Upload 11 files
Browse files- adapter_config.json +3 -3
- adapter_model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- tokenizer.json +6 -1
- trainer_state.json +196 -2744
- training_args.bin +1 -1
adapter_config.json
CHANGED
|
@@ -21,11 +21,11 @@
|
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
"fc1",
|
| 24 |
-
"dense",
|
| 25 |
"q_proj",
|
| 26 |
-
"v_proj",
|
| 27 |
"fc2",
|
| 28 |
-
"
|
|
|
|
|
|
|
| 29 |
],
|
| 30 |
"task_type": "CAUSAL_LM",
|
| 31 |
"use_dora": false,
|
|
|
|
| 21 |
"revision": null,
|
| 22 |
"target_modules": [
|
| 23 |
"fc1",
|
|
|
|
| 24 |
"q_proj",
|
|
|
|
| 25 |
"fc2",
|
| 26 |
+
"dense",
|
| 27 |
+
"k_proj",
|
| 28 |
+
"v_proj"
|
| 29 |
],
|
| 30 |
"task_type": "CAUSAL_LM",
|
| 31 |
"use_dora": false,
|
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 37774528
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0ed1c2ec7bba2c3a47f4604e98ca91feaae4ae45eeeb381c0f377bff0b4dbe7c
|
| 3 |
size 37774528
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 75611578
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:910dc9e1ee19b05944cf3c1f6c5db091e12c734a1586ef6e956ddd847bdbe150
|
| 3 |
size 75611578
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14244
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8caf4be913ba0c7cfadaae796e6d56eb5e1583e0b176f7995b566b543235b55
|
| 3 |
size 14244
|
tokenizer.json
CHANGED
|
@@ -1,6 +1,11 @@
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
-
"truncation":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
"padding": null,
|
| 5 |
"added_tokens": [
|
| 6 |
{
|
|
|
|
| 1 |
{
|
| 2 |
"version": "1.0",
|
| 3 |
+
"truncation": {
|
| 4 |
+
"direction": "Right",
|
| 5 |
+
"max_length": 2048,
|
| 6 |
+
"strategy": "LongestFirst",
|
| 7 |
+
"stride": 0
|
| 8 |
+
},
|
| 9 |
"padding": null,
|
| 10 |
"added_tokens": [
|
| 11 |
{
|
trainer_state.json
CHANGED
|
@@ -8,3193 +8,645 @@
|
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
-
{
|
| 12 |
-
"epoch": 0.01098901098901099,
|
| 13 |
-
"grad_norm": 3.238452911376953,
|
| 14 |
-
"learning_rate": 0.00019956043956043958,
|
| 15 |
-
"loss": 1.7274,
|
| 16 |
-
"step": 100
|
| 17 |
-
},
|
| 18 |
-
{
|
| 19 |
-
"epoch": 0.02197802197802198,
|
| 20 |
-
"grad_norm": 2.8424575328826904,
|
| 21 |
-
"learning_rate": 0.00019912087912087913,
|
| 22 |
-
"loss": 1.7205,
|
| 23 |
-
"step": 200
|
| 24 |
-
},
|
| 25 |
-
{
|
| 26 |
-
"epoch": 0.03296703296703297,
|
| 27 |
-
"grad_norm": 1.5604631900787354,
|
| 28 |
-
"learning_rate": 0.0001986813186813187,
|
| 29 |
-
"loss": 1.6773,
|
| 30 |
-
"step": 300
|
| 31 |
-
},
|
| 32 |
-
{
|
| 33 |
-
"epoch": 0.04395604395604396,
|
| 34 |
-
"grad_norm": 2.338499069213867,
|
| 35 |
-
"learning_rate": 0.00019824175824175827,
|
| 36 |
-
"loss": 1.7067,
|
| 37 |
-
"step": 400
|
| 38 |
-
},
|
| 39 |
{
|
| 40 |
"epoch": 0.054945054945054944,
|
| 41 |
-
"grad_norm":
|
| 42 |
"learning_rate": 0.0001978021978021978,
|
| 43 |
-
"loss": 1.
|
| 44 |
"step": 500
|
| 45 |
},
|
| 46 |
-
{
|
| 47 |
-
"epoch": 0.06593406593406594,
|
| 48 |
-
"grad_norm": 2.4433257579803467,
|
| 49 |
-
"learning_rate": 0.00019736263736263738,
|
| 50 |
-
"loss": 1.7072,
|
| 51 |
-
"step": 600
|
| 52 |
-
},
|
| 53 |
-
{
|
| 54 |
-
"epoch": 0.07692307692307693,
|
| 55 |
-
"grad_norm": 2.889246702194214,
|
| 56 |
-
"learning_rate": 0.00019692307692307696,
|
| 57 |
-
"loss": 1.7447,
|
| 58 |
-
"step": 700
|
| 59 |
-
},
|
| 60 |
-
{
|
| 61 |
-
"epoch": 0.08791208791208792,
|
| 62 |
-
"grad_norm": 3.244313955307007,
|
| 63 |
-
"learning_rate": 0.00019648351648351647,
|
| 64 |
-
"loss": 1.6999,
|
| 65 |
-
"step": 800
|
| 66 |
-
},
|
| 67 |
-
{
|
| 68 |
-
"epoch": 0.0989010989010989,
|
| 69 |
-
"grad_norm": 2.247797727584839,
|
| 70 |
-
"learning_rate": 0.00019604395604395604,
|
| 71 |
-
"loss": 1.7091,
|
| 72 |
-
"step": 900
|
| 73 |
-
},
|
| 74 |
{
|
| 75 |
"epoch": 0.10989010989010989,
|
| 76 |
-
"grad_norm":
|
| 77 |
"learning_rate": 0.00019560439560439562,
|
| 78 |
-
"loss": 1.
|
| 79 |
"step": 1000
|
| 80 |
},
|
| 81 |
-
{
|
| 82 |
-
"epoch": 0.12087912087912088,
|
| 83 |
-
"grad_norm": 4.7427167892456055,
|
| 84 |
-
"learning_rate": 0.00019516483516483516,
|
| 85 |
-
"loss": 1.7151,
|
| 86 |
-
"step": 1100
|
| 87 |
-
},
|
| 88 |
-
{
|
| 89 |
-
"epoch": 0.13186813186813187,
|
| 90 |
-
"grad_norm": 2.6960134506225586,
|
| 91 |
-
"learning_rate": 0.00019472527472527473,
|
| 92 |
-
"loss": 1.6853,
|
| 93 |
-
"step": 1200
|
| 94 |
-
},
|
| 95 |
-
{
|
| 96 |
-
"epoch": 0.14285714285714285,
|
| 97 |
-
"grad_norm": 3.6055240631103516,
|
| 98 |
-
"learning_rate": 0.0001942857142857143,
|
| 99 |
-
"loss": 1.6763,
|
| 100 |
-
"step": 1300
|
| 101 |
-
},
|
| 102 |
-
{
|
| 103 |
-
"epoch": 0.15384615384615385,
|
| 104 |
-
"grad_norm": 2.970062494277954,
|
| 105 |
-
"learning_rate": 0.00019384615384615385,
|
| 106 |
-
"loss": 1.7136,
|
| 107 |
-
"step": 1400
|
| 108 |
-
},
|
| 109 |
{
|
| 110 |
"epoch": 0.16483516483516483,
|
| 111 |
-
"grad_norm":
|
| 112 |
"learning_rate": 0.00019340659340659342,
|
| 113 |
-
"loss": 1.
|
| 114 |
"step": 1500
|
| 115 |
},
|
| 116 |
-
{
|
| 117 |
-
"epoch": 0.17582417582417584,
|
| 118 |
-
"grad_norm": 3.0747196674346924,
|
| 119 |
-
"learning_rate": 0.000192967032967033,
|
| 120 |
-
"loss": 1.7001,
|
| 121 |
-
"step": 1600
|
| 122 |
-
},
|
| 123 |
-
{
|
| 124 |
-
"epoch": 0.18681318681318682,
|
| 125 |
-
"grad_norm": 2.819571018218994,
|
| 126 |
-
"learning_rate": 0.00019252747252747254,
|
| 127 |
-
"loss": 1.6939,
|
| 128 |
-
"step": 1700
|
| 129 |
-
},
|
| 130 |
-
{
|
| 131 |
-
"epoch": 0.1978021978021978,
|
| 132 |
-
"grad_norm": 3.701860189437866,
|
| 133 |
-
"learning_rate": 0.0001920879120879121,
|
| 134 |
-
"loss": 1.6922,
|
| 135 |
-
"step": 1800
|
| 136 |
-
},
|
| 137 |
-
{
|
| 138 |
-
"epoch": 0.2087912087912088,
|
| 139 |
-
"grad_norm": 2.7427585124969482,
|
| 140 |
-
"learning_rate": 0.00019164835164835168,
|
| 141 |
-
"loss": 1.717,
|
| 142 |
-
"step": 1900
|
| 143 |
-
},
|
| 144 |
{
|
| 145 |
"epoch": 0.21978021978021978,
|
| 146 |
-
"grad_norm":
|
| 147 |
"learning_rate": 0.00019120879120879122,
|
| 148 |
-
"loss": 1.
|
| 149 |
"step": 2000
|
| 150 |
},
|
| 151 |
-
{
|
| 152 |
-
"epoch": 0.23076923076923078,
|
| 153 |
-
"grad_norm": 3.920029640197754,
|
| 154 |
-
"learning_rate": 0.0001907692307692308,
|
| 155 |
-
"loss": 1.7139,
|
| 156 |
-
"step": 2100
|
| 157 |
-
},
|
| 158 |
-
{
|
| 159 |
-
"epoch": 0.24175824175824176,
|
| 160 |
-
"grad_norm": 2.9013421535491943,
|
| 161 |
-
"learning_rate": 0.00019032967032967034,
|
| 162 |
-
"loss": 1.6954,
|
| 163 |
-
"step": 2200
|
| 164 |
-
},
|
| 165 |
-
{
|
| 166 |
-
"epoch": 0.25274725274725274,
|
| 167 |
-
"grad_norm": 2.9385688304901123,
|
| 168 |
-
"learning_rate": 0.00018989010989010988,
|
| 169 |
-
"loss": 1.7187,
|
| 170 |
-
"step": 2300
|
| 171 |
-
},
|
| 172 |
-
{
|
| 173 |
-
"epoch": 0.26373626373626374,
|
| 174 |
-
"grad_norm": 3.105120897293091,
|
| 175 |
-
"learning_rate": 0.00018945054945054945,
|
| 176 |
-
"loss": 1.6687,
|
| 177 |
-
"step": 2400
|
| 178 |
-
},
|
| 179 |
{
|
| 180 |
"epoch": 0.27472527472527475,
|
| 181 |
-
"grad_norm":
|
| 182 |
"learning_rate": 0.00018901098901098903,
|
| 183 |
-
"loss": 1.
|
| 184 |
"step": 2500
|
| 185 |
},
|
| 186 |
-
{
|
| 187 |
-
"epoch": 0.2857142857142857,
|
| 188 |
-
"grad_norm": 2.239258289337158,
|
| 189 |
-
"learning_rate": 0.00018857142857142857,
|
| 190 |
-
"loss": 1.6609,
|
| 191 |
-
"step": 2600
|
| 192 |
-
},
|
| 193 |
-
{
|
| 194 |
-
"epoch": 0.2967032967032967,
|
| 195 |
-
"grad_norm": 3.5539562702178955,
|
| 196 |
-
"learning_rate": 0.00018813186813186814,
|
| 197 |
-
"loss": 1.6778,
|
| 198 |
-
"step": 2700
|
| 199 |
-
},
|
| 200 |
-
{
|
| 201 |
-
"epoch": 0.3076923076923077,
|
| 202 |
-
"grad_norm": 3.821715831756592,
|
| 203 |
-
"learning_rate": 0.0001876923076923077,
|
| 204 |
-
"loss": 1.6924,
|
| 205 |
-
"step": 2800
|
| 206 |
-
},
|
| 207 |
-
{
|
| 208 |
-
"epoch": 0.31868131868131866,
|
| 209 |
-
"grad_norm": 2.939919948577881,
|
| 210 |
-
"learning_rate": 0.00018725274725274726,
|
| 211 |
-
"loss": 1.7103,
|
| 212 |
-
"step": 2900
|
| 213 |
-
},
|
| 214 |
{
|
| 215 |
"epoch": 0.32967032967032966,
|
| 216 |
-
"grad_norm":
|
| 217 |
"learning_rate": 0.00018681318681318683,
|
| 218 |
-
"loss": 1.
|
| 219 |
"step": 3000
|
| 220 |
},
|
| 221 |
-
{
|
| 222 |
-
"epoch": 0.34065934065934067,
|
| 223 |
-
"grad_norm": 5.75909423828125,
|
| 224 |
-
"learning_rate": 0.00018637362637362637,
|
| 225 |
-
"loss": 1.6928,
|
| 226 |
-
"step": 3100
|
| 227 |
-
},
|
| 228 |
-
{
|
| 229 |
-
"epoch": 0.3516483516483517,
|
| 230 |
-
"grad_norm": 4.105484485626221,
|
| 231 |
-
"learning_rate": 0.00018593406593406595,
|
| 232 |
-
"loss": 1.692,
|
| 233 |
-
"step": 3200
|
| 234 |
-
},
|
| 235 |
-
{
|
| 236 |
-
"epoch": 0.3626373626373626,
|
| 237 |
-
"grad_norm": 2.9766411781311035,
|
| 238 |
-
"learning_rate": 0.00018549450549450552,
|
| 239 |
-
"loss": 1.6781,
|
| 240 |
-
"step": 3300
|
| 241 |
-
},
|
| 242 |
-
{
|
| 243 |
-
"epoch": 0.37362637362637363,
|
| 244 |
-
"grad_norm": 3.47971510887146,
|
| 245 |
-
"learning_rate": 0.00018505494505494506,
|
| 246 |
-
"loss": 1.6668,
|
| 247 |
-
"step": 3400
|
| 248 |
-
},
|
| 249 |
{
|
| 250 |
"epoch": 0.38461538461538464,
|
| 251 |
-
"grad_norm":
|
| 252 |
"learning_rate": 0.00018461538461538463,
|
| 253 |
-
"loss": 1.
|
| 254 |
"step": 3500
|
| 255 |
},
|
| 256 |
-
{
|
| 257 |
-
"epoch": 0.3956043956043956,
|
| 258 |
-
"grad_norm": 2.745067834854126,
|
| 259 |
-
"learning_rate": 0.00018417582417582418,
|
| 260 |
-
"loss": 1.6739,
|
| 261 |
-
"step": 3600
|
| 262 |
-
},
|
| 263 |
-
{
|
| 264 |
-
"epoch": 0.4065934065934066,
|
| 265 |
-
"grad_norm": 4.118983268737793,
|
| 266 |
-
"learning_rate": 0.00018373626373626375,
|
| 267 |
-
"loss": 1.6597,
|
| 268 |
-
"step": 3700
|
| 269 |
-
},
|
| 270 |
-
{
|
| 271 |
-
"epoch": 0.4175824175824176,
|
| 272 |
-
"grad_norm": 3.8697848320007324,
|
| 273 |
-
"learning_rate": 0.0001832967032967033,
|
| 274 |
-
"loss": 1.6729,
|
| 275 |
-
"step": 3800
|
| 276 |
-
},
|
| 277 |
-
{
|
| 278 |
-
"epoch": 0.42857142857142855,
|
| 279 |
-
"grad_norm": 7.602184772491455,
|
| 280 |
-
"learning_rate": 0.00018285714285714286,
|
| 281 |
-
"loss": 1.6892,
|
| 282 |
-
"step": 3900
|
| 283 |
-
},
|
| 284 |
{
|
| 285 |
"epoch": 0.43956043956043955,
|
| 286 |
-
"grad_norm":
|
| 287 |
"learning_rate": 0.0001824175824175824,
|
| 288 |
-
"loss": 1.
|
| 289 |
"step": 4000
|
| 290 |
},
|
| 291 |
-
{
|
| 292 |
-
"epoch": 0.45054945054945056,
|
| 293 |
-
"grad_norm": 3.064992904663086,
|
| 294 |
-
"learning_rate": 0.00018197802197802198,
|
| 295 |
-
"loss": 1.6786,
|
| 296 |
-
"step": 4100
|
| 297 |
-
},
|
| 298 |
-
{
|
| 299 |
-
"epoch": 0.46153846153846156,
|
| 300 |
-
"grad_norm": 3.288447141647339,
|
| 301 |
-
"learning_rate": 0.00018153846153846155,
|
| 302 |
-
"loss": 1.6594,
|
| 303 |
-
"step": 4200
|
| 304 |
-
},
|
| 305 |
-
{
|
| 306 |
-
"epoch": 0.4725274725274725,
|
| 307 |
-
"grad_norm": 3.2647299766540527,
|
| 308 |
-
"learning_rate": 0.0001810989010989011,
|
| 309 |
-
"loss": 1.6593,
|
| 310 |
-
"step": 4300
|
| 311 |
-
},
|
| 312 |
-
{
|
| 313 |
-
"epoch": 0.4835164835164835,
|
| 314 |
-
"grad_norm": 2.6522679328918457,
|
| 315 |
-
"learning_rate": 0.00018065934065934067,
|
| 316 |
-
"loss": 1.6682,
|
| 317 |
-
"step": 4400
|
| 318 |
-
},
|
| 319 |
{
|
| 320 |
"epoch": 0.4945054945054945,
|
| 321 |
-
"grad_norm":
|
| 322 |
"learning_rate": 0.00018021978021978024,
|
| 323 |
-
"loss": 1.
|
| 324 |
"step": 4500
|
| 325 |
},
|
| 326 |
-
{
|
| 327 |
-
"epoch": 0.5054945054945055,
|
| 328 |
-
"grad_norm": 5.3127264976501465,
|
| 329 |
-
"learning_rate": 0.00017978021978021978,
|
| 330 |
-
"loss": 1.683,
|
| 331 |
-
"step": 4600
|
| 332 |
-
},
|
| 333 |
-
{
|
| 334 |
-
"epoch": 0.5164835164835165,
|
| 335 |
-
"grad_norm": 4.50441837310791,
|
| 336 |
-
"learning_rate": 0.00017934065934065936,
|
| 337 |
-
"loss": 1.6701,
|
| 338 |
-
"step": 4700
|
| 339 |
-
},
|
| 340 |
-
{
|
| 341 |
-
"epoch": 0.5274725274725275,
|
| 342 |
-
"grad_norm": 3.104093074798584,
|
| 343 |
-
"learning_rate": 0.00017890109890109893,
|
| 344 |
-
"loss": 1.6509,
|
| 345 |
-
"step": 4800
|
| 346 |
-
},
|
| 347 |
-
{
|
| 348 |
-
"epoch": 0.5384615384615384,
|
| 349 |
-
"grad_norm": 2.652597427368164,
|
| 350 |
-
"learning_rate": 0.00017846153846153847,
|
| 351 |
-
"loss": 1.6523,
|
| 352 |
-
"step": 4900
|
| 353 |
-
},
|
| 354 |
{
|
| 355 |
"epoch": 0.5494505494505495,
|
| 356 |
-
"grad_norm":
|
| 357 |
"learning_rate": 0.00017802197802197802,
|
| 358 |
-
"loss": 1.
|
| 359 |
"step": 5000
|
| 360 |
},
|
| 361 |
-
{
|
| 362 |
-
"epoch": 0.5604395604395604,
|
| 363 |
-
"grad_norm": 2.976457357406616,
|
| 364 |
-
"learning_rate": 0.0001775824175824176,
|
| 365 |
-
"loss": 1.6843,
|
| 366 |
-
"step": 5100
|
| 367 |
-
},
|
| 368 |
-
{
|
| 369 |
-
"epoch": 0.5714285714285714,
|
| 370 |
-
"grad_norm": 3.4616947174072266,
|
| 371 |
-
"learning_rate": 0.00017714285714285713,
|
| 372 |
-
"loss": 1.6989,
|
| 373 |
-
"step": 5200
|
| 374 |
-
},
|
| 375 |
-
{
|
| 376 |
-
"epoch": 0.5824175824175825,
|
| 377 |
-
"grad_norm": 3.0407989025115967,
|
| 378 |
-
"learning_rate": 0.0001767032967032967,
|
| 379 |
-
"loss": 1.6601,
|
| 380 |
-
"step": 5300
|
| 381 |
-
},
|
| 382 |
-
{
|
| 383 |
-
"epoch": 0.5934065934065934,
|
| 384 |
-
"grad_norm": 3.5591070652008057,
|
| 385 |
-
"learning_rate": 0.00017626373626373627,
|
| 386 |
-
"loss": 1.6641,
|
| 387 |
-
"step": 5400
|
| 388 |
-
},
|
| 389 |
{
|
| 390 |
"epoch": 0.6043956043956044,
|
| 391 |
-
"grad_norm":
|
| 392 |
"learning_rate": 0.00017582417582417582,
|
| 393 |
-
"loss": 1.
|
| 394 |
"step": 5500
|
| 395 |
},
|
| 396 |
-
{
|
| 397 |
-
"epoch": 0.6153846153846154,
|
| 398 |
-
"grad_norm": 5.115685939788818,
|
| 399 |
-
"learning_rate": 0.0001753846153846154,
|
| 400 |
-
"loss": 1.6617,
|
| 401 |
-
"step": 5600
|
| 402 |
-
},
|
| 403 |
-
{
|
| 404 |
-
"epoch": 0.6263736263736264,
|
| 405 |
-
"grad_norm": 4.381234645843506,
|
| 406 |
-
"learning_rate": 0.00017494505494505496,
|
| 407 |
-
"loss": 1.6844,
|
| 408 |
-
"step": 5700
|
| 409 |
-
},
|
| 410 |
-
{
|
| 411 |
-
"epoch": 0.6373626373626373,
|
| 412 |
-
"grad_norm": 3.331096887588501,
|
| 413 |
-
"learning_rate": 0.0001745054945054945,
|
| 414 |
-
"loss": 1.6842,
|
| 415 |
-
"step": 5800
|
| 416 |
-
},
|
| 417 |
-
{
|
| 418 |
-
"epoch": 0.6483516483516484,
|
| 419 |
-
"grad_norm": 4.683435916900635,
|
| 420 |
-
"learning_rate": 0.00017406593406593408,
|
| 421 |
-
"loss": 1.6583,
|
| 422 |
-
"step": 5900
|
| 423 |
-
},
|
| 424 |
{
|
| 425 |
"epoch": 0.6593406593406593,
|
| 426 |
-
"grad_norm":
|
| 427 |
"learning_rate": 0.00017362637362637365,
|
| 428 |
-
"loss": 1.
|
| 429 |
"step": 6000
|
| 430 |
},
|
| 431 |
-
{
|
| 432 |
-
"epoch": 0.6703296703296703,
|
| 433 |
-
"grad_norm": 3.5134270191192627,
|
| 434 |
-
"learning_rate": 0.0001731868131868132,
|
| 435 |
-
"loss": 1.6623,
|
| 436 |
-
"step": 6100
|
| 437 |
-
},
|
| 438 |
-
{
|
| 439 |
-
"epoch": 0.6813186813186813,
|
| 440 |
-
"grad_norm": 3.438363790512085,
|
| 441 |
-
"learning_rate": 0.00017274725274725276,
|
| 442 |
-
"loss": 1.6458,
|
| 443 |
-
"step": 6200
|
| 444 |
-
},
|
| 445 |
-
{
|
| 446 |
-
"epoch": 0.6923076923076923,
|
| 447 |
-
"grad_norm": 3.1804699897766113,
|
| 448 |
-
"learning_rate": 0.00017230769230769234,
|
| 449 |
-
"loss": 1.7014,
|
| 450 |
-
"step": 6300
|
| 451 |
-
},
|
| 452 |
-
{
|
| 453 |
-
"epoch": 0.7032967032967034,
|
| 454 |
-
"grad_norm": 2.5642647743225098,
|
| 455 |
-
"learning_rate": 0.00017186813186813185,
|
| 456 |
-
"loss": 1.6356,
|
| 457 |
-
"step": 6400
|
| 458 |
-
},
|
| 459 |
{
|
| 460 |
"epoch": 0.7142857142857143,
|
| 461 |
-
"grad_norm":
|
| 462 |
"learning_rate": 0.00017142857142857143,
|
| 463 |
-
"loss": 1.
|
| 464 |
"step": 6500
|
| 465 |
},
|
| 466 |
-
{
|
| 467 |
-
"epoch": 0.7252747252747253,
|
| 468 |
-
"grad_norm": 2.9326508045196533,
|
| 469 |
-
"learning_rate": 0.000170989010989011,
|
| 470 |
-
"loss": 1.6406,
|
| 471 |
-
"step": 6600
|
| 472 |
-
},
|
| 473 |
-
{
|
| 474 |
-
"epoch": 0.7362637362637363,
|
| 475 |
-
"grad_norm": 3.912691593170166,
|
| 476 |
-
"learning_rate": 0.00017054945054945054,
|
| 477 |
-
"loss": 1.6604,
|
| 478 |
-
"step": 6700
|
| 479 |
-
},
|
| 480 |
-
{
|
| 481 |
-
"epoch": 0.7472527472527473,
|
| 482 |
-
"grad_norm": 4.329016208648682,
|
| 483 |
-
"learning_rate": 0.0001701098901098901,
|
| 484 |
-
"loss": 1.7025,
|
| 485 |
-
"step": 6800
|
| 486 |
-
},
|
| 487 |
-
{
|
| 488 |
-
"epoch": 0.7582417582417582,
|
| 489 |
-
"grad_norm": 4.046297073364258,
|
| 490 |
-
"learning_rate": 0.00016967032967032968,
|
| 491 |
-
"loss": 1.6578,
|
| 492 |
-
"step": 6900
|
| 493 |
-
},
|
| 494 |
{
|
| 495 |
"epoch": 0.7692307692307693,
|
| 496 |
-
"grad_norm":
|
| 497 |
"learning_rate": 0.00016923076923076923,
|
| 498 |
-
"loss": 1.
|
| 499 |
"step": 7000
|
| 500 |
},
|
| 501 |
-
{
|
| 502 |
-
"epoch": 0.7802197802197802,
|
| 503 |
-
"grad_norm": 4.125152587890625,
|
| 504 |
-
"learning_rate": 0.0001687912087912088,
|
| 505 |
-
"loss": 1.6808,
|
| 506 |
-
"step": 7100
|
| 507 |
-
},
|
| 508 |
-
{
|
| 509 |
-
"epoch": 0.7912087912087912,
|
| 510 |
-
"grad_norm": 3.0072073936462402,
|
| 511 |
-
"learning_rate": 0.00016835164835164837,
|
| 512 |
-
"loss": 1.6935,
|
| 513 |
-
"step": 7200
|
| 514 |
-
},
|
| 515 |
-
{
|
| 516 |
-
"epoch": 0.8021978021978022,
|
| 517 |
-
"grad_norm": 3.7964279651641846,
|
| 518 |
-
"learning_rate": 0.00016791208791208792,
|
| 519 |
-
"loss": 1.6772,
|
| 520 |
-
"step": 7300
|
| 521 |
-
},
|
| 522 |
-
{
|
| 523 |
-
"epoch": 0.8131868131868132,
|
| 524 |
-
"grad_norm": 4.463648796081543,
|
| 525 |
-
"learning_rate": 0.0001674725274725275,
|
| 526 |
-
"loss": 1.6704,
|
| 527 |
-
"step": 7400
|
| 528 |
-
},
|
| 529 |
{
|
| 530 |
"epoch": 0.8241758241758241,
|
| 531 |
-
"grad_norm":
|
| 532 |
"learning_rate": 0.00016703296703296706,
|
| 533 |
-
"loss": 1.
|
| 534 |
"step": 7500
|
| 535 |
},
|
| 536 |
-
{
|
| 537 |
-
"epoch": 0.8351648351648352,
|
| 538 |
-
"grad_norm": 4.4003214836120605,
|
| 539 |
-
"learning_rate": 0.0001665934065934066,
|
| 540 |
-
"loss": 1.6491,
|
| 541 |
-
"step": 7600
|
| 542 |
-
},
|
| 543 |
-
{
|
| 544 |
-
"epoch": 0.8461538461538461,
|
| 545 |
-
"grad_norm": 2.7502541542053223,
|
| 546 |
-
"learning_rate": 0.00016615384615384617,
|
| 547 |
-
"loss": 1.6882,
|
| 548 |
-
"step": 7700
|
| 549 |
-
},
|
| 550 |
-
{
|
| 551 |
-
"epoch": 0.8571428571428571,
|
| 552 |
-
"grad_norm": 3.120934247970581,
|
| 553 |
-
"learning_rate": 0.00016571428571428575,
|
| 554 |
-
"loss": 1.6457,
|
| 555 |
-
"step": 7800
|
| 556 |
-
},
|
| 557 |
-
{
|
| 558 |
-
"epoch": 0.8681318681318682,
|
| 559 |
-
"grad_norm": 4.110687732696533,
|
| 560 |
-
"learning_rate": 0.00016527472527472526,
|
| 561 |
-
"loss": 1.6817,
|
| 562 |
-
"step": 7900
|
| 563 |
-
},
|
| 564 |
{
|
| 565 |
"epoch": 0.8791208791208791,
|
| 566 |
-
"grad_norm":
|
| 567 |
"learning_rate": 0.00016483516483516484,
|
| 568 |
-
"loss": 1.
|
| 569 |
"step": 8000
|
| 570 |
},
|
| 571 |
-
{
|
| 572 |
-
"epoch": 0.8901098901098901,
|
| 573 |
-
"grad_norm": 3.775438070297241,
|
| 574 |
-
"learning_rate": 0.0001643956043956044,
|
| 575 |
-
"loss": 1.6811,
|
| 576 |
-
"step": 8100
|
| 577 |
-
},
|
| 578 |
-
{
|
| 579 |
-
"epoch": 0.9010989010989011,
|
| 580 |
-
"grad_norm": 3.679263114929199,
|
| 581 |
-
"learning_rate": 0.00016395604395604395,
|
| 582 |
-
"loss": 1.6827,
|
| 583 |
-
"step": 8200
|
| 584 |
-
},
|
| 585 |
-
{
|
| 586 |
-
"epoch": 0.9120879120879121,
|
| 587 |
-
"grad_norm": 3.706937313079834,
|
| 588 |
-
"learning_rate": 0.00016351648351648352,
|
| 589 |
-
"loss": 1.6554,
|
| 590 |
-
"step": 8300
|
| 591 |
-
},
|
| 592 |
-
{
|
| 593 |
-
"epoch": 0.9230769230769231,
|
| 594 |
-
"grad_norm": 3.64091420173645,
|
| 595 |
-
"learning_rate": 0.0001630769230769231,
|
| 596 |
-
"loss": 1.6496,
|
| 597 |
-
"step": 8400
|
| 598 |
-
},
|
| 599 |
{
|
| 600 |
"epoch": 0.9340659340659341,
|
| 601 |
-
"grad_norm":
|
| 602 |
"learning_rate": 0.00016263736263736264,
|
| 603 |
-
"loss": 1.
|
| 604 |
"step": 8500
|
| 605 |
},
|
| 606 |
-
{
|
| 607 |
-
"epoch": 0.945054945054945,
|
| 608 |
-
"grad_norm": 3.5718305110931396,
|
| 609 |
-
"learning_rate": 0.0001621978021978022,
|
| 610 |
-
"loss": 1.6597,
|
| 611 |
-
"step": 8600
|
| 612 |
-
},
|
| 613 |
-
{
|
| 614 |
-
"epoch": 0.9560439560439561,
|
| 615 |
-
"grad_norm": 3.4999473094940186,
|
| 616 |
-
"learning_rate": 0.00016175824175824178,
|
| 617 |
-
"loss": 1.6563,
|
| 618 |
-
"step": 8700
|
| 619 |
-
},
|
| 620 |
-
{
|
| 621 |
-
"epoch": 0.967032967032967,
|
| 622 |
-
"grad_norm": 4.831845760345459,
|
| 623 |
-
"learning_rate": 0.00016131868131868133,
|
| 624 |
-
"loss": 1.645,
|
| 625 |
-
"step": 8800
|
| 626 |
-
},
|
| 627 |
-
{
|
| 628 |
-
"epoch": 0.978021978021978,
|
| 629 |
-
"grad_norm": 2.883336067199707,
|
| 630 |
-
"learning_rate": 0.0001608791208791209,
|
| 631 |
-
"loss": 1.6712,
|
| 632 |
-
"step": 8900
|
| 633 |
-
},
|
| 634 |
{
|
| 635 |
"epoch": 0.989010989010989,
|
| 636 |
-
"grad_norm":
|
| 637 |
"learning_rate": 0.00016043956043956044,
|
| 638 |
-
"loss": 1.
|
| 639 |
"step": 9000
|
| 640 |
},
|
| 641 |
-
{
|
| 642 |
-
"epoch": 1.0,
|
| 643 |
-
"grad_norm": 3.8828916549682617,
|
| 644 |
-
"learning_rate": 0.00016,
|
| 645 |
-
"loss": 1.6529,
|
| 646 |
-
"step": 9100
|
| 647 |
-
},
|
| 648 |
-
{
|
| 649 |
-
"epoch": 1.010989010989011,
|
| 650 |
-
"grad_norm": 3.094627857208252,
|
| 651 |
-
"learning_rate": 0.00015956043956043958,
|
| 652 |
-
"loss": 1.5436,
|
| 653 |
-
"step": 9200
|
| 654 |
-
},
|
| 655 |
-
{
|
| 656 |
-
"epoch": 1.021978021978022,
|
| 657 |
-
"grad_norm": 3.886928081512451,
|
| 658 |
-
"learning_rate": 0.00015912087912087913,
|
| 659 |
-
"loss": 1.5088,
|
| 660 |
-
"step": 9300
|
| 661 |
-
},
|
| 662 |
-
{
|
| 663 |
-
"epoch": 1.032967032967033,
|
| 664 |
-
"grad_norm": 2.9219202995300293,
|
| 665 |
-
"learning_rate": 0.00015868131868131867,
|
| 666 |
-
"loss": 1.567,
|
| 667 |
-
"step": 9400
|
| 668 |
-
},
|
| 669 |
{
|
| 670 |
"epoch": 1.043956043956044,
|
| 671 |
-
"grad_norm":
|
| 672 |
"learning_rate": 0.00015824175824175824,
|
| 673 |
-
"loss": 1.
|
| 674 |
"step": 9500
|
| 675 |
},
|
| 676 |
-
{
|
| 677 |
-
"epoch": 1.054945054945055,
|
| 678 |
-
"grad_norm": 3.9126474857330322,
|
| 679 |
-
"learning_rate": 0.00015780219780219782,
|
| 680 |
-
"loss": 1.5968,
|
| 681 |
-
"step": 9600
|
| 682 |
-
},
|
| 683 |
-
{
|
| 684 |
-
"epoch": 1.065934065934066,
|
| 685 |
-
"grad_norm": 4.429248332977295,
|
| 686 |
-
"learning_rate": 0.00015736263736263736,
|
| 687 |
-
"loss": 1.5518,
|
| 688 |
-
"step": 9700
|
| 689 |
-
},
|
| 690 |
-
{
|
| 691 |
-
"epoch": 1.0769230769230769,
|
| 692 |
-
"grad_norm": 3.1986796855926514,
|
| 693 |
-
"learning_rate": 0.00015692307692307693,
|
| 694 |
-
"loss": 1.566,
|
| 695 |
-
"step": 9800
|
| 696 |
-
},
|
| 697 |
-
{
|
| 698 |
-
"epoch": 1.0879120879120878,
|
| 699 |
-
"grad_norm": 7.024231910705566,
|
| 700 |
-
"learning_rate": 0.00015648351648351648,
|
| 701 |
-
"loss": 1.5857,
|
| 702 |
-
"step": 9900
|
| 703 |
-
},
|
| 704 |
{
|
| 705 |
"epoch": 1.098901098901099,
|
| 706 |
-
"grad_norm":
|
| 707 |
"learning_rate": 0.00015604395604395605,
|
| 708 |
-
"loss": 1.
|
| 709 |
"step": 10000
|
| 710 |
},
|
| 711 |
-
{
|
| 712 |
-
"epoch": 1.10989010989011,
|
| 713 |
-
"grad_norm": 4.216164588928223,
|
| 714 |
-
"learning_rate": 0.00015560439560439562,
|
| 715 |
-
"loss": 1.6087,
|
| 716 |
-
"step": 10100
|
| 717 |
-
},
|
| 718 |
-
{
|
| 719 |
-
"epoch": 1.120879120879121,
|
| 720 |
-
"grad_norm": 4.030162334442139,
|
| 721 |
-
"learning_rate": 0.00015516483516483516,
|
| 722 |
-
"loss": 1.5605,
|
| 723 |
-
"step": 10200
|
| 724 |
-
},
|
| 725 |
-
{
|
| 726 |
-
"epoch": 1.1318681318681318,
|
| 727 |
-
"grad_norm": 4.694043159484863,
|
| 728 |
-
"learning_rate": 0.00015472527472527474,
|
| 729 |
-
"loss": 1.5485,
|
| 730 |
-
"step": 10300
|
| 731 |
-
},
|
| 732 |
-
{
|
| 733 |
-
"epoch": 1.1428571428571428,
|
| 734 |
-
"grad_norm": 4.568836688995361,
|
| 735 |
-
"learning_rate": 0.0001542857142857143,
|
| 736 |
-
"loss": 1.5422,
|
| 737 |
-
"step": 10400
|
| 738 |
-
},
|
| 739 |
{
|
| 740 |
"epoch": 1.1538461538461537,
|
| 741 |
-
"grad_norm":
|
| 742 |
"learning_rate": 0.00015384615384615385,
|
| 743 |
-
"loss": 1.
|
| 744 |
"step": 10500
|
| 745 |
},
|
| 746 |
-
{
|
| 747 |
-
"epoch": 1.164835164835165,
|
| 748 |
-
"grad_norm": 5.639504432678223,
|
| 749 |
-
"learning_rate": 0.00015340659340659342,
|
| 750 |
-
"loss": 1.5752,
|
| 751 |
-
"step": 10600
|
| 752 |
-
},
|
| 753 |
-
{
|
| 754 |
-
"epoch": 1.1758241758241759,
|
| 755 |
-
"grad_norm": 4.555628299713135,
|
| 756 |
-
"learning_rate": 0.00015296703296703297,
|
| 757 |
-
"loss": 1.5716,
|
| 758 |
-
"step": 10700
|
| 759 |
-
},
|
| 760 |
-
{
|
| 761 |
-
"epoch": 1.1868131868131868,
|
| 762 |
-
"grad_norm": 4.831645488739014,
|
| 763 |
-
"learning_rate": 0.00015252747252747254,
|
| 764 |
-
"loss": 1.5895,
|
| 765 |
-
"step": 10800
|
| 766 |
-
},
|
| 767 |
-
{
|
| 768 |
-
"epoch": 1.1978021978021978,
|
| 769 |
-
"grad_norm": 5.440847873687744,
|
| 770 |
-
"learning_rate": 0.00015208791208791208,
|
| 771 |
-
"loss": 1.5499,
|
| 772 |
-
"step": 10900
|
| 773 |
-
},
|
| 774 |
{
|
| 775 |
"epoch": 1.2087912087912087,
|
| 776 |
-
"grad_norm":
|
| 777 |
"learning_rate": 0.00015164835164835165,
|
| 778 |
-
"loss": 1.
|
| 779 |
"step": 11000
|
| 780 |
},
|
| 781 |
-
{
|
| 782 |
-
"epoch": 1.2197802197802199,
|
| 783 |
-
"grad_norm": 4.695743560791016,
|
| 784 |
-
"learning_rate": 0.0001512087912087912,
|
| 785 |
-
"loss": 1.607,
|
| 786 |
-
"step": 11100
|
| 787 |
-
},
|
| 788 |
-
{
|
| 789 |
-
"epoch": 1.2307692307692308,
|
| 790 |
-
"grad_norm": 3.870124101638794,
|
| 791 |
-
"learning_rate": 0.00015076923076923077,
|
| 792 |
-
"loss": 1.5658,
|
| 793 |
-
"step": 11200
|
| 794 |
-
},
|
| 795 |
-
{
|
| 796 |
-
"epoch": 1.2417582417582418,
|
| 797 |
-
"grad_norm": 3.653447389602661,
|
| 798 |
-
"learning_rate": 0.00015032967032967034,
|
| 799 |
-
"loss": 1.5651,
|
| 800 |
-
"step": 11300
|
| 801 |
-
},
|
| 802 |
-
{
|
| 803 |
-
"epoch": 1.2527472527472527,
|
| 804 |
-
"grad_norm": 3.547433853149414,
|
| 805 |
-
"learning_rate": 0.0001498901098901099,
|
| 806 |
-
"loss": 1.5628,
|
| 807 |
-
"step": 11400
|
| 808 |
-
},
|
| 809 |
{
|
| 810 |
"epoch": 1.2637362637362637,
|
| 811 |
-
"grad_norm":
|
| 812 |
"learning_rate": 0.00014945054945054946,
|
| 813 |
-
"loss": 1.
|
| 814 |
"step": 11500
|
| 815 |
},
|
| 816 |
-
{
|
| 817 |
-
"epoch": 1.2747252747252746,
|
| 818 |
-
"grad_norm": 3.197800636291504,
|
| 819 |
-
"learning_rate": 0.00014901098901098903,
|
| 820 |
-
"loss": 1.5782,
|
| 821 |
-
"step": 11600
|
| 822 |
-
},
|
| 823 |
-
{
|
| 824 |
-
"epoch": 1.2857142857142856,
|
| 825 |
-
"grad_norm": 3.305222988128662,
|
| 826 |
-
"learning_rate": 0.00014857142857142857,
|
| 827 |
-
"loss": 1.5998,
|
| 828 |
-
"step": 11700
|
| 829 |
-
},
|
| 830 |
-
{
|
| 831 |
-
"epoch": 1.2967032967032968,
|
| 832 |
-
"grad_norm": 3.467773914337158,
|
| 833 |
-
"learning_rate": 0.00014813186813186815,
|
| 834 |
-
"loss": 1.5951,
|
| 835 |
-
"step": 11800
|
| 836 |
-
},
|
| 837 |
-
{
|
| 838 |
-
"epoch": 1.3076923076923077,
|
| 839 |
-
"grad_norm": 4.193416595458984,
|
| 840 |
-
"learning_rate": 0.00014769230769230772,
|
| 841 |
-
"loss": 1.592,
|
| 842 |
-
"step": 11900
|
| 843 |
-
},
|
| 844 |
{
|
| 845 |
"epoch": 1.3186813186813187,
|
| 846 |
-
"grad_norm":
|
| 847 |
"learning_rate": 0.00014725274725274726,
|
| 848 |
-
"loss": 1.
|
| 849 |
"step": 12000
|
| 850 |
},
|
| 851 |
-
{
|
| 852 |
-
"epoch": 1.3296703296703296,
|
| 853 |
-
"grad_norm": 3.4815001487731934,
|
| 854 |
-
"learning_rate": 0.0001468131868131868,
|
| 855 |
-
"loss": 1.5546,
|
| 856 |
-
"step": 12100
|
| 857 |
-
},
|
| 858 |
-
{
|
| 859 |
-
"epoch": 1.3406593406593408,
|
| 860 |
-
"grad_norm": 8.390545845031738,
|
| 861 |
-
"learning_rate": 0.00014637362637362638,
|
| 862 |
-
"loss": 1.5886,
|
| 863 |
-
"step": 12200
|
| 864 |
-
},
|
| 865 |
-
{
|
| 866 |
-
"epoch": 1.3516483516483517,
|
| 867 |
-
"grad_norm": 4.278797626495361,
|
| 868 |
-
"learning_rate": 0.00014593406593406592,
|
| 869 |
-
"loss": 1.5537,
|
| 870 |
-
"step": 12300
|
| 871 |
-
},
|
| 872 |
-
{
|
| 873 |
-
"epoch": 1.3626373626373627,
|
| 874 |
-
"grad_norm": 4.196005821228027,
|
| 875 |
-
"learning_rate": 0.0001454945054945055,
|
| 876 |
-
"loss": 1.5602,
|
| 877 |
-
"step": 12400
|
| 878 |
-
},
|
| 879 |
{
|
| 880 |
"epoch": 1.3736263736263736,
|
| 881 |
-
"grad_norm": 2.
|
| 882 |
"learning_rate": 0.00014505494505494506,
|
| 883 |
-
"loss": 1.
|
| 884 |
"step": 12500
|
| 885 |
},
|
| 886 |
-
{
|
| 887 |
-
"epoch": 1.3846153846153846,
|
| 888 |
-
"grad_norm": 3.346822500228882,
|
| 889 |
-
"learning_rate": 0.0001446153846153846,
|
| 890 |
-
"loss": 1.5658,
|
| 891 |
-
"step": 12600
|
| 892 |
-
},
|
| 893 |
-
{
|
| 894 |
-
"epoch": 1.3956043956043955,
|
| 895 |
-
"grad_norm": 4.593723773956299,
|
| 896 |
-
"learning_rate": 0.00014417582417582418,
|
| 897 |
-
"loss": 1.5846,
|
| 898 |
-
"step": 12700
|
| 899 |
-
},
|
| 900 |
-
{
|
| 901 |
-
"epoch": 1.4065934065934065,
|
| 902 |
-
"grad_norm": 4.338139057159424,
|
| 903 |
-
"learning_rate": 0.00014373626373626375,
|
| 904 |
-
"loss": 1.5682,
|
| 905 |
-
"step": 12800
|
| 906 |
-
},
|
| 907 |
-
{
|
| 908 |
-
"epoch": 1.4175824175824177,
|
| 909 |
-
"grad_norm": 4.005186557769775,
|
| 910 |
-
"learning_rate": 0.0001432967032967033,
|
| 911 |
-
"loss": 1.5411,
|
| 912 |
-
"step": 12900
|
| 913 |
-
},
|
| 914 |
{
|
| 915 |
"epoch": 1.4285714285714286,
|
| 916 |
-
"grad_norm":
|
| 917 |
"learning_rate": 0.00014285714285714287,
|
| 918 |
-
"loss": 1.
|
| 919 |
"step": 13000
|
| 920 |
},
|
| 921 |
-
{
|
| 922 |
-
"epoch": 1.4395604395604396,
|
| 923 |
-
"grad_norm": 3.864833354949951,
|
| 924 |
-
"learning_rate": 0.00014241758241758244,
|
| 925 |
-
"loss": 1.5794,
|
| 926 |
-
"step": 13100
|
| 927 |
-
},
|
| 928 |
-
{
|
| 929 |
-
"epoch": 1.4505494505494505,
|
| 930 |
-
"grad_norm": 5.973376750946045,
|
| 931 |
-
"learning_rate": 0.00014197802197802198,
|
| 932 |
-
"loss": 1.5585,
|
| 933 |
-
"step": 13200
|
| 934 |
-
},
|
| 935 |
-
{
|
| 936 |
-
"epoch": 1.4615384615384617,
|
| 937 |
-
"grad_norm": 4.784486770629883,
|
| 938 |
-
"learning_rate": 0.00014153846153846156,
|
| 939 |
-
"loss": 1.5396,
|
| 940 |
-
"step": 13300
|
| 941 |
-
},
|
| 942 |
-
{
|
| 943 |
-
"epoch": 1.4725274725274726,
|
| 944 |
-
"grad_norm": 3.6531691551208496,
|
| 945 |
-
"learning_rate": 0.00014109890109890113,
|
| 946 |
-
"loss": 1.5427,
|
| 947 |
-
"step": 13400
|
| 948 |
-
},
|
| 949 |
{
|
| 950 |
"epoch": 1.4835164835164836,
|
| 951 |
-
"grad_norm":
|
| 952 |
"learning_rate": 0.00014065934065934067,
|
| 953 |
-
"loss": 1.
|
| 954 |
"step": 13500
|
| 955 |
},
|
| 956 |
-
{
|
| 957 |
-
"epoch": 1.4945054945054945,
|
| 958 |
-
"grad_norm": 4.377908706665039,
|
| 959 |
-
"learning_rate": 0.00014021978021978022,
|
| 960 |
-
"loss": 1.5393,
|
| 961 |
-
"step": 13600
|
| 962 |
-
},
|
| 963 |
-
{
|
| 964 |
-
"epoch": 1.5054945054945055,
|
| 965 |
-
"grad_norm": 10.914487838745117,
|
| 966 |
-
"learning_rate": 0.0001397802197802198,
|
| 967 |
-
"loss": 1.564,
|
| 968 |
-
"step": 13700
|
| 969 |
-
},
|
| 970 |
-
{
|
| 971 |
-
"epoch": 1.5164835164835164,
|
| 972 |
-
"grad_norm": 5.559396743774414,
|
| 973 |
-
"learning_rate": 0.00013934065934065933,
|
| 974 |
-
"loss": 1.5925,
|
| 975 |
-
"step": 13800
|
| 976 |
-
},
|
| 977 |
-
{
|
| 978 |
-
"epoch": 1.5274725274725274,
|
| 979 |
-
"grad_norm": 4.534477710723877,
|
| 980 |
-
"learning_rate": 0.0001389010989010989,
|
| 981 |
-
"loss": 1.5753,
|
| 982 |
-
"step": 13900
|
| 983 |
-
},
|
| 984 |
{
|
| 985 |
"epoch": 1.5384615384615383,
|
| 986 |
-
"grad_norm":
|
| 987 |
"learning_rate": 0.00013846153846153847,
|
| 988 |
-
"loss": 1.
|
| 989 |
"step": 14000
|
| 990 |
},
|
| 991 |
-
{
|
| 992 |
-
"epoch": 1.5494505494505495,
|
| 993 |
-
"grad_norm": 5.009109020233154,
|
| 994 |
-
"learning_rate": 0.00013802197802197802,
|
| 995 |
-
"loss": 1.5559,
|
| 996 |
-
"step": 14100
|
| 997 |
-
},
|
| 998 |
-
{
|
| 999 |
-
"epoch": 1.5604395604395604,
|
| 1000 |
-
"grad_norm": 3.9765381813049316,
|
| 1001 |
-
"learning_rate": 0.0001375824175824176,
|
| 1002 |
-
"loss": 1.5458,
|
| 1003 |
-
"step": 14200
|
| 1004 |
-
},
|
| 1005 |
-
{
|
| 1006 |
-
"epoch": 1.5714285714285714,
|
| 1007 |
-
"grad_norm": 5.045168399810791,
|
| 1008 |
-
"learning_rate": 0.00013714285714285716,
|
| 1009 |
-
"loss": 1.5594,
|
| 1010 |
-
"step": 14300
|
| 1011 |
-
},
|
| 1012 |
-
{
|
| 1013 |
-
"epoch": 1.5824175824175826,
|
| 1014 |
-
"grad_norm": 5.065695285797119,
|
| 1015 |
-
"learning_rate": 0.0001367032967032967,
|
| 1016 |
-
"loss": 1.5523,
|
| 1017 |
-
"step": 14400
|
| 1018 |
-
},
|
| 1019 |
{
|
| 1020 |
"epoch": 1.5934065934065935,
|
| 1021 |
-
"grad_norm":
|
| 1022 |
"learning_rate": 0.00013626373626373628,
|
| 1023 |
-
"loss": 1.
|
| 1024 |
"step": 14500
|
| 1025 |
},
|
| 1026 |
-
{
|
| 1027 |
-
"epoch": 1.6043956043956045,
|
| 1028 |
-
"grad_norm": 3.814141273498535,
|
| 1029 |
-
"learning_rate": 0.00013582417582417585,
|
| 1030 |
-
"loss": 1.5937,
|
| 1031 |
-
"step": 14600
|
| 1032 |
-
},
|
| 1033 |
-
{
|
| 1034 |
-
"epoch": 1.6153846153846154,
|
| 1035 |
-
"grad_norm": 3.292379379272461,
|
| 1036 |
-
"learning_rate": 0.0001353846153846154,
|
| 1037 |
-
"loss": 1.5506,
|
| 1038 |
-
"step": 14700
|
| 1039 |
-
},
|
| 1040 |
-
{
|
| 1041 |
-
"epoch": 1.6263736263736264,
|
| 1042 |
-
"grad_norm": 3.32590389251709,
|
| 1043 |
-
"learning_rate": 0.00013494505494505497,
|
| 1044 |
-
"loss": 1.5501,
|
| 1045 |
-
"step": 14800
|
| 1046 |
-
},
|
| 1047 |
-
{
|
| 1048 |
-
"epoch": 1.6373626373626373,
|
| 1049 |
-
"grad_norm": 4.293301582336426,
|
| 1050 |
-
"learning_rate": 0.0001345054945054945,
|
| 1051 |
-
"loss": 1.5684,
|
| 1052 |
-
"step": 14900
|
| 1053 |
-
},
|
| 1054 |
{
|
| 1055 |
"epoch": 1.6483516483516483,
|
| 1056 |
-
"grad_norm":
|
| 1057 |
"learning_rate": 0.00013406593406593405,
|
| 1058 |
-
"loss": 1.
|
| 1059 |
"step": 15000
|
| 1060 |
},
|
| 1061 |
{
|
| 1062 |
-
"epoch": 1.
|
| 1063 |
-
"grad_norm":
|
| 1064 |
-
"learning_rate": 0.
|
| 1065 |
-
"loss": 1.
|
| 1066 |
-
"step":
|
| 1067 |
},
|
| 1068 |
{
|
| 1069 |
-
"epoch": 1.
|
| 1070 |
-
"grad_norm":
|
| 1071 |
-
"learning_rate": 0.
|
| 1072 |
-
"loss": 1.
|
| 1073 |
-
"step":
|
| 1074 |
},
|
| 1075 |
{
|
| 1076 |
-
"epoch": 1.
|
| 1077 |
-
"grad_norm":
|
| 1078 |
-
"learning_rate": 0.
|
| 1079 |
-
"loss": 1.
|
| 1080 |
-
"step":
|
| 1081 |
},
|
| 1082 |
{
|
| 1083 |
-
"epoch": 1.
|
| 1084 |
-
"grad_norm":
|
| 1085 |
-
"learning_rate": 0.
|
| 1086 |
-
"loss": 1.
|
| 1087 |
-
"step":
|
| 1088 |
-
},
|
| 1089 |
-
{
|
| 1090 |
-
"epoch": 1.7032967032967035,
|
| 1091 |
-
"grad_norm": 3.9485526084899902,
|
| 1092 |
-
"learning_rate": 0.00013186813186813188,
|
| 1093 |
-
"loss": 1.5811,
|
| 1094 |
-
"step": 15500
|
| 1095 |
-
},
|
| 1096 |
-
{
|
| 1097 |
-
"epoch": 1.7142857142857144,
|
| 1098 |
-
"grad_norm": 5.078463554382324,
|
| 1099 |
-
"learning_rate": 0.00013142857142857143,
|
| 1100 |
-
"loss": 1.5576,
|
| 1101 |
-
"step": 15600
|
| 1102 |
-
},
|
| 1103 |
-
{
|
| 1104 |
-
"epoch": 1.7252747252747254,
|
| 1105 |
-
"grad_norm": 3.7762668132781982,
|
| 1106 |
-
"learning_rate": 0.000130989010989011,
|
| 1107 |
-
"loss": 1.5502,
|
| 1108 |
-
"step": 15700
|
| 1109 |
-
},
|
| 1110 |
-
{
|
| 1111 |
-
"epoch": 1.7362637362637363,
|
| 1112 |
-
"grad_norm": 4.067686080932617,
|
| 1113 |
-
"learning_rate": 0.00013054945054945057,
|
| 1114 |
-
"loss": 1.5486,
|
| 1115 |
-
"step": 15800
|
| 1116 |
-
},
|
| 1117 |
-
{
|
| 1118 |
-
"epoch": 1.7472527472527473,
|
| 1119 |
-
"grad_norm": 3.324613094329834,
|
| 1120 |
-
"learning_rate": 0.00013010989010989012,
|
| 1121 |
-
"loss": 1.5786,
|
| 1122 |
-
"step": 15900
|
| 1123 |
-
},
|
| 1124 |
-
{
|
| 1125 |
-
"epoch": 1.7582417582417582,
|
| 1126 |
-
"grad_norm": 3.8743014335632324,
|
| 1127 |
-
"learning_rate": 0.0001296703296703297,
|
| 1128 |
-
"loss": 1.5341,
|
| 1129 |
-
"step": 16000
|
| 1130 |
-
},
|
| 1131 |
-
{
|
| 1132 |
-
"epoch": 1.7692307692307692,
|
| 1133 |
-
"grad_norm": 3.096376419067383,
|
| 1134 |
-
"learning_rate": 0.00012923076923076923,
|
| 1135 |
-
"loss": 1.5428,
|
| 1136 |
-
"step": 16100
|
| 1137 |
-
},
|
| 1138 |
-
{
|
| 1139 |
-
"epoch": 1.7802197802197801,
|
| 1140 |
-
"grad_norm": 5.382870197296143,
|
| 1141 |
-
"learning_rate": 0.0001287912087912088,
|
| 1142 |
-
"loss": 1.5364,
|
| 1143 |
-
"step": 16200
|
| 1144 |
-
},
|
| 1145 |
-
{
|
| 1146 |
-
"epoch": 1.791208791208791,
|
| 1147 |
-
"grad_norm": 7.159231185913086,
|
| 1148 |
-
"learning_rate": 0.00012835164835164837,
|
| 1149 |
-
"loss": 1.5559,
|
| 1150 |
-
"step": 16300
|
| 1151 |
-
},
|
| 1152 |
-
{
|
| 1153 |
-
"epoch": 1.8021978021978022,
|
| 1154 |
-
"grad_norm": 3.7771100997924805,
|
| 1155 |
-
"learning_rate": 0.00012791208791208792,
|
| 1156 |
-
"loss": 1.5576,
|
| 1157 |
-
"step": 16400
|
| 1158 |
-
},
|
| 1159 |
-
{
|
| 1160 |
-
"epoch": 1.8131868131868132,
|
| 1161 |
-
"grad_norm": 3.727796792984009,
|
| 1162 |
-
"learning_rate": 0.00012747252747252746,
|
| 1163 |
-
"loss": 1.5741,
|
| 1164 |
-
"step": 16500
|
| 1165 |
-
},
|
| 1166 |
-
{
|
| 1167 |
-
"epoch": 1.8241758241758241,
|
| 1168 |
-
"grad_norm": 4.148353576660156,
|
| 1169 |
-
"learning_rate": 0.00012703296703296704,
|
| 1170 |
-
"loss": 1.5618,
|
| 1171 |
-
"step": 16600
|
| 1172 |
-
},
|
| 1173 |
-
{
|
| 1174 |
-
"epoch": 1.8351648351648353,
|
| 1175 |
-
"grad_norm": 4.795083999633789,
|
| 1176 |
-
"learning_rate": 0.0001265934065934066,
|
| 1177 |
-
"loss": 1.5203,
|
| 1178 |
-
"step": 16700
|
| 1179 |
-
},
|
| 1180 |
-
{
|
| 1181 |
-
"epoch": 1.8461538461538463,
|
| 1182 |
-
"grad_norm": 4.537530422210693,
|
| 1183 |
-
"learning_rate": 0.00012615384615384615,
|
| 1184 |
-
"loss": 1.5464,
|
| 1185 |
-
"step": 16800
|
| 1186 |
-
},
|
| 1187 |
-
{
|
| 1188 |
-
"epoch": 1.8571428571428572,
|
| 1189 |
-
"grad_norm": 6.016224384307861,
|
| 1190 |
-
"learning_rate": 0.00012571428571428572,
|
| 1191 |
-
"loss": 1.5484,
|
| 1192 |
-
"step": 16900
|
| 1193 |
-
},
|
| 1194 |
-
{
|
| 1195 |
-
"epoch": 1.8681318681318682,
|
| 1196 |
-
"grad_norm": 4.713044166564941,
|
| 1197 |
-
"learning_rate": 0.00012527472527472527,
|
| 1198 |
-
"loss": 1.5361,
|
| 1199 |
-
"step": 17000
|
| 1200 |
-
},
|
| 1201 |
-
{
|
| 1202 |
-
"epoch": 1.879120879120879,
|
| 1203 |
-
"grad_norm": 4.3841094970703125,
|
| 1204 |
-
"learning_rate": 0.00012483516483516484,
|
| 1205 |
-
"loss": 1.5591,
|
| 1206 |
-
"step": 17100
|
| 1207 |
-
},
|
| 1208 |
-
{
|
| 1209 |
-
"epoch": 1.89010989010989,
|
| 1210 |
-
"grad_norm": 5.991598129272461,
|
| 1211 |
-
"learning_rate": 0.0001243956043956044,
|
| 1212 |
-
"loss": 1.5474,
|
| 1213 |
-
"step": 17200
|
| 1214 |
-
},
|
| 1215 |
-
{
|
| 1216 |
-
"epoch": 1.901098901098901,
|
| 1217 |
-
"grad_norm": 4.520913600921631,
|
| 1218 |
-
"learning_rate": 0.00012395604395604395,
|
| 1219 |
-
"loss": 1.5269,
|
| 1220 |
-
"step": 17300
|
| 1221 |
-
},
|
| 1222 |
-
{
|
| 1223 |
-
"epoch": 1.912087912087912,
|
| 1224 |
-
"grad_norm": 4.3360066413879395,
|
| 1225 |
-
"learning_rate": 0.00012351648351648353,
|
| 1226 |
-
"loss": 1.5665,
|
| 1227 |
-
"step": 17400
|
| 1228 |
},
|
| 1229 |
{
|
| 1230 |
"epoch": 1.9230769230769231,
|
| 1231 |
-
"grad_norm":
|
| 1232 |
"learning_rate": 0.0001230769230769231,
|
| 1233 |
-
"loss": 1.
|
| 1234 |
"step": 17500
|
| 1235 |
},
|
| 1236 |
-
{
|
| 1237 |
-
"epoch": 1.934065934065934,
|
| 1238 |
-
"grad_norm": 3.756524085998535,
|
| 1239 |
-
"learning_rate": 0.00012263736263736264,
|
| 1240 |
-
"loss": 1.5602,
|
| 1241 |
-
"step": 17600
|
| 1242 |
-
},
|
| 1243 |
-
{
|
| 1244 |
-
"epoch": 1.945054945054945,
|
| 1245 |
-
"grad_norm": 3.8551275730133057,
|
| 1246 |
-
"learning_rate": 0.0001221978021978022,
|
| 1247 |
-
"loss": 1.55,
|
| 1248 |
-
"step": 17700
|
| 1249 |
-
},
|
| 1250 |
-
{
|
| 1251 |
-
"epoch": 1.9560439560439562,
|
| 1252 |
-
"grad_norm": 5.236588001251221,
|
| 1253 |
-
"learning_rate": 0.00012175824175824176,
|
| 1254 |
-
"loss": 1.5409,
|
| 1255 |
-
"step": 17800
|
| 1256 |
-
},
|
| 1257 |
-
{
|
| 1258 |
-
"epoch": 1.9670329670329672,
|
| 1259 |
-
"grad_norm": 4.134451389312744,
|
| 1260 |
-
"learning_rate": 0.00012131868131868132,
|
| 1261 |
-
"loss": 1.5126,
|
| 1262 |
-
"step": 17900
|
| 1263 |
-
},
|
| 1264 |
{
|
| 1265 |
"epoch": 1.978021978021978,
|
| 1266 |
-
"grad_norm":
|
| 1267 |
"learning_rate": 0.00012087912087912087,
|
| 1268 |
-
"loss": 1.
|
| 1269 |
"step": 18000
|
| 1270 |
},
|
| 1271 |
-
{
|
| 1272 |
-
"epoch": 1.989010989010989,
|
| 1273 |
-
"grad_norm": 4.457215309143066,
|
| 1274 |
-
"learning_rate": 0.00012043956043956045,
|
| 1275 |
-
"loss": 1.554,
|
| 1276 |
-
"step": 18100
|
| 1277 |
-
},
|
| 1278 |
-
{
|
| 1279 |
-
"epoch": 2.0,
|
| 1280 |
-
"grad_norm": 3.8229620456695557,
|
| 1281 |
-
"learning_rate": 0.00012,
|
| 1282 |
-
"loss": 1.5436,
|
| 1283 |
-
"step": 18200
|
| 1284 |
-
},
|
| 1285 |
-
{
|
| 1286 |
-
"epoch": 2.010989010989011,
|
| 1287 |
-
"grad_norm": 3.4859657287597656,
|
| 1288 |
-
"learning_rate": 0.00011956043956043956,
|
| 1289 |
-
"loss": 1.4374,
|
| 1290 |
-
"step": 18300
|
| 1291 |
-
},
|
| 1292 |
-
{
|
| 1293 |
-
"epoch": 2.021978021978022,
|
| 1294 |
-
"grad_norm": 4.995362758636475,
|
| 1295 |
-
"learning_rate": 0.00011912087912087912,
|
| 1296 |
-
"loss": 1.4391,
|
| 1297 |
-
"step": 18400
|
| 1298 |
-
},
|
| 1299 |
{
|
| 1300 |
"epoch": 2.032967032967033,
|
| 1301 |
-
"grad_norm":
|
| 1302 |
"learning_rate": 0.00011868131868131869,
|
| 1303 |
-
"loss": 1.
|
| 1304 |
"step": 18500
|
| 1305 |
},
|
| 1306 |
-
{
|
| 1307 |
-
"epoch": 2.043956043956044,
|
| 1308 |
-
"grad_norm": 2.1358206272125244,
|
| 1309 |
-
"learning_rate": 0.00011824175824175825,
|
| 1310 |
-
"loss": 1.4027,
|
| 1311 |
-
"step": 18600
|
| 1312 |
-
},
|
| 1313 |
-
{
|
| 1314 |
-
"epoch": 2.0549450549450547,
|
| 1315 |
-
"grad_norm": 6.7210187911987305,
|
| 1316 |
-
"learning_rate": 0.0001178021978021978,
|
| 1317 |
-
"loss": 1.4537,
|
| 1318 |
-
"step": 18700
|
| 1319 |
-
},
|
| 1320 |
-
{
|
| 1321 |
-
"epoch": 2.065934065934066,
|
| 1322 |
-
"grad_norm": 3.9406368732452393,
|
| 1323 |
-
"learning_rate": 0.00011736263736263738,
|
| 1324 |
-
"loss": 1.433,
|
| 1325 |
-
"step": 18800
|
| 1326 |
-
},
|
| 1327 |
-
{
|
| 1328 |
-
"epoch": 2.076923076923077,
|
| 1329 |
-
"grad_norm": 4.559463977813721,
|
| 1330 |
-
"learning_rate": 0.00011692307692307694,
|
| 1331 |
-
"loss": 1.4321,
|
| 1332 |
-
"step": 18900
|
| 1333 |
-
},
|
| 1334 |
{
|
| 1335 |
"epoch": 2.087912087912088,
|
| 1336 |
-
"grad_norm":
|
| 1337 |
"learning_rate": 0.0001164835164835165,
|
| 1338 |
-
"loss": 1.
|
| 1339 |
"step": 19000
|
| 1340 |
},
|
| 1341 |
-
{
|
| 1342 |
-
"epoch": 2.098901098901099,
|
| 1343 |
-
"grad_norm": 4.5423760414123535,
|
| 1344 |
-
"learning_rate": 0.00011604395604395607,
|
| 1345 |
-
"loss": 1.4326,
|
| 1346 |
-
"step": 19100
|
| 1347 |
-
},
|
| 1348 |
-
{
|
| 1349 |
-
"epoch": 2.10989010989011,
|
| 1350 |
-
"grad_norm": 4.078341484069824,
|
| 1351 |
-
"learning_rate": 0.0001156043956043956,
|
| 1352 |
-
"loss": 1.4392,
|
| 1353 |
-
"step": 19200
|
| 1354 |
-
},
|
| 1355 |
-
{
|
| 1356 |
-
"epoch": 2.120879120879121,
|
| 1357 |
-
"grad_norm": 4.639175891876221,
|
| 1358 |
-
"learning_rate": 0.00011516483516483517,
|
| 1359 |
-
"loss": 1.4325,
|
| 1360 |
-
"step": 19300
|
| 1361 |
-
},
|
| 1362 |
-
{
|
| 1363 |
-
"epoch": 2.131868131868132,
|
| 1364 |
-
"grad_norm": 3.981069326400757,
|
| 1365 |
-
"learning_rate": 0.00011472527472527473,
|
| 1366 |
-
"loss": 1.4833,
|
| 1367 |
-
"step": 19400
|
| 1368 |
-
},
|
| 1369 |
{
|
| 1370 |
"epoch": 2.142857142857143,
|
| 1371 |
-
"grad_norm":
|
| 1372 |
"learning_rate": 0.00011428571428571428,
|
| 1373 |
-
"loss": 1.
|
| 1374 |
"step": 19500
|
| 1375 |
},
|
| 1376 |
-
{
|
| 1377 |
-
"epoch": 2.1538461538461537,
|
| 1378 |
-
"grad_norm": 5.009954929351807,
|
| 1379 |
-
"learning_rate": 0.00011384615384615384,
|
| 1380 |
-
"loss": 1.4562,
|
| 1381 |
-
"step": 19600
|
| 1382 |
-
},
|
| 1383 |
-
{
|
| 1384 |
-
"epoch": 2.1648351648351647,
|
| 1385 |
-
"grad_norm": 6.226076126098633,
|
| 1386 |
-
"learning_rate": 0.00011340659340659341,
|
| 1387 |
-
"loss": 1.4481,
|
| 1388 |
-
"step": 19700
|
| 1389 |
-
},
|
| 1390 |
-
{
|
| 1391 |
-
"epoch": 2.1758241758241756,
|
| 1392 |
-
"grad_norm": 5.82151985168457,
|
| 1393 |
-
"learning_rate": 0.00011296703296703297,
|
| 1394 |
-
"loss": 1.4518,
|
| 1395 |
-
"step": 19800
|
| 1396 |
-
},
|
| 1397 |
-
{
|
| 1398 |
-
"epoch": 2.186813186813187,
|
| 1399 |
-
"grad_norm": 4.184169292449951,
|
| 1400 |
-
"learning_rate": 0.00011252747252747253,
|
| 1401 |
-
"loss": 1.428,
|
| 1402 |
-
"step": 19900
|
| 1403 |
-
},
|
| 1404 |
{
|
| 1405 |
"epoch": 2.197802197802198,
|
| 1406 |
-
"grad_norm":
|
| 1407 |
"learning_rate": 0.0001120879120879121,
|
| 1408 |
-
"loss": 1.
|
| 1409 |
"step": 20000
|
| 1410 |
},
|
| 1411 |
-
{
|
| 1412 |
-
"epoch": 2.208791208791209,
|
| 1413 |
-
"grad_norm": 4.715268611907959,
|
| 1414 |
-
"learning_rate": 0.00011164835164835166,
|
| 1415 |
-
"loss": 1.4184,
|
| 1416 |
-
"step": 20100
|
| 1417 |
-
},
|
| 1418 |
-
{
|
| 1419 |
-
"epoch": 2.21978021978022,
|
| 1420 |
-
"grad_norm": 4.748434543609619,
|
| 1421 |
-
"learning_rate": 0.00011120879120879122,
|
| 1422 |
-
"loss": 1.4969,
|
| 1423 |
-
"step": 20200
|
| 1424 |
-
},
|
| 1425 |
-
{
|
| 1426 |
-
"epoch": 2.230769230769231,
|
| 1427 |
-
"grad_norm": 4.664528846740723,
|
| 1428 |
-
"learning_rate": 0.00011076923076923077,
|
| 1429 |
-
"loss": 1.4535,
|
| 1430 |
-
"step": 20300
|
| 1431 |
-
},
|
| 1432 |
-
{
|
| 1433 |
-
"epoch": 2.241758241758242,
|
| 1434 |
-
"grad_norm": 4.951592445373535,
|
| 1435 |
-
"learning_rate": 0.00011032967032967035,
|
| 1436 |
-
"loss": 1.4665,
|
| 1437 |
-
"step": 20400
|
| 1438 |
-
},
|
| 1439 |
{
|
| 1440 |
"epoch": 2.2527472527472527,
|
| 1441 |
-
"grad_norm":
|
| 1442 |
"learning_rate": 0.0001098901098901099,
|
| 1443 |
-
"loss": 1.
|
| 1444 |
"step": 20500
|
| 1445 |
},
|
| 1446 |
-
{
|
| 1447 |
-
"epoch": 2.2637362637362637,
|
| 1448 |
-
"grad_norm": 4.944486618041992,
|
| 1449 |
-
"learning_rate": 0.00010945054945054946,
|
| 1450 |
-
"loss": 1.4623,
|
| 1451 |
-
"step": 20600
|
| 1452 |
-
},
|
| 1453 |
-
{
|
| 1454 |
-
"epoch": 2.2747252747252746,
|
| 1455 |
-
"grad_norm": 5.376631259918213,
|
| 1456 |
-
"learning_rate": 0.000109010989010989,
|
| 1457 |
-
"loss": 1.4411,
|
| 1458 |
-
"step": 20700
|
| 1459 |
-
},
|
| 1460 |
-
{
|
| 1461 |
-
"epoch": 2.2857142857142856,
|
| 1462 |
-
"grad_norm": 5.888622760772705,
|
| 1463 |
-
"learning_rate": 0.00010857142857142856,
|
| 1464 |
-
"loss": 1.4515,
|
| 1465 |
-
"step": 20800
|
| 1466 |
-
},
|
| 1467 |
-
{
|
| 1468 |
-
"epoch": 2.2967032967032965,
|
| 1469 |
-
"grad_norm": 4.544143199920654,
|
| 1470 |
-
"learning_rate": 0.00010813186813186814,
|
| 1471 |
-
"loss": 1.4128,
|
| 1472 |
-
"step": 20900
|
| 1473 |
-
},
|
| 1474 |
{
|
| 1475 |
"epoch": 2.3076923076923075,
|
| 1476 |
-
"grad_norm":
|
| 1477 |
"learning_rate": 0.0001076923076923077,
|
| 1478 |
-
"loss": 1.
|
| 1479 |
"step": 21000
|
| 1480 |
},
|
| 1481 |
-
{
|
| 1482 |
-
"epoch": 2.3186813186813184,
|
| 1483 |
-
"grad_norm": 5.327804088592529,
|
| 1484 |
-
"learning_rate": 0.00010725274725274725,
|
| 1485 |
-
"loss": 1.4718,
|
| 1486 |
-
"step": 21100
|
| 1487 |
-
},
|
| 1488 |
-
{
|
| 1489 |
-
"epoch": 2.32967032967033,
|
| 1490 |
-
"grad_norm": 5.668950080871582,
|
| 1491 |
-
"learning_rate": 0.00010681318681318682,
|
| 1492 |
-
"loss": 1.4485,
|
| 1493 |
-
"step": 21200
|
| 1494 |
-
},
|
| 1495 |
-
{
|
| 1496 |
-
"epoch": 2.340659340659341,
|
| 1497 |
-
"grad_norm": 4.657803058624268,
|
| 1498 |
-
"learning_rate": 0.00010637362637362638,
|
| 1499 |
-
"loss": 1.4781,
|
| 1500 |
-
"step": 21300
|
| 1501 |
-
},
|
| 1502 |
-
{
|
| 1503 |
-
"epoch": 2.3516483516483517,
|
| 1504 |
-
"grad_norm": 5.519957065582275,
|
| 1505 |
-
"learning_rate": 0.00010593406593406594,
|
| 1506 |
-
"loss": 1.4356,
|
| 1507 |
-
"step": 21400
|
| 1508 |
-
},
|
| 1509 |
{
|
| 1510 |
"epoch": 2.3626373626373627,
|
| 1511 |
-
"grad_norm":
|
| 1512 |
"learning_rate": 0.0001054945054945055,
|
| 1513 |
-
"loss": 1.
|
| 1514 |
"step": 21500
|
| 1515 |
},
|
| 1516 |
-
{
|
| 1517 |
-
"epoch": 2.3736263736263736,
|
| 1518 |
-
"grad_norm": 7.092361927032471,
|
| 1519 |
-
"learning_rate": 0.00010505494505494507,
|
| 1520 |
-
"loss": 1.4482,
|
| 1521 |
-
"step": 21600
|
| 1522 |
-
},
|
| 1523 |
-
{
|
| 1524 |
-
"epoch": 2.3846153846153846,
|
| 1525 |
-
"grad_norm": 3.946814775466919,
|
| 1526 |
-
"learning_rate": 0.00010461538461538463,
|
| 1527 |
-
"loss": 1.45,
|
| 1528 |
-
"step": 21700
|
| 1529 |
-
},
|
| 1530 |
-
{
|
| 1531 |
-
"epoch": 2.3956043956043955,
|
| 1532 |
-
"grad_norm": 7.119655132293701,
|
| 1533 |
-
"learning_rate": 0.00010417582417582418,
|
| 1534 |
-
"loss": 1.4431,
|
| 1535 |
-
"step": 21800
|
| 1536 |
-
},
|
| 1537 |
-
{
|
| 1538 |
-
"epoch": 2.4065934065934065,
|
| 1539 |
-
"grad_norm": 4.780879974365234,
|
| 1540 |
-
"learning_rate": 0.00010373626373626376,
|
| 1541 |
-
"loss": 1.4483,
|
| 1542 |
-
"step": 21900
|
| 1543 |
-
},
|
| 1544 |
{
|
| 1545 |
"epoch": 2.4175824175824174,
|
| 1546 |
-
"grad_norm":
|
| 1547 |
"learning_rate": 0.00010329670329670331,
|
| 1548 |
-
"loss": 1.
|
| 1549 |
"step": 22000
|
| 1550 |
},
|
| 1551 |
-
{
|
| 1552 |
-
"epoch": 2.4285714285714284,
|
| 1553 |
-
"grad_norm": 3.923065662384033,
|
| 1554 |
-
"learning_rate": 0.00010285714285714286,
|
| 1555 |
-
"loss": 1.4625,
|
| 1556 |
-
"step": 22100
|
| 1557 |
-
},
|
| 1558 |
-
{
|
| 1559 |
-
"epoch": 2.4395604395604398,
|
| 1560 |
-
"grad_norm": 4.095515727996826,
|
| 1561 |
-
"learning_rate": 0.00010241758241758242,
|
| 1562 |
-
"loss": 1.4558,
|
| 1563 |
-
"step": 22200
|
| 1564 |
-
},
|
| 1565 |
-
{
|
| 1566 |
-
"epoch": 2.4505494505494507,
|
| 1567 |
-
"grad_norm": 3.7886769771575928,
|
| 1568 |
-
"learning_rate": 0.00010197802197802197,
|
| 1569 |
-
"loss": 1.4966,
|
| 1570 |
-
"step": 22300
|
| 1571 |
-
},
|
| 1572 |
-
{
|
| 1573 |
-
"epoch": 2.4615384615384617,
|
| 1574 |
-
"grad_norm": 5.163322925567627,
|
| 1575 |
-
"learning_rate": 0.00010153846153846153,
|
| 1576 |
-
"loss": 1.4471,
|
| 1577 |
-
"step": 22400
|
| 1578 |
-
},
|
| 1579 |
{
|
| 1580 |
"epoch": 2.4725274725274726,
|
| 1581 |
-
"grad_norm": 3.
|
| 1582 |
"learning_rate": 0.0001010989010989011,
|
| 1583 |
-
"loss": 1.
|
| 1584 |
"step": 22500
|
| 1585 |
},
|
| 1586 |
-
{
|
| 1587 |
-
"epoch": 2.4835164835164836,
|
| 1588 |
-
"grad_norm": 6.910905838012695,
|
| 1589 |
-
"learning_rate": 0.00010065934065934066,
|
| 1590 |
-
"loss": 1.4391,
|
| 1591 |
-
"step": 22600
|
| 1592 |
-
},
|
| 1593 |
-
{
|
| 1594 |
-
"epoch": 2.4945054945054945,
|
| 1595 |
-
"grad_norm": 4.383417129516602,
|
| 1596 |
-
"learning_rate": 0.00010021978021978022,
|
| 1597 |
-
"loss": 1.4457,
|
| 1598 |
-
"step": 22700
|
| 1599 |
-
},
|
| 1600 |
-
{
|
| 1601 |
-
"epoch": 2.5054945054945055,
|
| 1602 |
-
"grad_norm": 4.188605308532715,
|
| 1603 |
-
"learning_rate": 9.978021978021979e-05,
|
| 1604 |
-
"loss": 1.4623,
|
| 1605 |
-
"step": 22800
|
| 1606 |
-
},
|
| 1607 |
-
{
|
| 1608 |
-
"epoch": 2.5164835164835164,
|
| 1609 |
-
"grad_norm": 3.304708242416382,
|
| 1610 |
-
"learning_rate": 9.934065934065935e-05,
|
| 1611 |
-
"loss": 1.4539,
|
| 1612 |
-
"step": 22900
|
| 1613 |
-
},
|
| 1614 |
{
|
| 1615 |
"epoch": 2.5274725274725274,
|
| 1616 |
-
"grad_norm":
|
| 1617 |
"learning_rate": 9.89010989010989e-05,
|
| 1618 |
-
"loss": 1.
|
| 1619 |
"step": 23000
|
| 1620 |
},
|
| 1621 |
-
{
|
| 1622 |
-
"epoch": 2.5384615384615383,
|
| 1623 |
-
"grad_norm": 4.69569730758667,
|
| 1624 |
-
"learning_rate": 9.846153846153848e-05,
|
| 1625 |
-
"loss": 1.4674,
|
| 1626 |
-
"step": 23100
|
| 1627 |
-
},
|
| 1628 |
-
{
|
| 1629 |
-
"epoch": 2.5494505494505493,
|
| 1630 |
-
"grad_norm": 5.309410572052002,
|
| 1631 |
-
"learning_rate": 9.802197802197802e-05,
|
| 1632 |
-
"loss": 1.4758,
|
| 1633 |
-
"step": 23200
|
| 1634 |
-
},
|
| 1635 |
-
{
|
| 1636 |
-
"epoch": 2.5604395604395602,
|
| 1637 |
-
"grad_norm": 5.060328483581543,
|
| 1638 |
-
"learning_rate": 9.758241758241758e-05,
|
| 1639 |
-
"loss": 1.4705,
|
| 1640 |
-
"step": 23300
|
| 1641 |
-
},
|
| 1642 |
-
{
|
| 1643 |
-
"epoch": 2.571428571428571,
|
| 1644 |
-
"grad_norm": 5.109841346740723,
|
| 1645 |
-
"learning_rate": 9.714285714285715e-05,
|
| 1646 |
-
"loss": 1.4226,
|
| 1647 |
-
"step": 23400
|
| 1648 |
-
},
|
| 1649 |
{
|
| 1650 |
"epoch": 2.5824175824175826,
|
| 1651 |
-
"grad_norm":
|
| 1652 |
"learning_rate": 9.670329670329671e-05,
|
| 1653 |
-
"loss": 1.
|
| 1654 |
"step": 23500
|
| 1655 |
},
|
| 1656 |
-
{
|
| 1657 |
-
"epoch": 2.5934065934065935,
|
| 1658 |
-
"grad_norm": 5.292356967926025,
|
| 1659 |
-
"learning_rate": 9.626373626373627e-05,
|
| 1660 |
-
"loss": 1.4619,
|
| 1661 |
-
"step": 23600
|
| 1662 |
-
},
|
| 1663 |
-
{
|
| 1664 |
-
"epoch": 2.6043956043956045,
|
| 1665 |
-
"grad_norm": 4.887085914611816,
|
| 1666 |
-
"learning_rate": 9.582417582417584e-05,
|
| 1667 |
-
"loss": 1.4379,
|
| 1668 |
-
"step": 23700
|
| 1669 |
-
},
|
| 1670 |
-
{
|
| 1671 |
-
"epoch": 2.6153846153846154,
|
| 1672 |
-
"grad_norm": 4.668257713317871,
|
| 1673 |
-
"learning_rate": 9.53846153846154e-05,
|
| 1674 |
-
"loss": 1.4249,
|
| 1675 |
-
"step": 23800
|
| 1676 |
-
},
|
| 1677 |
-
{
|
| 1678 |
-
"epoch": 2.6263736263736264,
|
| 1679 |
-
"grad_norm": 6.5269293785095215,
|
| 1680 |
-
"learning_rate": 9.494505494505494e-05,
|
| 1681 |
-
"loss": 1.447,
|
| 1682 |
-
"step": 23900
|
| 1683 |
-
},
|
| 1684 |
{
|
| 1685 |
"epoch": 2.6373626373626373,
|
| 1686 |
-
"grad_norm":
|
| 1687 |
"learning_rate": 9.450549450549451e-05,
|
| 1688 |
-
"loss": 1.
|
| 1689 |
"step": 24000
|
| 1690 |
},
|
| 1691 |
-
{
|
| 1692 |
-
"epoch": 2.6483516483516483,
|
| 1693 |
-
"grad_norm": 5.63477897644043,
|
| 1694 |
-
"learning_rate": 9.406593406593407e-05,
|
| 1695 |
-
"loss": 1.4754,
|
| 1696 |
-
"step": 24100
|
| 1697 |
-
},
|
| 1698 |
-
{
|
| 1699 |
-
"epoch": 2.659340659340659,
|
| 1700 |
-
"grad_norm": 4.150704383850098,
|
| 1701 |
-
"learning_rate": 9.362637362637363e-05,
|
| 1702 |
-
"loss": 1.4467,
|
| 1703 |
-
"step": 24200
|
| 1704 |
-
},
|
| 1705 |
-
{
|
| 1706 |
-
"epoch": 2.67032967032967,
|
| 1707 |
-
"grad_norm": 6.260478973388672,
|
| 1708 |
-
"learning_rate": 9.318681318681319e-05,
|
| 1709 |
-
"loss": 1.4454,
|
| 1710 |
-
"step": 24300
|
| 1711 |
-
},
|
| 1712 |
-
{
|
| 1713 |
-
"epoch": 2.6813186813186816,
|
| 1714 |
-
"grad_norm": 6.3307929039001465,
|
| 1715 |
-
"learning_rate": 9.274725274725276e-05,
|
| 1716 |
-
"loss": 1.4387,
|
| 1717 |
-
"step": 24400
|
| 1718 |
-
},
|
| 1719 |
{
|
| 1720 |
"epoch": 2.6923076923076925,
|
| 1721 |
-
"grad_norm":
|
| 1722 |
"learning_rate": 9.230769230769232e-05,
|
| 1723 |
-
"loss": 1.
|
| 1724 |
"step": 24500
|
| 1725 |
},
|
| 1726 |
-
{
|
| 1727 |
-
"epoch": 2.7032967032967035,
|
| 1728 |
-
"grad_norm": 5.355330467224121,
|
| 1729 |
-
"learning_rate": 9.186813186813187e-05,
|
| 1730 |
-
"loss": 1.4282,
|
| 1731 |
-
"step": 24600
|
| 1732 |
-
},
|
| 1733 |
-
{
|
| 1734 |
-
"epoch": 2.7142857142857144,
|
| 1735 |
-
"grad_norm": 4.880154132843018,
|
| 1736 |
-
"learning_rate": 9.142857142857143e-05,
|
| 1737 |
-
"loss": 1.4258,
|
| 1738 |
-
"step": 24700
|
| 1739 |
-
},
|
| 1740 |
-
{
|
| 1741 |
-
"epoch": 2.7252747252747254,
|
| 1742 |
-
"grad_norm": 6.773056983947754,
|
| 1743 |
-
"learning_rate": 9.098901098901099e-05,
|
| 1744 |
-
"loss": 1.4809,
|
| 1745 |
-
"step": 24800
|
| 1746 |
-
},
|
| 1747 |
-
{
|
| 1748 |
-
"epoch": 2.7362637362637363,
|
| 1749 |
-
"grad_norm": 4.716909408569336,
|
| 1750 |
-
"learning_rate": 9.054945054945055e-05,
|
| 1751 |
-
"loss": 1.4571,
|
| 1752 |
-
"step": 24900
|
| 1753 |
-
},
|
| 1754 |
{
|
| 1755 |
"epoch": 2.7472527472527473,
|
| 1756 |
-
"grad_norm":
|
| 1757 |
"learning_rate": 9.010989010989012e-05,
|
| 1758 |
-
"loss": 1.
|
| 1759 |
"step": 25000
|
| 1760 |
},
|
| 1761 |
-
{
|
| 1762 |
-
"epoch": 2.758241758241758,
|
| 1763 |
-
"grad_norm": 7.815650939941406,
|
| 1764 |
-
"learning_rate": 8.967032967032968e-05,
|
| 1765 |
-
"loss": 1.4285,
|
| 1766 |
-
"step": 25100
|
| 1767 |
-
},
|
| 1768 |
-
{
|
| 1769 |
-
"epoch": 2.769230769230769,
|
| 1770 |
-
"grad_norm": 3.7082254886627197,
|
| 1771 |
-
"learning_rate": 8.923076923076924e-05,
|
| 1772 |
-
"loss": 1.4619,
|
| 1773 |
-
"step": 25200
|
| 1774 |
-
},
|
| 1775 |
-
{
|
| 1776 |
-
"epoch": 2.78021978021978,
|
| 1777 |
-
"grad_norm": 5.004034996032715,
|
| 1778 |
-
"learning_rate": 8.87912087912088e-05,
|
| 1779 |
-
"loss": 1.4389,
|
| 1780 |
-
"step": 25300
|
| 1781 |
-
},
|
| 1782 |
-
{
|
| 1783 |
-
"epoch": 2.791208791208791,
|
| 1784 |
-
"grad_norm": 4.239031791687012,
|
| 1785 |
-
"learning_rate": 8.835164835164835e-05,
|
| 1786 |
-
"loss": 1.4405,
|
| 1787 |
-
"step": 25400
|
| 1788 |
-
},
|
| 1789 |
{
|
| 1790 |
"epoch": 2.802197802197802,
|
| 1791 |
-
"grad_norm":
|
| 1792 |
"learning_rate": 8.791208791208791e-05,
|
| 1793 |
-
"loss": 1.
|
| 1794 |
"step": 25500
|
| 1795 |
},
|
| 1796 |
-
{
|
| 1797 |
-
"epoch": 2.813186813186813,
|
| 1798 |
-
"grad_norm": 2.9742252826690674,
|
| 1799 |
-
"learning_rate": 8.747252747252748e-05,
|
| 1800 |
-
"loss": 1.4213,
|
| 1801 |
-
"step": 25600
|
| 1802 |
-
},
|
| 1803 |
-
{
|
| 1804 |
-
"epoch": 2.824175824175824,
|
| 1805 |
-
"grad_norm": 4.981147289276123,
|
| 1806 |
-
"learning_rate": 8.703296703296704e-05,
|
| 1807 |
-
"loss": 1.4565,
|
| 1808 |
-
"step": 25700
|
| 1809 |
-
},
|
| 1810 |
-
{
|
| 1811 |
-
"epoch": 2.8351648351648353,
|
| 1812 |
-
"grad_norm": 4.742037296295166,
|
| 1813 |
-
"learning_rate": 8.65934065934066e-05,
|
| 1814 |
-
"loss": 1.44,
|
| 1815 |
-
"step": 25800
|
| 1816 |
-
},
|
| 1817 |
-
{
|
| 1818 |
-
"epoch": 2.8461538461538463,
|
| 1819 |
-
"grad_norm": 5.7680253982543945,
|
| 1820 |
-
"learning_rate": 8.615384615384617e-05,
|
| 1821 |
-
"loss": 1.4416,
|
| 1822 |
-
"step": 25900
|
| 1823 |
-
},
|
| 1824 |
{
|
| 1825 |
"epoch": 2.857142857142857,
|
| 1826 |
-
"grad_norm":
|
| 1827 |
"learning_rate": 8.571428571428571e-05,
|
| 1828 |
-
"loss": 1.
|
| 1829 |
"step": 26000
|
| 1830 |
},
|
| 1831 |
-
{
|
| 1832 |
-
"epoch": 2.868131868131868,
|
| 1833 |
-
"grad_norm": 6.970800399780273,
|
| 1834 |
-
"learning_rate": 8.527472527472527e-05,
|
| 1835 |
-
"loss": 1.4754,
|
| 1836 |
-
"step": 26100
|
| 1837 |
-
},
|
| 1838 |
-
{
|
| 1839 |
-
"epoch": 2.879120879120879,
|
| 1840 |
-
"grad_norm": 4.084986686706543,
|
| 1841 |
-
"learning_rate": 8.483516483516484e-05,
|
| 1842 |
-
"loss": 1.4369,
|
| 1843 |
-
"step": 26200
|
| 1844 |
-
},
|
| 1845 |
-
{
|
| 1846 |
-
"epoch": 2.89010989010989,
|
| 1847 |
-
"grad_norm": 5.708220481872559,
|
| 1848 |
-
"learning_rate": 8.43956043956044e-05,
|
| 1849 |
-
"loss": 1.4711,
|
| 1850 |
-
"step": 26300
|
| 1851 |
-
},
|
| 1852 |
-
{
|
| 1853 |
-
"epoch": 2.901098901098901,
|
| 1854 |
-
"grad_norm": 4.949507713317871,
|
| 1855 |
-
"learning_rate": 8.395604395604396e-05,
|
| 1856 |
-
"loss": 1.4651,
|
| 1857 |
-
"step": 26400
|
| 1858 |
-
},
|
| 1859 |
{
|
| 1860 |
"epoch": 2.912087912087912,
|
| 1861 |
-
"grad_norm":
|
| 1862 |
"learning_rate": 8.351648351648353e-05,
|
| 1863 |
-
"loss": 1.
|
| 1864 |
"step": 26500
|
| 1865 |
},
|
| 1866 |
-
{
|
| 1867 |
-
"epoch": 2.9230769230769234,
|
| 1868 |
-
"grad_norm": 3.862776756286621,
|
| 1869 |
-
"learning_rate": 8.307692307692309e-05,
|
| 1870 |
-
"loss": 1.437,
|
| 1871 |
-
"step": 26600
|
| 1872 |
-
},
|
| 1873 |
-
{
|
| 1874 |
-
"epoch": 2.9340659340659343,
|
| 1875 |
-
"grad_norm": 4.923271656036377,
|
| 1876 |
-
"learning_rate": 8.263736263736263e-05,
|
| 1877 |
-
"loss": 1.4416,
|
| 1878 |
-
"step": 26700
|
| 1879 |
-
},
|
| 1880 |
-
{
|
| 1881 |
-
"epoch": 2.9450549450549453,
|
| 1882 |
-
"grad_norm": 4.893067359924316,
|
| 1883 |
-
"learning_rate": 8.21978021978022e-05,
|
| 1884 |
-
"loss": 1.4585,
|
| 1885 |
-
"step": 26800
|
| 1886 |
-
},
|
| 1887 |
-
{
|
| 1888 |
-
"epoch": 2.956043956043956,
|
| 1889 |
-
"grad_norm": 4.829097747802734,
|
| 1890 |
-
"learning_rate": 8.175824175824176e-05,
|
| 1891 |
-
"loss": 1.4411,
|
| 1892 |
-
"step": 26900
|
| 1893 |
-
},
|
| 1894 |
{
|
| 1895 |
"epoch": 2.967032967032967,
|
| 1896 |
-
"grad_norm":
|
| 1897 |
"learning_rate": 8.131868131868132e-05,
|
| 1898 |
-
"loss": 1.
|
| 1899 |
"step": 27000
|
| 1900 |
},
|
| 1901 |
-
{
|
| 1902 |
-
"epoch": 2.978021978021978,
|
| 1903 |
-
"grad_norm": 5.461435317993164,
|
| 1904 |
-
"learning_rate": 8.087912087912089e-05,
|
| 1905 |
-
"loss": 1.4576,
|
| 1906 |
-
"step": 27100
|
| 1907 |
-
},
|
| 1908 |
-
{
|
| 1909 |
-
"epoch": 2.989010989010989,
|
| 1910 |
-
"grad_norm": 4.598509311676025,
|
| 1911 |
-
"learning_rate": 8.043956043956045e-05,
|
| 1912 |
-
"loss": 1.4681,
|
| 1913 |
-
"step": 27200
|
| 1914 |
-
},
|
| 1915 |
-
{
|
| 1916 |
-
"epoch": 3.0,
|
| 1917 |
-
"grad_norm": 5.4833149909973145,
|
| 1918 |
-
"learning_rate": 8e-05,
|
| 1919 |
-
"loss": 1.4727,
|
| 1920 |
-
"step": 27300
|
| 1921 |
-
},
|
| 1922 |
-
{
|
| 1923 |
-
"epoch": 3.010989010989011,
|
| 1924 |
-
"grad_norm": 4.663654327392578,
|
| 1925 |
-
"learning_rate": 7.956043956043956e-05,
|
| 1926 |
-
"loss": 1.3761,
|
| 1927 |
-
"step": 27400
|
| 1928 |
-
},
|
| 1929 |
{
|
| 1930 |
"epoch": 3.021978021978022,
|
| 1931 |
-
"grad_norm":
|
| 1932 |
"learning_rate": 7.912087912087912e-05,
|
| 1933 |
-
"loss": 1.
|
| 1934 |
"step": 27500
|
| 1935 |
},
|
| 1936 |
-
{
|
| 1937 |
-
"epoch": 3.032967032967033,
|
| 1938 |
-
"grad_norm": 4.271589279174805,
|
| 1939 |
-
"learning_rate": 7.868131868131868e-05,
|
| 1940 |
-
"loss": 1.3362,
|
| 1941 |
-
"step": 27600
|
| 1942 |
-
},
|
| 1943 |
-
{
|
| 1944 |
-
"epoch": 3.043956043956044,
|
| 1945 |
-
"grad_norm": 6.143280506134033,
|
| 1946 |
-
"learning_rate": 7.824175824175824e-05,
|
| 1947 |
-
"loss": 1.3565,
|
| 1948 |
-
"step": 27700
|
| 1949 |
-
},
|
| 1950 |
-
{
|
| 1951 |
-
"epoch": 3.0549450549450547,
|
| 1952 |
-
"grad_norm": 8.49393081665039,
|
| 1953 |
-
"learning_rate": 7.780219780219781e-05,
|
| 1954 |
-
"loss": 1.3434,
|
| 1955 |
-
"step": 27800
|
| 1956 |
-
},
|
| 1957 |
-
{
|
| 1958 |
-
"epoch": 3.065934065934066,
|
| 1959 |
-
"grad_norm": 5.024664878845215,
|
| 1960 |
-
"learning_rate": 7.736263736263737e-05,
|
| 1961 |
-
"loss": 1.3385,
|
| 1962 |
-
"step": 27900
|
| 1963 |
-
},
|
| 1964 |
{
|
| 1965 |
"epoch": 3.076923076923077,
|
| 1966 |
-
"grad_norm":
|
| 1967 |
"learning_rate": 7.692307692307693e-05,
|
| 1968 |
-
"loss": 1.
|
| 1969 |
"step": 28000
|
| 1970 |
},
|
| 1971 |
-
{
|
| 1972 |
-
"epoch": 3.087912087912088,
|
| 1973 |
-
"grad_norm": 5.72268009185791,
|
| 1974 |
-
"learning_rate": 7.648351648351648e-05,
|
| 1975 |
-
"loss": 1.3673,
|
| 1976 |
-
"step": 28100
|
| 1977 |
-
},
|
| 1978 |
-
{
|
| 1979 |
-
"epoch": 3.098901098901099,
|
| 1980 |
-
"grad_norm": 5.348138332366943,
|
| 1981 |
-
"learning_rate": 7.604395604395604e-05,
|
| 1982 |
-
"loss": 1.3505,
|
| 1983 |
-
"step": 28200
|
| 1984 |
-
},
|
| 1985 |
-
{
|
| 1986 |
-
"epoch": 3.10989010989011,
|
| 1987 |
-
"grad_norm": 4.440935134887695,
|
| 1988 |
-
"learning_rate": 7.56043956043956e-05,
|
| 1989 |
-
"loss": 1.3583,
|
| 1990 |
-
"step": 28300
|
| 1991 |
-
},
|
| 1992 |
-
{
|
| 1993 |
-
"epoch": 3.120879120879121,
|
| 1994 |
-
"grad_norm": 4.01630163192749,
|
| 1995 |
-
"learning_rate": 7.516483516483517e-05,
|
| 1996 |
-
"loss": 1.3524,
|
| 1997 |
-
"step": 28400
|
| 1998 |
-
},
|
| 1999 |
{
|
| 2000 |
"epoch": 3.131868131868132,
|
| 2001 |
-
"grad_norm":
|
| 2002 |
"learning_rate": 7.472527472527473e-05,
|
| 2003 |
-
"loss": 1.
|
| 2004 |
"step": 28500
|
| 2005 |
},
|
| 2006 |
-
{
|
| 2007 |
-
"epoch": 3.142857142857143,
|
| 2008 |
-
"grad_norm": 5.873259544372559,
|
| 2009 |
-
"learning_rate": 7.428571428571429e-05,
|
| 2010 |
-
"loss": 1.3408,
|
| 2011 |
-
"step": 28600
|
| 2012 |
-
},
|
| 2013 |
-
{
|
| 2014 |
-
"epoch": 3.1538461538461537,
|
| 2015 |
-
"grad_norm": 4.558487892150879,
|
| 2016 |
-
"learning_rate": 7.384615384615386e-05,
|
| 2017 |
-
"loss": 1.3017,
|
| 2018 |
-
"step": 28700
|
| 2019 |
-
},
|
| 2020 |
-
{
|
| 2021 |
-
"epoch": 3.1648351648351647,
|
| 2022 |
-
"grad_norm": 7.020089149475098,
|
| 2023 |
-
"learning_rate": 7.34065934065934e-05,
|
| 2024 |
-
"loss": 1.3585,
|
| 2025 |
-
"step": 28800
|
| 2026 |
-
},
|
| 2027 |
-
{
|
| 2028 |
-
"epoch": 3.1758241758241756,
|
| 2029 |
-
"grad_norm": 4.015739917755127,
|
| 2030 |
-
"learning_rate": 7.296703296703296e-05,
|
| 2031 |
-
"loss": 1.3393,
|
| 2032 |
-
"step": 28900
|
| 2033 |
-
},
|
| 2034 |
{
|
| 2035 |
"epoch": 3.186813186813187,
|
| 2036 |
-
"grad_norm":
|
| 2037 |
"learning_rate": 7.252747252747253e-05,
|
| 2038 |
-
"loss": 1.
|
| 2039 |
"step": 29000
|
| 2040 |
},
|
| 2041 |
-
{
|
| 2042 |
-
"epoch": 3.197802197802198,
|
| 2043 |
-
"grad_norm": 9.16402816772461,
|
| 2044 |
-
"learning_rate": 7.208791208791209e-05,
|
| 2045 |
-
"loss": 1.341,
|
| 2046 |
-
"step": 29100
|
| 2047 |
-
},
|
| 2048 |
-
{
|
| 2049 |
-
"epoch": 3.208791208791209,
|
| 2050 |
-
"grad_norm": 5.517131328582764,
|
| 2051 |
-
"learning_rate": 7.164835164835165e-05,
|
| 2052 |
-
"loss": 1.3631,
|
| 2053 |
-
"step": 29200
|
| 2054 |
-
},
|
| 2055 |
-
{
|
| 2056 |
-
"epoch": 3.21978021978022,
|
| 2057 |
-
"grad_norm": 9.678403854370117,
|
| 2058 |
-
"learning_rate": 7.120879120879122e-05,
|
| 2059 |
-
"loss": 1.3739,
|
| 2060 |
-
"step": 29300
|
| 2061 |
-
},
|
| 2062 |
-
{
|
| 2063 |
-
"epoch": 3.230769230769231,
|
| 2064 |
-
"grad_norm": 5.7049880027771,
|
| 2065 |
-
"learning_rate": 7.076923076923078e-05,
|
| 2066 |
-
"loss": 1.3897,
|
| 2067 |
-
"step": 29400
|
| 2068 |
-
},
|
| 2069 |
{
|
| 2070 |
"epoch": 3.241758241758242,
|
| 2071 |
-
"grad_norm":
|
| 2072 |
"learning_rate": 7.032967032967034e-05,
|
| 2073 |
-
"loss": 1.
|
| 2074 |
"step": 29500
|
| 2075 |
},
|
| 2076 |
-
{
|
| 2077 |
-
"epoch": 3.2527472527472527,
|
| 2078 |
-
"grad_norm": 7.1398115158081055,
|
| 2079 |
-
"learning_rate": 6.98901098901099e-05,
|
| 2080 |
-
"loss": 1.3695,
|
| 2081 |
-
"step": 29600
|
| 2082 |
-
},
|
| 2083 |
-
{
|
| 2084 |
-
"epoch": 3.2637362637362637,
|
| 2085 |
-
"grad_norm": 5.985105991363525,
|
| 2086 |
-
"learning_rate": 6.945054945054945e-05,
|
| 2087 |
-
"loss": 1.3642,
|
| 2088 |
-
"step": 29700
|
| 2089 |
-
},
|
| 2090 |
-
{
|
| 2091 |
-
"epoch": 3.2747252747252746,
|
| 2092 |
-
"grad_norm": 6.806448459625244,
|
| 2093 |
-
"learning_rate": 6.901098901098901e-05,
|
| 2094 |
-
"loss": 1.3625,
|
| 2095 |
-
"step": 29800
|
| 2096 |
-
},
|
| 2097 |
-
{
|
| 2098 |
-
"epoch": 3.2857142857142856,
|
| 2099 |
-
"grad_norm": 5.546496868133545,
|
| 2100 |
-
"learning_rate": 6.857142857142858e-05,
|
| 2101 |
-
"loss": 1.3715,
|
| 2102 |
-
"step": 29900
|
| 2103 |
-
},
|
| 2104 |
{
|
| 2105 |
"epoch": 3.2967032967032965,
|
| 2106 |
-
"grad_norm":
|
| 2107 |
"learning_rate": 6.813186813186814e-05,
|
| 2108 |
-
"loss": 1.
|
| 2109 |
"step": 30000
|
| 2110 |
},
|
| 2111 |
-
{
|
| 2112 |
-
"epoch": 3.3076923076923075,
|
| 2113 |
-
"grad_norm": 5.186868190765381,
|
| 2114 |
-
"learning_rate": 6.76923076923077e-05,
|
| 2115 |
-
"loss": 1.3715,
|
| 2116 |
-
"step": 30100
|
| 2117 |
-
},
|
| 2118 |
-
{
|
| 2119 |
-
"epoch": 3.3186813186813184,
|
| 2120 |
-
"grad_norm": 5.597959995269775,
|
| 2121 |
-
"learning_rate": 6.725274725274725e-05,
|
| 2122 |
-
"loss": 1.357,
|
| 2123 |
-
"step": 30200
|
| 2124 |
-
},
|
| 2125 |
-
{
|
| 2126 |
-
"epoch": 3.32967032967033,
|
| 2127 |
-
"grad_norm": 5.903227806091309,
|
| 2128 |
-
"learning_rate": 6.681318681318681e-05,
|
| 2129 |
-
"loss": 1.3479,
|
| 2130 |
-
"step": 30300
|
| 2131 |
-
},
|
| 2132 |
-
{
|
| 2133 |
-
"epoch": 3.340659340659341,
|
| 2134 |
-
"grad_norm": 5.141580581665039,
|
| 2135 |
-
"learning_rate": 6.637362637362637e-05,
|
| 2136 |
-
"loss": 1.3734,
|
| 2137 |
-
"step": 30400
|
| 2138 |
-
},
|
| 2139 |
{
|
| 2140 |
"epoch": 3.3516483516483517,
|
| 2141 |
-
"grad_norm":
|
| 2142 |
"learning_rate": 6.593406593406594e-05,
|
| 2143 |
-
"loss": 1.
|
| 2144 |
"step": 30500
|
| 2145 |
},
|
| 2146 |
-
{
|
| 2147 |
-
"epoch": 3.3626373626373627,
|
| 2148 |
-
"grad_norm": 5.92864465713501,
|
| 2149 |
-
"learning_rate": 6.54945054945055e-05,
|
| 2150 |
-
"loss": 1.358,
|
| 2151 |
-
"step": 30600
|
| 2152 |
-
},
|
| 2153 |
-
{
|
| 2154 |
-
"epoch": 3.3736263736263736,
|
| 2155 |
-
"grad_norm": 5.521232604980469,
|
| 2156 |
-
"learning_rate": 6.505494505494506e-05,
|
| 2157 |
-
"loss": 1.3117,
|
| 2158 |
-
"step": 30700
|
| 2159 |
-
},
|
| 2160 |
-
{
|
| 2161 |
-
"epoch": 3.3846153846153846,
|
| 2162 |
-
"grad_norm": 6.751832962036133,
|
| 2163 |
-
"learning_rate": 6.461538461538462e-05,
|
| 2164 |
-
"loss": 1.3345,
|
| 2165 |
-
"step": 30800
|
| 2166 |
-
},
|
| 2167 |
-
{
|
| 2168 |
-
"epoch": 3.3956043956043955,
|
| 2169 |
-
"grad_norm": 7.770620346069336,
|
| 2170 |
-
"learning_rate": 6.417582417582419e-05,
|
| 2171 |
-
"loss": 1.3192,
|
| 2172 |
-
"step": 30900
|
| 2173 |
-
},
|
| 2174 |
{
|
| 2175 |
"epoch": 3.4065934065934065,
|
| 2176 |
-
"grad_norm":
|
| 2177 |
"learning_rate": 6.373626373626373e-05,
|
| 2178 |
-
"loss": 1.
|
| 2179 |
"step": 31000
|
| 2180 |
},
|
| 2181 |
-
{
|
| 2182 |
-
"epoch": 3.4175824175824174,
|
| 2183 |
-
"grad_norm": 6.34819221496582,
|
| 2184 |
-
"learning_rate": 6.32967032967033e-05,
|
| 2185 |
-
"loss": 1.3726,
|
| 2186 |
-
"step": 31100
|
| 2187 |
-
},
|
| 2188 |
-
{
|
| 2189 |
-
"epoch": 3.4285714285714284,
|
| 2190 |
-
"grad_norm": 7.109829425811768,
|
| 2191 |
-
"learning_rate": 6.285714285714286e-05,
|
| 2192 |
-
"loss": 1.3753,
|
| 2193 |
-
"step": 31200
|
| 2194 |
-
},
|
| 2195 |
-
{
|
| 2196 |
-
"epoch": 3.4395604395604398,
|
| 2197 |
-
"grad_norm": 6.445315361022949,
|
| 2198 |
-
"learning_rate": 6.241758241758242e-05,
|
| 2199 |
-
"loss": 1.3552,
|
| 2200 |
-
"step": 31300
|
| 2201 |
-
},
|
| 2202 |
-
{
|
| 2203 |
-
"epoch": 3.4505494505494507,
|
| 2204 |
-
"grad_norm": 4.920654296875,
|
| 2205 |
-
"learning_rate": 6.197802197802198e-05,
|
| 2206 |
-
"loss": 1.356,
|
| 2207 |
-
"step": 31400
|
| 2208 |
-
},
|
| 2209 |
{
|
| 2210 |
"epoch": 3.4615384615384617,
|
| 2211 |
-
"grad_norm": 3.
|
| 2212 |
"learning_rate": 6.153846153846155e-05,
|
| 2213 |
-
"loss": 1.
|
| 2214 |
"step": 31500
|
| 2215 |
},
|
| 2216 |
-
{
|
| 2217 |
-
"epoch": 3.4725274725274726,
|
| 2218 |
-
"grad_norm": 7.977355480194092,
|
| 2219 |
-
"learning_rate": 6.10989010989011e-05,
|
| 2220 |
-
"loss": 1.3677,
|
| 2221 |
-
"step": 31600
|
| 2222 |
-
},
|
| 2223 |
-
{
|
| 2224 |
-
"epoch": 3.4835164835164836,
|
| 2225 |
-
"grad_norm": 7.270201206207275,
|
| 2226 |
-
"learning_rate": 6.065934065934066e-05,
|
| 2227 |
-
"loss": 1.3782,
|
| 2228 |
-
"step": 31700
|
| 2229 |
-
},
|
| 2230 |
-
{
|
| 2231 |
-
"epoch": 3.4945054945054945,
|
| 2232 |
-
"grad_norm": 4.333983898162842,
|
| 2233 |
-
"learning_rate": 6.021978021978022e-05,
|
| 2234 |
-
"loss": 1.3477,
|
| 2235 |
-
"step": 31800
|
| 2236 |
-
},
|
| 2237 |
-
{
|
| 2238 |
-
"epoch": 3.5054945054945055,
|
| 2239 |
-
"grad_norm": 5.169381141662598,
|
| 2240 |
-
"learning_rate": 5.978021978021978e-05,
|
| 2241 |
-
"loss": 1.3181,
|
| 2242 |
-
"step": 31900
|
| 2243 |
-
},
|
| 2244 |
{
|
| 2245 |
"epoch": 3.5164835164835164,
|
| 2246 |
-
"grad_norm":
|
| 2247 |
"learning_rate": 5.9340659340659345e-05,
|
| 2248 |
-
"loss": 1.
|
| 2249 |
"step": 32000
|
| 2250 |
},
|
| 2251 |
-
{
|
| 2252 |
-
"epoch": 3.5274725274725274,
|
| 2253 |
-
"grad_norm": 5.986196041107178,
|
| 2254 |
-
"learning_rate": 5.89010989010989e-05,
|
| 2255 |
-
"loss": 1.3586,
|
| 2256 |
-
"step": 32100
|
| 2257 |
-
},
|
| 2258 |
-
{
|
| 2259 |
-
"epoch": 3.5384615384615383,
|
| 2260 |
-
"grad_norm": 5.404372215270996,
|
| 2261 |
-
"learning_rate": 5.846153846153847e-05,
|
| 2262 |
-
"loss": 1.3575,
|
| 2263 |
-
"step": 32200
|
| 2264 |
-
},
|
| 2265 |
-
{
|
| 2266 |
-
"epoch": 3.5494505494505493,
|
| 2267 |
-
"grad_norm": 4.96329927444458,
|
| 2268 |
-
"learning_rate": 5.802197802197803e-05,
|
| 2269 |
-
"loss": 1.3882,
|
| 2270 |
-
"step": 32300
|
| 2271 |
-
},
|
| 2272 |
-
{
|
| 2273 |
-
"epoch": 3.5604395604395602,
|
| 2274 |
-
"grad_norm": 8.159017562866211,
|
| 2275 |
-
"learning_rate": 5.7582417582417584e-05,
|
| 2276 |
-
"loss": 1.3239,
|
| 2277 |
-
"step": 32400
|
| 2278 |
-
},
|
| 2279 |
{
|
| 2280 |
"epoch": 3.571428571428571,
|
| 2281 |
-
"grad_norm":
|
| 2282 |
"learning_rate": 5.714285714285714e-05,
|
| 2283 |
-
"loss": 1.
|
| 2284 |
"step": 32500
|
| 2285 |
},
|
| 2286 |
-
{
|
| 2287 |
-
"epoch": 3.5824175824175826,
|
| 2288 |
-
"grad_norm": 4.539822578430176,
|
| 2289 |
-
"learning_rate": 5.6703296703296706e-05,
|
| 2290 |
-
"loss": 1.3353,
|
| 2291 |
-
"step": 32600
|
| 2292 |
-
},
|
| 2293 |
-
{
|
| 2294 |
-
"epoch": 3.5934065934065935,
|
| 2295 |
-
"grad_norm": 7.850346088409424,
|
| 2296 |
-
"learning_rate": 5.6263736263736264e-05,
|
| 2297 |
-
"loss": 1.3523,
|
| 2298 |
-
"step": 32700
|
| 2299 |
-
},
|
| 2300 |
-
{
|
| 2301 |
-
"epoch": 3.6043956043956045,
|
| 2302 |
-
"grad_norm": 6.533629417419434,
|
| 2303 |
-
"learning_rate": 5.582417582417583e-05,
|
| 2304 |
-
"loss": 1.3777,
|
| 2305 |
-
"step": 32800
|
| 2306 |
-
},
|
| 2307 |
-
{
|
| 2308 |
-
"epoch": 3.6153846153846154,
|
| 2309 |
-
"grad_norm": 4.991445064544678,
|
| 2310 |
-
"learning_rate": 5.538461538461539e-05,
|
| 2311 |
-
"loss": 1.358,
|
| 2312 |
-
"step": 32900
|
| 2313 |
-
},
|
| 2314 |
{
|
| 2315 |
"epoch": 3.6263736263736264,
|
| 2316 |
-
"grad_norm":
|
| 2317 |
"learning_rate": 5.494505494505495e-05,
|
| 2318 |
-
"loss": 1.
|
| 2319 |
"step": 33000
|
| 2320 |
},
|
| 2321 |
-
{
|
| 2322 |
-
"epoch": 3.6373626373626373,
|
| 2323 |
-
"grad_norm": 15.688408851623535,
|
| 2324 |
-
"learning_rate": 5.45054945054945e-05,
|
| 2325 |
-
"loss": 1.3492,
|
| 2326 |
-
"step": 33100
|
| 2327 |
-
},
|
| 2328 |
-
{
|
| 2329 |
-
"epoch": 3.6483516483516483,
|
| 2330 |
-
"grad_norm": 5.074567794799805,
|
| 2331 |
-
"learning_rate": 5.406593406593407e-05,
|
| 2332 |
-
"loss": 1.3764,
|
| 2333 |
-
"step": 33200
|
| 2334 |
-
},
|
| 2335 |
-
{
|
| 2336 |
-
"epoch": 3.659340659340659,
|
| 2337 |
-
"grad_norm": 6.356872081756592,
|
| 2338 |
-
"learning_rate": 5.3626373626373626e-05,
|
| 2339 |
-
"loss": 1.3701,
|
| 2340 |
-
"step": 33300
|
| 2341 |
-
},
|
| 2342 |
-
{
|
| 2343 |
-
"epoch": 3.67032967032967,
|
| 2344 |
-
"grad_norm": 4.222293376922607,
|
| 2345 |
-
"learning_rate": 5.318681318681319e-05,
|
| 2346 |
-
"loss": 1.3443,
|
| 2347 |
-
"step": 33400
|
| 2348 |
-
},
|
| 2349 |
{
|
| 2350 |
"epoch": 3.6813186813186816,
|
| 2351 |
-
"grad_norm":
|
| 2352 |
"learning_rate": 5.274725274725275e-05,
|
| 2353 |
-
"loss": 1.
|
| 2354 |
"step": 33500
|
| 2355 |
},
|
| 2356 |
-
{
|
| 2357 |
-
"epoch": 3.6923076923076925,
|
| 2358 |
-
"grad_norm": 5.048154354095459,
|
| 2359 |
-
"learning_rate": 5.230769230769231e-05,
|
| 2360 |
-
"loss": 1.318,
|
| 2361 |
-
"step": 33600
|
| 2362 |
-
},
|
| 2363 |
-
{
|
| 2364 |
-
"epoch": 3.7032967032967035,
|
| 2365 |
-
"grad_norm": 4.227085590362549,
|
| 2366 |
-
"learning_rate": 5.186813186813188e-05,
|
| 2367 |
-
"loss": 1.3581,
|
| 2368 |
-
"step": 33700
|
| 2369 |
-
},
|
| 2370 |
-
{
|
| 2371 |
-
"epoch": 3.7142857142857144,
|
| 2372 |
-
"grad_norm": 9.587520599365234,
|
| 2373 |
-
"learning_rate": 5.142857142857143e-05,
|
| 2374 |
-
"loss": 1.3228,
|
| 2375 |
-
"step": 33800
|
| 2376 |
-
},
|
| 2377 |
-
{
|
| 2378 |
-
"epoch": 3.7252747252747254,
|
| 2379 |
-
"grad_norm": 6.039963722229004,
|
| 2380 |
-
"learning_rate": 5.098901098901099e-05,
|
| 2381 |
-
"loss": 1.349,
|
| 2382 |
-
"step": 33900
|
| 2383 |
-
},
|
| 2384 |
{
|
| 2385 |
"epoch": 3.7362637362637363,
|
| 2386 |
-
"grad_norm":
|
| 2387 |
"learning_rate": 5.054945054945055e-05,
|
| 2388 |
-
"loss": 1.
|
| 2389 |
"step": 34000
|
| 2390 |
},
|
| 2391 |
-
{
|
| 2392 |
-
"epoch": 3.7472527472527473,
|
| 2393 |
-
"grad_norm": 10.933670043945312,
|
| 2394 |
-
"learning_rate": 5.010989010989011e-05,
|
| 2395 |
-
"loss": 1.3313,
|
| 2396 |
-
"step": 34100
|
| 2397 |
-
},
|
| 2398 |
-
{
|
| 2399 |
-
"epoch": 3.758241758241758,
|
| 2400 |
-
"grad_norm": 6.5416364669799805,
|
| 2401 |
-
"learning_rate": 4.9670329670329674e-05,
|
| 2402 |
-
"loss": 1.3401,
|
| 2403 |
-
"step": 34200
|
| 2404 |
-
},
|
| 2405 |
-
{
|
| 2406 |
-
"epoch": 3.769230769230769,
|
| 2407 |
-
"grad_norm": 4.092519283294678,
|
| 2408 |
-
"learning_rate": 4.923076923076924e-05,
|
| 2409 |
-
"loss": 1.349,
|
| 2410 |
-
"step": 34300
|
| 2411 |
-
},
|
| 2412 |
-
{
|
| 2413 |
-
"epoch": 3.78021978021978,
|
| 2414 |
-
"grad_norm": 6.520788669586182,
|
| 2415 |
-
"learning_rate": 4.879120879120879e-05,
|
| 2416 |
-
"loss": 1.3346,
|
| 2417 |
-
"step": 34400
|
| 2418 |
-
},
|
| 2419 |
{
|
| 2420 |
"epoch": 3.791208791208791,
|
| 2421 |
-
"grad_norm":
|
| 2422 |
"learning_rate": 4.8351648351648355e-05,
|
| 2423 |
-
"loss": 1.
|
| 2424 |
"step": 34500
|
| 2425 |
},
|
| 2426 |
-
{
|
| 2427 |
-
"epoch": 3.802197802197802,
|
| 2428 |
-
"grad_norm": 4.871542453765869,
|
| 2429 |
-
"learning_rate": 4.791208791208792e-05,
|
| 2430 |
-
"loss": 1.3517,
|
| 2431 |
-
"step": 34600
|
| 2432 |
-
},
|
| 2433 |
-
{
|
| 2434 |
-
"epoch": 3.813186813186813,
|
| 2435 |
-
"grad_norm": 6.653789520263672,
|
| 2436 |
-
"learning_rate": 4.747252747252747e-05,
|
| 2437 |
-
"loss": 1.356,
|
| 2438 |
-
"step": 34700
|
| 2439 |
-
},
|
| 2440 |
-
{
|
| 2441 |
-
"epoch": 3.824175824175824,
|
| 2442 |
-
"grad_norm": 7.272556781768799,
|
| 2443 |
-
"learning_rate": 4.7032967032967035e-05,
|
| 2444 |
-
"loss": 1.3392,
|
| 2445 |
-
"step": 34800
|
| 2446 |
-
},
|
| 2447 |
-
{
|
| 2448 |
-
"epoch": 3.8351648351648353,
|
| 2449 |
-
"grad_norm": 9.819136619567871,
|
| 2450 |
-
"learning_rate": 4.6593406593406593e-05,
|
| 2451 |
-
"loss": 1.3772,
|
| 2452 |
-
"step": 34900
|
| 2453 |
-
},
|
| 2454 |
{
|
| 2455 |
"epoch": 3.8461538461538463,
|
| 2456 |
-
"grad_norm":
|
| 2457 |
"learning_rate": 4.615384615384616e-05,
|
| 2458 |
-
"loss": 1.
|
| 2459 |
"step": 35000
|
| 2460 |
},
|
| 2461 |
-
{
|
| 2462 |
-
"epoch": 3.857142857142857,
|
| 2463 |
-
"grad_norm": 5.265222549438477,
|
| 2464 |
-
"learning_rate": 4.5714285714285716e-05,
|
| 2465 |
-
"loss": 1.3609,
|
| 2466 |
-
"step": 35100
|
| 2467 |
-
},
|
| 2468 |
-
{
|
| 2469 |
-
"epoch": 3.868131868131868,
|
| 2470 |
-
"grad_norm": 6.421568870544434,
|
| 2471 |
-
"learning_rate": 4.5274725274725274e-05,
|
| 2472 |
-
"loss": 1.3848,
|
| 2473 |
-
"step": 35200
|
| 2474 |
-
},
|
| 2475 |
-
{
|
| 2476 |
-
"epoch": 3.879120879120879,
|
| 2477 |
-
"grad_norm": 5.183870315551758,
|
| 2478 |
-
"learning_rate": 4.483516483516484e-05,
|
| 2479 |
-
"loss": 1.3554,
|
| 2480 |
-
"step": 35300
|
| 2481 |
-
},
|
| 2482 |
-
{
|
| 2483 |
-
"epoch": 3.89010989010989,
|
| 2484 |
-
"grad_norm": 6.2706170082092285,
|
| 2485 |
-
"learning_rate": 4.43956043956044e-05,
|
| 2486 |
-
"loss": 1.3603,
|
| 2487 |
-
"step": 35400
|
| 2488 |
-
},
|
| 2489 |
{
|
| 2490 |
"epoch": 3.901098901098901,
|
| 2491 |
-
"grad_norm":
|
| 2492 |
"learning_rate": 4.3956043956043955e-05,
|
| 2493 |
-
"loss": 1.
|
| 2494 |
"step": 35500
|
| 2495 |
},
|
| 2496 |
-
{
|
| 2497 |
-
"epoch": 3.912087912087912,
|
| 2498 |
-
"grad_norm": 6.289906024932861,
|
| 2499 |
-
"learning_rate": 4.351648351648352e-05,
|
| 2500 |
-
"loss": 1.3188,
|
| 2501 |
-
"step": 35600
|
| 2502 |
-
},
|
| 2503 |
-
{
|
| 2504 |
-
"epoch": 3.9230769230769234,
|
| 2505 |
-
"grad_norm": 8.751489639282227,
|
| 2506 |
-
"learning_rate": 4.3076923076923084e-05,
|
| 2507 |
-
"loss": 1.3744,
|
| 2508 |
-
"step": 35700
|
| 2509 |
-
},
|
| 2510 |
-
{
|
| 2511 |
-
"epoch": 3.9340659340659343,
|
| 2512 |
-
"grad_norm": 8.174962043762207,
|
| 2513 |
-
"learning_rate": 4.2637362637362635e-05,
|
| 2514 |
-
"loss": 1.3422,
|
| 2515 |
-
"step": 35800
|
| 2516 |
-
},
|
| 2517 |
-
{
|
| 2518 |
-
"epoch": 3.9450549450549453,
|
| 2519 |
-
"grad_norm": 5.522929668426514,
|
| 2520 |
-
"learning_rate": 4.21978021978022e-05,
|
| 2521 |
-
"loss": 1.3361,
|
| 2522 |
-
"step": 35900
|
| 2523 |
-
},
|
| 2524 |
{
|
| 2525 |
"epoch": 3.956043956043956,
|
| 2526 |
-
"grad_norm":
|
| 2527 |
"learning_rate": 4.1758241758241765e-05,
|
| 2528 |
-
"loss": 1.
|
| 2529 |
"step": 36000
|
| 2530 |
},
|
| 2531 |
-
{
|
| 2532 |
-
"epoch": 3.967032967032967,
|
| 2533 |
-
"grad_norm": 5.191071510314941,
|
| 2534 |
-
"learning_rate": 4.1318681318681316e-05,
|
| 2535 |
-
"loss": 1.3534,
|
| 2536 |
-
"step": 36100
|
| 2537 |
-
},
|
| 2538 |
-
{
|
| 2539 |
-
"epoch": 3.978021978021978,
|
| 2540 |
-
"grad_norm": 5.726668357849121,
|
| 2541 |
-
"learning_rate": 4.087912087912088e-05,
|
| 2542 |
-
"loss": 1.3106,
|
| 2543 |
-
"step": 36200
|
| 2544 |
-
},
|
| 2545 |
-
{
|
| 2546 |
-
"epoch": 3.989010989010989,
|
| 2547 |
-
"grad_norm": 3.8736844062805176,
|
| 2548 |
-
"learning_rate": 4.0439560439560445e-05,
|
| 2549 |
-
"loss": 1.3477,
|
| 2550 |
-
"step": 36300
|
| 2551 |
-
},
|
| 2552 |
-
{
|
| 2553 |
-
"epoch": 4.0,
|
| 2554 |
-
"grad_norm": 5.525221824645996,
|
| 2555 |
-
"learning_rate": 4e-05,
|
| 2556 |
-
"loss": 1.3572,
|
| 2557 |
-
"step": 36400
|
| 2558 |
-
},
|
| 2559 |
{
|
| 2560 |
"epoch": 4.010989010989011,
|
| 2561 |
-
"grad_norm": 3.
|
| 2562 |
"learning_rate": 3.956043956043956e-05,
|
| 2563 |
-
"loss": 1.
|
| 2564 |
"step": 36500
|
| 2565 |
},
|
| 2566 |
-
{
|
| 2567 |
-
"epoch": 4.021978021978022,
|
| 2568 |
-
"grad_norm": 7.469931125640869,
|
| 2569 |
-
"learning_rate": 3.912087912087912e-05,
|
| 2570 |
-
"loss": 1.2631,
|
| 2571 |
-
"step": 36600
|
| 2572 |
-
},
|
| 2573 |
-
{
|
| 2574 |
-
"epoch": 4.032967032967033,
|
| 2575 |
-
"grad_norm": 5.60840368270874,
|
| 2576 |
-
"learning_rate": 3.8681318681318684e-05,
|
| 2577 |
-
"loss": 1.2658,
|
| 2578 |
-
"step": 36700
|
| 2579 |
-
},
|
| 2580 |
-
{
|
| 2581 |
-
"epoch": 4.043956043956044,
|
| 2582 |
-
"grad_norm": 5.429786205291748,
|
| 2583 |
-
"learning_rate": 3.824175824175824e-05,
|
| 2584 |
-
"loss": 1.2783,
|
| 2585 |
-
"step": 36800
|
| 2586 |
-
},
|
| 2587 |
-
{
|
| 2588 |
-
"epoch": 4.054945054945055,
|
| 2589 |
-
"grad_norm": 6.977511882781982,
|
| 2590 |
-
"learning_rate": 3.78021978021978e-05,
|
| 2591 |
-
"loss": 1.2748,
|
| 2592 |
-
"step": 36900
|
| 2593 |
-
},
|
| 2594 |
{
|
| 2595 |
"epoch": 4.065934065934066,
|
| 2596 |
-
"grad_norm": 5.
|
| 2597 |
"learning_rate": 3.7362637362637365e-05,
|
| 2598 |
-
"loss": 1.
|
| 2599 |
"step": 37000
|
| 2600 |
},
|
| 2601 |
-
{
|
| 2602 |
-
"epoch": 4.076923076923077,
|
| 2603 |
-
"grad_norm": 5.720318794250488,
|
| 2604 |
-
"learning_rate": 3.692307692307693e-05,
|
| 2605 |
-
"loss": 1.3083,
|
| 2606 |
-
"step": 37100
|
| 2607 |
-
},
|
| 2608 |
-
{
|
| 2609 |
-
"epoch": 4.087912087912088,
|
| 2610 |
-
"grad_norm": 7.204798698425293,
|
| 2611 |
-
"learning_rate": 3.648351648351648e-05,
|
| 2612 |
-
"loss": 1.2572,
|
| 2613 |
-
"step": 37200
|
| 2614 |
-
},
|
| 2615 |
-
{
|
| 2616 |
-
"epoch": 4.0989010989010985,
|
| 2617 |
-
"grad_norm": 5.261708736419678,
|
| 2618 |
-
"learning_rate": 3.6043956043956045e-05,
|
| 2619 |
-
"loss": 1.2744,
|
| 2620 |
-
"step": 37300
|
| 2621 |
-
},
|
| 2622 |
-
{
|
| 2623 |
-
"epoch": 4.1098901098901095,
|
| 2624 |
-
"grad_norm": 3.5920863151550293,
|
| 2625 |
-
"learning_rate": 3.560439560439561e-05,
|
| 2626 |
-
"loss": 1.2563,
|
| 2627 |
-
"step": 37400
|
| 2628 |
-
},
|
| 2629 |
{
|
| 2630 |
"epoch": 4.1208791208791204,
|
| 2631 |
-
"grad_norm":
|
| 2632 |
"learning_rate": 3.516483516483517e-05,
|
| 2633 |
-
"loss": 1.
|
| 2634 |
"step": 37500
|
| 2635 |
},
|
| 2636 |
-
{
|
| 2637 |
-
"epoch": 4.131868131868132,
|
| 2638 |
-
"grad_norm": 6.257133483886719,
|
| 2639 |
-
"learning_rate": 3.4725274725274726e-05,
|
| 2640 |
-
"loss": 1.2462,
|
| 2641 |
-
"step": 37600
|
| 2642 |
-
},
|
| 2643 |
-
{
|
| 2644 |
-
"epoch": 4.142857142857143,
|
| 2645 |
-
"grad_norm": 7.922190189361572,
|
| 2646 |
-
"learning_rate": 3.428571428571429e-05,
|
| 2647 |
-
"loss": 1.2428,
|
| 2648 |
-
"step": 37700
|
| 2649 |
-
},
|
| 2650 |
-
{
|
| 2651 |
-
"epoch": 4.153846153846154,
|
| 2652 |
-
"grad_norm": 6.363681316375732,
|
| 2653 |
-
"learning_rate": 3.384615384615385e-05,
|
| 2654 |
-
"loss": 1.2898,
|
| 2655 |
-
"step": 37800
|
| 2656 |
-
},
|
| 2657 |
-
{
|
| 2658 |
-
"epoch": 4.164835164835165,
|
| 2659 |
-
"grad_norm": 4.985836505889893,
|
| 2660 |
-
"learning_rate": 3.3406593406593406e-05,
|
| 2661 |
-
"loss": 1.2577,
|
| 2662 |
-
"step": 37900
|
| 2663 |
-
},
|
| 2664 |
{
|
| 2665 |
"epoch": 4.175824175824176,
|
| 2666 |
-
"grad_norm":
|
| 2667 |
"learning_rate": 3.296703296703297e-05,
|
| 2668 |
-
"loss": 1.
|
| 2669 |
"step": 38000
|
| 2670 |
},
|
| 2671 |
-
{
|
| 2672 |
-
"epoch": 4.186813186813187,
|
| 2673 |
-
"grad_norm": 9.412073135375977,
|
| 2674 |
-
"learning_rate": 3.252747252747253e-05,
|
| 2675 |
-
"loss": 1.253,
|
| 2676 |
-
"step": 38100
|
| 2677 |
-
},
|
| 2678 |
-
{
|
| 2679 |
-
"epoch": 4.197802197802198,
|
| 2680 |
-
"grad_norm": 6.068290710449219,
|
| 2681 |
-
"learning_rate": 3.2087912087912094e-05,
|
| 2682 |
-
"loss": 1.2902,
|
| 2683 |
-
"step": 38200
|
| 2684 |
-
},
|
| 2685 |
-
{
|
| 2686 |
-
"epoch": 4.208791208791209,
|
| 2687 |
-
"grad_norm": 6.8105010986328125,
|
| 2688 |
-
"learning_rate": 3.164835164835165e-05,
|
| 2689 |
-
"loss": 1.2962,
|
| 2690 |
-
"step": 38300
|
| 2691 |
-
},
|
| 2692 |
-
{
|
| 2693 |
-
"epoch": 4.21978021978022,
|
| 2694 |
-
"grad_norm": 6.277568340301514,
|
| 2695 |
-
"learning_rate": 3.120879120879121e-05,
|
| 2696 |
-
"loss": 1.2482,
|
| 2697 |
-
"step": 38400
|
| 2698 |
-
},
|
| 2699 |
{
|
| 2700 |
"epoch": 4.230769230769231,
|
| 2701 |
-
"grad_norm":
|
| 2702 |
"learning_rate": 3.0769230769230774e-05,
|
| 2703 |
-
"loss": 1.
|
| 2704 |
"step": 38500
|
| 2705 |
},
|
| 2706 |
-
{
|
| 2707 |
-
"epoch": 4.241758241758242,
|
| 2708 |
-
"grad_norm": 10.061241149902344,
|
| 2709 |
-
"learning_rate": 3.032967032967033e-05,
|
| 2710 |
-
"loss": 1.2396,
|
| 2711 |
-
"step": 38600
|
| 2712 |
-
},
|
| 2713 |
-
{
|
| 2714 |
-
"epoch": 4.252747252747253,
|
| 2715 |
-
"grad_norm": 5.88912296295166,
|
| 2716 |
-
"learning_rate": 2.989010989010989e-05,
|
| 2717 |
-
"loss": 1.2677,
|
| 2718 |
-
"step": 38700
|
| 2719 |
-
},
|
| 2720 |
-
{
|
| 2721 |
-
"epoch": 4.263736263736264,
|
| 2722 |
-
"grad_norm": 7.349645614624023,
|
| 2723 |
-
"learning_rate": 2.945054945054945e-05,
|
| 2724 |
-
"loss": 1.2397,
|
| 2725 |
-
"step": 38800
|
| 2726 |
-
},
|
| 2727 |
-
{
|
| 2728 |
-
"epoch": 4.274725274725275,
|
| 2729 |
-
"grad_norm": 7.222245216369629,
|
| 2730 |
-
"learning_rate": 2.9010989010989016e-05,
|
| 2731 |
-
"loss": 1.266,
|
| 2732 |
-
"step": 38900
|
| 2733 |
-
},
|
| 2734 |
{
|
| 2735 |
"epoch": 4.285714285714286,
|
| 2736 |
-
"grad_norm":
|
| 2737 |
"learning_rate": 2.857142857142857e-05,
|
| 2738 |
-
"loss": 1.
|
| 2739 |
"step": 39000
|
| 2740 |
},
|
| 2741 |
-
{
|
| 2742 |
-
"epoch": 4.2967032967032965,
|
| 2743 |
-
"grad_norm": 5.563273906707764,
|
| 2744 |
-
"learning_rate": 2.8131868131868132e-05,
|
| 2745 |
-
"loss": 1.2662,
|
| 2746 |
-
"step": 39100
|
| 2747 |
-
},
|
| 2748 |
-
{
|
| 2749 |
-
"epoch": 4.3076923076923075,
|
| 2750 |
-
"grad_norm": 5.392216205596924,
|
| 2751 |
-
"learning_rate": 2.7692307692307694e-05,
|
| 2752 |
-
"loss": 1.2758,
|
| 2753 |
-
"step": 39200
|
| 2754 |
-
},
|
| 2755 |
-
{
|
| 2756 |
-
"epoch": 4.318681318681318,
|
| 2757 |
-
"grad_norm": 6.449277400970459,
|
| 2758 |
-
"learning_rate": 2.725274725274725e-05,
|
| 2759 |
-
"loss": 1.2747,
|
| 2760 |
-
"step": 39300
|
| 2761 |
-
},
|
| 2762 |
-
{
|
| 2763 |
-
"epoch": 4.329670329670329,
|
| 2764 |
-
"grad_norm": 8.933030128479004,
|
| 2765 |
-
"learning_rate": 2.6813186813186813e-05,
|
| 2766 |
-
"loss": 1.253,
|
| 2767 |
-
"step": 39400
|
| 2768 |
-
},
|
| 2769 |
{
|
| 2770 |
"epoch": 4.34065934065934,
|
| 2771 |
-
"grad_norm":
|
| 2772 |
"learning_rate": 2.6373626373626374e-05,
|
| 2773 |
-
"loss": 1.
|
| 2774 |
"step": 39500
|
| 2775 |
},
|
| 2776 |
-
{
|
| 2777 |
-
"epoch": 4.351648351648351,
|
| 2778 |
-
"grad_norm": 5.006295204162598,
|
| 2779 |
-
"learning_rate": 2.593406593406594e-05,
|
| 2780 |
-
"loss": 1.2288,
|
| 2781 |
-
"step": 39600
|
| 2782 |
-
},
|
| 2783 |
-
{
|
| 2784 |
-
"epoch": 4.362637362637362,
|
| 2785 |
-
"grad_norm": 7.589299201965332,
|
| 2786 |
-
"learning_rate": 2.5494505494505493e-05,
|
| 2787 |
-
"loss": 1.2758,
|
| 2788 |
-
"step": 39700
|
| 2789 |
-
},
|
| 2790 |
-
{
|
| 2791 |
-
"epoch": 4.373626373626374,
|
| 2792 |
-
"grad_norm": 7.450462818145752,
|
| 2793 |
-
"learning_rate": 2.5054945054945055e-05,
|
| 2794 |
-
"loss": 1.2594,
|
| 2795 |
-
"step": 39800
|
| 2796 |
-
},
|
| 2797 |
-
{
|
| 2798 |
-
"epoch": 4.384615384615385,
|
| 2799 |
-
"grad_norm": 6.71153450012207,
|
| 2800 |
-
"learning_rate": 2.461538461538462e-05,
|
| 2801 |
-
"loss": 1.2853,
|
| 2802 |
-
"step": 39900
|
| 2803 |
-
},
|
| 2804 |
{
|
| 2805 |
"epoch": 4.395604395604396,
|
| 2806 |
-
"grad_norm":
|
| 2807 |
"learning_rate": 2.4175824175824177e-05,
|
| 2808 |
-
"loss": 1.
|
| 2809 |
"step": 40000
|
| 2810 |
},
|
| 2811 |
-
{
|
| 2812 |
-
"epoch": 4.406593406593407,
|
| 2813 |
-
"grad_norm": 6.07938814163208,
|
| 2814 |
-
"learning_rate": 2.3736263736263735e-05,
|
| 2815 |
-
"loss": 1.3094,
|
| 2816 |
-
"step": 40100
|
| 2817 |
-
},
|
| 2818 |
-
{
|
| 2819 |
-
"epoch": 4.417582417582418,
|
| 2820 |
-
"grad_norm": 5.3800530433654785,
|
| 2821 |
-
"learning_rate": 2.3296703296703297e-05,
|
| 2822 |
-
"loss": 1.2926,
|
| 2823 |
-
"step": 40200
|
| 2824 |
-
},
|
| 2825 |
-
{
|
| 2826 |
-
"epoch": 4.428571428571429,
|
| 2827 |
-
"grad_norm": 8.152463912963867,
|
| 2828 |
-
"learning_rate": 2.2857142857142858e-05,
|
| 2829 |
-
"loss": 1.2626,
|
| 2830 |
-
"step": 40300
|
| 2831 |
-
},
|
| 2832 |
-
{
|
| 2833 |
-
"epoch": 4.43956043956044,
|
| 2834 |
-
"grad_norm": 6.699965000152588,
|
| 2835 |
-
"learning_rate": 2.241758241758242e-05,
|
| 2836 |
-
"loss": 1.2763,
|
| 2837 |
-
"step": 40400
|
| 2838 |
-
},
|
| 2839 |
{
|
| 2840 |
"epoch": 4.450549450549451,
|
| 2841 |
-
"grad_norm":
|
| 2842 |
"learning_rate": 2.1978021978021977e-05,
|
| 2843 |
-
"loss": 1.
|
| 2844 |
"step": 40500
|
| 2845 |
},
|
| 2846 |
-
{
|
| 2847 |
-
"epoch": 4.461538461538462,
|
| 2848 |
-
"grad_norm": 5.403158187866211,
|
| 2849 |
-
"learning_rate": 2.1538461538461542e-05,
|
| 2850 |
-
"loss": 1.2842,
|
| 2851 |
-
"step": 40600
|
| 2852 |
-
},
|
| 2853 |
-
{
|
| 2854 |
-
"epoch": 4.472527472527473,
|
| 2855 |
-
"grad_norm": 6.363950252532959,
|
| 2856 |
-
"learning_rate": 2.10989010989011e-05,
|
| 2857 |
-
"loss": 1.2822,
|
| 2858 |
-
"step": 40700
|
| 2859 |
-
},
|
| 2860 |
-
{
|
| 2861 |
-
"epoch": 4.483516483516484,
|
| 2862 |
-
"grad_norm": 8.08597183227539,
|
| 2863 |
-
"learning_rate": 2.0659340659340658e-05,
|
| 2864 |
-
"loss": 1.2914,
|
| 2865 |
-
"step": 40800
|
| 2866 |
-
},
|
| 2867 |
-
{
|
| 2868 |
-
"epoch": 4.4945054945054945,
|
| 2869 |
-
"grad_norm": 5.750666618347168,
|
| 2870 |
-
"learning_rate": 2.0219780219780223e-05,
|
| 2871 |
-
"loss": 1.2919,
|
| 2872 |
-
"step": 40900
|
| 2873 |
-
},
|
| 2874 |
{
|
| 2875 |
"epoch": 4.5054945054945055,
|
| 2876 |
-
"grad_norm":
|
| 2877 |
"learning_rate": 1.978021978021978e-05,
|
| 2878 |
-
"loss": 1.
|
| 2879 |
"step": 41000
|
| 2880 |
},
|
| 2881 |
-
{
|
| 2882 |
-
"epoch": 4.516483516483516,
|
| 2883 |
-
"grad_norm": 4.417567253112793,
|
| 2884 |
-
"learning_rate": 1.9340659340659342e-05,
|
| 2885 |
-
"loss": 1.2891,
|
| 2886 |
-
"step": 41100
|
| 2887 |
-
},
|
| 2888 |
-
{
|
| 2889 |
-
"epoch": 4.527472527472527,
|
| 2890 |
-
"grad_norm": 8.629948616027832,
|
| 2891 |
-
"learning_rate": 1.89010989010989e-05,
|
| 2892 |
-
"loss": 1.3173,
|
| 2893 |
-
"step": 41200
|
| 2894 |
-
},
|
| 2895 |
-
{
|
| 2896 |
-
"epoch": 4.538461538461538,
|
| 2897 |
-
"grad_norm": 8.524527549743652,
|
| 2898 |
-
"learning_rate": 1.8461538461538465e-05,
|
| 2899 |
-
"loss": 1.2716,
|
| 2900 |
-
"step": 41300
|
| 2901 |
-
},
|
| 2902 |
-
{
|
| 2903 |
-
"epoch": 4.549450549450549,
|
| 2904 |
-
"grad_norm": 7.150579929351807,
|
| 2905 |
-
"learning_rate": 1.8021978021978023e-05,
|
| 2906 |
-
"loss": 1.2893,
|
| 2907 |
-
"step": 41400
|
| 2908 |
-
},
|
| 2909 |
{
|
| 2910 |
"epoch": 4.56043956043956,
|
| 2911 |
-
"grad_norm": 4.
|
| 2912 |
"learning_rate": 1.7582417582417584e-05,
|
| 2913 |
-
"loss": 1.
|
| 2914 |
"step": 41500
|
| 2915 |
},
|
| 2916 |
-
{
|
| 2917 |
-
"epoch": 4.571428571428571,
|
| 2918 |
-
"grad_norm": 10.751927375793457,
|
| 2919 |
-
"learning_rate": 1.7142857142857145e-05,
|
| 2920 |
-
"loss": 1.2755,
|
| 2921 |
-
"step": 41600
|
| 2922 |
-
},
|
| 2923 |
-
{
|
| 2924 |
-
"epoch": 4.582417582417582,
|
| 2925 |
-
"grad_norm": 6.685251712799072,
|
| 2926 |
-
"learning_rate": 1.6703296703296703e-05,
|
| 2927 |
-
"loss": 1.2961,
|
| 2928 |
-
"step": 41700
|
| 2929 |
-
},
|
| 2930 |
-
{
|
| 2931 |
-
"epoch": 4.593406593406593,
|
| 2932 |
-
"grad_norm": 6.694922924041748,
|
| 2933 |
-
"learning_rate": 1.6263736263736265e-05,
|
| 2934 |
-
"loss": 1.2787,
|
| 2935 |
-
"step": 41800
|
| 2936 |
-
},
|
| 2937 |
-
{
|
| 2938 |
-
"epoch": 4.604395604395604,
|
| 2939 |
-
"grad_norm": 7.051819324493408,
|
| 2940 |
-
"learning_rate": 1.5824175824175826e-05,
|
| 2941 |
-
"loss": 1.2991,
|
| 2942 |
-
"step": 41900
|
| 2943 |
-
},
|
| 2944 |
{
|
| 2945 |
"epoch": 4.615384615384615,
|
| 2946 |
-
"grad_norm":
|
| 2947 |
"learning_rate": 1.5384615384615387e-05,
|
| 2948 |
-
"loss": 1.
|
| 2949 |
"step": 42000
|
| 2950 |
},
|
| 2951 |
-
{
|
| 2952 |
-
"epoch": 4.626373626373626,
|
| 2953 |
-
"grad_norm": 4.073245525360107,
|
| 2954 |
-
"learning_rate": 1.4945054945054945e-05,
|
| 2955 |
-
"loss": 1.264,
|
| 2956 |
-
"step": 42100
|
| 2957 |
-
},
|
| 2958 |
-
{
|
| 2959 |
-
"epoch": 4.637362637362637,
|
| 2960 |
-
"grad_norm": 5.903039455413818,
|
| 2961 |
-
"learning_rate": 1.4505494505494508e-05,
|
| 2962 |
-
"loss": 1.2788,
|
| 2963 |
-
"step": 42200
|
| 2964 |
-
},
|
| 2965 |
-
{
|
| 2966 |
-
"epoch": 4.648351648351649,
|
| 2967 |
-
"grad_norm": 5.781927585601807,
|
| 2968 |
-
"learning_rate": 1.4065934065934066e-05,
|
| 2969 |
-
"loss": 1.2667,
|
| 2970 |
-
"step": 42300
|
| 2971 |
-
},
|
| 2972 |
-
{
|
| 2973 |
-
"epoch": 4.65934065934066,
|
| 2974 |
-
"grad_norm": 8.685062408447266,
|
| 2975 |
-
"learning_rate": 1.3626373626373626e-05,
|
| 2976 |
-
"loss": 1.2774,
|
| 2977 |
-
"step": 42400
|
| 2978 |
-
},
|
| 2979 |
{
|
| 2980 |
"epoch": 4.670329670329671,
|
| 2981 |
-
"grad_norm":
|
| 2982 |
"learning_rate": 1.3186813186813187e-05,
|
| 2983 |
-
"loss": 1.
|
| 2984 |
"step": 42500
|
| 2985 |
},
|
| 2986 |
-
{
|
| 2987 |
-
"epoch": 4.681318681318682,
|
| 2988 |
-
"grad_norm": 5.177548885345459,
|
| 2989 |
-
"learning_rate": 1.2747252747252747e-05,
|
| 2990 |
-
"loss": 1.2885,
|
| 2991 |
-
"step": 42600
|
| 2992 |
-
},
|
| 2993 |
-
{
|
| 2994 |
-
"epoch": 4.6923076923076925,
|
| 2995 |
-
"grad_norm": 5.23708963394165,
|
| 2996 |
-
"learning_rate": 1.230769230769231e-05,
|
| 2997 |
-
"loss": 1.2728,
|
| 2998 |
-
"step": 42700
|
| 2999 |
-
},
|
| 3000 |
-
{
|
| 3001 |
-
"epoch": 4.7032967032967035,
|
| 3002 |
-
"grad_norm": 6.084412574768066,
|
| 3003 |
-
"learning_rate": 1.1868131868131868e-05,
|
| 3004 |
-
"loss": 1.2841,
|
| 3005 |
-
"step": 42800
|
| 3006 |
-
},
|
| 3007 |
-
{
|
| 3008 |
-
"epoch": 4.714285714285714,
|
| 3009 |
-
"grad_norm": 7.0549492835998535,
|
| 3010 |
-
"learning_rate": 1.1428571428571429e-05,
|
| 3011 |
-
"loss": 1.2945,
|
| 3012 |
-
"step": 42900
|
| 3013 |
-
},
|
| 3014 |
{
|
| 3015 |
"epoch": 4.725274725274725,
|
| 3016 |
-
"grad_norm":
|
| 3017 |
"learning_rate": 1.0989010989010989e-05,
|
| 3018 |
-
"loss": 1.
|
| 3019 |
"step": 43000
|
| 3020 |
},
|
| 3021 |
-
{
|
| 3022 |
-
"epoch": 4.736263736263736,
|
| 3023 |
-
"grad_norm": 6.6250128746032715,
|
| 3024 |
-
"learning_rate": 1.054945054945055e-05,
|
| 3025 |
-
"loss": 1.2687,
|
| 3026 |
-
"step": 43100
|
| 3027 |
-
},
|
| 3028 |
-
{
|
| 3029 |
-
"epoch": 4.747252747252747,
|
| 3030 |
-
"grad_norm": 7.570830345153809,
|
| 3031 |
-
"learning_rate": 1.0109890109890111e-05,
|
| 3032 |
-
"loss": 1.2593,
|
| 3033 |
-
"step": 43200
|
| 3034 |
-
},
|
| 3035 |
-
{
|
| 3036 |
-
"epoch": 4.758241758241758,
|
| 3037 |
-
"grad_norm": 7.158848285675049,
|
| 3038 |
-
"learning_rate": 9.670329670329671e-06,
|
| 3039 |
-
"loss": 1.3019,
|
| 3040 |
-
"step": 43300
|
| 3041 |
-
},
|
| 3042 |
-
{
|
| 3043 |
-
"epoch": 4.769230769230769,
|
| 3044 |
-
"grad_norm": 4.113587856292725,
|
| 3045 |
-
"learning_rate": 9.230769230769232e-06,
|
| 3046 |
-
"loss": 1.2655,
|
| 3047 |
-
"step": 43400
|
| 3048 |
-
},
|
| 3049 |
{
|
| 3050 |
"epoch": 4.78021978021978,
|
| 3051 |
-
"grad_norm":
|
| 3052 |
"learning_rate": 8.791208791208792e-06,
|
| 3053 |
-
"loss": 1.
|
| 3054 |
"step": 43500
|
| 3055 |
},
|
| 3056 |
-
{
|
| 3057 |
-
"epoch": 4.791208791208791,
|
| 3058 |
-
"grad_norm": 6.412448406219482,
|
| 3059 |
-
"learning_rate": 8.351648351648352e-06,
|
| 3060 |
-
"loss": 1.2663,
|
| 3061 |
-
"step": 43600
|
| 3062 |
-
},
|
| 3063 |
-
{
|
| 3064 |
-
"epoch": 4.802197802197802,
|
| 3065 |
-
"grad_norm": 6.107234001159668,
|
| 3066 |
-
"learning_rate": 7.912087912087913e-06,
|
| 3067 |
-
"loss": 1.2722,
|
| 3068 |
-
"step": 43700
|
| 3069 |
-
},
|
| 3070 |
-
{
|
| 3071 |
-
"epoch": 4.813186813186813,
|
| 3072 |
-
"grad_norm": 10.841856002807617,
|
| 3073 |
-
"learning_rate": 7.4725274725274726e-06,
|
| 3074 |
-
"loss": 1.2749,
|
| 3075 |
-
"step": 43800
|
| 3076 |
-
},
|
| 3077 |
-
{
|
| 3078 |
-
"epoch": 4.824175824175824,
|
| 3079 |
-
"grad_norm": 5.174416542053223,
|
| 3080 |
-
"learning_rate": 7.032967032967033e-06,
|
| 3081 |
-
"loss": 1.2589,
|
| 3082 |
-
"step": 43900
|
| 3083 |
-
},
|
| 3084 |
{
|
| 3085 |
"epoch": 4.835164835164835,
|
| 3086 |
-
"grad_norm":
|
| 3087 |
"learning_rate": 6.5934065934065935e-06,
|
| 3088 |
-
"loss": 1.
|
| 3089 |
"step": 44000
|
| 3090 |
},
|
| 3091 |
-
{
|
| 3092 |
-
"epoch": 4.846153846153846,
|
| 3093 |
-
"grad_norm": 6.337531089782715,
|
| 3094 |
-
"learning_rate": 6.153846153846155e-06,
|
| 3095 |
-
"loss": 1.2489,
|
| 3096 |
-
"step": 44100
|
| 3097 |
-
},
|
| 3098 |
-
{
|
| 3099 |
-
"epoch": 4.857142857142857,
|
| 3100 |
-
"grad_norm": 4.3771562576293945,
|
| 3101 |
-
"learning_rate": 5.7142857142857145e-06,
|
| 3102 |
-
"loss": 1.2597,
|
| 3103 |
-
"step": 44200
|
| 3104 |
-
},
|
| 3105 |
-
{
|
| 3106 |
-
"epoch": 4.868131868131869,
|
| 3107 |
-
"grad_norm": 4.407495021820068,
|
| 3108 |
-
"learning_rate": 5.274725274725275e-06,
|
| 3109 |
-
"loss": 1.2695,
|
| 3110 |
-
"step": 44300
|
| 3111 |
-
},
|
| 3112 |
-
{
|
| 3113 |
-
"epoch": 4.8791208791208796,
|
| 3114 |
-
"grad_norm": 6.802704811096191,
|
| 3115 |
-
"learning_rate": 4.8351648351648355e-06,
|
| 3116 |
-
"loss": 1.2716,
|
| 3117 |
-
"step": 44400
|
| 3118 |
-
},
|
| 3119 |
{
|
| 3120 |
"epoch": 4.8901098901098905,
|
| 3121 |
-
"grad_norm":
|
| 3122 |
"learning_rate": 4.395604395604396e-06,
|
| 3123 |
-
"loss": 1.
|
| 3124 |
"step": 44500
|
| 3125 |
},
|
| 3126 |
-
{
|
| 3127 |
-
"epoch": 4.9010989010989015,
|
| 3128 |
-
"grad_norm": 6.069747447967529,
|
| 3129 |
-
"learning_rate": 3.9560439560439565e-06,
|
| 3130 |
-
"loss": 1.2504,
|
| 3131 |
-
"step": 44600
|
| 3132 |
-
},
|
| 3133 |
-
{
|
| 3134 |
-
"epoch": 4.912087912087912,
|
| 3135 |
-
"grad_norm": 6.654344081878662,
|
| 3136 |
-
"learning_rate": 3.5164835164835165e-06,
|
| 3137 |
-
"loss": 1.2859,
|
| 3138 |
-
"step": 44700
|
| 3139 |
-
},
|
| 3140 |
-
{
|
| 3141 |
-
"epoch": 4.923076923076923,
|
| 3142 |
-
"grad_norm": 5.364152431488037,
|
| 3143 |
-
"learning_rate": 3.0769230769230774e-06,
|
| 3144 |
-
"loss": 1.2886,
|
| 3145 |
-
"step": 44800
|
| 3146 |
-
},
|
| 3147 |
-
{
|
| 3148 |
-
"epoch": 4.934065934065934,
|
| 3149 |
-
"grad_norm": 6.0342607498168945,
|
| 3150 |
-
"learning_rate": 2.6373626373626375e-06,
|
| 3151 |
-
"loss": 1.2876,
|
| 3152 |
-
"step": 44900
|
| 3153 |
-
},
|
| 3154 |
{
|
| 3155 |
"epoch": 4.945054945054945,
|
| 3156 |
-
"grad_norm":
|
| 3157 |
"learning_rate": 2.197802197802198e-06,
|
| 3158 |
-
"loss": 1.
|
| 3159 |
"step": 45000
|
| 3160 |
},
|
| 3161 |
-
{
|
| 3162 |
-
"epoch": 4.956043956043956,
|
| 3163 |
-
"grad_norm": 5.942348480224609,
|
| 3164 |
-
"learning_rate": 1.7582417582417583e-06,
|
| 3165 |
-
"loss": 1.2745,
|
| 3166 |
-
"step": 45100
|
| 3167 |
-
},
|
| 3168 |
-
{
|
| 3169 |
-
"epoch": 4.967032967032967,
|
| 3170 |
-
"grad_norm": 7.537661075592041,
|
| 3171 |
-
"learning_rate": 1.3186813186813187e-06,
|
| 3172 |
-
"loss": 1.2399,
|
| 3173 |
-
"step": 45200
|
| 3174 |
-
},
|
| 3175 |
-
{
|
| 3176 |
-
"epoch": 4.978021978021978,
|
| 3177 |
-
"grad_norm": 6.031464576721191,
|
| 3178 |
-
"learning_rate": 8.791208791208791e-07,
|
| 3179 |
-
"loss": 1.2424,
|
| 3180 |
-
"step": 45300
|
| 3181 |
-
},
|
| 3182 |
-
{
|
| 3183 |
-
"epoch": 4.989010989010989,
|
| 3184 |
-
"grad_norm": 6.415643692016602,
|
| 3185 |
-
"learning_rate": 4.3956043956043957e-07,
|
| 3186 |
-
"loss": 1.2585,
|
| 3187 |
-
"step": 45400
|
| 3188 |
-
},
|
| 3189 |
{
|
| 3190 |
"epoch": 5.0,
|
| 3191 |
-
"grad_norm":
|
| 3192 |
"learning_rate": 0.0,
|
| 3193 |
-
"loss": 1.
|
| 3194 |
"step": 45500
|
| 3195 |
}
|
| 3196 |
],
|
| 3197 |
-
"logging_steps":
|
| 3198 |
"max_steps": 45500,
|
| 3199 |
"num_input_tokens_seen": 0,
|
| 3200 |
"num_train_epochs": 5,
|
|
@@ -3211,7 +663,7 @@
|
|
| 3211 |
"attributes": {}
|
| 3212 |
}
|
| 3213 |
},
|
| 3214 |
-
"total_flos": 3.
|
| 3215 |
"train_batch_size": 4,
|
| 3216 |
"trial_name": null,
|
| 3217 |
"trial_params": null
|
|
|
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
{
|
| 12 |
"epoch": 0.054945054945054944,
|
| 13 |
+
"grad_norm": 1.780437707901001,
|
| 14 |
"learning_rate": 0.0001978021978021978,
|
| 15 |
+
"loss": 1.5906,
|
| 16 |
"step": 500
|
| 17 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18 |
{
|
| 19 |
"epoch": 0.10989010989010989,
|
| 20 |
+
"grad_norm": 2.4972989559173584,
|
| 21 |
"learning_rate": 0.00019560439560439562,
|
| 22 |
+
"loss": 1.5616,
|
| 23 |
"step": 1000
|
| 24 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 25 |
{
|
| 26 |
"epoch": 0.16483516483516483,
|
| 27 |
+
"grad_norm": 1.6102287769317627,
|
| 28 |
"learning_rate": 0.00019340659340659342,
|
| 29 |
+
"loss": 1.5549,
|
| 30 |
"step": 1500
|
| 31 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 32 |
{
|
| 33 |
"epoch": 0.21978021978021978,
|
| 34 |
+
"grad_norm": 1.6323072910308838,
|
| 35 |
"learning_rate": 0.00019120879120879122,
|
| 36 |
+
"loss": 1.5446,
|
| 37 |
"step": 2000
|
| 38 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 39 |
{
|
| 40 |
"epoch": 0.27472527472527475,
|
| 41 |
+
"grad_norm": 1.6228586435317993,
|
| 42 |
"learning_rate": 0.00018901098901098903,
|
| 43 |
+
"loss": 1.5549,
|
| 44 |
"step": 2500
|
| 45 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 46 |
{
|
| 47 |
"epoch": 0.32967032967032966,
|
| 48 |
+
"grad_norm": 1.433160662651062,
|
| 49 |
"learning_rate": 0.00018681318681318683,
|
| 50 |
+
"loss": 1.5521,
|
| 51 |
"step": 3000
|
| 52 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 53 |
{
|
| 54 |
"epoch": 0.38461538461538464,
|
| 55 |
+
"grad_norm": 1.4724050760269165,
|
| 56 |
"learning_rate": 0.00018461538461538463,
|
| 57 |
+
"loss": 1.5335,
|
| 58 |
"step": 3500
|
| 59 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 60 |
{
|
| 61 |
"epoch": 0.43956043956043955,
|
| 62 |
+
"grad_norm": 1.2267966270446777,
|
| 63 |
"learning_rate": 0.0001824175824175824,
|
| 64 |
+
"loss": 1.5359,
|
| 65 |
"step": 4000
|
| 66 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 67 |
{
|
| 68 |
"epoch": 0.4945054945054945,
|
| 69 |
+
"grad_norm": 1.3739681243896484,
|
| 70 |
"learning_rate": 0.00018021978021978024,
|
| 71 |
+
"loss": 1.5362,
|
| 72 |
"step": 4500
|
| 73 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 74 |
{
|
| 75 |
"epoch": 0.5494505494505495,
|
| 76 |
+
"grad_norm": 1.3636854887008667,
|
| 77 |
"learning_rate": 0.00017802197802197802,
|
| 78 |
+
"loss": 1.5257,
|
| 79 |
"step": 5000
|
| 80 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 81 |
{
|
| 82 |
"epoch": 0.6043956043956044,
|
| 83 |
+
"grad_norm": 1.9858155250549316,
|
| 84 |
"learning_rate": 0.00017582417582417582,
|
| 85 |
+
"loss": 1.5378,
|
| 86 |
"step": 5500
|
| 87 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 88 |
{
|
| 89 |
"epoch": 0.6593406593406593,
|
| 90 |
+
"grad_norm": 1.6531416177749634,
|
| 91 |
"learning_rate": 0.00017362637362637365,
|
| 92 |
+
"loss": 1.5239,
|
| 93 |
"step": 6000
|
| 94 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 95 |
{
|
| 96 |
"epoch": 0.7142857142857143,
|
| 97 |
+
"grad_norm": 1.638098120689392,
|
| 98 |
"learning_rate": 0.00017142857142857143,
|
| 99 |
+
"loss": 1.5331,
|
| 100 |
"step": 6500
|
| 101 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 102 |
{
|
| 103 |
"epoch": 0.7692307692307693,
|
| 104 |
+
"grad_norm": 1.8147164583206177,
|
| 105 |
"learning_rate": 0.00016923076923076923,
|
| 106 |
+
"loss": 1.5291,
|
| 107 |
"step": 7000
|
| 108 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 109 |
{
|
| 110 |
"epoch": 0.8241758241758241,
|
| 111 |
+
"grad_norm": 1.4330575466156006,
|
| 112 |
"learning_rate": 0.00016703296703296706,
|
| 113 |
+
"loss": 1.5328,
|
| 114 |
"step": 7500
|
| 115 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 116 |
{
|
| 117 |
"epoch": 0.8791208791208791,
|
| 118 |
+
"grad_norm": 1.3865350484848022,
|
| 119 |
"learning_rate": 0.00016483516483516484,
|
| 120 |
+
"loss": 1.5266,
|
| 121 |
"step": 8000
|
| 122 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 123 |
{
|
| 124 |
"epoch": 0.9340659340659341,
|
| 125 |
+
"grad_norm": 1.8921843767166138,
|
| 126 |
"learning_rate": 0.00016263736263736264,
|
| 127 |
+
"loss": 1.5323,
|
| 128 |
"step": 8500
|
| 129 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 130 |
{
|
| 131 |
"epoch": 0.989010989010989,
|
| 132 |
+
"grad_norm": 2.120455026626587,
|
| 133 |
"learning_rate": 0.00016043956043956044,
|
| 134 |
+
"loss": 1.5254,
|
| 135 |
"step": 9000
|
| 136 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 137 |
{
|
| 138 |
"epoch": 1.043956043956044,
|
| 139 |
+
"grad_norm": 1.6023536920547485,
|
| 140 |
"learning_rate": 0.00015824175824175824,
|
| 141 |
+
"loss": 1.4525,
|
| 142 |
"step": 9500
|
| 143 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 144 |
{
|
| 145 |
"epoch": 1.098901098901099,
|
| 146 |
+
"grad_norm": 1.4707773923873901,
|
| 147 |
"learning_rate": 0.00015604395604395605,
|
| 148 |
+
"loss": 1.4538,
|
| 149 |
"step": 10000
|
| 150 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 151 |
{
|
| 152 |
"epoch": 1.1538461538461537,
|
| 153 |
+
"grad_norm": 1.7984991073608398,
|
| 154 |
"learning_rate": 0.00015384615384615385,
|
| 155 |
+
"loss": 1.4421,
|
| 156 |
"step": 10500
|
| 157 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 158 |
{
|
| 159 |
"epoch": 1.2087912087912087,
|
| 160 |
+
"grad_norm": 1.70187509059906,
|
| 161 |
"learning_rate": 0.00015164835164835165,
|
| 162 |
+
"loss": 1.4454,
|
| 163 |
"step": 11000
|
| 164 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 165 |
{
|
| 166 |
"epoch": 1.2637362637362637,
|
| 167 |
+
"grad_norm": 2.5541532039642334,
|
| 168 |
"learning_rate": 0.00014945054945054946,
|
| 169 |
+
"loss": 1.4512,
|
| 170 |
"step": 11500
|
| 171 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 172 |
{
|
| 173 |
"epoch": 1.3186813186813187,
|
| 174 |
+
"grad_norm": 1.5920054912567139,
|
| 175 |
"learning_rate": 0.00014725274725274726,
|
| 176 |
+
"loss": 1.4557,
|
| 177 |
"step": 12000
|
| 178 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 179 |
{
|
| 180 |
"epoch": 1.3736263736263736,
|
| 181 |
+
"grad_norm": 2.0435898303985596,
|
| 182 |
"learning_rate": 0.00014505494505494506,
|
| 183 |
+
"loss": 1.4496,
|
| 184 |
"step": 12500
|
| 185 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 186 |
{
|
| 187 |
"epoch": 1.4285714285714286,
|
| 188 |
+
"grad_norm": 2.3308217525482178,
|
| 189 |
"learning_rate": 0.00014285714285714287,
|
| 190 |
+
"loss": 1.437,
|
| 191 |
"step": 13000
|
| 192 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 193 |
{
|
| 194 |
"epoch": 1.4835164835164836,
|
| 195 |
+
"grad_norm": 2.7433712482452393,
|
| 196 |
"learning_rate": 0.00014065934065934067,
|
| 197 |
+
"loss": 1.437,
|
| 198 |
"step": 13500
|
| 199 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 200 |
{
|
| 201 |
"epoch": 1.5384615384615383,
|
| 202 |
+
"grad_norm": 2.6499412059783936,
|
| 203 |
"learning_rate": 0.00013846153846153847,
|
| 204 |
+
"loss": 1.4419,
|
| 205 |
"step": 14000
|
| 206 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 207 |
{
|
| 208 |
"epoch": 1.5934065934065935,
|
| 209 |
+
"grad_norm": 2.6177027225494385,
|
| 210 |
"learning_rate": 0.00013626373626373628,
|
| 211 |
+
"loss": 1.4408,
|
| 212 |
"step": 14500
|
| 213 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 214 |
{
|
| 215 |
"epoch": 1.6483516483516483,
|
| 216 |
+
"grad_norm": 1.9484443664550781,
|
| 217 |
"learning_rate": 0.00013406593406593405,
|
| 218 |
+
"loss": 1.4617,
|
| 219 |
"step": 15000
|
| 220 |
},
|
| 221 |
{
|
| 222 |
+
"epoch": 1.7032967032967035,
|
| 223 |
+
"grad_norm": 2.649055242538452,
|
| 224 |
+
"learning_rate": 0.00013186813186813188,
|
| 225 |
+
"loss": 1.4482,
|
| 226 |
+
"step": 15500
|
| 227 |
},
|
| 228 |
{
|
| 229 |
+
"epoch": 1.7582417582417582,
|
| 230 |
+
"grad_norm": 1.9473294019699097,
|
| 231 |
+
"learning_rate": 0.0001296703296703297,
|
| 232 |
+
"loss": 1.4358,
|
| 233 |
+
"step": 16000
|
| 234 |
},
|
| 235 |
{
|
| 236 |
+
"epoch": 1.8131868131868132,
|
| 237 |
+
"grad_norm": 2.2678744792938232,
|
| 238 |
+
"learning_rate": 0.00012747252747252746,
|
| 239 |
+
"loss": 1.4333,
|
| 240 |
+
"step": 16500
|
| 241 |
},
|
| 242 |
{
|
| 243 |
+
"epoch": 1.8681318681318682,
|
| 244 |
+
"grad_norm": 1.629854440689087,
|
| 245 |
+
"learning_rate": 0.00012527472527472527,
|
| 246 |
+
"loss": 1.4305,
|
| 247 |
+
"step": 17000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 248 |
},
|
| 249 |
{
|
| 250 |
"epoch": 1.9230769230769231,
|
| 251 |
+
"grad_norm": 2.318514823913574,
|
| 252 |
"learning_rate": 0.0001230769230769231,
|
| 253 |
+
"loss": 1.4292,
|
| 254 |
"step": 17500
|
| 255 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 256 |
{
|
| 257 |
"epoch": 1.978021978021978,
|
| 258 |
+
"grad_norm": 2.1276426315307617,
|
| 259 |
"learning_rate": 0.00012087912087912087,
|
| 260 |
+
"loss": 1.435,
|
| 261 |
"step": 18000
|
| 262 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 263 |
{
|
| 264 |
"epoch": 2.032967032967033,
|
| 265 |
+
"grad_norm": 2.3333852291107178,
|
| 266 |
"learning_rate": 0.00011868131868131869,
|
| 267 |
+
"loss": 1.3761,
|
| 268 |
"step": 18500
|
| 269 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 270 |
{
|
| 271 |
"epoch": 2.087912087912088,
|
| 272 |
+
"grad_norm": 3.305957078933716,
|
| 273 |
"learning_rate": 0.0001164835164835165,
|
| 274 |
+
"loss": 1.3318,
|
| 275 |
"step": 19000
|
| 276 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 277 |
{
|
| 278 |
"epoch": 2.142857142857143,
|
| 279 |
+
"grad_norm": 5.382218837738037,
|
| 280 |
"learning_rate": 0.00011428571428571428,
|
| 281 |
+
"loss": 1.3463,
|
| 282 |
"step": 19500
|
| 283 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 284 |
{
|
| 285 |
"epoch": 2.197802197802198,
|
| 286 |
+
"grad_norm": 3.35638427734375,
|
| 287 |
"learning_rate": 0.0001120879120879121,
|
| 288 |
+
"loss": 1.3418,
|
| 289 |
"step": 20000
|
| 290 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 291 |
{
|
| 292 |
"epoch": 2.2527472527472527,
|
| 293 |
+
"grad_norm": 3.04825496673584,
|
| 294 |
"learning_rate": 0.0001098901098901099,
|
| 295 |
+
"loss": 1.3568,
|
| 296 |
"step": 20500
|
| 297 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 298 |
{
|
| 299 |
"epoch": 2.3076923076923075,
|
| 300 |
+
"grad_norm": 2.9136993885040283,
|
| 301 |
"learning_rate": 0.0001076923076923077,
|
| 302 |
+
"loss": 1.3507,
|
| 303 |
"step": 21000
|
| 304 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 305 |
{
|
| 306 |
"epoch": 2.3626373626373627,
|
| 307 |
+
"grad_norm": 2.5451395511627197,
|
| 308 |
"learning_rate": 0.0001054945054945055,
|
| 309 |
+
"loss": 1.353,
|
| 310 |
"step": 21500
|
| 311 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 312 |
{
|
| 313 |
"epoch": 2.4175824175824174,
|
| 314 |
+
"grad_norm": 2.779684066772461,
|
| 315 |
"learning_rate": 0.00010329670329670331,
|
| 316 |
+
"loss": 1.3456,
|
| 317 |
"step": 22000
|
| 318 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 319 |
{
|
| 320 |
"epoch": 2.4725274725274726,
|
| 321 |
+
"grad_norm": 3.8211276531219482,
|
| 322 |
"learning_rate": 0.0001010989010989011,
|
| 323 |
+
"loss": 1.3519,
|
| 324 |
"step": 22500
|
| 325 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 326 |
{
|
| 327 |
"epoch": 2.5274725274725274,
|
| 328 |
+
"grad_norm": 2.877223253250122,
|
| 329 |
"learning_rate": 9.89010989010989e-05,
|
| 330 |
+
"loss": 1.3506,
|
| 331 |
"step": 23000
|
| 332 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 333 |
{
|
| 334 |
"epoch": 2.5824175824175826,
|
| 335 |
+
"grad_norm": 2.53200101852417,
|
| 336 |
"learning_rate": 9.670329670329671e-05,
|
| 337 |
+
"loss": 1.3564,
|
| 338 |
"step": 23500
|
| 339 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 340 |
{
|
| 341 |
"epoch": 2.6373626373626373,
|
| 342 |
+
"grad_norm": 3.249028205871582,
|
| 343 |
"learning_rate": 9.450549450549451e-05,
|
| 344 |
+
"loss": 1.3393,
|
| 345 |
"step": 24000
|
| 346 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 347 |
{
|
| 348 |
"epoch": 2.6923076923076925,
|
| 349 |
+
"grad_norm": 2.819939613342285,
|
| 350 |
"learning_rate": 9.230769230769232e-05,
|
| 351 |
+
"loss": 1.3413,
|
| 352 |
"step": 24500
|
| 353 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 354 |
{
|
| 355 |
"epoch": 2.7472527472527473,
|
| 356 |
+
"grad_norm": 3.4391469955444336,
|
| 357 |
"learning_rate": 9.010989010989012e-05,
|
| 358 |
+
"loss": 1.3517,
|
| 359 |
"step": 25000
|
| 360 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 361 |
{
|
| 362 |
"epoch": 2.802197802197802,
|
| 363 |
+
"grad_norm": 2.9319329261779785,
|
| 364 |
"learning_rate": 8.791208791208791e-05,
|
| 365 |
+
"loss": 1.3362,
|
| 366 |
"step": 25500
|
| 367 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 368 |
{
|
| 369 |
"epoch": 2.857142857142857,
|
| 370 |
+
"grad_norm": 2.257335901260376,
|
| 371 |
"learning_rate": 8.571428571428571e-05,
|
| 372 |
+
"loss": 1.3402,
|
| 373 |
"step": 26000
|
| 374 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 375 |
{
|
| 376 |
"epoch": 2.912087912087912,
|
| 377 |
+
"grad_norm": 2.6121294498443604,
|
| 378 |
"learning_rate": 8.351648351648353e-05,
|
| 379 |
+
"loss": 1.3615,
|
| 380 |
"step": 26500
|
| 381 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 382 |
{
|
| 383 |
"epoch": 2.967032967032967,
|
| 384 |
+
"grad_norm": 2.716127395629883,
|
| 385 |
"learning_rate": 8.131868131868132e-05,
|
| 386 |
+
"loss": 1.3418,
|
| 387 |
"step": 27000
|
| 388 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 389 |
{
|
| 390 |
"epoch": 3.021978021978022,
|
| 391 |
+
"grad_norm": 2.227154493331909,
|
| 392 |
"learning_rate": 7.912087912087912e-05,
|
| 393 |
+
"loss": 1.3199,
|
| 394 |
"step": 27500
|
| 395 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 396 |
{
|
| 397 |
"epoch": 3.076923076923077,
|
| 398 |
+
"grad_norm": 3.0606658458709717,
|
| 399 |
"learning_rate": 7.692307692307693e-05,
|
| 400 |
+
"loss": 1.2553,
|
| 401 |
"step": 28000
|
| 402 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 403 |
{
|
| 404 |
"epoch": 3.131868131868132,
|
| 405 |
+
"grad_norm": 2.8261947631835938,
|
| 406 |
"learning_rate": 7.472527472527473e-05,
|
| 407 |
+
"loss": 1.2688,
|
| 408 |
"step": 28500
|
| 409 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 410 |
{
|
| 411 |
"epoch": 3.186813186813187,
|
| 412 |
+
"grad_norm": 3.738393783569336,
|
| 413 |
"learning_rate": 7.252747252747253e-05,
|
| 414 |
+
"loss": 1.2513,
|
| 415 |
"step": 29000
|
| 416 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 417 |
{
|
| 418 |
"epoch": 3.241758241758242,
|
| 419 |
+
"grad_norm": 2.5947108268737793,
|
| 420 |
"learning_rate": 7.032967032967034e-05,
|
| 421 |
+
"loss": 1.2757,
|
| 422 |
"step": 29500
|
| 423 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 424 |
{
|
| 425 |
"epoch": 3.2967032967032965,
|
| 426 |
+
"grad_norm": 2.707387924194336,
|
| 427 |
"learning_rate": 6.813186813186814e-05,
|
| 428 |
+
"loss": 1.274,
|
| 429 |
"step": 30000
|
| 430 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 431 |
{
|
| 432 |
"epoch": 3.3516483516483517,
|
| 433 |
+
"grad_norm": 3.4624383449554443,
|
| 434 |
"learning_rate": 6.593406593406594e-05,
|
| 435 |
+
"loss": 1.2702,
|
| 436 |
"step": 30500
|
| 437 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 438 |
{
|
| 439 |
"epoch": 3.4065934065934065,
|
| 440 |
+
"grad_norm": 3.7795355319976807,
|
| 441 |
"learning_rate": 6.373626373626373e-05,
|
| 442 |
+
"loss": 1.2409,
|
| 443 |
"step": 31000
|
| 444 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 445 |
{
|
| 446 |
"epoch": 3.4615384615384617,
|
| 447 |
+
"grad_norm": 3.521811008453369,
|
| 448 |
"learning_rate": 6.153846153846155e-05,
|
| 449 |
+
"loss": 1.267,
|
| 450 |
"step": 31500
|
| 451 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 452 |
{
|
| 453 |
"epoch": 3.5164835164835164,
|
| 454 |
+
"grad_norm": 2.7402503490448,
|
| 455 |
"learning_rate": 5.9340659340659345e-05,
|
| 456 |
+
"loss": 1.2641,
|
| 457 |
"step": 32000
|
| 458 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 459 |
{
|
| 460 |
"epoch": 3.571428571428571,
|
| 461 |
+
"grad_norm": 2.937671661376953,
|
| 462 |
"learning_rate": 5.714285714285714e-05,
|
| 463 |
+
"loss": 1.2626,
|
| 464 |
"step": 32500
|
| 465 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 466 |
{
|
| 467 |
"epoch": 3.6263736263736264,
|
| 468 |
+
"grad_norm": 2.886845350265503,
|
| 469 |
"learning_rate": 5.494505494505495e-05,
|
| 470 |
+
"loss": 1.269,
|
| 471 |
"step": 33000
|
| 472 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 473 |
{
|
| 474 |
"epoch": 3.6813186813186816,
|
| 475 |
+
"grad_norm": 3.9152579307556152,
|
| 476 |
"learning_rate": 5.274725274725275e-05,
|
| 477 |
+
"loss": 1.2659,
|
| 478 |
"step": 33500
|
| 479 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 480 |
{
|
| 481 |
"epoch": 3.7362637362637363,
|
| 482 |
+
"grad_norm": 2.7805376052856445,
|
| 483 |
"learning_rate": 5.054945054945055e-05,
|
| 484 |
+
"loss": 1.2523,
|
| 485 |
"step": 34000
|
| 486 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 487 |
{
|
| 488 |
"epoch": 3.791208791208791,
|
| 489 |
+
"grad_norm": 3.7305562496185303,
|
| 490 |
"learning_rate": 4.8351648351648355e-05,
|
| 491 |
+
"loss": 1.2528,
|
| 492 |
"step": 34500
|
| 493 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 494 |
{
|
| 495 |
"epoch": 3.8461538461538463,
|
| 496 |
+
"grad_norm": 3.0859031677246094,
|
| 497 |
"learning_rate": 4.615384615384616e-05,
|
| 498 |
+
"loss": 1.273,
|
| 499 |
"step": 35000
|
| 500 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 501 |
{
|
| 502 |
"epoch": 3.901098901098901,
|
| 503 |
+
"grad_norm": 4.31486177444458,
|
| 504 |
"learning_rate": 4.3956043956043955e-05,
|
| 505 |
+
"loss": 1.2739,
|
| 506 |
"step": 35500
|
| 507 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 508 |
{
|
| 509 |
"epoch": 3.956043956043956,
|
| 510 |
+
"grad_norm": 2.46838116645813,
|
| 511 |
"learning_rate": 4.1758241758241765e-05,
|
| 512 |
+
"loss": 1.2555,
|
| 513 |
"step": 36000
|
| 514 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 515 |
{
|
| 516 |
"epoch": 4.010989010989011,
|
| 517 |
+
"grad_norm": 3.3113248348236084,
|
| 518 |
"learning_rate": 3.956043956043956e-05,
|
| 519 |
+
"loss": 1.2388,
|
| 520 |
"step": 36500
|
| 521 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 522 |
{
|
| 523 |
"epoch": 4.065934065934066,
|
| 524 |
+
"grad_norm": 5.830134868621826,
|
| 525 |
"learning_rate": 3.7362637362637365e-05,
|
| 526 |
+
"loss": 1.1952,
|
| 527 |
"step": 37000
|
| 528 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 529 |
{
|
| 530 |
"epoch": 4.1208791208791204,
|
| 531 |
+
"grad_norm": 2.97209095954895,
|
| 532 |
"learning_rate": 3.516483516483517e-05,
|
| 533 |
+
"loss": 1.1902,
|
| 534 |
"step": 37500
|
| 535 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 536 |
{
|
| 537 |
"epoch": 4.175824175824176,
|
| 538 |
+
"grad_norm": 3.9511055946350098,
|
| 539 |
"learning_rate": 3.296703296703297e-05,
|
| 540 |
+
"loss": 1.1837,
|
| 541 |
"step": 38000
|
| 542 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 543 |
{
|
| 544 |
"epoch": 4.230769230769231,
|
| 545 |
+
"grad_norm": 6.446171760559082,
|
| 546 |
"learning_rate": 3.0769230769230774e-05,
|
| 547 |
+
"loss": 1.1888,
|
| 548 |
"step": 38500
|
| 549 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 550 |
{
|
| 551 |
"epoch": 4.285714285714286,
|
| 552 |
+
"grad_norm": 4.0232744216918945,
|
| 553 |
"learning_rate": 2.857142857142857e-05,
|
| 554 |
+
"loss": 1.1752,
|
| 555 |
"step": 39000
|
| 556 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 557 |
{
|
| 558 |
"epoch": 4.34065934065934,
|
| 559 |
+
"grad_norm": 4.632090091705322,
|
| 560 |
"learning_rate": 2.6373626373626374e-05,
|
| 561 |
+
"loss": 1.1964,
|
| 562 |
"step": 39500
|
| 563 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 564 |
{
|
| 565 |
"epoch": 4.395604395604396,
|
| 566 |
+
"grad_norm": 3.9423835277557373,
|
| 567 |
"learning_rate": 2.4175824175824177e-05,
|
| 568 |
+
"loss": 1.1861,
|
| 569 |
"step": 40000
|
| 570 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 571 |
{
|
| 572 |
"epoch": 4.450549450549451,
|
| 573 |
+
"grad_norm": 4.528479099273682,
|
| 574 |
"learning_rate": 2.1978021978021977e-05,
|
| 575 |
+
"loss": 1.1999,
|
| 576 |
"step": 40500
|
| 577 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 578 |
{
|
| 579 |
"epoch": 4.5054945054945055,
|
| 580 |
+
"grad_norm": 4.554440021514893,
|
| 581 |
"learning_rate": 1.978021978021978e-05,
|
| 582 |
+
"loss": 1.2045,
|
| 583 |
"step": 41000
|
| 584 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 585 |
{
|
| 586 |
"epoch": 4.56043956043956,
|
| 587 |
+
"grad_norm": 4.417350769042969,
|
| 588 |
"learning_rate": 1.7582417582417584e-05,
|
| 589 |
+
"loss": 1.2074,
|
| 590 |
"step": 41500
|
| 591 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 592 |
{
|
| 593 |
"epoch": 4.615384615384615,
|
| 594 |
+
"grad_norm": 4.048484802246094,
|
| 595 |
"learning_rate": 1.5384615384615387e-05,
|
| 596 |
+
"loss": 1.203,
|
| 597 |
"step": 42000
|
| 598 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 599 |
{
|
| 600 |
"epoch": 4.670329670329671,
|
| 601 |
+
"grad_norm": 4.659012317657471,
|
| 602 |
"learning_rate": 1.3186813186813187e-05,
|
| 603 |
+
"loss": 1.1917,
|
| 604 |
"step": 42500
|
| 605 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 606 |
{
|
| 607 |
"epoch": 4.725274725274725,
|
| 608 |
+
"grad_norm": 3.7058045864105225,
|
| 609 |
"learning_rate": 1.0989010989010989e-05,
|
| 610 |
+
"loss": 1.1981,
|
| 611 |
"step": 43000
|
| 612 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 613 |
{
|
| 614 |
"epoch": 4.78021978021978,
|
| 615 |
+
"grad_norm": 3.214338779449463,
|
| 616 |
"learning_rate": 8.791208791208792e-06,
|
| 617 |
+
"loss": 1.1868,
|
| 618 |
"step": 43500
|
| 619 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 620 |
{
|
| 621 |
"epoch": 4.835164835164835,
|
| 622 |
+
"grad_norm": 3.2473630905151367,
|
| 623 |
"learning_rate": 6.5934065934065935e-06,
|
| 624 |
+
"loss": 1.1979,
|
| 625 |
"step": 44000
|
| 626 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 627 |
{
|
| 628 |
"epoch": 4.8901098901098905,
|
| 629 |
+
"grad_norm": 3.6867451667785645,
|
| 630 |
"learning_rate": 4.395604395604396e-06,
|
| 631 |
+
"loss": 1.1873,
|
| 632 |
"step": 44500
|
| 633 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 634 |
{
|
| 635 |
"epoch": 4.945054945054945,
|
| 636 |
+
"grad_norm": 4.085856914520264,
|
| 637 |
"learning_rate": 2.197802197802198e-06,
|
| 638 |
+
"loss": 1.1997,
|
| 639 |
"step": 45000
|
| 640 |
},
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 641 |
{
|
| 642 |
"epoch": 5.0,
|
| 643 |
+
"grad_norm": 3.7809948921203613,
|
| 644 |
"learning_rate": 0.0,
|
| 645 |
+
"loss": 1.1797,
|
| 646 |
"step": 45500
|
| 647 |
}
|
| 648 |
],
|
| 649 |
+
"logging_steps": 500,
|
| 650 |
"max_steps": 45500,
|
| 651 |
"num_input_tokens_seen": 0,
|
| 652 |
"num_train_epochs": 5,
|
|
|
|
| 663 |
"attributes": {}
|
| 664 |
}
|
| 665 |
},
|
| 666 |
+
"total_flos": 3.6063170257526784e+18,
|
| 667 |
"train_batch_size": 4,
|
| 668 |
"trial_name": null,
|
| 669 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 5496
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:09c8450b96bf0aa312a253c10156ec3e2c10a6f8f4c4de9d82477109290b625d
|
| 3 |
size 5496
|