Upload folder using huggingface_hub
Browse files- adapter_model.safetensors +1 -1
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +2133 -0
adapter_model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3705288
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:dc0bc753d549873778279b4352a2fef1178f8b18e419b345700de66966f94f38
|
| 3 |
size 3705288
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:399e81a7072d05f6bd91a40835348ac36d6e7a76eea6f1c41f12f661c7013db2
|
| 3 |
+
size 2213690
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c282cd48f6ae31c73c753236ccc00d4589da3b2940ecbca74e49b6784205ce15
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:4b8618bf2b16b696607893880c2e64585cb7c07b1c54a6d87b831038f8cd2066
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2133 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 29.286150091519218,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 6000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.09762050030506407,
|
| 13 |
+
"grad_norm": 0.5132213234901428,
|
| 14 |
+
"learning_rate": 4e-05,
|
| 15 |
+
"loss": 2.1313,
|
| 16 |
+
"step": 20
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.19524100061012814,
|
| 20 |
+
"grad_norm": 0.5252098441123962,
|
| 21 |
+
"learning_rate": 8e-05,
|
| 22 |
+
"loss": 2.0977,
|
| 23 |
+
"step": 40
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.2928615009151922,
|
| 27 |
+
"grad_norm": 0.5683884024620056,
|
| 28 |
+
"learning_rate": 0.00012,
|
| 29 |
+
"loss": 1.9168,
|
| 30 |
+
"step": 60
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.3904820012202563,
|
| 34 |
+
"grad_norm": 0.6169227957725525,
|
| 35 |
+
"learning_rate": 0.00016,
|
| 36 |
+
"loss": 1.8795,
|
| 37 |
+
"step": 80
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.4881025015253203,
|
| 41 |
+
"grad_norm": 0.8657425045967102,
|
| 42 |
+
"learning_rate": 0.0002,
|
| 43 |
+
"loss": 1.8257,
|
| 44 |
+
"step": 100
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.5857230018303844,
|
| 48 |
+
"grad_norm": 1.1708447933197021,
|
| 49 |
+
"learning_rate": 0.00019980295566502464,
|
| 50 |
+
"loss": 1.8,
|
| 51 |
+
"step": 120
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.6833435021354485,
|
| 55 |
+
"grad_norm": 1.4427006244659424,
|
| 56 |
+
"learning_rate": 0.00019960591133004926,
|
| 57 |
+
"loss": 1.75,
|
| 58 |
+
"step": 140
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.7809640024405126,
|
| 62 |
+
"grad_norm": 1.190605640411377,
|
| 63 |
+
"learning_rate": 0.00019940886699507392,
|
| 64 |
+
"loss": 1.6796,
|
| 65 |
+
"step": 160
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.8785845027455765,
|
| 69 |
+
"grad_norm": 0.9052116274833679,
|
| 70 |
+
"learning_rate": 0.00019921182266009852,
|
| 71 |
+
"loss": 1.6721,
|
| 72 |
+
"step": 180
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.9762050030506406,
|
| 76 |
+
"grad_norm": 1.3986254930496216,
|
| 77 |
+
"learning_rate": 0.00019901477832512317,
|
| 78 |
+
"loss": 1.5752,
|
| 79 |
+
"step": 200
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 1.0738255033557047,
|
| 83 |
+
"grad_norm": 1.1646902561187744,
|
| 84 |
+
"learning_rate": 0.0001988177339901478,
|
| 85 |
+
"loss": 1.6187,
|
| 86 |
+
"step": 220
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 1.1714460036607688,
|
| 90 |
+
"grad_norm": 1.3168349266052246,
|
| 91 |
+
"learning_rate": 0.00019862068965517243,
|
| 92 |
+
"loss": 1.518,
|
| 93 |
+
"step": 240
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 1.2690665039658329,
|
| 97 |
+
"grad_norm": 1.5947434902191162,
|
| 98 |
+
"learning_rate": 0.00019842364532019705,
|
| 99 |
+
"loss": 1.5919,
|
| 100 |
+
"step": 260
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 1.366687004270897,
|
| 104 |
+
"grad_norm": 1.550020694732666,
|
| 105 |
+
"learning_rate": 0.00019822660098522168,
|
| 106 |
+
"loss": 1.5565,
|
| 107 |
+
"step": 280
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 1.4643075045759608,
|
| 111 |
+
"grad_norm": 1.587833285331726,
|
| 112 |
+
"learning_rate": 0.0001980295566502463,
|
| 113 |
+
"loss": 1.4928,
|
| 114 |
+
"step": 300
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 1.561928004881025,
|
| 118 |
+
"grad_norm": 1.3913565874099731,
|
| 119 |
+
"learning_rate": 0.00019783251231527093,
|
| 120 |
+
"loss": 1.4603,
|
| 121 |
+
"step": 320
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 1.659548505186089,
|
| 125 |
+
"grad_norm": 1.6511396169662476,
|
| 126 |
+
"learning_rate": 0.00019763546798029556,
|
| 127 |
+
"loss": 1.3946,
|
| 128 |
+
"step": 340
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 1.757169005491153,
|
| 132 |
+
"grad_norm": 1.6132158041000366,
|
| 133 |
+
"learning_rate": 0.00019743842364532022,
|
| 134 |
+
"loss": 1.4182,
|
| 135 |
+
"step": 360
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 1.8547895057962172,
|
| 139 |
+
"grad_norm": 2.1648366451263428,
|
| 140 |
+
"learning_rate": 0.00019724137931034484,
|
| 141 |
+
"loss": 1.4717,
|
| 142 |
+
"step": 380
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 1.9524100061012812,
|
| 146 |
+
"grad_norm": 1.5196492671966553,
|
| 147 |
+
"learning_rate": 0.00019704433497536947,
|
| 148 |
+
"loss": 1.4192,
|
| 149 |
+
"step": 400
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 2.0500305064063453,
|
| 153 |
+
"grad_norm": 1.676941990852356,
|
| 154 |
+
"learning_rate": 0.0001968472906403941,
|
| 155 |
+
"loss": 1.378,
|
| 156 |
+
"step": 420
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 2.1476510067114094,
|
| 160 |
+
"grad_norm": 1.834902286529541,
|
| 161 |
+
"learning_rate": 0.00019665024630541872,
|
| 162 |
+
"loss": 1.3741,
|
| 163 |
+
"step": 440
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 2.2452715070164735,
|
| 167 |
+
"grad_norm": 1.9870941638946533,
|
| 168 |
+
"learning_rate": 0.00019645320197044338,
|
| 169 |
+
"loss": 1.3165,
|
| 170 |
+
"step": 460
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 2.3428920073215376,
|
| 174 |
+
"grad_norm": 1.682267427444458,
|
| 175 |
+
"learning_rate": 0.00019625615763546798,
|
| 176 |
+
"loss": 1.3828,
|
| 177 |
+
"step": 480
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 2.4405125076266017,
|
| 181 |
+
"grad_norm": 2.291842222213745,
|
| 182 |
+
"learning_rate": 0.00019605911330049263,
|
| 183 |
+
"loss": 1.3857,
|
| 184 |
+
"step": 500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 2.5381330079316657,
|
| 188 |
+
"grad_norm": 2.0962560176849365,
|
| 189 |
+
"learning_rate": 0.00019586206896551723,
|
| 190 |
+
"loss": 1.2759,
|
| 191 |
+
"step": 520
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 2.63575350823673,
|
| 195 |
+
"grad_norm": 1.6451084613800049,
|
| 196 |
+
"learning_rate": 0.0001956650246305419,
|
| 197 |
+
"loss": 1.3086,
|
| 198 |
+
"step": 540
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 2.733374008541794,
|
| 202 |
+
"grad_norm": 1.8540750741958618,
|
| 203 |
+
"learning_rate": 0.00019546798029556651,
|
| 204 |
+
"loss": 1.3416,
|
| 205 |
+
"step": 560
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 2.830994508846858,
|
| 209 |
+
"grad_norm": 1.8368124961853027,
|
| 210 |
+
"learning_rate": 0.00019527093596059114,
|
| 211 |
+
"loss": 1.2716,
|
| 212 |
+
"step": 580
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 2.9286150091519216,
|
| 216 |
+
"grad_norm": 3.194183588027954,
|
| 217 |
+
"learning_rate": 0.00019507389162561577,
|
| 218 |
+
"loss": 1.2964,
|
| 219 |
+
"step": 600
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 3.026235509456986,
|
| 223 |
+
"grad_norm": 1.751219630241394,
|
| 224 |
+
"learning_rate": 0.0001948768472906404,
|
| 225 |
+
"loss": 1.3046,
|
| 226 |
+
"step": 620
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 3.1238560097620502,
|
| 230 |
+
"grad_norm": 1.834823489189148,
|
| 231 |
+
"learning_rate": 0.00019467980295566505,
|
| 232 |
+
"loss": 1.2149,
|
| 233 |
+
"step": 640
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 3.221476510067114,
|
| 237 |
+
"grad_norm": 1.9562243223190308,
|
| 238 |
+
"learning_rate": 0.00019448275862068965,
|
| 239 |
+
"loss": 1.209,
|
| 240 |
+
"step": 660
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 3.319097010372178,
|
| 244 |
+
"grad_norm": 2.012437582015991,
|
| 245 |
+
"learning_rate": 0.0001942857142857143,
|
| 246 |
+
"loss": 1.2196,
|
| 247 |
+
"step": 680
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 3.416717510677242,
|
| 251 |
+
"grad_norm": 2.47426176071167,
|
| 252 |
+
"learning_rate": 0.00019408866995073893,
|
| 253 |
+
"loss": 1.2974,
|
| 254 |
+
"step": 700
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 3.514338010982306,
|
| 258 |
+
"grad_norm": 2.1828153133392334,
|
| 259 |
+
"learning_rate": 0.00019389162561576356,
|
| 260 |
+
"loss": 1.2287,
|
| 261 |
+
"step": 720
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 3.61195851128737,
|
| 265 |
+
"grad_norm": 2.335744619369507,
|
| 266 |
+
"learning_rate": 0.00019369458128078818,
|
| 267 |
+
"loss": 1.2018,
|
| 268 |
+
"step": 740
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 3.7095790115924343,
|
| 272 |
+
"grad_norm": 1.731418490409851,
|
| 273 |
+
"learning_rate": 0.0001934975369458128,
|
| 274 |
+
"loss": 1.2568,
|
| 275 |
+
"step": 760
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 3.8071995118974984,
|
| 279 |
+
"grad_norm": 2.0934510231018066,
|
| 280 |
+
"learning_rate": 0.00019330049261083744,
|
| 281 |
+
"loss": 1.2206,
|
| 282 |
+
"step": 780
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 3.9048200122025625,
|
| 286 |
+
"grad_norm": 2.2060680389404297,
|
| 287 |
+
"learning_rate": 0.0001931034482758621,
|
| 288 |
+
"loss": 1.1898,
|
| 289 |
+
"step": 800
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 4.002440512507627,
|
| 293 |
+
"grad_norm": 3.0342836380004883,
|
| 294 |
+
"learning_rate": 0.0001929064039408867,
|
| 295 |
+
"loss": 1.2248,
|
| 296 |
+
"step": 820
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 4.100061012812691,
|
| 300 |
+
"grad_norm": 2.1768083572387695,
|
| 301 |
+
"learning_rate": 0.00019270935960591135,
|
| 302 |
+
"loss": 1.1721,
|
| 303 |
+
"step": 840
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 4.197681513117755,
|
| 307 |
+
"grad_norm": 2.2883739471435547,
|
| 308 |
+
"learning_rate": 0.00019251231527093597,
|
| 309 |
+
"loss": 1.1117,
|
| 310 |
+
"step": 860
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 4.295302013422819,
|
| 314 |
+
"grad_norm": 2.45024037361145,
|
| 315 |
+
"learning_rate": 0.0001923152709359606,
|
| 316 |
+
"loss": 1.1392,
|
| 317 |
+
"step": 880
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 4.392922513727883,
|
| 321 |
+
"grad_norm": 1.9696956872940063,
|
| 322 |
+
"learning_rate": 0.00019211822660098523,
|
| 323 |
+
"loss": 1.0958,
|
| 324 |
+
"step": 900
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 4.490543014032947,
|
| 328 |
+
"grad_norm": 2.3901145458221436,
|
| 329 |
+
"learning_rate": 0.00019192118226600986,
|
| 330 |
+
"loss": 1.1693,
|
| 331 |
+
"step": 920
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 4.588163514338011,
|
| 335 |
+
"grad_norm": 2.003532648086548,
|
| 336 |
+
"learning_rate": 0.0001917241379310345,
|
| 337 |
+
"loss": 1.1301,
|
| 338 |
+
"step": 940
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 4.685784014643075,
|
| 342 |
+
"grad_norm": 1.990051031112671,
|
| 343 |
+
"learning_rate": 0.0001915270935960591,
|
| 344 |
+
"loss": 1.1315,
|
| 345 |
+
"step": 960
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 4.783404514948139,
|
| 349 |
+
"grad_norm": 2.517423152923584,
|
| 350 |
+
"learning_rate": 0.00019133004926108376,
|
| 351 |
+
"loss": 1.208,
|
| 352 |
+
"step": 980
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 4.881025015253203,
|
| 356 |
+
"grad_norm": 2.311152458190918,
|
| 357 |
+
"learning_rate": 0.0001911330049261084,
|
| 358 |
+
"loss": 1.1494,
|
| 359 |
+
"step": 1000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 4.978645515558267,
|
| 363 |
+
"grad_norm": 2.327719211578369,
|
| 364 |
+
"learning_rate": 0.00019093596059113302,
|
| 365 |
+
"loss": 1.1459,
|
| 366 |
+
"step": 1020
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 5.0762660158633315,
|
| 370 |
+
"grad_norm": 3.1623075008392334,
|
| 371 |
+
"learning_rate": 0.00019073891625615765,
|
| 372 |
+
"loss": 1.1417,
|
| 373 |
+
"step": 1040
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 5.173886516168396,
|
| 377 |
+
"grad_norm": 2.418928384780884,
|
| 378 |
+
"learning_rate": 0.00019054187192118227,
|
| 379 |
+
"loss": 1.091,
|
| 380 |
+
"step": 1060
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 5.27150701647346,
|
| 384 |
+
"grad_norm": 2.6035215854644775,
|
| 385 |
+
"learning_rate": 0.0001903448275862069,
|
| 386 |
+
"loss": 1.0851,
|
| 387 |
+
"step": 1080
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 5.369127516778524,
|
| 391 |
+
"grad_norm": 3.089789628982544,
|
| 392 |
+
"learning_rate": 0.00019014778325123153,
|
| 393 |
+
"loss": 1.0592,
|
| 394 |
+
"step": 1100
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 5.466748017083588,
|
| 398 |
+
"grad_norm": 2.885105609893799,
|
| 399 |
+
"learning_rate": 0.00018995073891625615,
|
| 400 |
+
"loss": 1.0781,
|
| 401 |
+
"step": 1120
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 5.564368517388652,
|
| 405 |
+
"grad_norm": 2.3023903369903564,
|
| 406 |
+
"learning_rate": 0.0001897536945812808,
|
| 407 |
+
"loss": 1.0706,
|
| 408 |
+
"step": 1140
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 5.661989017693716,
|
| 412 |
+
"grad_norm": 2.873560905456543,
|
| 413 |
+
"learning_rate": 0.00018955665024630543,
|
| 414 |
+
"loss": 1.092,
|
| 415 |
+
"step": 1160
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 5.75960951799878,
|
| 419 |
+
"grad_norm": 2.4178314208984375,
|
| 420 |
+
"learning_rate": 0.00018935960591133006,
|
| 421 |
+
"loss": 1.0896,
|
| 422 |
+
"step": 1180
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 5.857230018303844,
|
| 426 |
+
"grad_norm": 2.150630474090576,
|
| 427 |
+
"learning_rate": 0.0001891625615763547,
|
| 428 |
+
"loss": 1.1056,
|
| 429 |
+
"step": 1200
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 5.954850518608908,
|
| 433 |
+
"grad_norm": 3.347947359085083,
|
| 434 |
+
"learning_rate": 0.00018896551724137932,
|
| 435 |
+
"loss": 1.0832,
|
| 436 |
+
"step": 1220
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 6.052471018913972,
|
| 440 |
+
"grad_norm": 2.737258195877075,
|
| 441 |
+
"learning_rate": 0.00018876847290640397,
|
| 442 |
+
"loss": 1.0573,
|
| 443 |
+
"step": 1240
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 6.150091519219036,
|
| 447 |
+
"grad_norm": 2.3305180072784424,
|
| 448 |
+
"learning_rate": 0.00018857142857142857,
|
| 449 |
+
"loss": 0.9491,
|
| 450 |
+
"step": 1260
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 6.2477120195241005,
|
| 454 |
+
"grad_norm": 3.0475850105285645,
|
| 455 |
+
"learning_rate": 0.00018837438423645322,
|
| 456 |
+
"loss": 1.0221,
|
| 457 |
+
"step": 1280
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 6.345332519829164,
|
| 461 |
+
"grad_norm": 2.7141025066375732,
|
| 462 |
+
"learning_rate": 0.00018817733990147782,
|
| 463 |
+
"loss": 1.0566,
|
| 464 |
+
"step": 1300
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 6.442953020134228,
|
| 468 |
+
"grad_norm": 2.931290626525879,
|
| 469 |
+
"learning_rate": 0.00018798029556650248,
|
| 470 |
+
"loss": 1.0178,
|
| 471 |
+
"step": 1320
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 6.540573520439292,
|
| 475 |
+
"grad_norm": 2.9428722858428955,
|
| 476 |
+
"learning_rate": 0.0001877832512315271,
|
| 477 |
+
"loss": 1.0642,
|
| 478 |
+
"step": 1340
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 6.638194020744356,
|
| 482 |
+
"grad_norm": 2.452775001525879,
|
| 483 |
+
"learning_rate": 0.00018758620689655173,
|
| 484 |
+
"loss": 1.0928,
|
| 485 |
+
"step": 1360
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 6.73581452104942,
|
| 489 |
+
"grad_norm": 3.380108594894409,
|
| 490 |
+
"learning_rate": 0.00018738916256157636,
|
| 491 |
+
"loss": 1.0091,
|
| 492 |
+
"step": 1380
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 6.833435021354484,
|
| 496 |
+
"grad_norm": 2.9912617206573486,
|
| 497 |
+
"learning_rate": 0.000187192118226601,
|
| 498 |
+
"loss": 0.9958,
|
| 499 |
+
"step": 1400
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 6.931055521659548,
|
| 503 |
+
"grad_norm": 2.5559194087982178,
|
| 504 |
+
"learning_rate": 0.00018699507389162561,
|
| 505 |
+
"loss": 1.0891,
|
| 506 |
+
"step": 1420
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 7.028676021964612,
|
| 510 |
+
"grad_norm": 2.728987693786621,
|
| 511 |
+
"learning_rate": 0.00018679802955665024,
|
| 512 |
+
"loss": 0.9723,
|
| 513 |
+
"step": 1440
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 7.126296522269676,
|
| 517 |
+
"grad_norm": 2.4664106369018555,
|
| 518 |
+
"learning_rate": 0.0001866009852216749,
|
| 519 |
+
"loss": 0.9712,
|
| 520 |
+
"step": 1460
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 7.22391702257474,
|
| 524 |
+
"grad_norm": 2.6810712814331055,
|
| 525 |
+
"learning_rate": 0.00018640394088669952,
|
| 526 |
+
"loss": 0.9408,
|
| 527 |
+
"step": 1480
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 7.3215375228798045,
|
| 531 |
+
"grad_norm": 2.690723419189453,
|
| 532 |
+
"learning_rate": 0.00018620689655172415,
|
| 533 |
+
"loss": 0.9579,
|
| 534 |
+
"step": 1500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 7.419158023184869,
|
| 538 |
+
"grad_norm": 2.751676321029663,
|
| 539 |
+
"learning_rate": 0.00018600985221674878,
|
| 540 |
+
"loss": 0.9959,
|
| 541 |
+
"step": 1520
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 7.516778523489933,
|
| 545 |
+
"grad_norm": 2.6251280307769775,
|
| 546 |
+
"learning_rate": 0.0001858128078817734,
|
| 547 |
+
"loss": 0.9908,
|
| 548 |
+
"step": 1540
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 7.614399023794997,
|
| 552 |
+
"grad_norm": 2.897099733352661,
|
| 553 |
+
"learning_rate": 0.00018561576354679803,
|
| 554 |
+
"loss": 0.9631,
|
| 555 |
+
"step": 1560
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 7.712019524100061,
|
| 559 |
+
"grad_norm": 2.0911786556243896,
|
| 560 |
+
"learning_rate": 0.00018541871921182269,
|
| 561 |
+
"loss": 0.9963,
|
| 562 |
+
"step": 1580
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 7.809640024405125,
|
| 566 |
+
"grad_norm": 2.6954994201660156,
|
| 567 |
+
"learning_rate": 0.00018522167487684729,
|
| 568 |
+
"loss": 1.0134,
|
| 569 |
+
"step": 1600
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 7.907260524710189,
|
| 573 |
+
"grad_norm": 2.8063347339630127,
|
| 574 |
+
"learning_rate": 0.00018502463054187194,
|
| 575 |
+
"loss": 0.9732,
|
| 576 |
+
"step": 1620
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 8.004881025015253,
|
| 580 |
+
"grad_norm": 2.0492053031921387,
|
| 581 |
+
"learning_rate": 0.00018482758620689654,
|
| 582 |
+
"loss": 1.0361,
|
| 583 |
+
"step": 1640
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 8.102501525320317,
|
| 587 |
+
"grad_norm": 3.0692152976989746,
|
| 588 |
+
"learning_rate": 0.0001846305418719212,
|
| 589 |
+
"loss": 0.9293,
|
| 590 |
+
"step": 1660
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 8.200122025625381,
|
| 594 |
+
"grad_norm": 2.7933707237243652,
|
| 595 |
+
"learning_rate": 0.00018443349753694582,
|
| 596 |
+
"loss": 0.9101,
|
| 597 |
+
"step": 1680
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 8.297742525930445,
|
| 601 |
+
"grad_norm": 3.628946542739868,
|
| 602 |
+
"learning_rate": 0.00018423645320197045,
|
| 603 |
+
"loss": 0.9215,
|
| 604 |
+
"step": 1700
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 8.39536302623551,
|
| 608 |
+
"grad_norm": 2.892118215560913,
|
| 609 |
+
"learning_rate": 0.0001840394088669951,
|
| 610 |
+
"loss": 0.9008,
|
| 611 |
+
"step": 1720
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 8.492983526540574,
|
| 615 |
+
"grad_norm": 3.5419254302978516,
|
| 616 |
+
"learning_rate": 0.0001838423645320197,
|
| 617 |
+
"loss": 0.9577,
|
| 618 |
+
"step": 1740
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 8.590604026845638,
|
| 622 |
+
"grad_norm": 2.785578489303589,
|
| 623 |
+
"learning_rate": 0.00018364532019704436,
|
| 624 |
+
"loss": 0.8979,
|
| 625 |
+
"step": 1760
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 8.688224527150702,
|
| 629 |
+
"grad_norm": 3.6454851627349854,
|
| 630 |
+
"learning_rate": 0.00018344827586206896,
|
| 631 |
+
"loss": 0.9424,
|
| 632 |
+
"step": 1780
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 8.785845027455766,
|
| 636 |
+
"grad_norm": 3.1077752113342285,
|
| 637 |
+
"learning_rate": 0.0001832512315270936,
|
| 638 |
+
"loss": 0.976,
|
| 639 |
+
"step": 1800
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 8.88346552776083,
|
| 643 |
+
"grad_norm": 2.1347529888153076,
|
| 644 |
+
"learning_rate": 0.00018305418719211824,
|
| 645 |
+
"loss": 0.9889,
|
| 646 |
+
"step": 1820
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 8.981086028065894,
|
| 650 |
+
"grad_norm": 1.8763928413391113,
|
| 651 |
+
"learning_rate": 0.00018285714285714286,
|
| 652 |
+
"loss": 0.9595,
|
| 653 |
+
"step": 1840
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 9.078706528370958,
|
| 657 |
+
"grad_norm": 2.5731394290924072,
|
| 658 |
+
"learning_rate": 0.0001826600985221675,
|
| 659 |
+
"loss": 0.9346,
|
| 660 |
+
"step": 1860
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 9.176327028676022,
|
| 664 |
+
"grad_norm": 2.75944447517395,
|
| 665 |
+
"learning_rate": 0.00018246305418719212,
|
| 666 |
+
"loss": 0.8926,
|
| 667 |
+
"step": 1880
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 9.273947528981086,
|
| 671 |
+
"grad_norm": 2.7548296451568604,
|
| 672 |
+
"learning_rate": 0.00018226600985221675,
|
| 673 |
+
"loss": 0.8835,
|
| 674 |
+
"step": 1900
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 9.37156802928615,
|
| 678 |
+
"grad_norm": 3.4645333290100098,
|
| 679 |
+
"learning_rate": 0.0001820689655172414,
|
| 680 |
+
"loss": 0.8753,
|
| 681 |
+
"step": 1920
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 9.469188529591214,
|
| 685 |
+
"grad_norm": 2.7922091484069824,
|
| 686 |
+
"learning_rate": 0.00018187192118226603,
|
| 687 |
+
"loss": 0.9187,
|
| 688 |
+
"step": 1940
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 9.566809029896278,
|
| 692 |
+
"grad_norm": 2.257009506225586,
|
| 693 |
+
"learning_rate": 0.00018167487684729065,
|
| 694 |
+
"loss": 0.9294,
|
| 695 |
+
"step": 1960
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 9.664429530201343,
|
| 699 |
+
"grad_norm": 4.195834159851074,
|
| 700 |
+
"learning_rate": 0.00018147783251231528,
|
| 701 |
+
"loss": 0.9022,
|
| 702 |
+
"step": 1980
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 9.762050030506407,
|
| 706 |
+
"grad_norm": 2.8687057495117188,
|
| 707 |
+
"learning_rate": 0.0001812807881773399,
|
| 708 |
+
"loss": 0.8744,
|
| 709 |
+
"step": 2000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 9.85967053081147,
|
| 713 |
+
"grad_norm": 3.758493661880493,
|
| 714 |
+
"learning_rate": 0.00018108374384236456,
|
| 715 |
+
"loss": 0.9117,
|
| 716 |
+
"step": 2020
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 9.957291031116535,
|
| 720 |
+
"grad_norm": 3.2609262466430664,
|
| 721 |
+
"learning_rate": 0.00018088669950738916,
|
| 722 |
+
"loss": 0.9261,
|
| 723 |
+
"step": 2040
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 10.054911531421599,
|
| 727 |
+
"grad_norm": 3.5481553077697754,
|
| 728 |
+
"learning_rate": 0.00018068965517241382,
|
| 729 |
+
"loss": 0.8786,
|
| 730 |
+
"step": 2060
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 10.152532031726663,
|
| 734 |
+
"grad_norm": 2.8181192874908447,
|
| 735 |
+
"learning_rate": 0.00018049261083743842,
|
| 736 |
+
"loss": 0.8153,
|
| 737 |
+
"step": 2080
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 10.250152532031727,
|
| 741 |
+
"grad_norm": 2.582590341567993,
|
| 742 |
+
"learning_rate": 0.00018029556650246307,
|
| 743 |
+
"loss": 0.8763,
|
| 744 |
+
"step": 2100
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 10.347773032336791,
|
| 748 |
+
"grad_norm": 2.50076961517334,
|
| 749 |
+
"learning_rate": 0.0001800985221674877,
|
| 750 |
+
"loss": 0.8512,
|
| 751 |
+
"step": 2120
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 10.445393532641855,
|
| 755 |
+
"grad_norm": 3.2371861934661865,
|
| 756 |
+
"learning_rate": 0.00017990147783251232,
|
| 757 |
+
"loss": 0.823,
|
| 758 |
+
"step": 2140
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 10.54301403294692,
|
| 762 |
+
"grad_norm": 2.688570976257324,
|
| 763 |
+
"learning_rate": 0.00017970443349753695,
|
| 764 |
+
"loss": 0.8853,
|
| 765 |
+
"step": 2160
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 10.640634533251983,
|
| 769 |
+
"grad_norm": 2.4727838039398193,
|
| 770 |
+
"learning_rate": 0.00017950738916256158,
|
| 771 |
+
"loss": 0.8257,
|
| 772 |
+
"step": 2180
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 10.738255033557047,
|
| 776 |
+
"grad_norm": 3.330667495727539,
|
| 777 |
+
"learning_rate": 0.0001793103448275862,
|
| 778 |
+
"loss": 0.923,
|
| 779 |
+
"step": 2200
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 10.835875533862112,
|
| 783 |
+
"grad_norm": 2.5213732719421387,
|
| 784 |
+
"learning_rate": 0.00017911330049261083,
|
| 785 |
+
"loss": 0.8946,
|
| 786 |
+
"step": 2220
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 10.933496034167176,
|
| 790 |
+
"grad_norm": 2.6011056900024414,
|
| 791 |
+
"learning_rate": 0.0001789162561576355,
|
| 792 |
+
"loss": 0.9194,
|
| 793 |
+
"step": 2240
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 11.03111653447224,
|
| 797 |
+
"grad_norm": 3.4423539638519287,
|
| 798 |
+
"learning_rate": 0.00017871921182266011,
|
| 799 |
+
"loss": 0.8529,
|
| 800 |
+
"step": 2260
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 11.128737034777304,
|
| 804 |
+
"grad_norm": 3.608583927154541,
|
| 805 |
+
"learning_rate": 0.00017852216748768474,
|
| 806 |
+
"loss": 0.7944,
|
| 807 |
+
"step": 2280
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 11.226357535082368,
|
| 811 |
+
"grad_norm": 2.567775249481201,
|
| 812 |
+
"learning_rate": 0.00017832512315270937,
|
| 813 |
+
"loss": 0.7843,
|
| 814 |
+
"step": 2300
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 11.323978035387432,
|
| 818 |
+
"grad_norm": 3.0681939125061035,
|
| 819 |
+
"learning_rate": 0.000178128078817734,
|
| 820 |
+
"loss": 0.8238,
|
| 821 |
+
"step": 2320
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 11.421598535692496,
|
| 825 |
+
"grad_norm": 2.489577293395996,
|
| 826 |
+
"learning_rate": 0.00017793103448275862,
|
| 827 |
+
"loss": 0.8829,
|
| 828 |
+
"step": 2340
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 11.51921903599756,
|
| 832 |
+
"grad_norm": 2.9147262573242188,
|
| 833 |
+
"learning_rate": 0.00017773399014778328,
|
| 834 |
+
"loss": 0.8246,
|
| 835 |
+
"step": 2360
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 11.616839536302624,
|
| 839 |
+
"grad_norm": 2.5094566345214844,
|
| 840 |
+
"learning_rate": 0.00017753694581280788,
|
| 841 |
+
"loss": 0.8277,
|
| 842 |
+
"step": 2380
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 11.714460036607688,
|
| 846 |
+
"grad_norm": 2.4408226013183594,
|
| 847 |
+
"learning_rate": 0.00017733990147783253,
|
| 848 |
+
"loss": 0.8722,
|
| 849 |
+
"step": 2400
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 11.812080536912752,
|
| 853 |
+
"grad_norm": 2.5982508659362793,
|
| 854 |
+
"learning_rate": 0.00017714285714285713,
|
| 855 |
+
"loss": 0.8285,
|
| 856 |
+
"step": 2420
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 11.909701037217816,
|
| 860 |
+
"grad_norm": 4.408588409423828,
|
| 861 |
+
"learning_rate": 0.00017694581280788179,
|
| 862 |
+
"loss": 0.8042,
|
| 863 |
+
"step": 2440
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 12.00732153752288,
|
| 867 |
+
"grad_norm": 3.4463417530059814,
|
| 868 |
+
"learning_rate": 0.0001767487684729064,
|
| 869 |
+
"loss": 0.8606,
|
| 870 |
+
"step": 2460
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 12.104942037827945,
|
| 874 |
+
"grad_norm": 3.192249059677124,
|
| 875 |
+
"learning_rate": 0.00017655172413793104,
|
| 876 |
+
"loss": 0.7847,
|
| 877 |
+
"step": 2480
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 12.202562538133009,
|
| 881 |
+
"grad_norm": 2.760958671569824,
|
| 882 |
+
"learning_rate": 0.00017635467980295567,
|
| 883 |
+
"loss": 0.7968,
|
| 884 |
+
"step": 2500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 12.300183038438073,
|
| 888 |
+
"grad_norm": 2.8952383995056152,
|
| 889 |
+
"learning_rate": 0.0001761576354679803,
|
| 890 |
+
"loss": 0.8226,
|
| 891 |
+
"step": 2520
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 12.397803538743137,
|
| 895 |
+
"grad_norm": 3.6324946880340576,
|
| 896 |
+
"learning_rate": 0.00017596059113300495,
|
| 897 |
+
"loss": 0.7592,
|
| 898 |
+
"step": 2540
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 12.495424039048201,
|
| 902 |
+
"grad_norm": 4.0287885665893555,
|
| 903 |
+
"learning_rate": 0.00017576354679802955,
|
| 904 |
+
"loss": 0.8112,
|
| 905 |
+
"step": 2560
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 12.593044539353265,
|
| 909 |
+
"grad_norm": 3.1734702587127686,
|
| 910 |
+
"learning_rate": 0.0001755665024630542,
|
| 911 |
+
"loss": 0.7847,
|
| 912 |
+
"step": 2580
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 12.690665039658327,
|
| 916 |
+
"grad_norm": 2.9449315071105957,
|
| 917 |
+
"learning_rate": 0.00017536945812807883,
|
| 918 |
+
"loss": 0.8264,
|
| 919 |
+
"step": 2600
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 12.788285539963393,
|
| 923 |
+
"grad_norm": 3.1391289234161377,
|
| 924 |
+
"learning_rate": 0.00017517241379310346,
|
| 925 |
+
"loss": 0.8058,
|
| 926 |
+
"step": 2620
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 12.885906040268456,
|
| 930 |
+
"grad_norm": 3.2317001819610596,
|
| 931 |
+
"learning_rate": 0.00017497536945812808,
|
| 932 |
+
"loss": 0.767,
|
| 933 |
+
"step": 2640
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 12.98352654057352,
|
| 937 |
+
"grad_norm": 3.2640392780303955,
|
| 938 |
+
"learning_rate": 0.0001747783251231527,
|
| 939 |
+
"loss": 0.8314,
|
| 940 |
+
"step": 2660
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 13.081147040878584,
|
| 944 |
+
"grad_norm": 4.71024227142334,
|
| 945 |
+
"learning_rate": 0.00017458128078817734,
|
| 946 |
+
"loss": 0.756,
|
| 947 |
+
"step": 2680
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 13.178767541183648,
|
| 951 |
+
"grad_norm": 3.621242046356201,
|
| 952 |
+
"learning_rate": 0.000174384236453202,
|
| 953 |
+
"loss": 0.7309,
|
| 954 |
+
"step": 2700
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 13.276388041488712,
|
| 958 |
+
"grad_norm": 3.6408748626708984,
|
| 959 |
+
"learning_rate": 0.00017418719211822662,
|
| 960 |
+
"loss": 0.7143,
|
| 961 |
+
"step": 2720
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 13.374008541793776,
|
| 965 |
+
"grad_norm": 3.296096086502075,
|
| 966 |
+
"learning_rate": 0.00017399014778325125,
|
| 967 |
+
"loss": 0.7965,
|
| 968 |
+
"step": 2740
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 13.47162904209884,
|
| 972 |
+
"grad_norm": 2.74519944190979,
|
| 973 |
+
"learning_rate": 0.00017379310344827587,
|
| 974 |
+
"loss": 0.7654,
|
| 975 |
+
"step": 2760
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 13.569249542403904,
|
| 979 |
+
"grad_norm": 2.9242568016052246,
|
| 980 |
+
"learning_rate": 0.0001735960591133005,
|
| 981 |
+
"loss": 0.7875,
|
| 982 |
+
"step": 2780
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 13.666870042708968,
|
| 986 |
+
"grad_norm": 2.5848984718322754,
|
| 987 |
+
"learning_rate": 0.00017339901477832515,
|
| 988 |
+
"loss": 0.7594,
|
| 989 |
+
"step": 2800
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 13.764490543014032,
|
| 993 |
+
"grad_norm": 3.9295613765716553,
|
| 994 |
+
"learning_rate": 0.00017320197044334975,
|
| 995 |
+
"loss": 0.75,
|
| 996 |
+
"step": 2820
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 13.862111043319096,
|
| 1000 |
+
"grad_norm": 3.6406261920928955,
|
| 1001 |
+
"learning_rate": 0.0001730049261083744,
|
| 1002 |
+
"loss": 0.8149,
|
| 1003 |
+
"step": 2840
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 13.95973154362416,
|
| 1007 |
+
"grad_norm": 3.069199323654175,
|
| 1008 |
+
"learning_rate": 0.000172807881773399,
|
| 1009 |
+
"loss": 0.8217,
|
| 1010 |
+
"step": 2860
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 14.057352043929225,
|
| 1014 |
+
"grad_norm": 2.788712739944458,
|
| 1015 |
+
"learning_rate": 0.00017261083743842366,
|
| 1016 |
+
"loss": 0.7755,
|
| 1017 |
+
"step": 2880
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 14.154972544234289,
|
| 1021 |
+
"grad_norm": 3.468480110168457,
|
| 1022 |
+
"learning_rate": 0.00017241379310344826,
|
| 1023 |
+
"loss": 0.7071,
|
| 1024 |
+
"step": 2900
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 14.252593044539353,
|
| 1028 |
+
"grad_norm": 2.899951696395874,
|
| 1029 |
+
"learning_rate": 0.00017221674876847292,
|
| 1030 |
+
"loss": 0.7368,
|
| 1031 |
+
"step": 2920
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 14.350213544844417,
|
| 1035 |
+
"grad_norm": 3.6109790802001953,
|
| 1036 |
+
"learning_rate": 0.00017201970443349754,
|
| 1037 |
+
"loss": 0.7012,
|
| 1038 |
+
"step": 2940
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 14.44783404514948,
|
| 1042 |
+
"grad_norm": 3.448408842086792,
|
| 1043 |
+
"learning_rate": 0.00017182266009852217,
|
| 1044 |
+
"loss": 0.743,
|
| 1045 |
+
"step": 2960
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 14.545454545454545,
|
| 1049 |
+
"grad_norm": 2.819427013397217,
|
| 1050 |
+
"learning_rate": 0.0001716256157635468,
|
| 1051 |
+
"loss": 0.7552,
|
| 1052 |
+
"step": 2980
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 14.643075045759609,
|
| 1056 |
+
"grad_norm": 4.412954807281494,
|
| 1057 |
+
"learning_rate": 0.00017142857142857143,
|
| 1058 |
+
"loss": 0.7838,
|
| 1059 |
+
"step": 3000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 14.740695546064673,
|
| 1063 |
+
"grad_norm": 2.7720842361450195,
|
| 1064 |
+
"learning_rate": 0.00017123152709359608,
|
| 1065 |
+
"loss": 0.7589,
|
| 1066 |
+
"step": 3020
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 14.838316046369737,
|
| 1070 |
+
"grad_norm": 3.3187596797943115,
|
| 1071 |
+
"learning_rate": 0.0001710344827586207,
|
| 1072 |
+
"loss": 0.7812,
|
| 1073 |
+
"step": 3040
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 14.935936546674801,
|
| 1077 |
+
"grad_norm": 2.3551273345947266,
|
| 1078 |
+
"learning_rate": 0.00017083743842364533,
|
| 1079 |
+
"loss": 0.764,
|
| 1080 |
+
"step": 3060
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 15.033557046979865,
|
| 1084 |
+
"grad_norm": 2.663290023803711,
|
| 1085 |
+
"learning_rate": 0.00017064039408866996,
|
| 1086 |
+
"loss": 0.7034,
|
| 1087 |
+
"step": 3080
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 15.13117754728493,
|
| 1091 |
+
"grad_norm": 3.2227704524993896,
|
| 1092 |
+
"learning_rate": 0.0001704433497536946,
|
| 1093 |
+
"loss": 0.6878,
|
| 1094 |
+
"step": 3100
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 15.228798047589994,
|
| 1098 |
+
"grad_norm": 2.819664478302002,
|
| 1099 |
+
"learning_rate": 0.00017024630541871921,
|
| 1100 |
+
"loss": 0.6731,
|
| 1101 |
+
"step": 3120
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 15.326418547895058,
|
| 1105 |
+
"grad_norm": 2.9787933826446533,
|
| 1106 |
+
"learning_rate": 0.00017004926108374387,
|
| 1107 |
+
"loss": 0.7036,
|
| 1108 |
+
"step": 3140
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 15.424039048200122,
|
| 1112 |
+
"grad_norm": 2.4379117488861084,
|
| 1113 |
+
"learning_rate": 0.00016985221674876847,
|
| 1114 |
+
"loss": 0.7323,
|
| 1115 |
+
"step": 3160
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 15.521659548505186,
|
| 1119 |
+
"grad_norm": 1.9959620237350464,
|
| 1120 |
+
"learning_rate": 0.00016965517241379312,
|
| 1121 |
+
"loss": 0.7155,
|
| 1122 |
+
"step": 3180
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 15.61928004881025,
|
| 1126 |
+
"grad_norm": 2.856109619140625,
|
| 1127 |
+
"learning_rate": 0.00016945812807881772,
|
| 1128 |
+
"loss": 0.6876,
|
| 1129 |
+
"step": 3200
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 15.716900549115314,
|
| 1133 |
+
"grad_norm": 3.9589807987213135,
|
| 1134 |
+
"learning_rate": 0.00016926108374384238,
|
| 1135 |
+
"loss": 0.7316,
|
| 1136 |
+
"step": 3220
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 15.814521049420378,
|
| 1140 |
+
"grad_norm": 2.921196460723877,
|
| 1141 |
+
"learning_rate": 0.000169064039408867,
|
| 1142 |
+
"loss": 0.7306,
|
| 1143 |
+
"step": 3240
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 15.912141549725442,
|
| 1147 |
+
"grad_norm": 2.862910270690918,
|
| 1148 |
+
"learning_rate": 0.00016886699507389163,
|
| 1149 |
+
"loss": 0.7829,
|
| 1150 |
+
"step": 3260
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 16.009762050030506,
|
| 1154 |
+
"grad_norm": 2.988609552383423,
|
| 1155 |
+
"learning_rate": 0.00016866995073891626,
|
| 1156 |
+
"loss": 0.75,
|
| 1157 |
+
"step": 3280
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 16.107382550335572,
|
| 1161 |
+
"grad_norm": 3.728930950164795,
|
| 1162 |
+
"learning_rate": 0.00016847290640394089,
|
| 1163 |
+
"loss": 0.6083,
|
| 1164 |
+
"step": 3300
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 16.205003050640634,
|
| 1168 |
+
"grad_norm": 3.5626068115234375,
|
| 1169 |
+
"learning_rate": 0.00016827586206896554,
|
| 1170 |
+
"loss": 0.6849,
|
| 1171 |
+
"step": 3320
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 16.3026235509457,
|
| 1175 |
+
"grad_norm": 2.754389524459839,
|
| 1176 |
+
"learning_rate": 0.00016807881773399014,
|
| 1177 |
+
"loss": 0.6635,
|
| 1178 |
+
"step": 3340
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 16.400244051250763,
|
| 1182 |
+
"grad_norm": 3.2776389122009277,
|
| 1183 |
+
"learning_rate": 0.0001678817733990148,
|
| 1184 |
+
"loss": 0.6999,
|
| 1185 |
+
"step": 3360
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 16.49786455155583,
|
| 1189 |
+
"grad_norm": 3.0710105895996094,
|
| 1190 |
+
"learning_rate": 0.00016768472906403942,
|
| 1191 |
+
"loss": 0.6911,
|
| 1192 |
+
"step": 3380
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 16.59548505186089,
|
| 1196 |
+
"grad_norm": 3.1727585792541504,
|
| 1197 |
+
"learning_rate": 0.00016748768472906405,
|
| 1198 |
+
"loss": 0.7238,
|
| 1199 |
+
"step": 3400
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 16.693105552165953,
|
| 1203 |
+
"grad_norm": 2.671583652496338,
|
| 1204 |
+
"learning_rate": 0.00016729064039408868,
|
| 1205 |
+
"loss": 0.6925,
|
| 1206 |
+
"step": 3420
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 16.79072605247102,
|
| 1210 |
+
"grad_norm": 2.9183971881866455,
|
| 1211 |
+
"learning_rate": 0.0001670935960591133,
|
| 1212 |
+
"loss": 0.703,
|
| 1213 |
+
"step": 3440
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 16.888346552776085,
|
| 1217 |
+
"grad_norm": 3.785710334777832,
|
| 1218 |
+
"learning_rate": 0.00016689655172413793,
|
| 1219 |
+
"loss": 0.7245,
|
| 1220 |
+
"step": 3460
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 16.985967053081147,
|
| 1224 |
+
"grad_norm": 3.435655355453491,
|
| 1225 |
+
"learning_rate": 0.00016669950738916258,
|
| 1226 |
+
"loss": 0.7483,
|
| 1227 |
+
"step": 3480
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 17.08358755338621,
|
| 1231 |
+
"grad_norm": 3.7350969314575195,
|
| 1232 |
+
"learning_rate": 0.00016650246305418718,
|
| 1233 |
+
"loss": 0.639,
|
| 1234 |
+
"step": 3500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 17.181208053691275,
|
| 1238 |
+
"grad_norm": 3.0420546531677246,
|
| 1239 |
+
"learning_rate": 0.00016630541871921184,
|
| 1240 |
+
"loss": 0.675,
|
| 1241 |
+
"step": 3520
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 17.278828553996338,
|
| 1245 |
+
"grad_norm": 2.1023027896881104,
|
| 1246 |
+
"learning_rate": 0.00016610837438423646,
|
| 1247 |
+
"loss": 0.6857,
|
| 1248 |
+
"step": 3540
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 17.376449054301403,
|
| 1252 |
+
"grad_norm": 2.282754898071289,
|
| 1253 |
+
"learning_rate": 0.0001659113300492611,
|
| 1254 |
+
"loss": 0.7028,
|
| 1255 |
+
"step": 3560
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 17.474069554606466,
|
| 1259 |
+
"grad_norm": 4.962581634521484,
|
| 1260 |
+
"learning_rate": 0.00016571428571428575,
|
| 1261 |
+
"loss": 0.6297,
|
| 1262 |
+
"step": 3580
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 17.57169005491153,
|
| 1266 |
+
"grad_norm": 2.602381944656372,
|
| 1267 |
+
"learning_rate": 0.00016551724137931035,
|
| 1268 |
+
"loss": 0.7003,
|
| 1269 |
+
"step": 3600
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 17.669310555216594,
|
| 1273 |
+
"grad_norm": 4.691868782043457,
|
| 1274 |
+
"learning_rate": 0.000165320197044335,
|
| 1275 |
+
"loss": 0.6993,
|
| 1276 |
+
"step": 3620
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 17.76693105552166,
|
| 1280 |
+
"grad_norm": 3.7989959716796875,
|
| 1281 |
+
"learning_rate": 0.0001651231527093596,
|
| 1282 |
+
"loss": 0.6644,
|
| 1283 |
+
"step": 3640
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 17.864551555826722,
|
| 1287 |
+
"grad_norm": 3.188518524169922,
|
| 1288 |
+
"learning_rate": 0.00016492610837438425,
|
| 1289 |
+
"loss": 0.6713,
|
| 1290 |
+
"step": 3660
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 17.962172056131788,
|
| 1294 |
+
"grad_norm": 3.8618476390838623,
|
| 1295 |
+
"learning_rate": 0.00016472906403940885,
|
| 1296 |
+
"loss": 0.6652,
|
| 1297 |
+
"step": 3680
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 18.05979255643685,
|
| 1301 |
+
"grad_norm": 3.6163158416748047,
|
| 1302 |
+
"learning_rate": 0.0001645320197044335,
|
| 1303 |
+
"loss": 0.667,
|
| 1304 |
+
"step": 3700
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 18.157413056741916,
|
| 1308 |
+
"grad_norm": 3.723688840866089,
|
| 1309 |
+
"learning_rate": 0.00016433497536945814,
|
| 1310 |
+
"loss": 0.6456,
|
| 1311 |
+
"step": 3720
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 18.25503355704698,
|
| 1315 |
+
"grad_norm": 4.452234268188477,
|
| 1316 |
+
"learning_rate": 0.00016413793103448276,
|
| 1317 |
+
"loss": 0.627,
|
| 1318 |
+
"step": 3740
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 18.352654057352044,
|
| 1322 |
+
"grad_norm": 3.0752596855163574,
|
| 1323 |
+
"learning_rate": 0.0001639408866995074,
|
| 1324 |
+
"loss": 0.6755,
|
| 1325 |
+
"step": 3760
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 18.450274557657107,
|
| 1329 |
+
"grad_norm": 3.043836832046509,
|
| 1330 |
+
"learning_rate": 0.00016374384236453202,
|
| 1331 |
+
"loss": 0.6861,
|
| 1332 |
+
"step": 3780
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 18.547895057962172,
|
| 1336 |
+
"grad_norm": 4.210402011871338,
|
| 1337 |
+
"learning_rate": 0.00016354679802955667,
|
| 1338 |
+
"loss": 0.6206,
|
| 1339 |
+
"step": 3800
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 18.645515558267235,
|
| 1343 |
+
"grad_norm": 3.4578044414520264,
|
| 1344 |
+
"learning_rate": 0.0001633497536945813,
|
| 1345 |
+
"loss": 0.633,
|
| 1346 |
+
"step": 3820
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 18.7431360585723,
|
| 1350 |
+
"grad_norm": 3.9487128257751465,
|
| 1351 |
+
"learning_rate": 0.00016315270935960593,
|
| 1352 |
+
"loss": 0.6479,
|
| 1353 |
+
"step": 3840
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 18.840756558877363,
|
| 1357 |
+
"grad_norm": 3.114673376083374,
|
| 1358 |
+
"learning_rate": 0.00016295566502463055,
|
| 1359 |
+
"loss": 0.6468,
|
| 1360 |
+
"step": 3860
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 18.93837705918243,
|
| 1364 |
+
"grad_norm": 3.7751824855804443,
|
| 1365 |
+
"learning_rate": 0.00016275862068965518,
|
| 1366 |
+
"loss": 0.6695,
|
| 1367 |
+
"step": 3880
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 19.03599755948749,
|
| 1371 |
+
"grad_norm": 2.7188830375671387,
|
| 1372 |
+
"learning_rate": 0.0001625615763546798,
|
| 1373 |
+
"loss": 0.6507,
|
| 1374 |
+
"step": 3900
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 19.133618059792557,
|
| 1378 |
+
"grad_norm": 3.5054094791412354,
|
| 1379 |
+
"learning_rate": 0.00016236453201970446,
|
| 1380 |
+
"loss": 0.5542,
|
| 1381 |
+
"step": 3920
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 19.23123856009762,
|
| 1385 |
+
"grad_norm": 2.4097495079040527,
|
| 1386 |
+
"learning_rate": 0.00016216748768472906,
|
| 1387 |
+
"loss": 0.602,
|
| 1388 |
+
"step": 3940
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 19.328859060402685,
|
| 1392 |
+
"grad_norm": 2.925482749938965,
|
| 1393 |
+
"learning_rate": 0.00016197044334975372,
|
| 1394 |
+
"loss": 0.6493,
|
| 1395 |
+
"step": 3960
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 19.426479560707747,
|
| 1399 |
+
"grad_norm": 4.706211566925049,
|
| 1400 |
+
"learning_rate": 0.00016177339901477832,
|
| 1401 |
+
"loss": 0.6285,
|
| 1402 |
+
"step": 3980
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 19.524100061012813,
|
| 1406 |
+
"grad_norm": 3.257904052734375,
|
| 1407 |
+
"learning_rate": 0.00016157635467980297,
|
| 1408 |
+
"loss": 0.6515,
|
| 1409 |
+
"step": 4000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 19.621720561317876,
|
| 1413 |
+
"grad_norm": 3.0172128677368164,
|
| 1414 |
+
"learning_rate": 0.0001613793103448276,
|
| 1415 |
+
"loss": 0.6426,
|
| 1416 |
+
"step": 4020
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 19.71934106162294,
|
| 1420 |
+
"grad_norm": 2.948984146118164,
|
| 1421 |
+
"learning_rate": 0.00016118226600985222,
|
| 1422 |
+
"loss": 0.6487,
|
| 1423 |
+
"step": 4040
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 19.816961561928004,
|
| 1427 |
+
"grad_norm": 3.070138931274414,
|
| 1428 |
+
"learning_rate": 0.00016098522167487685,
|
| 1429 |
+
"loss": 0.6695,
|
| 1430 |
+
"step": 4060
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 19.91458206223307,
|
| 1434 |
+
"grad_norm": 3.364335060119629,
|
| 1435 |
+
"learning_rate": 0.00016078817733990148,
|
| 1436 |
+
"loss": 0.6443,
|
| 1437 |
+
"step": 4080
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 20.012202562538132,
|
| 1441 |
+
"grad_norm": 3.131267547607422,
|
| 1442 |
+
"learning_rate": 0.00016059113300492613,
|
| 1443 |
+
"loss": 0.6403,
|
| 1444 |
+
"step": 4100
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 20.109823062843198,
|
| 1448 |
+
"grad_norm": 2.4083542823791504,
|
| 1449 |
+
"learning_rate": 0.00016039408866995073,
|
| 1450 |
+
"loss": 0.5922,
|
| 1451 |
+
"step": 4120
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 20.20744356314826,
|
| 1455 |
+
"grad_norm": 4.872425556182861,
|
| 1456 |
+
"learning_rate": 0.00016019704433497539,
|
| 1457 |
+
"loss": 0.6166,
|
| 1458 |
+
"step": 4140
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 20.305064063453326,
|
| 1462 |
+
"grad_norm": 6.9143853187561035,
|
| 1463 |
+
"learning_rate": 0.00016,
|
| 1464 |
+
"loss": 0.6023,
|
| 1465 |
+
"step": 4160
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 20.40268456375839,
|
| 1469 |
+
"grad_norm": 2.4565210342407227,
|
| 1470 |
+
"learning_rate": 0.00015980295566502464,
|
| 1471 |
+
"loss": 0.6154,
|
| 1472 |
+
"step": 4180
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 20.500305064063454,
|
| 1476 |
+
"grad_norm": 2.886202096939087,
|
| 1477 |
+
"learning_rate": 0.00015960591133004927,
|
| 1478 |
+
"loss": 0.5861,
|
| 1479 |
+
"step": 4200
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 20.597925564368516,
|
| 1483 |
+
"grad_norm": 3.0811331272125244,
|
| 1484 |
+
"learning_rate": 0.0001594088669950739,
|
| 1485 |
+
"loss": 0.6445,
|
| 1486 |
+
"step": 4220
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 20.695546064673582,
|
| 1490 |
+
"grad_norm": 3.5066580772399902,
|
| 1491 |
+
"learning_rate": 0.00015921182266009852,
|
| 1492 |
+
"loss": 0.6133,
|
| 1493 |
+
"step": 4240
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 20.793166564978645,
|
| 1497 |
+
"grad_norm": 3.8073158264160156,
|
| 1498 |
+
"learning_rate": 0.00015901477832512318,
|
| 1499 |
+
"loss": 0.6133,
|
| 1500 |
+
"step": 4260
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 20.89078706528371,
|
| 1504 |
+
"grad_norm": 4.436833381652832,
|
| 1505 |
+
"learning_rate": 0.00015881773399014778,
|
| 1506 |
+
"loss": 0.6243,
|
| 1507 |
+
"step": 4280
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 20.988407565588773,
|
| 1511 |
+
"grad_norm": 2.7935214042663574,
|
| 1512 |
+
"learning_rate": 0.00015862068965517243,
|
| 1513 |
+
"loss": 0.6349,
|
| 1514 |
+
"step": 4300
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 21.08602806589384,
|
| 1518 |
+
"grad_norm": 3.224860668182373,
|
| 1519 |
+
"learning_rate": 0.00015842364532019706,
|
| 1520 |
+
"loss": 0.5906,
|
| 1521 |
+
"step": 4320
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 21.1836485661989,
|
| 1525 |
+
"grad_norm": 2.9267752170562744,
|
| 1526 |
+
"learning_rate": 0.00015822660098522168,
|
| 1527 |
+
"loss": 0.5512,
|
| 1528 |
+
"step": 4340
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 21.281269066503967,
|
| 1532 |
+
"grad_norm": 3.137066125869751,
|
| 1533 |
+
"learning_rate": 0.0001580295566502463,
|
| 1534 |
+
"loss": 0.5764,
|
| 1535 |
+
"step": 4360
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 21.37888956680903,
|
| 1539 |
+
"grad_norm": 3.112293004989624,
|
| 1540 |
+
"learning_rate": 0.00015783251231527094,
|
| 1541 |
+
"loss": 0.6045,
|
| 1542 |
+
"step": 4380
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 21.476510067114095,
|
| 1546 |
+
"grad_norm": 2.6162259578704834,
|
| 1547 |
+
"learning_rate": 0.0001576354679802956,
|
| 1548 |
+
"loss": 0.6009,
|
| 1549 |
+
"step": 4400
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 21.574130567419157,
|
| 1553 |
+
"grad_norm": 2.924473285675049,
|
| 1554 |
+
"learning_rate": 0.0001574384236453202,
|
| 1555 |
+
"loss": 0.589,
|
| 1556 |
+
"step": 4420
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 21.671751067724223,
|
| 1560 |
+
"grad_norm": 3.2589287757873535,
|
| 1561 |
+
"learning_rate": 0.00015724137931034485,
|
| 1562 |
+
"loss": 0.6078,
|
| 1563 |
+
"step": 4440
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 21.769371568029285,
|
| 1567 |
+
"grad_norm": 3.4130911827087402,
|
| 1568 |
+
"learning_rate": 0.00015704433497536945,
|
| 1569 |
+
"loss": 0.6177,
|
| 1570 |
+
"step": 4460
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 21.86699206833435,
|
| 1574 |
+
"grad_norm": 3.0816001892089844,
|
| 1575 |
+
"learning_rate": 0.0001568472906403941,
|
| 1576 |
+
"loss": 0.6077,
|
| 1577 |
+
"step": 4480
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 21.964612568639414,
|
| 1581 |
+
"grad_norm": 2.875441789627075,
|
| 1582 |
+
"learning_rate": 0.00015665024630541873,
|
| 1583 |
+
"loss": 0.6127,
|
| 1584 |
+
"step": 4500
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 22.06223306894448,
|
| 1588 |
+
"grad_norm": 4.020274639129639,
|
| 1589 |
+
"learning_rate": 0.00015645320197044335,
|
| 1590 |
+
"loss": 0.5673,
|
| 1591 |
+
"step": 4520
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 22.15985356924954,
|
| 1595 |
+
"grad_norm": 3.365691661834717,
|
| 1596 |
+
"learning_rate": 0.00015625615763546798,
|
| 1597 |
+
"loss": 0.5201,
|
| 1598 |
+
"step": 4540
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 22.257474069554608,
|
| 1602 |
+
"grad_norm": 3.449277400970459,
|
| 1603 |
+
"learning_rate": 0.0001560591133004926,
|
| 1604 |
+
"loss": 0.5657,
|
| 1605 |
+
"step": 4560
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 22.35509456985967,
|
| 1609 |
+
"grad_norm": 3.7012288570404053,
|
| 1610 |
+
"learning_rate": 0.00015586206896551724,
|
| 1611 |
+
"loss": 0.6035,
|
| 1612 |
+
"step": 4580
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 22.452715070164736,
|
| 1616 |
+
"grad_norm": 3.5211081504821777,
|
| 1617 |
+
"learning_rate": 0.0001556650246305419,
|
| 1618 |
+
"loss": 0.6173,
|
| 1619 |
+
"step": 4600
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 22.550335570469798,
|
| 1623 |
+
"grad_norm": 3.026588201522827,
|
| 1624 |
+
"learning_rate": 0.00015546798029556652,
|
| 1625 |
+
"loss": 0.6004,
|
| 1626 |
+
"step": 4620
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 22.647956070774864,
|
| 1630 |
+
"grad_norm": 2.7548885345458984,
|
| 1631 |
+
"learning_rate": 0.00015527093596059114,
|
| 1632 |
+
"loss": 0.5633,
|
| 1633 |
+
"step": 4640
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 22.745576571079926,
|
| 1637 |
+
"grad_norm": 5.050055027008057,
|
| 1638 |
+
"learning_rate": 0.00015507389162561577,
|
| 1639 |
+
"loss": 0.6061,
|
| 1640 |
+
"step": 4660
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 22.843197071384992,
|
| 1644 |
+
"grad_norm": 3.0278573036193848,
|
| 1645 |
+
"learning_rate": 0.0001548768472906404,
|
| 1646 |
+
"loss": 0.5607,
|
| 1647 |
+
"step": 4680
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 22.940817571690054,
|
| 1651 |
+
"grad_norm": 3.17149019241333,
|
| 1652 |
+
"learning_rate": 0.00015467980295566505,
|
| 1653 |
+
"loss": 0.5829,
|
| 1654 |
+
"step": 4700
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 23.03843807199512,
|
| 1658 |
+
"grad_norm": 2.5521585941314697,
|
| 1659 |
+
"learning_rate": 0.00015448275862068965,
|
| 1660 |
+
"loss": 0.5723,
|
| 1661 |
+
"step": 4720
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 23.136058572300183,
|
| 1665 |
+
"grad_norm": 2.7798378467559814,
|
| 1666 |
+
"learning_rate": 0.0001542857142857143,
|
| 1667 |
+
"loss": 0.5373,
|
| 1668 |
+
"step": 4740
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 23.23367907260525,
|
| 1672 |
+
"grad_norm": 3.4025466442108154,
|
| 1673 |
+
"learning_rate": 0.0001540886699507389,
|
| 1674 |
+
"loss": 0.5445,
|
| 1675 |
+
"step": 4760
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 23.33129957291031,
|
| 1679 |
+
"grad_norm": 3.9419145584106445,
|
| 1680 |
+
"learning_rate": 0.00015389162561576356,
|
| 1681 |
+
"loss": 0.5677,
|
| 1682 |
+
"step": 4780
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 23.428920073215377,
|
| 1686 |
+
"grad_norm": 2.300863265991211,
|
| 1687 |
+
"learning_rate": 0.00015369458128078816,
|
| 1688 |
+
"loss": 0.5941,
|
| 1689 |
+
"step": 4800
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 23.52654057352044,
|
| 1693 |
+
"grad_norm": 3.25654673576355,
|
| 1694 |
+
"learning_rate": 0.00015349753694581282,
|
| 1695 |
+
"loss": 0.5688,
|
| 1696 |
+
"step": 4820
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 23.624161073825505,
|
| 1700 |
+
"grad_norm": 3.1517579555511475,
|
| 1701 |
+
"learning_rate": 0.00015330049261083744,
|
| 1702 |
+
"loss": 0.5481,
|
| 1703 |
+
"step": 4840
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 23.721781574130567,
|
| 1707 |
+
"grad_norm": 2.5366251468658447,
|
| 1708 |
+
"learning_rate": 0.00015310344827586207,
|
| 1709 |
+
"loss": 0.5725,
|
| 1710 |
+
"step": 4860
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 23.819402074435633,
|
| 1714 |
+
"grad_norm": 4.309774875640869,
|
| 1715 |
+
"learning_rate": 0.00015290640394088672,
|
| 1716 |
+
"loss": 0.574,
|
| 1717 |
+
"step": 4880
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 23.917022574740695,
|
| 1721 |
+
"grad_norm": 3.031926155090332,
|
| 1722 |
+
"learning_rate": 0.00015270935960591132,
|
| 1723 |
+
"loss": 0.5431,
|
| 1724 |
+
"step": 4900
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 24.01464307504576,
|
| 1728 |
+
"grad_norm": 2.574500560760498,
|
| 1729 |
+
"learning_rate": 0.00015251231527093598,
|
| 1730 |
+
"loss": 0.5967,
|
| 1731 |
+
"step": 4920
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 24.112263575350823,
|
| 1735 |
+
"grad_norm": 2.556105136871338,
|
| 1736 |
+
"learning_rate": 0.0001523152709359606,
|
| 1737 |
+
"loss": 0.5419,
|
| 1738 |
+
"step": 4940
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 24.20988407565589,
|
| 1742 |
+
"grad_norm": 2.412322998046875,
|
| 1743 |
+
"learning_rate": 0.00015211822660098523,
|
| 1744 |
+
"loss": 0.5342,
|
| 1745 |
+
"step": 4960
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 24.30750457596095,
|
| 1749 |
+
"grad_norm": 2.39802622795105,
|
| 1750 |
+
"learning_rate": 0.00015192118226600986,
|
| 1751 |
+
"loss": 0.5249,
|
| 1752 |
+
"step": 4980
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 24.405125076266017,
|
| 1756 |
+
"grad_norm": 2.854398727416992,
|
| 1757 |
+
"learning_rate": 0.00015172413793103449,
|
| 1758 |
+
"loss": 0.5468,
|
| 1759 |
+
"step": 5000
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 24.50274557657108,
|
| 1763 |
+
"grad_norm": 2.8961057662963867,
|
| 1764 |
+
"learning_rate": 0.0001515270935960591,
|
| 1765 |
+
"loss": 0.5313,
|
| 1766 |
+
"step": 5020
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 24.600366076876146,
|
| 1770 |
+
"grad_norm": 3.2031073570251465,
|
| 1771 |
+
"learning_rate": 0.00015133004926108377,
|
| 1772 |
+
"loss": 0.5718,
|
| 1773 |
+
"step": 5040
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 24.697986577181208,
|
| 1777 |
+
"grad_norm": 4.338870525360107,
|
| 1778 |
+
"learning_rate": 0.00015113300492610837,
|
| 1779 |
+
"loss": 0.5415,
|
| 1780 |
+
"step": 5060
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 24.795607077486274,
|
| 1784 |
+
"grad_norm": 3.46842360496521,
|
| 1785 |
+
"learning_rate": 0.00015093596059113302,
|
| 1786 |
+
"loss": 0.5546,
|
| 1787 |
+
"step": 5080
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 24.893227577791336,
|
| 1791 |
+
"grad_norm": 2.853489637374878,
|
| 1792 |
+
"learning_rate": 0.00015073891625615765,
|
| 1793 |
+
"loss": 0.5691,
|
| 1794 |
+
"step": 5100
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 24.990848078096402,
|
| 1798 |
+
"grad_norm": 3.427720785140991,
|
| 1799 |
+
"learning_rate": 0.00015054187192118228,
|
| 1800 |
+
"loss": 0.5795,
|
| 1801 |
+
"step": 5120
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 25.088468578401464,
|
| 1805 |
+
"grad_norm": 3.2862656116485596,
|
| 1806 |
+
"learning_rate": 0.0001503448275862069,
|
| 1807 |
+
"loss": 0.5109,
|
| 1808 |
+
"step": 5140
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 25.18608907870653,
|
| 1812 |
+
"grad_norm": 3.383563756942749,
|
| 1813 |
+
"learning_rate": 0.00015014778325123153,
|
| 1814 |
+
"loss": 0.4983,
|
| 1815 |
+
"step": 5160
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 25.283709579011592,
|
| 1819 |
+
"grad_norm": 3.3909354209899902,
|
| 1820 |
+
"learning_rate": 0.00014995073891625618,
|
| 1821 |
+
"loss": 0.5164,
|
| 1822 |
+
"step": 5180
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 25.381330079316655,
|
| 1826 |
+
"grad_norm": 2.616955041885376,
|
| 1827 |
+
"learning_rate": 0.00014975369458128078,
|
| 1828 |
+
"loss": 0.5347,
|
| 1829 |
+
"step": 5200
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 25.47895057962172,
|
| 1833 |
+
"grad_norm": 2.7965965270996094,
|
| 1834 |
+
"learning_rate": 0.00014955665024630544,
|
| 1835 |
+
"loss": 0.5386,
|
| 1836 |
+
"step": 5220
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 25.576571079926783,
|
| 1840 |
+
"grad_norm": 2.9817397594451904,
|
| 1841 |
+
"learning_rate": 0.00014935960591133004,
|
| 1842 |
+
"loss": 0.5001,
|
| 1843 |
+
"step": 5240
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 25.67419158023185,
|
| 1847 |
+
"grad_norm": 2.527992010116577,
|
| 1848 |
+
"learning_rate": 0.0001491625615763547,
|
| 1849 |
+
"loss": 0.5572,
|
| 1850 |
+
"step": 5260
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 25.77181208053691,
|
| 1854 |
+
"grad_norm": 4.047604560852051,
|
| 1855 |
+
"learning_rate": 0.00014896551724137932,
|
| 1856 |
+
"loss": 0.5429,
|
| 1857 |
+
"step": 5280
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 25.869432580841977,
|
| 1861 |
+
"grad_norm": 3.2753515243530273,
|
| 1862 |
+
"learning_rate": 0.00014876847290640395,
|
| 1863 |
+
"loss": 0.5461,
|
| 1864 |
+
"step": 5300
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 25.96705308114704,
|
| 1868 |
+
"grad_norm": 3.5623252391815186,
|
| 1869 |
+
"learning_rate": 0.00014857142857142857,
|
| 1870 |
+
"loss": 0.571,
|
| 1871 |
+
"step": 5320
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 26.064673581452105,
|
| 1875 |
+
"grad_norm": 4.602993965148926,
|
| 1876 |
+
"learning_rate": 0.0001483743842364532,
|
| 1877 |
+
"loss": 0.4858,
|
| 1878 |
+
"step": 5340
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 26.162294081757167,
|
| 1882 |
+
"grad_norm": 3.4932191371917725,
|
| 1883 |
+
"learning_rate": 0.00014817733990147783,
|
| 1884 |
+
"loss": 0.5374,
|
| 1885 |
+
"step": 5360
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 26.259914582062233,
|
| 1889 |
+
"grad_norm": 2.595555305480957,
|
| 1890 |
+
"learning_rate": 0.00014798029556650248,
|
| 1891 |
+
"loss": 0.5217,
|
| 1892 |
+
"step": 5380
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 26.357535082367296,
|
| 1896 |
+
"grad_norm": 2.3642492294311523,
|
| 1897 |
+
"learning_rate": 0.0001477832512315271,
|
| 1898 |
+
"loss": 0.5055,
|
| 1899 |
+
"step": 5400
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 26.45515558267236,
|
| 1903 |
+
"grad_norm": 3.9272634983062744,
|
| 1904 |
+
"learning_rate": 0.00014758620689655174,
|
| 1905 |
+
"loss": 0.5535,
|
| 1906 |
+
"step": 5420
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 26.552776082977424,
|
| 1910 |
+
"grad_norm": 4.050607204437256,
|
| 1911 |
+
"learning_rate": 0.00014738916256157636,
|
| 1912 |
+
"loss": 0.5019,
|
| 1913 |
+
"step": 5440
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 26.65039658328249,
|
| 1917 |
+
"grad_norm": 3.2770299911499023,
|
| 1918 |
+
"learning_rate": 0.000147192118226601,
|
| 1919 |
+
"loss": 0.4922,
|
| 1920 |
+
"step": 5460
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 26.748017083587552,
|
| 1924 |
+
"grad_norm": 3.96409273147583,
|
| 1925 |
+
"learning_rate": 0.00014699507389162562,
|
| 1926 |
+
"loss": 0.5165,
|
| 1927 |
+
"step": 5480
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 26.845637583892618,
|
| 1931 |
+
"grad_norm": 4.587811470031738,
|
| 1932 |
+
"learning_rate": 0.00014679802955665024,
|
| 1933 |
+
"loss": 0.5513,
|
| 1934 |
+
"step": 5500
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 26.94325808419768,
|
| 1938 |
+
"grad_norm": 4.558196067810059,
|
| 1939 |
+
"learning_rate": 0.0001466009852216749,
|
| 1940 |
+
"loss": 0.5227,
|
| 1941 |
+
"step": 5520
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 27.040878584502746,
|
| 1945 |
+
"grad_norm": 3.807441473007202,
|
| 1946 |
+
"learning_rate": 0.0001464039408866995,
|
| 1947 |
+
"loss": 0.5141,
|
| 1948 |
+
"step": 5540
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 27.13849908480781,
|
| 1952 |
+
"grad_norm": 2.2902328968048096,
|
| 1953 |
+
"learning_rate": 0.00014620689655172415,
|
| 1954 |
+
"loss": 0.4822,
|
| 1955 |
+
"step": 5560
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 27.236119585112874,
|
| 1959 |
+
"grad_norm": 4.3950886726379395,
|
| 1960 |
+
"learning_rate": 0.00014600985221674875,
|
| 1961 |
+
"loss": 0.5136,
|
| 1962 |
+
"step": 5580
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 27.333740085417936,
|
| 1966 |
+
"grad_norm": 4.0127482414245605,
|
| 1967 |
+
"learning_rate": 0.0001458128078817734,
|
| 1968 |
+
"loss": 0.5299,
|
| 1969 |
+
"step": 5600
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 27.431360585723002,
|
| 1973 |
+
"grad_norm": 4.659334182739258,
|
| 1974 |
+
"learning_rate": 0.00014561576354679803,
|
| 1975 |
+
"loss": 0.4764,
|
| 1976 |
+
"step": 5620
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 27.528981086028065,
|
| 1980 |
+
"grad_norm": 4.769715785980225,
|
| 1981 |
+
"learning_rate": 0.00014541871921182266,
|
| 1982 |
+
"loss": 0.5236,
|
| 1983 |
+
"step": 5640
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 27.62660158633313,
|
| 1987 |
+
"grad_norm": 3.8856427669525146,
|
| 1988 |
+
"learning_rate": 0.00014522167487684732,
|
| 1989 |
+
"loss": 0.5028,
|
| 1990 |
+
"step": 5660
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 27.724222086638193,
|
| 1994 |
+
"grad_norm": 3.183850049972534,
|
| 1995 |
+
"learning_rate": 0.00014502463054187192,
|
| 1996 |
+
"loss": 0.4945,
|
| 1997 |
+
"step": 5680
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 27.82184258694326,
|
| 2001 |
+
"grad_norm": 3.1610593795776367,
|
| 2002 |
+
"learning_rate": 0.00014482758620689657,
|
| 2003 |
+
"loss": 0.4963,
|
| 2004 |
+
"step": 5700
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 27.91946308724832,
|
| 2008 |
+
"grad_norm": 4.054819107055664,
|
| 2009 |
+
"learning_rate": 0.0001446305418719212,
|
| 2010 |
+
"loss": 0.547,
|
| 2011 |
+
"step": 5720
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 28.017083587553387,
|
| 2015 |
+
"grad_norm": 2.7358503341674805,
|
| 2016 |
+
"learning_rate": 0.00014443349753694582,
|
| 2017 |
+
"loss": 0.5387,
|
| 2018 |
+
"step": 5740
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 28.11470408785845,
|
| 2022 |
+
"grad_norm": 2.403042793273926,
|
| 2023 |
+
"learning_rate": 0.00014423645320197045,
|
| 2024 |
+
"loss": 0.4593,
|
| 2025 |
+
"step": 5760
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 28.212324588163515,
|
| 2029 |
+
"grad_norm": 3.3207452297210693,
|
| 2030 |
+
"learning_rate": 0.00014403940886699508,
|
| 2031 |
+
"loss": 0.4842,
|
| 2032 |
+
"step": 5780
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 28.309945088468577,
|
| 2036 |
+
"grad_norm": 3.0579757690429688,
|
| 2037 |
+
"learning_rate": 0.0001438423645320197,
|
| 2038 |
+
"loss": 0.4754,
|
| 2039 |
+
"step": 5800
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 28.407565588773643,
|
| 2043 |
+
"grad_norm": 4.5140700340271,
|
| 2044 |
+
"learning_rate": 0.00014364532019704436,
|
| 2045 |
+
"loss": 0.5128,
|
| 2046 |
+
"step": 5820
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 28.505186089078705,
|
| 2050 |
+
"grad_norm": 3.541874885559082,
|
| 2051 |
+
"learning_rate": 0.00014344827586206896,
|
| 2052 |
+
"loss": 0.5187,
|
| 2053 |
+
"step": 5840
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 28.60280658938377,
|
| 2057 |
+
"grad_norm": 3.214235782623291,
|
| 2058 |
+
"learning_rate": 0.00014325123152709361,
|
| 2059 |
+
"loss": 0.475,
|
| 2060 |
+
"step": 5860
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 28.700427089688834,
|
| 2064 |
+
"grad_norm": 4.037768363952637,
|
| 2065 |
+
"learning_rate": 0.00014305418719211824,
|
| 2066 |
+
"loss": 0.4733,
|
| 2067 |
+
"step": 5880
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 28.7980475899939,
|
| 2071 |
+
"grad_norm": 3.0469048023223877,
|
| 2072 |
+
"learning_rate": 0.00014285714285714287,
|
| 2073 |
+
"loss": 0.5181,
|
| 2074 |
+
"step": 5900
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 28.89566809029896,
|
| 2078 |
+
"grad_norm": 3.3396294116973877,
|
| 2079 |
+
"learning_rate": 0.0001426600985221675,
|
| 2080 |
+
"loss": 0.5062,
|
| 2081 |
+
"step": 5920
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 28.993288590604028,
|
| 2085 |
+
"grad_norm": 3.4280455112457275,
|
| 2086 |
+
"learning_rate": 0.00014246305418719212,
|
| 2087 |
+
"loss": 0.5232,
|
| 2088 |
+
"step": 5940
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 29.09090909090909,
|
| 2092 |
+
"grad_norm": 3.8690781593322754,
|
| 2093 |
+
"learning_rate": 0.00014226600985221678,
|
| 2094 |
+
"loss": 0.4744,
|
| 2095 |
+
"step": 5960
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 29.188529591214156,
|
| 2099 |
+
"grad_norm": 3.1680831909179688,
|
| 2100 |
+
"learning_rate": 0.00014206896551724138,
|
| 2101 |
+
"loss": 0.4679,
|
| 2102 |
+
"step": 5980
|
| 2103 |
+
},
|
| 2104 |
+
{
|
| 2105 |
+
"epoch": 29.286150091519218,
|
| 2106 |
+
"grad_norm": 3.752593755722046,
|
| 2107 |
+
"learning_rate": 0.00014187192118226603,
|
| 2108 |
+
"loss": 0.444,
|
| 2109 |
+
"step": 6000
|
| 2110 |
+
}
|
| 2111 |
+
],
|
| 2112 |
+
"logging_steps": 20,
|
| 2113 |
+
"max_steps": 20400,
|
| 2114 |
+
"num_input_tokens_seen": 0,
|
| 2115 |
+
"num_train_epochs": 100,
|
| 2116 |
+
"save_steps": 500,
|
| 2117 |
+
"stateful_callbacks": {
|
| 2118 |
+
"TrainerControl": {
|
| 2119 |
+
"args": {
|
| 2120 |
+
"should_epoch_stop": false,
|
| 2121 |
+
"should_evaluate": false,
|
| 2122 |
+
"should_log": false,
|
| 2123 |
+
"should_save": true,
|
| 2124 |
+
"should_training_stop": false
|
| 2125 |
+
},
|
| 2126 |
+
"attributes": {}
|
| 2127 |
+
}
|
| 2128 |
+
},
|
| 2129 |
+
"total_flos": 1.1953936764223488e+17,
|
| 2130 |
+
"train_batch_size": 1,
|
| 2131 |
+
"trial_name": null,
|
| 2132 |
+
"trial_params": null
|
| 2133 |
+
}
|