Training in progress, step 13000, checkpoint
Browse files- last-checkpoint/global_step13000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt +3 -0
- last-checkpoint/global_step13000/mp_rank_00_model_states.pt +3 -0
- last-checkpoint/latest +1 -1
- last-checkpoint/model.safetensors +1 -1
- last-checkpoint/rng_state.pth +1 -1
- last-checkpoint/scheduler.pt +1 -1
- last-checkpoint/trainer_state.json +585 -5
last-checkpoint/global_step13000/bf16_zero_pp_rank_0_mp_rank_00_optim_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:60c1976b9749dd1ad8141542585466ba658edc3a7f62269542e003ae2aeb36ba
|
| 3 |
+
size 5117197489
|
last-checkpoint/global_step13000/mp_rank_00_model_states.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aa193a4fee965a9d1f6e272e88f494d12dddce81bc90c031f850c8285378f5d1
|
| 3 |
+
size 859127933
|
last-checkpoint/latest
CHANGED
|
@@ -1 +1 @@
|
|
| 1 |
-
|
|
|
|
| 1 |
+
global_step13000
|
last-checkpoint/model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 962205216
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dcf44af2fa113c91fad9fddc778811862704ae01a3e9b4a00d3f5dafecefce33
|
| 3 |
size 962205216
|
last-checkpoint/rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14709
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b43eb4b8d99ea61b9872ece72aa8086e069fc0c1aa3eba317a3f22a8741d1801
|
| 3 |
size 14709
|
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:c94d26cf1abc007265682f393879be534350558ed7250bc2fea3a5b1499b77e9
|
| 3 |
size 1465
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"best_metric": 83.
|
| 3 |
-
"best_model_checkpoint": "./iteboshi_student_model_temp/checkpoint-
|
| 4 |
-
"epoch":
|
| 5 |
"eval_steps": 1000,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -3197,6 +3197,586 @@
|
|
| 3197 |
"eval_steps_per_second": 2.024,
|
| 3198 |
"eval_wer": 83.86610089580387,
|
| 3199 |
"step": 11000
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3200 |
}
|
| 3201 |
],
|
| 3202 |
"logging_steps": 25,
|
|
@@ -3216,7 +3796,7 @@
|
|
| 3216 |
"attributes": {}
|
| 3217 |
}
|
| 3218 |
},
|
| 3219 |
-
"total_flos":
|
| 3220 |
"train_batch_size": 4,
|
| 3221 |
"trial_name": null,
|
| 3222 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
+
"best_metric": 83.48892032060348,
|
| 3 |
+
"best_model_checkpoint": "./iteboshi_student_model_temp/checkpoint-12000",
|
| 4 |
+
"epoch": 14.317180616740089,
|
| 5 |
"eval_steps": 1000,
|
| 6 |
+
"global_step": 13000,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 3197 |
"eval_steps_per_second": 2.024,
|
| 3198 |
"eval_wer": 83.86610089580387,
|
| 3199 |
"step": 11000
|
| 3200 |
+
},
|
| 3201 |
+
{
|
| 3202 |
+
"epoch": 12.142070484581497,
|
| 3203 |
+
"grad_norm": 0.24581079185009003,
|
| 3204 |
+
"learning_rate": 9.205128205128206e-06,
|
| 3205 |
+
"loss": 0.0142,
|
| 3206 |
+
"step": 11025
|
| 3207 |
+
},
|
| 3208 |
+
{
|
| 3209 |
+
"epoch": 12.169603524229075,
|
| 3210 |
+
"grad_norm": 0.1644822657108307,
|
| 3211 |
+
"learning_rate": 9.17948717948718e-06,
|
| 3212 |
+
"loss": 0.0115,
|
| 3213 |
+
"step": 11050
|
| 3214 |
+
},
|
| 3215 |
+
{
|
| 3216 |
+
"epoch": 12.197136563876652,
|
| 3217 |
+
"grad_norm": 0.0707065761089325,
|
| 3218 |
+
"learning_rate": 9.153846153846155e-06,
|
| 3219 |
+
"loss": 0.0146,
|
| 3220 |
+
"step": 11075
|
| 3221 |
+
},
|
| 3222 |
+
{
|
| 3223 |
+
"epoch": 12.224669603524228,
|
| 3224 |
+
"grad_norm": 0.20994262397289276,
|
| 3225 |
+
"learning_rate": 9.128205128205129e-06,
|
| 3226 |
+
"loss": 0.0118,
|
| 3227 |
+
"step": 11100
|
| 3228 |
+
},
|
| 3229 |
+
{
|
| 3230 |
+
"epoch": 12.252202643171806,
|
| 3231 |
+
"grad_norm": 0.04189275577664375,
|
| 3232 |
+
"learning_rate": 9.102564102564104e-06,
|
| 3233 |
+
"loss": 0.0098,
|
| 3234 |
+
"step": 11125
|
| 3235 |
+
},
|
| 3236 |
+
{
|
| 3237 |
+
"epoch": 12.279735682819384,
|
| 3238 |
+
"grad_norm": 0.2779877781867981,
|
| 3239 |
+
"learning_rate": 9.076923076923078e-06,
|
| 3240 |
+
"loss": 0.0132,
|
| 3241 |
+
"step": 11150
|
| 3242 |
+
},
|
| 3243 |
+
{
|
| 3244 |
+
"epoch": 12.30726872246696,
|
| 3245 |
+
"grad_norm": 0.04877633601427078,
|
| 3246 |
+
"learning_rate": 9.051282051282051e-06,
|
| 3247 |
+
"loss": 0.0125,
|
| 3248 |
+
"step": 11175
|
| 3249 |
+
},
|
| 3250 |
+
{
|
| 3251 |
+
"epoch": 12.334801762114537,
|
| 3252 |
+
"grad_norm": 0.15319667756557465,
|
| 3253 |
+
"learning_rate": 9.025641025641027e-06,
|
| 3254 |
+
"loss": 0.0111,
|
| 3255 |
+
"step": 11200
|
| 3256 |
+
},
|
| 3257 |
+
{
|
| 3258 |
+
"epoch": 12.362334801762115,
|
| 3259 |
+
"grad_norm": 0.5198453664779663,
|
| 3260 |
+
"learning_rate": 9e-06,
|
| 3261 |
+
"loss": 0.0119,
|
| 3262 |
+
"step": 11225
|
| 3263 |
+
},
|
| 3264 |
+
{
|
| 3265 |
+
"epoch": 12.389867841409691,
|
| 3266 |
+
"grad_norm": 0.1989358365535736,
|
| 3267 |
+
"learning_rate": 8.974358974358976e-06,
|
| 3268 |
+
"loss": 0.0112,
|
| 3269 |
+
"step": 11250
|
| 3270 |
+
},
|
| 3271 |
+
{
|
| 3272 |
+
"epoch": 12.417400881057269,
|
| 3273 |
+
"grad_norm": 0.5313608646392822,
|
| 3274 |
+
"learning_rate": 8.94871794871795e-06,
|
| 3275 |
+
"loss": 0.0094,
|
| 3276 |
+
"step": 11275
|
| 3277 |
+
},
|
| 3278 |
+
{
|
| 3279 |
+
"epoch": 12.444933920704846,
|
| 3280 |
+
"grad_norm": 0.17706328630447388,
|
| 3281 |
+
"learning_rate": 8.923076923076925e-06,
|
| 3282 |
+
"loss": 0.0128,
|
| 3283 |
+
"step": 11300
|
| 3284 |
+
},
|
| 3285 |
+
{
|
| 3286 |
+
"epoch": 12.472466960352422,
|
| 3287 |
+
"grad_norm": 0.45485684275627136,
|
| 3288 |
+
"learning_rate": 8.897435897435898e-06,
|
| 3289 |
+
"loss": 0.0136,
|
| 3290 |
+
"step": 11325
|
| 3291 |
+
},
|
| 3292 |
+
{
|
| 3293 |
+
"epoch": 12.5,
|
| 3294 |
+
"grad_norm": 0.14598797261714935,
|
| 3295 |
+
"learning_rate": 8.871794871794872e-06,
|
| 3296 |
+
"loss": 0.0129,
|
| 3297 |
+
"step": 11350
|
| 3298 |
+
},
|
| 3299 |
+
{
|
| 3300 |
+
"epoch": 12.527533039647578,
|
| 3301 |
+
"grad_norm": 0.20234304666519165,
|
| 3302 |
+
"learning_rate": 8.846153846153847e-06,
|
| 3303 |
+
"loss": 0.0133,
|
| 3304 |
+
"step": 11375
|
| 3305 |
+
},
|
| 3306 |
+
{
|
| 3307 |
+
"epoch": 12.555066079295154,
|
| 3308 |
+
"grad_norm": 0.15167205035686493,
|
| 3309 |
+
"learning_rate": 8.820512820512821e-06,
|
| 3310 |
+
"loss": 0.013,
|
| 3311 |
+
"step": 11400
|
| 3312 |
+
},
|
| 3313 |
+
{
|
| 3314 |
+
"epoch": 12.582599118942731,
|
| 3315 |
+
"grad_norm": 0.10332977026700974,
|
| 3316 |
+
"learning_rate": 8.794871794871796e-06,
|
| 3317 |
+
"loss": 0.012,
|
| 3318 |
+
"step": 11425
|
| 3319 |
+
},
|
| 3320 |
+
{
|
| 3321 |
+
"epoch": 12.610132158590309,
|
| 3322 |
+
"grad_norm": 0.06746497750282288,
|
| 3323 |
+
"learning_rate": 8.76923076923077e-06,
|
| 3324 |
+
"loss": 0.0102,
|
| 3325 |
+
"step": 11450
|
| 3326 |
+
},
|
| 3327 |
+
{
|
| 3328 |
+
"epoch": 12.637665198237885,
|
| 3329 |
+
"grad_norm": 0.21585692465305328,
|
| 3330 |
+
"learning_rate": 8.743589743589743e-06,
|
| 3331 |
+
"loss": 0.0131,
|
| 3332 |
+
"step": 11475
|
| 3333 |
+
},
|
| 3334 |
+
{
|
| 3335 |
+
"epoch": 12.665198237885463,
|
| 3336 |
+
"grad_norm": 0.11972617357969284,
|
| 3337 |
+
"learning_rate": 8.717948717948719e-06,
|
| 3338 |
+
"loss": 0.0124,
|
| 3339 |
+
"step": 11500
|
| 3340 |
+
},
|
| 3341 |
+
{
|
| 3342 |
+
"epoch": 12.69273127753304,
|
| 3343 |
+
"grad_norm": 0.16483718156814575,
|
| 3344 |
+
"learning_rate": 8.692307692307692e-06,
|
| 3345 |
+
"loss": 0.0136,
|
| 3346 |
+
"step": 11525
|
| 3347 |
+
},
|
| 3348 |
+
{
|
| 3349 |
+
"epoch": 12.720264317180616,
|
| 3350 |
+
"grad_norm": 0.04187220335006714,
|
| 3351 |
+
"learning_rate": 8.666666666666668e-06,
|
| 3352 |
+
"loss": 0.0122,
|
| 3353 |
+
"step": 11550
|
| 3354 |
+
},
|
| 3355 |
+
{
|
| 3356 |
+
"epoch": 12.747797356828194,
|
| 3357 |
+
"grad_norm": 0.37281060218811035,
|
| 3358 |
+
"learning_rate": 8.641025641025641e-06,
|
| 3359 |
+
"loss": 0.0117,
|
| 3360 |
+
"step": 11575
|
| 3361 |
+
},
|
| 3362 |
+
{
|
| 3363 |
+
"epoch": 12.775330396475772,
|
| 3364 |
+
"grad_norm": 0.42664897441864014,
|
| 3365 |
+
"learning_rate": 8.615384615384617e-06,
|
| 3366 |
+
"loss": 0.0132,
|
| 3367 |
+
"step": 11600
|
| 3368 |
+
},
|
| 3369 |
+
{
|
| 3370 |
+
"epoch": 12.802863436123348,
|
| 3371 |
+
"grad_norm": 0.08013510704040527,
|
| 3372 |
+
"learning_rate": 8.58974358974359e-06,
|
| 3373 |
+
"loss": 0.0129,
|
| 3374 |
+
"step": 11625
|
| 3375 |
+
},
|
| 3376 |
+
{
|
| 3377 |
+
"epoch": 12.830396475770925,
|
| 3378 |
+
"grad_norm": 0.24972140789031982,
|
| 3379 |
+
"learning_rate": 8.564102564102564e-06,
|
| 3380 |
+
"loss": 0.0132,
|
| 3381 |
+
"step": 11650
|
| 3382 |
+
},
|
| 3383 |
+
{
|
| 3384 |
+
"epoch": 12.857929515418503,
|
| 3385 |
+
"grad_norm": 1.2854641675949097,
|
| 3386 |
+
"learning_rate": 8.53846153846154e-06,
|
| 3387 |
+
"loss": 0.0129,
|
| 3388 |
+
"step": 11675
|
| 3389 |
+
},
|
| 3390 |
+
{
|
| 3391 |
+
"epoch": 12.885462555066079,
|
| 3392 |
+
"grad_norm": 0.5620148777961731,
|
| 3393 |
+
"learning_rate": 8.512820512820513e-06,
|
| 3394 |
+
"loss": 0.0165,
|
| 3395 |
+
"step": 11700
|
| 3396 |
+
},
|
| 3397 |
+
{
|
| 3398 |
+
"epoch": 12.912995594713657,
|
| 3399 |
+
"grad_norm": 0.1447388082742691,
|
| 3400 |
+
"learning_rate": 8.487179487179488e-06,
|
| 3401 |
+
"loss": 0.0152,
|
| 3402 |
+
"step": 11725
|
| 3403 |
+
},
|
| 3404 |
+
{
|
| 3405 |
+
"epoch": 12.940528634361234,
|
| 3406 |
+
"grad_norm": 0.17876878380775452,
|
| 3407 |
+
"learning_rate": 8.461538461538462e-06,
|
| 3408 |
+
"loss": 0.015,
|
| 3409 |
+
"step": 11750
|
| 3410 |
+
},
|
| 3411 |
+
{
|
| 3412 |
+
"epoch": 12.96806167400881,
|
| 3413 |
+
"grad_norm": 0.29610171914100647,
|
| 3414 |
+
"learning_rate": 8.435897435897436e-06,
|
| 3415 |
+
"loss": 0.0153,
|
| 3416 |
+
"step": 11775
|
| 3417 |
+
},
|
| 3418 |
+
{
|
| 3419 |
+
"epoch": 12.995594713656388,
|
| 3420 |
+
"grad_norm": 0.1498490869998932,
|
| 3421 |
+
"learning_rate": 8.410256410256411e-06,
|
| 3422 |
+
"loss": 0.0144,
|
| 3423 |
+
"step": 11800
|
| 3424 |
+
},
|
| 3425 |
+
{
|
| 3426 |
+
"epoch": 13.023127753303966,
|
| 3427 |
+
"grad_norm": 0.15733398497104645,
|
| 3428 |
+
"learning_rate": 8.384615384615385e-06,
|
| 3429 |
+
"loss": 0.0099,
|
| 3430 |
+
"step": 11825
|
| 3431 |
+
},
|
| 3432 |
+
{
|
| 3433 |
+
"epoch": 13.050660792951541,
|
| 3434 |
+
"grad_norm": 0.13376560807228088,
|
| 3435 |
+
"learning_rate": 8.35897435897436e-06,
|
| 3436 |
+
"loss": 0.0093,
|
| 3437 |
+
"step": 11850
|
| 3438 |
+
},
|
| 3439 |
+
{
|
| 3440 |
+
"epoch": 13.07819383259912,
|
| 3441 |
+
"grad_norm": 0.30212774872779846,
|
| 3442 |
+
"learning_rate": 8.333333333333334e-06,
|
| 3443 |
+
"loss": 0.0117,
|
| 3444 |
+
"step": 11875
|
| 3445 |
+
},
|
| 3446 |
+
{
|
| 3447 |
+
"epoch": 13.105726872246697,
|
| 3448 |
+
"grad_norm": 0.6191049814224243,
|
| 3449 |
+
"learning_rate": 8.307692307692309e-06,
|
| 3450 |
+
"loss": 0.0088,
|
| 3451 |
+
"step": 11900
|
| 3452 |
+
},
|
| 3453 |
+
{
|
| 3454 |
+
"epoch": 13.133259911894273,
|
| 3455 |
+
"grad_norm": 0.12833338975906372,
|
| 3456 |
+
"learning_rate": 8.282051282051283e-06,
|
| 3457 |
+
"loss": 0.0081,
|
| 3458 |
+
"step": 11925
|
| 3459 |
+
},
|
| 3460 |
+
{
|
| 3461 |
+
"epoch": 13.16079295154185,
|
| 3462 |
+
"grad_norm": 0.02582469768822193,
|
| 3463 |
+
"learning_rate": 8.256410256410256e-06,
|
| 3464 |
+
"loss": 0.0075,
|
| 3465 |
+
"step": 11950
|
| 3466 |
+
},
|
| 3467 |
+
{
|
| 3468 |
+
"epoch": 13.188325991189428,
|
| 3469 |
+
"grad_norm": 0.40551483631134033,
|
| 3470 |
+
"learning_rate": 8.230769230769232e-06,
|
| 3471 |
+
"loss": 0.0106,
|
| 3472 |
+
"step": 11975
|
| 3473 |
+
},
|
| 3474 |
+
{
|
| 3475 |
+
"epoch": 13.215859030837004,
|
| 3476 |
+
"grad_norm": 0.178267240524292,
|
| 3477 |
+
"learning_rate": 8.205128205128205e-06,
|
| 3478 |
+
"loss": 0.0105,
|
| 3479 |
+
"step": 12000
|
| 3480 |
+
},
|
| 3481 |
+
{
|
| 3482 |
+
"epoch": 13.215859030837004,
|
| 3483 |
+
"eval_cer": 23.92385909002327,
|
| 3484 |
+
"eval_loss": 0.8548922538757324,
|
| 3485 |
+
"eval_runtime": 1301.0663,
|
| 3486 |
+
"eval_samples_per_second": 8.133,
|
| 3487 |
+
"eval_steps_per_second": 2.034,
|
| 3488 |
+
"eval_wer": 83.48892032060348,
|
| 3489 |
+
"step": 12000
|
| 3490 |
+
},
|
| 3491 |
+
{
|
| 3492 |
+
"epoch": 13.243392070484582,
|
| 3493 |
+
"grad_norm": 0.25860708951950073,
|
| 3494 |
+
"learning_rate": 8.17948717948718e-06,
|
| 3495 |
+
"loss": 0.0107,
|
| 3496 |
+
"step": 12025
|
| 3497 |
+
},
|
| 3498 |
+
{
|
| 3499 |
+
"epoch": 13.270925110132158,
|
| 3500 |
+
"grad_norm": 0.0770430937409401,
|
| 3501 |
+
"learning_rate": 8.153846153846154e-06,
|
| 3502 |
+
"loss": 0.011,
|
| 3503 |
+
"step": 12050
|
| 3504 |
+
},
|
| 3505 |
+
{
|
| 3506 |
+
"epoch": 13.298458149779735,
|
| 3507 |
+
"grad_norm": 0.2660870850086212,
|
| 3508 |
+
"learning_rate": 8.12820512820513e-06,
|
| 3509 |
+
"loss": 0.0105,
|
| 3510 |
+
"step": 12075
|
| 3511 |
+
},
|
| 3512 |
+
{
|
| 3513 |
+
"epoch": 13.325991189427313,
|
| 3514 |
+
"grad_norm": 0.0884290263056755,
|
| 3515 |
+
"learning_rate": 8.102564102564103e-06,
|
| 3516 |
+
"loss": 0.0116,
|
| 3517 |
+
"step": 12100
|
| 3518 |
+
},
|
| 3519 |
+
{
|
| 3520 |
+
"epoch": 13.353524229074889,
|
| 3521 |
+
"grad_norm": 0.10695531964302063,
|
| 3522 |
+
"learning_rate": 8.076923076923077e-06,
|
| 3523 |
+
"loss": 0.0107,
|
| 3524 |
+
"step": 12125
|
| 3525 |
+
},
|
| 3526 |
+
{
|
| 3527 |
+
"epoch": 13.381057268722467,
|
| 3528 |
+
"grad_norm": 0.03739326447248459,
|
| 3529 |
+
"learning_rate": 8.051282051282052e-06,
|
| 3530 |
+
"loss": 0.0157,
|
| 3531 |
+
"step": 12150
|
| 3532 |
+
},
|
| 3533 |
+
{
|
| 3534 |
+
"epoch": 13.408590308370044,
|
| 3535 |
+
"grad_norm": 0.4565247595310211,
|
| 3536 |
+
"learning_rate": 8.025641025641026e-06,
|
| 3537 |
+
"loss": 0.0126,
|
| 3538 |
+
"step": 12175
|
| 3539 |
+
},
|
| 3540 |
+
{
|
| 3541 |
+
"epoch": 13.43612334801762,
|
| 3542 |
+
"grad_norm": 0.06541293859481812,
|
| 3543 |
+
"learning_rate": 8.000000000000001e-06,
|
| 3544 |
+
"loss": 0.0131,
|
| 3545 |
+
"step": 12200
|
| 3546 |
+
},
|
| 3547 |
+
{
|
| 3548 |
+
"epoch": 13.463656387665198,
|
| 3549 |
+
"grad_norm": 0.10597793757915497,
|
| 3550 |
+
"learning_rate": 7.974358974358975e-06,
|
| 3551 |
+
"loss": 0.0107,
|
| 3552 |
+
"step": 12225
|
| 3553 |
+
},
|
| 3554 |
+
{
|
| 3555 |
+
"epoch": 13.491189427312776,
|
| 3556 |
+
"grad_norm": 0.2990114390850067,
|
| 3557 |
+
"learning_rate": 7.948717948717949e-06,
|
| 3558 |
+
"loss": 0.0107,
|
| 3559 |
+
"step": 12250
|
| 3560 |
+
},
|
| 3561 |
+
{
|
| 3562 |
+
"epoch": 13.518722466960352,
|
| 3563 |
+
"grad_norm": 0.1924736499786377,
|
| 3564 |
+
"learning_rate": 7.923076923076924e-06,
|
| 3565 |
+
"loss": 0.0126,
|
| 3566 |
+
"step": 12275
|
| 3567 |
+
},
|
| 3568 |
+
{
|
| 3569 |
+
"epoch": 13.54625550660793,
|
| 3570 |
+
"grad_norm": 0.13384470343589783,
|
| 3571 |
+
"learning_rate": 7.897435897435898e-06,
|
| 3572 |
+
"loss": 0.0144,
|
| 3573 |
+
"step": 12300
|
| 3574 |
+
},
|
| 3575 |
+
{
|
| 3576 |
+
"epoch": 13.573788546255507,
|
| 3577 |
+
"grad_norm": 0.195987269282341,
|
| 3578 |
+
"learning_rate": 7.871794871794873e-06,
|
| 3579 |
+
"loss": 0.011,
|
| 3580 |
+
"step": 12325
|
| 3581 |
+
},
|
| 3582 |
+
{
|
| 3583 |
+
"epoch": 13.601321585903083,
|
| 3584 |
+
"grad_norm": 0.6262577772140503,
|
| 3585 |
+
"learning_rate": 7.846153846153847e-06,
|
| 3586 |
+
"loss": 0.0102,
|
| 3587 |
+
"step": 12350
|
| 3588 |
+
},
|
| 3589 |
+
{
|
| 3590 |
+
"epoch": 13.62885462555066,
|
| 3591 |
+
"grad_norm": 0.09502260386943817,
|
| 3592 |
+
"learning_rate": 7.820512820512822e-06,
|
| 3593 |
+
"loss": 0.0129,
|
| 3594 |
+
"step": 12375
|
| 3595 |
+
},
|
| 3596 |
+
{
|
| 3597 |
+
"epoch": 13.656387665198238,
|
| 3598 |
+
"grad_norm": 0.14882908761501312,
|
| 3599 |
+
"learning_rate": 7.794871794871796e-06,
|
| 3600 |
+
"loss": 0.0103,
|
| 3601 |
+
"step": 12400
|
| 3602 |
+
},
|
| 3603 |
+
{
|
| 3604 |
+
"epoch": 13.683920704845814,
|
| 3605 |
+
"grad_norm": 0.421539306640625,
|
| 3606 |
+
"learning_rate": 7.76923076923077e-06,
|
| 3607 |
+
"loss": 0.0092,
|
| 3608 |
+
"step": 12425
|
| 3609 |
+
},
|
| 3610 |
+
{
|
| 3611 |
+
"epoch": 13.711453744493392,
|
| 3612 |
+
"grad_norm": 0.04390239343047142,
|
| 3613 |
+
"learning_rate": 7.743589743589745e-06,
|
| 3614 |
+
"loss": 0.0141,
|
| 3615 |
+
"step": 12450
|
| 3616 |
+
},
|
| 3617 |
+
{
|
| 3618 |
+
"epoch": 13.73898678414097,
|
| 3619 |
+
"grad_norm": 0.14362525939941406,
|
| 3620 |
+
"learning_rate": 7.717948717948718e-06,
|
| 3621 |
+
"loss": 0.0108,
|
| 3622 |
+
"step": 12475
|
| 3623 |
+
},
|
| 3624 |
+
{
|
| 3625 |
+
"epoch": 13.766519823788546,
|
| 3626 |
+
"grad_norm": 0.25303810834884644,
|
| 3627 |
+
"learning_rate": 7.692307692307694e-06,
|
| 3628 |
+
"loss": 0.0151,
|
| 3629 |
+
"step": 12500
|
| 3630 |
+
},
|
| 3631 |
+
{
|
| 3632 |
+
"epoch": 13.794052863436123,
|
| 3633 |
+
"grad_norm": 0.37157806754112244,
|
| 3634 |
+
"learning_rate": 7.666666666666667e-06,
|
| 3635 |
+
"loss": 0.0136,
|
| 3636 |
+
"step": 12525
|
| 3637 |
+
},
|
| 3638 |
+
{
|
| 3639 |
+
"epoch": 13.821585903083701,
|
| 3640 |
+
"grad_norm": 0.36168116331100464,
|
| 3641 |
+
"learning_rate": 7.641025641025641e-06,
|
| 3642 |
+
"loss": 0.0139,
|
| 3643 |
+
"step": 12550
|
| 3644 |
+
},
|
| 3645 |
+
{
|
| 3646 |
+
"epoch": 13.849118942731277,
|
| 3647 |
+
"grad_norm": 0.32778996229171753,
|
| 3648 |
+
"learning_rate": 7.615384615384615e-06,
|
| 3649 |
+
"loss": 0.0111,
|
| 3650 |
+
"step": 12575
|
| 3651 |
+
},
|
| 3652 |
+
{
|
| 3653 |
+
"epoch": 13.876651982378855,
|
| 3654 |
+
"grad_norm": 0.16989374160766602,
|
| 3655 |
+
"learning_rate": 7.58974358974359e-06,
|
| 3656 |
+
"loss": 0.0111,
|
| 3657 |
+
"step": 12600
|
| 3658 |
+
},
|
| 3659 |
+
{
|
| 3660 |
+
"epoch": 13.904185022026432,
|
| 3661 |
+
"grad_norm": 0.05724957212805748,
|
| 3662 |
+
"learning_rate": 7.564102564102564e-06,
|
| 3663 |
+
"loss": 0.0125,
|
| 3664 |
+
"step": 12625
|
| 3665 |
+
},
|
| 3666 |
+
{
|
| 3667 |
+
"epoch": 13.931718061674008,
|
| 3668 |
+
"grad_norm": 0.35101988911628723,
|
| 3669 |
+
"learning_rate": 7.538461538461539e-06,
|
| 3670 |
+
"loss": 0.0147,
|
| 3671 |
+
"step": 12650
|
| 3672 |
+
},
|
| 3673 |
+
{
|
| 3674 |
+
"epoch": 13.959251101321586,
|
| 3675 |
+
"grad_norm": 0.0790194496512413,
|
| 3676 |
+
"learning_rate": 7.512820512820513e-06,
|
| 3677 |
+
"loss": 0.013,
|
| 3678 |
+
"step": 12675
|
| 3679 |
+
},
|
| 3680 |
+
{
|
| 3681 |
+
"epoch": 13.986784140969164,
|
| 3682 |
+
"grad_norm": 0.2358320951461792,
|
| 3683 |
+
"learning_rate": 7.487179487179488e-06,
|
| 3684 |
+
"loss": 0.0147,
|
| 3685 |
+
"step": 12700
|
| 3686 |
+
},
|
| 3687 |
+
{
|
| 3688 |
+
"epoch": 14.01431718061674,
|
| 3689 |
+
"grad_norm": 0.06501065939664841,
|
| 3690 |
+
"learning_rate": 7.461538461538462e-06,
|
| 3691 |
+
"loss": 0.0094,
|
| 3692 |
+
"step": 12725
|
| 3693 |
+
},
|
| 3694 |
+
{
|
| 3695 |
+
"epoch": 14.041850220264317,
|
| 3696 |
+
"grad_norm": 0.028322290629148483,
|
| 3697 |
+
"learning_rate": 7.435897435897437e-06,
|
| 3698 |
+
"loss": 0.01,
|
| 3699 |
+
"step": 12750
|
| 3700 |
+
},
|
| 3701 |
+
{
|
| 3702 |
+
"epoch": 14.069383259911895,
|
| 3703 |
+
"grad_norm": 0.04854853078722954,
|
| 3704 |
+
"learning_rate": 7.410256410256411e-06,
|
| 3705 |
+
"loss": 0.0109,
|
| 3706 |
+
"step": 12775
|
| 3707 |
+
},
|
| 3708 |
+
{
|
| 3709 |
+
"epoch": 14.09691629955947,
|
| 3710 |
+
"grad_norm": 0.042353082448244095,
|
| 3711 |
+
"learning_rate": 7.384615384615386e-06,
|
| 3712 |
+
"loss": 0.0069,
|
| 3713 |
+
"step": 12800
|
| 3714 |
+
},
|
| 3715 |
+
{
|
| 3716 |
+
"epoch": 14.124449339207048,
|
| 3717 |
+
"grad_norm": 0.24730762839317322,
|
| 3718 |
+
"learning_rate": 7.35897435897436e-06,
|
| 3719 |
+
"loss": 0.009,
|
| 3720 |
+
"step": 12825
|
| 3721 |
+
},
|
| 3722 |
+
{
|
| 3723 |
+
"epoch": 14.151982378854626,
|
| 3724 |
+
"grad_norm": 0.13546280562877655,
|
| 3725 |
+
"learning_rate": 7.333333333333333e-06,
|
| 3726 |
+
"loss": 0.0082,
|
| 3727 |
+
"step": 12850
|
| 3728 |
+
},
|
| 3729 |
+
{
|
| 3730 |
+
"epoch": 14.179515418502202,
|
| 3731 |
+
"grad_norm": 0.1111743226647377,
|
| 3732 |
+
"learning_rate": 7.307692307692308e-06,
|
| 3733 |
+
"loss": 0.0083,
|
| 3734 |
+
"step": 12875
|
| 3735 |
+
},
|
| 3736 |
+
{
|
| 3737 |
+
"epoch": 14.20704845814978,
|
| 3738 |
+
"grad_norm": 0.0262732096016407,
|
| 3739 |
+
"learning_rate": 7.282051282051282e-06,
|
| 3740 |
+
"loss": 0.007,
|
| 3741 |
+
"step": 12900
|
| 3742 |
+
},
|
| 3743 |
+
{
|
| 3744 |
+
"epoch": 14.234581497797357,
|
| 3745 |
+
"grad_norm": 0.3866499960422516,
|
| 3746 |
+
"learning_rate": 7.256410256410257e-06,
|
| 3747 |
+
"loss": 0.008,
|
| 3748 |
+
"step": 12925
|
| 3749 |
+
},
|
| 3750 |
+
{
|
| 3751 |
+
"epoch": 14.262114537444933,
|
| 3752 |
+
"grad_norm": 0.07846901565790176,
|
| 3753 |
+
"learning_rate": 7.230769230769231e-06,
|
| 3754 |
+
"loss": 0.0069,
|
| 3755 |
+
"step": 12950
|
| 3756 |
+
},
|
| 3757 |
+
{
|
| 3758 |
+
"epoch": 14.289647577092511,
|
| 3759 |
+
"grad_norm": 0.044436316937208176,
|
| 3760 |
+
"learning_rate": 7.205128205128206e-06,
|
| 3761 |
+
"loss": 0.0081,
|
| 3762 |
+
"step": 12975
|
| 3763 |
+
},
|
| 3764 |
+
{
|
| 3765 |
+
"epoch": 14.317180616740089,
|
| 3766 |
+
"grad_norm": 0.18789108097553253,
|
| 3767 |
+
"learning_rate": 7.17948717948718e-06,
|
| 3768 |
+
"loss": 0.0074,
|
| 3769 |
+
"step": 13000
|
| 3770 |
+
},
|
| 3771 |
+
{
|
| 3772 |
+
"epoch": 14.317180616740089,
|
| 3773 |
+
"eval_cer": 24.13134937466326,
|
| 3774 |
+
"eval_loss": 0.8608656525611877,
|
| 3775 |
+
"eval_runtime": 1301.5926,
|
| 3776 |
+
"eval_samples_per_second": 8.129,
|
| 3777 |
+
"eval_steps_per_second": 2.033,
|
| 3778 |
+
"eval_wer": 83.63979255068365,
|
| 3779 |
+
"step": 13000
|
| 3780 |
}
|
| 3781 |
],
|
| 3782 |
"logging_steps": 25,
|
|
|
|
| 3796 |
"attributes": {}
|
| 3797 |
}
|
| 3798 |
},
|
| 3799 |
+
"total_flos": 2.233586142437127e+20,
|
| 3800 |
"train_batch_size": 4,
|
| 3801 |
"trial_name": null,
|
| 3802 |
"trial_params": null
|