End of training
Browse files- README.md +1 -1
- all_results.json +22 -22
- eval_nbest_predictions.json +2 -2
- eval_null_odds.json +0 -0
- eval_predictions.json +0 -0
- eval_results.json +16 -16
- train_results.json +7 -7
- trainer_state.json +1055 -12
README.md
CHANGED
|
@@ -14,7 +14,7 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 14 |
|
| 15 |
# tiny-clinicalbert-qa
|
| 16 |
|
| 17 |
-
This model is a fine-tuned version of [nlpie/tiny-clinicalbert](https://huggingface.co/nlpie/tiny-clinicalbert) on the
|
| 18 |
|
| 19 |
## Model description
|
| 20 |
|
|
|
|
| 14 |
|
| 15 |
# tiny-clinicalbert-qa
|
| 16 |
|
| 17 |
+
This model is a fine-tuned version of [nlpie/tiny-clinicalbert](https://huggingface.co/nlpie/tiny-clinicalbert) on the Eladio/emrqa-msquad and the rajpurkar/squad_v2 datasets.
|
| 18 |
|
| 19 |
## Model description
|
| 20 |
|
all_results.json
CHANGED
|
@@ -1,26 +1,26 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_HasAns_exact":
|
| 4 |
-
"eval_HasAns_f1":
|
| 5 |
-
"eval_HasAns_total":
|
| 6 |
-
"eval_NoAns_exact":
|
| 7 |
-
"eval_NoAns_f1":
|
| 8 |
-
"eval_NoAns_total":
|
| 9 |
-
"eval_best_exact":
|
| 10 |
"eval_best_exact_thresh": 0.0,
|
| 11 |
-
"eval_best_f1":
|
| 12 |
"eval_best_f1_thresh": 0.0,
|
| 13 |
-
"eval_exact":
|
| 14 |
-
"eval_f1":
|
| 15 |
-
"eval_runtime":
|
| 16 |
-
"eval_samples":
|
| 17 |
-
"eval_samples_per_second":
|
| 18 |
-
"eval_steps_per_second":
|
| 19 |
-
"eval_total":
|
| 20 |
-
"total_flos":
|
| 21 |
-
"train_loss":
|
| 22 |
-
"train_runtime":
|
| 23 |
-
"train_samples":
|
| 24 |
-
"train_samples_per_second":
|
| 25 |
-
"train_steps_per_second":
|
| 26 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_HasAns_exact": 77.5803656865027,
|
| 4 |
+
"eval_HasAns_f1": 86.16666643213416,
|
| 5 |
+
"eval_HasAns_total": 38667,
|
| 6 |
+
"eval_NoAns_exact": 56.13120269133726,
|
| 7 |
+
"eval_NoAns_f1": 56.13120269133726,
|
| 8 |
+
"eval_NoAns_total": 5945,
|
| 9 |
+
"eval_best_exact": 74.72204787949431,
|
| 10 |
"eval_best_exact_thresh": 0.0,
|
| 11 |
+
"eval_best_f1": 82.16413724852865,
|
| 12 |
"eval_best_f1_thresh": 0.0,
|
| 13 |
+
"eval_exact": 74.72204787949431,
|
| 14 |
+
"eval_f1": 82.16413724852801,
|
| 15 |
+
"eval_runtime": 366.7395,
|
| 16 |
+
"eval_samples": 78873,
|
| 17 |
+
"eval_samples_per_second": 215.065,
|
| 18 |
+
"eval_steps_per_second": 26.886,
|
| 19 |
+
"eval_total": 44612,
|
| 20 |
+
"total_flos": 1.6739482297393152e+16,
|
| 21 |
+
"train_loss": 1.0437313645233104,
|
| 22 |
+
"train_runtime": 12605.9624,
|
| 23 |
+
"train_samples": 397452,
|
| 24 |
+
"train_samples_per_second": 94.587,
|
| 25 |
+
"train_steps_per_second": 5.912
|
| 26 |
}
|
eval_nbest_predictions.json
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7779eebb1aef7212e3a442866d0c496e1218f65c1fd507c77488fc2cefb99583
|
| 3 |
+
size 216086768
|
eval_null_odds.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
eval_predictions.json
CHANGED
|
The diff for this file is too large to render.
See raw diff
|
|
|
eval_results.json
CHANGED
|
@@ -1,20 +1,20 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"eval_HasAns_exact":
|
| 4 |
-
"eval_HasAns_f1":
|
| 5 |
-
"eval_HasAns_total":
|
| 6 |
-
"eval_NoAns_exact":
|
| 7 |
-
"eval_NoAns_f1":
|
| 8 |
-
"eval_NoAns_total":
|
| 9 |
-
"eval_best_exact":
|
| 10 |
"eval_best_exact_thresh": 0.0,
|
| 11 |
-
"eval_best_f1":
|
| 12 |
"eval_best_f1_thresh": 0.0,
|
| 13 |
-
"eval_exact":
|
| 14 |
-
"eval_f1":
|
| 15 |
-
"eval_runtime":
|
| 16 |
-
"eval_samples":
|
| 17 |
-
"eval_samples_per_second":
|
| 18 |
-
"eval_steps_per_second":
|
| 19 |
-
"eval_total":
|
| 20 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"eval_HasAns_exact": 77.5803656865027,
|
| 4 |
+
"eval_HasAns_f1": 86.16666643213416,
|
| 5 |
+
"eval_HasAns_total": 38667,
|
| 6 |
+
"eval_NoAns_exact": 56.13120269133726,
|
| 7 |
+
"eval_NoAns_f1": 56.13120269133726,
|
| 8 |
+
"eval_NoAns_total": 5945,
|
| 9 |
+
"eval_best_exact": 74.72204787949431,
|
| 10 |
"eval_best_exact_thresh": 0.0,
|
| 11 |
+
"eval_best_f1": 82.16413724852865,
|
| 12 |
"eval_best_f1_thresh": 0.0,
|
| 13 |
+
"eval_exact": 74.72204787949431,
|
| 14 |
+
"eval_f1": 82.16413724852801,
|
| 15 |
+
"eval_runtime": 366.7395,
|
| 16 |
+
"eval_samples": 78873,
|
| 17 |
+
"eval_samples_per_second": 215.065,
|
| 18 |
+
"eval_steps_per_second": 26.886,
|
| 19 |
+
"eval_total": 44612
|
| 20 |
}
|
train_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"epoch":
|
| 3 |
-
"total_flos":
|
| 4 |
-
"train_loss":
|
| 5 |
-
"train_runtime":
|
| 6 |
-
"train_samples":
|
| 7 |
-
"train_samples_per_second":
|
| 8 |
-
"train_steps_per_second":
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"epoch": 3.0,
|
| 3 |
+
"total_flos": 1.6739482297393152e+16,
|
| 4 |
+
"train_loss": 1.0437313645233104,
|
| 5 |
+
"train_runtime": 12605.9624,
|
| 6 |
+
"train_samples": 397452,
|
| 7 |
+
"train_samples_per_second": 94.587,
|
| 8 |
+
"train_steps_per_second": 5.912
|
| 9 |
}
|
trainer_state.json
CHANGED
|
@@ -2,27 +2,1070 @@
|
|
| 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,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
-
"epoch":
|
| 14 |
-
"
|
| 15 |
-
"
|
| 16 |
-
"
|
| 17 |
-
"
|
| 18 |
-
|
| 19 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
}
|
| 21 |
],
|
| 22 |
"logging_steps": 500,
|
| 23 |
-
"max_steps":
|
| 24 |
"num_input_tokens_seen": 0,
|
| 25 |
-
"num_train_epochs":
|
| 26 |
"save_steps": 500,
|
| 27 |
"stateful_callbacks": {
|
| 28 |
"TrainerControl": {
|
|
@@ -36,7 +1079,7 @@
|
|
| 36 |
"attributes": {}
|
| 37 |
}
|
| 38 |
},
|
| 39 |
-
"total_flos":
|
| 40 |
"train_batch_size": 16,
|
| 41 |
"trial_name": null,
|
| 42 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 3.0,
|
| 6 |
"eval_steps": 500,
|
| 7 |
+
"global_step": 74523,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
| 11 |
"log_history": [
|
| 12 |
{
|
| 13 |
+
"epoch": 0.020128014170121975,
|
| 14 |
+
"grad_norm": 4.302577018737793,
|
| 15 |
+
"learning_rate": 7.94643264495525e-05,
|
| 16 |
+
"loss": 3.4263,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.04025602834024395,
|
| 21 |
+
"grad_norm": 4.915486812591553,
|
| 22 |
+
"learning_rate": 7.892757940501591e-05,
|
| 23 |
+
"loss": 3.0255,
|
| 24 |
+
"step": 1000
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.06038404251036593,
|
| 28 |
+
"grad_norm": 6.236612796783447,
|
| 29 |
+
"learning_rate": 7.839083236047933e-05,
|
| 30 |
+
"loss": 2.8404,
|
| 31 |
+
"step": 1500
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.0805120566804879,
|
| 35 |
+
"grad_norm": 8.648529052734375,
|
| 36 |
+
"learning_rate": 7.785408531594274e-05,
|
| 37 |
+
"loss": 2.5965,
|
| 38 |
+
"step": 2000
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.10064007085060987,
|
| 42 |
+
"grad_norm": 9.566611289978027,
|
| 43 |
+
"learning_rate": 7.731733827140614e-05,
|
| 44 |
+
"loss": 2.3771,
|
| 45 |
+
"step": 2500
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.12076808502073186,
|
| 49 |
+
"grad_norm": 11.978983879089355,
|
| 50 |
+
"learning_rate": 7.678059122686957e-05,
|
| 51 |
+
"loss": 2.2719,
|
| 52 |
+
"step": 3000
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.14089609919085383,
|
| 56 |
+
"grad_norm": 9.598978996276855,
|
| 57 |
+
"learning_rate": 7.624384418233297e-05,
|
| 58 |
+
"loss": 2.1877,
|
| 59 |
+
"step": 3500
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.1610241133609758,
|
| 63 |
+
"grad_norm": 13.149515151977539,
|
| 64 |
+
"learning_rate": 7.57070971377964e-05,
|
| 65 |
+
"loss": 2.1664,
|
| 66 |
+
"step": 4000
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.18115212753109777,
|
| 70 |
+
"grad_norm": 7.942020416259766,
|
| 71 |
+
"learning_rate": 7.51703500932598e-05,
|
| 72 |
+
"loss": 2.0778,
|
| 73 |
+
"step": 4500
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.20128014170121974,
|
| 77 |
+
"grad_norm": 9.4037504196167,
|
| 78 |
+
"learning_rate": 7.463360304872321e-05,
|
| 79 |
+
"loss": 2.064,
|
| 80 |
+
"step": 5000
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.22140815587134174,
|
| 84 |
+
"grad_norm": 8.469627380371094,
|
| 85 |
+
"learning_rate": 7.409685600418663e-05,
|
| 86 |
+
"loss": 1.9481,
|
| 87 |
+
"step": 5500
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.24153617004146372,
|
| 91 |
+
"grad_norm": 13.245635032653809,
|
| 92 |
+
"learning_rate": 7.356010895965004e-05,
|
| 93 |
+
"loss": 1.9162,
|
| 94 |
+
"step": 6000
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.26166418421158566,
|
| 98 |
+
"grad_norm": 14.120226860046387,
|
| 99 |
+
"learning_rate": 7.302336191511346e-05,
|
| 100 |
+
"loss": 1.8887,
|
| 101 |
+
"step": 6500
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.28179219838170766,
|
| 105 |
+
"grad_norm": 10.026687622070312,
|
| 106 |
+
"learning_rate": 7.248661487057687e-05,
|
| 107 |
+
"loss": 1.8418,
|
| 108 |
+
"step": 7000
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.30192021255182966,
|
| 112 |
+
"grad_norm": 11.791897773742676,
|
| 113 |
+
"learning_rate": 7.194986782604029e-05,
|
| 114 |
+
"loss": 1.8123,
|
| 115 |
+
"step": 7500
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.3220482267219516,
|
| 119 |
+
"grad_norm": 8.655497550964355,
|
| 120 |
+
"learning_rate": 7.141312078150371e-05,
|
| 121 |
+
"loss": 1.7765,
|
| 122 |
+
"step": 8000
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.3421762408920736,
|
| 126 |
+
"grad_norm": 10.105185508728027,
|
| 127 |
+
"learning_rate": 7.087637373696712e-05,
|
| 128 |
+
"loss": 1.7373,
|
| 129 |
+
"step": 8500
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.36230425506219555,
|
| 133 |
+
"grad_norm": 10.553141593933105,
|
| 134 |
+
"learning_rate": 7.033962669243054e-05,
|
| 135 |
+
"loss": 1.7426,
|
| 136 |
+
"step": 9000
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.38243226923231755,
|
| 140 |
+
"grad_norm": 10.911526679992676,
|
| 141 |
+
"learning_rate": 6.980287964789395e-05,
|
| 142 |
+
"loss": 1.666,
|
| 143 |
+
"step": 9500
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.4025602834024395,
|
| 147 |
+
"grad_norm": 14.893563270568848,
|
| 148 |
+
"learning_rate": 6.926613260335736e-05,
|
| 149 |
+
"loss": 1.664,
|
| 150 |
+
"step": 10000
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.4226882975725615,
|
| 154 |
+
"grad_norm": 11.424589157104492,
|
| 155 |
+
"learning_rate": 6.872938555882078e-05,
|
| 156 |
+
"loss": 1.6313,
|
| 157 |
+
"step": 10500
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.4428163117426835,
|
| 161 |
+
"grad_norm": 11.619711875915527,
|
| 162 |
+
"learning_rate": 6.819263851428418e-05,
|
| 163 |
+
"loss": 1.6002,
|
| 164 |
+
"step": 11000
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.46294432591280543,
|
| 168 |
+
"grad_norm": 10.775858879089355,
|
| 169 |
+
"learning_rate": 6.76558914697476e-05,
|
| 170 |
+
"loss": 1.5659,
|
| 171 |
+
"step": 11500
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.48307234008292743,
|
| 175 |
+
"grad_norm": 10.806571960449219,
|
| 176 |
+
"learning_rate": 6.711914442521101e-05,
|
| 177 |
+
"loss": 1.5198,
|
| 178 |
+
"step": 12000
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.5032003542530494,
|
| 182 |
+
"grad_norm": 9.920073509216309,
|
| 183 |
+
"learning_rate": 6.658239738067442e-05,
|
| 184 |
+
"loss": 1.4948,
|
| 185 |
+
"step": 12500
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.5233283684231713,
|
| 189 |
+
"grad_norm": 18.91466522216797,
|
| 190 |
+
"learning_rate": 6.604565033613784e-05,
|
| 191 |
+
"loss": 1.4626,
|
| 192 |
+
"step": 13000
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.5434563825932933,
|
| 196 |
+
"grad_norm": 13.546745300292969,
|
| 197 |
+
"learning_rate": 6.550890329160125e-05,
|
| 198 |
+
"loss": 1.4395,
|
| 199 |
+
"step": 13500
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.5635843967634153,
|
| 203 |
+
"grad_norm": 17.364229202270508,
|
| 204 |
+
"learning_rate": 6.497215624706467e-05,
|
| 205 |
+
"loss": 1.4607,
|
| 206 |
+
"step": 14000
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.5837124109335373,
|
| 210 |
+
"grad_norm": 6.96090841293335,
|
| 211 |
+
"learning_rate": 6.443540920252808e-05,
|
| 212 |
+
"loss": 1.4053,
|
| 213 |
+
"step": 14500
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.6038404251036593,
|
| 217 |
+
"grad_norm": 11.06962776184082,
|
| 218 |
+
"learning_rate": 6.389866215799149e-05,
|
| 219 |
+
"loss": 1.3991,
|
| 220 |
+
"step": 15000
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.6239684392737812,
|
| 224 |
+
"grad_norm": 7.865960597991943,
|
| 225 |
+
"learning_rate": 6.336191511345491e-05,
|
| 226 |
+
"loss": 1.3924,
|
| 227 |
+
"step": 15500
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.6440964534439032,
|
| 231 |
+
"grad_norm": 11.192516326904297,
|
| 232 |
+
"learning_rate": 6.282516806891833e-05,
|
| 233 |
+
"loss": 1.3526,
|
| 234 |
+
"step": 16000
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.6642244676140252,
|
| 238 |
+
"grad_norm": 16.678586959838867,
|
| 239 |
+
"learning_rate": 6.228842102438174e-05,
|
| 240 |
+
"loss": 1.3077,
|
| 241 |
+
"step": 16500
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.6843524817841472,
|
| 245 |
+
"grad_norm": 10.391363143920898,
|
| 246 |
+
"learning_rate": 6.175167397984516e-05,
|
| 247 |
+
"loss": 1.2956,
|
| 248 |
+
"step": 17000
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.7044804959542692,
|
| 252 |
+
"grad_norm": 14.913769721984863,
|
| 253 |
+
"learning_rate": 6.121492693530857e-05,
|
| 254 |
+
"loss": 1.2836,
|
| 255 |
+
"step": 17500
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.7246085101243911,
|
| 259 |
+
"grad_norm": 8.673684120178223,
|
| 260 |
+
"learning_rate": 6.067817989077198e-05,
|
| 261 |
+
"loss": 1.3183,
|
| 262 |
+
"step": 18000
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.7447365242945131,
|
| 266 |
+
"grad_norm": 16.5867977142334,
|
| 267 |
+
"learning_rate": 6.0141432846235395e-05,
|
| 268 |
+
"loss": 1.2462,
|
| 269 |
+
"step": 18500
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.7648645384646351,
|
| 273 |
+
"grad_norm": 10.297107696533203,
|
| 274 |
+
"learning_rate": 5.960468580169881e-05,
|
| 275 |
+
"loss": 1.2573,
|
| 276 |
+
"step": 19000
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.7849925526347571,
|
| 280 |
+
"grad_norm": 11.904420852661133,
|
| 281 |
+
"learning_rate": 5.9067938757162224e-05,
|
| 282 |
+
"loss": 1.2147,
|
| 283 |
+
"step": 19500
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.805120566804879,
|
| 287 |
+
"grad_norm": 11.955267906188965,
|
| 288 |
+
"learning_rate": 5.853119171262564e-05,
|
| 289 |
+
"loss": 1.2139,
|
| 290 |
+
"step": 20000
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.825248580975001,
|
| 294 |
+
"grad_norm": 15.51834487915039,
|
| 295 |
+
"learning_rate": 5.7994444668089053e-05,
|
| 296 |
+
"loss": 1.1777,
|
| 297 |
+
"step": 20500
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.845376595145123,
|
| 301 |
+
"grad_norm": 18.978769302368164,
|
| 302 |
+
"learning_rate": 5.745769762355246e-05,
|
| 303 |
+
"loss": 1.18,
|
| 304 |
+
"step": 21000
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.865504609315245,
|
| 308 |
+
"grad_norm": 10.92957592010498,
|
| 309 |
+
"learning_rate": 5.692095057901588e-05,
|
| 310 |
+
"loss": 1.1882,
|
| 311 |
+
"step": 21500
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.885632623485367,
|
| 315 |
+
"grad_norm": 13.877032279968262,
|
| 316 |
+
"learning_rate": 5.638420353447929e-05,
|
| 317 |
+
"loss": 1.1327,
|
| 318 |
+
"step": 22000
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.9057606376554889,
|
| 322 |
+
"grad_norm": 10.083035469055176,
|
| 323 |
+
"learning_rate": 5.584745648994271e-05,
|
| 324 |
+
"loss": 1.1209,
|
| 325 |
+
"step": 22500
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.9258886518256109,
|
| 329 |
+
"grad_norm": 11.10254955291748,
|
| 330 |
+
"learning_rate": 5.531070944540612e-05,
|
| 331 |
+
"loss": 1.1248,
|
| 332 |
+
"step": 23000
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.9460166659957329,
|
| 336 |
+
"grad_norm": 10.709171295166016,
|
| 337 |
+
"learning_rate": 5.4773962400869534e-05,
|
| 338 |
+
"loss": 1.1159,
|
| 339 |
+
"step": 23500
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.9661446801658549,
|
| 343 |
+
"grad_norm": 6.209949016571045,
|
| 344 |
+
"learning_rate": 5.423721535633295e-05,
|
| 345 |
+
"loss": 1.1064,
|
| 346 |
+
"step": 24000
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.9862726943359769,
|
| 350 |
+
"grad_norm": 9.580244064331055,
|
| 351 |
+
"learning_rate": 5.370046831179636e-05,
|
| 352 |
+
"loss": 1.099,
|
| 353 |
+
"step": 24500
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 1.0064007085060989,
|
| 357 |
+
"grad_norm": 7.485245704650879,
|
| 358 |
+
"learning_rate": 5.316372126725978e-05,
|
| 359 |
+
"loss": 1.0562,
|
| 360 |
+
"step": 25000
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 1.0265287226762208,
|
| 364 |
+
"grad_norm": 12.7084321975708,
|
| 365 |
+
"learning_rate": 5.262697422272319e-05,
|
| 366 |
+
"loss": 1.0241,
|
| 367 |
+
"step": 25500
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 1.0466567368463426,
|
| 371 |
+
"grad_norm": 11.621143341064453,
|
| 372 |
+
"learning_rate": 5.20902271781866e-05,
|
| 373 |
+
"loss": 0.9977,
|
| 374 |
+
"step": 26000
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 1.0667847510164647,
|
| 378 |
+
"grad_norm": 11.446859359741211,
|
| 379 |
+
"learning_rate": 5.155348013365002e-05,
|
| 380 |
+
"loss": 0.9963,
|
| 381 |
+
"step": 26500
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 1.0869127651865866,
|
| 385 |
+
"grad_norm": 18.473350524902344,
|
| 386 |
+
"learning_rate": 5.101673308911343e-05,
|
| 387 |
+
"loss": 1.0238,
|
| 388 |
+
"step": 27000
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 1.1070407793567087,
|
| 392 |
+
"grad_norm": 9.876648902893066,
|
| 393 |
+
"learning_rate": 5.047998604457685e-05,
|
| 394 |
+
"loss": 0.9826,
|
| 395 |
+
"step": 27500
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 1.1271687935268306,
|
| 399 |
+
"grad_norm": 8.748170852661133,
|
| 400 |
+
"learning_rate": 4.994323900004026e-05,
|
| 401 |
+
"loss": 0.9525,
|
| 402 |
+
"step": 28000
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 1.1472968076969525,
|
| 406 |
+
"grad_norm": 15.020739555358887,
|
| 407 |
+
"learning_rate": 4.940649195550367e-05,
|
| 408 |
+
"loss": 0.9606,
|
| 409 |
+
"step": 28500
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 1.1674248218670746,
|
| 413 |
+
"grad_norm": 12.53282642364502,
|
| 414 |
+
"learning_rate": 4.886974491096709e-05,
|
| 415 |
+
"loss": 0.9428,
|
| 416 |
+
"step": 29000
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 1.1875528360371965,
|
| 420 |
+
"grad_norm": 22.736968994140625,
|
| 421 |
+
"learning_rate": 4.83329978664305e-05,
|
| 422 |
+
"loss": 0.9665,
|
| 423 |
+
"step": 29500
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 1.2076808502073186,
|
| 427 |
+
"grad_norm": 10.772764205932617,
|
| 428 |
+
"learning_rate": 4.779625082189392e-05,
|
| 429 |
+
"loss": 0.9221,
|
| 430 |
+
"step": 30000
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 1.2278088643774405,
|
| 434 |
+
"grad_norm": 19.697633743286133,
|
| 435 |
+
"learning_rate": 4.725950377735733e-05,
|
| 436 |
+
"loss": 0.9411,
|
| 437 |
+
"step": 30500
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 1.2479368785475624,
|
| 441 |
+
"grad_norm": 12.230613708496094,
|
| 442 |
+
"learning_rate": 4.672275673282074e-05,
|
| 443 |
+
"loss": 0.9224,
|
| 444 |
+
"step": 31000
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.2680648927176845,
|
| 448 |
+
"grad_norm": 12.695961952209473,
|
| 449 |
+
"learning_rate": 4.618600968828416e-05,
|
| 450 |
+
"loss": 0.911,
|
| 451 |
+
"step": 31500
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.2881929068878064,
|
| 455 |
+
"grad_norm": 10.137328147888184,
|
| 456 |
+
"learning_rate": 4.564926264374757e-05,
|
| 457 |
+
"loss": 0.8909,
|
| 458 |
+
"step": 32000
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.3083209210579283,
|
| 462 |
+
"grad_norm": 14.675042152404785,
|
| 463 |
+
"learning_rate": 4.511251559921099e-05,
|
| 464 |
+
"loss": 0.8834,
|
| 465 |
+
"step": 32500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.3284489352280504,
|
| 469 |
+
"grad_norm": 9.264530181884766,
|
| 470 |
+
"learning_rate": 4.4575768554674396e-05,
|
| 471 |
+
"loss": 0.8855,
|
| 472 |
+
"step": 33000
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.3485769493981723,
|
| 476 |
+
"grad_norm": 18.74481201171875,
|
| 477 |
+
"learning_rate": 4.403902151013781e-05,
|
| 478 |
+
"loss": 0.8788,
|
| 479 |
+
"step": 33500
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.3687049635682944,
|
| 483 |
+
"grad_norm": 8.973100662231445,
|
| 484 |
+
"learning_rate": 4.350227446560123e-05,
|
| 485 |
+
"loss": 0.863,
|
| 486 |
+
"step": 34000
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.3888329777384163,
|
| 490 |
+
"grad_norm": 7.543423175811768,
|
| 491 |
+
"learning_rate": 4.296552742106464e-05,
|
| 492 |
+
"loss": 0.888,
|
| 493 |
+
"step": 34500
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.4089609919085384,
|
| 497 |
+
"grad_norm": 11.946640014648438,
|
| 498 |
+
"learning_rate": 4.242878037652806e-05,
|
| 499 |
+
"loss": 0.8757,
|
| 500 |
+
"step": 35000
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.4290890060786603,
|
| 504 |
+
"grad_norm": 13.646244049072266,
|
| 505 |
+
"learning_rate": 4.189203333199147e-05,
|
| 506 |
+
"loss": 0.848,
|
| 507 |
+
"step": 35500
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.4492170202487822,
|
| 511 |
+
"grad_norm": 12.406970977783203,
|
| 512 |
+
"learning_rate": 4.1355286287454876e-05,
|
| 513 |
+
"loss": 0.8628,
|
| 514 |
+
"step": 36000
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.4693450344189043,
|
| 518 |
+
"grad_norm": 8.028214454650879,
|
| 519 |
+
"learning_rate": 4.08185392429183e-05,
|
| 520 |
+
"loss": 0.8727,
|
| 521 |
+
"step": 36500
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.4894730485890262,
|
| 525 |
+
"grad_norm": 13.901910781860352,
|
| 526 |
+
"learning_rate": 4.0281792198381705e-05,
|
| 527 |
+
"loss": 0.8496,
|
| 528 |
+
"step": 37000
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.509601062759148,
|
| 532 |
+
"grad_norm": 7.470578193664551,
|
| 533 |
+
"learning_rate": 3.974504515384512e-05,
|
| 534 |
+
"loss": 0.8504,
|
| 535 |
+
"step": 37500
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.5297290769292702,
|
| 539 |
+
"grad_norm": 7.381893634796143,
|
| 540 |
+
"learning_rate": 3.920829810930854e-05,
|
| 541 |
+
"loss": 0.8462,
|
| 542 |
+
"step": 38000
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.5498570910993923,
|
| 546 |
+
"grad_norm": 8.749372482299805,
|
| 547 |
+
"learning_rate": 3.8671551064771956e-05,
|
| 548 |
+
"loss": 0.805,
|
| 549 |
+
"step": 38500
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.569985105269514,
|
| 553 |
+
"grad_norm": 15.834980964660645,
|
| 554 |
+
"learning_rate": 3.813480402023537e-05,
|
| 555 |
+
"loss": 0.815,
|
| 556 |
+
"step": 39000
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.590113119439636,
|
| 560 |
+
"grad_norm": 8.48942756652832,
|
| 561 |
+
"learning_rate": 3.7598056975698785e-05,
|
| 562 |
+
"loss": 0.811,
|
| 563 |
+
"step": 39500
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.6102411336097582,
|
| 567 |
+
"grad_norm": 17.017009735107422,
|
| 568 |
+
"learning_rate": 3.706130993116219e-05,
|
| 569 |
+
"loss": 0.8117,
|
| 570 |
+
"step": 40000
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.63036914777988,
|
| 574 |
+
"grad_norm": 10.621646881103516,
|
| 575 |
+
"learning_rate": 3.652456288662561e-05,
|
| 576 |
+
"loss": 0.794,
|
| 577 |
+
"step": 40500
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.650497161950002,
|
| 581 |
+
"grad_norm": 12.966856002807617,
|
| 582 |
+
"learning_rate": 3.598781584208902e-05,
|
| 583 |
+
"loss": 0.8064,
|
| 584 |
+
"step": 41000
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.670625176120124,
|
| 588 |
+
"grad_norm": 5.664041519165039,
|
| 589 |
+
"learning_rate": 3.5451068797552436e-05,
|
| 590 |
+
"loss": 0.8125,
|
| 591 |
+
"step": 41500
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.690753190290246,
|
| 595 |
+
"grad_norm": 9.313810348510742,
|
| 596 |
+
"learning_rate": 3.491432175301585e-05,
|
| 597 |
+
"loss": 0.8086,
|
| 598 |
+
"step": 42000
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.7108812044603678,
|
| 602 |
+
"grad_norm": 7.844326019287109,
|
| 603 |
+
"learning_rate": 3.4377574708479265e-05,
|
| 604 |
+
"loss": 0.7769,
|
| 605 |
+
"step": 42500
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.73100921863049,
|
| 609 |
+
"grad_norm": 11.637877464294434,
|
| 610 |
+
"learning_rate": 3.384082766394268e-05,
|
| 611 |
+
"loss": 0.7954,
|
| 612 |
+
"step": 43000
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.7511372328006118,
|
| 616 |
+
"grad_norm": 10.841877937316895,
|
| 617 |
+
"learning_rate": 3.3304080619406094e-05,
|
| 618 |
+
"loss": 0.7756,
|
| 619 |
+
"step": 43500
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.7712652469707337,
|
| 623 |
+
"grad_norm": 8.754426002502441,
|
| 624 |
+
"learning_rate": 3.276733357486951e-05,
|
| 625 |
+
"loss": 0.7891,
|
| 626 |
+
"step": 44000
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.7913932611408558,
|
| 630 |
+
"grad_norm": 7.297679901123047,
|
| 631 |
+
"learning_rate": 3.223058653033292e-05,
|
| 632 |
+
"loss": 0.7789,
|
| 633 |
+
"step": 44500
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.811521275310978,
|
| 637 |
+
"grad_norm": 11.88018798828125,
|
| 638 |
+
"learning_rate": 3.169383948579634e-05,
|
| 639 |
+
"loss": 0.7613,
|
| 640 |
+
"step": 45000
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.8316492894810998,
|
| 644 |
+
"grad_norm": 13.821844100952148,
|
| 645 |
+
"learning_rate": 3.1157092441259745e-05,
|
| 646 |
+
"loss": 0.7653,
|
| 647 |
+
"step": 45500
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.8517773036512217,
|
| 651 |
+
"grad_norm": 12.671835899353027,
|
| 652 |
+
"learning_rate": 3.062034539672316e-05,
|
| 653 |
+
"loss": 0.7838,
|
| 654 |
+
"step": 46000
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.8719053178213438,
|
| 658 |
+
"grad_norm": 12.358460426330566,
|
| 659 |
+
"learning_rate": 3.0083598352186574e-05,
|
| 660 |
+
"loss": 0.7845,
|
| 661 |
+
"step": 46500
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.8920333319914657,
|
| 665 |
+
"grad_norm": 11.423474311828613,
|
| 666 |
+
"learning_rate": 2.9546851307649992e-05,
|
| 667 |
+
"loss": 0.757,
|
| 668 |
+
"step": 47000
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.9121613461615876,
|
| 672 |
+
"grad_norm": 13.875541687011719,
|
| 673 |
+
"learning_rate": 2.9010104263113407e-05,
|
| 674 |
+
"loss": 0.776,
|
| 675 |
+
"step": 47500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.9322893603317097,
|
| 679 |
+
"grad_norm": 8.722099304199219,
|
| 680 |
+
"learning_rate": 2.8473357218576815e-05,
|
| 681 |
+
"loss": 0.7385,
|
| 682 |
+
"step": 48000
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.9524173745018316,
|
| 686 |
+
"grad_norm": 18.633831024169922,
|
| 687 |
+
"learning_rate": 2.793661017404023e-05,
|
| 688 |
+
"loss": 0.7401,
|
| 689 |
+
"step": 48500
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.9725453886719535,
|
| 693 |
+
"grad_norm": 8.491524696350098,
|
| 694 |
+
"learning_rate": 2.7399863129503647e-05,
|
| 695 |
+
"loss": 0.7236,
|
| 696 |
+
"step": 49000
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.9926734028420756,
|
| 700 |
+
"grad_norm": 3.7893919944763184,
|
| 701 |
+
"learning_rate": 2.686311608496706e-05,
|
| 702 |
+
"loss": 0.7242,
|
| 703 |
+
"step": 49500
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 2.0128014170121977,
|
| 707 |
+
"grad_norm": 8.858596801757812,
|
| 708 |
+
"learning_rate": 2.6326369040430476e-05,
|
| 709 |
+
"loss": 0.6995,
|
| 710 |
+
"step": 50000
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 2.0329294311823194,
|
| 714 |
+
"grad_norm": 9.343021392822266,
|
| 715 |
+
"learning_rate": 2.5789621995893884e-05,
|
| 716 |
+
"loss": 0.6833,
|
| 717 |
+
"step": 50500
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 2.0530574453524415,
|
| 721 |
+
"grad_norm": 11.943695068359375,
|
| 722 |
+
"learning_rate": 2.5252874951357302e-05,
|
| 723 |
+
"loss": 0.6796,
|
| 724 |
+
"step": 51000
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 2.0731854595225636,
|
| 728 |
+
"grad_norm": 8.132844924926758,
|
| 729 |
+
"learning_rate": 2.4716127906820716e-05,
|
| 730 |
+
"loss": 0.6655,
|
| 731 |
+
"step": 51500
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 2.0933134736926853,
|
| 735 |
+
"grad_norm": 14.167099952697754,
|
| 736 |
+
"learning_rate": 2.417938086228413e-05,
|
| 737 |
+
"loss": 0.6843,
|
| 738 |
+
"step": 52000
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 2.1134414878628074,
|
| 742 |
+
"grad_norm": 7.767176628112793,
|
| 743 |
+
"learning_rate": 2.3642633817747545e-05,
|
| 744 |
+
"loss": 0.6785,
|
| 745 |
+
"step": 52500
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 2.1335695020329295,
|
| 749 |
+
"grad_norm": 12.753235816955566,
|
| 750 |
+
"learning_rate": 2.3105886773210956e-05,
|
| 751 |
+
"loss": 0.6591,
|
| 752 |
+
"step": 53000
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 2.1536975162030516,
|
| 756 |
+
"grad_norm": 15.783185005187988,
|
| 757 |
+
"learning_rate": 2.256913972867437e-05,
|
| 758 |
+
"loss": 0.6455,
|
| 759 |
+
"step": 53500
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 2.1738255303731733,
|
| 763 |
+
"grad_norm": 17.965497970581055,
|
| 764 |
+
"learning_rate": 2.2032392684137785e-05,
|
| 765 |
+
"loss": 0.6674,
|
| 766 |
+
"step": 54000
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 2.1939535445432954,
|
| 770 |
+
"grad_norm": 18.289630889892578,
|
| 771 |
+
"learning_rate": 2.14956456396012e-05,
|
| 772 |
+
"loss": 0.6584,
|
| 773 |
+
"step": 54500
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 2.2140815587134175,
|
| 777 |
+
"grad_norm": 17.599037170410156,
|
| 778 |
+
"learning_rate": 2.0958898595064615e-05,
|
| 779 |
+
"loss": 0.6432,
|
| 780 |
+
"step": 55000
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 2.234209572883539,
|
| 784 |
+
"grad_norm": 8.277852058410645,
|
| 785 |
+
"learning_rate": 2.042215155052803e-05,
|
| 786 |
+
"loss": 0.6373,
|
| 787 |
+
"step": 55500
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 2.2543375870536613,
|
| 791 |
+
"grad_norm": 11.451165199279785,
|
| 792 |
+
"learning_rate": 1.988540450599144e-05,
|
| 793 |
+
"loss": 0.651,
|
| 794 |
+
"step": 56000
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 2.2744656012237834,
|
| 798 |
+
"grad_norm": 7.643725872039795,
|
| 799 |
+
"learning_rate": 1.9348657461454855e-05,
|
| 800 |
+
"loss": 0.6928,
|
| 801 |
+
"step": 56500
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 2.294593615393905,
|
| 805 |
+
"grad_norm": 11.526747703552246,
|
| 806 |
+
"learning_rate": 1.881191041691827e-05,
|
| 807 |
+
"loss": 0.6581,
|
| 808 |
+
"step": 57000
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 2.314721629564027,
|
| 812 |
+
"grad_norm": 19.298818588256836,
|
| 813 |
+
"learning_rate": 1.827516337238168e-05,
|
| 814 |
+
"loss": 0.625,
|
| 815 |
+
"step": 57500
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 2.3348496437341493,
|
| 819 |
+
"grad_norm": 14.259527206420898,
|
| 820 |
+
"learning_rate": 1.7738416327845095e-05,
|
| 821 |
+
"loss": 0.6402,
|
| 822 |
+
"step": 58000
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 2.3549776579042714,
|
| 826 |
+
"grad_norm": 11.155771255493164,
|
| 827 |
+
"learning_rate": 1.720166928330851e-05,
|
| 828 |
+
"loss": 0.6362,
|
| 829 |
+
"step": 58500
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 2.375105672074393,
|
| 833 |
+
"grad_norm": 18.191736221313477,
|
| 834 |
+
"learning_rate": 1.6664922238771924e-05,
|
| 835 |
+
"loss": 0.6282,
|
| 836 |
+
"step": 59000
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 2.395233686244515,
|
| 840 |
+
"grad_norm": 14.766573905944824,
|
| 841 |
+
"learning_rate": 1.612817519423534e-05,
|
| 842 |
+
"loss": 0.6315,
|
| 843 |
+
"step": 59500
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 2.4153617004146373,
|
| 847 |
+
"grad_norm": 13.01074504852295,
|
| 848 |
+
"learning_rate": 1.5591428149698753e-05,
|
| 849 |
+
"loss": 0.6321,
|
| 850 |
+
"step": 60000
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 2.435489714584759,
|
| 854 |
+
"grad_norm": 8.57907485961914,
|
| 855 |
+
"learning_rate": 1.5054681105162166e-05,
|
| 856 |
+
"loss": 0.6049,
|
| 857 |
+
"step": 60500
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 2.455617728754881,
|
| 861 |
+
"grad_norm": 9.666173934936523,
|
| 862 |
+
"learning_rate": 1.451793406062558e-05,
|
| 863 |
+
"loss": 0.6389,
|
| 864 |
+
"step": 61000
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 2.475745742925003,
|
| 868 |
+
"grad_norm": 11.621274948120117,
|
| 869 |
+
"learning_rate": 1.3981187016088993e-05,
|
| 870 |
+
"loss": 0.6134,
|
| 871 |
+
"step": 61500
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 2.495873757095125,
|
| 875 |
+
"grad_norm": 6.330214500427246,
|
| 876 |
+
"learning_rate": 1.3444439971552408e-05,
|
| 877 |
+
"loss": 0.6183,
|
| 878 |
+
"step": 62000
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 2.516001771265247,
|
| 882 |
+
"grad_norm": 5.045897006988525,
|
| 883 |
+
"learning_rate": 1.2907692927015822e-05,
|
| 884 |
+
"loss": 0.6184,
|
| 885 |
+
"step": 62500
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 2.536129785435369,
|
| 889 |
+
"grad_norm": 5.904106140136719,
|
| 890 |
+
"learning_rate": 1.2370945882479235e-05,
|
| 891 |
+
"loss": 0.6352,
|
| 892 |
+
"step": 63000
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 2.5562577996054907,
|
| 896 |
+
"grad_norm": 8.777071952819824,
|
| 897 |
+
"learning_rate": 1.183419883794265e-05,
|
| 898 |
+
"loss": 0.6343,
|
| 899 |
+
"step": 63500
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 2.576385813775613,
|
| 903 |
+
"grad_norm": 14.809232711791992,
|
| 904 |
+
"learning_rate": 1.1297451793406064e-05,
|
| 905 |
+
"loss": 0.635,
|
| 906 |
+
"step": 64000
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 2.596513827945735,
|
| 910 |
+
"grad_norm": 9.59287166595459,
|
| 911 |
+
"learning_rate": 1.0760704748869477e-05,
|
| 912 |
+
"loss": 0.6307,
|
| 913 |
+
"step": 64500
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 2.6166418421158566,
|
| 917 |
+
"grad_norm": 4.077784538269043,
|
| 918 |
+
"learning_rate": 1.0223957704332891e-05,
|
| 919 |
+
"loss": 0.614,
|
| 920 |
+
"step": 65000
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.6367698562859787,
|
| 924 |
+
"grad_norm": 17.284059524536133,
|
| 925 |
+
"learning_rate": 9.687210659796306e-06,
|
| 926 |
+
"loss": 0.6556,
|
| 927 |
+
"step": 65500
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 2.656897870456101,
|
| 931 |
+
"grad_norm": 13.630504608154297,
|
| 932 |
+
"learning_rate": 9.150463615259719e-06,
|
| 933 |
+
"loss": 0.5926,
|
| 934 |
+
"step": 66000
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.677025884626223,
|
| 938 |
+
"grad_norm": 12.159080505371094,
|
| 939 |
+
"learning_rate": 8.613716570723133e-06,
|
| 940 |
+
"loss": 0.6097,
|
| 941 |
+
"step": 66500
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 2.6971538987963446,
|
| 945 |
+
"grad_norm": 11.280355453491211,
|
| 946 |
+
"learning_rate": 8.076969526186546e-06,
|
| 947 |
+
"loss": 0.614,
|
| 948 |
+
"step": 67000
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.7172819129664667,
|
| 952 |
+
"grad_norm": 16.78717613220215,
|
| 953 |
+
"learning_rate": 7.540222481649961e-06,
|
| 954 |
+
"loss": 0.6232,
|
| 955 |
+
"step": 67500
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 2.737409927136589,
|
| 959 |
+
"grad_norm": 10.830788612365723,
|
| 960 |
+
"learning_rate": 7.003475437113375e-06,
|
| 961 |
+
"loss": 0.5923,
|
| 962 |
+
"step": 68000
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.757537941306711,
|
| 966 |
+
"grad_norm": 6.981812000274658,
|
| 967 |
+
"learning_rate": 6.466728392576789e-06,
|
| 968 |
+
"loss": 0.5949,
|
| 969 |
+
"step": 68500
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 2.7776659554768326,
|
| 973 |
+
"grad_norm": 9.136080741882324,
|
| 974 |
+
"learning_rate": 5.929981348040202e-06,
|
| 975 |
+
"loss": 0.6128,
|
| 976 |
+
"step": 69000
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 2.7977939696469547,
|
| 980 |
+
"grad_norm": 10.217459678649902,
|
| 981 |
+
"learning_rate": 5.393234303503617e-06,
|
| 982 |
+
"loss": 0.5998,
|
| 983 |
+
"step": 69500
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 2.817921983817077,
|
| 987 |
+
"grad_norm": 10.00894832611084,
|
| 988 |
+
"learning_rate": 4.8564872589670306e-06,
|
| 989 |
+
"loss": 0.5922,
|
| 990 |
+
"step": 70000
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.8380499979871985,
|
| 994 |
+
"grad_norm": 12.962599754333496,
|
| 995 |
+
"learning_rate": 4.319740214430444e-06,
|
| 996 |
+
"loss": 0.5883,
|
| 997 |
+
"step": 70500
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.8581780121573206,
|
| 1001 |
+
"grad_norm": 13.627707481384277,
|
| 1002 |
+
"learning_rate": 3.7829931698938588e-06,
|
| 1003 |
+
"loss": 0.5994,
|
| 1004 |
+
"step": 71000
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.8783060263274427,
|
| 1008 |
+
"grad_norm": 11.855643272399902,
|
| 1009 |
+
"learning_rate": 3.2462461253572724e-06,
|
| 1010 |
+
"loss": 0.6179,
|
| 1011 |
+
"step": 71500
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.8984340404975644,
|
| 1015 |
+
"grad_norm": 9.207647323608398,
|
| 1016 |
+
"learning_rate": 2.7094990808206865e-06,
|
| 1017 |
+
"loss": 0.6169,
|
| 1018 |
+
"step": 72000
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.9185620546676865,
|
| 1022 |
+
"grad_norm": 13.871038436889648,
|
| 1023 |
+
"learning_rate": 2.1727520362841e-06,
|
| 1024 |
+
"loss": 0.6037,
|
| 1025 |
+
"step": 72500
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.9386900688378086,
|
| 1029 |
+
"grad_norm": 10.001399040222168,
|
| 1030 |
+
"learning_rate": 1.6360049917475143e-06,
|
| 1031 |
+
"loss": 0.619,
|
| 1032 |
+
"step": 73000
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.9588180830079303,
|
| 1036 |
+
"grad_norm": 15.267190933227539,
|
| 1037 |
+
"learning_rate": 1.0992579472109282e-06,
|
| 1038 |
+
"loss": 0.6055,
|
| 1039 |
+
"step": 73500
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.9789460971780524,
|
| 1043 |
+
"grad_norm": 3.722136974334717,
|
| 1044 |
+
"learning_rate": 5.625109026743423e-07,
|
| 1045 |
+
"loss": 0.591,
|
| 1046 |
+
"step": 74000
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.9990741113481745,
|
| 1050 |
+
"grad_norm": 9.667593955993652,
|
| 1051 |
+
"learning_rate": 2.576385813775613e-08,
|
| 1052 |
+
"loss": 0.6077,
|
| 1053 |
+
"step": 74500
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 3.0,
|
| 1057 |
+
"step": 74523,
|
| 1058 |
+
"total_flos": 1.6739482297393152e+16,
|
| 1059 |
+
"train_loss": 1.0437313645233104,
|
| 1060 |
+
"train_runtime": 12605.9624,
|
| 1061 |
+
"train_samples_per_second": 94.587,
|
| 1062 |
+
"train_steps_per_second": 5.912
|
| 1063 |
}
|
| 1064 |
],
|
| 1065 |
"logging_steps": 500,
|
| 1066 |
+
"max_steps": 74523,
|
| 1067 |
"num_input_tokens_seen": 0,
|
| 1068 |
+
"num_train_epochs": 3,
|
| 1069 |
"save_steps": 500,
|
| 1070 |
"stateful_callbacks": {
|
| 1071 |
"TrainerControl": {
|
|
|
|
| 1079 |
"attributes": {}
|
| 1080 |
}
|
| 1081 |
},
|
| 1082 |
+
"total_flos": 1.6739482297393152e+16,
|
| 1083 |
"train_batch_size": 16,
|
| 1084 |
"trial_name": null,
|
| 1085 |
"trial_params": null
|