Upload checkpoint-129030
Browse files- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1919 -0
- training_args.bin +3 -0
optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:4b930d5bcbe3df6d5ee9e15433eeacb1cff6f66fe5522d292d659eb4c38da3c0
|
3 |
+
size 866895354
|
rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:aa280e2e6dbd3a37c8dc2dc5fe9c7782cf7832367d7c106285ca3d9da6e271f7
|
3 |
+
size 14244
|
scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:1fe476619a2e0e15fce441e7fcb5a118b511efee40d7b844292db4a9762411b5
|
3 |
+
size 1064
|
trainer_state.json
ADDED
@@ -0,0 +1,1919 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": null,
|
3 |
+
"best_model_checkpoint": null,
|
4 |
+
"epoch": 10.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 129030,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.038750678136867396,
|
13 |
+
"grad_norm": 7.044970512390137,
|
14 |
+
"learning_rate": 1.9922498643726265e-05,
|
15 |
+
"loss": 5.0771,
|
16 |
+
"step": 500
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.07750135627373479,
|
20 |
+
"grad_norm": 6.129168510437012,
|
21 |
+
"learning_rate": 1.9844997287452532e-05,
|
22 |
+
"loss": 4.5935,
|
23 |
+
"step": 1000
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.11625203441060218,
|
27 |
+
"grad_norm": 6.968145847320557,
|
28 |
+
"learning_rate": 1.9767495931178796e-05,
|
29 |
+
"loss": 4.3872,
|
30 |
+
"step": 1500
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.15500271254746958,
|
34 |
+
"grad_norm": 6.270899295806885,
|
35 |
+
"learning_rate": 1.9689994574905063e-05,
|
36 |
+
"loss": 4.2507,
|
37 |
+
"step": 2000
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.19375339068433697,
|
41 |
+
"grad_norm": 6.4452104568481445,
|
42 |
+
"learning_rate": 1.9612493218631326e-05,
|
43 |
+
"loss": 4.135,
|
44 |
+
"step": 2500
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.23250406882120436,
|
48 |
+
"grad_norm": 6.564201354980469,
|
49 |
+
"learning_rate": 1.9534991862357593e-05,
|
50 |
+
"loss": 4.0704,
|
51 |
+
"step": 3000
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.2712547469580718,
|
55 |
+
"grad_norm": 6.716593265533447,
|
56 |
+
"learning_rate": 1.9457490506083857e-05,
|
57 |
+
"loss": 3.9814,
|
58 |
+
"step": 3500
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.31000542509493917,
|
62 |
+
"grad_norm": 6.578341484069824,
|
63 |
+
"learning_rate": 1.9379989149810124e-05,
|
64 |
+
"loss": 3.9078,
|
65 |
+
"step": 4000
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 0.34875610323180656,
|
69 |
+
"grad_norm": 6.6163482666015625,
|
70 |
+
"learning_rate": 1.9302487793536387e-05,
|
71 |
+
"loss": 3.858,
|
72 |
+
"step": 4500
|
73 |
+
},
|
74 |
+
{
|
75 |
+
"epoch": 0.38750678136867395,
|
76 |
+
"grad_norm": 5.90346097946167,
|
77 |
+
"learning_rate": 1.9224986437262654e-05,
|
78 |
+
"loss": 3.8036,
|
79 |
+
"step": 5000
|
80 |
+
},
|
81 |
+
{
|
82 |
+
"epoch": 0.42625745950554134,
|
83 |
+
"grad_norm": 5.8487467765808105,
|
84 |
+
"learning_rate": 1.9147485080988918e-05,
|
85 |
+
"loss": 3.7518,
|
86 |
+
"step": 5500
|
87 |
+
},
|
88 |
+
{
|
89 |
+
"epoch": 0.4650081376424087,
|
90 |
+
"grad_norm": 6.573217391967773,
|
91 |
+
"learning_rate": 1.9069983724715185e-05,
|
92 |
+
"loss": 3.7164,
|
93 |
+
"step": 6000
|
94 |
+
},
|
95 |
+
{
|
96 |
+
"epoch": 0.5037588157792762,
|
97 |
+
"grad_norm": 5.970231056213379,
|
98 |
+
"learning_rate": 1.899248236844145e-05,
|
99 |
+
"loss": 3.673,
|
100 |
+
"step": 6500
|
101 |
+
},
|
102 |
+
{
|
103 |
+
"epoch": 0.5425094939161436,
|
104 |
+
"grad_norm": 6.683649063110352,
|
105 |
+
"learning_rate": 1.8914981012167715e-05,
|
106 |
+
"loss": 3.6405,
|
107 |
+
"step": 7000
|
108 |
+
},
|
109 |
+
{
|
110 |
+
"epoch": 0.581260172053011,
|
111 |
+
"grad_norm": 6.538488864898682,
|
112 |
+
"learning_rate": 1.883747965589398e-05,
|
113 |
+
"loss": 3.6133,
|
114 |
+
"step": 7500
|
115 |
+
},
|
116 |
+
{
|
117 |
+
"epoch": 0.6200108501898783,
|
118 |
+
"grad_norm": 6.528162479400635,
|
119 |
+
"learning_rate": 1.8759978299620246e-05,
|
120 |
+
"loss": 3.5754,
|
121 |
+
"step": 8000
|
122 |
+
},
|
123 |
+
{
|
124 |
+
"epoch": 0.6587615283267457,
|
125 |
+
"grad_norm": 6.43408203125,
|
126 |
+
"learning_rate": 1.868247694334651e-05,
|
127 |
+
"loss": 3.5548,
|
128 |
+
"step": 8500
|
129 |
+
},
|
130 |
+
{
|
131 |
+
"epoch": 0.6975122064636131,
|
132 |
+
"grad_norm": 6.131889820098877,
|
133 |
+
"learning_rate": 1.8604975587072776e-05,
|
134 |
+
"loss": 3.5228,
|
135 |
+
"step": 9000
|
136 |
+
},
|
137 |
+
{
|
138 |
+
"epoch": 0.7362628846004805,
|
139 |
+
"grad_norm": 6.320891857147217,
|
140 |
+
"learning_rate": 1.852747423079904e-05,
|
141 |
+
"loss": 3.4999,
|
142 |
+
"step": 9500
|
143 |
+
},
|
144 |
+
{
|
145 |
+
"epoch": 0.7750135627373479,
|
146 |
+
"grad_norm": 6.105418682098389,
|
147 |
+
"learning_rate": 1.8449972874525307e-05,
|
148 |
+
"loss": 3.5071,
|
149 |
+
"step": 10000
|
150 |
+
},
|
151 |
+
{
|
152 |
+
"epoch": 0.8137642408742153,
|
153 |
+
"grad_norm": 6.774458885192871,
|
154 |
+
"learning_rate": 1.837247151825157e-05,
|
155 |
+
"loss": 3.4616,
|
156 |
+
"step": 10500
|
157 |
+
},
|
158 |
+
{
|
159 |
+
"epoch": 0.8525149190110827,
|
160 |
+
"grad_norm": 6.263659477233887,
|
161 |
+
"learning_rate": 1.8294970161977838e-05,
|
162 |
+
"loss": 3.4499,
|
163 |
+
"step": 11000
|
164 |
+
},
|
165 |
+
{
|
166 |
+
"epoch": 0.8912655971479501,
|
167 |
+
"grad_norm": 6.58251428604126,
|
168 |
+
"learning_rate": 1.82174688057041e-05,
|
169 |
+
"loss": 3.4157,
|
170 |
+
"step": 11500
|
171 |
+
},
|
172 |
+
{
|
173 |
+
"epoch": 0.9300162752848175,
|
174 |
+
"grad_norm": 6.030143737792969,
|
175 |
+
"learning_rate": 1.8139967449430368e-05,
|
176 |
+
"loss": 3.3842,
|
177 |
+
"step": 12000
|
178 |
+
},
|
179 |
+
{
|
180 |
+
"epoch": 0.9687669534216848,
|
181 |
+
"grad_norm": 6.361506462097168,
|
182 |
+
"learning_rate": 1.806246609315663e-05,
|
183 |
+
"loss": 3.3707,
|
184 |
+
"step": 12500
|
185 |
+
},
|
186 |
+
{
|
187 |
+
"epoch": 1.0,
|
188 |
+
"eval_loss": 3.2508351802825928,
|
189 |
+
"eval_runtime": 267.1986,
|
190 |
+
"eval_samples_per_second": 772.702,
|
191 |
+
"eval_steps_per_second": 12.077,
|
192 |
+
"step": 12903
|
193 |
+
},
|
194 |
+
{
|
195 |
+
"epoch": 1.0075176315585523,
|
196 |
+
"grad_norm": 6.418643474578857,
|
197 |
+
"learning_rate": 1.79849647368829e-05,
|
198 |
+
"loss": 3.3673,
|
199 |
+
"step": 13000
|
200 |
+
},
|
201 |
+
{
|
202 |
+
"epoch": 1.0462683096954197,
|
203 |
+
"grad_norm": 6.310774326324463,
|
204 |
+
"learning_rate": 1.7907463380609162e-05,
|
205 |
+
"loss": 3.3276,
|
206 |
+
"step": 13500
|
207 |
+
},
|
208 |
+
{
|
209 |
+
"epoch": 1.0850189878322871,
|
210 |
+
"grad_norm": 6.517366409301758,
|
211 |
+
"learning_rate": 1.782996202433543e-05,
|
212 |
+
"loss": 3.3288,
|
213 |
+
"step": 14000
|
214 |
+
},
|
215 |
+
{
|
216 |
+
"epoch": 1.1237696659691545,
|
217 |
+
"grad_norm": 6.407958984375,
|
218 |
+
"learning_rate": 1.7752460668061693e-05,
|
219 |
+
"loss": 3.3003,
|
220 |
+
"step": 14500
|
221 |
+
},
|
222 |
+
{
|
223 |
+
"epoch": 1.162520344106022,
|
224 |
+
"grad_norm": 6.145129203796387,
|
225 |
+
"learning_rate": 1.767495931178796e-05,
|
226 |
+
"loss": 3.2694,
|
227 |
+
"step": 15000
|
228 |
+
},
|
229 |
+
{
|
230 |
+
"epoch": 1.2012710222428893,
|
231 |
+
"grad_norm": 6.586604118347168,
|
232 |
+
"learning_rate": 1.7597457955514223e-05,
|
233 |
+
"loss": 3.2627,
|
234 |
+
"step": 15500
|
235 |
+
},
|
236 |
+
{
|
237 |
+
"epoch": 1.2400217003797567,
|
238 |
+
"grad_norm": 6.122056007385254,
|
239 |
+
"learning_rate": 1.751995659924049e-05,
|
240 |
+
"loss": 3.2631,
|
241 |
+
"step": 16000
|
242 |
+
},
|
243 |
+
{
|
244 |
+
"epoch": 1.278772378516624,
|
245 |
+
"grad_norm": 6.545727252960205,
|
246 |
+
"learning_rate": 1.7442455242966754e-05,
|
247 |
+
"loss": 3.2324,
|
248 |
+
"step": 16500
|
249 |
+
},
|
250 |
+
{
|
251 |
+
"epoch": 1.3175230566534915,
|
252 |
+
"grad_norm": 6.427816390991211,
|
253 |
+
"learning_rate": 1.7364953886693017e-05,
|
254 |
+
"loss": 3.227,
|
255 |
+
"step": 17000
|
256 |
+
},
|
257 |
+
{
|
258 |
+
"epoch": 1.3562737347903588,
|
259 |
+
"grad_norm": 6.253689765930176,
|
260 |
+
"learning_rate": 1.7287452530419284e-05,
|
261 |
+
"loss": 3.2099,
|
262 |
+
"step": 17500
|
263 |
+
},
|
264 |
+
{
|
265 |
+
"epoch": 1.3950244129272262,
|
266 |
+
"grad_norm": 6.5702080726623535,
|
267 |
+
"learning_rate": 1.7209951174145548e-05,
|
268 |
+
"loss": 3.2102,
|
269 |
+
"step": 18000
|
270 |
+
},
|
271 |
+
{
|
272 |
+
"epoch": 1.4337750910640936,
|
273 |
+
"grad_norm": 6.4822564125061035,
|
274 |
+
"learning_rate": 1.7132449817871815e-05,
|
275 |
+
"loss": 3.1935,
|
276 |
+
"step": 18500
|
277 |
+
},
|
278 |
+
{
|
279 |
+
"epoch": 1.472525769200961,
|
280 |
+
"grad_norm": 6.524315357208252,
|
281 |
+
"learning_rate": 1.705494846159808e-05,
|
282 |
+
"loss": 3.1955,
|
283 |
+
"step": 19000
|
284 |
+
},
|
285 |
+
{
|
286 |
+
"epoch": 1.5112764473378284,
|
287 |
+
"grad_norm": 6.302344799041748,
|
288 |
+
"learning_rate": 1.6977447105324345e-05,
|
289 |
+
"loss": 3.1726,
|
290 |
+
"step": 19500
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 1.5500271254746958,
|
294 |
+
"grad_norm": 5.837028503417969,
|
295 |
+
"learning_rate": 1.689994574905061e-05,
|
296 |
+
"loss": 3.1277,
|
297 |
+
"step": 20000
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 1.5887778036115632,
|
301 |
+
"grad_norm": 6.489377975463867,
|
302 |
+
"learning_rate": 1.6822444392776876e-05,
|
303 |
+
"loss": 3.1414,
|
304 |
+
"step": 20500
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 1.6275284817484306,
|
308 |
+
"grad_norm": 6.543872833251953,
|
309 |
+
"learning_rate": 1.674494303650314e-05,
|
310 |
+
"loss": 3.104,
|
311 |
+
"step": 21000
|
312 |
+
},
|
313 |
+
{
|
314 |
+
"epoch": 1.666279159885298,
|
315 |
+
"grad_norm": 6.05628776550293,
|
316 |
+
"learning_rate": 1.6667441680229406e-05,
|
317 |
+
"loss": 3.1459,
|
318 |
+
"step": 21500
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"epoch": 1.7050298380221653,
|
322 |
+
"grad_norm": 6.027078151702881,
|
323 |
+
"learning_rate": 1.658994032395567e-05,
|
324 |
+
"loss": 3.0963,
|
325 |
+
"step": 22000
|
326 |
+
},
|
327 |
+
{
|
328 |
+
"epoch": 1.7437805161590327,
|
329 |
+
"grad_norm": 6.577582359313965,
|
330 |
+
"learning_rate": 1.6512438967681937e-05,
|
331 |
+
"loss": 3.118,
|
332 |
+
"step": 22500
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 1.7825311942959001,
|
336 |
+
"grad_norm": 5.9164204597473145,
|
337 |
+
"learning_rate": 1.64349376114082e-05,
|
338 |
+
"loss": 3.0928,
|
339 |
+
"step": 23000
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 1.8212818724327677,
|
343 |
+
"grad_norm": 6.155348300933838,
|
344 |
+
"learning_rate": 1.6357436255134468e-05,
|
345 |
+
"loss": 3.0885,
|
346 |
+
"step": 23500
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"epoch": 1.8600325505696351,
|
350 |
+
"grad_norm": 6.302849769592285,
|
351 |
+
"learning_rate": 1.627993489886073e-05,
|
352 |
+
"loss": 3.0741,
|
353 |
+
"step": 24000
|
354 |
+
},
|
355 |
+
{
|
356 |
+
"epoch": 1.8987832287065025,
|
357 |
+
"grad_norm": 6.140907287597656,
|
358 |
+
"learning_rate": 1.6202433542586998e-05,
|
359 |
+
"loss": 3.0633,
|
360 |
+
"step": 24500
|
361 |
+
},
|
362 |
+
{
|
363 |
+
"epoch": 1.93753390684337,
|
364 |
+
"grad_norm": 5.85639762878418,
|
365 |
+
"learning_rate": 1.612493218631326e-05,
|
366 |
+
"loss": 3.0401,
|
367 |
+
"step": 25000
|
368 |
+
},
|
369 |
+
{
|
370 |
+
"epoch": 1.9762845849802373,
|
371 |
+
"grad_norm": 6.558920383453369,
|
372 |
+
"learning_rate": 1.604743083003953e-05,
|
373 |
+
"loss": 3.05,
|
374 |
+
"step": 25500
|
375 |
+
},
|
376 |
+
{
|
377 |
+
"epoch": 2.0,
|
378 |
+
"eval_loss": 2.9521713256835938,
|
379 |
+
"eval_runtime": 258.4886,
|
380 |
+
"eval_samples_per_second": 798.739,
|
381 |
+
"eval_steps_per_second": 12.484,
|
382 |
+
"step": 25806
|
383 |
+
},
|
384 |
+
{
|
385 |
+
"epoch": 2.0150352631171047,
|
386 |
+
"grad_norm": 6.003655433654785,
|
387 |
+
"learning_rate": 1.5969929473765792e-05,
|
388 |
+
"loss": 3.0292,
|
389 |
+
"step": 26000
|
390 |
+
},
|
391 |
+
{
|
392 |
+
"epoch": 2.053785941253972,
|
393 |
+
"grad_norm": 6.43280029296875,
|
394 |
+
"learning_rate": 1.589242811749206e-05,
|
395 |
+
"loss": 3.0205,
|
396 |
+
"step": 26500
|
397 |
+
},
|
398 |
+
{
|
399 |
+
"epoch": 2.0925366193908395,
|
400 |
+
"grad_norm": 6.051511287689209,
|
401 |
+
"learning_rate": 1.5814926761218323e-05,
|
402 |
+
"loss": 3.0152,
|
403 |
+
"step": 27000
|
404 |
+
},
|
405 |
+
{
|
406 |
+
"epoch": 2.131287297527707,
|
407 |
+
"grad_norm": 7.381418704986572,
|
408 |
+
"learning_rate": 1.573742540494459e-05,
|
409 |
+
"loss": 3.0067,
|
410 |
+
"step": 27500
|
411 |
+
},
|
412 |
+
{
|
413 |
+
"epoch": 2.1700379756645742,
|
414 |
+
"grad_norm": 6.032004356384277,
|
415 |
+
"learning_rate": 1.5659924048670853e-05,
|
416 |
+
"loss": 2.9821,
|
417 |
+
"step": 28000
|
418 |
+
},
|
419 |
+
{
|
420 |
+
"epoch": 2.2087886538014416,
|
421 |
+
"grad_norm": 6.481622695922852,
|
422 |
+
"learning_rate": 1.558242269239712e-05,
|
423 |
+
"loss": 2.9824,
|
424 |
+
"step": 28500
|
425 |
+
},
|
426 |
+
{
|
427 |
+
"epoch": 2.247539331938309,
|
428 |
+
"grad_norm": 5.934979438781738,
|
429 |
+
"learning_rate": 1.5504921336123384e-05,
|
430 |
+
"loss": 2.9708,
|
431 |
+
"step": 29000
|
432 |
+
},
|
433 |
+
{
|
434 |
+
"epoch": 2.2862900100751764,
|
435 |
+
"grad_norm": 7.498392581939697,
|
436 |
+
"learning_rate": 1.542741997984965e-05,
|
437 |
+
"loss": 2.9836,
|
438 |
+
"step": 29500
|
439 |
+
},
|
440 |
+
{
|
441 |
+
"epoch": 2.325040688212044,
|
442 |
+
"grad_norm": 6.350077152252197,
|
443 |
+
"learning_rate": 1.5349918623575914e-05,
|
444 |
+
"loss": 2.9608,
|
445 |
+
"step": 30000
|
446 |
+
},
|
447 |
+
{
|
448 |
+
"epoch": 2.363791366348911,
|
449 |
+
"grad_norm": 5.6795783042907715,
|
450 |
+
"learning_rate": 1.527241726730218e-05,
|
451 |
+
"loss": 2.9551,
|
452 |
+
"step": 30500
|
453 |
+
},
|
454 |
+
{
|
455 |
+
"epoch": 2.4025420444857786,
|
456 |
+
"grad_norm": 6.395376682281494,
|
457 |
+
"learning_rate": 1.5194915911028445e-05,
|
458 |
+
"loss": 2.9551,
|
459 |
+
"step": 31000
|
460 |
+
},
|
461 |
+
{
|
462 |
+
"epoch": 2.441292722622646,
|
463 |
+
"grad_norm": 6.238061904907227,
|
464 |
+
"learning_rate": 1.511741455475471e-05,
|
465 |
+
"loss": 2.9527,
|
466 |
+
"step": 31500
|
467 |
+
},
|
468 |
+
{
|
469 |
+
"epoch": 2.4800434007595134,
|
470 |
+
"grad_norm": 6.641284465789795,
|
471 |
+
"learning_rate": 1.5039913198480975e-05,
|
472 |
+
"loss": 2.9444,
|
473 |
+
"step": 32000
|
474 |
+
},
|
475 |
+
{
|
476 |
+
"epoch": 2.5187940788963807,
|
477 |
+
"grad_norm": 6.30321741104126,
|
478 |
+
"learning_rate": 1.496241184220724e-05,
|
479 |
+
"loss": 2.9346,
|
480 |
+
"step": 32500
|
481 |
+
},
|
482 |
+
{
|
483 |
+
"epoch": 2.557544757033248,
|
484 |
+
"grad_norm": 8.681157112121582,
|
485 |
+
"learning_rate": 1.4884910485933506e-05,
|
486 |
+
"loss": 2.9204,
|
487 |
+
"step": 33000
|
488 |
+
},
|
489 |
+
{
|
490 |
+
"epoch": 2.5962954351701155,
|
491 |
+
"grad_norm": 6.423407077789307,
|
492 |
+
"learning_rate": 1.4807409129659771e-05,
|
493 |
+
"loss": 2.9066,
|
494 |
+
"step": 33500
|
495 |
+
},
|
496 |
+
{
|
497 |
+
"epoch": 2.635046113306983,
|
498 |
+
"grad_norm": 6.697604179382324,
|
499 |
+
"learning_rate": 1.4729907773386036e-05,
|
500 |
+
"loss": 2.9079,
|
501 |
+
"step": 34000
|
502 |
+
},
|
503 |
+
{
|
504 |
+
"epoch": 2.6737967914438503,
|
505 |
+
"grad_norm": 6.646244049072266,
|
506 |
+
"learning_rate": 1.4652406417112302e-05,
|
507 |
+
"loss": 2.9148,
|
508 |
+
"step": 34500
|
509 |
+
},
|
510 |
+
{
|
511 |
+
"epoch": 2.7125474695807177,
|
512 |
+
"grad_norm": 6.80411958694458,
|
513 |
+
"learning_rate": 1.4574905060838567e-05,
|
514 |
+
"loss": 2.9005,
|
515 |
+
"step": 35000
|
516 |
+
},
|
517 |
+
{
|
518 |
+
"epoch": 2.751298147717585,
|
519 |
+
"grad_norm": 6.345988750457764,
|
520 |
+
"learning_rate": 1.4497403704564832e-05,
|
521 |
+
"loss": 2.888,
|
522 |
+
"step": 35500
|
523 |
+
},
|
524 |
+
{
|
525 |
+
"epoch": 2.7900488258544525,
|
526 |
+
"grad_norm": 5.965686798095703,
|
527 |
+
"learning_rate": 1.4419902348291098e-05,
|
528 |
+
"loss": 2.8845,
|
529 |
+
"step": 36000
|
530 |
+
},
|
531 |
+
{
|
532 |
+
"epoch": 2.82879950399132,
|
533 |
+
"grad_norm": 6.068357944488525,
|
534 |
+
"learning_rate": 1.4342400992017363e-05,
|
535 |
+
"loss": 2.8835,
|
536 |
+
"step": 36500
|
537 |
+
},
|
538 |
+
{
|
539 |
+
"epoch": 2.8675501821281872,
|
540 |
+
"grad_norm": 5.874370098114014,
|
541 |
+
"learning_rate": 1.4264899635743628e-05,
|
542 |
+
"loss": 2.8904,
|
543 |
+
"step": 37000
|
544 |
+
},
|
545 |
+
{
|
546 |
+
"epoch": 2.9063008602650546,
|
547 |
+
"grad_norm": 6.0566935539245605,
|
548 |
+
"learning_rate": 1.4187398279469893e-05,
|
549 |
+
"loss": 2.8823,
|
550 |
+
"step": 37500
|
551 |
+
},
|
552 |
+
{
|
553 |
+
"epoch": 2.945051538401922,
|
554 |
+
"grad_norm": 6.21787691116333,
|
555 |
+
"learning_rate": 1.4109896923196159e-05,
|
556 |
+
"loss": 2.867,
|
557 |
+
"step": 38000
|
558 |
+
},
|
559 |
+
{
|
560 |
+
"epoch": 2.9838022165387894,
|
561 |
+
"grad_norm": 6.055897235870361,
|
562 |
+
"learning_rate": 1.4032395566922424e-05,
|
563 |
+
"loss": 2.867,
|
564 |
+
"step": 38500
|
565 |
+
},
|
566 |
+
{
|
567 |
+
"epoch": 3.0,
|
568 |
+
"eval_loss": 2.775702953338623,
|
569 |
+
"eval_runtime": 259.0101,
|
570 |
+
"eval_samples_per_second": 797.131,
|
571 |
+
"eval_steps_per_second": 12.459,
|
572 |
+
"step": 38709
|
573 |
+
},
|
574 |
+
{
|
575 |
+
"epoch": 3.022552894675657,
|
576 |
+
"grad_norm": 5.503760814666748,
|
577 |
+
"learning_rate": 1.3954894210648689e-05,
|
578 |
+
"loss": 2.8428,
|
579 |
+
"step": 39000
|
580 |
+
},
|
581 |
+
{
|
582 |
+
"epoch": 3.061303572812524,
|
583 |
+
"grad_norm": 6.250561714172363,
|
584 |
+
"learning_rate": 1.3877392854374954e-05,
|
585 |
+
"loss": 2.842,
|
586 |
+
"step": 39500
|
587 |
+
},
|
588 |
+
{
|
589 |
+
"epoch": 3.1000542509493916,
|
590 |
+
"grad_norm": 6.394408226013184,
|
591 |
+
"learning_rate": 1.379989149810122e-05,
|
592 |
+
"loss": 2.8368,
|
593 |
+
"step": 40000
|
594 |
+
},
|
595 |
+
{
|
596 |
+
"epoch": 3.138804929086259,
|
597 |
+
"grad_norm": 5.7096428871154785,
|
598 |
+
"learning_rate": 1.3722390141827483e-05,
|
599 |
+
"loss": 2.8253,
|
600 |
+
"step": 40500
|
601 |
+
},
|
602 |
+
{
|
603 |
+
"epoch": 3.1775556072231264,
|
604 |
+
"grad_norm": 6.807374000549316,
|
605 |
+
"learning_rate": 1.3644888785553749e-05,
|
606 |
+
"loss": 2.821,
|
607 |
+
"step": 41000
|
608 |
+
},
|
609 |
+
{
|
610 |
+
"epoch": 3.2163062853599937,
|
611 |
+
"grad_norm": 6.367000102996826,
|
612 |
+
"learning_rate": 1.3567387429280014e-05,
|
613 |
+
"loss": 2.8302,
|
614 |
+
"step": 41500
|
615 |
+
},
|
616 |
+
{
|
617 |
+
"epoch": 3.255056963496861,
|
618 |
+
"grad_norm": 6.30033540725708,
|
619 |
+
"learning_rate": 1.3489886073006279e-05,
|
620 |
+
"loss": 2.8191,
|
621 |
+
"step": 42000
|
622 |
+
},
|
623 |
+
{
|
624 |
+
"epoch": 3.2938076416337285,
|
625 |
+
"grad_norm": 7.257653713226318,
|
626 |
+
"learning_rate": 1.3412384716732544e-05,
|
627 |
+
"loss": 2.8196,
|
628 |
+
"step": 42500
|
629 |
+
},
|
630 |
+
{
|
631 |
+
"epoch": 3.332558319770596,
|
632 |
+
"grad_norm": 7.1162109375,
|
633 |
+
"learning_rate": 1.333488336045881e-05,
|
634 |
+
"loss": 2.817,
|
635 |
+
"step": 43000
|
636 |
+
},
|
637 |
+
{
|
638 |
+
"epoch": 3.3713089979074633,
|
639 |
+
"grad_norm": 6.336881160736084,
|
640 |
+
"learning_rate": 1.3257382004185075e-05,
|
641 |
+
"loss": 2.8064,
|
642 |
+
"step": 43500
|
643 |
+
},
|
644 |
+
{
|
645 |
+
"epoch": 3.4100596760443307,
|
646 |
+
"grad_norm": 6.641462326049805,
|
647 |
+
"learning_rate": 1.317988064791134e-05,
|
648 |
+
"loss": 2.8035,
|
649 |
+
"step": 44000
|
650 |
+
},
|
651 |
+
{
|
652 |
+
"epoch": 3.448810354181198,
|
653 |
+
"grad_norm": 6.033754348754883,
|
654 |
+
"learning_rate": 1.3102379291637605e-05,
|
655 |
+
"loss": 2.7976,
|
656 |
+
"step": 44500
|
657 |
+
},
|
658 |
+
{
|
659 |
+
"epoch": 3.4875610323180655,
|
660 |
+
"grad_norm": 6.544773101806641,
|
661 |
+
"learning_rate": 1.302487793536387e-05,
|
662 |
+
"loss": 2.8048,
|
663 |
+
"step": 45000
|
664 |
+
},
|
665 |
+
{
|
666 |
+
"epoch": 3.526311710454933,
|
667 |
+
"grad_norm": 6.382020950317383,
|
668 |
+
"learning_rate": 1.2947376579090136e-05,
|
669 |
+
"loss": 2.7982,
|
670 |
+
"step": 45500
|
671 |
+
},
|
672 |
+
{
|
673 |
+
"epoch": 3.5650623885918002,
|
674 |
+
"grad_norm": 6.194632053375244,
|
675 |
+
"learning_rate": 1.2869875222816401e-05,
|
676 |
+
"loss": 2.7749,
|
677 |
+
"step": 46000
|
678 |
+
},
|
679 |
+
{
|
680 |
+
"epoch": 3.6038130667286676,
|
681 |
+
"grad_norm": 6.429641246795654,
|
682 |
+
"learning_rate": 1.2792373866542665e-05,
|
683 |
+
"loss": 2.7853,
|
684 |
+
"step": 46500
|
685 |
+
},
|
686 |
+
{
|
687 |
+
"epoch": 3.642563744865535,
|
688 |
+
"grad_norm": 6.209822177886963,
|
689 |
+
"learning_rate": 1.271487251026893e-05,
|
690 |
+
"loss": 2.7841,
|
691 |
+
"step": 47000
|
692 |
+
},
|
693 |
+
{
|
694 |
+
"epoch": 3.6813144230024024,
|
695 |
+
"grad_norm": 6.935910701751709,
|
696 |
+
"learning_rate": 1.2637371153995195e-05,
|
697 |
+
"loss": 2.7681,
|
698 |
+
"step": 47500
|
699 |
+
},
|
700 |
+
{
|
701 |
+
"epoch": 3.72006510113927,
|
702 |
+
"grad_norm": 7.021639347076416,
|
703 |
+
"learning_rate": 1.255986979772146e-05,
|
704 |
+
"loss": 2.7658,
|
705 |
+
"step": 48000
|
706 |
+
},
|
707 |
+
{
|
708 |
+
"epoch": 3.758815779276137,
|
709 |
+
"grad_norm": 6.242121696472168,
|
710 |
+
"learning_rate": 1.2482368441447726e-05,
|
711 |
+
"loss": 2.7698,
|
712 |
+
"step": 48500
|
713 |
+
},
|
714 |
+
{
|
715 |
+
"epoch": 3.7975664574130046,
|
716 |
+
"grad_norm": 6.123905658721924,
|
717 |
+
"learning_rate": 1.2404867085173991e-05,
|
718 |
+
"loss": 2.7711,
|
719 |
+
"step": 49000
|
720 |
+
},
|
721 |
+
{
|
722 |
+
"epoch": 3.836317135549872,
|
723 |
+
"grad_norm": 6.735771179199219,
|
724 |
+
"learning_rate": 1.2327365728900256e-05,
|
725 |
+
"loss": 2.726,
|
726 |
+
"step": 49500
|
727 |
+
},
|
728 |
+
{
|
729 |
+
"epoch": 3.8750678136867394,
|
730 |
+
"grad_norm": 6.921602725982666,
|
731 |
+
"learning_rate": 1.2249864372626522e-05,
|
732 |
+
"loss": 2.7545,
|
733 |
+
"step": 50000
|
734 |
+
},
|
735 |
+
{
|
736 |
+
"epoch": 3.9138184918236067,
|
737 |
+
"grad_norm": 6.343456745147705,
|
738 |
+
"learning_rate": 1.2172363016352787e-05,
|
739 |
+
"loss": 2.7474,
|
740 |
+
"step": 50500
|
741 |
+
},
|
742 |
+
{
|
743 |
+
"epoch": 3.9525691699604746,
|
744 |
+
"grad_norm": 6.30169677734375,
|
745 |
+
"learning_rate": 1.2094861660079052e-05,
|
746 |
+
"loss": 2.7467,
|
747 |
+
"step": 51000
|
748 |
+
},
|
749 |
+
{
|
750 |
+
"epoch": 3.9913198480973415,
|
751 |
+
"grad_norm": 6.6629767417907715,
|
752 |
+
"learning_rate": 1.2017360303805317e-05,
|
753 |
+
"loss": 2.7475,
|
754 |
+
"step": 51500
|
755 |
+
},
|
756 |
+
{
|
757 |
+
"epoch": 4.0,
|
758 |
+
"eval_loss": 2.6640822887420654,
|
759 |
+
"eval_runtime": 260.2494,
|
760 |
+
"eval_samples_per_second": 793.335,
|
761 |
+
"eval_steps_per_second": 12.4,
|
762 |
+
"step": 51612
|
763 |
+
},
|
764 |
+
{
|
765 |
+
"epoch": 4.030070526234209,
|
766 |
+
"grad_norm": 6.397671222686768,
|
767 |
+
"learning_rate": 1.1939858947531581e-05,
|
768 |
+
"loss": 2.7311,
|
769 |
+
"step": 52000
|
770 |
+
},
|
771 |
+
{
|
772 |
+
"epoch": 4.068821204371076,
|
773 |
+
"grad_norm": 6.374961853027344,
|
774 |
+
"learning_rate": 1.1862357591257846e-05,
|
775 |
+
"loss": 2.7119,
|
776 |
+
"step": 52500
|
777 |
+
},
|
778 |
+
{
|
779 |
+
"epoch": 4.107571882507944,
|
780 |
+
"grad_norm": 5.920938968658447,
|
781 |
+
"learning_rate": 1.1784856234984112e-05,
|
782 |
+
"loss": 2.7217,
|
783 |
+
"step": 53000
|
784 |
+
},
|
785 |
+
{
|
786 |
+
"epoch": 4.146322560644811,
|
787 |
+
"grad_norm": 6.377143859863281,
|
788 |
+
"learning_rate": 1.1707354878710377e-05,
|
789 |
+
"loss": 2.7044,
|
790 |
+
"step": 53500
|
791 |
+
},
|
792 |
+
{
|
793 |
+
"epoch": 4.185073238781679,
|
794 |
+
"grad_norm": 7.047250270843506,
|
795 |
+
"learning_rate": 1.1629853522436642e-05,
|
796 |
+
"loss": 2.7213,
|
797 |
+
"step": 54000
|
798 |
+
},
|
799 |
+
{
|
800 |
+
"epoch": 4.223823916918546,
|
801 |
+
"grad_norm": 6.682352066040039,
|
802 |
+
"learning_rate": 1.1552352166162907e-05,
|
803 |
+
"loss": 2.7025,
|
804 |
+
"step": 54500
|
805 |
+
},
|
806 |
+
{
|
807 |
+
"epoch": 4.262574595055414,
|
808 |
+
"grad_norm": 6.547230243682861,
|
809 |
+
"learning_rate": 1.1474850809889173e-05,
|
810 |
+
"loss": 2.7068,
|
811 |
+
"step": 55000
|
812 |
+
},
|
813 |
+
{
|
814 |
+
"epoch": 4.301325273192281,
|
815 |
+
"grad_norm": 6.038912296295166,
|
816 |
+
"learning_rate": 1.1397349453615438e-05,
|
817 |
+
"loss": 2.7061,
|
818 |
+
"step": 55500
|
819 |
+
},
|
820 |
+
{
|
821 |
+
"epoch": 4.3400759513291485,
|
822 |
+
"grad_norm": 6.072612762451172,
|
823 |
+
"learning_rate": 1.1319848097341703e-05,
|
824 |
+
"loss": 2.7037,
|
825 |
+
"step": 56000
|
826 |
+
},
|
827 |
+
{
|
828 |
+
"epoch": 4.378826629466015,
|
829 |
+
"grad_norm": 5.6306281089782715,
|
830 |
+
"learning_rate": 1.1242346741067968e-05,
|
831 |
+
"loss": 2.6999,
|
832 |
+
"step": 56500
|
833 |
+
},
|
834 |
+
{
|
835 |
+
"epoch": 4.417577307602883,
|
836 |
+
"grad_norm": 6.18297004699707,
|
837 |
+
"learning_rate": 1.1164845384794234e-05,
|
838 |
+
"loss": 2.6974,
|
839 |
+
"step": 57000
|
840 |
+
},
|
841 |
+
{
|
842 |
+
"epoch": 4.45632798573975,
|
843 |
+
"grad_norm": 6.371115207672119,
|
844 |
+
"learning_rate": 1.1087344028520499e-05,
|
845 |
+
"loss": 2.6918,
|
846 |
+
"step": 57500
|
847 |
+
},
|
848 |
+
{
|
849 |
+
"epoch": 4.495078663876618,
|
850 |
+
"grad_norm": 6.444944381713867,
|
851 |
+
"learning_rate": 1.1009842672246764e-05,
|
852 |
+
"loss": 2.6874,
|
853 |
+
"step": 58000
|
854 |
+
},
|
855 |
+
{
|
856 |
+
"epoch": 4.533829342013485,
|
857 |
+
"grad_norm": 6.176960468292236,
|
858 |
+
"learning_rate": 1.093234131597303e-05,
|
859 |
+
"loss": 2.68,
|
860 |
+
"step": 58500
|
861 |
+
},
|
862 |
+
{
|
863 |
+
"epoch": 4.572580020150353,
|
864 |
+
"grad_norm": 6.731847763061523,
|
865 |
+
"learning_rate": 1.0854839959699295e-05,
|
866 |
+
"loss": 2.6919,
|
867 |
+
"step": 59000
|
868 |
+
},
|
869 |
+
{
|
870 |
+
"epoch": 4.61133069828722,
|
871 |
+
"grad_norm": 7.826213836669922,
|
872 |
+
"learning_rate": 1.077733860342556e-05,
|
873 |
+
"loss": 2.6824,
|
874 |
+
"step": 59500
|
875 |
+
},
|
876 |
+
{
|
877 |
+
"epoch": 4.650081376424088,
|
878 |
+
"grad_norm": 7.052020072937012,
|
879 |
+
"learning_rate": 1.0699837247151825e-05,
|
880 |
+
"loss": 2.6616,
|
881 |
+
"step": 60000
|
882 |
+
},
|
883 |
+
{
|
884 |
+
"epoch": 4.6888320545609545,
|
885 |
+
"grad_norm": 5.36915922164917,
|
886 |
+
"learning_rate": 1.062233589087809e-05,
|
887 |
+
"loss": 2.667,
|
888 |
+
"step": 60500
|
889 |
+
},
|
890 |
+
{
|
891 |
+
"epoch": 4.727582732697822,
|
892 |
+
"grad_norm": 6.491717338562012,
|
893 |
+
"learning_rate": 1.0544834534604356e-05,
|
894 |
+
"loss": 2.6896,
|
895 |
+
"step": 61000
|
896 |
+
},
|
897 |
+
{
|
898 |
+
"epoch": 4.766333410834689,
|
899 |
+
"grad_norm": 7.702902793884277,
|
900 |
+
"learning_rate": 1.0467333178330621e-05,
|
901 |
+
"loss": 2.6712,
|
902 |
+
"step": 61500
|
903 |
+
},
|
904 |
+
{
|
905 |
+
"epoch": 4.805084088971557,
|
906 |
+
"grad_norm": 6.359930992126465,
|
907 |
+
"learning_rate": 1.0389831822056886e-05,
|
908 |
+
"loss": 2.6704,
|
909 |
+
"step": 62000
|
910 |
+
},
|
911 |
+
{
|
912 |
+
"epoch": 4.843834767108424,
|
913 |
+
"grad_norm": 6.2874531745910645,
|
914 |
+
"learning_rate": 1.0312330465783152e-05,
|
915 |
+
"loss": 2.6757,
|
916 |
+
"step": 62500
|
917 |
+
},
|
918 |
+
{
|
919 |
+
"epoch": 4.882585445245292,
|
920 |
+
"grad_norm": 6.827906131744385,
|
921 |
+
"learning_rate": 1.0234829109509417e-05,
|
922 |
+
"loss": 2.6567,
|
923 |
+
"step": 63000
|
924 |
+
},
|
925 |
+
{
|
926 |
+
"epoch": 4.921336123382159,
|
927 |
+
"grad_norm": 6.620416164398193,
|
928 |
+
"learning_rate": 1.0157327753235682e-05,
|
929 |
+
"loss": 2.6615,
|
930 |
+
"step": 63500
|
931 |
+
},
|
932 |
+
{
|
933 |
+
"epoch": 4.960086801519027,
|
934 |
+
"grad_norm": 6.6219162940979,
|
935 |
+
"learning_rate": 1.0079826396961947e-05,
|
936 |
+
"loss": 2.657,
|
937 |
+
"step": 64000
|
938 |
+
},
|
939 |
+
{
|
940 |
+
"epoch": 4.998837479655894,
|
941 |
+
"grad_norm": 6.214903831481934,
|
942 |
+
"learning_rate": 1.0002325040688213e-05,
|
943 |
+
"loss": 2.6549,
|
944 |
+
"step": 64500
|
945 |
+
},
|
946 |
+
{
|
947 |
+
"epoch": 5.0,
|
948 |
+
"eval_loss": 2.578911066055298,
|
949 |
+
"eval_runtime": 265.1883,
|
950 |
+
"eval_samples_per_second": 778.56,
|
951 |
+
"eval_steps_per_second": 12.169,
|
952 |
+
"step": 64515
|
953 |
+
},
|
954 |
+
{
|
955 |
+
"epoch": 5.0375881577927615,
|
956 |
+
"grad_norm": 6.627685546875,
|
957 |
+
"learning_rate": 9.924823684414478e-06,
|
958 |
+
"loss": 2.6203,
|
959 |
+
"step": 65000
|
960 |
+
},
|
961 |
+
{
|
962 |
+
"epoch": 5.076338835929628,
|
963 |
+
"grad_norm": 6.23040771484375,
|
964 |
+
"learning_rate": 9.847322328140743e-06,
|
965 |
+
"loss": 2.6349,
|
966 |
+
"step": 65500
|
967 |
+
},
|
968 |
+
{
|
969 |
+
"epoch": 5.115089514066496,
|
970 |
+
"grad_norm": 6.667369365692139,
|
971 |
+
"learning_rate": 9.769820971867009e-06,
|
972 |
+
"loss": 2.647,
|
973 |
+
"step": 66000
|
974 |
+
},
|
975 |
+
{
|
976 |
+
"epoch": 5.153840192203363,
|
977 |
+
"grad_norm": 6.694558620452881,
|
978 |
+
"learning_rate": 9.692319615593274e-06,
|
979 |
+
"loss": 2.6214,
|
980 |
+
"step": 66500
|
981 |
+
},
|
982 |
+
{
|
983 |
+
"epoch": 5.192590870340231,
|
984 |
+
"grad_norm": 6.280242443084717,
|
985 |
+
"learning_rate": 9.614818259319539e-06,
|
986 |
+
"loss": 2.6206,
|
987 |
+
"step": 67000
|
988 |
+
},
|
989 |
+
{
|
990 |
+
"epoch": 5.231341548477098,
|
991 |
+
"grad_norm": 6.660119533538818,
|
992 |
+
"learning_rate": 9.537316903045804e-06,
|
993 |
+
"loss": 2.6307,
|
994 |
+
"step": 67500
|
995 |
+
},
|
996 |
+
{
|
997 |
+
"epoch": 5.270092226613966,
|
998 |
+
"grad_norm": 6.439652919769287,
|
999 |
+
"learning_rate": 9.45981554677207e-06,
|
1000 |
+
"loss": 2.6431,
|
1001 |
+
"step": 68000
|
1002 |
+
},
|
1003 |
+
{
|
1004 |
+
"epoch": 5.308842904750833,
|
1005 |
+
"grad_norm": 6.055843830108643,
|
1006 |
+
"learning_rate": 9.382314190498335e-06,
|
1007 |
+
"loss": 2.6144,
|
1008 |
+
"step": 68500
|
1009 |
+
},
|
1010 |
+
{
|
1011 |
+
"epoch": 5.347593582887701,
|
1012 |
+
"grad_norm": 6.519714832305908,
|
1013 |
+
"learning_rate": 9.3048128342246e-06,
|
1014 |
+
"loss": 2.6056,
|
1015 |
+
"step": 69000
|
1016 |
+
},
|
1017 |
+
{
|
1018 |
+
"epoch": 5.3863442610245675,
|
1019 |
+
"grad_norm": 6.72304630279541,
|
1020 |
+
"learning_rate": 9.227311477950864e-06,
|
1021 |
+
"loss": 2.623,
|
1022 |
+
"step": 69500
|
1023 |
+
},
|
1024 |
+
{
|
1025 |
+
"epoch": 5.425094939161435,
|
1026 |
+
"grad_norm": 7.048790454864502,
|
1027 |
+
"learning_rate": 9.149810121677129e-06,
|
1028 |
+
"loss": 2.6043,
|
1029 |
+
"step": 70000
|
1030 |
+
},
|
1031 |
+
{
|
1032 |
+
"epoch": 5.463845617298302,
|
1033 |
+
"grad_norm": 6.654219627380371,
|
1034 |
+
"learning_rate": 9.072308765403394e-06,
|
1035 |
+
"loss": 2.6135,
|
1036 |
+
"step": 70500
|
1037 |
+
},
|
1038 |
+
{
|
1039 |
+
"epoch": 5.50259629543517,
|
1040 |
+
"grad_norm": 5.948112487792969,
|
1041 |
+
"learning_rate": 8.99480740912966e-06,
|
1042 |
+
"loss": 2.6295,
|
1043 |
+
"step": 71000
|
1044 |
+
},
|
1045 |
+
{
|
1046 |
+
"epoch": 5.541346973572038,
|
1047 |
+
"grad_norm": 7.8044328689575195,
|
1048 |
+
"learning_rate": 8.917306052855925e-06,
|
1049 |
+
"loss": 2.6104,
|
1050 |
+
"step": 71500
|
1051 |
+
},
|
1052 |
+
{
|
1053 |
+
"epoch": 5.580097651708905,
|
1054 |
+
"grad_norm": 6.743612766265869,
|
1055 |
+
"learning_rate": 8.83980469658219e-06,
|
1056 |
+
"loss": 2.6216,
|
1057 |
+
"step": 72000
|
1058 |
+
},
|
1059 |
+
{
|
1060 |
+
"epoch": 5.618848329845772,
|
1061 |
+
"grad_norm": 6.346240043640137,
|
1062 |
+
"learning_rate": 8.762303340308455e-06,
|
1063 |
+
"loss": 2.6238,
|
1064 |
+
"step": 72500
|
1065 |
+
},
|
1066 |
+
{
|
1067 |
+
"epoch": 5.65759900798264,
|
1068 |
+
"grad_norm": 6.496920108795166,
|
1069 |
+
"learning_rate": 8.68480198403472e-06,
|
1070 |
+
"loss": 2.6334,
|
1071 |
+
"step": 73000
|
1072 |
+
},
|
1073 |
+
{
|
1074 |
+
"epoch": 5.6963496861195075,
|
1075 |
+
"grad_norm": 6.356810569763184,
|
1076 |
+
"learning_rate": 8.607300627760986e-06,
|
1077 |
+
"loss": 2.5995,
|
1078 |
+
"step": 73500
|
1079 |
+
},
|
1080 |
+
{
|
1081 |
+
"epoch": 5.7351003642563745,
|
1082 |
+
"grad_norm": 6.226792812347412,
|
1083 |
+
"learning_rate": 8.529799271487251e-06,
|
1084 |
+
"loss": 2.5974,
|
1085 |
+
"step": 74000
|
1086 |
+
},
|
1087 |
+
{
|
1088 |
+
"epoch": 5.773851042393241,
|
1089 |
+
"grad_norm": 6.6555962562561035,
|
1090 |
+
"learning_rate": 8.452297915213516e-06,
|
1091 |
+
"loss": 2.6285,
|
1092 |
+
"step": 74500
|
1093 |
+
},
|
1094 |
+
{
|
1095 |
+
"epoch": 5.812601720530109,
|
1096 |
+
"grad_norm": 6.32110595703125,
|
1097 |
+
"learning_rate": 8.374796558939782e-06,
|
1098 |
+
"loss": 2.6035,
|
1099 |
+
"step": 75000
|
1100 |
+
},
|
1101 |
+
{
|
1102 |
+
"epoch": 5.851352398666977,
|
1103 |
+
"grad_norm": 6.651345252990723,
|
1104 |
+
"learning_rate": 8.297295202666047e-06,
|
1105 |
+
"loss": 2.5886,
|
1106 |
+
"step": 75500
|
1107 |
+
},
|
1108 |
+
{
|
1109 |
+
"epoch": 5.890103076803844,
|
1110 |
+
"grad_norm": 6.736583232879639,
|
1111 |
+
"learning_rate": 8.219793846392312e-06,
|
1112 |
+
"loss": 2.5903,
|
1113 |
+
"step": 76000
|
1114 |
+
},
|
1115 |
+
{
|
1116 |
+
"epoch": 5.928853754940711,
|
1117 |
+
"grad_norm": 6.635737895965576,
|
1118 |
+
"learning_rate": 8.142292490118577e-06,
|
1119 |
+
"loss": 2.597,
|
1120 |
+
"step": 76500
|
1121 |
+
},
|
1122 |
+
{
|
1123 |
+
"epoch": 5.967604433077579,
|
1124 |
+
"grad_norm": 6.3186492919921875,
|
1125 |
+
"learning_rate": 8.064791133844843e-06,
|
1126 |
+
"loss": 2.5732,
|
1127 |
+
"step": 77000
|
1128 |
+
},
|
1129 |
+
{
|
1130 |
+
"epoch": 6.0,
|
1131 |
+
"eval_loss": 2.5146169662475586,
|
1132 |
+
"eval_runtime": 259.2569,
|
1133 |
+
"eval_samples_per_second": 796.372,
|
1134 |
+
"eval_steps_per_second": 12.447,
|
1135 |
+
"step": 77418
|
1136 |
+
},
|
1137 |
+
{
|
1138 |
+
"epoch": 6.006355111214447,
|
1139 |
+
"grad_norm": 6.408041000366211,
|
1140 |
+
"learning_rate": 7.987289777571108e-06,
|
1141 |
+
"loss": 2.5742,
|
1142 |
+
"step": 77500
|
1143 |
+
},
|
1144 |
+
{
|
1145 |
+
"epoch": 6.045105789351314,
|
1146 |
+
"grad_norm": 6.398166656494141,
|
1147 |
+
"learning_rate": 7.909788421297373e-06,
|
1148 |
+
"loss": 2.5829,
|
1149 |
+
"step": 78000
|
1150 |
+
},
|
1151 |
+
{
|
1152 |
+
"epoch": 6.083856467488181,
|
1153 |
+
"grad_norm": 6.89434289932251,
|
1154 |
+
"learning_rate": 7.832287065023639e-06,
|
1155 |
+
"loss": 2.58,
|
1156 |
+
"step": 78500
|
1157 |
+
},
|
1158 |
+
{
|
1159 |
+
"epoch": 6.122607145625048,
|
1160 |
+
"grad_norm": 5.935701847076416,
|
1161 |
+
"learning_rate": 7.754785708749904e-06,
|
1162 |
+
"loss": 2.5853,
|
1163 |
+
"step": 79000
|
1164 |
+
},
|
1165 |
+
{
|
1166 |
+
"epoch": 6.161357823761916,
|
1167 |
+
"grad_norm": 7.224461555480957,
|
1168 |
+
"learning_rate": 7.677284352476169e-06,
|
1169 |
+
"loss": 2.5597,
|
1170 |
+
"step": 79500
|
1171 |
+
},
|
1172 |
+
{
|
1173 |
+
"epoch": 6.200108501898783,
|
1174 |
+
"grad_norm": 6.59751033782959,
|
1175 |
+
"learning_rate": 7.5997829962024335e-06,
|
1176 |
+
"loss": 2.5821,
|
1177 |
+
"step": 80000
|
1178 |
+
},
|
1179 |
+
{
|
1180 |
+
"epoch": 6.238859180035651,
|
1181 |
+
"grad_norm": 6.414103031158447,
|
1182 |
+
"learning_rate": 7.522281639928699e-06,
|
1183 |
+
"loss": 2.5542,
|
1184 |
+
"step": 80500
|
1185 |
+
},
|
1186 |
+
{
|
1187 |
+
"epoch": 6.277609858172518,
|
1188 |
+
"grad_norm": 6.270075798034668,
|
1189 |
+
"learning_rate": 7.444780283654964e-06,
|
1190 |
+
"loss": 2.5735,
|
1191 |
+
"step": 81000
|
1192 |
+
},
|
1193 |
+
{
|
1194 |
+
"epoch": 6.316360536309386,
|
1195 |
+
"grad_norm": 6.3846306800842285,
|
1196 |
+
"learning_rate": 7.367278927381229e-06,
|
1197 |
+
"loss": 2.5563,
|
1198 |
+
"step": 81500
|
1199 |
+
},
|
1200 |
+
{
|
1201 |
+
"epoch": 6.355111214446253,
|
1202 |
+
"grad_norm": 6.725887298583984,
|
1203 |
+
"learning_rate": 7.2897775711074945e-06,
|
1204 |
+
"loss": 2.5582,
|
1205 |
+
"step": 82000
|
1206 |
+
},
|
1207 |
+
{
|
1208 |
+
"epoch": 6.3938618925831205,
|
1209 |
+
"grad_norm": 6.913090229034424,
|
1210 |
+
"learning_rate": 7.21227621483376e-06,
|
1211 |
+
"loss": 2.5681,
|
1212 |
+
"step": 82500
|
1213 |
+
},
|
1214 |
+
{
|
1215 |
+
"epoch": 6.4326125707199875,
|
1216 |
+
"grad_norm": 6.630814075469971,
|
1217 |
+
"learning_rate": 7.134774858560025e-06,
|
1218 |
+
"loss": 2.5493,
|
1219 |
+
"step": 83000
|
1220 |
+
},
|
1221 |
+
{
|
1222 |
+
"epoch": 6.471363248856855,
|
1223 |
+
"grad_norm": 7.482264518737793,
|
1224 |
+
"learning_rate": 7.05727350228629e-06,
|
1225 |
+
"loss": 2.5672,
|
1226 |
+
"step": 83500
|
1227 |
+
},
|
1228 |
+
{
|
1229 |
+
"epoch": 6.510113926993722,
|
1230 |
+
"grad_norm": 5.896800518035889,
|
1231 |
+
"learning_rate": 6.979772146012556e-06,
|
1232 |
+
"loss": 2.5563,
|
1233 |
+
"step": 84000
|
1234 |
+
},
|
1235 |
+
{
|
1236 |
+
"epoch": 6.54886460513059,
|
1237 |
+
"grad_norm": 6.603734016418457,
|
1238 |
+
"learning_rate": 6.902270789738821e-06,
|
1239 |
+
"loss": 2.5358,
|
1240 |
+
"step": 84500
|
1241 |
+
},
|
1242 |
+
{
|
1243 |
+
"epoch": 6.587615283267457,
|
1244 |
+
"grad_norm": 6.386889457702637,
|
1245 |
+
"learning_rate": 6.824769433465086e-06,
|
1246 |
+
"loss": 2.5449,
|
1247 |
+
"step": 85000
|
1248 |
+
},
|
1249 |
+
{
|
1250 |
+
"epoch": 6.626365961404325,
|
1251 |
+
"grad_norm": 6.661931037902832,
|
1252 |
+
"learning_rate": 6.747268077191351e-06,
|
1253 |
+
"loss": 2.5405,
|
1254 |
+
"step": 85500
|
1255 |
+
},
|
1256 |
+
{
|
1257 |
+
"epoch": 6.665116639541192,
|
1258 |
+
"grad_norm": 6.331045627593994,
|
1259 |
+
"learning_rate": 6.669766720917617e-06,
|
1260 |
+
"loss": 2.5419,
|
1261 |
+
"step": 86000
|
1262 |
+
},
|
1263 |
+
{
|
1264 |
+
"epoch": 6.70386731767806,
|
1265 |
+
"grad_norm": 7.050119400024414,
|
1266 |
+
"learning_rate": 6.592265364643882e-06,
|
1267 |
+
"loss": 2.5196,
|
1268 |
+
"step": 86500
|
1269 |
+
},
|
1270 |
+
{
|
1271 |
+
"epoch": 6.742617995814927,
|
1272 |
+
"grad_norm": 6.065616130828857,
|
1273 |
+
"learning_rate": 6.514764008370147e-06,
|
1274 |
+
"loss": 2.539,
|
1275 |
+
"step": 87000
|
1276 |
+
},
|
1277 |
+
{
|
1278 |
+
"epoch": 6.781368673951794,
|
1279 |
+
"grad_norm": 5.768097877502441,
|
1280 |
+
"learning_rate": 6.4372626520964125e-06,
|
1281 |
+
"loss": 2.5245,
|
1282 |
+
"step": 87500
|
1283 |
+
},
|
1284 |
+
{
|
1285 |
+
"epoch": 6.820119352088661,
|
1286 |
+
"grad_norm": 6.785781383514404,
|
1287 |
+
"learning_rate": 6.359761295822677e-06,
|
1288 |
+
"loss": 2.5473,
|
1289 |
+
"step": 88000
|
1290 |
+
},
|
1291 |
+
{
|
1292 |
+
"epoch": 6.858870030225529,
|
1293 |
+
"grad_norm": 6.658846855163574,
|
1294 |
+
"learning_rate": 6.282259939548942e-06,
|
1295 |
+
"loss": 2.5385,
|
1296 |
+
"step": 88500
|
1297 |
+
},
|
1298 |
+
{
|
1299 |
+
"epoch": 6.897620708362396,
|
1300 |
+
"grad_norm": 5.932773590087891,
|
1301 |
+
"learning_rate": 6.2047585832752074e-06,
|
1302 |
+
"loss": 2.528,
|
1303 |
+
"step": 89000
|
1304 |
+
},
|
1305 |
+
{
|
1306 |
+
"epoch": 6.936371386499264,
|
1307 |
+
"grad_norm": 6.457767963409424,
|
1308 |
+
"learning_rate": 6.127257227001473e-06,
|
1309 |
+
"loss": 2.5327,
|
1310 |
+
"step": 89500
|
1311 |
+
},
|
1312 |
+
{
|
1313 |
+
"epoch": 6.975122064636131,
|
1314 |
+
"grad_norm": 6.143023490905762,
|
1315 |
+
"learning_rate": 6.049755870727738e-06,
|
1316 |
+
"loss": 2.5352,
|
1317 |
+
"step": 90000
|
1318 |
+
},
|
1319 |
+
{
|
1320 |
+
"epoch": 7.0,
|
1321 |
+
"eval_loss": 2.4585013389587402,
|
1322 |
+
"eval_runtime": 258.9573,
|
1323 |
+
"eval_samples_per_second": 797.294,
|
1324 |
+
"eval_steps_per_second": 12.462,
|
1325 |
+
"step": 90321
|
1326 |
+
},
|
1327 |
+
{
|
1328 |
+
"epoch": 7.013872742772999,
|
1329 |
+
"grad_norm": 6.153046607971191,
|
1330 |
+
"learning_rate": 5.972254514454003e-06,
|
1331 |
+
"loss": 2.5315,
|
1332 |
+
"step": 90500
|
1333 |
+
},
|
1334 |
+
{
|
1335 |
+
"epoch": 7.052623420909866,
|
1336 |
+
"grad_norm": 7.131119728088379,
|
1337 |
+
"learning_rate": 5.8947531581802685e-06,
|
1338 |
+
"loss": 2.5431,
|
1339 |
+
"step": 91000
|
1340 |
+
},
|
1341 |
+
{
|
1342 |
+
"epoch": 7.0913740990467335,
|
1343 |
+
"grad_norm": 6.677100658416748,
|
1344 |
+
"learning_rate": 5.817251801906534e-06,
|
1345 |
+
"loss": 2.5204,
|
1346 |
+
"step": 91500
|
1347 |
+
},
|
1348 |
+
{
|
1349 |
+
"epoch": 7.1301247771836005,
|
1350 |
+
"grad_norm": 6.799976348876953,
|
1351 |
+
"learning_rate": 5.739750445632799e-06,
|
1352 |
+
"loss": 2.5221,
|
1353 |
+
"step": 92000
|
1354 |
+
},
|
1355 |
+
{
|
1356 |
+
"epoch": 7.168875455320468,
|
1357 |
+
"grad_norm": 6.515171051025391,
|
1358 |
+
"learning_rate": 5.662249089359064e-06,
|
1359 |
+
"loss": 2.5222,
|
1360 |
+
"step": 92500
|
1361 |
+
},
|
1362 |
+
{
|
1363 |
+
"epoch": 7.207626133457335,
|
1364 |
+
"grad_norm": 7.057505130767822,
|
1365 |
+
"learning_rate": 5.58474773308533e-06,
|
1366 |
+
"loss": 2.5262,
|
1367 |
+
"step": 93000
|
1368 |
+
},
|
1369 |
+
{
|
1370 |
+
"epoch": 7.246376811594203,
|
1371 |
+
"grad_norm": 5.927343368530273,
|
1372 |
+
"learning_rate": 5.507246376811595e-06,
|
1373 |
+
"loss": 2.5272,
|
1374 |
+
"step": 93500
|
1375 |
+
},
|
1376 |
+
{
|
1377 |
+
"epoch": 7.28512748973107,
|
1378 |
+
"grad_norm": 6.7214155197143555,
|
1379 |
+
"learning_rate": 5.42974502053786e-06,
|
1380 |
+
"loss": 2.5195,
|
1381 |
+
"step": 94000
|
1382 |
+
},
|
1383 |
+
{
|
1384 |
+
"epoch": 7.323878167867938,
|
1385 |
+
"grad_norm": 6.162799835205078,
|
1386 |
+
"learning_rate": 5.352243664264125e-06,
|
1387 |
+
"loss": 2.5117,
|
1388 |
+
"step": 94500
|
1389 |
+
},
|
1390 |
+
{
|
1391 |
+
"epoch": 7.362628846004805,
|
1392 |
+
"grad_norm": 6.725783824920654,
|
1393 |
+
"learning_rate": 5.274742307990391e-06,
|
1394 |
+
"loss": 2.522,
|
1395 |
+
"step": 95000
|
1396 |
+
},
|
1397 |
+
{
|
1398 |
+
"epoch": 7.401379524141673,
|
1399 |
+
"grad_norm": 5.721879959106445,
|
1400 |
+
"learning_rate": 5.197240951716656e-06,
|
1401 |
+
"loss": 2.5047,
|
1402 |
+
"step": 95500
|
1403 |
+
},
|
1404 |
+
{
|
1405 |
+
"epoch": 7.44013020227854,
|
1406 |
+
"grad_norm": 7.531757354736328,
|
1407 |
+
"learning_rate": 5.11973959544292e-06,
|
1408 |
+
"loss": 2.4981,
|
1409 |
+
"step": 96000
|
1410 |
+
},
|
1411 |
+
{
|
1412 |
+
"epoch": 7.478880880415407,
|
1413 |
+
"grad_norm": 6.200819492340088,
|
1414 |
+
"learning_rate": 5.042238239169186e-06,
|
1415 |
+
"loss": 2.5016,
|
1416 |
+
"step": 96500
|
1417 |
+
},
|
1418 |
+
{
|
1419 |
+
"epoch": 7.517631558552274,
|
1420 |
+
"grad_norm": 6.8695597648620605,
|
1421 |
+
"learning_rate": 4.964736882895451e-06,
|
1422 |
+
"loss": 2.5085,
|
1423 |
+
"step": 97000
|
1424 |
+
},
|
1425 |
+
{
|
1426 |
+
"epoch": 7.556382236689142,
|
1427 |
+
"grad_norm": 6.3883843421936035,
|
1428 |
+
"learning_rate": 4.887235526621716e-06,
|
1429 |
+
"loss": 2.5092,
|
1430 |
+
"step": 97500
|
1431 |
+
},
|
1432 |
+
{
|
1433 |
+
"epoch": 7.595132914826009,
|
1434 |
+
"grad_norm": 6.085172653198242,
|
1435 |
+
"learning_rate": 4.809734170347981e-06,
|
1436 |
+
"loss": 2.4957,
|
1437 |
+
"step": 98000
|
1438 |
+
},
|
1439 |
+
{
|
1440 |
+
"epoch": 7.633883592962877,
|
1441 |
+
"grad_norm": 6.23600435256958,
|
1442 |
+
"learning_rate": 4.732232814074247e-06,
|
1443 |
+
"loss": 2.4876,
|
1444 |
+
"step": 98500
|
1445 |
+
},
|
1446 |
+
{
|
1447 |
+
"epoch": 7.672634271099744,
|
1448 |
+
"grad_norm": 6.483453750610352,
|
1449 |
+
"learning_rate": 4.654731457800512e-06,
|
1450 |
+
"loss": 2.5029,
|
1451 |
+
"step": 99000
|
1452 |
+
},
|
1453 |
+
{
|
1454 |
+
"epoch": 7.711384949236612,
|
1455 |
+
"grad_norm": 6.627302646636963,
|
1456 |
+
"learning_rate": 4.577230101526777e-06,
|
1457 |
+
"loss": 2.4989,
|
1458 |
+
"step": 99500
|
1459 |
+
},
|
1460 |
+
{
|
1461 |
+
"epoch": 7.750135627373479,
|
1462 |
+
"grad_norm": 7.044070243835449,
|
1463 |
+
"learning_rate": 4.4997287452530425e-06,
|
1464 |
+
"loss": 2.5085,
|
1465 |
+
"step": 100000
|
1466 |
+
},
|
1467 |
+
{
|
1468 |
+
"epoch": 7.7888863055103466,
|
1469 |
+
"grad_norm": 5.986552715301514,
|
1470 |
+
"learning_rate": 4.422227388979308e-06,
|
1471 |
+
"loss": 2.4842,
|
1472 |
+
"step": 100500
|
1473 |
+
},
|
1474 |
+
{
|
1475 |
+
"epoch": 7.8276369836472135,
|
1476 |
+
"grad_norm": 6.3408708572387695,
|
1477 |
+
"learning_rate": 4.344726032705573e-06,
|
1478 |
+
"loss": 2.4973,
|
1479 |
+
"step": 101000
|
1480 |
+
},
|
1481 |
+
{
|
1482 |
+
"epoch": 7.866387661784081,
|
1483 |
+
"grad_norm": 6.100359916687012,
|
1484 |
+
"learning_rate": 4.267224676431838e-06,
|
1485 |
+
"loss": 2.5111,
|
1486 |
+
"step": 101500
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 7.905138339920948,
|
1490 |
+
"grad_norm": 6.7454833984375,
|
1491 |
+
"learning_rate": 4.1897233201581036e-06,
|
1492 |
+
"loss": 2.4766,
|
1493 |
+
"step": 102000
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 7.943889018057816,
|
1497 |
+
"grad_norm": 6.790141582489014,
|
1498 |
+
"learning_rate": 4.112221963884369e-06,
|
1499 |
+
"loss": 2.4788,
|
1500 |
+
"step": 102500
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 7.982639696194683,
|
1504 |
+
"grad_norm": 6.926203727722168,
|
1505 |
+
"learning_rate": 4.034720607610634e-06,
|
1506 |
+
"loss": 2.4875,
|
1507 |
+
"step": 103000
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 8.0,
|
1511 |
+
"eval_loss": 2.435317277908325,
|
1512 |
+
"eval_runtime": 258.5225,
|
1513 |
+
"eval_samples_per_second": 798.634,
|
1514 |
+
"eval_steps_per_second": 12.482,
|
1515 |
+
"step": 103224
|
1516 |
+
},
|
1517 |
+
{
|
1518 |
+
"epoch": 8.02139037433155,
|
1519 |
+
"grad_norm": 6.832672119140625,
|
1520 |
+
"learning_rate": 3.957219251336899e-06,
|
1521 |
+
"loss": 2.4812,
|
1522 |
+
"step": 103500
|
1523 |
+
},
|
1524 |
+
{
|
1525 |
+
"epoch": 8.060141052468419,
|
1526 |
+
"grad_norm": 6.771292209625244,
|
1527 |
+
"learning_rate": 3.879717895063164e-06,
|
1528 |
+
"loss": 2.4945,
|
1529 |
+
"step": 104000
|
1530 |
+
},
|
1531 |
+
{
|
1532 |
+
"epoch": 8.098891730605285,
|
1533 |
+
"grad_norm": 6.624267101287842,
|
1534 |
+
"learning_rate": 3.802216538789429e-06,
|
1535 |
+
"loss": 2.4813,
|
1536 |
+
"step": 104500
|
1537 |
+
},
|
1538 |
+
{
|
1539 |
+
"epoch": 8.137642408742153,
|
1540 |
+
"grad_norm": 6.566524028778076,
|
1541 |
+
"learning_rate": 3.724715182515694e-06,
|
1542 |
+
"loss": 2.5087,
|
1543 |
+
"step": 105000
|
1544 |
+
},
|
1545 |
+
{
|
1546 |
+
"epoch": 8.17639308687902,
|
1547 |
+
"grad_norm": 6.612277507781982,
|
1548 |
+
"learning_rate": 3.647213826241959e-06,
|
1549 |
+
"loss": 2.481,
|
1550 |
+
"step": 105500
|
1551 |
+
},
|
1552 |
+
{
|
1553 |
+
"epoch": 8.215143765015888,
|
1554 |
+
"grad_norm": 6.12284517288208,
|
1555 |
+
"learning_rate": 3.5697124699682244e-06,
|
1556 |
+
"loss": 2.4825,
|
1557 |
+
"step": 106000
|
1558 |
+
},
|
1559 |
+
{
|
1560 |
+
"epoch": 8.253894443152754,
|
1561 |
+
"grad_norm": 6.495052814483643,
|
1562 |
+
"learning_rate": 3.4922111136944897e-06,
|
1563 |
+
"loss": 2.4883,
|
1564 |
+
"step": 106500
|
1565 |
+
},
|
1566 |
+
{
|
1567 |
+
"epoch": 8.292645121289622,
|
1568 |
+
"grad_norm": 7.689423561096191,
|
1569 |
+
"learning_rate": 3.414709757420755e-06,
|
1570 |
+
"loss": 2.4857,
|
1571 |
+
"step": 107000
|
1572 |
+
},
|
1573 |
+
{
|
1574 |
+
"epoch": 8.33139579942649,
|
1575 |
+
"grad_norm": 6.188397407531738,
|
1576 |
+
"learning_rate": 3.3372084011470202e-06,
|
1577 |
+
"loss": 2.4788,
|
1578 |
+
"step": 107500
|
1579 |
+
},
|
1580 |
+
{
|
1581 |
+
"epoch": 8.370146477563358,
|
1582 |
+
"grad_norm": 6.282194137573242,
|
1583 |
+
"learning_rate": 3.2597070448732855e-06,
|
1584 |
+
"loss": 2.4856,
|
1585 |
+
"step": 108000
|
1586 |
+
},
|
1587 |
+
{
|
1588 |
+
"epoch": 8.408897155700224,
|
1589 |
+
"grad_norm": 6.457098007202148,
|
1590 |
+
"learning_rate": 3.1822056885995508e-06,
|
1591 |
+
"loss": 2.4623,
|
1592 |
+
"step": 108500
|
1593 |
+
},
|
1594 |
+
{
|
1595 |
+
"epoch": 8.447647833837092,
|
1596 |
+
"grad_norm": 7.726540565490723,
|
1597 |
+
"learning_rate": 3.1047043323258156e-06,
|
1598 |
+
"loss": 2.4671,
|
1599 |
+
"step": 109000
|
1600 |
+
},
|
1601 |
+
{
|
1602 |
+
"epoch": 8.48639851197396,
|
1603 |
+
"grad_norm": 6.308920383453369,
|
1604 |
+
"learning_rate": 3.027202976052081e-06,
|
1605 |
+
"loss": 2.4808,
|
1606 |
+
"step": 109500
|
1607 |
+
},
|
1608 |
+
{
|
1609 |
+
"epoch": 8.525149190110827,
|
1610 |
+
"grad_norm": 6.501667499542236,
|
1611 |
+
"learning_rate": 2.949701619778346e-06,
|
1612 |
+
"loss": 2.4736,
|
1613 |
+
"step": 110000
|
1614 |
+
},
|
1615 |
+
{
|
1616 |
+
"epoch": 8.563899868247695,
|
1617 |
+
"grad_norm": 7.358393669128418,
|
1618 |
+
"learning_rate": 2.8722002635046114e-06,
|
1619 |
+
"loss": 2.4697,
|
1620 |
+
"step": 110500
|
1621 |
+
},
|
1622 |
+
{
|
1623 |
+
"epoch": 8.602650546384561,
|
1624 |
+
"grad_norm": 6.261012554168701,
|
1625 |
+
"learning_rate": 2.7946989072308767e-06,
|
1626 |
+
"loss": 2.4631,
|
1627 |
+
"step": 111000
|
1628 |
+
},
|
1629 |
+
{
|
1630 |
+
"epoch": 8.64140122452143,
|
1631 |
+
"grad_norm": 6.515717029571533,
|
1632 |
+
"learning_rate": 2.717197550957142e-06,
|
1633 |
+
"loss": 2.4915,
|
1634 |
+
"step": 111500
|
1635 |
+
},
|
1636 |
+
{
|
1637 |
+
"epoch": 8.680151902658297,
|
1638 |
+
"grad_norm": 6.8307600021362305,
|
1639 |
+
"learning_rate": 2.6396961946834072e-06,
|
1640 |
+
"loss": 2.48,
|
1641 |
+
"step": 112000
|
1642 |
+
},
|
1643 |
+
{
|
1644 |
+
"epoch": 8.718902580795163,
|
1645 |
+
"grad_norm": 6.784819602966309,
|
1646 |
+
"learning_rate": 2.5621948384096725e-06,
|
1647 |
+
"loss": 2.4748,
|
1648 |
+
"step": 112500
|
1649 |
+
},
|
1650 |
+
{
|
1651 |
+
"epoch": 8.75765325893203,
|
1652 |
+
"grad_norm": 7.1304473876953125,
|
1653 |
+
"learning_rate": 2.4846934821359373e-06,
|
1654 |
+
"loss": 2.4723,
|
1655 |
+
"step": 113000
|
1656 |
+
},
|
1657 |
+
{
|
1658 |
+
"epoch": 8.796403937068899,
|
1659 |
+
"grad_norm": 6.297511100769043,
|
1660 |
+
"learning_rate": 2.4071921258622026e-06,
|
1661 |
+
"loss": 2.463,
|
1662 |
+
"step": 113500
|
1663 |
+
},
|
1664 |
+
{
|
1665 |
+
"epoch": 8.835154615205767,
|
1666 |
+
"grad_norm": 6.689960479736328,
|
1667 |
+
"learning_rate": 2.329690769588468e-06,
|
1668 |
+
"loss": 2.4621,
|
1669 |
+
"step": 114000
|
1670 |
+
},
|
1671 |
+
{
|
1672 |
+
"epoch": 8.873905293342634,
|
1673 |
+
"grad_norm": 6.450560569763184,
|
1674 |
+
"learning_rate": 2.252189413314733e-06,
|
1675 |
+
"loss": 2.4559,
|
1676 |
+
"step": 114500
|
1677 |
+
},
|
1678 |
+
{
|
1679 |
+
"epoch": 8.9126559714795,
|
1680 |
+
"grad_norm": 6.459935665130615,
|
1681 |
+
"learning_rate": 2.1746880570409984e-06,
|
1682 |
+
"loss": 2.4646,
|
1683 |
+
"step": 115000
|
1684 |
+
},
|
1685 |
+
{
|
1686 |
+
"epoch": 8.951406649616368,
|
1687 |
+
"grad_norm": 6.182426452636719,
|
1688 |
+
"learning_rate": 2.0971867007672637e-06,
|
1689 |
+
"loss": 2.4665,
|
1690 |
+
"step": 115500
|
1691 |
+
},
|
1692 |
+
{
|
1693 |
+
"epoch": 8.990157327753236,
|
1694 |
+
"grad_norm": 7.122648239135742,
|
1695 |
+
"learning_rate": 2.019685344493529e-06,
|
1696 |
+
"loss": 2.475,
|
1697 |
+
"step": 116000
|
1698 |
+
},
|
1699 |
+
{
|
1700 |
+
"epoch": 9.0,
|
1701 |
+
"eval_loss": 2.406507968902588,
|
1702 |
+
"eval_runtime": 258.9009,
|
1703 |
+
"eval_samples_per_second": 797.467,
|
1704 |
+
"eval_steps_per_second": 12.464,
|
1705 |
+
"step": 116127
|
1706 |
+
},
|
1707 |
+
{
|
1708 |
+
"epoch": 9.028908005890104,
|
1709 |
+
"grad_norm": 7.267585754394531,
|
1710 |
+
"learning_rate": 1.942183988219794e-06,
|
1711 |
+
"loss": 2.447,
|
1712 |
+
"step": 116500
|
1713 |
+
},
|
1714 |
+
{
|
1715 |
+
"epoch": 9.06765868402697,
|
1716 |
+
"grad_norm": 6.2447991371154785,
|
1717 |
+
"learning_rate": 1.8646826319460593e-06,
|
1718 |
+
"loss": 2.4609,
|
1719 |
+
"step": 117000
|
1720 |
+
},
|
1721 |
+
{
|
1722 |
+
"epoch": 9.106409362163838,
|
1723 |
+
"grad_norm": 6.521481037139893,
|
1724 |
+
"learning_rate": 1.7871812756723245e-06,
|
1725 |
+
"loss": 2.4418,
|
1726 |
+
"step": 117500
|
1727 |
+
},
|
1728 |
+
{
|
1729 |
+
"epoch": 9.145160040300706,
|
1730 |
+
"grad_norm": 6.647397041320801,
|
1731 |
+
"learning_rate": 1.7096799193985896e-06,
|
1732 |
+
"loss": 2.4665,
|
1733 |
+
"step": 118000
|
1734 |
+
},
|
1735 |
+
{
|
1736 |
+
"epoch": 9.183910718437573,
|
1737 |
+
"grad_norm": 6.247033596038818,
|
1738 |
+
"learning_rate": 1.6321785631248548e-06,
|
1739 |
+
"loss": 2.4647,
|
1740 |
+
"step": 118500
|
1741 |
+
},
|
1742 |
+
{
|
1743 |
+
"epoch": 9.22266139657444,
|
1744 |
+
"grad_norm": 6.595357894897461,
|
1745 |
+
"learning_rate": 1.5546772068511201e-06,
|
1746 |
+
"loss": 2.4705,
|
1747 |
+
"step": 119000
|
1748 |
+
},
|
1749 |
+
{
|
1750 |
+
"epoch": 9.261412074711307,
|
1751 |
+
"grad_norm": 8.117677688598633,
|
1752 |
+
"learning_rate": 1.4771758505773854e-06,
|
1753 |
+
"loss": 2.4629,
|
1754 |
+
"step": 119500
|
1755 |
+
},
|
1756 |
+
{
|
1757 |
+
"epoch": 9.300162752848175,
|
1758 |
+
"grad_norm": 6.991618633270264,
|
1759 |
+
"learning_rate": 1.3996744943036504e-06,
|
1760 |
+
"loss": 2.4498,
|
1761 |
+
"step": 120000
|
1762 |
+
},
|
1763 |
+
{
|
1764 |
+
"epoch": 9.338913430985043,
|
1765 |
+
"grad_norm": 6.236393451690674,
|
1766 |
+
"learning_rate": 1.3221731380299157e-06,
|
1767 |
+
"loss": 2.467,
|
1768 |
+
"step": 120500
|
1769 |
+
},
|
1770 |
+
{
|
1771 |
+
"epoch": 9.377664109121909,
|
1772 |
+
"grad_norm": 6.595478534698486,
|
1773 |
+
"learning_rate": 1.2446717817561808e-06,
|
1774 |
+
"loss": 2.4547,
|
1775 |
+
"step": 121000
|
1776 |
+
},
|
1777 |
+
{
|
1778 |
+
"epoch": 9.416414787258777,
|
1779 |
+
"grad_norm": 7.194475173950195,
|
1780 |
+
"learning_rate": 1.167170425482446e-06,
|
1781 |
+
"loss": 2.4669,
|
1782 |
+
"step": 121500
|
1783 |
+
},
|
1784 |
+
{
|
1785 |
+
"epoch": 9.455165465395645,
|
1786 |
+
"grad_norm": 6.341099262237549,
|
1787 |
+
"learning_rate": 1.0896690692087113e-06,
|
1788 |
+
"loss": 2.4661,
|
1789 |
+
"step": 122000
|
1790 |
+
},
|
1791 |
+
{
|
1792 |
+
"epoch": 9.493916143532513,
|
1793 |
+
"grad_norm": 7.257521629333496,
|
1794 |
+
"learning_rate": 1.0121677129349766e-06,
|
1795 |
+
"loss": 2.4629,
|
1796 |
+
"step": 122500
|
1797 |
+
},
|
1798 |
+
{
|
1799 |
+
"epoch": 9.532666821669379,
|
1800 |
+
"grad_norm": 6.399875164031982,
|
1801 |
+
"learning_rate": 9.346663566612417e-07,
|
1802 |
+
"loss": 2.4555,
|
1803 |
+
"step": 123000
|
1804 |
+
},
|
1805 |
+
{
|
1806 |
+
"epoch": 9.571417499806246,
|
1807 |
+
"grad_norm": 7.292248249053955,
|
1808 |
+
"learning_rate": 8.571650003875069e-07,
|
1809 |
+
"loss": 2.4646,
|
1810 |
+
"step": 123500
|
1811 |
+
},
|
1812 |
+
{
|
1813 |
+
"epoch": 9.610168177943114,
|
1814 |
+
"grad_norm": 6.8132548332214355,
|
1815 |
+
"learning_rate": 7.79663644113772e-07,
|
1816 |
+
"loss": 2.4521,
|
1817 |
+
"step": 124000
|
1818 |
+
},
|
1819 |
+
{
|
1820 |
+
"epoch": 9.648918856079982,
|
1821 |
+
"grad_norm": 6.302210330963135,
|
1822 |
+
"learning_rate": 7.021622878400372e-07,
|
1823 |
+
"loss": 2.451,
|
1824 |
+
"step": 124500
|
1825 |
+
},
|
1826 |
+
{
|
1827 |
+
"epoch": 9.687669534216848,
|
1828 |
+
"grad_norm": 6.902337551116943,
|
1829 |
+
"learning_rate": 6.246609315663025e-07,
|
1830 |
+
"loss": 2.4515,
|
1831 |
+
"step": 125000
|
1832 |
+
},
|
1833 |
+
{
|
1834 |
+
"epoch": 9.726420212353716,
|
1835 |
+
"grad_norm": 6.4049296379089355,
|
1836 |
+
"learning_rate": 5.471595752925676e-07,
|
1837 |
+
"loss": 2.454,
|
1838 |
+
"step": 125500
|
1839 |
+
},
|
1840 |
+
{
|
1841 |
+
"epoch": 9.765170890490584,
|
1842 |
+
"grad_norm": 7.109240531921387,
|
1843 |
+
"learning_rate": 4.6965821901883286e-07,
|
1844 |
+
"loss": 2.4379,
|
1845 |
+
"step": 126000
|
1846 |
+
},
|
1847 |
+
{
|
1848 |
+
"epoch": 9.803921568627452,
|
1849 |
+
"grad_norm": 6.1289873123168945,
|
1850 |
+
"learning_rate": 3.921568627450981e-07,
|
1851 |
+
"loss": 2.4438,
|
1852 |
+
"step": 126500
|
1853 |
+
},
|
1854 |
+
{
|
1855 |
+
"epoch": 9.842672246764318,
|
1856 |
+
"grad_norm": 6.873955726623535,
|
1857 |
+
"learning_rate": 3.146555064713633e-07,
|
1858 |
+
"loss": 2.4526,
|
1859 |
+
"step": 127000
|
1860 |
+
},
|
1861 |
+
{
|
1862 |
+
"epoch": 9.881422924901186,
|
1863 |
+
"grad_norm": 6.842904090881348,
|
1864 |
+
"learning_rate": 2.3715415019762845e-07,
|
1865 |
+
"loss": 2.4471,
|
1866 |
+
"step": 127500
|
1867 |
+
},
|
1868 |
+
{
|
1869 |
+
"epoch": 9.920173603038053,
|
1870 |
+
"grad_norm": 9.636740684509277,
|
1871 |
+
"learning_rate": 1.5965279392389367e-07,
|
1872 |
+
"loss": 2.4469,
|
1873 |
+
"step": 128000
|
1874 |
+
},
|
1875 |
+
{
|
1876 |
+
"epoch": 9.958924281174921,
|
1877 |
+
"grad_norm": 6.161515235900879,
|
1878 |
+
"learning_rate": 8.21514376501589e-08,
|
1879 |
+
"loss": 2.4608,
|
1880 |
+
"step": 128500
|
1881 |
+
},
|
1882 |
+
{
|
1883 |
+
"epoch": 9.997674959311787,
|
1884 |
+
"grad_norm": 6.582516193389893,
|
1885 |
+
"learning_rate": 4.6500813764240875e-09,
|
1886 |
+
"loss": 2.4411,
|
1887 |
+
"step": 129000
|
1888 |
+
},
|
1889 |
+
{
|
1890 |
+
"epoch": 10.0,
|
1891 |
+
"eval_loss": 2.3977291584014893,
|
1892 |
+
"eval_runtime": 258.9982,
|
1893 |
+
"eval_samples_per_second": 797.168,
|
1894 |
+
"eval_steps_per_second": 12.46,
|
1895 |
+
"step": 129030
|
1896 |
+
}
|
1897 |
+
],
|
1898 |
+
"logging_steps": 500,
|
1899 |
+
"max_steps": 129030,
|
1900 |
+
"num_input_tokens_seen": 0,
|
1901 |
+
"num_train_epochs": 10,
|
1902 |
+
"save_steps": 500,
|
1903 |
+
"stateful_callbacks": {
|
1904 |
+
"TrainerControl": {
|
1905 |
+
"args": {
|
1906 |
+
"should_epoch_stop": false,
|
1907 |
+
"should_evaluate": false,
|
1908 |
+
"should_log": false,
|
1909 |
+
"should_save": true,
|
1910 |
+
"should_training_stop": true
|
1911 |
+
},
|
1912 |
+
"attributes": {}
|
1913 |
+
}
|
1914 |
+
},
|
1915 |
+
"total_flos": 3.235049087048045e+17,
|
1916 |
+
"train_batch_size": 64,
|
1917 |
+
"trial_name": null,
|
1918 |
+
"trial_params": null
|
1919 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7c29e483a17b334874faa05ddef49273d268f86505fb1cbbeea2a4cddc04a790
|
3 |
+
size 5048
|