897 |
+
"grad_norm": 0.06997792694817997,
|
| 1898 |
"learning_rate": 2.701082472212879e-07,
|
| 1899 |
+
"loss": 0.3302,
|
| 1900 |
"step": 1340
|
| 1901 |
},
|
| 1902 |
{
|
| 1903 |
"epoch": 1.8157568936409678,
|
| 1904 |
+
"grad_norm": 0.07837574138979145,
|
| 1905 |
"learning_rate": 2.5131974864734063e-07,
|
| 1906 |
"loss": 0.3425,
|
| 1907 |
"step": 1345
|
| 1908 |
},
|
| 1909 |
{
|
| 1910 |
"epoch": 1.8225098480585256,
|
| 1911 |
+
"grad_norm": 0.07239388554199513,
|
| 1912 |
"learning_rate": 2.331916381128535e-07,
|
| 1913 |
+
"loss": 0.3054,
|
| 1914 |
"step": 1350
|
| 1915 |
},
|
| 1916 |
{
|
| 1917 |
"epoch": 1.8292628024760833,
|
| 1918 |
+
"grad_norm": 0.0674988666661417,
|
| 1919 |
"learning_rate": 2.157264366524603e-07,
|
| 1920 |
+
"loss": 0.28,
|
| 1921 |
"step": 1355
|
| 1922 |
},
|
| 1923 |
{
|
| 1924 |
"epoch": 1.8360157568936408,
|
| 1925 |
+
"grad_norm": 0.07181164697688498,
|
| 1926 |
"learning_rate": 1.989265731115525e-07,
|
| 1927 |
+
"loss": 0.3307,
|
| 1928 |
"step": 1360
|
| 1929 |
},
|
| 1930 |
{
|
| 1931 |
"epoch": 1.8427687113111988,
|
| 1932 |
+
"grad_norm": 0.07548202922443094,
|
| 1933 |
"learning_rate": 1.827943838085111e-07,
|
| 1934 |
"loss": 0.306,
|
| 1935 |
"step": 1365
|
| 1936 |
},
|
| 1937 |
{
|
| 1938 |
"epoch": 1.8495216657287563,
|
| 1939 |
+
"grad_norm": 0.05982075375718766,
|
| 1940 |
"learning_rate": 1.6733211220979316e-07,
|
| 1941 |
+
"loss": 0.2981,
|
| 1942 |
"step": 1370
|
| 1943 |
},
|
| 1944 |
{
|
| 1945 |
"epoch": 1.856274620146314,
|
| 1946 |
+
"grad_norm": 0.08355451454560252,
|
| 1947 |
"learning_rate": 1.5254190861794415e-07,
|
| 1948 |
+
"loss": 0.3083,
|
| 1949 |
"step": 1375
|
| 1950 |
},
|
| 1951 |
{
|
| 1952 |
"epoch": 1.8630275745638718,
|
| 1953 |
+
"grad_norm": 0.07033662219869818,
|
| 1954 |
"learning_rate": 1.3842582987255494e-07,
|
| 1955 |
+
"loss": 0.293,
|
| 1956 |
"step": 1380
|
| 1957 |
},
|
| 1958 |
{
|
| 1959 |
"epoch": 1.8697805289814293,
|
| 1960 |
+
"grad_norm": 0.06615537508166416,
|
| 1961 |
"learning_rate": 1.2498583906422779e-07,
|
| 1962 |
+
"loss": 0.3072,
|
| 1963 |
"step": 1385
|
| 1964 |
},
|
| 1965 |
{
|
| 1966 |
"epoch": 1.876533483398987,
|
| 1967 |
+
"grad_norm": 0.059237664909188886,
|
| 1968 |
"learning_rate": 1.1222380526156929e-07,
|
| 1969 |
+
"loss": 0.313,
|
| 1970 |
"step": 1390
|
| 1971 |
},
|
| 1972 |
{
|
| 1973 |
"epoch": 1.8832864378165448,
|
| 1974 |
+
"grad_norm": 0.08185957993390121,
|
| 1975 |
"learning_rate": 1.0014150325126315e-07,
|
| 1976 |
+
"loss": 0.3071,
|
| 1977 |
"step": 1395
|
| 1978 |
},
|
| 1979 |
{
|
| 1980 |
"epoch": 1.8900393922341023,
|
| 1981 |
+
"grad_norm": 0.06113218790264024,
|
| 1982 |
"learning_rate": 8.874061329125939e-08,
|
| 1983 |
+
"loss": 0.3116,
|
| 1984 |
"step": 1400
|
| 1985 |
},
|
| 1986 |
{
|
| 1987 |
"epoch": 1.89679234665166,
|
| 1988 |
+
"grad_norm": 0.08593985138096456,
|
| 1989 |
"learning_rate": 7.802272087709951e-08,
|
| 1990 |
+
"loss": 0.3106,
|
| 1991 |
"step": 1405
|
| 1992 |
},
|
| 1993 |
{
|
| 1994 |
"epoch": 1.9035453010692178,
|
| 1995 |
+
"grad_norm": 0.0753339474248732,
|
| 1996 |
"learning_rate": 6.798931652142738e-08,
|
| 1997 |
+
"loss": 0.3118,
|
| 1998 |
"step": 1410
|
| 1999 |
},
|
| 2000 |
{
|
| 2001 |
"epoch": 1.9102982554867753,
|
| 2002 |
+
"grad_norm": 0.06406340322576629,
|
| 2003 |
"learning_rate": 5.864179554670724e-08,
|
| 2004 |
+
"loss": 0.3384,
|
| 2005 |
"step": 1415
|
| 2006 |
},
|
| 2007 |
{
|
| 2008 |
"epoch": 1.9170512099043333,
|
| 2009 |
+
"grad_norm": 0.06739406393286214,
|
| 2010 |
"learning_rate": 4.998145789118114e-08,
|
| 2011 |
+
"loss": 0.2755,
|
| 2012 |
"step": 1420
|
| 2013 |
},
|
| 2014 |
{
|
| 2015 |
"epoch": 1.9238041643218908,
|
| 2016 |
+
"grad_norm": 0.07101405306951455,
|
| 2017 |
"learning_rate": 4.2009507928084116e-08,
|
| 2018 |
+
"loss": 0.2802,
|
| 2019 |
"step": 1425
|
| 2020 |
},
|
| 2021 |
{
|
| 2022 |
"epoch": 1.9305571187394484,
|
| 2023 |
+
"grad_norm": 0.07509273619426268,
|
| 2024 |
"learning_rate": 3.4727054298161475e-08,
|
| 2025 |
+
"loss": 0.2778,
|
| 2026 |
"step": 1430
|
| 2027 |
},
|
| 2028 |
{
|
| 2029 |
"epoch": 1.9373100731570063,
|
| 2030 |
+
"grad_norm": 0.0719074660592959,
|
| 2031 |
"learning_rate": 2.8135109755487723e-08,
|
| 2032 |
+
"loss": 0.3238,
|
| 2033 |
"step": 1435
|
| 2034 |
},
|
| 2035 |
{
|
| 2036 |
"epoch": 1.9440630275745638,
|
| 2037 |
+
"grad_norm": 0.07042162359075328,
|
| 2038 |
"learning_rate": 2.223459102662695e-08,
|
| 2039 |
+
"loss": 0.3235,
|
| 2040 |
"step": 1440
|
| 2041 |
},
|
| 2042 |
{
|
| 2043 |
"epoch": 1.9508159819921216,
|
| 2044 |
+
"grad_norm": 0.07906944364159181,
|
| 2045 |
"learning_rate": 1.7026318683147082e-08,
|
| 2046 |
+
"loss": 0.3115,
|
| 2047 |
"step": 1445
|
| 2048 |
},
|
| 2049 |
{
|
| 2050 |
"epoch": 1.9575689364096793,
|
| 2051 |
+
"grad_norm": 0.08613249297162148,
|
| 2052 |
"learning_rate": 1.2511017027501682e-08,
|
| 2053 |
"loss": 0.3032,
|
| 2054 |
"step": 1450
|
| 2055 |
},
|
| 2056 |
{
|
| 2057 |
"epoch": 1.9643218908272368,
|
| 2058 |
+
"grad_norm": 0.07379471104805907,
|
| 2059 |
"learning_rate": 8.689313992306104e-09,
|
| 2060 |
+
"loss": 0.303,
|
| 2061 |
"step": 1455
|
| 2062 |
},
|
| 2063 |
{
|
| 2064 |
"epoch": 1.9710748452447946,
|
| 2065 |
+
"grad_norm": 0.07085802534414035,
|
| 2066 |
"learning_rate": 5.561741053010661e-09,
|
| 2067 |
+
"loss": 0.3059,
|
| 2068 |
"step": 1460
|
| 2069 |
},
|
| 2070 |
{
|
| 2071 |
"epoch": 1.9778277996623523,
|
| 2072 |
+
"grad_norm": 0.07455102650353135,
|
| 2073 |
"learning_rate": 3.1287331539903155e-09,
|
| 2074 |
+
"loss": 0.3165,
|
| 2075 |
"step": 1465
|
| 2076 |
},
|
| 2077 |
{
|
| 2078 |
"epoch": 1.9845807540799099,
|
| 2079 |
+
"grad_norm": 0.07699115693937987,
|
| 2080 |
"learning_rate": 1.3906286480563913e-09,
|
| 2081 |
+
"loss": 0.3108,
|
| 2082 |
"step": 1470
|
| 2083 |
},
|
| 2084 |
{
|
| 2085 |
"epoch": 1.9913337084974676,
|
| 2086 |
+
"grad_norm": 0.06672257472875329,
|
| 2087 |
"learning_rate": 3.4766924940476954e-10,
|
| 2088 |
+
"loss": 0.3134,
|
| 2089 |
"step": 1475
|
| 2090 |
},
|
| 2091 |
{
|
| 2092 |
"epoch": 1.9980866629150253,
|
| 2093 |
+
"grad_norm": 0.06196469052553269,
|
| 2094 |
"learning_rate": 0.0,
|
| 2095 |
+
"loss": 0.3054,
|
| 2096 |
"step": 1480
|
| 2097 |
},
|
| 2098 |
{
|
| 2099 |
"epoch": 1.9980866629150253,
|
| 2100 |
+
"eval_loss": 0.3086739182472229,
|
| 2101 |
+
"eval_runtime": 17.7246,
|
| 2102 |
+
"eval_samples_per_second": 20.254,
|
| 2103 |
+
"eval_steps_per_second": 5.078,
|
| 2104 |
"step": 1480
|
| 2105 |
},
|
| 2106 |
{
|
| 2107 |
"epoch": 1.9980866629150253,
|
| 2108 |
"step": 1480,
|
| 2109 |
"total_flos": 3.6719988787579453e+18,
|
| 2110 |
+
"train_loss": 0.37467364362768224,
|
| 2111 |
+
"train_runtime": 13260.7751,
|
| 2112 |
+
"train_samples_per_second": 5.36,
|
| 2113 |
"train_steps_per_second": 0.112
|
| 2114 |
}
|
| 2115 |
],
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 7288
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6f6a7799de28c93d654b3137724722a60a4634ceedd183783e265a4c550d87c3
|
| 3 |
size 7288
|