Training in progress, step 588, checkpoint
Browse files
last-checkpoint/adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 83945296
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1b8cc864cc1f54d10890b9acd2551c323b0e3438a9fe18de8b0fb09ba6d10dda
|
| 3 |
size 83945296
|
last-checkpoint/optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 43123028
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3f743e096d531a90f19e6fda22899fe87bc41c458d260888bce8d18557ec563b
|
| 3 |
size 43123028
|
last-checkpoint/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 sha256:5dbbd2dfd1f3a5044a31ea5a030e11aeba92fcd37cb867f40e061211bcde3e30
|
| 3 |
size 14244
|
last-checkpoint/scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1064
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fe12bf308471af9518165c28c516df6c67b2311f209af1fe824f9e028f69f79c
|
| 3 |
size 1064
|
last-checkpoint/trainer_state.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 294,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
@@ -2073,6 +2073,2072 @@
|
|
| 2073 |
"eval_samples_per_second": 12.278,
|
| 2074 |
"eval_steps_per_second": 6.139,
|
| 2075 |
"step": 294
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 2076 |
}
|
| 2077 |
],
|
| 2078 |
"logging_steps": 1,
|
|
@@ -2092,7 +4158,7 @@
|
|
| 2092 |
"attributes": {}
|
| 2093 |
}
|
| 2094 |
},
|
| 2095 |
-
"total_flos":
|
| 2096 |
"train_batch_size": 2,
|
| 2097 |
"trial_name": null,
|
| 2098 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.5001594896331738,
|
| 5 |
"eval_steps": 294,
|
| 6 |
+
"global_step": 588,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
|
|
|
| 2073 |
"eval_samples_per_second": 12.278,
|
| 2074 |
"eval_steps_per_second": 6.139,
|
| 2075 |
"step": 294
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 0.2509303561935141,
|
| 2079 |
+
"grad_norm": 0.6918848156929016,
|
| 2080 |
+
"learning_rate": 0.00017193813910758,
|
| 2081 |
+
"loss": 2.6556,
|
| 2082 |
+
"step": 295
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 0.25178096757044127,
|
| 2086 |
+
"grad_norm": 0.5682982802391052,
|
| 2087 |
+
"learning_rate": 0.00017175072570443312,
|
| 2088 |
+
"loss": 2.6581,
|
| 2089 |
+
"step": 296
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 0.25263157894736843,
|
| 2093 |
+
"grad_norm": 0.6087559461593628,
|
| 2094 |
+
"learning_rate": 0.00017156279143208352,
|
| 2095 |
+
"loss": 2.5665,
|
| 2096 |
+
"step": 297
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 0.2534821903242956,
|
| 2100 |
+
"grad_norm": 0.6545628309249878,
|
| 2101 |
+
"learning_rate": 0.00017137433765482642,
|
| 2102 |
+
"loss": 2.8215,
|
| 2103 |
+
"step": 298
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 0.25433280170122274,
|
| 2107 |
+
"grad_norm": 0.6754540801048279,
|
| 2108 |
+
"learning_rate": 0.00017118536574072842,
|
| 2109 |
+
"loss": 2.7991,
|
| 2110 |
+
"step": 299
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 0.2551834130781499,
|
| 2114 |
+
"grad_norm": 0.6926846504211426,
|
| 2115 |
+
"learning_rate": 0.0001709958770616174,
|
| 2116 |
+
"loss": 2.7371,
|
| 2117 |
+
"step": 300
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 0.2560340244550771,
|
| 2121 |
+
"grad_norm": 0.32997071743011475,
|
| 2122 |
+
"learning_rate": 0.00017080587299307283,
|
| 2123 |
+
"loss": 2.7739,
|
| 2124 |
+
"step": 301
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 0.25688463583200427,
|
| 2128 |
+
"grad_norm": 0.3260290324687958,
|
| 2129 |
+
"learning_rate": 0.0001706153549144154,
|
| 2130 |
+
"loss": 2.5971,
|
| 2131 |
+
"step": 302
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 0.25773524720893143,
|
| 2135 |
+
"grad_norm": 0.3378421366214752,
|
| 2136 |
+
"learning_rate": 0.00017042432420869732,
|
| 2137 |
+
"loss": 2.7588,
|
| 2138 |
+
"step": 303
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 0.2585858585858586,
|
| 2142 |
+
"grad_norm": 0.27844691276550293,
|
| 2143 |
+
"learning_rate": 0.0001702327822626922,
|
| 2144 |
+
"loss": 2.6493,
|
| 2145 |
+
"step": 304
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 0.25943646996278574,
|
| 2149 |
+
"grad_norm": 0.29100462794303894,
|
| 2150 |
+
"learning_rate": 0.00017004073046688497,
|
| 2151 |
+
"loss": 2.6397,
|
| 2152 |
+
"step": 305
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 0.2602870813397129,
|
| 2156 |
+
"grad_norm": 0.2764577865600586,
|
| 2157 |
+
"learning_rate": 0.00016984817021546177,
|
| 2158 |
+
"loss": 2.4199,
|
| 2159 |
+
"step": 306
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 0.26113769271664006,
|
| 2163 |
+
"grad_norm": 0.26870644092559814,
|
| 2164 |
+
"learning_rate": 0.00016965510290629972,
|
| 2165 |
+
"loss": 2.2552,
|
| 2166 |
+
"step": 307
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 0.26198830409356727,
|
| 2170 |
+
"grad_norm": 0.2770349979400635,
|
| 2171 |
+
"learning_rate": 0.00016946152994095704,
|
| 2172 |
+
"loss": 2.6118,
|
| 2173 |
+
"step": 308
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 0.2628389154704944,
|
| 2177 |
+
"grad_norm": 0.27041003108024597,
|
| 2178 |
+
"learning_rate": 0.00016926745272466268,
|
| 2179 |
+
"loss": 2.5329,
|
| 2180 |
+
"step": 309
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 0.2636895268474216,
|
| 2184 |
+
"grad_norm": 0.25608015060424805,
|
| 2185 |
+
"learning_rate": 0.00016907287266630614,
|
| 2186 |
+
"loss": 2.3411,
|
| 2187 |
+
"step": 310
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 0.26454013822434874,
|
| 2191 |
+
"grad_norm": 0.2750420570373535,
|
| 2192 |
+
"learning_rate": 0.00016887779117842725,
|
| 2193 |
+
"loss": 2.6393,
|
| 2194 |
+
"step": 311
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 0.2653907496012759,
|
| 2198 |
+
"grad_norm": 0.26276537775993347,
|
| 2199 |
+
"learning_rate": 0.00016868220967720604,
|
| 2200 |
+
"loss": 2.3616,
|
| 2201 |
+
"step": 312
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 0.26624136097820306,
|
| 2205 |
+
"grad_norm": 0.2735307514667511,
|
| 2206 |
+
"learning_rate": 0.00016848612958245216,
|
| 2207 |
+
"loss": 2.5156,
|
| 2208 |
+
"step": 313
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 0.26709197235513027,
|
| 2212 |
+
"grad_norm": 0.32951095700263977,
|
| 2213 |
+
"learning_rate": 0.00016828955231759497,
|
| 2214 |
+
"loss": 2.5329,
|
| 2215 |
+
"step": 314
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 0.2679425837320574,
|
| 2219 |
+
"grad_norm": 0.2762184143066406,
|
| 2220 |
+
"learning_rate": 0.00016809247930967282,
|
| 2221 |
+
"loss": 2.6873,
|
| 2222 |
+
"step": 315
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 0.2687931951089846,
|
| 2226 |
+
"grad_norm": 0.2802570164203644,
|
| 2227 |
+
"learning_rate": 0.000167894911989323,
|
| 2228 |
+
"loss": 2.5532,
|
| 2229 |
+
"step": 316
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 0.26964380648591174,
|
| 2233 |
+
"grad_norm": 0.26968276500701904,
|
| 2234 |
+
"learning_rate": 0.0001676968517907712,
|
| 2235 |
+
"loss": 2.602,
|
| 2236 |
+
"step": 317
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 0.2704944178628389,
|
| 2240 |
+
"grad_norm": 0.27560874819755554,
|
| 2241 |
+
"learning_rate": 0.00016749830015182107,
|
| 2242 |
+
"loss": 2.5003,
|
| 2243 |
+
"step": 318
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 0.27134502923976606,
|
| 2247 |
+
"grad_norm": 0.288411021232605,
|
| 2248 |
+
"learning_rate": 0.00016729925851384386,
|
| 2249 |
+
"loss": 2.6859,
|
| 2250 |
+
"step": 319
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 0.27219564061669327,
|
| 2254 |
+
"grad_norm": 0.2999224364757538,
|
| 2255 |
+
"learning_rate": 0.00016709972832176797,
|
| 2256 |
+
"loss": 2.8356,
|
| 2257 |
+
"step": 320
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 0.2730462519936204,
|
| 2261 |
+
"grad_norm": 0.2956329882144928,
|
| 2262 |
+
"learning_rate": 0.0001668997110240684,
|
| 2263 |
+
"loss": 2.6157,
|
| 2264 |
+
"step": 321
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 0.2738968633705476,
|
| 2268 |
+
"grad_norm": 0.30924192070961,
|
| 2269 |
+
"learning_rate": 0.00016669920807275623,
|
| 2270 |
+
"loss": 2.8421,
|
| 2271 |
+
"step": 322
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 0.27474747474747474,
|
| 2275 |
+
"grad_norm": 0.31185418367385864,
|
| 2276 |
+
"learning_rate": 0.00016649822092336812,
|
| 2277 |
+
"loss": 2.758,
|
| 2278 |
+
"step": 323
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 0.2755980861244019,
|
| 2282 |
+
"grad_norm": 0.3129124045372009,
|
| 2283 |
+
"learning_rate": 0.0001662967510349558,
|
| 2284 |
+
"loss": 2.6861,
|
| 2285 |
+
"step": 324
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 0.27644869750132905,
|
| 2289 |
+
"grad_norm": 0.3300238251686096,
|
| 2290 |
+
"learning_rate": 0.00016609479987007527,
|
| 2291 |
+
"loss": 2.8284,
|
| 2292 |
+
"step": 325
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 0.27729930887825627,
|
| 2296 |
+
"grad_norm": 0.3459511399269104,
|
| 2297 |
+
"learning_rate": 0.00016589236889477646,
|
| 2298 |
+
"loss": 2.6454,
|
| 2299 |
+
"step": 326
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 0.2781499202551834,
|
| 2303 |
+
"grad_norm": 0.3566714823246002,
|
| 2304 |
+
"learning_rate": 0.00016568945957859236,
|
| 2305 |
+
"loss": 2.399,
|
| 2306 |
+
"step": 327
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 0.2790005316321106,
|
| 2310 |
+
"grad_norm": 0.3650771379470825,
|
| 2311 |
+
"learning_rate": 0.00016548607339452853,
|
| 2312 |
+
"loss": 2.7529,
|
| 2313 |
+
"step": 328
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 0.27985114300903774,
|
| 2317 |
+
"grad_norm": 0.35940608382225037,
|
| 2318 |
+
"learning_rate": 0.00016528221181905217,
|
| 2319 |
+
"loss": 2.5894,
|
| 2320 |
+
"step": 329
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 0.2807017543859649,
|
| 2324 |
+
"grad_norm": 0.4018422067165375,
|
| 2325 |
+
"learning_rate": 0.0001650778763320817,
|
| 2326 |
+
"loss": 2.7001,
|
| 2327 |
+
"step": 330
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 0.28155236576289205,
|
| 2331 |
+
"grad_norm": 0.3867095708847046,
|
| 2332 |
+
"learning_rate": 0.00016487306841697578,
|
| 2333 |
+
"loss": 2.9919,
|
| 2334 |
+
"step": 331
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 0.28240297713981927,
|
| 2338 |
+
"grad_norm": 0.5714160799980164,
|
| 2339 |
+
"learning_rate": 0.0001646677895605227,
|
| 2340 |
+
"loss": 2.9366,
|
| 2341 |
+
"step": 332
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 0.2832535885167464,
|
| 2345 |
+
"grad_norm": 0.3697023391723633,
|
| 2346 |
+
"learning_rate": 0.00016446204125292942,
|
| 2347 |
+
"loss": 2.653,
|
| 2348 |
+
"step": 333
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 0.2841041998936736,
|
| 2352 |
+
"grad_norm": 0.4222877323627472,
|
| 2353 |
+
"learning_rate": 0.00016425582498781087,
|
| 2354 |
+
"loss": 2.8557,
|
| 2355 |
+
"step": 334
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 0.28495481127060074,
|
| 2359 |
+
"grad_norm": 0.40899163484573364,
|
| 2360 |
+
"learning_rate": 0.0001640491422621792,
|
| 2361 |
+
"loss": 2.7875,
|
| 2362 |
+
"step": 335
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 0.2858054226475279,
|
| 2366 |
+
"grad_norm": 0.44694027304649353,
|
| 2367 |
+
"learning_rate": 0.00016384199457643262,
|
| 2368 |
+
"loss": 2.8616,
|
| 2369 |
+
"step": 336
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 0.28665603402445505,
|
| 2373 |
+
"grad_norm": 0.45921215415000916,
|
| 2374 |
+
"learning_rate": 0.00016363438343434483,
|
| 2375 |
+
"loss": 2.5821,
|
| 2376 |
+
"step": 337
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 0.28750664540138227,
|
| 2380 |
+
"grad_norm": 0.4156991243362427,
|
| 2381 |
+
"learning_rate": 0.00016342631034305384,
|
| 2382 |
+
"loss": 2.7228,
|
| 2383 |
+
"step": 338
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 0.2883572567783094,
|
| 2387 |
+
"grad_norm": 0.43912479281425476,
|
| 2388 |
+
"learning_rate": 0.00016321777681305125,
|
| 2389 |
+
"loss": 2.7119,
|
| 2390 |
+
"step": 339
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 0.2892078681552366,
|
| 2394 |
+
"grad_norm": 0.4638505280017853,
|
| 2395 |
+
"learning_rate": 0.00016300878435817113,
|
| 2396 |
+
"loss": 2.6832,
|
| 2397 |
+
"step": 340
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 0.29005847953216374,
|
| 2401 |
+
"grad_norm": 0.4912424087524414,
|
| 2402 |
+
"learning_rate": 0.00016279933449557906,
|
| 2403 |
+
"loss": 2.6583,
|
| 2404 |
+
"step": 341
|
| 2405 |
+
},
|
| 2406 |
+
{
|
| 2407 |
+
"epoch": 0.2909090909090909,
|
| 2408 |
+
"grad_norm": 0.5067815780639648,
|
| 2409 |
+
"learning_rate": 0.00016258942874576118,
|
| 2410 |
+
"loss": 2.7723,
|
| 2411 |
+
"step": 342
|
| 2412 |
+
},
|
| 2413 |
+
{
|
| 2414 |
+
"epoch": 0.29175970228601805,
|
| 2415 |
+
"grad_norm": 0.49635544419288635,
|
| 2416 |
+
"learning_rate": 0.0001623790686325131,
|
| 2417 |
+
"loss": 2.7351,
|
| 2418 |
+
"step": 343
|
| 2419 |
+
},
|
| 2420 |
+
{
|
| 2421 |
+
"epoch": 0.29261031366294527,
|
| 2422 |
+
"grad_norm": 0.49465370178222656,
|
| 2423 |
+
"learning_rate": 0.00016216825568292885,
|
| 2424 |
+
"loss": 2.7251,
|
| 2425 |
+
"step": 344
|
| 2426 |
+
},
|
| 2427 |
+
{
|
| 2428 |
+
"epoch": 0.2934609250398724,
|
| 2429 |
+
"grad_norm": 0.558665931224823,
|
| 2430 |
+
"learning_rate": 0.00016195699142738975,
|
| 2431 |
+
"loss": 2.5475,
|
| 2432 |
+
"step": 345
|
| 2433 |
+
},
|
| 2434 |
+
{
|
| 2435 |
+
"epoch": 0.2943115364167996,
|
| 2436 |
+
"grad_norm": 0.6477727890014648,
|
| 2437 |
+
"learning_rate": 0.00016174527739955342,
|
| 2438 |
+
"loss": 2.7069,
|
| 2439 |
+
"step": 346
|
| 2440 |
+
},
|
| 2441 |
+
{
|
| 2442 |
+
"epoch": 0.29516214779372674,
|
| 2443 |
+
"grad_norm": 0.6656542420387268,
|
| 2444 |
+
"learning_rate": 0.00016153311513634257,
|
| 2445 |
+
"loss": 3.0185,
|
| 2446 |
+
"step": 347
|
| 2447 |
+
},
|
| 2448 |
+
{
|
| 2449 |
+
"epoch": 0.2960127591706539,
|
| 2450 |
+
"grad_norm": 0.7209298610687256,
|
| 2451 |
+
"learning_rate": 0.0001613205061779337,
|
| 2452 |
+
"loss": 2.8225,
|
| 2453 |
+
"step": 348
|
| 2454 |
+
},
|
| 2455 |
+
{
|
| 2456 |
+
"epoch": 0.29686337054758105,
|
| 2457 |
+
"grad_norm": 0.6285322308540344,
|
| 2458 |
+
"learning_rate": 0.0001611074520677462,
|
| 2459 |
+
"loss": 2.9088,
|
| 2460 |
+
"step": 349
|
| 2461 |
+
},
|
| 2462 |
+
{
|
| 2463 |
+
"epoch": 0.29771398192450826,
|
| 2464 |
+
"grad_norm": 0.7612189054489136,
|
| 2465 |
+
"learning_rate": 0.00016089395435243105,
|
| 2466 |
+
"loss": 2.9744,
|
| 2467 |
+
"step": 350
|
| 2468 |
+
},
|
| 2469 |
+
{
|
| 2470 |
+
"epoch": 0.2985645933014354,
|
| 2471 |
+
"grad_norm": 0.3265992999076843,
|
| 2472 |
+
"learning_rate": 0.00016068001458185936,
|
| 2473 |
+
"loss": 2.3931,
|
| 2474 |
+
"step": 351
|
| 2475 |
+
},
|
| 2476 |
+
{
|
| 2477 |
+
"epoch": 0.2994152046783626,
|
| 2478 |
+
"grad_norm": 0.3179806172847748,
|
| 2479 |
+
"learning_rate": 0.00016046563430911146,
|
| 2480 |
+
"loss": 2.6212,
|
| 2481 |
+
"step": 352
|
| 2482 |
+
},
|
| 2483 |
+
{
|
| 2484 |
+
"epoch": 0.30026581605528974,
|
| 2485 |
+
"grad_norm": 0.31296634674072266,
|
| 2486 |
+
"learning_rate": 0.00016025081509046544,
|
| 2487 |
+
"loss": 2.5008,
|
| 2488 |
+
"step": 353
|
| 2489 |
+
},
|
| 2490 |
+
{
|
| 2491 |
+
"epoch": 0.3011164274322169,
|
| 2492 |
+
"grad_norm": 0.283408522605896,
|
| 2493 |
+
"learning_rate": 0.00016003555848538586,
|
| 2494 |
+
"loss": 2.3946,
|
| 2495 |
+
"step": 354
|
| 2496 |
+
},
|
| 2497 |
+
{
|
| 2498 |
+
"epoch": 0.30196703880914405,
|
| 2499 |
+
"grad_norm": 0.2742927074432373,
|
| 2500 |
+
"learning_rate": 0.00015981986605651248,
|
| 2501 |
+
"loss": 2.5154,
|
| 2502 |
+
"step": 355
|
| 2503 |
+
},
|
| 2504 |
+
{
|
| 2505 |
+
"epoch": 0.30281765018607126,
|
| 2506 |
+
"grad_norm": 0.29107552766799927,
|
| 2507 |
+
"learning_rate": 0.00015960373936964892,
|
| 2508 |
+
"loss": 2.505,
|
| 2509 |
+
"step": 356
|
| 2510 |
+
},
|
| 2511 |
+
{
|
| 2512 |
+
"epoch": 0.3036682615629984,
|
| 2513 |
+
"grad_norm": 0.3244473338127136,
|
| 2514 |
+
"learning_rate": 0.0001593871799937512,
|
| 2515 |
+
"loss": 2.5244,
|
| 2516 |
+
"step": 357
|
| 2517 |
+
},
|
| 2518 |
+
{
|
| 2519 |
+
"epoch": 0.3045188729399256,
|
| 2520 |
+
"grad_norm": 0.27494120597839355,
|
| 2521 |
+
"learning_rate": 0.0001591701895009164,
|
| 2522 |
+
"loss": 2.4161,
|
| 2523 |
+
"step": 358
|
| 2524 |
+
},
|
| 2525 |
+
{
|
| 2526 |
+
"epoch": 0.30536948431685274,
|
| 2527 |
+
"grad_norm": 0.28784480690956116,
|
| 2528 |
+
"learning_rate": 0.00015895276946637136,
|
| 2529 |
+
"loss": 2.6434,
|
| 2530 |
+
"step": 359
|
| 2531 |
+
},
|
| 2532 |
+
{
|
| 2533 |
+
"epoch": 0.3062200956937799,
|
| 2534 |
+
"grad_norm": 0.28096505999565125,
|
| 2535 |
+
"learning_rate": 0.00015873492146846108,
|
| 2536 |
+
"loss": 2.6784,
|
| 2537 |
+
"step": 360
|
| 2538 |
+
},
|
| 2539 |
+
{
|
| 2540 |
+
"epoch": 0.30707070707070705,
|
| 2541 |
+
"grad_norm": 0.2602795362472534,
|
| 2542 |
+
"learning_rate": 0.00015851664708863735,
|
| 2543 |
+
"loss": 2.2712,
|
| 2544 |
+
"step": 361
|
| 2545 |
+
},
|
| 2546 |
+
{
|
| 2547 |
+
"epoch": 0.30792131844763426,
|
| 2548 |
+
"grad_norm": 0.28796225786209106,
|
| 2549 |
+
"learning_rate": 0.0001582979479114472,
|
| 2550 |
+
"loss": 2.8673,
|
| 2551 |
+
"step": 362
|
| 2552 |
+
},
|
| 2553 |
+
{
|
| 2554 |
+
"epoch": 0.3087719298245614,
|
| 2555 |
+
"grad_norm": 0.273456871509552,
|
| 2556 |
+
"learning_rate": 0.00015807882552452154,
|
| 2557 |
+
"loss": 2.6931,
|
| 2558 |
+
"step": 363
|
| 2559 |
+
},
|
| 2560 |
+
{
|
| 2561 |
+
"epoch": 0.3096225412014886,
|
| 2562 |
+
"grad_norm": 0.29190319776535034,
|
| 2563 |
+
"learning_rate": 0.00015785928151856347,
|
| 2564 |
+
"loss": 2.5932,
|
| 2565 |
+
"step": 364
|
| 2566 |
+
},
|
| 2567 |
+
{
|
| 2568 |
+
"epoch": 0.31047315257841573,
|
| 2569 |
+
"grad_norm": 0.290829598903656,
|
| 2570 |
+
"learning_rate": 0.0001576393174873368,
|
| 2571 |
+
"loss": 2.6889,
|
| 2572 |
+
"step": 365
|
| 2573 |
+
},
|
| 2574 |
+
{
|
| 2575 |
+
"epoch": 0.3113237639553429,
|
| 2576 |
+
"grad_norm": 0.27631402015686035,
|
| 2577 |
+
"learning_rate": 0.0001574189350276545,
|
| 2578 |
+
"loss": 2.6508,
|
| 2579 |
+
"step": 366
|
| 2580 |
+
},
|
| 2581 |
+
{
|
| 2582 |
+
"epoch": 0.31217437533227005,
|
| 2583 |
+
"grad_norm": 0.30464449524879456,
|
| 2584 |
+
"learning_rate": 0.00015719813573936712,
|
| 2585 |
+
"loss": 2.4918,
|
| 2586 |
+
"step": 367
|
| 2587 |
+
},
|
| 2588 |
+
{
|
| 2589 |
+
"epoch": 0.31302498670919726,
|
| 2590 |
+
"grad_norm": 0.2622525095939636,
|
| 2591 |
+
"learning_rate": 0.00015697692122535107,
|
| 2592 |
+
"loss": 2.4657,
|
| 2593 |
+
"step": 368
|
| 2594 |
+
},
|
| 2595 |
+
{
|
| 2596 |
+
"epoch": 0.3138755980861244,
|
| 2597 |
+
"grad_norm": 0.2790607810020447,
|
| 2598 |
+
"learning_rate": 0.0001567552930914972,
|
| 2599 |
+
"loss": 2.78,
|
| 2600 |
+
"step": 369
|
| 2601 |
+
},
|
| 2602 |
+
{
|
| 2603 |
+
"epoch": 0.3147262094630516,
|
| 2604 |
+
"grad_norm": 0.27601122856140137,
|
| 2605 |
+
"learning_rate": 0.00015653325294669884,
|
| 2606 |
+
"loss": 2.6908,
|
| 2607 |
+
"step": 370
|
| 2608 |
+
},
|
| 2609 |
+
{
|
| 2610 |
+
"epoch": 0.31557682083997873,
|
| 2611 |
+
"grad_norm": 0.3054703176021576,
|
| 2612 |
+
"learning_rate": 0.0001563108024028404,
|
| 2613 |
+
"loss": 2.6602,
|
| 2614 |
+
"step": 371
|
| 2615 |
+
},
|
| 2616 |
+
{
|
| 2617 |
+
"epoch": 0.3164274322169059,
|
| 2618 |
+
"grad_norm": 0.2979956567287445,
|
| 2619 |
+
"learning_rate": 0.00015608794307478546,
|
| 2620 |
+
"loss": 2.8034,
|
| 2621 |
+
"step": 372
|
| 2622 |
+
},
|
| 2623 |
+
{
|
| 2624 |
+
"epoch": 0.31727804359383305,
|
| 2625 |
+
"grad_norm": 0.3022500276565552,
|
| 2626 |
+
"learning_rate": 0.00015586467658036524,
|
| 2627 |
+
"loss": 2.6491,
|
| 2628 |
+
"step": 373
|
| 2629 |
+
},
|
| 2630 |
+
{
|
| 2631 |
+
"epoch": 0.31812865497076026,
|
| 2632 |
+
"grad_norm": 0.3278852701187134,
|
| 2633 |
+
"learning_rate": 0.0001556410045403667,
|
| 2634 |
+
"loss": 2.5928,
|
| 2635 |
+
"step": 374
|
| 2636 |
+
},
|
| 2637 |
+
{
|
| 2638 |
+
"epoch": 0.3189792663476874,
|
| 2639 |
+
"grad_norm": 0.3023948073387146,
|
| 2640 |
+
"learning_rate": 0.0001554169285785208,
|
| 2641 |
+
"loss": 2.6165,
|
| 2642 |
+
"step": 375
|
| 2643 |
+
},
|
| 2644 |
+
{
|
| 2645 |
+
"epoch": 0.3198298777246146,
|
| 2646 |
+
"grad_norm": 0.33025047183036804,
|
| 2647 |
+
"learning_rate": 0.00015519245032149083,
|
| 2648 |
+
"loss": 2.7669,
|
| 2649 |
+
"step": 376
|
| 2650 |
+
},
|
| 2651 |
+
{
|
| 2652 |
+
"epoch": 0.32068048910154173,
|
| 2653 |
+
"grad_norm": 0.3081379532814026,
|
| 2654 |
+
"learning_rate": 0.0001549675713988604,
|
| 2655 |
+
"loss": 2.5991,
|
| 2656 |
+
"step": 377
|
| 2657 |
+
},
|
| 2658 |
+
{
|
| 2659 |
+
"epoch": 0.3215311004784689,
|
| 2660 |
+
"grad_norm": 0.35036811232566833,
|
| 2661 |
+
"learning_rate": 0.0001547422934431218,
|
| 2662 |
+
"loss": 2.6891,
|
| 2663 |
+
"step": 378
|
| 2664 |
+
},
|
| 2665 |
+
{
|
| 2666 |
+
"epoch": 0.32238171185539605,
|
| 2667 |
+
"grad_norm": 0.399915486574173,
|
| 2668 |
+
"learning_rate": 0.00015451661808966405,
|
| 2669 |
+
"loss": 2.8271,
|
| 2670 |
+
"step": 379
|
| 2671 |
+
},
|
| 2672 |
+
{
|
| 2673 |
+
"epoch": 0.32323232323232326,
|
| 2674 |
+
"grad_norm": 0.3355950713157654,
|
| 2675 |
+
"learning_rate": 0.00015429054697676107,
|
| 2676 |
+
"loss": 2.3574,
|
| 2677 |
+
"step": 380
|
| 2678 |
+
},
|
| 2679 |
+
{
|
| 2680 |
+
"epoch": 0.3240829346092504,
|
| 2681 |
+
"grad_norm": 0.35686787962913513,
|
| 2682 |
+
"learning_rate": 0.00015406408174555976,
|
| 2683 |
+
"loss": 2.6926,
|
| 2684 |
+
"step": 381
|
| 2685 |
+
},
|
| 2686 |
+
{
|
| 2687 |
+
"epoch": 0.3249335459861776,
|
| 2688 |
+
"grad_norm": 0.3730961084365845,
|
| 2689 |
+
"learning_rate": 0.00015383722404006806,
|
| 2690 |
+
"loss": 2.7418,
|
| 2691 |
+
"step": 382
|
| 2692 |
+
},
|
| 2693 |
+
{
|
| 2694 |
+
"epoch": 0.32578415736310473,
|
| 2695 |
+
"grad_norm": 0.36210712790489197,
|
| 2696 |
+
"learning_rate": 0.00015360997550714305,
|
| 2697 |
+
"loss": 2.7188,
|
| 2698 |
+
"step": 383
|
| 2699 |
+
},
|
| 2700 |
+
{
|
| 2701 |
+
"epoch": 0.3266347687400319,
|
| 2702 |
+
"grad_norm": 0.4164154827594757,
|
| 2703 |
+
"learning_rate": 0.0001533823377964791,
|
| 2704 |
+
"loss": 2.8182,
|
| 2705 |
+
"step": 384
|
| 2706 |
+
},
|
| 2707 |
+
{
|
| 2708 |
+
"epoch": 0.32748538011695905,
|
| 2709 |
+
"grad_norm": 0.3890111446380615,
|
| 2710 |
+
"learning_rate": 0.0001531543125605956,
|
| 2711 |
+
"loss": 2.6873,
|
| 2712 |
+
"step": 385
|
| 2713 |
+
},
|
| 2714 |
+
{
|
| 2715 |
+
"epoch": 0.3283359914938862,
|
| 2716 |
+
"grad_norm": 0.3929746747016907,
|
| 2717 |
+
"learning_rate": 0.0001529259014548253,
|
| 2718 |
+
"loss": 2.6169,
|
| 2719 |
+
"step": 386
|
| 2720 |
+
},
|
| 2721 |
+
{
|
| 2722 |
+
"epoch": 0.3291866028708134,
|
| 2723 |
+
"grad_norm": 0.4354581832885742,
|
| 2724 |
+
"learning_rate": 0.0001526971061373021,
|
| 2725 |
+
"loss": 2.681,
|
| 2726 |
+
"step": 387
|
| 2727 |
+
},
|
| 2728 |
+
{
|
| 2729 |
+
"epoch": 0.3300372142477406,
|
| 2730 |
+
"grad_norm": 0.4014440178871155,
|
| 2731 |
+
"learning_rate": 0.00015246792826894906,
|
| 2732 |
+
"loss": 2.6601,
|
| 2733 |
+
"step": 388
|
| 2734 |
+
},
|
| 2735 |
+
{
|
| 2736 |
+
"epoch": 0.33088782562466773,
|
| 2737 |
+
"grad_norm": 0.41521206498146057,
|
| 2738 |
+
"learning_rate": 0.00015223836951346634,
|
| 2739 |
+
"loss": 2.7763,
|
| 2740 |
+
"step": 389
|
| 2741 |
+
},
|
| 2742 |
+
{
|
| 2743 |
+
"epoch": 0.3317384370015949,
|
| 2744 |
+
"grad_norm": 0.43811067938804626,
|
| 2745 |
+
"learning_rate": 0.00015200843153731906,
|
| 2746 |
+
"loss": 2.7373,
|
| 2747 |
+
"step": 390
|
| 2748 |
+
},
|
| 2749 |
+
{
|
| 2750 |
+
"epoch": 0.33258904837852205,
|
| 2751 |
+
"grad_norm": 0.4687197208404541,
|
| 2752 |
+
"learning_rate": 0.0001517781160097254,
|
| 2753 |
+
"loss": 2.6432,
|
| 2754 |
+
"step": 391
|
| 2755 |
+
},
|
| 2756 |
+
{
|
| 2757 |
+
"epoch": 0.3334396597554492,
|
| 2758 |
+
"grad_norm": 0.4998420178890228,
|
| 2759 |
+
"learning_rate": 0.00015154742460264425,
|
| 2760 |
+
"loss": 2.6434,
|
| 2761 |
+
"step": 392
|
| 2762 |
+
},
|
| 2763 |
+
{
|
| 2764 |
+
"epoch": 0.3342902711323764,
|
| 2765 |
+
"grad_norm": 0.4983424246311188,
|
| 2766 |
+
"learning_rate": 0.0001513163589907632,
|
| 2767 |
+
"loss": 2.7619,
|
| 2768 |
+
"step": 393
|
| 2769 |
+
},
|
| 2770 |
+
{
|
| 2771 |
+
"epoch": 0.3351408825093036,
|
| 2772 |
+
"grad_norm": 0.5066902041435242,
|
| 2773 |
+
"learning_rate": 0.00015108492085148632,
|
| 2774 |
+
"loss": 2.6515,
|
| 2775 |
+
"step": 394
|
| 2776 |
+
},
|
| 2777 |
+
{
|
| 2778 |
+
"epoch": 0.33599149388623073,
|
| 2779 |
+
"grad_norm": 0.5106616020202637,
|
| 2780 |
+
"learning_rate": 0.00015085311186492206,
|
| 2781 |
+
"loss": 2.6288,
|
| 2782 |
+
"step": 395
|
| 2783 |
+
},
|
| 2784 |
+
{
|
| 2785 |
+
"epoch": 0.3368421052631579,
|
| 2786 |
+
"grad_norm": 0.6036600470542908,
|
| 2787 |
+
"learning_rate": 0.00015062093371387097,
|
| 2788 |
+
"loss": 2.6391,
|
| 2789 |
+
"step": 396
|
| 2790 |
+
},
|
| 2791 |
+
{
|
| 2792 |
+
"epoch": 0.33769271664008504,
|
| 2793 |
+
"grad_norm": 0.6260906457901001,
|
| 2794 |
+
"learning_rate": 0.00015038838808381354,
|
| 2795 |
+
"loss": 2.8623,
|
| 2796 |
+
"step": 397
|
| 2797 |
+
},
|
| 2798 |
+
{
|
| 2799 |
+
"epoch": 0.3385433280170122,
|
| 2800 |
+
"grad_norm": 0.6070663332939148,
|
| 2801 |
+
"learning_rate": 0.00015015547666289797,
|
| 2802 |
+
"loss": 2.615,
|
| 2803 |
+
"step": 398
|
| 2804 |
+
},
|
| 2805 |
+
{
|
| 2806 |
+
"epoch": 0.3393939393939394,
|
| 2807 |
+
"grad_norm": 0.66013503074646,
|
| 2808 |
+
"learning_rate": 0.00014992220114192785,
|
| 2809 |
+
"loss": 2.6865,
|
| 2810 |
+
"step": 399
|
| 2811 |
+
},
|
| 2812 |
+
{
|
| 2813 |
+
"epoch": 0.34024455077086657,
|
| 2814 |
+
"grad_norm": 0.7371327877044678,
|
| 2815 |
+
"learning_rate": 0.00014968856321434998,
|
| 2816 |
+
"loss": 2.6959,
|
| 2817 |
+
"step": 400
|
| 2818 |
+
},
|
| 2819 |
+
{
|
| 2820 |
+
"epoch": 0.34109516214779373,
|
| 2821 |
+
"grad_norm": 0.3081108629703522,
|
| 2822 |
+
"learning_rate": 0.00014945456457624197,
|
| 2823 |
+
"loss": 2.321,
|
| 2824 |
+
"step": 401
|
| 2825 |
+
},
|
| 2826 |
+
{
|
| 2827 |
+
"epoch": 0.3419457735247209,
|
| 2828 |
+
"grad_norm": 0.37835538387298584,
|
| 2829 |
+
"learning_rate": 0.0001492202069263,
|
| 2830 |
+
"loss": 2.5144,
|
| 2831 |
+
"step": 402
|
| 2832 |
+
},
|
| 2833 |
+
{
|
| 2834 |
+
"epoch": 0.34279638490164804,
|
| 2835 |
+
"grad_norm": 0.3184291422367096,
|
| 2836 |
+
"learning_rate": 0.00014898549196582645,
|
| 2837 |
+
"loss": 2.5783,
|
| 2838 |
+
"step": 403
|
| 2839 |
+
},
|
| 2840 |
+
{
|
| 2841 |
+
"epoch": 0.3436469962785752,
|
| 2842 |
+
"grad_norm": 0.3378361165523529,
|
| 2843 |
+
"learning_rate": 0.00014875042139871766,
|
| 2844 |
+
"loss": 2.667,
|
| 2845 |
+
"step": 404
|
| 2846 |
+
},
|
| 2847 |
+
{
|
| 2848 |
+
"epoch": 0.3444976076555024,
|
| 2849 |
+
"grad_norm": 0.285756379365921,
|
| 2850 |
+
"learning_rate": 0.00014851499693145135,
|
| 2851 |
+
"loss": 2.5316,
|
| 2852 |
+
"step": 405
|
| 2853 |
+
},
|
| 2854 |
+
{
|
| 2855 |
+
"epoch": 0.34534821903242957,
|
| 2856 |
+
"grad_norm": 0.28837427496910095,
|
| 2857 |
+
"learning_rate": 0.00014827922027307451,
|
| 2858 |
+
"loss": 2.5012,
|
| 2859 |
+
"step": 406
|
| 2860 |
+
},
|
| 2861 |
+
{
|
| 2862 |
+
"epoch": 0.34619883040935673,
|
| 2863 |
+
"grad_norm": 0.30004727840423584,
|
| 2864 |
+
"learning_rate": 0.0001480430931351906,
|
| 2865 |
+
"loss": 2.5237,
|
| 2866 |
+
"step": 407
|
| 2867 |
+
},
|
| 2868 |
+
{
|
| 2869 |
+
"epoch": 0.3470494417862839,
|
| 2870 |
+
"grad_norm": 0.28527382016181946,
|
| 2871 |
+
"learning_rate": 0.00014780661723194757,
|
| 2872 |
+
"loss": 2.5067,
|
| 2873 |
+
"step": 408
|
| 2874 |
+
},
|
| 2875 |
+
{
|
| 2876 |
+
"epoch": 0.34790005316321104,
|
| 2877 |
+
"grad_norm": 0.26983842253685,
|
| 2878 |
+
"learning_rate": 0.00014756979428002514,
|
| 2879 |
+
"loss": 2.5577,
|
| 2880 |
+
"step": 409
|
| 2881 |
+
},
|
| 2882 |
+
{
|
| 2883 |
+
"epoch": 0.3487506645401382,
|
| 2884 |
+
"grad_norm": 0.2686617076396942,
|
| 2885 |
+
"learning_rate": 0.00014733262599862234,
|
| 2886 |
+
"loss": 2.5267,
|
| 2887 |
+
"step": 410
|
| 2888 |
+
},
|
| 2889 |
+
{
|
| 2890 |
+
"epoch": 0.3496012759170654,
|
| 2891 |
+
"grad_norm": 0.26147857308387756,
|
| 2892 |
+
"learning_rate": 0.00014709511410944523,
|
| 2893 |
+
"loss": 2.4459,
|
| 2894 |
+
"step": 411
|
| 2895 |
+
},
|
| 2896 |
+
{
|
| 2897 |
+
"epoch": 0.35045188729399257,
|
| 2898 |
+
"grad_norm": 0.29142996668815613,
|
| 2899 |
+
"learning_rate": 0.00014685726033669412,
|
| 2900 |
+
"loss": 2.7246,
|
| 2901 |
+
"step": 412
|
| 2902 |
+
},
|
| 2903 |
+
{
|
| 2904 |
+
"epoch": 0.3513024986709197,
|
| 2905 |
+
"grad_norm": 0.29214030504226685,
|
| 2906 |
+
"learning_rate": 0.00014661906640705129,
|
| 2907 |
+
"loss": 2.6422,
|
| 2908 |
+
"step": 413
|
| 2909 |
+
},
|
| 2910 |
+
{
|
| 2911 |
+
"epoch": 0.3521531100478469,
|
| 2912 |
+
"grad_norm": 0.2727803885936737,
|
| 2913 |
+
"learning_rate": 0.00014638053404966836,
|
| 2914 |
+
"loss": 2.6416,
|
| 2915 |
+
"step": 414
|
| 2916 |
+
},
|
| 2917 |
+
{
|
| 2918 |
+
"epoch": 0.35300372142477404,
|
| 2919 |
+
"grad_norm": 0.2637098431587219,
|
| 2920 |
+
"learning_rate": 0.0001461416649961537,
|
| 2921 |
+
"loss": 2.5555,
|
| 2922 |
+
"step": 415
|
| 2923 |
+
},
|
| 2924 |
+
{
|
| 2925 |
+
"epoch": 0.3538543328017012,
|
| 2926 |
+
"grad_norm": 0.2736997902393341,
|
| 2927 |
+
"learning_rate": 0.00014590246098055996,
|
| 2928 |
+
"loss": 2.7423,
|
| 2929 |
+
"step": 416
|
| 2930 |
+
},
|
| 2931 |
+
{
|
| 2932 |
+
"epoch": 0.3547049441786284,
|
| 2933 |
+
"grad_norm": 0.2691054940223694,
|
| 2934 |
+
"learning_rate": 0.0001456629237393713,
|
| 2935 |
+
"loss": 2.7818,
|
| 2936 |
+
"step": 417
|
| 2937 |
+
},
|
| 2938 |
+
{
|
| 2939 |
+
"epoch": 0.35555555555555557,
|
| 2940 |
+
"grad_norm": 0.28304579854011536,
|
| 2941 |
+
"learning_rate": 0.0001454230550114911,
|
| 2942 |
+
"loss": 2.554,
|
| 2943 |
+
"step": 418
|
| 2944 |
+
},
|
| 2945 |
+
{
|
| 2946 |
+
"epoch": 0.3564061669324827,
|
| 2947 |
+
"grad_norm": 0.2698013484477997,
|
| 2948 |
+
"learning_rate": 0.00014518285653822898,
|
| 2949 |
+
"loss": 2.5155,
|
| 2950 |
+
"step": 419
|
| 2951 |
+
},
|
| 2952 |
+
{
|
| 2953 |
+
"epoch": 0.3572567783094099,
|
| 2954 |
+
"grad_norm": 0.25874024629592896,
|
| 2955 |
+
"learning_rate": 0.00014494233006328837,
|
| 2956 |
+
"loss": 2.4841,
|
| 2957 |
+
"step": 420
|
| 2958 |
+
},
|
| 2959 |
+
{
|
| 2960 |
+
"epoch": 0.35810738968633704,
|
| 2961 |
+
"grad_norm": 0.2918216586112976,
|
| 2962 |
+
"learning_rate": 0.00014470147733275387,
|
| 2963 |
+
"loss": 2.7489,
|
| 2964 |
+
"step": 421
|
| 2965 |
+
},
|
| 2966 |
+
{
|
| 2967 |
+
"epoch": 0.3589580010632642,
|
| 2968 |
+
"grad_norm": 0.28201210498809814,
|
| 2969 |
+
"learning_rate": 0.0001444603000950784,
|
| 2970 |
+
"loss": 2.5709,
|
| 2971 |
+
"step": 422
|
| 2972 |
+
},
|
| 2973 |
+
{
|
| 2974 |
+
"epoch": 0.3598086124401914,
|
| 2975 |
+
"grad_norm": 0.3077748119831085,
|
| 2976 |
+
"learning_rate": 0.0001442188001010707,
|
| 2977 |
+
"loss": 2.6089,
|
| 2978 |
+
"step": 423
|
| 2979 |
+
},
|
| 2980 |
+
{
|
| 2981 |
+
"epoch": 0.36065922381711857,
|
| 2982 |
+
"grad_norm": 0.2970117926597595,
|
| 2983 |
+
"learning_rate": 0.00014397697910388248,
|
| 2984 |
+
"loss": 2.6171,
|
| 2985 |
+
"step": 424
|
| 2986 |
+
},
|
| 2987 |
+
{
|
| 2988 |
+
"epoch": 0.3615098351940457,
|
| 2989 |
+
"grad_norm": 0.302947074174881,
|
| 2990 |
+
"learning_rate": 0.00014373483885899582,
|
| 2991 |
+
"loss": 2.4033,
|
| 2992 |
+
"step": 425
|
| 2993 |
+
},
|
| 2994 |
+
{
|
| 2995 |
+
"epoch": 0.3623604465709729,
|
| 2996 |
+
"grad_norm": 0.31938815116882324,
|
| 2997 |
+
"learning_rate": 0.00014349238112421024,
|
| 2998 |
+
"loss": 2.7042,
|
| 2999 |
+
"step": 426
|
| 3000 |
+
},
|
| 3001 |
+
{
|
| 3002 |
+
"epoch": 0.36321105794790004,
|
| 3003 |
+
"grad_norm": 0.3481922447681427,
|
| 3004 |
+
"learning_rate": 0.00014324960765963018,
|
| 3005 |
+
"loss": 2.6479,
|
| 3006 |
+
"step": 427
|
| 3007 |
+
},
|
| 3008 |
+
{
|
| 3009 |
+
"epoch": 0.3640616693248272,
|
| 3010 |
+
"grad_norm": 0.3142179846763611,
|
| 3011 |
+
"learning_rate": 0.00014300652022765207,
|
| 3012 |
+
"loss": 2.5285,
|
| 3013 |
+
"step": 428
|
| 3014 |
+
},
|
| 3015 |
+
{
|
| 3016 |
+
"epoch": 0.3649122807017544,
|
| 3017 |
+
"grad_norm": 0.3226439952850342,
|
| 3018 |
+
"learning_rate": 0.00014276312059295147,
|
| 3019 |
+
"loss": 2.5389,
|
| 3020 |
+
"step": 429
|
| 3021 |
+
},
|
| 3022 |
+
{
|
| 3023 |
+
"epoch": 0.36576289207868157,
|
| 3024 |
+
"grad_norm": 0.401301771402359,
|
| 3025 |
+
"learning_rate": 0.00014251941052247045,
|
| 3026 |
+
"loss": 2.8749,
|
| 3027 |
+
"step": 430
|
| 3028 |
+
},
|
| 3029 |
+
{
|
| 3030 |
+
"epoch": 0.3666135034556087,
|
| 3031 |
+
"grad_norm": 0.3587849736213684,
|
| 3032 |
+
"learning_rate": 0.00014227539178540463,
|
| 3033 |
+
"loss": 2.7549,
|
| 3034 |
+
"step": 431
|
| 3035 |
+
},
|
| 3036 |
+
{
|
| 3037 |
+
"epoch": 0.3674641148325359,
|
| 3038 |
+
"grad_norm": 0.38098907470703125,
|
| 3039 |
+
"learning_rate": 0.00014203106615319038,
|
| 3040 |
+
"loss": 2.6159,
|
| 3041 |
+
"step": 432
|
| 3042 |
+
},
|
| 3043 |
+
{
|
| 3044 |
+
"epoch": 0.36831472620946304,
|
| 3045 |
+
"grad_norm": 0.402971088886261,
|
| 3046 |
+
"learning_rate": 0.00014178643539949196,
|
| 3047 |
+
"loss": 2.6626,
|
| 3048 |
+
"step": 433
|
| 3049 |
+
},
|
| 3050 |
+
{
|
| 3051 |
+
"epoch": 0.3691653375863902,
|
| 3052 |
+
"grad_norm": 0.3960564136505127,
|
| 3053 |
+
"learning_rate": 0.00014154150130018866,
|
| 3054 |
+
"loss": 2.8099,
|
| 3055 |
+
"step": 434
|
| 3056 |
+
},
|
| 3057 |
+
{
|
| 3058 |
+
"epoch": 0.3700159489633174,
|
| 3059 |
+
"grad_norm": 0.41555696725845337,
|
| 3060 |
+
"learning_rate": 0.00014129626563336178,
|
| 3061 |
+
"loss": 2.7282,
|
| 3062 |
+
"step": 435
|
| 3063 |
+
},
|
| 3064 |
+
{
|
| 3065 |
+
"epoch": 0.37086656034024457,
|
| 3066 |
+
"grad_norm": 0.3991285264492035,
|
| 3067 |
+
"learning_rate": 0.000141050730179282,
|
| 3068 |
+
"loss": 2.5909,
|
| 3069 |
+
"step": 436
|
| 3070 |
+
},
|
| 3071 |
+
{
|
| 3072 |
+
"epoch": 0.3717171717171717,
|
| 3073 |
+
"grad_norm": 0.442220002412796,
|
| 3074 |
+
"learning_rate": 0.00014080489672039606,
|
| 3075 |
+
"loss": 2.8671,
|
| 3076 |
+
"step": 437
|
| 3077 |
+
},
|
| 3078 |
+
{
|
| 3079 |
+
"epoch": 0.3725677830940989,
|
| 3080 |
+
"grad_norm": 0.4263823628425598,
|
| 3081 |
+
"learning_rate": 0.0001405587670413143,
|
| 3082 |
+
"loss": 2.5901,
|
| 3083 |
+
"step": 438
|
| 3084 |
+
},
|
| 3085 |
+
{
|
| 3086 |
+
"epoch": 0.37341839447102604,
|
| 3087 |
+
"grad_norm": 0.45683711767196655,
|
| 3088 |
+
"learning_rate": 0.00014031234292879725,
|
| 3089 |
+
"loss": 2.7801,
|
| 3090 |
+
"step": 439
|
| 3091 |
+
},
|
| 3092 |
+
{
|
| 3093 |
+
"epoch": 0.3742690058479532,
|
| 3094 |
+
"grad_norm": 0.4672732353210449,
|
| 3095 |
+
"learning_rate": 0.00014006562617174294,
|
| 3096 |
+
"loss": 2.696,
|
| 3097 |
+
"step": 440
|
| 3098 |
+
},
|
| 3099 |
+
{
|
| 3100 |
+
"epoch": 0.3751196172248804,
|
| 3101 |
+
"grad_norm": 0.45676231384277344,
|
| 3102 |
+
"learning_rate": 0.0001398186185611738,
|
| 3103 |
+
"loss": 2.6197,
|
| 3104 |
+
"step": 441
|
| 3105 |
+
},
|
| 3106 |
+
{
|
| 3107 |
+
"epoch": 0.37597022860180757,
|
| 3108 |
+
"grad_norm": 0.4717809855937958,
|
| 3109 |
+
"learning_rate": 0.00013957132189022374,
|
| 3110 |
+
"loss": 2.7676,
|
| 3111 |
+
"step": 442
|
| 3112 |
+
},
|
| 3113 |
+
{
|
| 3114 |
+
"epoch": 0.3768208399787347,
|
| 3115 |
+
"grad_norm": 0.4931057393550873,
|
| 3116 |
+
"learning_rate": 0.00013932373795412503,
|
| 3117 |
+
"loss": 2.7968,
|
| 3118 |
+
"step": 443
|
| 3119 |
+
},
|
| 3120 |
+
{
|
| 3121 |
+
"epoch": 0.3776714513556619,
|
| 3122 |
+
"grad_norm": 0.5112429857254028,
|
| 3123 |
+
"learning_rate": 0.0001390758685501954,
|
| 3124 |
+
"loss": 2.5947,
|
| 3125 |
+
"step": 444
|
| 3126 |
+
},
|
| 3127 |
+
{
|
| 3128 |
+
"epoch": 0.37852206273258904,
|
| 3129 |
+
"grad_norm": 0.5512686371803284,
|
| 3130 |
+
"learning_rate": 0.00013882771547782475,
|
| 3131 |
+
"loss": 2.6041,
|
| 3132 |
+
"step": 445
|
| 3133 |
+
},
|
| 3134 |
+
{
|
| 3135 |
+
"epoch": 0.3793726741095162,
|
| 3136 |
+
"grad_norm": 0.5506839752197266,
|
| 3137 |
+
"learning_rate": 0.0001385792805384625,
|
| 3138 |
+
"loss": 2.7389,
|
| 3139 |
+
"step": 446
|
| 3140 |
+
},
|
| 3141 |
+
{
|
| 3142 |
+
"epoch": 0.3802232854864434,
|
| 3143 |
+
"grad_norm": 0.5831094980239868,
|
| 3144 |
+
"learning_rate": 0.00013833056553560398,
|
| 3145 |
+
"loss": 2.8867,
|
| 3146 |
+
"step": 447
|
| 3147 |
+
},
|
| 3148 |
+
{
|
| 3149 |
+
"epoch": 0.38107389686337056,
|
| 3150 |
+
"grad_norm": 0.6127706170082092,
|
| 3151 |
+
"learning_rate": 0.00013808157227477788,
|
| 3152 |
+
"loss": 2.689,
|
| 3153 |
+
"step": 448
|
| 3154 |
+
},
|
| 3155 |
+
{
|
| 3156 |
+
"epoch": 0.3819245082402977,
|
| 3157 |
+
"grad_norm": 0.6911614537239075,
|
| 3158 |
+
"learning_rate": 0.00013783230256353266,
|
| 3159 |
+
"loss": 2.9826,
|
| 3160 |
+
"step": 449
|
| 3161 |
+
},
|
| 3162 |
+
{
|
| 3163 |
+
"epoch": 0.3827751196172249,
|
| 3164 |
+
"grad_norm": 0.7095205187797546,
|
| 3165 |
+
"learning_rate": 0.00013758275821142382,
|
| 3166 |
+
"loss": 2.7725,
|
| 3167 |
+
"step": 450
|
| 3168 |
+
},
|
| 3169 |
+
{
|
| 3170 |
+
"epoch": 0.38362573099415204,
|
| 3171 |
+
"grad_norm": 0.31148967146873474,
|
| 3172 |
+
"learning_rate": 0.00013733294103000055,
|
| 3173 |
+
"loss": 2.5203,
|
| 3174 |
+
"step": 451
|
| 3175 |
+
},
|
| 3176 |
+
{
|
| 3177 |
+
"epoch": 0.3844763423710792,
|
| 3178 |
+
"grad_norm": 0.3002786338329315,
|
| 3179 |
+
"learning_rate": 0.00013708285283279252,
|
| 3180 |
+
"loss": 2.5911,
|
| 3181 |
+
"step": 452
|
| 3182 |
+
},
|
| 3183 |
+
{
|
| 3184 |
+
"epoch": 0.3853269537480064,
|
| 3185 |
+
"grad_norm": 0.29820379614830017,
|
| 3186 |
+
"learning_rate": 0.00013683249543529696,
|
| 3187 |
+
"loss": 2.3441,
|
| 3188 |
+
"step": 453
|
| 3189 |
+
},
|
| 3190 |
+
{
|
| 3191 |
+
"epoch": 0.38617756512493356,
|
| 3192 |
+
"grad_norm": 0.29162371158599854,
|
| 3193 |
+
"learning_rate": 0.00013658187065496532,
|
| 3194 |
+
"loss": 2.4791,
|
| 3195 |
+
"step": 454
|
| 3196 |
+
},
|
| 3197 |
+
{
|
| 3198 |
+
"epoch": 0.3870281765018607,
|
| 3199 |
+
"grad_norm": 0.2905353307723999,
|
| 3200 |
+
"learning_rate": 0.00013633098031119002,
|
| 3201 |
+
"loss": 2.4208,
|
| 3202 |
+
"step": 455
|
| 3203 |
+
},
|
| 3204 |
+
{
|
| 3205 |
+
"epoch": 0.3878787878787879,
|
| 3206 |
+
"grad_norm": 0.28394779562950134,
|
| 3207 |
+
"learning_rate": 0.00013607982622529133,
|
| 3208 |
+
"loss": 2.6468,
|
| 3209 |
+
"step": 456
|
| 3210 |
+
},
|
| 3211 |
+
{
|
| 3212 |
+
"epoch": 0.38872939925571504,
|
| 3213 |
+
"grad_norm": 0.2683579921722412,
|
| 3214 |
+
"learning_rate": 0.00013582841022050424,
|
| 3215 |
+
"loss": 2.5199,
|
| 3216 |
+
"step": 457
|
| 3217 |
+
},
|
| 3218 |
+
{
|
| 3219 |
+
"epoch": 0.3895800106326422,
|
| 3220 |
+
"grad_norm": 0.2719745934009552,
|
| 3221 |
+
"learning_rate": 0.00013557673412196503,
|
| 3222 |
+
"loss": 2.4813,
|
| 3223 |
+
"step": 458
|
| 3224 |
+
},
|
| 3225 |
+
{
|
| 3226 |
+
"epoch": 0.39043062200956935,
|
| 3227 |
+
"grad_norm": 0.26781052350997925,
|
| 3228 |
+
"learning_rate": 0.00013532479975669808,
|
| 3229 |
+
"loss": 2.5296,
|
| 3230 |
+
"step": 459
|
| 3231 |
+
},
|
| 3232 |
+
{
|
| 3233 |
+
"epoch": 0.39128123338649656,
|
| 3234 |
+
"grad_norm": 0.26786506175994873,
|
| 3235 |
+
"learning_rate": 0.00013507260895360274,
|
| 3236 |
+
"loss": 2.5254,
|
| 3237 |
+
"step": 460
|
| 3238 |
+
},
|
| 3239 |
+
{
|
| 3240 |
+
"epoch": 0.3921318447634237,
|
| 3241 |
+
"grad_norm": 0.27148234844207764,
|
| 3242 |
+
"learning_rate": 0.0001348201635434399,
|
| 3243 |
+
"loss": 2.6415,
|
| 3244 |
+
"step": 461
|
| 3245 |
+
},
|
| 3246 |
+
{
|
| 3247 |
+
"epoch": 0.3929824561403509,
|
| 3248 |
+
"grad_norm": 0.29129862785339355,
|
| 3249 |
+
"learning_rate": 0.00013456746535881871,
|
| 3250 |
+
"loss": 2.6098,
|
| 3251 |
+
"step": 462
|
| 3252 |
+
},
|
| 3253 |
+
{
|
| 3254 |
+
"epoch": 0.39383306751727803,
|
| 3255 |
+
"grad_norm": 0.2636375427246094,
|
| 3256 |
+
"learning_rate": 0.00013431451623418343,
|
| 3257 |
+
"loss": 2.4875,
|
| 3258 |
+
"step": 463
|
| 3259 |
+
},
|
| 3260 |
+
{
|
| 3261 |
+
"epoch": 0.3946836788942052,
|
| 3262 |
+
"grad_norm": 0.2701190114021301,
|
| 3263 |
+
"learning_rate": 0.00013406131800579985,
|
| 3264 |
+
"loss": 2.7151,
|
| 3265 |
+
"step": 464
|
| 3266 |
+
},
|
| 3267 |
+
{
|
| 3268 |
+
"epoch": 0.39553429027113235,
|
| 3269 |
+
"grad_norm": 0.2739466428756714,
|
| 3270 |
+
"learning_rate": 0.00013380787251174225,
|
| 3271 |
+
"loss": 2.6465,
|
| 3272 |
+
"step": 465
|
| 3273 |
+
},
|
| 3274 |
+
{
|
| 3275 |
+
"epoch": 0.39638490164805956,
|
| 3276 |
+
"grad_norm": 0.2672434151172638,
|
| 3277 |
+
"learning_rate": 0.00013355418159187985,
|
| 3278 |
+
"loss": 2.6823,
|
| 3279 |
+
"step": 466
|
| 3280 |
+
},
|
| 3281 |
+
{
|
| 3282 |
+
"epoch": 0.3972355130249867,
|
| 3283 |
+
"grad_norm": 0.26863133907318115,
|
| 3284 |
+
"learning_rate": 0.00013330024708786353,
|
| 3285 |
+
"loss": 2.5357,
|
| 3286 |
+
"step": 467
|
| 3287 |
+
},
|
| 3288 |
+
{
|
| 3289 |
+
"epoch": 0.3980861244019139,
|
| 3290 |
+
"grad_norm": 0.2704436779022217,
|
| 3291 |
+
"learning_rate": 0.00013304607084311244,
|
| 3292 |
+
"loss": 2.6887,
|
| 3293 |
+
"step": 468
|
| 3294 |
+
},
|
| 3295 |
+
{
|
| 3296 |
+
"epoch": 0.39893673577884103,
|
| 3297 |
+
"grad_norm": 0.2772809863090515,
|
| 3298 |
+
"learning_rate": 0.00013279165470280065,
|
| 3299 |
+
"loss": 2.6992,
|
| 3300 |
+
"step": 469
|
| 3301 |
+
},
|
| 3302 |
+
{
|
| 3303 |
+
"epoch": 0.3997873471557682,
|
| 3304 |
+
"grad_norm": 0.3024834394454956,
|
| 3305 |
+
"learning_rate": 0.0001325370005138437,
|
| 3306 |
+
"loss": 2.753,
|
| 3307 |
+
"step": 470
|
| 3308 |
+
},
|
| 3309 |
+
{
|
| 3310 |
+
"epoch": 0.40063795853269535,
|
| 3311 |
+
"grad_norm": 0.2810865342617035,
|
| 3312 |
+
"learning_rate": 0.00013228211012488532,
|
| 3313 |
+
"loss": 2.7921,
|
| 3314 |
+
"step": 471
|
| 3315 |
+
},
|
| 3316 |
+
{
|
| 3317 |
+
"epoch": 0.40148856990962256,
|
| 3318 |
+
"grad_norm": 0.28023043274879456,
|
| 3319 |
+
"learning_rate": 0.00013202698538628376,
|
| 3320 |
+
"loss": 2.2935,
|
| 3321 |
+
"step": 472
|
| 3322 |
+
},
|
| 3323 |
+
{
|
| 3324 |
+
"epoch": 0.4023391812865497,
|
| 3325 |
+
"grad_norm": 0.30379030108451843,
|
| 3326 |
+
"learning_rate": 0.0001317716281500987,
|
| 3327 |
+
"loss": 2.5575,
|
| 3328 |
+
"step": 473
|
| 3329 |
+
},
|
| 3330 |
+
{
|
| 3331 |
+
"epoch": 0.4031897926634769,
|
| 3332 |
+
"grad_norm": 0.29648759961128235,
|
| 3333 |
+
"learning_rate": 0.00013151604027007745,
|
| 3334 |
+
"loss": 2.4384,
|
| 3335 |
+
"step": 474
|
| 3336 |
+
},
|
| 3337 |
+
{
|
| 3338 |
+
"epoch": 0.40404040404040403,
|
| 3339 |
+
"grad_norm": 0.32276931405067444,
|
| 3340 |
+
"learning_rate": 0.00013126022360164172,
|
| 3341 |
+
"loss": 2.7529,
|
| 3342 |
+
"step": 475
|
| 3343 |
+
},
|
| 3344 |
+
{
|
| 3345 |
+
"epoch": 0.4048910154173312,
|
| 3346 |
+
"grad_norm": 0.3146274983882904,
|
| 3347 |
+
"learning_rate": 0.00013100418000187419,
|
| 3348 |
+
"loss": 2.473,
|
| 3349 |
+
"step": 476
|
| 3350 |
+
},
|
| 3351 |
+
{
|
| 3352 |
+
"epoch": 0.40574162679425835,
|
| 3353 |
+
"grad_norm": 0.33162757754325867,
|
| 3354 |
+
"learning_rate": 0.00013074791132950485,
|
| 3355 |
+
"loss": 2.5832,
|
| 3356 |
+
"step": 477
|
| 3357 |
+
},
|
| 3358 |
+
{
|
| 3359 |
+
"epoch": 0.40659223817118556,
|
| 3360 |
+
"grad_norm": 0.33521875739097595,
|
| 3361 |
+
"learning_rate": 0.00013049141944489748,
|
| 3362 |
+
"loss": 2.4807,
|
| 3363 |
+
"step": 478
|
| 3364 |
+
},
|
| 3365 |
+
{
|
| 3366 |
+
"epoch": 0.4074428495481127,
|
| 3367 |
+
"grad_norm": 0.3724415898323059,
|
| 3368 |
+
"learning_rate": 0.00013023470621003643,
|
| 3369 |
+
"loss": 2.8245,
|
| 3370 |
+
"step": 479
|
| 3371 |
+
},
|
| 3372 |
+
{
|
| 3373 |
+
"epoch": 0.4082934609250399,
|
| 3374 |
+
"grad_norm": 0.3634830117225647,
|
| 3375 |
+
"learning_rate": 0.00012997777348851288,
|
| 3376 |
+
"loss": 2.5921,
|
| 3377 |
+
"step": 480
|
| 3378 |
+
},
|
| 3379 |
+
{
|
| 3380 |
+
"epoch": 0.40914407230196703,
|
| 3381 |
+
"grad_norm": 0.38656124472618103,
|
| 3382 |
+
"learning_rate": 0.0001297206231455113,
|
| 3383 |
+
"loss": 2.619,
|
| 3384 |
+
"step": 481
|
| 3385 |
+
},
|
| 3386 |
+
{
|
| 3387 |
+
"epoch": 0.4099946836788942,
|
| 3388 |
+
"grad_norm": 0.3939076066017151,
|
| 3389 |
+
"learning_rate": 0.00012946325704779602,
|
| 3390 |
+
"loss": 2.7466,
|
| 3391 |
+
"step": 482
|
| 3392 |
+
},
|
| 3393 |
+
{
|
| 3394 |
+
"epoch": 0.41084529505582135,
|
| 3395 |
+
"grad_norm": 0.37405261397361755,
|
| 3396 |
+
"learning_rate": 0.00012920567706369758,
|
| 3397 |
+
"loss": 2.764,
|
| 3398 |
+
"step": 483
|
| 3399 |
+
},
|
| 3400 |
+
{
|
| 3401 |
+
"epoch": 0.41169590643274856,
|
| 3402 |
+
"grad_norm": 0.3894766569137573,
|
| 3403 |
+
"learning_rate": 0.0001289478850630993,
|
| 3404 |
+
"loss": 2.7373,
|
| 3405 |
+
"step": 484
|
| 3406 |
+
},
|
| 3407 |
+
{
|
| 3408 |
+
"epoch": 0.4125465178096757,
|
| 3409 |
+
"grad_norm": 0.44012099504470825,
|
| 3410 |
+
"learning_rate": 0.00012868988291742347,
|
| 3411 |
+
"loss": 2.6475,
|
| 3412 |
+
"step": 485
|
| 3413 |
+
},
|
| 3414 |
+
{
|
| 3415 |
+
"epoch": 0.4133971291866029,
|
| 3416 |
+
"grad_norm": 0.41175583004951477,
|
| 3417 |
+
"learning_rate": 0.0001284316724996181,
|
| 3418 |
+
"loss": 2.7573,
|
| 3419 |
+
"step": 486
|
| 3420 |
+
},
|
| 3421 |
+
{
|
| 3422 |
+
"epoch": 0.41424774056353003,
|
| 3423 |
+
"grad_norm": 0.4406805634498596,
|
| 3424 |
+
"learning_rate": 0.00012817325568414297,
|
| 3425 |
+
"loss": 2.5429,
|
| 3426 |
+
"step": 487
|
| 3427 |
+
},
|
| 3428 |
+
{
|
| 3429 |
+
"epoch": 0.4150983519404572,
|
| 3430 |
+
"grad_norm": 0.4783489406108856,
|
| 3431 |
+
"learning_rate": 0.0001279146343469563,
|
| 3432 |
+
"loss": 2.8275,
|
| 3433 |
+
"step": 488
|
| 3434 |
+
},
|
| 3435 |
+
{
|
| 3436 |
+
"epoch": 0.41594896331738435,
|
| 3437 |
+
"grad_norm": 0.44115763902664185,
|
| 3438 |
+
"learning_rate": 0.00012765581036550095,
|
| 3439 |
+
"loss": 2.6858,
|
| 3440 |
+
"step": 489
|
| 3441 |
+
},
|
| 3442 |
+
{
|
| 3443 |
+
"epoch": 0.41679957469431156,
|
| 3444 |
+
"grad_norm": 0.5237467885017395,
|
| 3445 |
+
"learning_rate": 0.0001273967856186909,
|
| 3446 |
+
"loss": 2.8638,
|
| 3447 |
+
"step": 490
|
| 3448 |
+
},
|
| 3449 |
+
{
|
| 3450 |
+
"epoch": 0.4176501860712387,
|
| 3451 |
+
"grad_norm": 0.5071147680282593,
|
| 3452 |
+
"learning_rate": 0.00012713756198689757,
|
| 3453 |
+
"loss": 2.8603,
|
| 3454 |
+
"step": 491
|
| 3455 |
+
},
|
| 3456 |
+
{
|
| 3457 |
+
"epoch": 0.4185007974481659,
|
| 3458 |
+
"grad_norm": 0.5125464797019958,
|
| 3459 |
+
"learning_rate": 0.00012687814135193612,
|
| 3460 |
+
"loss": 3.0048,
|
| 3461 |
+
"step": 492
|
| 3462 |
+
},
|
| 3463 |
+
{
|
| 3464 |
+
"epoch": 0.41935140882509303,
|
| 3465 |
+
"grad_norm": 0.5373572707176208,
|
| 3466 |
+
"learning_rate": 0.0001266185255970519,
|
| 3467 |
+
"loss": 2.5727,
|
| 3468 |
+
"step": 493
|
| 3469 |
+
},
|
| 3470 |
+
{
|
| 3471 |
+
"epoch": 0.4202020202020202,
|
| 3472 |
+
"grad_norm": 0.4913314878940582,
|
| 3473 |
+
"learning_rate": 0.00012635871660690676,
|
| 3474 |
+
"loss": 2.6501,
|
| 3475 |
+
"step": 494
|
| 3476 |
+
},
|
| 3477 |
+
{
|
| 3478 |
+
"epoch": 0.42105263157894735,
|
| 3479 |
+
"grad_norm": 0.5953575968742371,
|
| 3480 |
+
"learning_rate": 0.00012609871626756522,
|
| 3481 |
+
"loss": 2.8674,
|
| 3482 |
+
"step": 495
|
| 3483 |
+
},
|
| 3484 |
+
{
|
| 3485 |
+
"epoch": 0.42190324295587456,
|
| 3486 |
+
"grad_norm": 0.5852685570716858,
|
| 3487 |
+
"learning_rate": 0.00012583852646648095,
|
| 3488 |
+
"loss": 2.6855,
|
| 3489 |
+
"step": 496
|
| 3490 |
+
},
|
| 3491 |
+
{
|
| 3492 |
+
"epoch": 0.4227538543328017,
|
| 3493 |
+
"grad_norm": 0.6449065804481506,
|
| 3494 |
+
"learning_rate": 0.00012557814909248296,
|
| 3495 |
+
"loss": 2.7688,
|
| 3496 |
+
"step": 497
|
| 3497 |
+
},
|
| 3498 |
+
{
|
| 3499 |
+
"epoch": 0.4236044657097289,
|
| 3500 |
+
"grad_norm": 0.6615833640098572,
|
| 3501 |
+
"learning_rate": 0.000125317586035762,
|
| 3502 |
+
"loss": 2.7446,
|
| 3503 |
+
"step": 498
|
| 3504 |
+
},
|
| 3505 |
+
{
|
| 3506 |
+
"epoch": 0.42445507708665603,
|
| 3507 |
+
"grad_norm": 0.6561222076416016,
|
| 3508 |
+
"learning_rate": 0.0001250568391878567,
|
| 3509 |
+
"loss": 2.801,
|
| 3510 |
+
"step": 499
|
| 3511 |
+
},
|
| 3512 |
+
{
|
| 3513 |
+
"epoch": 0.4253056884635832,
|
| 3514 |
+
"grad_norm": 0.8370924592018127,
|
| 3515 |
+
"learning_rate": 0.00012479591044163997,
|
| 3516 |
+
"loss": 2.9991,
|
| 3517 |
+
"step": 500
|
| 3518 |
+
},
|
| 3519 |
+
{
|
| 3520 |
+
"epoch": 0.42615629984051034,
|
| 3521 |
+
"grad_norm": 0.3273813724517822,
|
| 3522 |
+
"learning_rate": 0.0001245348016913051,
|
| 3523 |
+
"loss": 2.5914,
|
| 3524 |
+
"step": 501
|
| 3525 |
+
},
|
| 3526 |
+
{
|
| 3527 |
+
"epoch": 0.42700691121743756,
|
| 3528 |
+
"grad_norm": 0.3103164732456207,
|
| 3529 |
+
"learning_rate": 0.00012427351483235223,
|
| 3530 |
+
"loss": 2.6331,
|
| 3531 |
+
"step": 502
|
| 3532 |
+
},
|
| 3533 |
+
{
|
| 3534 |
+
"epoch": 0.4278575225943647,
|
| 3535 |
+
"grad_norm": 0.3061564862728119,
|
| 3536 |
+
"learning_rate": 0.00012401205176157447,
|
| 3537 |
+
"loss": 2.5927,
|
| 3538 |
+
"step": 503
|
| 3539 |
+
},
|
| 3540 |
+
{
|
| 3541 |
+
"epoch": 0.42870813397129187,
|
| 3542 |
+
"grad_norm": 0.2943616509437561,
|
| 3543 |
+
"learning_rate": 0.00012375041437704393,
|
| 3544 |
+
"loss": 2.4734,
|
| 3545 |
+
"step": 504
|
| 3546 |
+
},
|
| 3547 |
+
{
|
| 3548 |
+
"epoch": 0.42955874534821903,
|
| 3549 |
+
"grad_norm": 0.29883497953414917,
|
| 3550 |
+
"learning_rate": 0.00012348860457809838,
|
| 3551 |
+
"loss": 2.4734,
|
| 3552 |
+
"step": 505
|
| 3553 |
+
},
|
| 3554 |
+
{
|
| 3555 |
+
"epoch": 0.4304093567251462,
|
| 3556 |
+
"grad_norm": 0.295578271150589,
|
| 3557 |
+
"learning_rate": 0.00012322662426532708,
|
| 3558 |
+
"loss": 2.3561,
|
| 3559 |
+
"step": 506
|
| 3560 |
+
},
|
| 3561 |
+
{
|
| 3562 |
+
"epoch": 0.43125996810207334,
|
| 3563 |
+
"grad_norm": 0.3073442578315735,
|
| 3564 |
+
"learning_rate": 0.00012296447534055716,
|
| 3565 |
+
"loss": 2.5489,
|
| 3566 |
+
"step": 507
|
| 3567 |
+
},
|
| 3568 |
+
{
|
| 3569 |
+
"epoch": 0.43211057947900056,
|
| 3570 |
+
"grad_norm": 0.3207886517047882,
|
| 3571 |
+
"learning_rate": 0.00012270215970683977,
|
| 3572 |
+
"loss": 2.6984,
|
| 3573 |
+
"step": 508
|
| 3574 |
+
},
|
| 3575 |
+
{
|
| 3576 |
+
"epoch": 0.4329611908559277,
|
| 3577 |
+
"grad_norm": 0.26681843400001526,
|
| 3578 |
+
"learning_rate": 0.00012243967926843627,
|
| 3579 |
+
"loss": 2.4313,
|
| 3580 |
+
"step": 509
|
| 3581 |
+
},
|
| 3582 |
+
{
|
| 3583 |
+
"epoch": 0.43381180223285487,
|
| 3584 |
+
"grad_norm": 0.29009976983070374,
|
| 3585 |
+
"learning_rate": 0.00012217703593080445,
|
| 3586 |
+
"loss": 2.6128,
|
| 3587 |
+
"step": 510
|
| 3588 |
+
},
|
| 3589 |
+
{
|
| 3590 |
+
"epoch": 0.43466241360978203,
|
| 3591 |
+
"grad_norm": 0.27712225914001465,
|
| 3592 |
+
"learning_rate": 0.00012191423160058462,
|
| 3593 |
+
"loss": 2.4976,
|
| 3594 |
+
"step": 511
|
| 3595 |
+
},
|
| 3596 |
+
{
|
| 3597 |
+
"epoch": 0.4355130249867092,
|
| 3598 |
+
"grad_norm": 0.30184635519981384,
|
| 3599 |
+
"learning_rate": 0.00012165126818558572,
|
| 3600 |
+
"loss": 2.6912,
|
| 3601 |
+
"step": 512
|
| 3602 |
+
},
|
| 3603 |
+
{
|
| 3604 |
+
"epoch": 0.43636363636363634,
|
| 3605 |
+
"grad_norm": 0.28122004866600037,
|
| 3606 |
+
"learning_rate": 0.00012138814759477176,
|
| 3607 |
+
"loss": 2.6935,
|
| 3608 |
+
"step": 513
|
| 3609 |
+
},
|
| 3610 |
+
{
|
| 3611 |
+
"epoch": 0.43721424774056356,
|
| 3612 |
+
"grad_norm": 0.2680952250957489,
|
| 3613 |
+
"learning_rate": 0.00012112487173824753,
|
| 3614 |
+
"loss": 2.5607,
|
| 3615 |
+
"step": 514
|
| 3616 |
+
},
|
| 3617 |
+
{
|
| 3618 |
+
"epoch": 0.4380648591174907,
|
| 3619 |
+
"grad_norm": 0.26283374428749084,
|
| 3620 |
+
"learning_rate": 0.00012086144252724513,
|
| 3621 |
+
"loss": 2.5001,
|
| 3622 |
+
"step": 515
|
| 3623 |
+
},
|
| 3624 |
+
{
|
| 3625 |
+
"epoch": 0.43891547049441787,
|
| 3626 |
+
"grad_norm": 0.26960310339927673,
|
| 3627 |
+
"learning_rate": 0.00012059786187410984,
|
| 3628 |
+
"loss": 2.609,
|
| 3629 |
+
"step": 516
|
| 3630 |
+
},
|
| 3631 |
+
{
|
| 3632 |
+
"epoch": 0.439766081871345,
|
| 3633 |
+
"grad_norm": 0.2755014896392822,
|
| 3634 |
+
"learning_rate": 0.00012033413169228635,
|
| 3635 |
+
"loss": 2.6356,
|
| 3636 |
+
"step": 517
|
| 3637 |
+
},
|
| 3638 |
+
{
|
| 3639 |
+
"epoch": 0.4406166932482722,
|
| 3640 |
+
"grad_norm": 0.2706344723701477,
|
| 3641 |
+
"learning_rate": 0.00012007025389630484,
|
| 3642 |
+
"loss": 2.6909,
|
| 3643 |
+
"step": 518
|
| 3644 |
+
},
|
| 3645 |
+
{
|
| 3646 |
+
"epoch": 0.44146730462519934,
|
| 3647 |
+
"grad_norm": 0.2639751434326172,
|
| 3648 |
+
"learning_rate": 0.00011980623040176704,
|
| 3649 |
+
"loss": 2.542,
|
| 3650 |
+
"step": 519
|
| 3651 |
+
},
|
| 3652 |
+
{
|
| 3653 |
+
"epoch": 0.44231791600212655,
|
| 3654 |
+
"grad_norm": 0.2747778594493866,
|
| 3655 |
+
"learning_rate": 0.00011954206312533245,
|
| 3656 |
+
"loss": 2.4773,
|
| 3657 |
+
"step": 520
|
| 3658 |
+
},
|
| 3659 |
+
{
|
| 3660 |
+
"epoch": 0.4431685273790537,
|
| 3661 |
+
"grad_norm": 0.2990424931049347,
|
| 3662 |
+
"learning_rate": 0.0001192777539847043,
|
| 3663 |
+
"loss": 2.6881,
|
| 3664 |
+
"step": 521
|
| 3665 |
+
},
|
| 3666 |
+
{
|
| 3667 |
+
"epoch": 0.44401913875598087,
|
| 3668 |
+
"grad_norm": 0.27671095728874207,
|
| 3669 |
+
"learning_rate": 0.00011901330489861564,
|
| 3670 |
+
"loss": 2.5495,
|
| 3671 |
+
"step": 522
|
| 3672 |
+
},
|
| 3673 |
+
{
|
| 3674 |
+
"epoch": 0.444869750132908,
|
| 3675 |
+
"grad_norm": 0.3051941394805908,
|
| 3676 |
+
"learning_rate": 0.00011874871778681555,
|
| 3677 |
+
"loss": 2.7591,
|
| 3678 |
+
"step": 523
|
| 3679 |
+
},
|
| 3680 |
+
{
|
| 3681 |
+
"epoch": 0.4457203615098352,
|
| 3682 |
+
"grad_norm": 0.2990604043006897,
|
| 3683 |
+
"learning_rate": 0.00011848399457005495,
|
| 3684 |
+
"loss": 2.5765,
|
| 3685 |
+
"step": 524
|
| 3686 |
+
},
|
| 3687 |
+
{
|
| 3688 |
+
"epoch": 0.44657097288676234,
|
| 3689 |
+
"grad_norm": 0.3488616943359375,
|
| 3690 |
+
"learning_rate": 0.00011821913717007298,
|
| 3691 |
+
"loss": 2.634,
|
| 3692 |
+
"step": 525
|
| 3693 |
+
},
|
| 3694 |
+
{
|
| 3695 |
+
"epoch": 0.44742158426368955,
|
| 3696 |
+
"grad_norm": 0.3207804262638092,
|
| 3697 |
+
"learning_rate": 0.00011795414750958265,
|
| 3698 |
+
"loss": 2.5632,
|
| 3699 |
+
"step": 526
|
| 3700 |
+
},
|
| 3701 |
+
{
|
| 3702 |
+
"epoch": 0.4482721956406167,
|
| 3703 |
+
"grad_norm": 0.3244103491306305,
|
| 3704 |
+
"learning_rate": 0.0001176890275122573,
|
| 3705 |
+
"loss": 2.5784,
|
| 3706 |
+
"step": 527
|
| 3707 |
+
},
|
| 3708 |
+
{
|
| 3709 |
+
"epoch": 0.44912280701754387,
|
| 3710 |
+
"grad_norm": 0.33527323603630066,
|
| 3711 |
+
"learning_rate": 0.00011742377910271639,
|
| 3712 |
+
"loss": 2.7016,
|
| 3713 |
+
"step": 528
|
| 3714 |
+
},
|
| 3715 |
+
{
|
| 3716 |
+
"epoch": 0.449973418394471,
|
| 3717 |
+
"grad_norm": 0.3452233672142029,
|
| 3718 |
+
"learning_rate": 0.00011715840420651152,
|
| 3719 |
+
"loss": 2.5495,
|
| 3720 |
+
"step": 529
|
| 3721 |
+
},
|
| 3722 |
+
{
|
| 3723 |
+
"epoch": 0.4508240297713982,
|
| 3724 |
+
"grad_norm": 0.36391714215278625,
|
| 3725 |
+
"learning_rate": 0.00011689290475011259,
|
| 3726 |
+
"loss": 2.662,
|
| 3727 |
+
"step": 530
|
| 3728 |
+
},
|
| 3729 |
+
{
|
| 3730 |
+
"epoch": 0.45167464114832534,
|
| 3731 |
+
"grad_norm": 0.41914063692092896,
|
| 3732 |
+
"learning_rate": 0.00011662728266089364,
|
| 3733 |
+
"loss": 2.8876,
|
| 3734 |
+
"step": 531
|
| 3735 |
+
},
|
| 3736 |
+
{
|
| 3737 |
+
"epoch": 0.45252525252525255,
|
| 3738 |
+
"grad_norm": 0.4124680459499359,
|
| 3739 |
+
"learning_rate": 0.00011636153986711906,
|
| 3740 |
+
"loss": 2.7271,
|
| 3741 |
+
"step": 532
|
| 3742 |
+
},
|
| 3743 |
+
{
|
| 3744 |
+
"epoch": 0.4533758639021797,
|
| 3745 |
+
"grad_norm": 0.3957962393760681,
|
| 3746 |
+
"learning_rate": 0.00011609567829792944,
|
| 3747 |
+
"loss": 2.5991,
|
| 3748 |
+
"step": 533
|
| 3749 |
+
},
|
| 3750 |
+
{
|
| 3751 |
+
"epoch": 0.45422647527910687,
|
| 3752 |
+
"grad_norm": 0.401443213224411,
|
| 3753 |
+
"learning_rate": 0.00011582969988332757,
|
| 3754 |
+
"loss": 2.7777,
|
| 3755 |
+
"step": 534
|
| 3756 |
+
},
|
| 3757 |
+
{
|
| 3758 |
+
"epoch": 0.455077086656034,
|
| 3759 |
+
"grad_norm": 0.4349088966846466,
|
| 3760 |
+
"learning_rate": 0.00011556360655416457,
|
| 3761 |
+
"loss": 2.524,
|
| 3762 |
+
"step": 535
|
| 3763 |
+
},
|
| 3764 |
+
{
|
| 3765 |
+
"epoch": 0.4559276980329612,
|
| 3766 |
+
"grad_norm": 0.44450250267982483,
|
| 3767 |
+
"learning_rate": 0.00011529740024212565,
|
| 3768 |
+
"loss": 2.8296,
|
| 3769 |
+
"step": 536
|
| 3770 |
+
},
|
| 3771 |
+
{
|
| 3772 |
+
"epoch": 0.45677830940988834,
|
| 3773 |
+
"grad_norm": 0.44510000944137573,
|
| 3774 |
+
"learning_rate": 0.00011503108287971626,
|
| 3775 |
+
"loss": 2.8425,
|
| 3776 |
+
"step": 537
|
| 3777 |
+
},
|
| 3778 |
+
{
|
| 3779 |
+
"epoch": 0.4576289207868155,
|
| 3780 |
+
"grad_norm": 0.4766497313976288,
|
| 3781 |
+
"learning_rate": 0.00011476465640024814,
|
| 3782 |
+
"loss": 2.7471,
|
| 3783 |
+
"step": 538
|
| 3784 |
+
},
|
| 3785 |
+
{
|
| 3786 |
+
"epoch": 0.4584795321637427,
|
| 3787 |
+
"grad_norm": 0.458278089761734,
|
| 3788 |
+
"learning_rate": 0.00011449812273782492,
|
| 3789 |
+
"loss": 2.7274,
|
| 3790 |
+
"step": 539
|
| 3791 |
+
},
|
| 3792 |
+
{
|
| 3793 |
+
"epoch": 0.45933014354066987,
|
| 3794 |
+
"grad_norm": 0.46523377299308777,
|
| 3795 |
+
"learning_rate": 0.00011423148382732853,
|
| 3796 |
+
"loss": 2.4967,
|
| 3797 |
+
"step": 540
|
| 3798 |
+
},
|
| 3799 |
+
{
|
| 3800 |
+
"epoch": 0.460180754917597,
|
| 3801 |
+
"grad_norm": 0.4741576611995697,
|
| 3802 |
+
"learning_rate": 0.00011396474160440478,
|
| 3803 |
+
"loss": 2.6591,
|
| 3804 |
+
"step": 541
|
| 3805 |
+
},
|
| 3806 |
+
{
|
| 3807 |
+
"epoch": 0.4610313662945242,
|
| 3808 |
+
"grad_norm": 0.5007473230361938,
|
| 3809 |
+
"learning_rate": 0.00011369789800544959,
|
| 3810 |
+
"loss": 2.8139,
|
| 3811 |
+
"step": 542
|
| 3812 |
+
},
|
| 3813 |
+
{
|
| 3814 |
+
"epoch": 0.46188197767145134,
|
| 3815 |
+
"grad_norm": 0.45440518856048584,
|
| 3816 |
+
"learning_rate": 0.00011343095496759476,
|
| 3817 |
+
"loss": 2.6233,
|
| 3818 |
+
"step": 543
|
| 3819 |
+
},
|
| 3820 |
+
{
|
| 3821 |
+
"epoch": 0.4627325890483785,
|
| 3822 |
+
"grad_norm": 0.5588196516036987,
|
| 3823 |
+
"learning_rate": 0.00011316391442869394,
|
| 3824 |
+
"loss": 2.6369,
|
| 3825 |
+
"step": 544
|
| 3826 |
+
},
|
| 3827 |
+
{
|
| 3828 |
+
"epoch": 0.4635832004253057,
|
| 3829 |
+
"grad_norm": 0.5833231806755066,
|
| 3830 |
+
"learning_rate": 0.00011289677832730862,
|
| 3831 |
+
"loss": 2.8079,
|
| 3832 |
+
"step": 545
|
| 3833 |
+
},
|
| 3834 |
+
{
|
| 3835 |
+
"epoch": 0.46443381180223287,
|
| 3836 |
+
"grad_norm": 0.6029292345046997,
|
| 3837 |
+
"learning_rate": 0.00011262954860269399,
|
| 3838 |
+
"loss": 2.8007,
|
| 3839 |
+
"step": 546
|
| 3840 |
+
},
|
| 3841 |
+
{
|
| 3842 |
+
"epoch": 0.46528442317916,
|
| 3843 |
+
"grad_norm": 0.5825842022895813,
|
| 3844 |
+
"learning_rate": 0.00011236222719478491,
|
| 3845 |
+
"loss": 2.7767,
|
| 3846 |
+
"step": 547
|
| 3847 |
+
},
|
| 3848 |
+
{
|
| 3849 |
+
"epoch": 0.4661350345560872,
|
| 3850 |
+
"grad_norm": 0.6685728430747986,
|
| 3851 |
+
"learning_rate": 0.00011209481604418181,
|
| 3852 |
+
"loss": 2.6621,
|
| 3853 |
+
"step": 548
|
| 3854 |
+
},
|
| 3855 |
+
{
|
| 3856 |
+
"epoch": 0.46698564593301434,
|
| 3857 |
+
"grad_norm": 0.6266542077064514,
|
| 3858 |
+
"learning_rate": 0.00011182731709213659,
|
| 3859 |
+
"loss": 2.5234,
|
| 3860 |
+
"step": 549
|
| 3861 |
+
},
|
| 3862 |
+
{
|
| 3863 |
+
"epoch": 0.4678362573099415,
|
| 3864 |
+
"grad_norm": 0.7756956815719604,
|
| 3865 |
+
"learning_rate": 0.00011155973228053853,
|
| 3866 |
+
"loss": 2.9975,
|
| 3867 |
+
"step": 550
|
| 3868 |
+
},
|
| 3869 |
+
{
|
| 3870 |
+
"epoch": 0.4686868686868687,
|
| 3871 |
+
"grad_norm": 0.27434810996055603,
|
| 3872 |
+
"learning_rate": 0.00011129206355190025,
|
| 3873 |
+
"loss": 2.3867,
|
| 3874 |
+
"step": 551
|
| 3875 |
+
},
|
| 3876 |
+
{
|
| 3877 |
+
"epoch": 0.46953748006379586,
|
| 3878 |
+
"grad_norm": 0.28944891691207886,
|
| 3879 |
+
"learning_rate": 0.00011102431284934345,
|
| 3880 |
+
"loss": 2.6283,
|
| 3881 |
+
"step": 552
|
| 3882 |
+
},
|
| 3883 |
+
{
|
| 3884 |
+
"epoch": 0.470388091440723,
|
| 3885 |
+
"grad_norm": 0.2992161810398102,
|
| 3886 |
+
"learning_rate": 0.00011075648211658505,
|
| 3887 |
+
"loss": 2.5818,
|
| 3888 |
+
"step": 553
|
| 3889 |
+
},
|
| 3890 |
+
{
|
| 3891 |
+
"epoch": 0.4712387028176502,
|
| 3892 |
+
"grad_norm": 0.2830989956855774,
|
| 3893 |
+
"learning_rate": 0.00011048857329792284,
|
| 3894 |
+
"loss": 2.5738,
|
| 3895 |
+
"step": 554
|
| 3896 |
+
},
|
| 3897 |
+
{
|
| 3898 |
+
"epoch": 0.47208931419457734,
|
| 3899 |
+
"grad_norm": 0.2766299545764923,
|
| 3900 |
+
"learning_rate": 0.00011022058833822158,
|
| 3901 |
+
"loss": 2.5138,
|
| 3902 |
+
"step": 555
|
| 3903 |
+
},
|
| 3904 |
+
{
|
| 3905 |
+
"epoch": 0.4729399255715045,
|
| 3906 |
+
"grad_norm": 0.27325817942619324,
|
| 3907 |
+
"learning_rate": 0.0001099525291828986,
|
| 3908 |
+
"loss": 2.5166,
|
| 3909 |
+
"step": 556
|
| 3910 |
+
},
|
| 3911 |
+
{
|
| 3912 |
+
"epoch": 0.4737905369484317,
|
| 3913 |
+
"grad_norm": 0.27071428298950195,
|
| 3914 |
+
"learning_rate": 0.00010968439777790999,
|
| 3915 |
+
"loss": 2.4935,
|
| 3916 |
+
"step": 557
|
| 3917 |
+
},
|
| 3918 |
+
{
|
| 3919 |
+
"epoch": 0.47464114832535886,
|
| 3920 |
+
"grad_norm": 0.28781965374946594,
|
| 3921 |
+
"learning_rate": 0.00010941619606973632,
|
| 3922 |
+
"loss": 2.5851,
|
| 3923 |
+
"step": 558
|
| 3924 |
+
},
|
| 3925 |
+
{
|
| 3926 |
+
"epoch": 0.475491759702286,
|
| 3927 |
+
"grad_norm": 0.2613832354545593,
|
| 3928 |
+
"learning_rate": 0.00010914792600536843,
|
| 3929 |
+
"loss": 2.5289,
|
| 3930 |
+
"step": 559
|
| 3931 |
+
},
|
| 3932 |
+
{
|
| 3933 |
+
"epoch": 0.4763423710792132,
|
| 3934 |
+
"grad_norm": 0.2806464433670044,
|
| 3935 |
+
"learning_rate": 0.00010887958953229349,
|
| 3936 |
+
"loss": 2.659,
|
| 3937 |
+
"step": 560
|
| 3938 |
+
},
|
| 3939 |
+
{
|
| 3940 |
+
"epoch": 0.47719298245614034,
|
| 3941 |
+
"grad_norm": 0.2767505347728729,
|
| 3942 |
+
"learning_rate": 0.00010861118859848067,
|
| 3943 |
+
"loss": 2.6562,
|
| 3944 |
+
"step": 561
|
| 3945 |
+
},
|
| 3946 |
+
{
|
| 3947 |
+
"epoch": 0.4780435938330675,
|
| 3948 |
+
"grad_norm": 0.26740705966949463,
|
| 3949 |
+
"learning_rate": 0.0001083427251523672,
|
| 3950 |
+
"loss": 2.5689,
|
| 3951 |
+
"step": 562
|
| 3952 |
+
},
|
| 3953 |
+
{
|
| 3954 |
+
"epoch": 0.4788942052099947,
|
| 3955 |
+
"grad_norm": 0.2635597884654999,
|
| 3956 |
+
"learning_rate": 0.000108074201142844,
|
| 3957 |
+
"loss": 2.3973,
|
| 3958 |
+
"step": 563
|
| 3959 |
+
},
|
| 3960 |
+
{
|
| 3961 |
+
"epoch": 0.47974481658692186,
|
| 3962 |
+
"grad_norm": 0.2559509575366974,
|
| 3963 |
+
"learning_rate": 0.00010780561851924167,
|
| 3964 |
+
"loss": 2.4662,
|
| 3965 |
+
"step": 564
|
| 3966 |
+
},
|
| 3967 |
+
{
|
| 3968 |
+
"epoch": 0.480595427963849,
|
| 3969 |
+
"grad_norm": 0.2819572687149048,
|
| 3970 |
+
"learning_rate": 0.0001075369792313164,
|
| 3971 |
+
"loss": 2.7678,
|
| 3972 |
+
"step": 565
|
| 3973 |
+
},
|
| 3974 |
+
{
|
| 3975 |
+
"epoch": 0.4814460393407762,
|
| 3976 |
+
"grad_norm": 0.2618950605392456,
|
| 3977 |
+
"learning_rate": 0.00010726828522923562,
|
| 3978 |
+
"loss": 2.6463,
|
| 3979 |
+
"step": 566
|
| 3980 |
+
},
|
| 3981 |
+
{
|
| 3982 |
+
"epoch": 0.48229665071770333,
|
| 3983 |
+
"grad_norm": 0.2766074240207672,
|
| 3984 |
+
"learning_rate": 0.000106999538463564,
|
| 3985 |
+
"loss": 2.7133,
|
| 3986 |
+
"step": 567
|
| 3987 |
+
},
|
| 3988 |
+
{
|
| 3989 |
+
"epoch": 0.4831472620946305,
|
| 3990 |
+
"grad_norm": 0.280367910861969,
|
| 3991 |
+
"learning_rate": 0.00010673074088524926,
|
| 3992 |
+
"loss": 2.61,
|
| 3993 |
+
"step": 568
|
| 3994 |
+
},
|
| 3995 |
+
{
|
| 3996 |
+
"epoch": 0.4839978734715577,
|
| 3997 |
+
"grad_norm": 0.3028632402420044,
|
| 3998 |
+
"learning_rate": 0.00010646189444560799,
|
| 3999 |
+
"loss": 2.5465,
|
| 4000 |
+
"step": 569
|
| 4001 |
+
},
|
| 4002 |
+
{
|
| 4003 |
+
"epoch": 0.48484848484848486,
|
| 4004 |
+
"grad_norm": 0.2950790822505951,
|
| 4005 |
+
"learning_rate": 0.00010619300109631145,
|
| 4006 |
+
"loss": 2.6517,
|
| 4007 |
+
"step": 570
|
| 4008 |
+
},
|
| 4009 |
+
{
|
| 4010 |
+
"epoch": 0.485699096225412,
|
| 4011 |
+
"grad_norm": 0.2803073227405548,
|
| 4012 |
+
"learning_rate": 0.00010592406278937144,
|
| 4013 |
+
"loss": 2.6062,
|
| 4014 |
+
"step": 571
|
| 4015 |
+
},
|
| 4016 |
+
{
|
| 4017 |
+
"epoch": 0.4865497076023392,
|
| 4018 |
+
"grad_norm": 0.29916471242904663,
|
| 4019 |
+
"learning_rate": 0.00010565508147712617,
|
| 4020 |
+
"loss": 2.5532,
|
| 4021 |
+
"step": 572
|
| 4022 |
+
},
|
| 4023 |
+
{
|
| 4024 |
+
"epoch": 0.48740031897926633,
|
| 4025 |
+
"grad_norm": 0.32185062766075134,
|
| 4026 |
+
"learning_rate": 0.00010538605911222603,
|
| 4027 |
+
"loss": 2.722,
|
| 4028 |
+
"step": 573
|
| 4029 |
+
},
|
| 4030 |
+
{
|
| 4031 |
+
"epoch": 0.4882509303561935,
|
| 4032 |
+
"grad_norm": 0.3155268728733063,
|
| 4033 |
+
"learning_rate": 0.00010511699764761936,
|
| 4034 |
+
"loss": 2.6655,
|
| 4035 |
+
"step": 574
|
| 4036 |
+
},
|
| 4037 |
+
{
|
| 4038 |
+
"epoch": 0.4891015417331207,
|
| 4039 |
+
"grad_norm": 0.3241024315357208,
|
| 4040 |
+
"learning_rate": 0.00010484789903653846,
|
| 4041 |
+
"loss": 2.7093,
|
| 4042 |
+
"step": 575
|
| 4043 |
+
},
|
| 4044 |
+
{
|
| 4045 |
+
"epoch": 0.48995215311004786,
|
| 4046 |
+
"grad_norm": 0.3311263620853424,
|
| 4047 |
+
"learning_rate": 0.00010457876523248518,
|
| 4048 |
+
"loss": 2.6485,
|
| 4049 |
+
"step": 576
|
| 4050 |
+
},
|
| 4051 |
+
{
|
| 4052 |
+
"epoch": 0.490802764486975,
|
| 4053 |
+
"grad_norm": 0.34630128741264343,
|
| 4054 |
+
"learning_rate": 0.00010430959818921694,
|
| 4055 |
+
"loss": 2.8315,
|
| 4056 |
+
"step": 577
|
| 4057 |
+
},
|
| 4058 |
+
{
|
| 4059 |
+
"epoch": 0.4916533758639022,
|
| 4060 |
+
"grad_norm": 0.34841713309288025,
|
| 4061 |
+
"learning_rate": 0.00010404039986073244,
|
| 4062 |
+
"loss": 2.7484,
|
| 4063 |
+
"step": 578
|
| 4064 |
+
},
|
| 4065 |
+
{
|
| 4066 |
+
"epoch": 0.49250398724082933,
|
| 4067 |
+
"grad_norm": 0.35180601477622986,
|
| 4068 |
+
"learning_rate": 0.00010377117220125741,
|
| 4069 |
+
"loss": 2.6745,
|
| 4070 |
+
"step": 579
|
| 4071 |
+
},
|
| 4072 |
+
{
|
| 4073 |
+
"epoch": 0.4933545986177565,
|
| 4074 |
+
"grad_norm": 0.38369500637054443,
|
| 4075 |
+
"learning_rate": 0.00010350191716523059,
|
| 4076 |
+
"loss": 2.6623,
|
| 4077 |
+
"step": 580
|
| 4078 |
+
},
|
| 4079 |
+
{
|
| 4080 |
+
"epoch": 0.4942052099946837,
|
| 4081 |
+
"grad_norm": 0.3735206425189972,
|
| 4082 |
+
"learning_rate": 0.00010323263670728946,
|
| 4083 |
+
"loss": 2.5805,
|
| 4084 |
+
"step": 581
|
| 4085 |
+
},
|
| 4086 |
+
{
|
| 4087 |
+
"epoch": 0.49505582137161086,
|
| 4088 |
+
"grad_norm": 0.3994956314563751,
|
| 4089 |
+
"learning_rate": 0.00010296333278225599,
|
| 4090 |
+
"loss": 2.5622,
|
| 4091 |
+
"step": 582
|
| 4092 |
+
},
|
| 4093 |
+
{
|
| 4094 |
+
"epoch": 0.495906432748538,
|
| 4095 |
+
"grad_norm": 0.39151209592819214,
|
| 4096 |
+
"learning_rate": 0.00010269400734512256,
|
| 4097 |
+
"loss": 2.805,
|
| 4098 |
+
"step": 583
|
| 4099 |
+
},
|
| 4100 |
+
{
|
| 4101 |
+
"epoch": 0.4967570441254652,
|
| 4102 |
+
"grad_norm": 0.40469613671302795,
|
| 4103 |
+
"learning_rate": 0.0001024246623510377,
|
| 4104 |
+
"loss": 2.7588,
|
| 4105 |
+
"step": 584
|
| 4106 |
+
},
|
| 4107 |
+
{
|
| 4108 |
+
"epoch": 0.49760765550239233,
|
| 4109 |
+
"grad_norm": 0.4307393431663513,
|
| 4110 |
+
"learning_rate": 0.0001021552997552919,
|
| 4111 |
+
"loss": 2.8039,
|
| 4112 |
+
"step": 585
|
| 4113 |
+
},
|
| 4114 |
+
{
|
| 4115 |
+
"epoch": 0.4984582668793195,
|
| 4116 |
+
"grad_norm": 0.41100749373435974,
|
| 4117 |
+
"learning_rate": 0.00010188592151330343,
|
| 4118 |
+
"loss": 2.6842,
|
| 4119 |
+
"step": 586
|
| 4120 |
+
},
|
| 4121 |
+
{
|
| 4122 |
+
"epoch": 0.4993088782562467,
|
| 4123 |
+
"grad_norm": 0.45486176013946533,
|
| 4124 |
+
"learning_rate": 0.00010161652958060417,
|
| 4125 |
+
"loss": 2.8073,
|
| 4126 |
+
"step": 587
|
| 4127 |
+
},
|
| 4128 |
+
{
|
| 4129 |
+
"epoch": 0.5001594896331738,
|
| 4130 |
+
"grad_norm": 0.43648669123649597,
|
| 4131 |
+
"learning_rate": 0.00010134712591282538,
|
| 4132 |
+
"loss": 2.638,
|
| 4133 |
+
"step": 588
|
| 4134 |
+
},
|
| 4135 |
+
{
|
| 4136 |
+
"epoch": 0.5001594896331738,
|
| 4137 |
+
"eval_loss": 2.66949725151062,
|
| 4138 |
+
"eval_runtime": 80.5199,
|
| 4139 |
+
"eval_samples_per_second": 12.295,
|
| 4140 |
+
"eval_steps_per_second": 6.148,
|
| 4141 |
+
"step": 588
|
| 4142 |
}
|
| 4143 |
],
|
| 4144 |
"logging_steps": 1,
|
|
|
|
| 4158 |
"attributes": {}
|
| 4159 |
}
|
| 4160 |
},
|
| 4161 |
+
"total_flos": 5.999986420985364e+17,
|
| 4162 |
"train_batch_size": 2,
|
| 4163 |
"trial_name": null,
|
| 4164 |
"trial_params": null
|