Upload 11 files
Browse files- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1442 -0
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1e4b0ee667ce3d9b24472c6e5e7a7dab51f3f63174e1e5cc435a1487d4ad2645
|
| 3 |
+
size 3137574192
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6325268ebe7563149341153b5a7af45a0e5235e22e7f7b27c3ea5acbb847511f
|
| 3 |
+
size 14308
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0c835d3b1463b1684087991cc2048e9b35be194145ec7c5f7fffa5c5ef05d96d
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1442 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.07299865617200998,
|
| 3 |
+
"best_model_checkpoint": "d:\\\\whisper-large-v3-pt-cv19-fleurs\\checkpoint-5000",
|
| 4 |
+
"epoch": 2.288329519450801,
|
| 5 |
+
"eval_steps": 5000,
|
| 6 |
+
"global_step": 5000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.011441647597254004,
|
| 13 |
+
"grad_norm": 6.62877082824707,
|
| 14 |
+
"learning_rate": 1.5625000000000003e-08,
|
| 15 |
+
"loss": 0.7265,
|
| 16 |
+
"step": 25
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.02288329519450801,
|
| 20 |
+
"grad_norm": Infinity,
|
| 21 |
+
"learning_rate": 3e-08,
|
| 22 |
+
"loss": 1.2394,
|
| 23 |
+
"step": 50
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.034324942791762014,
|
| 27 |
+
"grad_norm": 5.83213996887207,
|
| 28 |
+
"learning_rate": 4.5625e-08,
|
| 29 |
+
"loss": 0.721,
|
| 30 |
+
"step": 75
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.04576659038901602,
|
| 34 |
+
"grad_norm": 19.50699806213379,
|
| 35 |
+
"learning_rate": 6.125e-08,
|
| 36 |
+
"loss": 1.2406,
|
| 37 |
+
"step": 100
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.057208237986270026,
|
| 41 |
+
"grad_norm": 6.239120960235596,
|
| 42 |
+
"learning_rate": 7.6875e-08,
|
| 43 |
+
"loss": 0.7169,
|
| 44 |
+
"step": 125
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.06864988558352403,
|
| 48 |
+
"grad_norm": 18.885549545288086,
|
| 49 |
+
"learning_rate": 9.250000000000001e-08,
|
| 50 |
+
"loss": 1.2154,
|
| 51 |
+
"step": 150
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.08009153318077804,
|
| 55 |
+
"grad_norm": 7.490365982055664,
|
| 56 |
+
"learning_rate": 1.08125e-07,
|
| 57 |
+
"loss": 0.6488,
|
| 58 |
+
"step": 175
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.09153318077803203,
|
| 62 |
+
"grad_norm": 15.990517616271973,
|
| 63 |
+
"learning_rate": 1.2375e-07,
|
| 64 |
+
"loss": 1.0847,
|
| 65 |
+
"step": 200
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.10297482837528604,
|
| 69 |
+
"grad_norm": 7.532449722290039,
|
| 70 |
+
"learning_rate": 1.3937500000000002e-07,
|
| 71 |
+
"loss": 0.5148,
|
| 72 |
+
"step": 225
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.11441647597254005,
|
| 76 |
+
"grad_norm": 15.39826774597168,
|
| 77 |
+
"learning_rate": 1.55e-07,
|
| 78 |
+
"loss": 0.6084,
|
| 79 |
+
"step": 250
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.12585812356979406,
|
| 83 |
+
"grad_norm": 4.829259872436523,
|
| 84 |
+
"learning_rate": 1.70625e-07,
|
| 85 |
+
"loss": 0.2526,
|
| 86 |
+
"step": 275
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.13729977116704806,
|
| 90 |
+
"grad_norm": 14.0956392288208,
|
| 91 |
+
"learning_rate": 1.8625000000000002e-07,
|
| 92 |
+
"loss": 0.3234,
|
| 93 |
+
"step": 300
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.14874141876430205,
|
| 97 |
+
"grad_norm": 3.5299148559570312,
|
| 98 |
+
"learning_rate": 2.0187500000000003e-07,
|
| 99 |
+
"loss": 0.1732,
|
| 100 |
+
"step": 325
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.16018306636155608,
|
| 104 |
+
"grad_norm": 14.003329277038574,
|
| 105 |
+
"learning_rate": 2.1749999999999998e-07,
|
| 106 |
+
"loss": 0.228,
|
| 107 |
+
"step": 350
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.17162471395881007,
|
| 111 |
+
"grad_norm": 3.2087361812591553,
|
| 112 |
+
"learning_rate": 2.3312500000000002e-07,
|
| 113 |
+
"loss": 0.1796,
|
| 114 |
+
"step": 375
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.18306636155606407,
|
| 118 |
+
"grad_norm": 13.039973258972168,
|
| 119 |
+
"learning_rate": 2.4875e-07,
|
| 120 |
+
"loss": 0.2275,
|
| 121 |
+
"step": 400
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.1945080091533181,
|
| 125 |
+
"grad_norm": 3.967797040939331,
|
| 126 |
+
"learning_rate": 2.64375e-07,
|
| 127 |
+
"loss": 0.1404,
|
| 128 |
+
"step": 425
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.20594965675057209,
|
| 132 |
+
"grad_norm": 13.055093765258789,
|
| 133 |
+
"learning_rate": 2.8e-07,
|
| 134 |
+
"loss": 0.1859,
|
| 135 |
+
"step": 450
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.21739130434782608,
|
| 139 |
+
"grad_norm": 2.9818637371063232,
|
| 140 |
+
"learning_rate": 2.9562500000000003e-07,
|
| 141 |
+
"loss": 0.1438,
|
| 142 |
+
"step": 475
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.2288329519450801,
|
| 146 |
+
"grad_norm": 8.552623748779297,
|
| 147 |
+
"learning_rate": 3.1125e-07,
|
| 148 |
+
"loss": 0.1653,
|
| 149 |
+
"step": 500
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.2402745995423341,
|
| 153 |
+
"grad_norm": 5.015272617340088,
|
| 154 |
+
"learning_rate": 3.26875e-07,
|
| 155 |
+
"loss": 0.1247,
|
| 156 |
+
"step": 525
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.2517162471395881,
|
| 160 |
+
"grad_norm": 8.722563743591309,
|
| 161 |
+
"learning_rate": 3.425e-07,
|
| 162 |
+
"loss": 0.1648,
|
| 163 |
+
"step": 550
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.2631578947368421,
|
| 167 |
+
"grad_norm": 3.586549997329712,
|
| 168 |
+
"learning_rate": 3.58125e-07,
|
| 169 |
+
"loss": 0.1415,
|
| 170 |
+
"step": 575
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.2745995423340961,
|
| 174 |
+
"grad_norm": 14.463945388793945,
|
| 175 |
+
"learning_rate": 3.7375e-07,
|
| 176 |
+
"loss": 0.1626,
|
| 177 |
+
"step": 600
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.28604118993135014,
|
| 181 |
+
"grad_norm": 4.430425643920898,
|
| 182 |
+
"learning_rate": 3.8937500000000003e-07,
|
| 183 |
+
"loss": 0.1349,
|
| 184 |
+
"step": 625
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.2974828375286041,
|
| 188 |
+
"grad_norm": 12.54709243774414,
|
| 189 |
+
"learning_rate": 4.05e-07,
|
| 190 |
+
"loss": 0.1453,
|
| 191 |
+
"step": 650
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.30892448512585813,
|
| 195 |
+
"grad_norm": 4.3557939529418945,
|
| 196 |
+
"learning_rate": 4.20625e-07,
|
| 197 |
+
"loss": 0.1385,
|
| 198 |
+
"step": 675
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.32036613272311215,
|
| 202 |
+
"grad_norm": 7.513607501983643,
|
| 203 |
+
"learning_rate": 4.3625e-07,
|
| 204 |
+
"loss": 0.1605,
|
| 205 |
+
"step": 700
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.3318077803203661,
|
| 209 |
+
"grad_norm": 4.362117767333984,
|
| 210 |
+
"learning_rate": 4.5187500000000007e-07,
|
| 211 |
+
"loss": 0.1219,
|
| 212 |
+
"step": 725
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.34324942791762014,
|
| 216 |
+
"grad_norm": 9.83281421661377,
|
| 217 |
+
"learning_rate": 4.675000000000001e-07,
|
| 218 |
+
"loss": 0.1569,
|
| 219 |
+
"step": 750
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.35469107551487417,
|
| 223 |
+
"grad_norm": 3.5240659713745117,
|
| 224 |
+
"learning_rate": 4.83125e-07,
|
| 225 |
+
"loss": 0.1357,
|
| 226 |
+
"step": 775
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.36613272311212813,
|
| 230 |
+
"grad_norm": 8.545495986938477,
|
| 231 |
+
"learning_rate": 4.9875e-07,
|
| 232 |
+
"loss": 0.1462,
|
| 233 |
+
"step": 800
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.37757437070938216,
|
| 237 |
+
"grad_norm": 4.000309467315674,
|
| 238 |
+
"learning_rate": 5.14375e-07,
|
| 239 |
+
"loss": 0.1222,
|
| 240 |
+
"step": 825
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.3890160183066362,
|
| 244 |
+
"grad_norm": 10.31431770324707,
|
| 245 |
+
"learning_rate": 5.3e-07,
|
| 246 |
+
"loss": 0.1344,
|
| 247 |
+
"step": 850
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.40045766590389015,
|
| 251 |
+
"grad_norm": 2.6014370918273926,
|
| 252 |
+
"learning_rate": 5.456250000000001e-07,
|
| 253 |
+
"loss": 0.1284,
|
| 254 |
+
"step": 875
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.41189931350114417,
|
| 258 |
+
"grad_norm": 7.926617622375488,
|
| 259 |
+
"learning_rate": 5.6125e-07,
|
| 260 |
+
"loss": 0.1414,
|
| 261 |
+
"step": 900
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.4233409610983982,
|
| 265 |
+
"grad_norm": 4.362987518310547,
|
| 266 |
+
"learning_rate": 5.76875e-07,
|
| 267 |
+
"loss": 0.1258,
|
| 268 |
+
"step": 925
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.43478260869565216,
|
| 272 |
+
"grad_norm": 8.642595291137695,
|
| 273 |
+
"learning_rate": 5.925e-07,
|
| 274 |
+
"loss": 0.1303,
|
| 275 |
+
"step": 950
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.4462242562929062,
|
| 279 |
+
"grad_norm": 2.2307755947113037,
|
| 280 |
+
"learning_rate": 6.08125e-07,
|
| 281 |
+
"loss": 0.1266,
|
| 282 |
+
"step": 975
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.4576659038901602,
|
| 286 |
+
"grad_norm": 10.151159286499023,
|
| 287 |
+
"learning_rate": 6.237500000000001e-07,
|
| 288 |
+
"loss": 0.1332,
|
| 289 |
+
"step": 1000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.4691075514874142,
|
| 293 |
+
"grad_norm": 3.0874152183532715,
|
| 294 |
+
"learning_rate": 6.39375e-07,
|
| 295 |
+
"loss": 0.1192,
|
| 296 |
+
"step": 1025
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.4805491990846682,
|
| 300 |
+
"grad_norm": 7.633583068847656,
|
| 301 |
+
"learning_rate": 6.550000000000001e-07,
|
| 302 |
+
"loss": 0.1204,
|
| 303 |
+
"step": 1050
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.4919908466819222,
|
| 307 |
+
"grad_norm": 2.0958948135375977,
|
| 308 |
+
"learning_rate": 6.70625e-07,
|
| 309 |
+
"loss": 0.1142,
|
| 310 |
+
"step": 1075
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.5034324942791762,
|
| 314 |
+
"grad_norm": 7.018194198608398,
|
| 315 |
+
"learning_rate": 6.8625e-07,
|
| 316 |
+
"loss": 0.1014,
|
| 317 |
+
"step": 1100
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.5148741418764302,
|
| 321 |
+
"grad_norm": 5.929811000823975,
|
| 322 |
+
"learning_rate": 7.018750000000001e-07,
|
| 323 |
+
"loss": 0.1251,
|
| 324 |
+
"step": 1125
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.5263157894736842,
|
| 328 |
+
"grad_norm": 11.825408935546875,
|
| 329 |
+
"learning_rate": 7.175e-07,
|
| 330 |
+
"loss": 0.1426,
|
| 331 |
+
"step": 1150
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.5377574370709383,
|
| 335 |
+
"grad_norm": 1.4950259923934937,
|
| 336 |
+
"learning_rate": 7.331250000000001e-07,
|
| 337 |
+
"loss": 0.1195,
|
| 338 |
+
"step": 1175
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.5491990846681922,
|
| 342 |
+
"grad_norm": 3.756826639175415,
|
| 343 |
+
"learning_rate": 7.4875e-07,
|
| 344 |
+
"loss": 0.1139,
|
| 345 |
+
"step": 1200
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.5606407322654462,
|
| 349 |
+
"grad_norm": 1.7584425210952759,
|
| 350 |
+
"learning_rate": 7.643750000000001e-07,
|
| 351 |
+
"loss": 0.1241,
|
| 352 |
+
"step": 1225
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.5720823798627003,
|
| 356 |
+
"grad_norm": 9.659570693969727,
|
| 357 |
+
"learning_rate": 7.8e-07,
|
| 358 |
+
"loss": 0.1175,
|
| 359 |
+
"step": 1250
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.5835240274599542,
|
| 363 |
+
"grad_norm": 2.339547634124756,
|
| 364 |
+
"learning_rate": 7.95625e-07,
|
| 365 |
+
"loss": 0.1153,
|
| 366 |
+
"step": 1275
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.5949656750572082,
|
| 370 |
+
"grad_norm": 8.946951866149902,
|
| 371 |
+
"learning_rate": 8.112500000000001e-07,
|
| 372 |
+
"loss": 0.1278,
|
| 373 |
+
"step": 1300
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.6064073226544623,
|
| 377 |
+
"grad_norm": 1.9447693824768066,
|
| 378 |
+
"learning_rate": 8.26875e-07,
|
| 379 |
+
"loss": 0.1213,
|
| 380 |
+
"step": 1325
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.6178489702517163,
|
| 384 |
+
"grad_norm": 16.7085018157959,
|
| 385 |
+
"learning_rate": 8.425000000000001e-07,
|
| 386 |
+
"loss": 0.14,
|
| 387 |
+
"step": 1350
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.6292906178489702,
|
| 391 |
+
"grad_norm": 2.692288875579834,
|
| 392 |
+
"learning_rate": 8.58125e-07,
|
| 393 |
+
"loss": 0.1279,
|
| 394 |
+
"step": 1375
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.6407322654462243,
|
| 398 |
+
"grad_norm": 8.288985252380371,
|
| 399 |
+
"learning_rate": 8.737500000000001e-07,
|
| 400 |
+
"loss": 0.1041,
|
| 401 |
+
"step": 1400
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.6521739130434783,
|
| 405 |
+
"grad_norm": 2.5615968704223633,
|
| 406 |
+
"learning_rate": 8.893750000000001e-07,
|
| 407 |
+
"loss": 0.1207,
|
| 408 |
+
"step": 1425
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.6636155606407322,
|
| 412 |
+
"grad_norm": 5.402917861938477,
|
| 413 |
+
"learning_rate": 9.050000000000001e-07,
|
| 414 |
+
"loss": 0.1209,
|
| 415 |
+
"step": 1450
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.6750572082379863,
|
| 419 |
+
"grad_norm": 3.1846108436584473,
|
| 420 |
+
"learning_rate": 9.20625e-07,
|
| 421 |
+
"loss": 0.1131,
|
| 422 |
+
"step": 1475
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.6864988558352403,
|
| 426 |
+
"grad_norm": 7.106233596801758,
|
| 427 |
+
"learning_rate": 9.362499999999999e-07,
|
| 428 |
+
"loss": 0.1206,
|
| 429 |
+
"step": 1500
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.6979405034324943,
|
| 433 |
+
"grad_norm": 1.574010968208313,
|
| 434 |
+
"learning_rate": 9.51875e-07,
|
| 435 |
+
"loss": 0.1169,
|
| 436 |
+
"step": 1525
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.7093821510297483,
|
| 440 |
+
"grad_norm": 12.645787239074707,
|
| 441 |
+
"learning_rate": 9.675e-07,
|
| 442 |
+
"loss": 0.153,
|
| 443 |
+
"step": 1550
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.7208237986270023,
|
| 447 |
+
"grad_norm": 2.3170979022979736,
|
| 448 |
+
"learning_rate": 9.83125e-07,
|
| 449 |
+
"loss": 0.1196,
|
| 450 |
+
"step": 1575
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.7322654462242563,
|
| 454 |
+
"grad_norm": 6.579622745513916,
|
| 455 |
+
"learning_rate": 9.9875e-07,
|
| 456 |
+
"loss": 0.1112,
|
| 457 |
+
"step": 1600
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.7437070938215103,
|
| 461 |
+
"grad_norm": 3.6276774406433105,
|
| 462 |
+
"learning_rate": 1.0143750000000001e-06,
|
| 463 |
+
"loss": 0.1174,
|
| 464 |
+
"step": 1625
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.7551487414187643,
|
| 468 |
+
"grad_norm": 10.94548511505127,
|
| 469 |
+
"learning_rate": 1.03e-06,
|
| 470 |
+
"loss": 0.1115,
|
| 471 |
+
"step": 1650
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.7665903890160183,
|
| 475 |
+
"grad_norm": 4.662591934204102,
|
| 476 |
+
"learning_rate": 1.045625e-06,
|
| 477 |
+
"loss": 0.103,
|
| 478 |
+
"step": 1675
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.7780320366132724,
|
| 482 |
+
"grad_norm": 11.371915817260742,
|
| 483 |
+
"learning_rate": 1.06125e-06,
|
| 484 |
+
"loss": 0.1163,
|
| 485 |
+
"step": 1700
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.7894736842105263,
|
| 489 |
+
"grad_norm": 2.4760260581970215,
|
| 490 |
+
"learning_rate": 1.0768750000000002e-06,
|
| 491 |
+
"loss": 0.1047,
|
| 492 |
+
"step": 1725
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.8009153318077803,
|
| 496 |
+
"grad_norm": 7.711414337158203,
|
| 497 |
+
"learning_rate": 1.0925000000000001e-06,
|
| 498 |
+
"loss": 0.1167,
|
| 499 |
+
"step": 1750
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.8123569794050344,
|
| 503 |
+
"grad_norm": 3.2793569564819336,
|
| 504 |
+
"learning_rate": 1.108125e-06,
|
| 505 |
+
"loss": 0.119,
|
| 506 |
+
"step": 1775
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.8237986270022883,
|
| 510 |
+
"grad_norm": 7.447999000549316,
|
| 511 |
+
"learning_rate": 1.12375e-06,
|
| 512 |
+
"loss": 0.1219,
|
| 513 |
+
"step": 1800
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.8352402745995423,
|
| 517 |
+
"grad_norm": 1.265462040901184,
|
| 518 |
+
"learning_rate": 1.139375e-06,
|
| 519 |
+
"loss": 0.1011,
|
| 520 |
+
"step": 1825
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.8466819221967964,
|
| 524 |
+
"grad_norm": 26.126134872436523,
|
| 525 |
+
"learning_rate": 1.155e-06,
|
| 526 |
+
"loss": 0.132,
|
| 527 |
+
"step": 1850
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.8581235697940504,
|
| 531 |
+
"grad_norm": 2.147458553314209,
|
| 532 |
+
"learning_rate": 1.1706250000000001e-06,
|
| 533 |
+
"loss": 0.1171,
|
| 534 |
+
"step": 1875
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.8695652173913043,
|
| 538 |
+
"grad_norm": 11.847962379455566,
|
| 539 |
+
"learning_rate": 1.18625e-06,
|
| 540 |
+
"loss": 0.1223,
|
| 541 |
+
"step": 1900
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.8810068649885584,
|
| 545 |
+
"grad_norm": 1.5704588890075684,
|
| 546 |
+
"learning_rate": 1.201875e-06,
|
| 547 |
+
"loss": 0.1207,
|
| 548 |
+
"step": 1925
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.8924485125858124,
|
| 552 |
+
"grad_norm": 12.10556411743164,
|
| 553 |
+
"learning_rate": 1.2175e-06,
|
| 554 |
+
"loss": 0.0989,
|
| 555 |
+
"step": 1950
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.9038901601830663,
|
| 559 |
+
"grad_norm": 3.2033963203430176,
|
| 560 |
+
"learning_rate": 1.2331250000000002e-06,
|
| 561 |
+
"loss": 0.1027,
|
| 562 |
+
"step": 1975
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.9153318077803204,
|
| 566 |
+
"grad_norm": 6.506619930267334,
|
| 567 |
+
"learning_rate": 1.2487500000000001e-06,
|
| 568 |
+
"loss": 0.1426,
|
| 569 |
+
"step": 2000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.9267734553775744,
|
| 573 |
+
"grad_norm": 2.171173095703125,
|
| 574 |
+
"learning_rate": 1.264375e-06,
|
| 575 |
+
"loss": 0.1036,
|
| 576 |
+
"step": 2025
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.9382151029748284,
|
| 580 |
+
"grad_norm": 8.255300521850586,
|
| 581 |
+
"learning_rate": 1.28e-06,
|
| 582 |
+
"loss": 0.1328,
|
| 583 |
+
"step": 2050
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.9496567505720824,
|
| 587 |
+
"grad_norm": 4.062456130981445,
|
| 588 |
+
"learning_rate": 1.2956250000000002e-06,
|
| 589 |
+
"loss": 0.1175,
|
| 590 |
+
"step": 2075
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.9610983981693364,
|
| 594 |
+
"grad_norm": 7.572447776794434,
|
| 595 |
+
"learning_rate": 1.31125e-06,
|
| 596 |
+
"loss": 0.1377,
|
| 597 |
+
"step": 2100
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.9725400457665904,
|
| 601 |
+
"grad_norm": 1.6148779392242432,
|
| 602 |
+
"learning_rate": 1.326875e-06,
|
| 603 |
+
"loss": 0.1134,
|
| 604 |
+
"step": 2125
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.9839816933638444,
|
| 608 |
+
"grad_norm": 5.8796186447143555,
|
| 609 |
+
"learning_rate": 1.3425e-06,
|
| 610 |
+
"loss": 0.1233,
|
| 611 |
+
"step": 2150
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.9954233409610984,
|
| 615 |
+
"grad_norm": 2.6012580394744873,
|
| 616 |
+
"learning_rate": 1.358125e-06,
|
| 617 |
+
"loss": 0.1141,
|
| 618 |
+
"step": 2175
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 1.0068649885583525,
|
| 622 |
+
"grad_norm": 2.369357109069824,
|
| 623 |
+
"learning_rate": 1.37375e-06,
|
| 624 |
+
"loss": 0.0964,
|
| 625 |
+
"step": 2200
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 1.0183066361556063,
|
| 629 |
+
"grad_norm": 3.426637887954712,
|
| 630 |
+
"learning_rate": 1.3893750000000001e-06,
|
| 631 |
+
"loss": 0.0844,
|
| 632 |
+
"step": 2225
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 1.0297482837528604,
|
| 636 |
+
"grad_norm": 2.0123000144958496,
|
| 637 |
+
"learning_rate": 1.405e-06,
|
| 638 |
+
"loss": 0.1125,
|
| 639 |
+
"step": 2250
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 1.0411899313501145,
|
| 643 |
+
"grad_norm": 2.8704395294189453,
|
| 644 |
+
"learning_rate": 1.420625e-06,
|
| 645 |
+
"loss": 0.0827,
|
| 646 |
+
"step": 2275
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 1.0526315789473684,
|
| 650 |
+
"grad_norm": 2.3413915634155273,
|
| 651 |
+
"learning_rate": 1.43625e-06,
|
| 652 |
+
"loss": 0.108,
|
| 653 |
+
"step": 2300
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 1.0640732265446224,
|
| 657 |
+
"grad_norm": 1.216963291168213,
|
| 658 |
+
"learning_rate": 1.4518750000000002e-06,
|
| 659 |
+
"loss": 0.0766,
|
| 660 |
+
"step": 2325
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 1.0755148741418765,
|
| 664 |
+
"grad_norm": 2.2277841567993164,
|
| 665 |
+
"learning_rate": 1.4675000000000001e-06,
|
| 666 |
+
"loss": 0.1035,
|
| 667 |
+
"step": 2350
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 1.0869565217391304,
|
| 671 |
+
"grad_norm": 3.6448659896850586,
|
| 672 |
+
"learning_rate": 1.483125e-06,
|
| 673 |
+
"loss": 0.0749,
|
| 674 |
+
"step": 2375
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 1.0983981693363845,
|
| 678 |
+
"grad_norm": 1.5013147592544556,
|
| 679 |
+
"learning_rate": 1.49875e-06,
|
| 680 |
+
"loss": 0.111,
|
| 681 |
+
"step": 2400
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 1.1098398169336385,
|
| 685 |
+
"grad_norm": 3.0823745727539062,
|
| 686 |
+
"learning_rate": 1.514375e-06,
|
| 687 |
+
"loss": 0.0842,
|
| 688 |
+
"step": 2425
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 1.1212814645308924,
|
| 692 |
+
"grad_norm": 1.238543152809143,
|
| 693 |
+
"learning_rate": 1.53e-06,
|
| 694 |
+
"loss": 0.0978,
|
| 695 |
+
"step": 2450
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 1.1327231121281465,
|
| 699 |
+
"grad_norm": 4.413508415222168,
|
| 700 |
+
"learning_rate": 1.545625e-06,
|
| 701 |
+
"loss": 0.0759,
|
| 702 |
+
"step": 2475
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 1.1441647597254005,
|
| 706 |
+
"grad_norm": 2.655590057373047,
|
| 707 |
+
"learning_rate": 1.56125e-06,
|
| 708 |
+
"loss": 0.1061,
|
| 709 |
+
"step": 2500
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 1.1556064073226544,
|
| 713 |
+
"grad_norm": 3.8559327125549316,
|
| 714 |
+
"learning_rate": 1.5768750000000003e-06,
|
| 715 |
+
"loss": 0.0876,
|
| 716 |
+
"step": 2525
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.1670480549199085,
|
| 720 |
+
"grad_norm": 2.452881336212158,
|
| 721 |
+
"learning_rate": 1.5925000000000002e-06,
|
| 722 |
+
"loss": 0.1068,
|
| 723 |
+
"step": 2550
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 1.1784897025171626,
|
| 727 |
+
"grad_norm": 3.86460280418396,
|
| 728 |
+
"learning_rate": 1.608125e-06,
|
| 729 |
+
"loss": 0.079,
|
| 730 |
+
"step": 2575
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 1.1899313501144164,
|
| 734 |
+
"grad_norm": 1.9224119186401367,
|
| 735 |
+
"learning_rate": 1.62375e-06,
|
| 736 |
+
"loss": 0.1139,
|
| 737 |
+
"step": 2600
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 1.2013729977116705,
|
| 741 |
+
"grad_norm": 7.143287181854248,
|
| 742 |
+
"learning_rate": 1.6393749999999999e-06,
|
| 743 |
+
"loss": 0.0802,
|
| 744 |
+
"step": 2625
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 1.2128146453089246,
|
| 748 |
+
"grad_norm": 1.8962572813034058,
|
| 749 |
+
"learning_rate": 1.655e-06,
|
| 750 |
+
"loss": 0.104,
|
| 751 |
+
"step": 2650
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 1.2242562929061784,
|
| 755 |
+
"grad_norm": 3.0214734077453613,
|
| 756 |
+
"learning_rate": 1.670625e-06,
|
| 757 |
+
"loss": 0.0711,
|
| 758 |
+
"step": 2675
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.2356979405034325,
|
| 762 |
+
"grad_norm": 2.490997314453125,
|
| 763 |
+
"learning_rate": 1.68625e-06,
|
| 764 |
+
"loss": 0.0987,
|
| 765 |
+
"step": 2700
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 1.2471395881006866,
|
| 769 |
+
"grad_norm": 3.6045401096343994,
|
| 770 |
+
"learning_rate": 1.701875e-06,
|
| 771 |
+
"loss": 0.0968,
|
| 772 |
+
"step": 2725
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 1.2585812356979404,
|
| 776 |
+
"grad_norm": 2.8490657806396484,
|
| 777 |
+
"learning_rate": 1.7175e-06,
|
| 778 |
+
"loss": 0.1124,
|
| 779 |
+
"step": 2750
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 1.2700228832951945,
|
| 783 |
+
"grad_norm": 6.608100414276123,
|
| 784 |
+
"learning_rate": 1.733125e-06,
|
| 785 |
+
"loss": 0.0855,
|
| 786 |
+
"step": 2775
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 1.2814645308924484,
|
| 790 |
+
"grad_norm": 2.4767751693725586,
|
| 791 |
+
"learning_rate": 1.74875e-06,
|
| 792 |
+
"loss": 0.1073,
|
| 793 |
+
"step": 2800
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 1.2929061784897025,
|
| 797 |
+
"grad_norm": 5.4979047775268555,
|
| 798 |
+
"learning_rate": 1.7643750000000002e-06,
|
| 799 |
+
"loss": 0.0772,
|
| 800 |
+
"step": 2825
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 1.3043478260869565,
|
| 804 |
+
"grad_norm": 2.7516839504241943,
|
| 805 |
+
"learning_rate": 1.7800000000000001e-06,
|
| 806 |
+
"loss": 0.111,
|
| 807 |
+
"step": 2850
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 1.3157894736842106,
|
| 811 |
+
"grad_norm": 3.251453161239624,
|
| 812 |
+
"learning_rate": 1.795625e-06,
|
| 813 |
+
"loss": 0.0759,
|
| 814 |
+
"step": 2875
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.3272311212814645,
|
| 818 |
+
"grad_norm": 3.175234079360962,
|
| 819 |
+
"learning_rate": 1.81125e-06,
|
| 820 |
+
"loss": 0.123,
|
| 821 |
+
"step": 2900
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.3386727688787186,
|
| 825 |
+
"grad_norm": 3.5862858295440674,
|
| 826 |
+
"learning_rate": 1.8268750000000002e-06,
|
| 827 |
+
"loss": 0.0735,
|
| 828 |
+
"step": 2925
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 1.3501144164759724,
|
| 832 |
+
"grad_norm": 2.170401096343994,
|
| 833 |
+
"learning_rate": 1.8425000000000001e-06,
|
| 834 |
+
"loss": 0.1189,
|
| 835 |
+
"step": 2950
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 1.3615560640732265,
|
| 839 |
+
"grad_norm": 3.134989023208618,
|
| 840 |
+
"learning_rate": 1.858125e-06,
|
| 841 |
+
"loss": 0.081,
|
| 842 |
+
"step": 2975
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 1.3729977116704806,
|
| 846 |
+
"grad_norm": 1.6760380268096924,
|
| 847 |
+
"learning_rate": 1.87375e-06,
|
| 848 |
+
"loss": 0.1074,
|
| 849 |
+
"step": 3000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 1.3844393592677346,
|
| 853 |
+
"grad_norm": 3.4386298656463623,
|
| 854 |
+
"learning_rate": 1.8893750000000002e-06,
|
| 855 |
+
"loss": 0.0704,
|
| 856 |
+
"step": 3025
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 1.3958810068649885,
|
| 860 |
+
"grad_norm": 2.0835421085357666,
|
| 861 |
+
"learning_rate": 1.9050000000000002e-06,
|
| 862 |
+
"loss": 0.1112,
|
| 863 |
+
"step": 3050
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 1.4073226544622426,
|
| 867 |
+
"grad_norm": 4.792001247406006,
|
| 868 |
+
"learning_rate": 1.9206250000000004e-06,
|
| 869 |
+
"loss": 0.0914,
|
| 870 |
+
"step": 3075
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 1.4187643020594964,
|
| 874 |
+
"grad_norm": 2.2697513103485107,
|
| 875 |
+
"learning_rate": 1.9362500000000003e-06,
|
| 876 |
+
"loss": 0.1044,
|
| 877 |
+
"step": 3100
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 1.4302059496567505,
|
| 881 |
+
"grad_norm": 1.7390592098236084,
|
| 882 |
+
"learning_rate": 1.9518750000000003e-06,
|
| 883 |
+
"loss": 0.0789,
|
| 884 |
+
"step": 3125
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 1.4416475972540046,
|
| 888 |
+
"grad_norm": 2.5395240783691406,
|
| 889 |
+
"learning_rate": 1.9675000000000002e-06,
|
| 890 |
+
"loss": 0.1103,
|
| 891 |
+
"step": 3150
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 1.4530892448512587,
|
| 895 |
+
"grad_norm": 9.415529251098633,
|
| 896 |
+
"learning_rate": 1.983125e-06,
|
| 897 |
+
"loss": 0.0815,
|
| 898 |
+
"step": 3175
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 1.4645308924485125,
|
| 902 |
+
"grad_norm": 3.2418212890625,
|
| 903 |
+
"learning_rate": 1.99875e-06,
|
| 904 |
+
"loss": 0.0877,
|
| 905 |
+
"step": 3200
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.4759725400457666,
|
| 909 |
+
"grad_norm": 2.5140135288238525,
|
| 910 |
+
"learning_rate": 2.014375e-06,
|
| 911 |
+
"loss": 0.072,
|
| 912 |
+
"step": 3225
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 1.4874141876430205,
|
| 916 |
+
"grad_norm": 1.8860526084899902,
|
| 917 |
+
"learning_rate": 2.03e-06,
|
| 918 |
+
"loss": 0.1082,
|
| 919 |
+
"step": 3250
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 1.4988558352402745,
|
| 923 |
+
"grad_norm": 3.322282075881958,
|
| 924 |
+
"learning_rate": 2.045625e-06,
|
| 925 |
+
"loss": 0.0703,
|
| 926 |
+
"step": 3275
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.5102974828375286,
|
| 930 |
+
"grad_norm": 1.9170950651168823,
|
| 931 |
+
"learning_rate": 2.06125e-06,
|
| 932 |
+
"loss": 0.1038,
|
| 933 |
+
"step": 3300
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 1.5217391304347827,
|
| 937 |
+
"grad_norm": 3.269129753112793,
|
| 938 |
+
"learning_rate": 2.076875e-06,
|
| 939 |
+
"loss": 0.0862,
|
| 940 |
+
"step": 3325
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 1.5331807780320366,
|
| 944 |
+
"grad_norm": 2.319660186767578,
|
| 945 |
+
"learning_rate": 2.0925e-06,
|
| 946 |
+
"loss": 0.1168,
|
| 947 |
+
"step": 3350
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 1.5446224256292906,
|
| 951 |
+
"grad_norm": 2.9521560668945312,
|
| 952 |
+
"learning_rate": 2.108125e-06,
|
| 953 |
+
"loss": 0.083,
|
| 954 |
+
"step": 3375
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 1.5560640732265445,
|
| 958 |
+
"grad_norm": 2.316248655319214,
|
| 959 |
+
"learning_rate": 2.12375e-06,
|
| 960 |
+
"loss": 0.1144,
|
| 961 |
+
"step": 3400
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.5675057208237986,
|
| 965 |
+
"grad_norm": 3.6519219875335693,
|
| 966 |
+
"learning_rate": 2.139375e-06,
|
| 967 |
+
"loss": 0.0833,
|
| 968 |
+
"step": 3425
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 1.5789473684210527,
|
| 972 |
+
"grad_norm": 1.8947266340255737,
|
| 973 |
+
"learning_rate": 2.155e-06,
|
| 974 |
+
"loss": 0.1178,
|
| 975 |
+
"step": 3450
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 1.5903890160183067,
|
| 979 |
+
"grad_norm": 5.7655158042907715,
|
| 980 |
+
"learning_rate": 2.170625e-06,
|
| 981 |
+
"loss": 0.0706,
|
| 982 |
+
"step": 3475
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 1.6018306636155606,
|
| 986 |
+
"grad_norm": 2.7136037349700928,
|
| 987 |
+
"learning_rate": 2.18625e-06,
|
| 988 |
+
"loss": 0.1206,
|
| 989 |
+
"step": 3500
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 1.6132723112128147,
|
| 993 |
+
"grad_norm": 2.735067844390869,
|
| 994 |
+
"learning_rate": 2.201875e-06,
|
| 995 |
+
"loss": 0.0799,
|
| 996 |
+
"step": 3525
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.6247139588100685,
|
| 1000 |
+
"grad_norm": 2.059816837310791,
|
| 1001 |
+
"learning_rate": 2.2175e-06,
|
| 1002 |
+
"loss": 0.094,
|
| 1003 |
+
"step": 3550
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 1.6361556064073226,
|
| 1007 |
+
"grad_norm": 2.8907694816589355,
|
| 1008 |
+
"learning_rate": 2.2331250000000003e-06,
|
| 1009 |
+
"loss": 0.0894,
|
| 1010 |
+
"step": 3575
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 1.6475972540045767,
|
| 1014 |
+
"grad_norm": 1.6282504796981812,
|
| 1015 |
+
"learning_rate": 2.2487500000000003e-06,
|
| 1016 |
+
"loss": 0.1034,
|
| 1017 |
+
"step": 3600
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 1.6590389016018308,
|
| 1021 |
+
"grad_norm": 2.726628303527832,
|
| 1022 |
+
"learning_rate": 2.2643750000000002e-06,
|
| 1023 |
+
"loss": 0.0789,
|
| 1024 |
+
"step": 3625
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 1.6704805491990846,
|
| 1028 |
+
"grad_norm": 2.466552257537842,
|
| 1029 |
+
"learning_rate": 2.28e-06,
|
| 1030 |
+
"loss": 0.0927,
|
| 1031 |
+
"step": 3650
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.6819221967963387,
|
| 1035 |
+
"grad_norm": 2.815573215484619,
|
| 1036 |
+
"learning_rate": 2.295625e-06,
|
| 1037 |
+
"loss": 0.086,
|
| 1038 |
+
"step": 3675
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 1.6933638443935926,
|
| 1042 |
+
"grad_norm": 2.731748104095459,
|
| 1043 |
+
"learning_rate": 2.31125e-06,
|
| 1044 |
+
"loss": 0.1175,
|
| 1045 |
+
"step": 3700
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 1.7048054919908466,
|
| 1049 |
+
"grad_norm": 3.349738836288452,
|
| 1050 |
+
"learning_rate": 2.326875e-06,
|
| 1051 |
+
"loss": 0.0763,
|
| 1052 |
+
"step": 3725
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.7162471395881007,
|
| 1056 |
+
"grad_norm": 2.117974042892456,
|
| 1057 |
+
"learning_rate": 2.3425000000000004e-06,
|
| 1058 |
+
"loss": 0.1058,
|
| 1059 |
+
"step": 3750
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 1.7276887871853548,
|
| 1063 |
+
"grad_norm": 3.7518179416656494,
|
| 1064 |
+
"learning_rate": 2.3581250000000004e-06,
|
| 1065 |
+
"loss": 0.0689,
|
| 1066 |
+
"step": 3775
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 1.7391304347826086,
|
| 1070 |
+
"grad_norm": 2.805143356323242,
|
| 1071 |
+
"learning_rate": 2.3737500000000004e-06,
|
| 1072 |
+
"loss": 0.1188,
|
| 1073 |
+
"step": 3800
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 1.7505720823798627,
|
| 1077 |
+
"grad_norm": 5.231779098510742,
|
| 1078 |
+
"learning_rate": 2.389375e-06,
|
| 1079 |
+
"loss": 0.0884,
|
| 1080 |
+
"step": 3825
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 1.7620137299771166,
|
| 1084 |
+
"grad_norm": 2.7431952953338623,
|
| 1085 |
+
"learning_rate": 2.405e-06,
|
| 1086 |
+
"loss": 0.1073,
|
| 1087 |
+
"step": 3850
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.7734553775743707,
|
| 1091 |
+
"grad_norm": 3.7958285808563232,
|
| 1092 |
+
"learning_rate": 2.420625e-06,
|
| 1093 |
+
"loss": 0.0754,
|
| 1094 |
+
"step": 3875
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 1.7848970251716247,
|
| 1098 |
+
"grad_norm": 3.2921016216278076,
|
| 1099 |
+
"learning_rate": 2.43625e-06,
|
| 1100 |
+
"loss": 0.1042,
|
| 1101 |
+
"step": 3900
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 1.7963386727688788,
|
| 1105 |
+
"grad_norm": 4.16481876373291,
|
| 1106 |
+
"learning_rate": 2.451875e-06,
|
| 1107 |
+
"loss": 0.0687,
|
| 1108 |
+
"step": 3925
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 1.8077803203661327,
|
| 1112 |
+
"grad_norm": 1.6314984560012817,
|
| 1113 |
+
"learning_rate": 2.4675e-06,
|
| 1114 |
+
"loss": 0.1265,
|
| 1115 |
+
"step": 3950
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 1.8192219679633868,
|
| 1119 |
+
"grad_norm": 3.358804702758789,
|
| 1120 |
+
"learning_rate": 2.483125e-06,
|
| 1121 |
+
"loss": 0.0683,
|
| 1122 |
+
"step": 3975
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 1.8306636155606406,
|
| 1126 |
+
"grad_norm": 2.8832662105560303,
|
| 1127 |
+
"learning_rate": 2.49875e-06,
|
| 1128 |
+
"loss": 0.0982,
|
| 1129 |
+
"step": 4000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 1.8421052631578947,
|
| 1133 |
+
"grad_norm": 2.2557880878448486,
|
| 1134 |
+
"learning_rate": 2.514375e-06,
|
| 1135 |
+
"loss": 0.0716,
|
| 1136 |
+
"step": 4025
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.8535469107551488,
|
| 1140 |
+
"grad_norm": 2.0373599529266357,
|
| 1141 |
+
"learning_rate": 2.53e-06,
|
| 1142 |
+
"loss": 0.1072,
|
| 1143 |
+
"step": 4050
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.8649885583524028,
|
| 1147 |
+
"grad_norm": 2.316436767578125,
|
| 1148 |
+
"learning_rate": 2.5456250000000003e-06,
|
| 1149 |
+
"loss": 0.0784,
|
| 1150 |
+
"step": 4075
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.8764302059496567,
|
| 1154 |
+
"grad_norm": 1.2671200037002563,
|
| 1155 |
+
"learning_rate": 2.5612500000000003e-06,
|
| 1156 |
+
"loss": 0.0996,
|
| 1157 |
+
"step": 4100
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.8878718535469108,
|
| 1161 |
+
"grad_norm": 3.7395412921905518,
|
| 1162 |
+
"learning_rate": 2.576875e-06,
|
| 1163 |
+
"loss": 0.0651,
|
| 1164 |
+
"step": 4125
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.8993135011441646,
|
| 1168 |
+
"grad_norm": 1.91667640209198,
|
| 1169 |
+
"learning_rate": 2.5925e-06,
|
| 1170 |
+
"loss": 0.1002,
|
| 1171 |
+
"step": 4150
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.9107551487414187,
|
| 1175 |
+
"grad_norm": 4.17740535736084,
|
| 1176 |
+
"learning_rate": 2.608125e-06,
|
| 1177 |
+
"loss": 0.0813,
|
| 1178 |
+
"step": 4175
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.9221967963386728,
|
| 1182 |
+
"grad_norm": 2.539367198944092,
|
| 1183 |
+
"learning_rate": 2.62375e-06,
|
| 1184 |
+
"loss": 0.1104,
|
| 1185 |
+
"step": 4200
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.9336384439359269,
|
| 1189 |
+
"grad_norm": 2.484417200088501,
|
| 1190 |
+
"learning_rate": 2.639375e-06,
|
| 1191 |
+
"loss": 0.0836,
|
| 1192 |
+
"step": 4225
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.9450800915331807,
|
| 1196 |
+
"grad_norm": 2.7852723598480225,
|
| 1197 |
+
"learning_rate": 2.655e-06,
|
| 1198 |
+
"loss": 0.1162,
|
| 1199 |
+
"step": 4250
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.9565217391304348,
|
| 1203 |
+
"grad_norm": 3.9952969551086426,
|
| 1204 |
+
"learning_rate": 2.6706250000000004e-06,
|
| 1205 |
+
"loss": 0.0832,
|
| 1206 |
+
"step": 4275
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.9679633867276887,
|
| 1210 |
+
"grad_norm": 2.994462490081787,
|
| 1211 |
+
"learning_rate": 2.6862500000000003e-06,
|
| 1212 |
+
"loss": 0.1299,
|
| 1213 |
+
"step": 4300
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.9794050343249427,
|
| 1217 |
+
"grad_norm": 2.588834047317505,
|
| 1218 |
+
"learning_rate": 2.7018750000000003e-06,
|
| 1219 |
+
"loss": 0.0821,
|
| 1220 |
+
"step": 4325
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.9908466819221968,
|
| 1224 |
+
"grad_norm": 1.583375334739685,
|
| 1225 |
+
"learning_rate": 2.7175000000000002e-06,
|
| 1226 |
+
"loss": 0.091,
|
| 1227 |
+
"step": 4350
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 2.002288329519451,
|
| 1231 |
+
"grad_norm": 6.827860355377197,
|
| 1232 |
+
"learning_rate": 2.733125e-06,
|
| 1233 |
+
"loss": 0.0953,
|
| 1234 |
+
"step": 4375
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 2.013729977116705,
|
| 1238 |
+
"grad_norm": 3.3871593475341797,
|
| 1239 |
+
"learning_rate": 2.74875e-06,
|
| 1240 |
+
"loss": 0.0436,
|
| 1241 |
+
"step": 4400
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 2.0251716247139586,
|
| 1245 |
+
"grad_norm": 3.1622588634490967,
|
| 1246 |
+
"learning_rate": 2.764375e-06,
|
| 1247 |
+
"loss": 0.0819,
|
| 1248 |
+
"step": 4425
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 2.0366132723112127,
|
| 1252 |
+
"grad_norm": 2.5629935264587402,
|
| 1253 |
+
"learning_rate": 2.78e-06,
|
| 1254 |
+
"loss": 0.0467,
|
| 1255 |
+
"step": 4450
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 2.0480549199084668,
|
| 1259 |
+
"grad_norm": 2.397113561630249,
|
| 1260 |
+
"learning_rate": 2.795625e-06,
|
| 1261 |
+
"loss": 0.0788,
|
| 1262 |
+
"step": 4475
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 2.059496567505721,
|
| 1266 |
+
"grad_norm": 3.9952454566955566,
|
| 1267 |
+
"learning_rate": 2.81125e-06,
|
| 1268 |
+
"loss": 0.0544,
|
| 1269 |
+
"step": 4500
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 2.070938215102975,
|
| 1273 |
+
"grad_norm": 2.8579368591308594,
|
| 1274 |
+
"learning_rate": 2.826875e-06,
|
| 1275 |
+
"loss": 0.0675,
|
| 1276 |
+
"step": 4525
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 2.082379862700229,
|
| 1280 |
+
"grad_norm": 2.419980525970459,
|
| 1281 |
+
"learning_rate": 2.8425e-06,
|
| 1282 |
+
"loss": 0.0479,
|
| 1283 |
+
"step": 4550
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 2.0938215102974826,
|
| 1287 |
+
"grad_norm": 2.3462846279144287,
|
| 1288 |
+
"learning_rate": 2.8575e-06,
|
| 1289 |
+
"loss": 0.0692,
|
| 1290 |
+
"step": 4575
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 2.1052631578947367,
|
| 1294 |
+
"grad_norm": 2.523151397705078,
|
| 1295 |
+
"learning_rate": 2.873125e-06,
|
| 1296 |
+
"loss": 0.0466,
|
| 1297 |
+
"step": 4600
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 2.116704805491991,
|
| 1301 |
+
"grad_norm": 1.8373658657073975,
|
| 1302 |
+
"learning_rate": 2.88875e-06,
|
| 1303 |
+
"loss": 0.0596,
|
| 1304 |
+
"step": 4625
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 2.128146453089245,
|
| 1308 |
+
"grad_norm": 2.7236149311065674,
|
| 1309 |
+
"learning_rate": 2.904375e-06,
|
| 1310 |
+
"loss": 0.0465,
|
| 1311 |
+
"step": 4650
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 2.139588100686499,
|
| 1315 |
+
"grad_norm": 3.013388156890869,
|
| 1316 |
+
"learning_rate": 2.92e-06,
|
| 1317 |
+
"loss": 0.0829,
|
| 1318 |
+
"step": 4675
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 2.151029748283753,
|
| 1322 |
+
"grad_norm": 1.6054365634918213,
|
| 1323 |
+
"learning_rate": 2.9356250000000004e-06,
|
| 1324 |
+
"loss": 0.0447,
|
| 1325 |
+
"step": 4700
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 2.1624713958810067,
|
| 1329 |
+
"grad_norm": 1.8882802724838257,
|
| 1330 |
+
"learning_rate": 2.9512500000000003e-06,
|
| 1331 |
+
"loss": 0.0634,
|
| 1332 |
+
"step": 4725
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 2.1739130434782608,
|
| 1336 |
+
"grad_norm": 4.886575698852539,
|
| 1337 |
+
"learning_rate": 2.9668750000000003e-06,
|
| 1338 |
+
"loss": 0.0543,
|
| 1339 |
+
"step": 4750
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 2.185354691075515,
|
| 1343 |
+
"grad_norm": 1.4552139043807983,
|
| 1344 |
+
"learning_rate": 2.9825000000000002e-06,
|
| 1345 |
+
"loss": 0.0697,
|
| 1346 |
+
"step": 4775
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 2.196796338672769,
|
| 1350 |
+
"grad_norm": 2.0052905082702637,
|
| 1351 |
+
"learning_rate": 2.998125e-06,
|
| 1352 |
+
"loss": 0.0562,
|
| 1353 |
+
"step": 4800
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 2.208237986270023,
|
| 1357 |
+
"grad_norm": 2.3724350929260254,
|
| 1358 |
+
"learning_rate": 3.01375e-06,
|
| 1359 |
+
"loss": 0.0718,
|
| 1360 |
+
"step": 4825
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 2.219679633867277,
|
| 1364 |
+
"grad_norm": 1.8292073011398315,
|
| 1365 |
+
"learning_rate": 3.029375e-06,
|
| 1366 |
+
"loss": 0.055,
|
| 1367 |
+
"step": 4850
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 2.2311212814645307,
|
| 1371 |
+
"grad_norm": 1.8585909605026245,
|
| 1372 |
+
"learning_rate": 3.0450000000000005e-06,
|
| 1373 |
+
"loss": 0.071,
|
| 1374 |
+
"step": 4875
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 2.242562929061785,
|
| 1378 |
+
"grad_norm": 3.8988568782806396,
|
| 1379 |
+
"learning_rate": 3.0606250000000004e-06,
|
| 1380 |
+
"loss": 0.0464,
|
| 1381 |
+
"step": 4900
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 2.254004576659039,
|
| 1385 |
+
"grad_norm": 2.759815216064453,
|
| 1386 |
+
"learning_rate": 3.0762500000000004e-06,
|
| 1387 |
+
"loss": 0.0684,
|
| 1388 |
+
"step": 4925
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 2.265446224256293,
|
| 1392 |
+
"grad_norm": 1.6180821657180786,
|
| 1393 |
+
"learning_rate": 3.091875e-06,
|
| 1394 |
+
"loss": 0.0684,
|
| 1395 |
+
"step": 4950
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 2.276887871853547,
|
| 1399 |
+
"grad_norm": 1.9958300590515137,
|
| 1400 |
+
"learning_rate": 3.1075e-06,
|
| 1401 |
+
"loss": 0.0676,
|
| 1402 |
+
"step": 4975
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 2.288329519450801,
|
| 1406 |
+
"grad_norm": 4.299381732940674,
|
| 1407 |
+
"learning_rate": 3.123125e-06,
|
| 1408 |
+
"loss": 0.0559,
|
| 1409 |
+
"step": 5000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 2.288329519450801,
|
| 1413 |
+
"eval_loss": 0.10962941497564316,
|
| 1414 |
+
"eval_runtime": 8536.2165,
|
| 1415 |
+
"eval_samples_per_second": 1.115,
|
| 1416 |
+
"eval_steps_per_second": 0.14,
|
| 1417 |
+
"eval_wer": 0.07299865617200998,
|
| 1418 |
+
"step": 5000
|
| 1419 |
+
}
|
| 1420 |
+
],
|
| 1421 |
+
"logging_steps": 25,
|
| 1422 |
+
"max_steps": 50000,
|
| 1423 |
+
"num_input_tokens_seen": 0,
|
| 1424 |
+
"num_train_epochs": 23,
|
| 1425 |
+
"save_steps": 5000,
|
| 1426 |
+
"stateful_callbacks": {
|
| 1427 |
+
"TrainerControl": {
|
| 1428 |
+
"args": {
|
| 1429 |
+
"should_epoch_stop": false,
|
| 1430 |
+
"should_evaluate": false,
|
| 1431 |
+
"should_log": false,
|
| 1432 |
+
"should_save": true,
|
| 1433 |
+
"should_training_stop": false
|
| 1434 |
+
},
|
| 1435 |
+
"attributes": {}
|
| 1436 |
+
}
|
| 1437 |
+
},
|
| 1438 |
+
"total_flos": 2.7177268453834752e+20,
|
| 1439 |
+
"train_batch_size": 8,
|
| 1440 |
+
"trial_name": null,
|
| 1441 |
+
"trial_params": null
|
| 1442 |
+
}
|