Training in progress, step 6500, checkpoint
Browse files
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4523108832
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f30fded0506f5abc758d4b29697667a6c05e1a95e6c9e0686c7f108dc11e49a0
|
| 3 |
size 4523108832
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2912179275
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62b7d9c7d41c6b73fcbd87de771044e3b905349cd837f4d94e3d38a8f1a95531
|
| 3 |
size 2912179275
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fd61fdcd042c7209212ed4986c4cd6aa5d57a730e78431a78580137d7a601038
|
| 3 |
size 14645
|
last-checkpoint/scaler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1383
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eb083ebf35d0e4448f128a01dd67912abe022f1092025a85ebc3f0f8d0fcb3ba
|
| 3 |
size 1383
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:41630e2df6dcc25df36d789c3c5fa21ee69d78ee6210c7879940bab6e3977772
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch":
|
| 6 |
"eval_steps": 500,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -4208,6 +4208,356 @@
|
|
| 4208 |
"learning_rate": 8.004e-05,
|
| 4209 |
"loss": 0.029340890049934388,
|
| 4210 |
"step": 6000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4211 |
}
|
| 4212 |
],
|
| 4213 |
"logging_steps": 10,
|
|
@@ -4227,7 +4577,7 @@
|
|
| 4227 |
"attributes": {}
|
| 4228 |
}
|
| 4229 |
},
|
| 4230 |
-
"total_flos": 5.
|
| 4231 |
"train_batch_size": 1,
|
| 4232 |
"trial_name": null,
|
| 4233 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 7.0128,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 6500,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 4208 |
"learning_rate": 8.004e-05,
|
| 4209 |
"loss": 0.029340890049934388,
|
| 4210 |
"step": 6000
|
| 4211 |
+
},
|
| 4212 |
+
{
|
| 4213 |
+
"epoch": 5.0262,
|
| 4214 |
+
"grad_norm": 0.21514344215393066,
|
| 4215 |
+
"learning_rate": 7.984000000000001e-05,
|
| 4216 |
+
"loss": 0.029882851243019103,
|
| 4217 |
+
"step": 6010
|
| 4218 |
+
},
|
| 4219 |
+
{
|
| 4220 |
+
"epoch": 5.0272,
|
| 4221 |
+
"grad_norm": 0.18462449312210083,
|
| 4222 |
+
"learning_rate": 7.964e-05,
|
| 4223 |
+
"loss": 0.03178698718547821,
|
| 4224 |
+
"step": 6020
|
| 4225 |
+
},
|
| 4226 |
+
{
|
| 4227 |
+
"epoch": 5.0282,
|
| 4228 |
+
"grad_norm": 0.17087651789188385,
|
| 4229 |
+
"learning_rate": 7.944e-05,
|
| 4230 |
+
"loss": 0.031465652585029605,
|
| 4231 |
+
"step": 6030
|
| 4232 |
+
},
|
| 4233 |
+
{
|
| 4234 |
+
"epoch": 5.0292,
|
| 4235 |
+
"grad_norm": 0.17649878561496735,
|
| 4236 |
+
"learning_rate": 7.924000000000001e-05,
|
| 4237 |
+
"loss": 0.030974376201629638,
|
| 4238 |
+
"step": 6040
|
| 4239 |
+
},
|
| 4240 |
+
{
|
| 4241 |
+
"epoch": 5.0302,
|
| 4242 |
+
"grad_norm": 0.17699715495109558,
|
| 4243 |
+
"learning_rate": 7.904e-05,
|
| 4244 |
+
"loss": 0.03126508891582489,
|
| 4245 |
+
"step": 6050
|
| 4246 |
+
},
|
| 4247 |
+
{
|
| 4248 |
+
"epoch": 5.0312,
|
| 4249 |
+
"grad_norm": 0.1981772631406784,
|
| 4250 |
+
"learning_rate": 7.884e-05,
|
| 4251 |
+
"loss": 0.03145314157009125,
|
| 4252 |
+
"step": 6060
|
| 4253 |
+
},
|
| 4254 |
+
{
|
| 4255 |
+
"epoch": 6.001,
|
| 4256 |
+
"grad_norm": 0.20157794654369354,
|
| 4257 |
+
"learning_rate": 7.864e-05,
|
| 4258 |
+
"loss": 0.034029459953308104,
|
| 4259 |
+
"step": 6070
|
| 4260 |
+
},
|
| 4261 |
+
{
|
| 4262 |
+
"epoch": 6.002,
|
| 4263 |
+
"grad_norm": 0.19972266256809235,
|
| 4264 |
+
"learning_rate": 7.844e-05,
|
| 4265 |
+
"loss": 0.03382774591445923,
|
| 4266 |
+
"step": 6080
|
| 4267 |
+
},
|
| 4268 |
+
{
|
| 4269 |
+
"epoch": 6.003,
|
| 4270 |
+
"grad_norm": 0.2004992961883545,
|
| 4271 |
+
"learning_rate": 7.824e-05,
|
| 4272 |
+
"loss": 0.033232077956199646,
|
| 4273 |
+
"step": 6090
|
| 4274 |
+
},
|
| 4275 |
+
{
|
| 4276 |
+
"epoch": 6.004,
|
| 4277 |
+
"grad_norm": 0.19882363080978394,
|
| 4278 |
+
"learning_rate": 7.804e-05,
|
| 4279 |
+
"loss": 0.03139175176620483,
|
| 4280 |
+
"step": 6100
|
| 4281 |
+
},
|
| 4282 |
+
{
|
| 4283 |
+
"epoch": 6.005,
|
| 4284 |
+
"grad_norm": 0.17804212868213654,
|
| 4285 |
+
"learning_rate": 7.784e-05,
|
| 4286 |
+
"loss": 0.03176693022251129,
|
| 4287 |
+
"step": 6110
|
| 4288 |
+
},
|
| 4289 |
+
{
|
| 4290 |
+
"epoch": 6.006,
|
| 4291 |
+
"grad_norm": 0.19885767996311188,
|
| 4292 |
+
"learning_rate": 7.764e-05,
|
| 4293 |
+
"loss": 0.03188234269618988,
|
| 4294 |
+
"step": 6120
|
| 4295 |
+
},
|
| 4296 |
+
{
|
| 4297 |
+
"epoch": 6.007,
|
| 4298 |
+
"grad_norm": 0.231368288397789,
|
| 4299 |
+
"learning_rate": 7.744e-05,
|
| 4300 |
+
"loss": 0.03077830970287323,
|
| 4301 |
+
"step": 6130
|
| 4302 |
+
},
|
| 4303 |
+
{
|
| 4304 |
+
"epoch": 6.008,
|
| 4305 |
+
"grad_norm": 0.21649253368377686,
|
| 4306 |
+
"learning_rate": 7.724e-05,
|
| 4307 |
+
"loss": 0.030950218439102173,
|
| 4308 |
+
"step": 6140
|
| 4309 |
+
},
|
| 4310 |
+
{
|
| 4311 |
+
"epoch": 6.009,
|
| 4312 |
+
"grad_norm": 0.18937084078788757,
|
| 4313 |
+
"learning_rate": 7.704000000000001e-05,
|
| 4314 |
+
"loss": 0.03231707811355591,
|
| 4315 |
+
"step": 6150
|
| 4316 |
+
},
|
| 4317 |
+
{
|
| 4318 |
+
"epoch": 6.01,
|
| 4319 |
+
"grad_norm": 0.1458628922700882,
|
| 4320 |
+
"learning_rate": 7.684e-05,
|
| 4321 |
+
"loss": 0.02982558310031891,
|
| 4322 |
+
"step": 6160
|
| 4323 |
+
},
|
| 4324 |
+
{
|
| 4325 |
+
"epoch": 6.011,
|
| 4326 |
+
"grad_norm": 0.18662185966968536,
|
| 4327 |
+
"learning_rate": 7.664e-05,
|
| 4328 |
+
"loss": 0.030502399802207945,
|
| 4329 |
+
"step": 6170
|
| 4330 |
+
},
|
| 4331 |
+
{
|
| 4332 |
+
"epoch": 6.012,
|
| 4333 |
+
"grad_norm": 0.9817757606506348,
|
| 4334 |
+
"learning_rate": 7.644e-05,
|
| 4335 |
+
"loss": 0.03566881716251373,
|
| 4336 |
+
"step": 6180
|
| 4337 |
+
},
|
| 4338 |
+
{
|
| 4339 |
+
"epoch": 6.013,
|
| 4340 |
+
"grad_norm": 0.21928751468658447,
|
| 4341 |
+
"learning_rate": 7.624e-05,
|
| 4342 |
+
"loss": 0.03193634152412415,
|
| 4343 |
+
"step": 6190
|
| 4344 |
+
},
|
| 4345 |
+
{
|
| 4346 |
+
"epoch": 6.014,
|
| 4347 |
+
"grad_norm": 0.2132958173751831,
|
| 4348 |
+
"learning_rate": 7.604e-05,
|
| 4349 |
+
"loss": 0.03002108633518219,
|
| 4350 |
+
"step": 6200
|
| 4351 |
+
},
|
| 4352 |
+
{
|
| 4353 |
+
"epoch": 6.015,
|
| 4354 |
+
"grad_norm": 0.19893419742584229,
|
| 4355 |
+
"learning_rate": 7.584e-05,
|
| 4356 |
+
"loss": 0.02982170581817627,
|
| 4357 |
+
"step": 6210
|
| 4358 |
+
},
|
| 4359 |
+
{
|
| 4360 |
+
"epoch": 6.016,
|
| 4361 |
+
"grad_norm": 0.17725782096385956,
|
| 4362 |
+
"learning_rate": 7.564e-05,
|
| 4363 |
+
"loss": 0.036422553658485415,
|
| 4364 |
+
"step": 6220
|
| 4365 |
+
},
|
| 4366 |
+
{
|
| 4367 |
+
"epoch": 6.017,
|
| 4368 |
+
"grad_norm": 0.1956397294998169,
|
| 4369 |
+
"learning_rate": 7.544e-05,
|
| 4370 |
+
"loss": 0.028982603549957277,
|
| 4371 |
+
"step": 6230
|
| 4372 |
+
},
|
| 4373 |
+
{
|
| 4374 |
+
"epoch": 6.018,
|
| 4375 |
+
"grad_norm": 0.1871957629919052,
|
| 4376 |
+
"learning_rate": 7.524e-05,
|
| 4377 |
+
"loss": 0.03049376904964447,
|
| 4378 |
+
"step": 6240
|
| 4379 |
+
},
|
| 4380 |
+
{
|
| 4381 |
+
"epoch": 6.019,
|
| 4382 |
+
"grad_norm": 0.15977296233177185,
|
| 4383 |
+
"learning_rate": 7.504e-05,
|
| 4384 |
+
"loss": 0.028836172819137574,
|
| 4385 |
+
"step": 6250
|
| 4386 |
+
},
|
| 4387 |
+
{
|
| 4388 |
+
"epoch": 6.02,
|
| 4389 |
+
"grad_norm": 0.1604572832584381,
|
| 4390 |
+
"learning_rate": 7.484e-05,
|
| 4391 |
+
"loss": 0.031694543361663816,
|
| 4392 |
+
"step": 6260
|
| 4393 |
+
},
|
| 4394 |
+
{
|
| 4395 |
+
"epoch": 6.021,
|
| 4396 |
+
"grad_norm": 0.19708256423473358,
|
| 4397 |
+
"learning_rate": 7.464e-05,
|
| 4398 |
+
"loss": 0.0299109548330307,
|
| 4399 |
+
"step": 6270
|
| 4400 |
+
},
|
| 4401 |
+
{
|
| 4402 |
+
"epoch": 6.022,
|
| 4403 |
+
"grad_norm": 0.15963052213191986,
|
| 4404 |
+
"learning_rate": 7.444e-05,
|
| 4405 |
+
"loss": 0.02851861119270325,
|
| 4406 |
+
"step": 6280
|
| 4407 |
+
},
|
| 4408 |
+
{
|
| 4409 |
+
"epoch": 6.023,
|
| 4410 |
+
"grad_norm": 0.19732700288295746,
|
| 4411 |
+
"learning_rate": 7.424e-05,
|
| 4412 |
+
"loss": 0.02974964678287506,
|
| 4413 |
+
"step": 6290
|
| 4414 |
+
},
|
| 4415 |
+
{
|
| 4416 |
+
"epoch": 6.024,
|
| 4417 |
+
"grad_norm": 0.22040079534053802,
|
| 4418 |
+
"learning_rate": 7.404e-05,
|
| 4419 |
+
"loss": 0.02835639417171478,
|
| 4420 |
+
"step": 6300
|
| 4421 |
+
},
|
| 4422 |
+
{
|
| 4423 |
+
"epoch": 6.025,
|
| 4424 |
+
"grad_norm": 0.19101078808307648,
|
| 4425 |
+
"learning_rate": 7.384e-05,
|
| 4426 |
+
"loss": 0.028555047512054444,
|
| 4427 |
+
"step": 6310
|
| 4428 |
+
},
|
| 4429 |
+
{
|
| 4430 |
+
"epoch": 6.026,
|
| 4431 |
+
"grad_norm": 0.17046277225017548,
|
| 4432 |
+
"learning_rate": 7.364e-05,
|
| 4433 |
+
"loss": 0.029239752888679506,
|
| 4434 |
+
"step": 6320
|
| 4435 |
+
},
|
| 4436 |
+
{
|
| 4437 |
+
"epoch": 6.027,
|
| 4438 |
+
"grad_norm": 0.17496943473815918,
|
| 4439 |
+
"learning_rate": 7.344000000000002e-05,
|
| 4440 |
+
"loss": 0.028895750641822815,
|
| 4441 |
+
"step": 6330
|
| 4442 |
+
},
|
| 4443 |
+
{
|
| 4444 |
+
"epoch": 6.028,
|
| 4445 |
+
"grad_norm": 0.1851808875799179,
|
| 4446 |
+
"learning_rate": 7.324000000000001e-05,
|
| 4447 |
+
"loss": 0.029496818780899048,
|
| 4448 |
+
"step": 6340
|
| 4449 |
+
},
|
| 4450 |
+
{
|
| 4451 |
+
"epoch": 6.029,
|
| 4452 |
+
"grad_norm": 0.1762508600950241,
|
| 4453 |
+
"learning_rate": 7.304e-05,
|
| 4454 |
+
"loss": 0.028394830226898194,
|
| 4455 |
+
"step": 6350
|
| 4456 |
+
},
|
| 4457 |
+
{
|
| 4458 |
+
"epoch": 6.03,
|
| 4459 |
+
"grad_norm": 0.2357683628797531,
|
| 4460 |
+
"learning_rate": 7.284000000000001e-05,
|
| 4461 |
+
"loss": 0.02871710658073425,
|
| 4462 |
+
"step": 6360
|
| 4463 |
+
},
|
| 4464 |
+
{
|
| 4465 |
+
"epoch": 6.031,
|
| 4466 |
+
"grad_norm": 0.16987094283103943,
|
| 4467 |
+
"learning_rate": 7.264000000000001e-05,
|
| 4468 |
+
"loss": 0.030784130096435547,
|
| 4469 |
+
"step": 6370
|
| 4470 |
+
},
|
| 4471 |
+
{
|
| 4472 |
+
"epoch": 7.0008,
|
| 4473 |
+
"grad_norm": 0.18693174421787262,
|
| 4474 |
+
"learning_rate": 7.244e-05,
|
| 4475 |
+
"loss": 0.03147324323654175,
|
| 4476 |
+
"step": 6380
|
| 4477 |
+
},
|
| 4478 |
+
{
|
| 4479 |
+
"epoch": 7.0018,
|
| 4480 |
+
"grad_norm": 0.21501696109771729,
|
| 4481 |
+
"learning_rate": 7.224000000000001e-05,
|
| 4482 |
+
"loss": 0.029030478000640868,
|
| 4483 |
+
"step": 6390
|
| 4484 |
+
},
|
| 4485 |
+
{
|
| 4486 |
+
"epoch": 7.0028,
|
| 4487 |
+
"grad_norm": 0.18433253467082977,
|
| 4488 |
+
"learning_rate": 7.204000000000001e-05,
|
| 4489 |
+
"loss": 0.031191644072532655,
|
| 4490 |
+
"step": 6400
|
| 4491 |
+
},
|
| 4492 |
+
{
|
| 4493 |
+
"epoch": 7.0038,
|
| 4494 |
+
"grad_norm": 0.22278091311454773,
|
| 4495 |
+
"learning_rate": 7.184e-05,
|
| 4496 |
+
"loss": 0.028717485070228577,
|
| 4497 |
+
"step": 6410
|
| 4498 |
+
},
|
| 4499 |
+
{
|
| 4500 |
+
"epoch": 7.0048,
|
| 4501 |
+
"grad_norm": 0.1910051554441452,
|
| 4502 |
+
"learning_rate": 7.164000000000001e-05,
|
| 4503 |
+
"loss": 0.027972865104675292,
|
| 4504 |
+
"step": 6420
|
| 4505 |
+
},
|
| 4506 |
+
{
|
| 4507 |
+
"epoch": 7.0058,
|
| 4508 |
+
"grad_norm": 0.227296844124794,
|
| 4509 |
+
"learning_rate": 7.144000000000001e-05,
|
| 4510 |
+
"loss": 0.027867498993873595,
|
| 4511 |
+
"step": 6430
|
| 4512 |
+
},
|
| 4513 |
+
{
|
| 4514 |
+
"epoch": 7.0068,
|
| 4515 |
+
"grad_norm": 0.21170516312122345,
|
| 4516 |
+
"learning_rate": 7.124e-05,
|
| 4517 |
+
"loss": 0.028536897897720338,
|
| 4518 |
+
"step": 6440
|
| 4519 |
+
},
|
| 4520 |
+
{
|
| 4521 |
+
"epoch": 7.0078,
|
| 4522 |
+
"grad_norm": 0.22283098101615906,
|
| 4523 |
+
"learning_rate": 7.104000000000001e-05,
|
| 4524 |
+
"loss": 0.027557718753814697,
|
| 4525 |
+
"step": 6450
|
| 4526 |
+
},
|
| 4527 |
+
{
|
| 4528 |
+
"epoch": 7.0088,
|
| 4529 |
+
"grad_norm": 0.19392724335193634,
|
| 4530 |
+
"learning_rate": 7.084e-05,
|
| 4531 |
+
"loss": 0.027928680181503296,
|
| 4532 |
+
"step": 6460
|
| 4533 |
+
},
|
| 4534 |
+
{
|
| 4535 |
+
"epoch": 7.0098,
|
| 4536 |
+
"grad_norm": 0.15986867249011993,
|
| 4537 |
+
"learning_rate": 7.064e-05,
|
| 4538 |
+
"loss": 0.028359243273735048,
|
| 4539 |
+
"step": 6470
|
| 4540 |
+
},
|
| 4541 |
+
{
|
| 4542 |
+
"epoch": 7.0108,
|
| 4543 |
+
"grad_norm": 0.16614486277103424,
|
| 4544 |
+
"learning_rate": 7.044000000000001e-05,
|
| 4545 |
+
"loss": 0.027036306262016297,
|
| 4546 |
+
"step": 6480
|
| 4547 |
+
},
|
| 4548 |
+
{
|
| 4549 |
+
"epoch": 7.0118,
|
| 4550 |
+
"grad_norm": 0.192184180021286,
|
| 4551 |
+
"learning_rate": 7.024e-05,
|
| 4552 |
+
"loss": 0.026826027035713195,
|
| 4553 |
+
"step": 6490
|
| 4554 |
+
},
|
| 4555 |
+
{
|
| 4556 |
+
"epoch": 7.0128,
|
| 4557 |
+
"grad_norm": 0.1559998244047165,
|
| 4558 |
+
"learning_rate": 7.004e-05,
|
| 4559 |
+
"loss": 0.030304345488548278,
|
| 4560 |
+
"step": 6500
|
| 4561 |
}
|
| 4562 |
],
|
| 4563 |
"logging_steps": 10,
|
|
|
|
| 4577 |
"attributes": {}
|
| 4578 |
}
|
| 4579 |
},
|
| 4580 |
+
"total_flos": 5.501953492276101e+17,
|
| 4581 |
"train_batch_size": 1,
|
| 4582 |
"trial_name": null,
|
| 4583 |
"trial_params": null
|