Training in progress, step 1200
Browse files- checkpoint-1200/optimizer.pt +3 -0
- checkpoint-1200/pytorch_model.bin +3 -0
- checkpoint-1200/rng_state.pth +3 -0
- checkpoint-1200/scaler.pt +3 -0
- checkpoint-1200/scheduler.pt +3 -0
- checkpoint-1200/trainer_state.json +1504 -0
- checkpoint-1200/training_args.bin +3 -0
- pytorch_model.bin +1 -1
- runs/Jun08_19-57-59_5b791b640f29/events.out.tfevents.1686254295.5b791b640f29.4078.0 +2 -2
checkpoint-1200/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8873482b7cfcbb338e2ec0df73bc89bc6b5026cceeda16a99cbc202bb93de888
|
| 3 |
+
size 25279045
|
checkpoint-1200/pytorch_model.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:906c9686996da6e301977bd38c2c83361146915ac6bd33d6e07e32d7953b4bbe
|
| 3 |
+
size 12653389
|
checkpoint-1200/rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b9ff675cce0eebd20f5cbc2ddbf423308cce155031b1547934916f977ba1087c
|
| 3 |
+
size 14575
|
checkpoint-1200/scaler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:741cefeca9ef427f92406d2d10b81996655e2a9d50eb7aaa9614e6fdd1c9f529
|
| 3 |
+
size 557
|
checkpoint-1200/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e06b5c38d3b44d051c04acffdd6ab7ad478cd27a815df3d67ead9b302b1f22be
|
| 3 |
+
size 627
|
checkpoint-1200/trainer_state.json
ADDED
|
@@ -0,0 +1,1504 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 1.318132996559143,
|
| 3 |
+
"best_model_checkpoint": "outputs/checkpoint-1200",
|
| 4 |
+
"epoch": 0.4106776180698152,
|
| 5 |
+
"global_step": 1200,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 1e-05,
|
| 13 |
+
"loss": 1.8606,
|
| 14 |
+
"step": 5
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.0,
|
| 18 |
+
"learning_rate": 2e-05,
|
| 19 |
+
"loss": 1.8806,
|
| 20 |
+
"step": 10
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 3e-05,
|
| 25 |
+
"loss": 1.8434,
|
| 26 |
+
"step": 15
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.01,
|
| 30 |
+
"learning_rate": 4e-05,
|
| 31 |
+
"loss": 1.7431,
|
| 32 |
+
"step": 20
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.01,
|
| 36 |
+
"learning_rate": 5e-05,
|
| 37 |
+
"loss": 1.7721,
|
| 38 |
+
"step": 25
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.01,
|
| 42 |
+
"learning_rate": 6e-05,
|
| 43 |
+
"loss": 1.6826,
|
| 44 |
+
"step": 30
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.01,
|
| 48 |
+
"learning_rate": 7e-05,
|
| 49 |
+
"loss": 1.5892,
|
| 50 |
+
"step": 35
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.01,
|
| 54 |
+
"learning_rate": 8e-05,
|
| 55 |
+
"loss": 1.6333,
|
| 56 |
+
"step": 40
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.02,
|
| 60 |
+
"learning_rate": 9e-05,
|
| 61 |
+
"loss": 1.5955,
|
| 62 |
+
"step": 45
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.02,
|
| 66 |
+
"learning_rate": 0.0001,
|
| 67 |
+
"loss": 1.5504,
|
| 68 |
+
"step": 50
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.02,
|
| 72 |
+
"learning_rate": 0.00011000000000000002,
|
| 73 |
+
"loss": 1.5621,
|
| 74 |
+
"step": 55
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.02,
|
| 78 |
+
"learning_rate": 0.00012,
|
| 79 |
+
"loss": 1.5142,
|
| 80 |
+
"step": 60
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.02,
|
| 84 |
+
"learning_rate": 0.00013000000000000002,
|
| 85 |
+
"loss": 1.4386,
|
| 86 |
+
"step": 65
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.02,
|
| 90 |
+
"learning_rate": 0.00014,
|
| 91 |
+
"loss": 1.5017,
|
| 92 |
+
"step": 70
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.03,
|
| 96 |
+
"learning_rate": 0.00015000000000000001,
|
| 97 |
+
"loss": 1.4666,
|
| 98 |
+
"step": 75
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.03,
|
| 102 |
+
"learning_rate": 0.00016,
|
| 103 |
+
"loss": 1.4295,
|
| 104 |
+
"step": 80
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.03,
|
| 108 |
+
"learning_rate": 0.00017,
|
| 109 |
+
"loss": 1.4402,
|
| 110 |
+
"step": 85
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.03,
|
| 114 |
+
"learning_rate": 0.00018,
|
| 115 |
+
"loss": 1.4795,
|
| 116 |
+
"step": 90
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.03,
|
| 120 |
+
"learning_rate": 0.00019,
|
| 121 |
+
"loss": 1.4138,
|
| 122 |
+
"step": 95
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.03,
|
| 126 |
+
"learning_rate": 0.0002,
|
| 127 |
+
"loss": 1.4493,
|
| 128 |
+
"step": 100
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.04,
|
| 132 |
+
"learning_rate": 0.0001998259052924791,
|
| 133 |
+
"loss": 1.4918,
|
| 134 |
+
"step": 105
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.04,
|
| 138 |
+
"learning_rate": 0.00019965181058495822,
|
| 139 |
+
"loss": 1.4659,
|
| 140 |
+
"step": 110
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.04,
|
| 144 |
+
"learning_rate": 0.00019947771587743734,
|
| 145 |
+
"loss": 1.4573,
|
| 146 |
+
"step": 115
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.04,
|
| 150 |
+
"learning_rate": 0.00019930362116991646,
|
| 151 |
+
"loss": 1.4274,
|
| 152 |
+
"step": 120
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.04,
|
| 156 |
+
"learning_rate": 0.00019912952646239555,
|
| 157 |
+
"loss": 1.4064,
|
| 158 |
+
"step": 125
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.04,
|
| 162 |
+
"learning_rate": 0.00019895543175487465,
|
| 163 |
+
"loss": 1.4048,
|
| 164 |
+
"step": 130
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.05,
|
| 168 |
+
"learning_rate": 0.00019878133704735376,
|
| 169 |
+
"loss": 1.4493,
|
| 170 |
+
"step": 135
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.05,
|
| 174 |
+
"learning_rate": 0.00019860724233983288,
|
| 175 |
+
"loss": 1.3317,
|
| 176 |
+
"step": 140
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.05,
|
| 180 |
+
"learning_rate": 0.00019843314763231198,
|
| 181 |
+
"loss": 1.4041,
|
| 182 |
+
"step": 145
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.05,
|
| 186 |
+
"learning_rate": 0.0001982590529247911,
|
| 187 |
+
"loss": 1.4311,
|
| 188 |
+
"step": 150
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.05,
|
| 192 |
+
"learning_rate": 0.0001980849582172702,
|
| 193 |
+
"loss": 1.3626,
|
| 194 |
+
"step": 155
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.05,
|
| 198 |
+
"learning_rate": 0.00019791086350974934,
|
| 199 |
+
"loss": 1.4296,
|
| 200 |
+
"step": 160
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.06,
|
| 204 |
+
"learning_rate": 0.00019773676880222843,
|
| 205 |
+
"loss": 1.3888,
|
| 206 |
+
"step": 165
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.06,
|
| 210 |
+
"learning_rate": 0.00019756267409470752,
|
| 211 |
+
"loss": 1.3389,
|
| 212 |
+
"step": 170
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.06,
|
| 216 |
+
"learning_rate": 0.00019738857938718664,
|
| 217 |
+
"loss": 1.4544,
|
| 218 |
+
"step": 175
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.06,
|
| 222 |
+
"learning_rate": 0.00019721448467966573,
|
| 223 |
+
"loss": 1.4382,
|
| 224 |
+
"step": 180
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.06,
|
| 228 |
+
"learning_rate": 0.00019704038997214485,
|
| 229 |
+
"loss": 1.4953,
|
| 230 |
+
"step": 185
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.07,
|
| 234 |
+
"learning_rate": 0.00019686629526462397,
|
| 235 |
+
"loss": 1.4026,
|
| 236 |
+
"step": 190
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.07,
|
| 240 |
+
"learning_rate": 0.00019669220055710306,
|
| 241 |
+
"loss": 1.3757,
|
| 242 |
+
"step": 195
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.07,
|
| 246 |
+
"learning_rate": 0.00019651810584958218,
|
| 247 |
+
"loss": 1.3247,
|
| 248 |
+
"step": 200
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.07,
|
| 252 |
+
"eval_loss": 1.3710952997207642,
|
| 253 |
+
"eval_runtime": 738.1164,
|
| 254 |
+
"eval_samples_per_second": 7.038,
|
| 255 |
+
"eval_steps_per_second": 0.881,
|
| 256 |
+
"step": 200
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.07,
|
| 260 |
+
"learning_rate": 0.0001963440111420613,
|
| 261 |
+
"loss": 1.4409,
|
| 262 |
+
"step": 205
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.07,
|
| 266 |
+
"learning_rate": 0.0001961699164345404,
|
| 267 |
+
"loss": 1.4023,
|
| 268 |
+
"step": 210
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.07,
|
| 272 |
+
"learning_rate": 0.00019599582172701951,
|
| 273 |
+
"loss": 1.3736,
|
| 274 |
+
"step": 215
|
| 275 |
+
},
|
| 276 |
+
{
|
| 277 |
+
"epoch": 0.08,
|
| 278 |
+
"learning_rate": 0.0001958217270194986,
|
| 279 |
+
"loss": 1.4321,
|
| 280 |
+
"step": 220
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.08,
|
| 284 |
+
"learning_rate": 0.00019564763231197773,
|
| 285 |
+
"loss": 1.4344,
|
| 286 |
+
"step": 225
|
| 287 |
+
},
|
| 288 |
+
{
|
| 289 |
+
"epoch": 0.08,
|
| 290 |
+
"learning_rate": 0.00019547353760445685,
|
| 291 |
+
"loss": 1.3539,
|
| 292 |
+
"step": 230
|
| 293 |
+
},
|
| 294 |
+
{
|
| 295 |
+
"epoch": 0.08,
|
| 296 |
+
"learning_rate": 0.00019529944289693594,
|
| 297 |
+
"loss": 1.438,
|
| 298 |
+
"step": 235
|
| 299 |
+
},
|
| 300 |
+
{
|
| 301 |
+
"epoch": 0.08,
|
| 302 |
+
"learning_rate": 0.00019512534818941506,
|
| 303 |
+
"loss": 1.457,
|
| 304 |
+
"step": 240
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.08,
|
| 308 |
+
"learning_rate": 0.00019495125348189415,
|
| 309 |
+
"loss": 1.429,
|
| 310 |
+
"step": 245
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.09,
|
| 314 |
+
"learning_rate": 0.00019477715877437327,
|
| 315 |
+
"loss": 1.4465,
|
| 316 |
+
"step": 250
|
| 317 |
+
},
|
| 318 |
+
{
|
| 319 |
+
"epoch": 0.09,
|
| 320 |
+
"learning_rate": 0.0001946030640668524,
|
| 321 |
+
"loss": 1.4202,
|
| 322 |
+
"step": 255
|
| 323 |
+
},
|
| 324 |
+
{
|
| 325 |
+
"epoch": 0.09,
|
| 326 |
+
"learning_rate": 0.00019442896935933148,
|
| 327 |
+
"loss": 1.343,
|
| 328 |
+
"step": 260
|
| 329 |
+
},
|
| 330 |
+
{
|
| 331 |
+
"epoch": 0.09,
|
| 332 |
+
"learning_rate": 0.0001942548746518106,
|
| 333 |
+
"loss": 1.4127,
|
| 334 |
+
"step": 265
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 0.09,
|
| 338 |
+
"learning_rate": 0.0001940807799442897,
|
| 339 |
+
"loss": 1.3171,
|
| 340 |
+
"step": 270
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.09,
|
| 344 |
+
"learning_rate": 0.0001939066852367688,
|
| 345 |
+
"loss": 1.4012,
|
| 346 |
+
"step": 275
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.1,
|
| 350 |
+
"learning_rate": 0.00019373259052924793,
|
| 351 |
+
"loss": 1.2933,
|
| 352 |
+
"step": 280
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.1,
|
| 356 |
+
"learning_rate": 0.00019355849582172702,
|
| 357 |
+
"loss": 1.326,
|
| 358 |
+
"step": 285
|
| 359 |
+
},
|
| 360 |
+
{
|
| 361 |
+
"epoch": 0.1,
|
| 362 |
+
"learning_rate": 0.00019338440111420614,
|
| 363 |
+
"loss": 1.3951,
|
| 364 |
+
"step": 290
|
| 365 |
+
},
|
| 366 |
+
{
|
| 367 |
+
"epoch": 0.1,
|
| 368 |
+
"learning_rate": 0.00019321030640668524,
|
| 369 |
+
"loss": 1.3149,
|
| 370 |
+
"step": 295
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.1,
|
| 374 |
+
"learning_rate": 0.00019303621169916436,
|
| 375 |
+
"loss": 1.4046,
|
| 376 |
+
"step": 300
|
| 377 |
+
},
|
| 378 |
+
{
|
| 379 |
+
"epoch": 0.1,
|
| 380 |
+
"learning_rate": 0.00019286211699164348,
|
| 381 |
+
"loss": 1.343,
|
| 382 |
+
"step": 305
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.11,
|
| 386 |
+
"learning_rate": 0.00019268802228412257,
|
| 387 |
+
"loss": 1.3676,
|
| 388 |
+
"step": 310
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.11,
|
| 392 |
+
"learning_rate": 0.00019251392757660166,
|
| 393 |
+
"loss": 1.3593,
|
| 394 |
+
"step": 315
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.11,
|
| 398 |
+
"learning_rate": 0.0001923398328690808,
|
| 399 |
+
"loss": 1.3869,
|
| 400 |
+
"step": 320
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 0.11,
|
| 404 |
+
"learning_rate": 0.0001921657381615599,
|
| 405 |
+
"loss": 1.3841,
|
| 406 |
+
"step": 325
|
| 407 |
+
},
|
| 408 |
+
{
|
| 409 |
+
"epoch": 0.11,
|
| 410 |
+
"learning_rate": 0.00019199164345403902,
|
| 411 |
+
"loss": 1.3806,
|
| 412 |
+
"step": 330
|
| 413 |
+
},
|
| 414 |
+
{
|
| 415 |
+
"epoch": 0.11,
|
| 416 |
+
"learning_rate": 0.0001918175487465181,
|
| 417 |
+
"loss": 1.4527,
|
| 418 |
+
"step": 335
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 0.12,
|
| 422 |
+
"learning_rate": 0.0001916434540389972,
|
| 423 |
+
"loss": 1.4161,
|
| 424 |
+
"step": 340
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 0.12,
|
| 428 |
+
"learning_rate": 0.00019146935933147635,
|
| 429 |
+
"loss": 1.3019,
|
| 430 |
+
"step": 345
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.12,
|
| 434 |
+
"learning_rate": 0.00019129526462395544,
|
| 435 |
+
"loss": 1.4082,
|
| 436 |
+
"step": 350
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.12,
|
| 440 |
+
"learning_rate": 0.00019112116991643454,
|
| 441 |
+
"loss": 1.4261,
|
| 442 |
+
"step": 355
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 0.12,
|
| 446 |
+
"learning_rate": 0.00019094707520891365,
|
| 447 |
+
"loss": 1.4152,
|
| 448 |
+
"step": 360
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.12,
|
| 452 |
+
"learning_rate": 0.00019077298050139277,
|
| 453 |
+
"loss": 1.4289,
|
| 454 |
+
"step": 365
|
| 455 |
+
},
|
| 456 |
+
{
|
| 457 |
+
"epoch": 0.13,
|
| 458 |
+
"learning_rate": 0.0001905988857938719,
|
| 459 |
+
"loss": 1.3951,
|
| 460 |
+
"step": 370
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 0.13,
|
| 464 |
+
"learning_rate": 0.00019042479108635099,
|
| 465 |
+
"loss": 1.3578,
|
| 466 |
+
"step": 375
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 0.13,
|
| 470 |
+
"learning_rate": 0.00019025069637883008,
|
| 471 |
+
"loss": 1.3342,
|
| 472 |
+
"step": 380
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.13,
|
| 476 |
+
"learning_rate": 0.0001900766016713092,
|
| 477 |
+
"loss": 1.3959,
|
| 478 |
+
"step": 385
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.13,
|
| 482 |
+
"learning_rate": 0.00018990250696378832,
|
| 483 |
+
"loss": 1.4336,
|
| 484 |
+
"step": 390
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 0.14,
|
| 488 |
+
"learning_rate": 0.00018972841225626744,
|
| 489 |
+
"loss": 1.464,
|
| 490 |
+
"step": 395
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 0.14,
|
| 494 |
+
"learning_rate": 0.00018955431754874653,
|
| 495 |
+
"loss": 1.4053,
|
| 496 |
+
"step": 400
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 0.14,
|
| 500 |
+
"eval_loss": 1.3470327854156494,
|
| 501 |
+
"eval_runtime": 738.9389,
|
| 502 |
+
"eval_samples_per_second": 7.03,
|
| 503 |
+
"eval_steps_per_second": 0.88,
|
| 504 |
+
"step": 400
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 0.14,
|
| 508 |
+
"learning_rate": 0.00018938022284122562,
|
| 509 |
+
"loss": 1.3582,
|
| 510 |
+
"step": 405
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 0.14,
|
| 514 |
+
"learning_rate": 0.00018920612813370474,
|
| 515 |
+
"loss": 1.375,
|
| 516 |
+
"step": 410
|
| 517 |
+
},
|
| 518 |
+
{
|
| 519 |
+
"epoch": 0.14,
|
| 520 |
+
"learning_rate": 0.00018903203342618386,
|
| 521 |
+
"loss": 1.4229,
|
| 522 |
+
"step": 415
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 0.14,
|
| 526 |
+
"learning_rate": 0.00018885793871866295,
|
| 527 |
+
"loss": 1.3843,
|
| 528 |
+
"step": 420
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.15,
|
| 532 |
+
"learning_rate": 0.00018868384401114207,
|
| 533 |
+
"loss": 1.39,
|
| 534 |
+
"step": 425
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.15,
|
| 538 |
+
"learning_rate": 0.00018850974930362116,
|
| 539 |
+
"loss": 1.3863,
|
| 540 |
+
"step": 430
|
| 541 |
+
},
|
| 542 |
+
{
|
| 543 |
+
"epoch": 0.15,
|
| 544 |
+
"learning_rate": 0.00018833565459610028,
|
| 545 |
+
"loss": 1.3597,
|
| 546 |
+
"step": 435
|
| 547 |
+
},
|
| 548 |
+
{
|
| 549 |
+
"epoch": 0.15,
|
| 550 |
+
"learning_rate": 0.0001881615598885794,
|
| 551 |
+
"loss": 1.4114,
|
| 552 |
+
"step": 440
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 0.15,
|
| 556 |
+
"learning_rate": 0.0001879874651810585,
|
| 557 |
+
"loss": 1.2929,
|
| 558 |
+
"step": 445
|
| 559 |
+
},
|
| 560 |
+
{
|
| 561 |
+
"epoch": 0.15,
|
| 562 |
+
"learning_rate": 0.00018781337047353762,
|
| 563 |
+
"loss": 1.3087,
|
| 564 |
+
"step": 450
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 0.16,
|
| 568 |
+
"learning_rate": 0.0001876392757660167,
|
| 569 |
+
"loss": 1.3748,
|
| 570 |
+
"step": 455
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.16,
|
| 574 |
+
"learning_rate": 0.00018746518105849583,
|
| 575 |
+
"loss": 1.3716,
|
| 576 |
+
"step": 460
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.16,
|
| 580 |
+
"learning_rate": 0.00018729108635097495,
|
| 581 |
+
"loss": 1.404,
|
| 582 |
+
"step": 465
|
| 583 |
+
},
|
| 584 |
+
{
|
| 585 |
+
"epoch": 0.16,
|
| 586 |
+
"learning_rate": 0.00018711699164345404,
|
| 587 |
+
"loss": 1.2931,
|
| 588 |
+
"step": 470
|
| 589 |
+
},
|
| 590 |
+
{
|
| 591 |
+
"epoch": 0.16,
|
| 592 |
+
"learning_rate": 0.00018694289693593316,
|
| 593 |
+
"loss": 1.3345,
|
| 594 |
+
"step": 475
|
| 595 |
+
},
|
| 596 |
+
{
|
| 597 |
+
"epoch": 0.16,
|
| 598 |
+
"learning_rate": 0.00018676880222841225,
|
| 599 |
+
"loss": 1.3836,
|
| 600 |
+
"step": 480
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.17,
|
| 604 |
+
"learning_rate": 0.00018659470752089137,
|
| 605 |
+
"loss": 1.3923,
|
| 606 |
+
"step": 485
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.17,
|
| 610 |
+
"learning_rate": 0.0001864206128133705,
|
| 611 |
+
"loss": 1.3166,
|
| 612 |
+
"step": 490
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.17,
|
| 616 |
+
"learning_rate": 0.00018624651810584958,
|
| 617 |
+
"loss": 1.3697,
|
| 618 |
+
"step": 495
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.17,
|
| 622 |
+
"learning_rate": 0.0001860724233983287,
|
| 623 |
+
"loss": 1.437,
|
| 624 |
+
"step": 500
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.17,
|
| 628 |
+
"learning_rate": 0.00018589832869080782,
|
| 629 |
+
"loss": 1.3579,
|
| 630 |
+
"step": 505
|
| 631 |
+
},
|
| 632 |
+
{
|
| 633 |
+
"epoch": 0.17,
|
| 634 |
+
"learning_rate": 0.00018572423398328691,
|
| 635 |
+
"loss": 1.3769,
|
| 636 |
+
"step": 510
|
| 637 |
+
},
|
| 638 |
+
{
|
| 639 |
+
"epoch": 0.18,
|
| 640 |
+
"learning_rate": 0.00018555013927576603,
|
| 641 |
+
"loss": 1.3989,
|
| 642 |
+
"step": 515
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 0.18,
|
| 646 |
+
"learning_rate": 0.00018537604456824513,
|
| 647 |
+
"loss": 1.331,
|
| 648 |
+
"step": 520
|
| 649 |
+
},
|
| 650 |
+
{
|
| 651 |
+
"epoch": 0.18,
|
| 652 |
+
"learning_rate": 0.00018520194986072425,
|
| 653 |
+
"loss": 1.3678,
|
| 654 |
+
"step": 525
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.18,
|
| 658 |
+
"learning_rate": 0.00018502785515320337,
|
| 659 |
+
"loss": 1.3475,
|
| 660 |
+
"step": 530
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.18,
|
| 664 |
+
"learning_rate": 0.00018485376044568246,
|
| 665 |
+
"loss": 1.4465,
|
| 666 |
+
"step": 535
|
| 667 |
+
},
|
| 668 |
+
{
|
| 669 |
+
"epoch": 0.18,
|
| 670 |
+
"learning_rate": 0.00018467966573816158,
|
| 671 |
+
"loss": 1.3304,
|
| 672 |
+
"step": 540
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 0.19,
|
| 676 |
+
"learning_rate": 0.00018450557103064067,
|
| 677 |
+
"loss": 1.4006,
|
| 678 |
+
"step": 545
|
| 679 |
+
},
|
| 680 |
+
{
|
| 681 |
+
"epoch": 0.19,
|
| 682 |
+
"learning_rate": 0.0001843314763231198,
|
| 683 |
+
"loss": 1.3645,
|
| 684 |
+
"step": 550
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 0.19,
|
| 688 |
+
"learning_rate": 0.0001841573816155989,
|
| 689 |
+
"loss": 1.3217,
|
| 690 |
+
"step": 555
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"epoch": 0.19,
|
| 694 |
+
"learning_rate": 0.000183983286908078,
|
| 695 |
+
"loss": 1.3421,
|
| 696 |
+
"step": 560
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.19,
|
| 700 |
+
"learning_rate": 0.0001838091922005571,
|
| 701 |
+
"loss": 1.3283,
|
| 702 |
+
"step": 565
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.2,
|
| 706 |
+
"learning_rate": 0.0001836350974930362,
|
| 707 |
+
"loss": 1.4101,
|
| 708 |
+
"step": 570
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 0.2,
|
| 712 |
+
"learning_rate": 0.00018346100278551533,
|
| 713 |
+
"loss": 1.3333,
|
| 714 |
+
"step": 575
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 0.2,
|
| 718 |
+
"learning_rate": 0.00018328690807799445,
|
| 719 |
+
"loss": 1.4574,
|
| 720 |
+
"step": 580
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.2,
|
| 724 |
+
"learning_rate": 0.00018311281337047354,
|
| 725 |
+
"loss": 1.3755,
|
| 726 |
+
"step": 585
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 0.2,
|
| 730 |
+
"learning_rate": 0.00018293871866295264,
|
| 731 |
+
"loss": 1.3537,
|
| 732 |
+
"step": 590
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.2,
|
| 736 |
+
"learning_rate": 0.00018276462395543176,
|
| 737 |
+
"loss": 1.3734,
|
| 738 |
+
"step": 595
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 0.21,
|
| 742 |
+
"learning_rate": 0.00018259052924791088,
|
| 743 |
+
"loss": 1.3431,
|
| 744 |
+
"step": 600
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.21,
|
| 748 |
+
"eval_loss": 1.3354507684707642,
|
| 749 |
+
"eval_runtime": 738.7441,
|
| 750 |
+
"eval_samples_per_second": 7.032,
|
| 751 |
+
"eval_steps_per_second": 0.88,
|
| 752 |
+
"step": 600
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.21,
|
| 756 |
+
"learning_rate": 0.00018241643454039,
|
| 757 |
+
"loss": 1.3685,
|
| 758 |
+
"step": 605
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.21,
|
| 762 |
+
"learning_rate": 0.0001822423398328691,
|
| 763 |
+
"loss": 1.377,
|
| 764 |
+
"step": 610
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.21,
|
| 768 |
+
"learning_rate": 0.00018206824512534818,
|
| 769 |
+
"loss": 1.4289,
|
| 770 |
+
"step": 615
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.21,
|
| 774 |
+
"learning_rate": 0.00018189415041782733,
|
| 775 |
+
"loss": 1.398,
|
| 776 |
+
"step": 620
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.21,
|
| 780 |
+
"learning_rate": 0.00018172005571030642,
|
| 781 |
+
"loss": 1.3915,
|
| 782 |
+
"step": 625
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.22,
|
| 786 |
+
"learning_rate": 0.0001815459610027855,
|
| 787 |
+
"loss": 1.4608,
|
| 788 |
+
"step": 630
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.22,
|
| 792 |
+
"learning_rate": 0.00018137186629526463,
|
| 793 |
+
"loss": 1.4408,
|
| 794 |
+
"step": 635
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.22,
|
| 798 |
+
"learning_rate": 0.00018119777158774372,
|
| 799 |
+
"loss": 1.3443,
|
| 800 |
+
"step": 640
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.22,
|
| 804 |
+
"learning_rate": 0.00018102367688022287,
|
| 805 |
+
"loss": 1.4033,
|
| 806 |
+
"step": 645
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.22,
|
| 810 |
+
"learning_rate": 0.00018084958217270196,
|
| 811 |
+
"loss": 1.3086,
|
| 812 |
+
"step": 650
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.22,
|
| 816 |
+
"learning_rate": 0.00018067548746518105,
|
| 817 |
+
"loss": 1.3581,
|
| 818 |
+
"step": 655
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.23,
|
| 822 |
+
"learning_rate": 0.00018050139275766017,
|
| 823 |
+
"loss": 1.3457,
|
| 824 |
+
"step": 660
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.23,
|
| 828 |
+
"learning_rate": 0.0001803272980501393,
|
| 829 |
+
"loss": 1.4488,
|
| 830 |
+
"step": 665
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.23,
|
| 834 |
+
"learning_rate": 0.0001801532033426184,
|
| 835 |
+
"loss": 1.3765,
|
| 836 |
+
"step": 670
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.23,
|
| 840 |
+
"learning_rate": 0.0001799791086350975,
|
| 841 |
+
"loss": 1.4061,
|
| 842 |
+
"step": 675
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.23,
|
| 846 |
+
"learning_rate": 0.0001798050139275766,
|
| 847 |
+
"loss": 1.3556,
|
| 848 |
+
"step": 680
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.23,
|
| 852 |
+
"learning_rate": 0.00017963091922005572,
|
| 853 |
+
"loss": 1.3359,
|
| 854 |
+
"step": 685
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.24,
|
| 858 |
+
"learning_rate": 0.00017945682451253484,
|
| 859 |
+
"loss": 1.2965,
|
| 860 |
+
"step": 690
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.24,
|
| 864 |
+
"learning_rate": 0.00017928272980501393,
|
| 865 |
+
"loss": 1.3211,
|
| 866 |
+
"step": 695
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.24,
|
| 870 |
+
"learning_rate": 0.00017910863509749305,
|
| 871 |
+
"loss": 1.4063,
|
| 872 |
+
"step": 700
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.24,
|
| 876 |
+
"learning_rate": 0.00017893454038997214,
|
| 877 |
+
"loss": 1.3837,
|
| 878 |
+
"step": 705
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.24,
|
| 882 |
+
"learning_rate": 0.00017876044568245126,
|
| 883 |
+
"loss": 1.397,
|
| 884 |
+
"step": 710
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.24,
|
| 888 |
+
"learning_rate": 0.00017858635097493038,
|
| 889 |
+
"loss": 1.3671,
|
| 890 |
+
"step": 715
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.25,
|
| 894 |
+
"learning_rate": 0.00017841225626740947,
|
| 895 |
+
"loss": 1.3666,
|
| 896 |
+
"step": 720
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.25,
|
| 900 |
+
"learning_rate": 0.0001782381615598886,
|
| 901 |
+
"loss": 1.3875,
|
| 902 |
+
"step": 725
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.25,
|
| 906 |
+
"learning_rate": 0.00017806406685236768,
|
| 907 |
+
"loss": 1.3613,
|
| 908 |
+
"step": 730
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.25,
|
| 912 |
+
"learning_rate": 0.0001778899721448468,
|
| 913 |
+
"loss": 1.3611,
|
| 914 |
+
"step": 735
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.25,
|
| 918 |
+
"learning_rate": 0.00017771587743732592,
|
| 919 |
+
"loss": 1.357,
|
| 920 |
+
"step": 740
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.25,
|
| 924 |
+
"learning_rate": 0.00017754178272980502,
|
| 925 |
+
"loss": 1.3561,
|
| 926 |
+
"step": 745
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.26,
|
| 930 |
+
"learning_rate": 0.00017736768802228414,
|
| 931 |
+
"loss": 1.4103,
|
| 932 |
+
"step": 750
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.26,
|
| 936 |
+
"learning_rate": 0.00017719359331476323,
|
| 937 |
+
"loss": 1.4355,
|
| 938 |
+
"step": 755
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.26,
|
| 942 |
+
"learning_rate": 0.00017701949860724235,
|
| 943 |
+
"loss": 1.3901,
|
| 944 |
+
"step": 760
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.26,
|
| 948 |
+
"learning_rate": 0.00017684540389972147,
|
| 949 |
+
"loss": 1.4161,
|
| 950 |
+
"step": 765
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.26,
|
| 954 |
+
"learning_rate": 0.00017667130919220056,
|
| 955 |
+
"loss": 1.4353,
|
| 956 |
+
"step": 770
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.27,
|
| 960 |
+
"learning_rate": 0.00017649721448467968,
|
| 961 |
+
"loss": 1.3647,
|
| 962 |
+
"step": 775
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.27,
|
| 966 |
+
"learning_rate": 0.0001763231197771588,
|
| 967 |
+
"loss": 1.3503,
|
| 968 |
+
"step": 780
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.27,
|
| 972 |
+
"learning_rate": 0.0001761490250696379,
|
| 973 |
+
"loss": 1.3034,
|
| 974 |
+
"step": 785
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.27,
|
| 978 |
+
"learning_rate": 0.000175974930362117,
|
| 979 |
+
"loss": 1.3908,
|
| 980 |
+
"step": 790
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.27,
|
| 984 |
+
"learning_rate": 0.0001758008356545961,
|
| 985 |
+
"loss": 1.3361,
|
| 986 |
+
"step": 795
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.27,
|
| 990 |
+
"learning_rate": 0.0001756267409470752,
|
| 991 |
+
"loss": 1.4012,
|
| 992 |
+
"step": 800
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.27,
|
| 996 |
+
"eval_loss": 1.3277837038040161,
|
| 997 |
+
"eval_runtime": 737.1439,
|
| 998 |
+
"eval_samples_per_second": 7.047,
|
| 999 |
+
"eval_steps_per_second": 0.882,
|
| 1000 |
+
"step": 800
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 0.28,
|
| 1004 |
+
"learning_rate": 0.00017545264623955434,
|
| 1005 |
+
"loss": 1.3631,
|
| 1006 |
+
"step": 805
|
| 1007 |
+
},
|
| 1008 |
+
{
|
| 1009 |
+
"epoch": 0.28,
|
| 1010 |
+
"learning_rate": 0.00017527855153203343,
|
| 1011 |
+
"loss": 1.3789,
|
| 1012 |
+
"step": 810
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 0.28,
|
| 1016 |
+
"learning_rate": 0.00017510445682451255,
|
| 1017 |
+
"loss": 1.3418,
|
| 1018 |
+
"step": 815
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 0.28,
|
| 1022 |
+
"learning_rate": 0.00017493036211699165,
|
| 1023 |
+
"loss": 1.3556,
|
| 1024 |
+
"step": 820
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.28,
|
| 1028 |
+
"learning_rate": 0.00017475626740947077,
|
| 1029 |
+
"loss": 1.339,
|
| 1030 |
+
"step": 825
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"epoch": 0.28,
|
| 1034 |
+
"learning_rate": 0.00017458217270194989,
|
| 1035 |
+
"loss": 1.3419,
|
| 1036 |
+
"step": 830
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 0.29,
|
| 1040 |
+
"learning_rate": 0.00017440807799442898,
|
| 1041 |
+
"loss": 1.2919,
|
| 1042 |
+
"step": 835
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 0.29,
|
| 1046 |
+
"learning_rate": 0.00017423398328690807,
|
| 1047 |
+
"loss": 1.3757,
|
| 1048 |
+
"step": 840
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 0.29,
|
| 1052 |
+
"learning_rate": 0.0001740598885793872,
|
| 1053 |
+
"loss": 1.3165,
|
| 1054 |
+
"step": 845
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 0.29,
|
| 1058 |
+
"learning_rate": 0.0001738857938718663,
|
| 1059 |
+
"loss": 1.325,
|
| 1060 |
+
"step": 850
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 0.29,
|
| 1064 |
+
"learning_rate": 0.00017371169916434543,
|
| 1065 |
+
"loss": 1.3846,
|
| 1066 |
+
"step": 855
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.29,
|
| 1070 |
+
"learning_rate": 0.00017353760445682452,
|
| 1071 |
+
"loss": 1.393,
|
| 1072 |
+
"step": 860
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 0.3,
|
| 1076 |
+
"learning_rate": 0.0001733635097493036,
|
| 1077 |
+
"loss": 1.3176,
|
| 1078 |
+
"step": 865
|
| 1079 |
+
},
|
| 1080 |
+
{
|
| 1081 |
+
"epoch": 0.3,
|
| 1082 |
+
"learning_rate": 0.00017318941504178273,
|
| 1083 |
+
"loss": 1.3603,
|
| 1084 |
+
"step": 870
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 0.3,
|
| 1088 |
+
"learning_rate": 0.00017301532033426185,
|
| 1089 |
+
"loss": 1.3858,
|
| 1090 |
+
"step": 875
|
| 1091 |
+
},
|
| 1092 |
+
{
|
| 1093 |
+
"epoch": 0.3,
|
| 1094 |
+
"learning_rate": 0.00017284122562674097,
|
| 1095 |
+
"loss": 1.3801,
|
| 1096 |
+
"step": 880
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 0.3,
|
| 1100 |
+
"learning_rate": 0.00017266713091922006,
|
| 1101 |
+
"loss": 1.3258,
|
| 1102 |
+
"step": 885
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 0.3,
|
| 1106 |
+
"learning_rate": 0.00017249303621169916,
|
| 1107 |
+
"loss": 1.3958,
|
| 1108 |
+
"step": 890
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.31,
|
| 1112 |
+
"learning_rate": 0.00017231894150417828,
|
| 1113 |
+
"loss": 1.3352,
|
| 1114 |
+
"step": 895
|
| 1115 |
+
},
|
| 1116 |
+
{
|
| 1117 |
+
"epoch": 0.31,
|
| 1118 |
+
"learning_rate": 0.0001721448467966574,
|
| 1119 |
+
"loss": 1.3546,
|
| 1120 |
+
"step": 900
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 0.31,
|
| 1124 |
+
"learning_rate": 0.0001719707520891365,
|
| 1125 |
+
"loss": 1.3576,
|
| 1126 |
+
"step": 905
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 0.31,
|
| 1130 |
+
"learning_rate": 0.0001717966573816156,
|
| 1131 |
+
"loss": 1.5061,
|
| 1132 |
+
"step": 910
|
| 1133 |
+
},
|
| 1134 |
+
{
|
| 1135 |
+
"epoch": 0.31,
|
| 1136 |
+
"learning_rate": 0.0001716225626740947,
|
| 1137 |
+
"loss": 1.3496,
|
| 1138 |
+
"step": 915
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.31,
|
| 1142 |
+
"learning_rate": 0.00017144846796657385,
|
| 1143 |
+
"loss": 1.4537,
|
| 1144 |
+
"step": 920
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 0.32,
|
| 1148 |
+
"learning_rate": 0.00017127437325905294,
|
| 1149 |
+
"loss": 1.4241,
|
| 1150 |
+
"step": 925
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.32,
|
| 1154 |
+
"learning_rate": 0.00017110027855153203,
|
| 1155 |
+
"loss": 1.4109,
|
| 1156 |
+
"step": 930
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 0.32,
|
| 1160 |
+
"learning_rate": 0.00017092618384401115,
|
| 1161 |
+
"loss": 1.4104,
|
| 1162 |
+
"step": 935
|
| 1163 |
+
},
|
| 1164 |
+
{
|
| 1165 |
+
"epoch": 0.32,
|
| 1166 |
+
"learning_rate": 0.00017075208913649024,
|
| 1167 |
+
"loss": 1.4555,
|
| 1168 |
+
"step": 940
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 0.32,
|
| 1172 |
+
"learning_rate": 0.00017057799442896936,
|
| 1173 |
+
"loss": 1.2743,
|
| 1174 |
+
"step": 945
|
| 1175 |
+
},
|
| 1176 |
+
{
|
| 1177 |
+
"epoch": 0.33,
|
| 1178 |
+
"learning_rate": 0.00017040389972144848,
|
| 1179 |
+
"loss": 1.3573,
|
| 1180 |
+
"step": 950
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 0.33,
|
| 1184 |
+
"learning_rate": 0.00017022980501392757,
|
| 1185 |
+
"loss": 1.3463,
|
| 1186 |
+
"step": 955
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 0.33,
|
| 1190 |
+
"learning_rate": 0.0001700557103064067,
|
| 1191 |
+
"loss": 1.3817,
|
| 1192 |
+
"step": 960
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 0.33,
|
| 1196 |
+
"learning_rate": 0.0001698816155988858,
|
| 1197 |
+
"loss": 1.4315,
|
| 1198 |
+
"step": 965
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 0.33,
|
| 1202 |
+
"learning_rate": 0.0001697075208913649,
|
| 1203 |
+
"loss": 1.3112,
|
| 1204 |
+
"step": 970
|
| 1205 |
+
},
|
| 1206 |
+
{
|
| 1207 |
+
"epoch": 0.33,
|
| 1208 |
+
"learning_rate": 0.00016953342618384403,
|
| 1209 |
+
"loss": 1.3998,
|
| 1210 |
+
"step": 975
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 0.34,
|
| 1214 |
+
"learning_rate": 0.00016935933147632312,
|
| 1215 |
+
"loss": 1.3697,
|
| 1216 |
+
"step": 980
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 0.34,
|
| 1220 |
+
"learning_rate": 0.00016918523676880224,
|
| 1221 |
+
"loss": 1.3726,
|
| 1222 |
+
"step": 985
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 0.34,
|
| 1226 |
+
"learning_rate": 0.00016901114206128136,
|
| 1227 |
+
"loss": 1.4137,
|
| 1228 |
+
"step": 990
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 0.34,
|
| 1232 |
+
"learning_rate": 0.00016883704735376045,
|
| 1233 |
+
"loss": 1.3371,
|
| 1234 |
+
"step": 995
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 0.34,
|
| 1238 |
+
"learning_rate": 0.00016866295264623957,
|
| 1239 |
+
"loss": 1.3189,
|
| 1240 |
+
"step": 1000
|
| 1241 |
+
},
|
| 1242 |
+
{
|
| 1243 |
+
"epoch": 0.34,
|
| 1244 |
+
"eval_loss": 1.3234615325927734,
|
| 1245 |
+
"eval_runtime": 736.9447,
|
| 1246 |
+
"eval_samples_per_second": 7.049,
|
| 1247 |
+
"eval_steps_per_second": 0.882,
|
| 1248 |
+
"step": 1000
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.34,
|
| 1252 |
+
"learning_rate": 0.00016848885793871866,
|
| 1253 |
+
"loss": 1.3207,
|
| 1254 |
+
"step": 1005
|
| 1255 |
+
},
|
| 1256 |
+
{
|
| 1257 |
+
"epoch": 0.35,
|
| 1258 |
+
"learning_rate": 0.00016831476323119778,
|
| 1259 |
+
"loss": 1.3369,
|
| 1260 |
+
"step": 1010
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 0.35,
|
| 1264 |
+
"learning_rate": 0.0001681406685236769,
|
| 1265 |
+
"loss": 1.3205,
|
| 1266 |
+
"step": 1015
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.35,
|
| 1270 |
+
"learning_rate": 0.000167966573816156,
|
| 1271 |
+
"loss": 1.3124,
|
| 1272 |
+
"step": 1020
|
| 1273 |
+
},
|
| 1274 |
+
{
|
| 1275 |
+
"epoch": 0.35,
|
| 1276 |
+
"learning_rate": 0.0001677924791086351,
|
| 1277 |
+
"loss": 1.3317,
|
| 1278 |
+
"step": 1025
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 0.35,
|
| 1282 |
+
"learning_rate": 0.0001676183844011142,
|
| 1283 |
+
"loss": 1.3713,
|
| 1284 |
+
"step": 1030
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 0.35,
|
| 1288 |
+
"learning_rate": 0.00016744428969359332,
|
| 1289 |
+
"loss": 1.3276,
|
| 1290 |
+
"step": 1035
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 0.36,
|
| 1294 |
+
"learning_rate": 0.00016727019498607244,
|
| 1295 |
+
"loss": 1.2705,
|
| 1296 |
+
"step": 1040
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 0.36,
|
| 1300 |
+
"learning_rate": 0.00016709610027855154,
|
| 1301 |
+
"loss": 1.4057,
|
| 1302 |
+
"step": 1045
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 0.36,
|
| 1306 |
+
"learning_rate": 0.00016692200557103066,
|
| 1307 |
+
"loss": 1.4044,
|
| 1308 |
+
"step": 1050
|
| 1309 |
+
},
|
| 1310 |
+
{
|
| 1311 |
+
"epoch": 0.36,
|
| 1312 |
+
"learning_rate": 0.00016674791086350975,
|
| 1313 |
+
"loss": 1.3438,
|
| 1314 |
+
"step": 1055
|
| 1315 |
+
},
|
| 1316 |
+
{
|
| 1317 |
+
"epoch": 0.36,
|
| 1318 |
+
"learning_rate": 0.00016657381615598887,
|
| 1319 |
+
"loss": 1.3871,
|
| 1320 |
+
"step": 1060
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 0.36,
|
| 1324 |
+
"learning_rate": 0.000166399721448468,
|
| 1325 |
+
"loss": 1.3212,
|
| 1326 |
+
"step": 1065
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 0.37,
|
| 1330 |
+
"learning_rate": 0.00016622562674094708,
|
| 1331 |
+
"loss": 1.4142,
|
| 1332 |
+
"step": 1070
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 0.37,
|
| 1336 |
+
"learning_rate": 0.00016605153203342617,
|
| 1337 |
+
"loss": 1.3637,
|
| 1338 |
+
"step": 1075
|
| 1339 |
+
},
|
| 1340 |
+
{
|
| 1341 |
+
"epoch": 0.37,
|
| 1342 |
+
"learning_rate": 0.00016587743732590532,
|
| 1343 |
+
"loss": 1.2748,
|
| 1344 |
+
"step": 1080
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 0.37,
|
| 1348 |
+
"learning_rate": 0.0001657033426183844,
|
| 1349 |
+
"loss": 1.4044,
|
| 1350 |
+
"step": 1085
|
| 1351 |
+
},
|
| 1352 |
+
{
|
| 1353 |
+
"epoch": 0.37,
|
| 1354 |
+
"learning_rate": 0.00016552924791086353,
|
| 1355 |
+
"loss": 1.3374,
|
| 1356 |
+
"step": 1090
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 0.37,
|
| 1360 |
+
"learning_rate": 0.00016535515320334262,
|
| 1361 |
+
"loss": 1.3647,
|
| 1362 |
+
"step": 1095
|
| 1363 |
+
},
|
| 1364 |
+
{
|
| 1365 |
+
"epoch": 0.38,
|
| 1366 |
+
"learning_rate": 0.00016518105849582171,
|
| 1367 |
+
"loss": 1.4259,
|
| 1368 |
+
"step": 1100
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 0.38,
|
| 1372 |
+
"learning_rate": 0.00016500696378830086,
|
| 1373 |
+
"loss": 1.2926,
|
| 1374 |
+
"step": 1105
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 0.38,
|
| 1378 |
+
"learning_rate": 0.00016483286908077995,
|
| 1379 |
+
"loss": 1.4621,
|
| 1380 |
+
"step": 1110
|
| 1381 |
+
},
|
| 1382 |
+
{
|
| 1383 |
+
"epoch": 0.38,
|
| 1384 |
+
"learning_rate": 0.00016465877437325905,
|
| 1385 |
+
"loss": 1.3746,
|
| 1386 |
+
"step": 1115
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 0.38,
|
| 1390 |
+
"learning_rate": 0.00016448467966573817,
|
| 1391 |
+
"loss": 1.4036,
|
| 1392 |
+
"step": 1120
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.39,
|
| 1396 |
+
"learning_rate": 0.00016431058495821729,
|
| 1397 |
+
"loss": 1.4062,
|
| 1398 |
+
"step": 1125
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 0.39,
|
| 1402 |
+
"learning_rate": 0.0001641364902506964,
|
| 1403 |
+
"loss": 1.348,
|
| 1404 |
+
"step": 1130
|
| 1405 |
+
},
|
| 1406 |
+
{
|
| 1407 |
+
"epoch": 0.39,
|
| 1408 |
+
"learning_rate": 0.0001639623955431755,
|
| 1409 |
+
"loss": 1.3271,
|
| 1410 |
+
"step": 1135
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 0.39,
|
| 1414 |
+
"learning_rate": 0.0001637883008356546,
|
| 1415 |
+
"loss": 1.3663,
|
| 1416 |
+
"step": 1140
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 0.39,
|
| 1420 |
+
"learning_rate": 0.0001636142061281337,
|
| 1421 |
+
"loss": 1.3276,
|
| 1422 |
+
"step": 1145
|
| 1423 |
+
},
|
| 1424 |
+
{
|
| 1425 |
+
"epoch": 0.39,
|
| 1426 |
+
"learning_rate": 0.00016344011142061283,
|
| 1427 |
+
"loss": 1.2809,
|
| 1428 |
+
"step": 1150
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 0.4,
|
| 1432 |
+
"learning_rate": 0.00016326601671309195,
|
| 1433 |
+
"loss": 1.3586,
|
| 1434 |
+
"step": 1155
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 0.4,
|
| 1438 |
+
"learning_rate": 0.00016309192200557104,
|
| 1439 |
+
"loss": 1.3864,
|
| 1440 |
+
"step": 1160
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 0.4,
|
| 1444 |
+
"learning_rate": 0.00016291782729805013,
|
| 1445 |
+
"loss": 1.3163,
|
| 1446 |
+
"step": 1165
|
| 1447 |
+
},
|
| 1448 |
+
{
|
| 1449 |
+
"epoch": 0.4,
|
| 1450 |
+
"learning_rate": 0.00016274373259052925,
|
| 1451 |
+
"loss": 1.3677,
|
| 1452 |
+
"step": 1170
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 0.4,
|
| 1456 |
+
"learning_rate": 0.00016256963788300837,
|
| 1457 |
+
"loss": 1.2984,
|
| 1458 |
+
"step": 1175
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 0.4,
|
| 1462 |
+
"learning_rate": 0.00016239554317548746,
|
| 1463 |
+
"loss": 1.4083,
|
| 1464 |
+
"step": 1180
|
| 1465 |
+
},
|
| 1466 |
+
{
|
| 1467 |
+
"epoch": 0.41,
|
| 1468 |
+
"learning_rate": 0.00016222144846796658,
|
| 1469 |
+
"loss": 1.445,
|
| 1470 |
+
"step": 1185
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 0.41,
|
| 1474 |
+
"learning_rate": 0.00016204735376044568,
|
| 1475 |
+
"loss": 1.3306,
|
| 1476 |
+
"step": 1190
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 0.41,
|
| 1480 |
+
"learning_rate": 0.0001618732590529248,
|
| 1481 |
+
"loss": 1.4185,
|
| 1482 |
+
"step": 1195
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 0.41,
|
| 1486 |
+
"learning_rate": 0.00016169916434540392,
|
| 1487 |
+
"loss": 1.3581,
|
| 1488 |
+
"step": 1200
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 0.41,
|
| 1492 |
+
"eval_loss": 1.318132996559143,
|
| 1493 |
+
"eval_runtime": 737.5524,
|
| 1494 |
+
"eval_samples_per_second": 7.044,
|
| 1495 |
+
"eval_steps_per_second": 0.881,
|
| 1496 |
+
"step": 1200
|
| 1497 |
+
}
|
| 1498 |
+
],
|
| 1499 |
+
"max_steps": 5844,
|
| 1500 |
+
"num_train_epochs": 2,
|
| 1501 |
+
"total_flos": 7.14718259970048e+16,
|
| 1502 |
+
"trial_name": null,
|
| 1503 |
+
"trial_params": null
|
| 1504 |
+
}
|
checkpoint-1200/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b1519d033b5a5a628e3283f98e5ba6662a6b6f8c76dfadfd8264558baafe9285
|
| 3 |
+
size 3963
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 12653389
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:906c9686996da6e301977bd38c2c83361146915ac6bd33d6e07e32d7953b4bbe
|
| 3 |
size 12653389
|
runs/Jun08_19-57-59_5b791b640f29/events.out.tfevents.1686254295.5b791b640f29.4078.0
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:b4ea3a630d828ec268790573a366e7db6946cf785e801feb82fe7e166095b728
|
| 3 |
+
size 43538
|