Upload checkpoint-9351
Browse files
checkpoint-9351/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c20bd18122349a95643d9e290b8658efb6f835f3bcf3f44ed94c6927b533f3b1
|
| 3 |
+
size 16389
|
checkpoint-9351/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0cd3faf17d9f001fabcda4acfa7acbf1dd0e38d7512567c37d9f9a1cb2624d5b
|
| 3 |
+
size 1465
|
checkpoint-9351/trainer_state.json
ADDED
|
@@ -0,0 +1,2731 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": 7273,
|
| 3 |
+
"best_metric": 2.507423162460327,
|
| 4 |
+
"best_model_checkpoint": "outputs/chess-explainer/checkpoint-7273",
|
| 5 |
+
"epoch": 9.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 9351,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.0009624639076034649,
|
| 14 |
+
"grad_norm": 883.1925048828125,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 17.491071701049805,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.02406159769008662,
|
| 21 |
+
"grad_norm": 55.103694915771484,
|
| 22 |
+
"learning_rate": 1.5384615384615387e-05,
|
| 23 |
+
"loss": 11.750035603841146,
|
| 24 |
+
"step": 25
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.04812319538017324,
|
| 28 |
+
"grad_norm": 9.345708847045898,
|
| 29 |
+
"learning_rate": 3.141025641025641e-05,
|
| 30 |
+
"loss": 4.889132080078125,
|
| 31 |
+
"step": 50
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.07218479307025986,
|
| 35 |
+
"grad_norm": 10.196633338928223,
|
| 36 |
+
"learning_rate": 4.7435897435897435e-05,
|
| 37 |
+
"loss": 3.7856689453125,
|
| 38 |
+
"step": 75
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.09624639076034648,
|
| 42 |
+
"grad_norm": 8.611700057983398,
|
| 43 |
+
"learning_rate": 6.346153846153847e-05,
|
| 44 |
+
"loss": 3.5517022705078123,
|
| 45 |
+
"step": 100
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.12030798845043311,
|
| 49 |
+
"grad_norm": 7.812922477722168,
|
| 50 |
+
"learning_rate": 7.948717948717948e-05,
|
| 51 |
+
"loss": 3.4541973876953125,
|
| 52 |
+
"step": 125
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.14436958614051973,
|
| 56 |
+
"grad_norm": 7.820157527923584,
|
| 57 |
+
"learning_rate": 9.551282051282052e-05,
|
| 58 |
+
"loss": 3.3625674438476563,
|
| 59 |
+
"step": 150
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.16843118383060635,
|
| 63 |
+
"grad_norm": 7.358847141265869,
|
| 64 |
+
"learning_rate": 0.00011153846153846154,
|
| 65 |
+
"loss": 3.33571533203125,
|
| 66 |
+
"step": 175
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.19249278152069296,
|
| 70 |
+
"grad_norm": 7.274508953094482,
|
| 71 |
+
"learning_rate": 0.00012756410256410257,
|
| 72 |
+
"loss": 3.3122216796875,
|
| 73 |
+
"step": 200
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.2165543792107796,
|
| 77 |
+
"grad_norm": 7.056718349456787,
|
| 78 |
+
"learning_rate": 0.0001435897435897436,
|
| 79 |
+
"loss": 3.28180419921875,
|
| 80 |
+
"step": 225
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.24061597690086622,
|
| 84 |
+
"grad_norm": 6.19630241394043,
|
| 85 |
+
"learning_rate": 0.00015961538461538462,
|
| 86 |
+
"loss": 3.2475381469726563,
|
| 87 |
+
"step": 250
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.2646775745909528,
|
| 91 |
+
"grad_norm": 6.5128397941589355,
|
| 92 |
+
"learning_rate": 0.00017564102564102566,
|
| 93 |
+
"loss": 3.234718017578125,
|
| 94 |
+
"step": 275
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.28873917228103946,
|
| 98 |
+
"grad_norm": 6.359344482421875,
|
| 99 |
+
"learning_rate": 0.00019166666666666667,
|
| 100 |
+
"loss": 3.2271856689453124,
|
| 101 |
+
"step": 300
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.3128007699711261,
|
| 105 |
+
"grad_norm": 6.695319175720215,
|
| 106 |
+
"learning_rate": 0.0001999993003464737,
|
| 107 |
+
"loss": 3.1996661376953126,
|
| 108 |
+
"step": 325
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.3368623676612127,
|
| 112 |
+
"grad_norm": 5.9124321937561035,
|
| 113 |
+
"learning_rate": 0.00019999334849877745,
|
| 114 |
+
"loss": 3.1934442138671875,
|
| 115 |
+
"step": 350
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.36092396535129934,
|
| 119 |
+
"grad_norm": 5.623684406280518,
|
| 120 |
+
"learning_rate": 0.00019998132369740206,
|
| 121 |
+
"loss": 3.1889007568359373,
|
| 122 |
+
"step": 375
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.3849855630413859,
|
| 126 |
+
"grad_norm": 5.664840221405029,
|
| 127 |
+
"learning_rate": 0.00019996322667265672,
|
| 128 |
+
"loss": 3.173824462890625,
|
| 129 |
+
"step": 400
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.40904716073147257,
|
| 133 |
+
"grad_norm": 6.227374076843262,
|
| 134 |
+
"learning_rate": 0.0001999390585236385,
|
| 135 |
+
"loss": 3.15881591796875,
|
| 136 |
+
"step": 425
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.4331087584215592,
|
| 140 |
+
"grad_norm": 5.214109420776367,
|
| 141 |
+
"learning_rate": 0.0001999088207181655,
|
| 142 |
+
"loss": 3.09910888671875,
|
| 143 |
+
"step": 450
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.4571703561116458,
|
| 147 |
+
"grad_norm": 5.23874568939209,
|
| 148 |
+
"learning_rate": 0.0001998725150926878,
|
| 149 |
+
"loss": 3.1576531982421874,
|
| 150 |
+
"step": 475
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.48123195380173245,
|
| 154 |
+
"grad_norm": 5.108809471130371,
|
| 155 |
+
"learning_rate": 0.0001998301438521759,
|
| 156 |
+
"loss": 3.1337042236328125,
|
| 157 |
+
"step": 500
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.5052935514918191,
|
| 161 |
+
"grad_norm": 5.708388805389404,
|
| 162 |
+
"learning_rate": 0.00019978170956998676,
|
| 163 |
+
"loss": 3.128594970703125,
|
| 164 |
+
"step": 525
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.5293551491819056,
|
| 168 |
+
"grad_norm": 5.174871921539307,
|
| 169 |
+
"learning_rate": 0.00019972721518770757,
|
| 170 |
+
"loss": 3.1022515869140626,
|
| 171 |
+
"step": 550
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.5534167468719923,
|
| 175 |
+
"grad_norm": 5.113072872161865,
|
| 176 |
+
"learning_rate": 0.00019966666401497705,
|
| 177 |
+
"loss": 3.063781433105469,
|
| 178 |
+
"step": 575
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.5774783445620789,
|
| 182 |
+
"grad_norm": 5.224897384643555,
|
| 183 |
+
"learning_rate": 0.00019960005972928445,
|
| 184 |
+
"loss": 3.0889013671875,
|
| 185 |
+
"step": 600
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.6015399422521656,
|
| 189 |
+
"grad_norm": 5.108465671539307,
|
| 190 |
+
"learning_rate": 0.00019952740637574633,
|
| 191 |
+
"loss": 3.0858636474609376,
|
| 192 |
+
"step": 625
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.6256015399422522,
|
| 196 |
+
"grad_norm": 5.350228786468506,
|
| 197 |
+
"learning_rate": 0.00019944870836686063,
|
| 198 |
+
"loss": 3.0400531005859377,
|
| 199 |
+
"step": 650
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.6496631376323387,
|
| 203 |
+
"grad_norm": 4.423816204071045,
|
| 204 |
+
"learning_rate": 0.0001993639704822389,
|
| 205 |
+
"loss": 3.06386474609375,
|
| 206 |
+
"step": 675
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.6737247353224254,
|
| 210 |
+
"grad_norm": 5.499719619750977,
|
| 211 |
+
"learning_rate": 0.00019927319786831594,
|
| 212 |
+
"loss": 3.0412890625,
|
| 213 |
+
"step": 700
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.697786333012512,
|
| 217 |
+
"grad_norm": 4.638278007507324,
|
| 218 |
+
"learning_rate": 0.0001991763960380373,
|
| 219 |
+
"loss": 3.0481976318359374,
|
| 220 |
+
"step": 725
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.7218479307025987,
|
| 224 |
+
"grad_norm": 5.008415222167969,
|
| 225 |
+
"learning_rate": 0.00019907357087052424,
|
| 226 |
+
"loss": 3.0042901611328126,
|
| 227 |
+
"step": 750
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.7459095283926853,
|
| 231 |
+
"grad_norm": 4.793009281158447,
|
| 232 |
+
"learning_rate": 0.00019896472861071702,
|
| 233 |
+
"loss": 2.9917413330078126,
|
| 234 |
+
"step": 775
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.7699711260827719,
|
| 238 |
+
"grad_norm": 5.019183158874512,
|
| 239 |
+
"learning_rate": 0.0001988498758689953,
|
| 240 |
+
"loss": 3.0368035888671874,
|
| 241 |
+
"step": 800
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.7940327237728585,
|
| 245 |
+
"grad_norm": 4.315158843994141,
|
| 246 |
+
"learning_rate": 0.00019872901962077687,
|
| 247 |
+
"loss": 2.9863958740234375,
|
| 248 |
+
"step": 825
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.8180943214629451,
|
| 252 |
+
"grad_norm": 5.23874044418335,
|
| 253 |
+
"learning_rate": 0.00019860216720609395,
|
| 254 |
+
"loss": 2.9992340087890623,
|
| 255 |
+
"step": 850
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.8421559191530318,
|
| 259 |
+
"grad_norm": 4.575445652008057,
|
| 260 |
+
"learning_rate": 0.00019846932632914733,
|
| 261 |
+
"loss": 3.0112255859375,
|
| 262 |
+
"step": 875
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.8662175168431184,
|
| 266 |
+
"grad_norm": 4.955294609069824,
|
| 267 |
+
"learning_rate": 0.0001983305050578386,
|
| 268 |
+
"loss": 3.0229232788085936,
|
| 269 |
+
"step": 900
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.890279114533205,
|
| 273 |
+
"grad_norm": 4.7536773681640625,
|
| 274 |
+
"learning_rate": 0.00019818571182328003,
|
| 275 |
+
"loss": 3.0158026123046877,
|
| 276 |
+
"step": 925
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.9143407122232916,
|
| 280 |
+
"grad_norm": 4.718550205230713,
|
| 281 |
+
"learning_rate": 0.00019803495541928262,
|
| 282 |
+
"loss": 2.983572998046875,
|
| 283 |
+
"step": 950
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.9384023099133783,
|
| 287 |
+
"grad_norm": 4.429309844970703,
|
| 288 |
+
"learning_rate": 0.000197878245001822,
|
| 289 |
+
"loss": 2.9714495849609377,
|
| 290 |
+
"step": 975
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.9624639076034649,
|
| 294 |
+
"grad_norm": 4.627259254455566,
|
| 295 |
+
"learning_rate": 0.00019771559008848223,
|
| 296 |
+
"loss": 2.961048889160156,
|
| 297 |
+
"step": 1000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.9865255052935515,
|
| 301 |
+
"grad_norm": 5.209016799926758,
|
| 302 |
+
"learning_rate": 0.00019754700055787797,
|
| 303 |
+
"loss": 2.9716583251953126,
|
| 304 |
+
"step": 1025
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 1.0,
|
| 308 |
+
"eval_loss": 2.934025287628174,
|
| 309 |
+
"eval_runtime": 12.8238,
|
| 310 |
+
"eval_samples_per_second": 116.97,
|
| 311 |
+
"eval_steps_per_second": 1.872,
|
| 312 |
+
"step": 1039
|
| 313 |
+
},
|
| 314 |
+
{
|
| 315 |
+
"epoch": 1.0105871029836382,
|
| 316 |
+
"grad_norm": 4.551164150238037,
|
| 317 |
+
"learning_rate": 0.00019737248664905435,
|
| 318 |
+
"loss": 2.9250759887695312,
|
| 319 |
+
"step": 1050
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 1.0346487006737248,
|
| 323 |
+
"grad_norm": 5.032167911529541,
|
| 324 |
+
"learning_rate": 0.00019719205896086522,
|
| 325 |
+
"loss": 2.8127606201171873,
|
| 326 |
+
"step": 1075
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 1.0587102983638113,
|
| 330 |
+
"grad_norm": 5.156538009643555,
|
| 331 |
+
"learning_rate": 0.00019700572845132936,
|
| 332 |
+
"loss": 2.836297607421875,
|
| 333 |
+
"step": 1100
|
| 334 |
+
},
|
| 335 |
+
{
|
| 336 |
+
"epoch": 1.082771896053898,
|
| 337 |
+
"grad_norm": 5.519596099853516,
|
| 338 |
+
"learning_rate": 0.00019681350643696506,
|
| 339 |
+
"loss": 2.84802490234375,
|
| 340 |
+
"step": 1125
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 1.1068334937439845,
|
| 344 |
+
"grad_norm": 4.649038314819336,
|
| 345 |
+
"learning_rate": 0.00019661540459210271,
|
| 346 |
+
"loss": 2.8203363037109375,
|
| 347 |
+
"step": 1150
|
| 348 |
+
},
|
| 349 |
+
{
|
| 350 |
+
"epoch": 1.1308950914340712,
|
| 351 |
+
"grad_norm": 4.99331521987915,
|
| 352 |
+
"learning_rate": 0.00019641143494817582,
|
| 353 |
+
"loss": 2.862440185546875,
|
| 354 |
+
"step": 1175
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 1.1549566891241578,
|
| 358 |
+
"grad_norm": 4.924983501434326,
|
| 359 |
+
"learning_rate": 0.0001962016098929904,
|
| 360 |
+
"loss": 2.8315557861328124,
|
| 361 |
+
"step": 1200
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 1.1790182868142445,
|
| 365 |
+
"grad_norm": 5.165991306304932,
|
| 366 |
+
"learning_rate": 0.0001959859421699725,
|
| 367 |
+
"loss": 2.8693508911132812,
|
| 368 |
+
"step": 1225
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 1.2030798845043311,
|
| 372 |
+
"grad_norm": 4.788819313049316,
|
| 373 |
+
"learning_rate": 0.00019576444487739425,
|
| 374 |
+
"loss": 2.836847229003906,
|
| 375 |
+
"step": 1250
|
| 376 |
+
},
|
| 377 |
+
{
|
| 378 |
+
"epoch": 1.2271414821944178,
|
| 379 |
+
"grad_norm": 5.095317840576172,
|
| 380 |
+
"learning_rate": 0.0001955371314675784,
|
| 381 |
+
"loss": 2.8328265380859374,
|
| 382 |
+
"step": 1275
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 1.2512030798845044,
|
| 386 |
+
"grad_norm": 5.162815570831299,
|
| 387 |
+
"learning_rate": 0.00019530401574608134,
|
| 388 |
+
"loss": 2.8390570068359375,
|
| 389 |
+
"step": 1300
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 1.275264677574591,
|
| 393 |
+
"grad_norm": 4.838395595550537,
|
| 394 |
+
"learning_rate": 0.00019506511187085456,
|
| 395 |
+
"loss": 2.855579833984375,
|
| 396 |
+
"step": 1325
|
| 397 |
+
},
|
| 398 |
+
{
|
| 399 |
+
"epoch": 1.2993262752646775,
|
| 400 |
+
"grad_norm": 4.820464611053467,
|
| 401 |
+
"learning_rate": 0.00019482043435138478,
|
| 402 |
+
"loss": 2.835367431640625,
|
| 403 |
+
"step": 1350
|
| 404 |
+
},
|
| 405 |
+
{
|
| 406 |
+
"epoch": 1.3233878729547641,
|
| 407 |
+
"grad_norm": 4.540008544921875,
|
| 408 |
+
"learning_rate": 0.00019456999804781297,
|
| 409 |
+
"loss": 2.822310485839844,
|
| 410 |
+
"step": 1375
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 1.3474494706448508,
|
| 414 |
+
"grad_norm": 4.731804847717285,
|
| 415 |
+
"learning_rate": 0.00019431381817003142,
|
| 416 |
+
"loss": 2.8431842041015627,
|
| 417 |
+
"step": 1400
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 1.3715110683349374,
|
| 421 |
+
"grad_norm": 4.5690741539001465,
|
| 422 |
+
"learning_rate": 0.00019405191027676034,
|
| 423 |
+
"loss": 2.8410440063476563,
|
| 424 |
+
"step": 1425
|
| 425 |
+
},
|
| 426 |
+
{
|
| 427 |
+
"epoch": 1.395572666025024,
|
| 428 |
+
"grad_norm": 4.800422668457031,
|
| 429 |
+
"learning_rate": 0.0001937842902746029,
|
| 430 |
+
"loss": 2.8300006103515627,
|
| 431 |
+
"step": 1450
|
| 432 |
+
},
|
| 433 |
+
{
|
| 434 |
+
"epoch": 1.4196342637151107,
|
| 435 |
+
"grad_norm": 5.346854209899902,
|
| 436 |
+
"learning_rate": 0.00019351097441707883,
|
| 437 |
+
"loss": 2.8334808349609375,
|
| 438 |
+
"step": 1475
|
| 439 |
+
},
|
| 440 |
+
{
|
| 441 |
+
"epoch": 1.4436958614051973,
|
| 442 |
+
"grad_norm": 5.628939628601074,
|
| 443 |
+
"learning_rate": 0.00019323197930363776,
|
| 444 |
+
"loss": 2.8730230712890625,
|
| 445 |
+
"step": 1500
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 1.467757459095284,
|
| 449 |
+
"grad_norm": 4.565985202789307,
|
| 450 |
+
"learning_rate": 0.00019294732187865072,
|
| 451 |
+
"loss": 2.856542053222656,
|
| 452 |
+
"step": 1525
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 1.4918190567853706,
|
| 456 |
+
"grad_norm": 5.309317111968994,
|
| 457 |
+
"learning_rate": 0.0001926570194303812,
|
| 458 |
+
"loss": 2.797247314453125,
|
| 459 |
+
"step": 1550
|
| 460 |
+
},
|
| 461 |
+
{
|
| 462 |
+
"epoch": 1.515880654475457,
|
| 463 |
+
"grad_norm": 5.3165693283081055,
|
| 464 |
+
"learning_rate": 0.00019236108958993519,
|
| 465 |
+
"loss": 2.81298828125,
|
| 466 |
+
"step": 1575
|
| 467 |
+
},
|
| 468 |
+
{
|
| 469 |
+
"epoch": 1.539942252165544,
|
| 470 |
+
"grad_norm": 4.712189674377441,
|
| 471 |
+
"learning_rate": 0.0001920595503301902,
|
| 472 |
+
"loss": 2.82627685546875,
|
| 473 |
+
"step": 1600
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 1.5640038498556303,
|
| 477 |
+
"grad_norm": 4.948224067687988,
|
| 478 |
+
"learning_rate": 0.00019175241996470402,
|
| 479 |
+
"loss": 2.820457763671875,
|
| 480 |
+
"step": 1625
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 1.588065447545717,
|
| 484 |
+
"grad_norm": 5.298257827758789,
|
| 485 |
+
"learning_rate": 0.00019143971714660212,
|
| 486 |
+
"loss": 2.8194976806640626,
|
| 487 |
+
"step": 1650
|
| 488 |
+
},
|
| 489 |
+
{
|
| 490 |
+
"epoch": 1.6121270452358036,
|
| 491 |
+
"grad_norm": 5.093235015869141,
|
| 492 |
+
"learning_rate": 0.00019112146086744505,
|
| 493 |
+
"loss": 2.81148193359375,
|
| 494 |
+
"step": 1675
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 1.6361886429258903,
|
| 498 |
+
"grad_norm": 4.273094654083252,
|
| 499 |
+
"learning_rate": 0.00019079767045607492,
|
| 500 |
+
"loss": 2.7786264038085937,
|
| 501 |
+
"step": 1700
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 1.660250240615977,
|
| 505 |
+
"grad_norm": 4.984681606292725,
|
| 506 |
+
"learning_rate": 0.00019046836557744135,
|
| 507 |
+
"loss": 2.815713195800781,
|
| 508 |
+
"step": 1725
|
| 509 |
+
},
|
| 510 |
+
{
|
| 511 |
+
"epoch": 1.6843118383060636,
|
| 512 |
+
"grad_norm": 5.150766849517822,
|
| 513 |
+
"learning_rate": 0.00019013356623140748,
|
| 514 |
+
"loss": 2.837911376953125,
|
| 515 |
+
"step": 1750
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"epoch": 1.7083734359961502,
|
| 519 |
+
"grad_norm": 5.059906959533691,
|
| 520 |
+
"learning_rate": 0.00018979329275153493,
|
| 521 |
+
"loss": 2.7880908203125,
|
| 522 |
+
"step": 1775
|
| 523 |
+
},
|
| 524 |
+
{
|
| 525 |
+
"epoch": 1.7324350336862366,
|
| 526 |
+
"grad_norm": 4.918034076690674,
|
| 527 |
+
"learning_rate": 0.00018944756580384914,
|
| 528 |
+
"loss": 2.800741882324219,
|
| 529 |
+
"step": 1800
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 1.7564966313763235,
|
| 533 |
+
"grad_norm": 5.184783458709717,
|
| 534 |
+
"learning_rate": 0.00018909640638558415,
|
| 535 |
+
"loss": 2.806388244628906,
|
| 536 |
+
"step": 1825
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 1.78055822906641,
|
| 540 |
+
"grad_norm": 4.884471416473389,
|
| 541 |
+
"learning_rate": 0.0001887398358239073,
|
| 542 |
+
"loss": 2.8272454833984373,
|
| 543 |
+
"step": 1850
|
| 544 |
+
},
|
| 545 |
+
{
|
| 546 |
+
"epoch": 1.8046198267564968,
|
| 547 |
+
"grad_norm": 5.7572922706604,
|
| 548 |
+
"learning_rate": 0.0001883778757746241,
|
| 549 |
+
"loss": 2.7949139404296877,
|
| 550 |
+
"step": 1875
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 1.8286814244465832,
|
| 554 |
+
"grad_norm": 5.056210517883301,
|
| 555 |
+
"learning_rate": 0.0001880105482208629,
|
| 556 |
+
"loss": 2.783699951171875,
|
| 557 |
+
"step": 1900
|
| 558 |
+
},
|
| 559 |
+
{
|
| 560 |
+
"epoch": 1.8527430221366699,
|
| 561 |
+
"grad_norm": 4.134166240692139,
|
| 562 |
+
"learning_rate": 0.00018763787547173975,
|
| 563 |
+
"loss": 2.7708892822265625,
|
| 564 |
+
"step": 1925
|
| 565 |
+
},
|
| 566 |
+
{
|
| 567 |
+
"epoch": 1.8768046198267565,
|
| 568 |
+
"grad_norm": 4.962014198303223,
|
| 569 |
+
"learning_rate": 0.00018725988016100357,
|
| 570 |
+
"loss": 2.8055865478515627,
|
| 571 |
+
"step": 1950
|
| 572 |
+
},
|
| 573 |
+
{
|
| 574 |
+
"epoch": 1.9008662175168431,
|
| 575 |
+
"grad_norm": 4.6821088790893555,
|
| 576 |
+
"learning_rate": 0.0001868765852456615,
|
| 577 |
+
"loss": 2.809751281738281,
|
| 578 |
+
"step": 1975
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 1.9249278152069298,
|
| 582 |
+
"grad_norm": 4.961895942687988,
|
| 583 |
+
"learning_rate": 0.0001864880140045846,
|
| 584 |
+
"loss": 2.837526550292969,
|
| 585 |
+
"step": 2000
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 1.9489894128970162,
|
| 589 |
+
"grad_norm": 4.305720806121826,
|
| 590 |
+
"learning_rate": 0.000186094190037094,
|
| 591 |
+
"loss": 2.8118356323242186,
|
| 592 |
+
"step": 2025
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 1.973051010587103,
|
| 596 |
+
"grad_norm": 4.6312174797058105,
|
| 597 |
+
"learning_rate": 0.00018569513726152782,
|
| 598 |
+
"loss": 2.8202804565429687,
|
| 599 |
+
"step": 2050
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 1.9971126082771895,
|
| 603 |
+
"grad_norm": 5.322890281677246,
|
| 604 |
+
"learning_rate": 0.0001852908799137883,
|
| 605 |
+
"loss": 2.7954449462890625,
|
| 606 |
+
"step": 2075
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 2.0,
|
| 610 |
+
"eval_loss": 2.796738386154175,
|
| 611 |
+
"eval_runtime": 10.4549,
|
| 612 |
+
"eval_samples_per_second": 143.473,
|
| 613 |
+
"eval_steps_per_second": 2.296,
|
| 614 |
+
"step": 2078
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 2.0211742059672764,
|
| 618 |
+
"grad_norm": 4.973223686218262,
|
| 619 |
+
"learning_rate": 0.00018488144254586998,
|
| 620 |
+
"loss": 2.634290771484375,
|
| 621 |
+
"step": 2100
|
| 622 |
+
},
|
| 623 |
+
{
|
| 624 |
+
"epoch": 2.045235803657363,
|
| 625 |
+
"grad_norm": 4.723742961883545,
|
| 626 |
+
"learning_rate": 0.0001844668500243686,
|
| 627 |
+
"loss": 2.6185684204101562,
|
| 628 |
+
"step": 2125
|
| 629 |
+
},
|
| 630 |
+
{
|
| 631 |
+
"epoch": 2.0692974013474497,
|
| 632 |
+
"grad_norm": 4.866776466369629,
|
| 633 |
+
"learning_rate": 0.0001840471275289707,
|
| 634 |
+
"loss": 2.63686767578125,
|
| 635 |
+
"step": 2150
|
| 636 |
+
},
|
| 637 |
+
{
|
| 638 |
+
"epoch": 2.093358999037536,
|
| 639 |
+
"grad_norm": 5.0636887550354,
|
| 640 |
+
"learning_rate": 0.00018362230055092465,
|
| 641 |
+
"loss": 2.6171771240234376,
|
| 642 |
+
"step": 2175
|
| 643 |
+
},
|
| 644 |
+
{
|
| 645 |
+
"epoch": 2.1174205967276225,
|
| 646 |
+
"grad_norm": 5.5221943855285645,
|
| 647 |
+
"learning_rate": 0.0001831923948914921,
|
| 648 |
+
"loss": 2.6215435791015627,
|
| 649 |
+
"step": 2200
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 2.1414821944177094,
|
| 653 |
+
"grad_norm": 4.635878562927246,
|
| 654 |
+
"learning_rate": 0.00018275743666038144,
|
| 655 |
+
"loss": 2.6706222534179687,
|
| 656 |
+
"step": 2225
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 2.165543792107796,
|
| 660 |
+
"grad_norm": 5.1727705001831055,
|
| 661 |
+
"learning_rate": 0.00018231745227416155,
|
| 662 |
+
"loss": 2.63311279296875,
|
| 663 |
+
"step": 2250
|
| 664 |
+
},
|
| 665 |
+
{
|
| 666 |
+
"epoch": 2.1896053897978827,
|
| 667 |
+
"grad_norm": 4.907919406890869,
|
| 668 |
+
"learning_rate": 0.00018187246845465787,
|
| 669 |
+
"loss": 2.64796630859375,
|
| 670 |
+
"step": 2275
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 2.213666987487969,
|
| 674 |
+
"grad_norm": 4.771077632904053,
|
| 675 |
+
"learning_rate": 0.0001814225122273292,
|
| 676 |
+
"loss": 2.6579470825195313,
|
| 677 |
+
"step": 2300
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"epoch": 2.237728585178056,
|
| 681 |
+
"grad_norm": 4.675436496734619,
|
| 682 |
+
"learning_rate": 0.0001809676109196265,
|
| 683 |
+
"loss": 2.655972900390625,
|
| 684 |
+
"step": 2325
|
| 685 |
+
},
|
| 686 |
+
{
|
| 687 |
+
"epoch": 2.2617901828681424,
|
| 688 |
+
"grad_norm": 5.554975986480713,
|
| 689 |
+
"learning_rate": 0.00018050779215933314,
|
| 690 |
+
"loss": 2.6388140869140626,
|
| 691 |
+
"step": 2350
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 2.2858517805582292,
|
| 695 |
+
"grad_norm": 5.329654216766357,
|
| 696 |
+
"learning_rate": 0.00018004308387288695,
|
| 697 |
+
"loss": 2.681984558105469,
|
| 698 |
+
"step": 2375
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 2.3099133782483157,
|
| 702 |
+
"grad_norm": 5.244048595428467,
|
| 703 |
+
"learning_rate": 0.0001795735142836842,
|
| 704 |
+
"loss": 2.65951416015625,
|
| 705 |
+
"step": 2400
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 2.3339749759384025,
|
| 709 |
+
"grad_norm": 4.824357509613037,
|
| 710 |
+
"learning_rate": 0.00017909911191036547,
|
| 711 |
+
"loss": 2.6685150146484373,
|
| 712 |
+
"step": 2425
|
| 713 |
+
},
|
| 714 |
+
{
|
| 715 |
+
"epoch": 2.358036573628489,
|
| 716 |
+
"grad_norm": 5.436433792114258,
|
| 717 |
+
"learning_rate": 0.00017861990556508356,
|
| 718 |
+
"loss": 2.639136962890625,
|
| 719 |
+
"step": 2450
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"epoch": 2.3820981713185754,
|
| 723 |
+
"grad_norm": 5.015676975250244,
|
| 724 |
+
"learning_rate": 0.0001781359243517537,
|
| 725 |
+
"loss": 2.652701416015625,
|
| 726 |
+
"step": 2475
|
| 727 |
+
},
|
| 728 |
+
{
|
| 729 |
+
"epoch": 2.4061597690086622,
|
| 730 |
+
"grad_norm": 5.517067909240723,
|
| 731 |
+
"learning_rate": 0.00017764719766428594,
|
| 732 |
+
"loss": 2.6934796142578126,
|
| 733 |
+
"step": 2500
|
| 734 |
+
},
|
| 735 |
+
{
|
| 736 |
+
"epoch": 2.4302213666987487,
|
| 737 |
+
"grad_norm": 4.892432689666748,
|
| 738 |
+
"learning_rate": 0.00017715375518479994,
|
| 739 |
+
"loss": 2.647161865234375,
|
| 740 |
+
"step": 2525
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 2.4542829643888355,
|
| 744 |
+
"grad_norm": 5.116722106933594,
|
| 745 |
+
"learning_rate": 0.0001766556268818222,
|
| 746 |
+
"loss": 2.6621990966796876,
|
| 747 |
+
"step": 2550
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 2.478344562078922,
|
| 751 |
+
"grad_norm": 4.984952926635742,
|
| 752 |
+
"learning_rate": 0.0001761528430084662,
|
| 753 |
+
"loss": 2.6337930297851564,
|
| 754 |
+
"step": 2575
|
| 755 |
+
},
|
| 756 |
+
{
|
| 757 |
+
"epoch": 2.502406159769009,
|
| 758 |
+
"grad_norm": 6.104625701904297,
|
| 759 |
+
"learning_rate": 0.00017564543410059473,
|
| 760 |
+
"loss": 2.6945245361328123,
|
| 761 |
+
"step": 2600
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 2.5264677574590952,
|
| 765 |
+
"grad_norm": 5.737064361572266,
|
| 766 |
+
"learning_rate": 0.0001751334309749655,
|
| 767 |
+
"loss": 2.6489166259765624,
|
| 768 |
+
"step": 2625
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 2.550529355149182,
|
| 772 |
+
"grad_norm": 5.924130916595459,
|
| 773 |
+
"learning_rate": 0.0001746168647273595,
|
| 774 |
+
"loss": 2.649996337890625,
|
| 775 |
+
"step": 2650
|
| 776 |
+
},
|
| 777 |
+
{
|
| 778 |
+
"epoch": 2.5745909528392685,
|
| 779 |
+
"grad_norm": 5.325296878814697,
|
| 780 |
+
"learning_rate": 0.00017409576673069245,
|
| 781 |
+
"loss": 2.6771316528320312,
|
| 782 |
+
"step": 2675
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 2.598652550529355,
|
| 786 |
+
"grad_norm": 5.358754634857178,
|
| 787 |
+
"learning_rate": 0.0001735701686331094,
|
| 788 |
+
"loss": 2.6654647827148437,
|
| 789 |
+
"step": 2700
|
| 790 |
+
},
|
| 791 |
+
{
|
| 792 |
+
"epoch": 2.622714148219442,
|
| 793 |
+
"grad_norm": 4.84153938293457,
|
| 794 |
+
"learning_rate": 0.00017304010235606257,
|
| 795 |
+
"loss": 2.63984619140625,
|
| 796 |
+
"step": 2725
|
| 797 |
+
},
|
| 798 |
+
{
|
| 799 |
+
"epoch": 2.6467757459095282,
|
| 800 |
+
"grad_norm": 5.001486778259277,
|
| 801 |
+
"learning_rate": 0.00017250560009237276,
|
| 802 |
+
"loss": 2.650990295410156,
|
| 803 |
+
"step": 2750
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 2.670837343599615,
|
| 807 |
+
"grad_norm": 5.516018867492676,
|
| 808 |
+
"learning_rate": 0.0001719666943042741,
|
| 809 |
+
"loss": 2.6812786865234375,
|
| 810 |
+
"step": 2775
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 2.6948989412897015,
|
| 814 |
+
"grad_norm": 4.497090816497803,
|
| 815 |
+
"learning_rate": 0.00017142341772144248,
|
| 816 |
+
"loss": 2.6683084106445314,
|
| 817 |
+
"step": 2800
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 2.7189605389797884,
|
| 821 |
+
"grad_norm": 5.212906360626221,
|
| 822 |
+
"learning_rate": 0.00017087580333900778,
|
| 823 |
+
"loss": 2.646448974609375,
|
| 824 |
+
"step": 2825
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 2.743022136669875,
|
| 828 |
+
"grad_norm": 5.7047600746154785,
|
| 829 |
+
"learning_rate": 0.00017032388441555,
|
| 830 |
+
"loss": 2.654071044921875,
|
| 831 |
+
"step": 2850
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 2.7670837343599617,
|
| 835 |
+
"grad_norm": 5.212134838104248,
|
| 836 |
+
"learning_rate": 0.00016976769447107928,
|
| 837 |
+
"loss": 2.635329284667969,
|
| 838 |
+
"step": 2875
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 2.791145332050048,
|
| 842 |
+
"grad_norm": 5.364657878875732,
|
| 843 |
+
"learning_rate": 0.00016920726728500017,
|
| 844 |
+
"loss": 2.6405084228515623,
|
| 845 |
+
"step": 2900
|
| 846 |
+
},
|
| 847 |
+
{
|
| 848 |
+
"epoch": 2.8152069297401345,
|
| 849 |
+
"grad_norm": 5.30003023147583,
|
| 850 |
+
"learning_rate": 0.00016864263689406002,
|
| 851 |
+
"loss": 2.656103210449219,
|
| 852 |
+
"step": 2925
|
| 853 |
+
},
|
| 854 |
+
{
|
| 855 |
+
"epoch": 2.8392685274302214,
|
| 856 |
+
"grad_norm": 5.683626174926758,
|
| 857 |
+
"learning_rate": 0.00016807383759028186,
|
| 858 |
+
"loss": 2.6557977294921873,
|
| 859 |
+
"step": 2950
|
| 860 |
+
},
|
| 861 |
+
{
|
| 862 |
+
"epoch": 2.863330125120308,
|
| 863 |
+
"grad_norm": 6.03591251373291,
|
| 864 |
+
"learning_rate": 0.00016750090391888173,
|
| 865 |
+
"loss": 2.651222839355469,
|
| 866 |
+
"step": 2975
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 2.8873917228103947,
|
| 870 |
+
"grad_norm": 5.038674354553223,
|
| 871 |
+
"learning_rate": 0.00016692387067617062,
|
| 872 |
+
"loss": 2.6435317993164062,
|
| 873 |
+
"step": 3000
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 2.911453320500481,
|
| 877 |
+
"grad_norm": 4.938141822814941,
|
| 878 |
+
"learning_rate": 0.00016634277290744103,
|
| 879 |
+
"loss": 2.64944580078125,
|
| 880 |
+
"step": 3025
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 2.935514918190568,
|
| 884 |
+
"grad_norm": 5.381541728973389,
|
| 885 |
+
"learning_rate": 0.0001657576459048388,
|
| 886 |
+
"loss": 2.665923767089844,
|
| 887 |
+
"step": 3050
|
| 888 |
+
},
|
| 889 |
+
{
|
| 890 |
+
"epoch": 2.9595765158806544,
|
| 891 |
+
"grad_norm": 5.302907943725586,
|
| 892 |
+
"learning_rate": 0.00016516852520521947,
|
| 893 |
+
"loss": 2.6620660400390626,
|
| 894 |
+
"step": 3075
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"epoch": 2.9836381135707413,
|
| 898 |
+
"grad_norm": 5.368643283843994,
|
| 899 |
+
"learning_rate": 0.00016457544658799011,
|
| 900 |
+
"loss": 2.6165301513671877,
|
| 901 |
+
"step": 3100
|
| 902 |
+
},
|
| 903 |
+
{
|
| 904 |
+
"epoch": 3.0,
|
| 905 |
+
"eval_loss": 2.7015655040740967,
|
| 906 |
+
"eval_runtime": 9.9742,
|
| 907 |
+
"eval_samples_per_second": 150.388,
|
| 908 |
+
"eval_steps_per_second": 2.406,
|
| 909 |
+
"step": 3117
|
| 910 |
+
},
|
| 911 |
+
{
|
| 912 |
+
"epoch": 3.0076997112608277,
|
| 913 |
+
"grad_norm": 5.0673828125,
|
| 914 |
+
"learning_rate": 0.00016397844607293627,
|
| 915 |
+
"loss": 2.61294677734375,
|
| 916 |
+
"step": 3125
|
| 917 |
+
},
|
| 918 |
+
{
|
| 919 |
+
"epoch": 3.0317613089509146,
|
| 920 |
+
"grad_norm": 5.165353775024414,
|
| 921 |
+
"learning_rate": 0.00016337755991803444,
|
| 922 |
+
"loss": 2.4642942810058592,
|
| 923 |
+
"step": 3150
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"epoch": 3.055822906641001,
|
| 927 |
+
"grad_norm": 5.3854079246521,
|
| 928 |
+
"learning_rate": 0.00016277282461724988,
|
| 929 |
+
"loss": 2.4820993041992185,
|
| 930 |
+
"step": 3175
|
| 931 |
+
},
|
| 932 |
+
{
|
| 933 |
+
"epoch": 3.0798845043310874,
|
| 934 |
+
"grad_norm": 5.312911510467529,
|
| 935 |
+
"learning_rate": 0.00016216427689832015,
|
| 936 |
+
"loss": 2.489494171142578,
|
| 937 |
+
"step": 3200
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 3.1039461020211743,
|
| 941 |
+
"grad_norm": 5.348562240600586,
|
| 942 |
+
"learning_rate": 0.0001615519537205247,
|
| 943 |
+
"loss": 2.4705026245117185,
|
| 944 |
+
"step": 3225
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 3.1280076997112607,
|
| 948 |
+
"grad_norm": 5.841189384460449,
|
| 949 |
+
"learning_rate": 0.00016093589227244017,
|
| 950 |
+
"loss": 2.46598388671875,
|
| 951 |
+
"step": 3250
|
| 952 |
+
},
|
| 953 |
+
{
|
| 954 |
+
"epoch": 3.1520692974013476,
|
| 955 |
+
"grad_norm": 5.61829948425293,
|
| 956 |
+
"learning_rate": 0.00016031612996968156,
|
| 957 |
+
"loss": 2.4969444274902344,
|
| 958 |
+
"step": 3275
|
| 959 |
+
},
|
| 960 |
+
{
|
| 961 |
+
"epoch": 3.176130895091434,
|
| 962 |
+
"grad_norm": 5.292537212371826,
|
| 963 |
+
"learning_rate": 0.0001596927044526301,
|
| 964 |
+
"loss": 2.473064422607422,
|
| 965 |
+
"step": 3300
|
| 966 |
+
},
|
| 967 |
+
{
|
| 968 |
+
"epoch": 3.200192492781521,
|
| 969 |
+
"grad_norm": 5.158451557159424,
|
| 970 |
+
"learning_rate": 0.0001590656535841471,
|
| 971 |
+
"loss": 2.4472772216796876,
|
| 972 |
+
"step": 3325
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 3.2242540904716073,
|
| 976 |
+
"grad_norm": 5.177428722381592,
|
| 977 |
+
"learning_rate": 0.00015843501544727446,
|
| 978 |
+
"loss": 2.5115626525878905,
|
| 979 |
+
"step": 3350
|
| 980 |
+
},
|
| 981 |
+
{
|
| 982 |
+
"epoch": 3.248315688161694,
|
| 983 |
+
"grad_norm": 4.971099853515625,
|
| 984 |
+
"learning_rate": 0.00015780082834292172,
|
| 985 |
+
"loss": 2.5359817504882813,
|
| 986 |
+
"step": 3375
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 3.2723772858517806,
|
| 990 |
+
"grad_norm": 5.984713554382324,
|
| 991 |
+
"learning_rate": 0.00015716313078753986,
|
| 992 |
+
"loss": 2.518363342285156,
|
| 993 |
+
"step": 3400
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 3.2964388835418674,
|
| 997 |
+
"grad_norm": 5.768255233764648,
|
| 998 |
+
"learning_rate": 0.00015652196151078217,
|
| 999 |
+
"loss": 2.4845159912109374,
|
| 1000 |
+
"step": 3425
|
| 1001 |
+
},
|
| 1002 |
+
{
|
| 1003 |
+
"epoch": 3.320500481231954,
|
| 1004 |
+
"grad_norm": 5.554891586303711,
|
| 1005 |
+
"learning_rate": 0.0001558773594531519,
|
| 1006 |
+
"loss": 2.521295166015625,
|
| 1007 |
+
"step": 3450
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 3.3445620789220403,
|
| 1011 |
+
"grad_norm": 5.35845422744751,
|
| 1012 |
+
"learning_rate": 0.00015522936376363746,
|
| 1013 |
+
"loss": 2.5144276428222656,
|
| 1014 |
+
"step": 3475
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 3.368623676612127,
|
| 1018 |
+
"grad_norm": 5.9035210609436035,
|
| 1019 |
+
"learning_rate": 0.00015457801379733464,
|
| 1020 |
+
"loss": 2.4923356628417968,
|
| 1021 |
+
"step": 3500
|
| 1022 |
+
},
|
| 1023 |
+
{
|
| 1024 |
+
"epoch": 3.3926852743022136,
|
| 1025 |
+
"grad_norm": 5.4024786949157715,
|
| 1026 |
+
"learning_rate": 0.0001539233491130564,
|
| 1027 |
+
"loss": 2.5232693481445314,
|
| 1028 |
+
"step": 3525
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 3.4167468719923004,
|
| 1032 |
+
"grad_norm": 5.661716461181641,
|
| 1033 |
+
"learning_rate": 0.00015326540947093042,
|
| 1034 |
+
"loss": 2.510073699951172,
|
| 1035 |
+
"step": 3550
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 3.440808469682387,
|
| 1039 |
+
"grad_norm": 5.541335582733154,
|
| 1040 |
+
"learning_rate": 0.00015260423482998427,
|
| 1041 |
+
"loss": 2.512746124267578,
|
| 1042 |
+
"step": 3575
|
| 1043 |
+
},
|
| 1044 |
+
{
|
| 1045 |
+
"epoch": 3.4648700673724737,
|
| 1046 |
+
"grad_norm": 5.722592353820801,
|
| 1047 |
+
"learning_rate": 0.00015193986534571863,
|
| 1048 |
+
"loss": 2.5191300964355468,
|
| 1049 |
+
"step": 3600
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 3.48893166506256,
|
| 1053 |
+
"grad_norm": 6.089288711547852,
|
| 1054 |
+
"learning_rate": 0.0001512723413676683,
|
| 1055 |
+
"loss": 2.4799041748046875,
|
| 1056 |
+
"step": 3625
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 3.512993262752647,
|
| 1060 |
+
"grad_norm": 5.490766525268555,
|
| 1061 |
+
"learning_rate": 0.00015060170343695196,
|
| 1062 |
+
"loss": 2.501981658935547,
|
| 1063 |
+
"step": 3650
|
| 1064 |
+
},
|
| 1065 |
+
{
|
| 1066 |
+
"epoch": 3.5370548604427334,
|
| 1067 |
+
"grad_norm": 5.429261684417725,
|
| 1068 |
+
"learning_rate": 0.0001499279922838096,
|
| 1069 |
+
"loss": 2.528828430175781,
|
| 1070 |
+
"step": 3675
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 3.56111645813282,
|
| 1074 |
+
"grad_norm": 5.104456424713135,
|
| 1075 |
+
"learning_rate": 0.00014925124882512908,
|
| 1076 |
+
"loss": 2.506761016845703,
|
| 1077 |
+
"step": 3700
|
| 1078 |
+
},
|
| 1079 |
+
{
|
| 1080 |
+
"epoch": 3.5851780558229067,
|
| 1081 |
+
"grad_norm": 5.07927942276001,
|
| 1082 |
+
"learning_rate": 0.000148571514161961,
|
| 1083 |
+
"loss": 2.476543121337891,
|
| 1084 |
+
"step": 3725
|
| 1085 |
+
},
|
| 1086 |
+
{
|
| 1087 |
+
"epoch": 3.609239653512993,
|
| 1088 |
+
"grad_norm": 5.794335842132568,
|
| 1089 |
+
"learning_rate": 0.00014788882957702257,
|
| 1090 |
+
"loss": 2.5332624816894533,
|
| 1091 |
+
"step": 3750
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 3.63330125120308,
|
| 1095 |
+
"grad_norm": 5.124516010284424,
|
| 1096 |
+
"learning_rate": 0.00014720323653219018,
|
| 1097 |
+
"loss": 2.5622378540039064,
|
| 1098 |
+
"step": 3775
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"epoch": 3.6573628488931664,
|
| 1102 |
+
"grad_norm": 5.550374984741211,
|
| 1103 |
+
"learning_rate": 0.00014651477666598151,
|
| 1104 |
+
"loss": 2.531949462890625,
|
| 1105 |
+
"step": 3800
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 3.681424446583253,
|
| 1109 |
+
"grad_norm": 5.177489280700684,
|
| 1110 |
+
"learning_rate": 0.00014582349179102644,
|
| 1111 |
+
"loss": 2.5247061157226565,
|
| 1112 |
+
"step": 3825
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 3.7054860442733397,
|
| 1116 |
+
"grad_norm": 5.406796932220459,
|
| 1117 |
+
"learning_rate": 0.00014512942389152783,
|
| 1118 |
+
"loss": 2.5437385559082033,
|
| 1119 |
+
"step": 3850
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 3.7295476419634266,
|
| 1123 |
+
"grad_norm": 5.12544584274292,
|
| 1124 |
+
"learning_rate": 0.00014443261512071156,
|
| 1125 |
+
"loss": 2.4803817749023436,
|
| 1126 |
+
"step": 3875
|
| 1127 |
+
},
|
| 1128 |
+
{
|
| 1129 |
+
"epoch": 3.753609239653513,
|
| 1130 |
+
"grad_norm": 4.869628429412842,
|
| 1131 |
+
"learning_rate": 0.00014373310779826638,
|
| 1132 |
+
"loss": 2.5459124755859377,
|
| 1133 |
+
"step": 3900
|
| 1134 |
+
},
|
| 1135 |
+
{
|
| 1136 |
+
"epoch": 3.7776708373435994,
|
| 1137 |
+
"grad_norm": 5.2594099044799805,
|
| 1138 |
+
"learning_rate": 0.0001430309444077738,
|
| 1139 |
+
"loss": 2.4942247009277345,
|
| 1140 |
+
"step": 3925
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"epoch": 3.8017324350336863,
|
| 1144 |
+
"grad_norm": 5.267584800720215,
|
| 1145 |
+
"learning_rate": 0.00014232616759412772,
|
| 1146 |
+
"loss": 2.5201406860351563,
|
| 1147 |
+
"step": 3950
|
| 1148 |
+
},
|
| 1149 |
+
{
|
| 1150 |
+
"epoch": 3.8257940327237727,
|
| 1151 |
+
"grad_norm": 5.06338357925415,
|
| 1152 |
+
"learning_rate": 0.00014161882016094473,
|
| 1153 |
+
"loss": 2.5072589111328125,
|
| 1154 |
+
"step": 3975
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 3.8498556304138596,
|
| 1158 |
+
"grad_norm": 5.17350435256958,
|
| 1159 |
+
"learning_rate": 0.00014090894506796426,
|
| 1160 |
+
"loss": 2.54510986328125,
|
| 1161 |
+
"step": 4000
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 3.873917228103946,
|
| 1165 |
+
"grad_norm": 4.9889678955078125,
|
| 1166 |
+
"learning_rate": 0.00014019658542843951,
|
| 1167 |
+
"loss": 2.4879167175292967,
|
| 1168 |
+
"step": 4025
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 3.897978825794033,
|
| 1172 |
+
"grad_norm": 6.220742225646973,
|
| 1173 |
+
"learning_rate": 0.0001394817845065192,
|
| 1174 |
+
"loss": 2.517736358642578,
|
| 1175 |
+
"step": 4050
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 3.9220404234841193,
|
| 1179 |
+
"grad_norm": 5.632843017578125,
|
| 1180 |
+
"learning_rate": 0.00013876458571461978,
|
| 1181 |
+
"loss": 2.4728121948242188,
|
| 1182 |
+
"step": 4075
|
| 1183 |
+
},
|
| 1184 |
+
{
|
| 1185 |
+
"epoch": 3.946102021174206,
|
| 1186 |
+
"grad_norm": 5.72349214553833,
|
| 1187 |
+
"learning_rate": 0.0001380450326107889,
|
| 1188 |
+
"loss": 2.5195367431640623,
|
| 1189 |
+
"step": 4100
|
| 1190 |
+
},
|
| 1191 |
+
{
|
| 1192 |
+
"epoch": 3.9701636188642926,
|
| 1193 |
+
"grad_norm": 6.055045127868652,
|
| 1194 |
+
"learning_rate": 0.00013732316889606005,
|
| 1195 |
+
"loss": 2.530181579589844,
|
| 1196 |
+
"step": 4125
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 3.994225216554379,
|
| 1200 |
+
"grad_norm": 4.998103141784668,
|
| 1201 |
+
"learning_rate": 0.00013659903841179832,
|
| 1202 |
+
"loss": 2.5351657104492187,
|
| 1203 |
+
"step": 4150
|
| 1204 |
+
},
|
| 1205 |
+
{
|
| 1206 |
+
"epoch": 4.0,
|
| 1207 |
+
"eval_loss": 2.6353492736816406,
|
| 1208 |
+
"eval_runtime": 10.2956,
|
| 1209 |
+
"eval_samples_per_second": 145.693,
|
| 1210 |
+
"eval_steps_per_second": 2.331,
|
| 1211 |
+
"step": 4156
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 4.018286814244465,
|
| 1215 |
+
"grad_norm": 5.625172138214111,
|
| 1216 |
+
"learning_rate": 0.00013587268513703794,
|
| 1217 |
+
"loss": 2.335594482421875,
|
| 1218 |
+
"step": 4175
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 4.042348411934553,
|
| 1222 |
+
"grad_norm": 5.737284183502197,
|
| 1223 |
+
"learning_rate": 0.00013514415318581103,
|
| 1224 |
+
"loss": 2.3087588500976564,
|
| 1225 |
+
"step": 4200
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 4.066410009624639,
|
| 1229 |
+
"grad_norm": 5.879457473754883,
|
| 1230 |
+
"learning_rate": 0.0001344134868044686,
|
| 1231 |
+
"loss": 2.3059092712402345,
|
| 1232 |
+
"step": 4225
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 4.090471607314726,
|
| 1236 |
+
"grad_norm": 5.486513137817383,
|
| 1237 |
+
"learning_rate": 0.00013368073036899313,
|
| 1238 |
+
"loss": 2.299142150878906,
|
| 1239 |
+
"step": 4250
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 4.114533205004812,
|
| 1243 |
+
"grad_norm": 5.92104959487915,
|
| 1244 |
+
"learning_rate": 0.00013294592838230374,
|
| 1245 |
+
"loss": 2.336262512207031,
|
| 1246 |
+
"step": 4275
|
| 1247 |
+
},
|
| 1248 |
+
{
|
| 1249 |
+
"epoch": 4.138594802694899,
|
| 1250 |
+
"grad_norm": 6.021739482879639,
|
| 1251 |
+
"learning_rate": 0.000132209125471553,
|
| 1252 |
+
"loss": 2.333507385253906,
|
| 1253 |
+
"step": 4300
|
| 1254 |
+
},
|
| 1255 |
+
{
|
| 1256 |
+
"epoch": 4.162656400384986,
|
| 1257 |
+
"grad_norm": 5.839460372924805,
|
| 1258 |
+
"learning_rate": 0.00013147036638541697,
|
| 1259 |
+
"loss": 2.369920349121094,
|
| 1260 |
+
"step": 4325
|
| 1261 |
+
},
|
| 1262 |
+
{
|
| 1263 |
+
"epoch": 4.186717998075072,
|
| 1264 |
+
"grad_norm": 5.582374572753906,
|
| 1265 |
+
"learning_rate": 0.00013072969599137697,
|
| 1266 |
+
"loss": 2.3627223205566406,
|
| 1267 |
+
"step": 4350
|
| 1268 |
+
},
|
| 1269 |
+
{
|
| 1270 |
+
"epoch": 4.210779595765159,
|
| 1271 |
+
"grad_norm": 5.355888843536377,
|
| 1272 |
+
"learning_rate": 0.00012998715927299513,
|
| 1273 |
+
"loss": 2.379355163574219,
|
| 1274 |
+
"step": 4375
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 4.234841193455245,
|
| 1278 |
+
"grad_norm": 5.518582820892334,
|
| 1279 |
+
"learning_rate": 0.00012924280132718197,
|
| 1280 |
+
"loss": 2.351807861328125,
|
| 1281 |
+
"step": 4400
|
| 1282 |
+
},
|
| 1283 |
+
{
|
| 1284 |
+
"epoch": 4.258902791145332,
|
| 1285 |
+
"grad_norm": 5.967413902282715,
|
| 1286 |
+
"learning_rate": 0.00012849666736145773,
|
| 1287 |
+
"loss": 2.3464089965820314,
|
| 1288 |
+
"step": 4425
|
| 1289 |
+
},
|
| 1290 |
+
{
|
| 1291 |
+
"epoch": 4.282964388835419,
|
| 1292 |
+
"grad_norm": 5.655083656311035,
|
| 1293 |
+
"learning_rate": 0.0001277488026912067,
|
| 1294 |
+
"loss": 2.3598651123046874,
|
| 1295 |
+
"step": 4450
|
| 1296 |
+
},
|
| 1297 |
+
{
|
| 1298 |
+
"epoch": 4.307025986525505,
|
| 1299 |
+
"grad_norm": 5.970212459564209,
|
| 1300 |
+
"learning_rate": 0.00012699925273692507,
|
| 1301 |
+
"loss": 2.3835829162597655,
|
| 1302 |
+
"step": 4475
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"epoch": 4.331087584215592,
|
| 1306 |
+
"grad_norm": 5.315586090087891,
|
| 1307 |
+
"learning_rate": 0.00012624806302146224,
|
| 1308 |
+
"loss": 2.3316766357421876,
|
| 1309 |
+
"step": 4500
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 4.355149181905679,
|
| 1313 |
+
"grad_norm": 5.3828229904174805,
|
| 1314 |
+
"learning_rate": 0.00012549527916725633,
|
| 1315 |
+
"loss": 2.3893943786621095,
|
| 1316 |
+
"step": 4525
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 4.379210779595765,
|
| 1320 |
+
"grad_norm": 5.529123306274414,
|
| 1321 |
+
"learning_rate": 0.00012474094689356307,
|
| 1322 |
+
"loss": 2.353526763916016,
|
| 1323 |
+
"step": 4550
|
| 1324 |
+
},
|
| 1325 |
+
{
|
| 1326 |
+
"epoch": 4.403272377285852,
|
| 1327 |
+
"grad_norm": 6.203123092651367,
|
| 1328 |
+
"learning_rate": 0.00012398511201367927,
|
| 1329 |
+
"loss": 2.392206726074219,
|
| 1330 |
+
"step": 4575
|
| 1331 |
+
},
|
| 1332 |
+
{
|
| 1333 |
+
"epoch": 4.427333974975938,
|
| 1334 |
+
"grad_norm": 5.528677940368652,
|
| 1335 |
+
"learning_rate": 0.00012322782043216053,
|
| 1336 |
+
"loss": 2.4033900451660157,
|
| 1337 |
+
"step": 4600
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 4.451395572666025,
|
| 1341 |
+
"grad_norm": 5.775765419006348,
|
| 1342 |
+
"learning_rate": 0.00012246911814203294,
|
| 1343 |
+
"loss": 2.3578640747070314,
|
| 1344 |
+
"step": 4625
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 4.475457170356112,
|
| 1348 |
+
"grad_norm": 5.77819299697876,
|
| 1349 |
+
"learning_rate": 0.00012170905122200011,
|
| 1350 |
+
"loss": 2.3865695190429688,
|
| 1351 |
+
"step": 4650
|
| 1352 |
+
},
|
| 1353 |
+
{
|
| 1354 |
+
"epoch": 4.499518768046198,
|
| 1355 |
+
"grad_norm": 5.330785751342773,
|
| 1356 |
+
"learning_rate": 0.00012094766583364445,
|
| 1357 |
+
"loss": 2.4133177185058594,
|
| 1358 |
+
"step": 4675
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 4.523580365736285,
|
| 1362 |
+
"grad_norm": 6.005075931549072,
|
| 1363 |
+
"learning_rate": 0.00012018500821862369,
|
| 1364 |
+
"loss": 2.3767886352539063,
|
| 1365 |
+
"step": 4700
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 4.547641963426371,
|
| 1369 |
+
"grad_norm": 5.781045436859131,
|
| 1370 |
+
"learning_rate": 0.00011942112469586242,
|
| 1371 |
+
"loss": 2.3768569946289064,
|
| 1372 |
+
"step": 4725
|
| 1373 |
+
},
|
| 1374 |
+
{
|
| 1375 |
+
"epoch": 4.5717035611164585,
|
| 1376 |
+
"grad_norm": 6.250720024108887,
|
| 1377 |
+
"learning_rate": 0.00011865606165873897,
|
| 1378 |
+
"loss": 2.367838134765625,
|
| 1379 |
+
"step": 4750
|
| 1380 |
+
},
|
| 1381 |
+
{
|
| 1382 |
+
"epoch": 4.595765158806545,
|
| 1383 |
+
"grad_norm": 5.679470539093018,
|
| 1384 |
+
"learning_rate": 0.00011788986557226787,
|
| 1385 |
+
"loss": 2.3778204345703124,
|
| 1386 |
+
"step": 4775
|
| 1387 |
+
},
|
| 1388 |
+
{
|
| 1389 |
+
"epoch": 4.619826756496631,
|
| 1390 |
+
"grad_norm": 5.813436985015869,
|
| 1391 |
+
"learning_rate": 0.00011712258297027772,
|
| 1392 |
+
"loss": 2.44917724609375,
|
| 1393 |
+
"step": 4800
|
| 1394 |
+
},
|
| 1395 |
+
{
|
| 1396 |
+
"epoch": 4.643888354186718,
|
| 1397 |
+
"grad_norm": 5.503162860870361,
|
| 1398 |
+
"learning_rate": 0.00011635426045258508,
|
| 1399 |
+
"loss": 2.3535726928710936,
|
| 1400 |
+
"step": 4825
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 4.667949951876805,
|
| 1404 |
+
"grad_norm": 5.896456718444824,
|
| 1405 |
+
"learning_rate": 0.00011558494468216439,
|
| 1406 |
+
"loss": 2.3747299194335936,
|
| 1407 |
+
"step": 4850
|
| 1408 |
+
},
|
| 1409 |
+
{
|
| 1410 |
+
"epoch": 4.6920115495668915,
|
| 1411 |
+
"grad_norm": 5.829119682312012,
|
| 1412 |
+
"learning_rate": 0.00011481468238231378,
|
| 1413 |
+
"loss": 2.371831359863281,
|
| 1414 |
+
"step": 4875
|
| 1415 |
+
},
|
| 1416 |
+
{
|
| 1417 |
+
"epoch": 4.716073147256978,
|
| 1418 |
+
"grad_norm": 5.434453010559082,
|
| 1419 |
+
"learning_rate": 0.00011404352033381764,
|
| 1420 |
+
"loss": 2.3506382751464843,
|
| 1421 |
+
"step": 4900
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 4.740134744947064,
|
| 1425 |
+
"grad_norm": 5.646220684051514,
|
| 1426 |
+
"learning_rate": 0.00011327150537210516,
|
| 1427 |
+
"loss": 2.408179931640625,
|
| 1428 |
+
"step": 4925
|
| 1429 |
+
},
|
| 1430 |
+
{
|
| 1431 |
+
"epoch": 4.764196342637151,
|
| 1432 |
+
"grad_norm": 5.965972900390625,
|
| 1433 |
+
"learning_rate": 0.0001124986843844061,
|
| 1434 |
+
"loss": 2.402149505615234,
|
| 1435 |
+
"step": 4950
|
| 1436 |
+
},
|
| 1437 |
+
{
|
| 1438 |
+
"epoch": 4.788257940327238,
|
| 1439 |
+
"grad_norm": 5.979576110839844,
|
| 1440 |
+
"learning_rate": 0.00011172510430690298,
|
| 1441 |
+
"loss": 2.3634904479980468,
|
| 1442 |
+
"step": 4975
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 4.8123195380173245,
|
| 1446 |
+
"grad_norm": 5.2546210289001465,
|
| 1447 |
+
"learning_rate": 0.00011095081212188061,
|
| 1448 |
+
"loss": 2.370100250244141,
|
| 1449 |
+
"step": 5000
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 4.836381135707411,
|
| 1453 |
+
"grad_norm": 5.855418682098389,
|
| 1454 |
+
"learning_rate": 0.0001101758548548727,
|
| 1455 |
+
"loss": 2.3603744506835938,
|
| 1456 |
+
"step": 5025
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 4.860442733397497,
|
| 1460 |
+
"grad_norm": 5.719959735870361,
|
| 1461 |
+
"learning_rate": 0.00010940027957180564,
|
| 1462 |
+
"loss": 2.3629183959960938,
|
| 1463 |
+
"step": 5050
|
| 1464 |
+
},
|
| 1465 |
+
{
|
| 1466 |
+
"epoch": 4.884504331087584,
|
| 1467 |
+
"grad_norm": 5.5997843742370605,
|
| 1468 |
+
"learning_rate": 0.00010862413337614022,
|
| 1469 |
+
"loss": 2.406308898925781,
|
| 1470 |
+
"step": 5075
|
| 1471 |
+
},
|
| 1472 |
+
{
|
| 1473 |
+
"epoch": 4.908565928777671,
|
| 1474 |
+
"grad_norm": 5.469118118286133,
|
| 1475 |
+
"learning_rate": 0.00010784746340601078,
|
| 1476 |
+
"loss": 2.3328375244140624,
|
| 1477 |
+
"step": 5100
|
| 1478 |
+
},
|
| 1479 |
+
{
|
| 1480 |
+
"epoch": 4.9326275264677575,
|
| 1481 |
+
"grad_norm": 6.1267781257629395,
|
| 1482 |
+
"learning_rate": 0.00010707031683136231,
|
| 1483 |
+
"loss": 2.3826669311523436,
|
| 1484 |
+
"step": 5125
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 4.956689124157844,
|
| 1488 |
+
"grad_norm": 5.942570686340332,
|
| 1489 |
+
"learning_rate": 0.00010629274085108582,
|
| 1490 |
+
"loss": 2.396967315673828,
|
| 1491 |
+
"step": 5150
|
| 1492 |
+
},
|
| 1493 |
+
{
|
| 1494 |
+
"epoch": 4.98075072184793,
|
| 1495 |
+
"grad_norm": 5.1797709465026855,
|
| 1496 |
+
"learning_rate": 0.00010551478269015153,
|
| 1497 |
+
"loss": 2.3436398315429687,
|
| 1498 |
+
"step": 5175
|
| 1499 |
+
},
|
| 1500 |
+
{
|
| 1501 |
+
"epoch": 5.0,
|
| 1502 |
+
"eval_loss": 2.573817491531372,
|
| 1503 |
+
"eval_runtime": 11.4346,
|
| 1504 |
+
"eval_samples_per_second": 131.181,
|
| 1505 |
+
"eval_steps_per_second": 2.099,
|
| 1506 |
+
"step": 5195
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"epoch": 5.004812319538018,
|
| 1510 |
+
"grad_norm": 6.427822113037109,
|
| 1511 |
+
"learning_rate": 0.00010473648959674096,
|
| 1512 |
+
"loss": 2.323263702392578,
|
| 1513 |
+
"step": 5200
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 5.028873917228104,
|
| 1517 |
+
"grad_norm": 6.104551315307617,
|
| 1518 |
+
"learning_rate": 0.00010395790883937718,
|
| 1519 |
+
"loss": 2.1802984619140626,
|
| 1520 |
+
"step": 5225
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 5.0529355149181905,
|
| 1524 |
+
"grad_norm": 6.109057426452637,
|
| 1525 |
+
"learning_rate": 0.00010317908770405415,
|
| 1526 |
+
"loss": 2.2006021118164063,
|
| 1527 |
+
"step": 5250
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 5.076997112608277,
|
| 1531 |
+
"grad_norm": 6.017938137054443,
|
| 1532 |
+
"learning_rate": 0.00010240007349136489,
|
| 1533 |
+
"loss": 2.1716549682617186,
|
| 1534 |
+
"step": 5275
|
| 1535 |
+
},
|
| 1536 |
+
{
|
| 1537 |
+
"epoch": 5.101058710298364,
|
| 1538 |
+
"grad_norm": 6.390856742858887,
|
| 1539 |
+
"learning_rate": 0.00010162091351362868,
|
| 1540 |
+
"loss": 2.1978250122070313,
|
| 1541 |
+
"step": 5300
|
| 1542 |
+
},
|
| 1543 |
+
{
|
| 1544 |
+
"epoch": 5.125120307988451,
|
| 1545 |
+
"grad_norm": 6.782686710357666,
|
| 1546 |
+
"learning_rate": 0.00010084165509201764,
|
| 1547 |
+
"loss": 2.2078280639648438,
|
| 1548 |
+
"step": 5325
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"epoch": 5.149181905678537,
|
| 1552 |
+
"grad_norm": 6.1608500480651855,
|
| 1553 |
+
"learning_rate": 0.00010006234555368265,
|
| 1554 |
+
"loss": 2.2312646484375,
|
| 1555 |
+
"step": 5350
|
| 1556 |
+
},
|
| 1557 |
+
{
|
| 1558 |
+
"epoch": 5.1732435033686235,
|
| 1559 |
+
"grad_norm": 6.1210551261901855,
|
| 1560 |
+
"learning_rate": 9.928303222887925e-05,
|
| 1561 |
+
"loss": 2.1966952514648437,
|
| 1562 |
+
"step": 5375
|
| 1563 |
+
},
|
| 1564 |
+
{
|
| 1565 |
+
"epoch": 5.19730510105871,
|
| 1566 |
+
"grad_norm": 6.189257621765137,
|
| 1567 |
+
"learning_rate": 9.850376244809278e-05,
|
| 1568 |
+
"loss": 2.204204406738281,
|
| 1569 |
+
"step": 5400
|
| 1570 |
+
},
|
| 1571 |
+
{
|
| 1572 |
+
"epoch": 5.221366698748797,
|
| 1573 |
+
"grad_norm": 6.595432758331299,
|
| 1574 |
+
"learning_rate": 9.772458353916412e-05,
|
| 1575 |
+
"loss": 2.218436279296875,
|
| 1576 |
+
"step": 5425
|
| 1577 |
+
},
|
| 1578 |
+
{
|
| 1579 |
+
"epoch": 5.245428296438884,
|
| 1580 |
+
"grad_norm": 6.288983345031738,
|
| 1581 |
+
"learning_rate": 9.694554282441514e-05,
|
| 1582 |
+
"loss": 2.220950927734375,
|
| 1583 |
+
"step": 5450
|
| 1584 |
+
},
|
| 1585 |
+
{
|
| 1586 |
+
"epoch": 5.26948989412897,
|
| 1587 |
+
"grad_norm": 5.690927028656006,
|
| 1588 |
+
"learning_rate": 9.616668761777464e-05,
|
| 1589 |
+
"loss": 2.2794346618652344,
|
| 1590 |
+
"step": 5475
|
| 1591 |
+
},
|
| 1592 |
+
{
|
| 1593 |
+
"epoch": 5.2935514918190565,
|
| 1594 |
+
"grad_norm": 6.080009937286377,
|
| 1595 |
+
"learning_rate": 9.53880652219049e-05,
|
| 1596 |
+
"loss": 2.2308926391601562,
|
| 1597 |
+
"step": 5500
|
| 1598 |
+
},
|
| 1599 |
+
{
|
| 1600 |
+
"epoch": 5.317613089509144,
|
| 1601 |
+
"grad_norm": 6.3759589195251465,
|
| 1602 |
+
"learning_rate": 9.460972292532875e-05,
|
| 1603 |
+
"loss": 2.226695251464844,
|
| 1604 |
+
"step": 5525
|
| 1605 |
+
},
|
| 1606 |
+
{
|
| 1607 |
+
"epoch": 5.34167468719923,
|
| 1608 |
+
"grad_norm": 6.490814685821533,
|
| 1609 |
+
"learning_rate": 9.383170799955761e-05,
|
| 1610 |
+
"loss": 2.235789794921875,
|
| 1611 |
+
"step": 5550
|
| 1612 |
+
},
|
| 1613 |
+
{
|
| 1614 |
+
"epoch": 5.365736284889317,
|
| 1615 |
+
"grad_norm": 6.471175193786621,
|
| 1616 |
+
"learning_rate": 9.305406769622046e-05,
|
| 1617 |
+
"loss": 2.2033010864257814,
|
| 1618 |
+
"step": 5575
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 5.389797882579403,
|
| 1622 |
+
"grad_norm": 6.464748859405518,
|
| 1623 |
+
"learning_rate": 9.227684924419414e-05,
|
| 1624 |
+
"loss": 2.250180511474609,
|
| 1625 |
+
"step": 5600
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 5.4138594802694895,
|
| 1629 |
+
"grad_norm": 6.329024791717529,
|
| 1630 |
+
"learning_rate": 9.150009984673498e-05,
|
| 1631 |
+
"loss": 2.223790588378906,
|
| 1632 |
+
"step": 5625
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 5.437921077959577,
|
| 1636 |
+
"grad_norm": 5.941084384918213,
|
| 1637 |
+
"learning_rate": 9.072386667861186e-05,
|
| 1638 |
+
"loss": 2.2452133178710936,
|
| 1639 |
+
"step": 5650
|
| 1640 |
+
},
|
| 1641 |
+
{
|
| 1642 |
+
"epoch": 5.461982675649663,
|
| 1643 |
+
"grad_norm": 6.427567481994629,
|
| 1644 |
+
"learning_rate": 8.994819688324133e-05,
|
| 1645 |
+
"loss": 2.192438507080078,
|
| 1646 |
+
"step": 5675
|
| 1647 |
+
},
|
| 1648 |
+
{
|
| 1649 |
+
"epoch": 5.48604427333975,
|
| 1650 |
+
"grad_norm": 5.633625030517578,
|
| 1651 |
+
"learning_rate": 8.917313756982422e-05,
|
| 1652 |
+
"loss": 2.231309661865234,
|
| 1653 |
+
"step": 5700
|
| 1654 |
+
},
|
| 1655 |
+
{
|
| 1656 |
+
"epoch": 5.510105871029836,
|
| 1657 |
+
"grad_norm": 6.054250717163086,
|
| 1658 |
+
"learning_rate": 8.839873581048461e-05,
|
| 1659 |
+
"loss": 2.200685577392578,
|
| 1660 |
+
"step": 5725
|
| 1661 |
+
},
|
| 1662 |
+
{
|
| 1663 |
+
"epoch": 5.534167468719923,
|
| 1664 |
+
"grad_norm": 7.056374549865723,
|
| 1665 |
+
"learning_rate": 8.7625038637411e-05,
|
| 1666 |
+
"loss": 2.26966796875,
|
| 1667 |
+
"step": 5750
|
| 1668 |
+
},
|
| 1669 |
+
{
|
| 1670 |
+
"epoch": 5.55822906641001,
|
| 1671 |
+
"grad_norm": 5.970189571380615,
|
| 1672 |
+
"learning_rate": 8.685209303999979e-05,
|
| 1673 |
+
"loss": 2.2440625,
|
| 1674 |
+
"step": 5775
|
| 1675 |
+
},
|
| 1676 |
+
{
|
| 1677 |
+
"epoch": 5.582290664100096,
|
| 1678 |
+
"grad_norm": 6.584507465362549,
|
| 1679 |
+
"learning_rate": 8.607994596200156e-05,
|
| 1680 |
+
"loss": 2.2399227905273436,
|
| 1681 |
+
"step": 5800
|
| 1682 |
+
},
|
| 1683 |
+
{
|
| 1684 |
+
"epoch": 5.606352261790183,
|
| 1685 |
+
"grad_norm": 6.909767150878906,
|
| 1686 |
+
"learning_rate": 8.530864429866996e-05,
|
| 1687 |
+
"loss": 2.273231201171875,
|
| 1688 |
+
"step": 5825
|
| 1689 |
+
},
|
| 1690 |
+
{
|
| 1691 |
+
"epoch": 5.630413859480269,
|
| 1692 |
+
"grad_norm": 6.671900749206543,
|
| 1693 |
+
"learning_rate": 8.453823489391352e-05,
|
| 1694 |
+
"loss": 2.249536895751953,
|
| 1695 |
+
"step": 5850
|
| 1696 |
+
},
|
| 1697 |
+
{
|
| 1698 |
+
"epoch": 5.654475457170356,
|
| 1699 |
+
"grad_norm": 6.130863189697266,
|
| 1700 |
+
"learning_rate": 8.376876453745071e-05,
|
| 1701 |
+
"loss": 2.226740417480469,
|
| 1702 |
+
"step": 5875
|
| 1703 |
+
},
|
| 1704 |
+
{
|
| 1705 |
+
"epoch": 5.678537054860443,
|
| 1706 |
+
"grad_norm": 6.046282768249512,
|
| 1707 |
+
"learning_rate": 8.300027996196837e-05,
|
| 1708 |
+
"loss": 2.2287556457519533,
|
| 1709 |
+
"step": 5900
|
| 1710 |
+
},
|
| 1711 |
+
{
|
| 1712 |
+
"epoch": 5.702598652550529,
|
| 1713 |
+
"grad_norm": 7.050989627838135,
|
| 1714 |
+
"learning_rate": 8.22328278402832e-05,
|
| 1715 |
+
"loss": 2.2470748901367186,
|
| 1716 |
+
"step": 5925
|
| 1717 |
+
},
|
| 1718 |
+
{
|
| 1719 |
+
"epoch": 5.726660250240616,
|
| 1720 |
+
"grad_norm": 5.74175500869751,
|
| 1721 |
+
"learning_rate": 8.146645478250733e-05,
|
| 1722 |
+
"loss": 2.205854797363281,
|
| 1723 |
+
"step": 5950
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 5.750721847930703,
|
| 1727 |
+
"grad_norm": 6.062502861022949,
|
| 1728 |
+
"learning_rate": 8.070120733321746e-05,
|
| 1729 |
+
"loss": 2.2029763793945314,
|
| 1730 |
+
"step": 5975
|
| 1731 |
+
},
|
| 1732 |
+
{
|
| 1733 |
+
"epoch": 5.774783445620789,
|
| 1734 |
+
"grad_norm": 6.140648365020752,
|
| 1735 |
+
"learning_rate": 7.993713196862813e-05,
|
| 1736 |
+
"loss": 2.240644836425781,
|
| 1737 |
+
"step": 6000
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 5.798845043310876,
|
| 1741 |
+
"grad_norm": 6.486162185668945,
|
| 1742 |
+
"learning_rate": 7.917427509376889e-05,
|
| 1743 |
+
"loss": 2.235250701904297,
|
| 1744 |
+
"step": 6025
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 5.822906641000962,
|
| 1748 |
+
"grad_norm": 6.734943866729736,
|
| 1749 |
+
"learning_rate": 7.841268303966612e-05,
|
| 1750 |
+
"loss": 2.2041876220703127,
|
| 1751 |
+
"step": 6050
|
| 1752 |
+
},
|
| 1753 |
+
{
|
| 1754 |
+
"epoch": 5.8469682386910495,
|
| 1755 |
+
"grad_norm": 5.4792351722717285,
|
| 1756 |
+
"learning_rate": 7.765240206052911e-05,
|
| 1757 |
+
"loss": 2.216496887207031,
|
| 1758 |
+
"step": 6075
|
| 1759 |
+
},
|
| 1760 |
+
{
|
| 1761 |
+
"epoch": 5.871029836381136,
|
| 1762 |
+
"grad_norm": 6.080085277557373,
|
| 1763 |
+
"learning_rate": 7.68934783309408e-05,
|
| 1764 |
+
"loss": 2.21278076171875,
|
| 1765 |
+
"step": 6100
|
| 1766 |
+
},
|
| 1767 |
+
{
|
| 1768 |
+
"epoch": 5.895091434071222,
|
| 1769 |
+
"grad_norm": 6.729173183441162,
|
| 1770 |
+
"learning_rate": 7.613595794305359e-05,
|
| 1771 |
+
"loss": 2.231321105957031,
|
| 1772 |
+
"step": 6125
|
| 1773 |
+
},
|
| 1774 |
+
{
|
| 1775 |
+
"epoch": 5.919153031761309,
|
| 1776 |
+
"grad_norm": 7.153930187225342,
|
| 1777 |
+
"learning_rate": 7.537988690378994e-05,
|
| 1778 |
+
"loss": 2.2208010864257814,
|
| 1779 |
+
"step": 6150
|
| 1780 |
+
},
|
| 1781 |
+
{
|
| 1782 |
+
"epoch": 5.943214629451395,
|
| 1783 |
+
"grad_norm": 6.1341023445129395,
|
| 1784 |
+
"learning_rate": 7.462531113204809e-05,
|
| 1785 |
+
"loss": 2.2468557739257813,
|
| 1786 |
+
"step": 6175
|
| 1787 |
+
},
|
| 1788 |
+
{
|
| 1789 |
+
"epoch": 5.9672762271414825,
|
| 1790 |
+
"grad_norm": 6.451995849609375,
|
| 1791 |
+
"learning_rate": 7.387227645591342e-05,
|
| 1792 |
+
"loss": 2.2257095336914063,
|
| 1793 |
+
"step": 6200
|
| 1794 |
+
},
|
| 1795 |
+
{
|
| 1796 |
+
"epoch": 5.991337824831569,
|
| 1797 |
+
"grad_norm": 5.904285907745361,
|
| 1798 |
+
"learning_rate": 7.312082860987501e-05,
|
| 1799 |
+
"loss": 2.2414822387695312,
|
| 1800 |
+
"step": 6225
|
| 1801 |
+
},
|
| 1802 |
+
{
|
| 1803 |
+
"epoch": 6.0,
|
| 1804 |
+
"eval_loss": 2.538609504699707,
|
| 1805 |
+
"eval_runtime": 10.5504,
|
| 1806 |
+
"eval_samples_per_second": 142.174,
|
| 1807 |
+
"eval_steps_per_second": 2.275,
|
| 1808 |
+
"step": 6234
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 6.015399422521655,
|
| 1812 |
+
"grad_norm": 6.037642478942871,
|
| 1813 |
+
"learning_rate": 7.23710132320481e-05,
|
| 1814 |
+
"loss": 2.1239503479003905,
|
| 1815 |
+
"step": 6250
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 6.039461020211742,
|
| 1819 |
+
"grad_norm": 6.592258453369141,
|
| 1820 |
+
"learning_rate": 7.162287586140229e-05,
|
| 1821 |
+
"loss": 2.0652923583984375,
|
| 1822 |
+
"step": 6275
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 6.063522617901829,
|
| 1826 |
+
"grad_norm": 6.6665167808532715,
|
| 1827 |
+
"learning_rate": 7.087646193499579e-05,
|
| 1828 |
+
"loss": 2.0848187255859374,
|
| 1829 |
+
"step": 6300
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 6.0875842155919155,
|
| 1833 |
+
"grad_norm": 6.209718227386475,
|
| 1834 |
+
"learning_rate": 7.013181678521592e-05,
|
| 1835 |
+
"loss": 2.0491201782226565,
|
| 1836 |
+
"step": 6325
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 6.111645813282002,
|
| 1840 |
+
"grad_norm": 7.076566696166992,
|
| 1841 |
+
"learning_rate": 6.938898563702581e-05,
|
| 1842 |
+
"loss": 2.071932678222656,
|
| 1843 |
+
"step": 6350
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 6.135707410972088,
|
| 1847 |
+
"grad_norm": 6.92293643951416,
|
| 1848 |
+
"learning_rate": 6.86480136052178e-05,
|
| 1849 |
+
"loss": 2.111164855957031,
|
| 1850 |
+
"step": 6375
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 6.159769008662175,
|
| 1854 |
+
"grad_norm": 6.941470146179199,
|
| 1855 |
+
"learning_rate": 6.79089456916735e-05,
|
| 1856 |
+
"loss": 2.08375732421875,
|
| 1857 |
+
"step": 6400
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 6.183830606352262,
|
| 1861 |
+
"grad_norm": 6.843981742858887,
|
| 1862 |
+
"learning_rate": 6.717182678263056e-05,
|
| 1863 |
+
"loss": 2.0809051513671877,
|
| 1864 |
+
"step": 6425
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 6.2078922040423485,
|
| 1868 |
+
"grad_norm": 7.375129222869873,
|
| 1869 |
+
"learning_rate": 6.643670164595659e-05,
|
| 1870 |
+
"loss": 2.0675634765625,
|
| 1871 |
+
"step": 6450
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 6.231953801732435,
|
| 1875 |
+
"grad_norm": 6.744461536407471,
|
| 1876 |
+
"learning_rate": 6.570361492843034e-05,
|
| 1877 |
+
"loss": 2.1084820556640627,
|
| 1878 |
+
"step": 6475
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 6.256015399422521,
|
| 1882 |
+
"grad_norm": 6.756580829620361,
|
| 1883 |
+
"learning_rate": 6.497261115303e-05,
|
| 1884 |
+
"loss": 2.082350769042969,
|
| 1885 |
+
"step": 6500
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 6.280076997112609,
|
| 1889 |
+
"grad_norm": 6.490564346313477,
|
| 1890 |
+
"learning_rate": 6.42437347162293e-05,
|
| 1891 |
+
"loss": 2.0561422729492187,
|
| 1892 |
+
"step": 6525
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 6.304138594802695,
|
| 1896 |
+
"grad_norm": 6.300208568572998,
|
| 1897 |
+
"learning_rate": 6.351702988530107e-05,
|
| 1898 |
+
"loss": 2.1108993530273437,
|
| 1899 |
+
"step": 6550
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 6.3282001924927815,
|
| 1903 |
+
"grad_norm": 6.731162071228027,
|
| 1904 |
+
"learning_rate": 6.279254079562873e-05,
|
| 1905 |
+
"loss": 2.0939395141601564,
|
| 1906 |
+
"step": 6575
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 6.352261790182868,
|
| 1910 |
+
"grad_norm": 6.744671821594238,
|
| 1911 |
+
"learning_rate": 6.207031144802584e-05,
|
| 1912 |
+
"loss": 2.1048483276367187,
|
| 1913 |
+
"step": 6600
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 6.376323387872954,
|
| 1917 |
+
"grad_norm": 6.09976863861084,
|
| 1918 |
+
"learning_rate": 6.135038570606369e-05,
|
| 1919 |
+
"loss": 2.0879864501953125,
|
| 1920 |
+
"step": 6625
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 6.400384985563042,
|
| 1924 |
+
"grad_norm": 6.7319865226745605,
|
| 1925 |
+
"learning_rate": 6.063280729340739e-05,
|
| 1926 |
+
"loss": 2.0968563842773436,
|
| 1927 |
+
"step": 6650
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 6.424446583253128,
|
| 1931 |
+
"grad_norm": 7.194849014282227,
|
| 1932 |
+
"learning_rate": 5.991761979116039e-05,
|
| 1933 |
+
"loss": 2.0894454956054687,
|
| 1934 |
+
"step": 6675
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 6.4485081809432145,
|
| 1938 |
+
"grad_norm": 7.315184593200684,
|
| 1939 |
+
"learning_rate": 5.920486663521754e-05,
|
| 1940 |
+
"loss": 2.0605679321289063,
|
| 1941 |
+
"step": 6700
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 6.472569778633301,
|
| 1945 |
+
"grad_norm": 7.200357913970947,
|
| 1946 |
+
"learning_rate": 5.849459111362716e-05,
|
| 1947 |
+
"loss": 2.13160888671875,
|
| 1948 |
+
"step": 6725
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 6.496631376323388,
|
| 1952 |
+
"grad_norm": 6.48388147354126,
|
| 1953 |
+
"learning_rate": 5.7786836363961925e-05,
|
| 1954 |
+
"loss": 2.06083984375,
|
| 1955 |
+
"step": 6750
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 6.520692974013475,
|
| 1959 |
+
"grad_norm": 6.48615026473999,
|
| 1960 |
+
"learning_rate": 5.708164537069911e-05,
|
| 1961 |
+
"loss": 2.1018710327148438,
|
| 1962 |
+
"step": 6775
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 6.544754571703561,
|
| 1966 |
+
"grad_norm": 6.623106479644775,
|
| 1967 |
+
"learning_rate": 5.637906096260983e-05,
|
| 1968 |
+
"loss": 2.108205108642578,
|
| 1969 |
+
"step": 6800
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 6.5688161693936475,
|
| 1973 |
+
"grad_norm": 6.922781467437744,
|
| 1974 |
+
"learning_rate": 5.567912581015796e-05,
|
| 1975 |
+
"loss": 2.122475891113281,
|
| 1976 |
+
"step": 6825
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 6.592877767083735,
|
| 1980 |
+
"grad_norm": 7.1996989250183105,
|
| 1981 |
+
"learning_rate": 5.4981882422908616e-05,
|
| 1982 |
+
"loss": 2.1007341003417968,
|
| 1983 |
+
"step": 6850
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 6.616939364773821,
|
| 1987 |
+
"grad_norm": 7.156136512756348,
|
| 1988 |
+
"learning_rate": 5.4287373146946383e-05,
|
| 1989 |
+
"loss": 2.059091339111328,
|
| 1990 |
+
"step": 6875
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 6.641000962463908,
|
| 1994 |
+
"grad_norm": 6.269294738769531,
|
| 1995 |
+
"learning_rate": 5.3595640162303515e-05,
|
| 1996 |
+
"loss": 2.1031985473632813,
|
| 1997 |
+
"step": 6900
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 6.665062560153994,
|
| 2001 |
+
"grad_norm": 6.408496856689453,
|
| 2002 |
+
"learning_rate": 5.2906725480398125e-05,
|
| 2003 |
+
"loss": 2.083514862060547,
|
| 2004 |
+
"step": 6925
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 6.6891241578440805,
|
| 2008 |
+
"grad_norm": 6.383841037750244,
|
| 2009 |
+
"learning_rate": 5.222067094148273e-05,
|
| 2010 |
+
"loss": 2.084517822265625,
|
| 2011 |
+
"step": 6950
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 6.713185755534168,
|
| 2015 |
+
"grad_norm": 6.838091850280762,
|
| 2016 |
+
"learning_rate": 5.1537518212103116e-05,
|
| 2017 |
+
"loss": 2.107218933105469,
|
| 2018 |
+
"step": 6975
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 6.737247353224254,
|
| 2022 |
+
"grad_norm": 5.92156982421875,
|
| 2023 |
+
"learning_rate": 5.085730878256791e-05,
|
| 2024 |
+
"loss": 2.0721734619140624,
|
| 2025 |
+
"step": 7000
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 6.761308950914341,
|
| 2029 |
+
"grad_norm": 6.9503278732299805,
|
| 2030 |
+
"learning_rate": 5.018008396442847e-05,
|
| 2031 |
+
"loss": 2.1143641662597656,
|
| 2032 |
+
"step": 7025
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 6.785370548604427,
|
| 2036 |
+
"grad_norm": 6.930166244506836,
|
| 2037 |
+
"learning_rate": 4.95058848879701e-05,
|
| 2038 |
+
"loss": 2.0891964721679686,
|
| 2039 |
+
"step": 7050
|
| 2040 |
+
},
|
| 2041 |
+
{
|
| 2042 |
+
"epoch": 6.8094321462945135,
|
| 2043 |
+
"grad_norm": 6.717877388000488,
|
| 2044 |
+
"learning_rate": 4.883475249971396e-05,
|
| 2045 |
+
"loss": 2.045329895019531,
|
| 2046 |
+
"step": 7075
|
| 2047 |
+
},
|
| 2048 |
+
{
|
| 2049 |
+
"epoch": 6.833493743984601,
|
| 2050 |
+
"grad_norm": 6.861194610595703,
|
| 2051 |
+
"learning_rate": 4.816672755993031e-05,
|
| 2052 |
+
"loss": 2.121363220214844,
|
| 2053 |
+
"step": 7100
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 6.857555341674687,
|
| 2057 |
+
"grad_norm": 6.723923683166504,
|
| 2058 |
+
"learning_rate": 4.7501850640162914e-05,
|
| 2059 |
+
"loss": 2.09227294921875,
|
| 2060 |
+
"step": 7125
|
| 2061 |
+
},
|
| 2062 |
+
{
|
| 2063 |
+
"epoch": 6.881616939364774,
|
| 2064 |
+
"grad_norm": 6.983868598937988,
|
| 2065 |
+
"learning_rate": 4.6840162120765026e-05,
|
| 2066 |
+
"loss": 2.0283517456054687,
|
| 2067 |
+
"step": 7150
|
| 2068 |
+
},
|
| 2069 |
+
{
|
| 2070 |
+
"epoch": 6.90567853705486,
|
| 2071 |
+
"grad_norm": 6.482507705688477,
|
| 2072 |
+
"learning_rate": 4.618170218844692e-05,
|
| 2073 |
+
"loss": 2.121909942626953,
|
| 2074 |
+
"step": 7175
|
| 2075 |
+
},
|
| 2076 |
+
{
|
| 2077 |
+
"epoch": 6.929740134744947,
|
| 2078 |
+
"grad_norm": 6.73229455947876,
|
| 2079 |
+
"learning_rate": 4.552651083383531e-05,
|
| 2080 |
+
"loss": 2.109517364501953,
|
| 2081 |
+
"step": 7200
|
| 2082 |
+
},
|
| 2083 |
+
{
|
| 2084 |
+
"epoch": 6.953801732435034,
|
| 2085 |
+
"grad_norm": 6.240851879119873,
|
| 2086 |
+
"learning_rate": 4.487462784904444e-05,
|
| 2087 |
+
"loss": 2.078155212402344,
|
| 2088 |
+
"step": 7225
|
| 2089 |
+
},
|
| 2090 |
+
{
|
| 2091 |
+
"epoch": 6.97786333012512,
|
| 2092 |
+
"grad_norm": 6.120413303375244,
|
| 2093 |
+
"learning_rate": 4.4226092825259415e-05,
|
| 2094 |
+
"loss": 2.056332550048828,
|
| 2095 |
+
"step": 7250
|
| 2096 |
+
},
|
| 2097 |
+
{
|
| 2098 |
+
"epoch": 7.0,
|
| 2099 |
+
"eval_loss": 2.507423162460327,
|
| 2100 |
+
"eval_runtime": 10.3065,
|
| 2101 |
+
"eval_samples_per_second": 145.539,
|
| 2102 |
+
"eval_steps_per_second": 2.329,
|
| 2103 |
+
"step": 7273
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 7.001924927815207,
|
| 2107 |
+
"grad_norm": 6.26215934753418,
|
| 2108 |
+
"learning_rate": 4.358094515033169e-05,
|
| 2109 |
+
"loss": 2.1014112854003906,
|
| 2110 |
+
"step": 7275
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 7.025986525505293,
|
| 2114 |
+
"grad_norm": 6.870030403137207,
|
| 2115 |
+
"learning_rate": 4.29392240063869e-05,
|
| 2116 |
+
"loss": 1.9020887756347655,
|
| 2117 |
+
"step": 7300
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 7.05004812319538,
|
| 2121 |
+
"grad_norm": 6.9596638679504395,
|
| 2122 |
+
"learning_rate": 4.2300968367445304e-05,
|
| 2123 |
+
"loss": 1.9513621520996094,
|
| 2124 |
+
"step": 7325
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 7.074109720885467,
|
| 2128 |
+
"grad_norm": 7.853752613067627,
|
| 2129 |
+
"learning_rate": 4.166621699705453e-05,
|
| 2130 |
+
"loss": 1.9907861328125,
|
| 2131 |
+
"step": 7350
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 7.098171318575553,
|
| 2135 |
+
"grad_norm": 7.6610822677612305,
|
| 2136 |
+
"learning_rate": 4.103500844593546e-05,
|
| 2137 |
+
"loss": 1.9352001953125,
|
| 2138 |
+
"step": 7375
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 7.12223291626564,
|
| 2142 |
+
"grad_norm": 6.845297336578369,
|
| 2143 |
+
"learning_rate": 4.040738104964094e-05,
|
| 2144 |
+
"loss": 1.9567535400390625,
|
| 2145 |
+
"step": 7400
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 7.146294513955727,
|
| 2149 |
+
"grad_norm": 7.249809741973877,
|
| 2150 |
+
"learning_rate": 3.9783372926227424e-05,
|
| 2151 |
+
"loss": 1.9297769165039063,
|
| 2152 |
+
"step": 7425
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 7.170356111645813,
|
| 2156 |
+
"grad_norm": 6.793155193328857,
|
| 2157 |
+
"learning_rate": 3.916302197394002e-05,
|
| 2158 |
+
"loss": 1.9665078735351562,
|
| 2159 |
+
"step": 7450
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 7.1944177093359,
|
| 2163 |
+
"grad_norm": 6.96444845199585,
|
| 2164 |
+
"learning_rate": 3.854636586891073e-05,
|
| 2165 |
+
"loss": 1.9770477294921875,
|
| 2166 |
+
"step": 7475
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 7.218479307025986,
|
| 2170 |
+
"grad_norm": 6.901854991912842,
|
| 2171 |
+
"learning_rate": 3.7933442062870275e-05,
|
| 2172 |
+
"loss": 1.9699441528320312,
|
| 2173 |
+
"step": 7500
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 7.242540904716073,
|
| 2177 |
+
"grad_norm": 7.236461162567139,
|
| 2178 |
+
"learning_rate": 3.73242877808735e-05,
|
| 2179 |
+
"loss": 1.9833169555664063,
|
| 2180 |
+
"step": 7525
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 7.26660250240616,
|
| 2184 |
+
"grad_norm": 7.724422454833984,
|
| 2185 |
+
"learning_rate": 3.6718940019038605e-05,
|
| 2186 |
+
"loss": 1.9778964233398437,
|
| 2187 |
+
"step": 7550
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 7.290664100096246,
|
| 2191 |
+
"grad_norm": 8.021563529968262,
|
| 2192 |
+
"learning_rate": 3.611743554230013e-05,
|
| 2193 |
+
"loss": 1.9492559814453125,
|
| 2194 |
+
"step": 7575
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 7.314725697786333,
|
| 2198 |
+
"grad_norm": 6.951311111450195,
|
| 2199 |
+
"learning_rate": 3.551981088217628e-05,
|
| 2200 |
+
"loss": 1.9887191772460937,
|
| 2201 |
+
"step": 7600
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 7.338787295476419,
|
| 2205 |
+
"grad_norm": 7.245166301727295,
|
| 2206 |
+
"learning_rate": 3.492610233455006e-05,
|
| 2207 |
+
"loss": 1.905584259033203,
|
| 2208 |
+
"step": 7625
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 7.362848893166507,
|
| 2212 |
+
"grad_norm": 7.547442436218262,
|
| 2213 |
+
"learning_rate": 3.433634595746493e-05,
|
| 2214 |
+
"loss": 1.9690737915039063,
|
| 2215 |
+
"step": 7650
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 7.386910490856593,
|
| 2219 |
+
"grad_norm": 7.36473274230957,
|
| 2220 |
+
"learning_rate": 3.375057756893494e-05,
|
| 2221 |
+
"loss": 2.005036163330078,
|
| 2222 |
+
"step": 7675
|
| 2223 |
+
},
|
| 2224 |
+
{
|
| 2225 |
+
"epoch": 7.410972088546679,
|
| 2226 |
+
"grad_norm": 7.059237480163574,
|
| 2227 |
+
"learning_rate": 3.31688327447693e-05,
|
| 2228 |
+
"loss": 1.9534266662597657,
|
| 2229 |
+
"step": 7700
|
| 2230 |
+
},
|
| 2231 |
+
{
|
| 2232 |
+
"epoch": 7.435033686236766,
|
| 2233 |
+
"grad_norm": 6.996562957763672,
|
| 2234 |
+
"learning_rate": 3.25911468164118e-05,
|
| 2235 |
+
"loss": 1.967351837158203,
|
| 2236 |
+
"step": 7725
|
| 2237 |
+
},
|
| 2238 |
+
{
|
| 2239 |
+
"epoch": 7.459095283926853,
|
| 2240 |
+
"grad_norm": 7.537076950073242,
|
| 2241 |
+
"learning_rate": 3.201755486879494e-05,
|
| 2242 |
+
"loss": 1.9477909851074218,
|
| 2243 |
+
"step": 7750
|
| 2244 |
+
},
|
| 2245 |
+
{
|
| 2246 |
+
"epoch": 7.48315688161694,
|
| 2247 |
+
"grad_norm": 6.252792835235596,
|
| 2248 |
+
"learning_rate": 3.144809173820916e-05,
|
| 2249 |
+
"loss": 1.9597027587890625,
|
| 2250 |
+
"step": 7775
|
| 2251 |
+
},
|
| 2252 |
+
{
|
| 2253 |
+
"epoch": 7.507218479307026,
|
| 2254 |
+
"grad_norm": 7.163032531738281,
|
| 2255 |
+
"learning_rate": 3.0882792010187055e-05,
|
| 2256 |
+
"loss": 1.9337490844726561,
|
| 2257 |
+
"step": 7800
|
| 2258 |
+
},
|
| 2259 |
+
{
|
| 2260 |
+
"epoch": 7.531280076997112,
|
| 2261 |
+
"grad_norm": 6.879805564880371,
|
| 2262 |
+
"learning_rate": 3.0321690017402848e-05,
|
| 2263 |
+
"loss": 1.960424041748047,
|
| 2264 |
+
"step": 7825
|
| 2265 |
+
},
|
| 2266 |
+
{
|
| 2267 |
+
"epoch": 7.555341674687199,
|
| 2268 |
+
"grad_norm": 7.573090553283691,
|
| 2269 |
+
"learning_rate": 2.976481983758741e-05,
|
| 2270 |
+
"loss": 1.9841940307617187,
|
| 2271 |
+
"step": 7850
|
| 2272 |
+
},
|
| 2273 |
+
{
|
| 2274 |
+
"epoch": 7.579403272377286,
|
| 2275 |
+
"grad_norm": 7.028075218200684,
|
| 2276 |
+
"learning_rate": 2.9212215291458368e-05,
|
| 2277 |
+
"loss": 1.9862918090820312,
|
| 2278 |
+
"step": 7875
|
| 2279 |
+
},
|
| 2280 |
+
{
|
| 2281 |
+
"epoch": 7.603464870067373,
|
| 2282 |
+
"grad_norm": 6.813868999481201,
|
| 2283 |
+
"learning_rate": 2.8663909940666178e-05,
|
| 2284 |
+
"loss": 1.9782305908203126,
|
| 2285 |
+
"step": 7900
|
| 2286 |
+
},
|
| 2287 |
+
{
|
| 2288 |
+
"epoch": 7.627526467757459,
|
| 2289 |
+
"grad_norm": 7.10428524017334,
|
| 2290 |
+
"learning_rate": 2.811993708575581e-05,
|
| 2291 |
+
"loss": 1.9762313842773438,
|
| 2292 |
+
"step": 7925
|
| 2293 |
+
},
|
| 2294 |
+
{
|
| 2295 |
+
"epoch": 7.651588065447545,
|
| 2296 |
+
"grad_norm": 7.511239528656006,
|
| 2297 |
+
"learning_rate": 2.7580329764144265e-05,
|
| 2298 |
+
"loss": 1.972908477783203,
|
| 2299 |
+
"step": 7950
|
| 2300 |
+
},
|
| 2301 |
+
{
|
| 2302 |
+
"epoch": 7.675649663137632,
|
| 2303 |
+
"grad_norm": 7.587350845336914,
|
| 2304 |
+
"learning_rate": 2.704512074811405e-05,
|
| 2305 |
+
"loss": 2.0077505493164063,
|
| 2306 |
+
"step": 7975
|
| 2307 |
+
},
|
| 2308 |
+
{
|
| 2309 |
+
"epoch": 7.699711260827719,
|
| 2310 |
+
"grad_norm": 7.37968111038208,
|
| 2311 |
+
"learning_rate": 2.6514342542822914e-05,
|
| 2312 |
+
"loss": 1.9805903625488281,
|
| 2313 |
+
"step": 8000
|
| 2314 |
+
},
|
| 2315 |
+
{
|
| 2316 |
+
"epoch": 7.723772858517806,
|
| 2317 |
+
"grad_norm": 7.895181179046631,
|
| 2318 |
+
"learning_rate": 2.598802738432956e-05,
|
| 2319 |
+
"loss": 1.9543246459960937,
|
| 2320 |
+
"step": 8025
|
| 2321 |
+
},
|
| 2322 |
+
{
|
| 2323 |
+
"epoch": 7.747834456207892,
|
| 2324 |
+
"grad_norm": 7.283754825592041,
|
| 2325 |
+
"learning_rate": 2.546620723763593e-05,
|
| 2326 |
+
"loss": 1.974208984375,
|
| 2327 |
+
"step": 8050
|
| 2328 |
+
},
|
| 2329 |
+
{
|
| 2330 |
+
"epoch": 7.771896053897978,
|
| 2331 |
+
"grad_norm": 6.7545647621154785,
|
| 2332 |
+
"learning_rate": 2.494891379474581e-05,
|
| 2333 |
+
"loss": 1.9493072509765625,
|
| 2334 |
+
"step": 8075
|
| 2335 |
+
},
|
| 2336 |
+
{
|
| 2337 |
+
"epoch": 7.795957651588066,
|
| 2338 |
+
"grad_norm": 6.595056533813477,
|
| 2339 |
+
"learning_rate": 2.4436178472740057e-05,
|
| 2340 |
+
"loss": 1.9800338745117188,
|
| 2341 |
+
"step": 8100
|
| 2342 |
+
},
|
| 2343 |
+
{
|
| 2344 |
+
"epoch": 7.820019249278152,
|
| 2345 |
+
"grad_norm": 7.348852634429932,
|
| 2346 |
+
"learning_rate": 2.3928032411868618e-05,
|
| 2347 |
+
"loss": 1.9982472229003907,
|
| 2348 |
+
"step": 8125
|
| 2349 |
+
},
|
| 2350 |
+
{
|
| 2351 |
+
"epoch": 7.844080846968239,
|
| 2352 |
+
"grad_norm": 7.632541179656982,
|
| 2353 |
+
"learning_rate": 2.3424506473659136e-05,
|
| 2354 |
+
"loss": 1.9603765869140626,
|
| 2355 |
+
"step": 8150
|
| 2356 |
+
},
|
| 2357 |
+
{
|
| 2358 |
+
"epoch": 7.868142444658325,
|
| 2359 |
+
"grad_norm": 6.745494842529297,
|
| 2360 |
+
"learning_rate": 2.2925631239042667e-05,
|
| 2361 |
+
"loss": 2.0219635009765624,
|
| 2362 |
+
"step": 8175
|
| 2363 |
+
},
|
| 2364 |
+
{
|
| 2365 |
+
"epoch": 7.892204042348412,
|
| 2366 |
+
"grad_norm": 7.200120449066162,
|
| 2367 |
+
"learning_rate": 2.2431437006496446e-05,
|
| 2368 |
+
"loss": 1.978095245361328,
|
| 2369 |
+
"step": 8200
|
| 2370 |
+
},
|
| 2371 |
+
{
|
| 2372 |
+
"epoch": 7.916265640038499,
|
| 2373 |
+
"grad_norm": 7.620790958404541,
|
| 2374 |
+
"learning_rate": 2.1941953790203683e-05,
|
| 2375 |
+
"loss": 1.9720132446289063,
|
| 2376 |
+
"step": 8225
|
| 2377 |
+
},
|
| 2378 |
+
{
|
| 2379 |
+
"epoch": 7.940327237728585,
|
| 2380 |
+
"grad_norm": 7.06905460357666,
|
| 2381 |
+
"learning_rate": 2.145721131823074e-05,
|
| 2382 |
+
"loss": 1.9642568969726562,
|
| 2383 |
+
"step": 8250
|
| 2384 |
+
},
|
| 2385 |
+
{
|
| 2386 |
+
"epoch": 7.964388835418672,
|
| 2387 |
+
"grad_norm": 7.65277624130249,
|
| 2388 |
+
"learning_rate": 2.0977239030721617e-05,
|
| 2389 |
+
"loss": 1.983006134033203,
|
| 2390 |
+
"step": 8275
|
| 2391 |
+
},
|
| 2392 |
+
{
|
| 2393 |
+
"epoch": 7.988450433108758,
|
| 2394 |
+
"grad_norm": 7.044776439666748,
|
| 2395 |
+
"learning_rate": 2.0502066078109948e-05,
|
| 2396 |
+
"loss": 1.965108642578125,
|
| 2397 |
+
"step": 8300
|
| 2398 |
+
},
|
| 2399 |
+
{
|
| 2400 |
+
"epoch": 8.0,
|
| 2401 |
+
"eval_loss": 2.5075199604034424,
|
| 2402 |
+
"eval_runtime": 11.2899,
|
| 2403 |
+
"eval_samples_per_second": 132.862,
|
| 2404 |
+
"eval_steps_per_second": 2.126,
|
| 2405 |
+
"step": 8312
|
| 2406 |
+
},
|
| 2407 |
+
{
|
| 2408 |
+
"epoch": 8.012512030798845,
|
| 2409 |
+
"grad_norm": 8.218935012817383,
|
| 2410 |
+
"learning_rate": 2.0031721319348596e-05,
|
| 2411 |
+
"loss": 1.9299472045898438,
|
| 2412 |
+
"step": 8325
|
| 2413 |
+
},
|
| 2414 |
+
{
|
| 2415 |
+
"epoch": 8.03657362848893,
|
| 2416 |
+
"grad_norm": 7.044556140899658,
|
| 2417 |
+
"learning_rate": 1.95662333201569e-05,
|
| 2418 |
+
"loss": 1.8376336669921876,
|
| 2419 |
+
"step": 8350
|
| 2420 |
+
},
|
| 2421 |
+
{
|
| 2422 |
+
"epoch": 8.060635226179018,
|
| 2423 |
+
"grad_norm": 7.477766990661621,
|
| 2424 |
+
"learning_rate": 1.910563035128593e-05,
|
| 2425 |
+
"loss": 1.8646517944335939,
|
| 2426 |
+
"step": 8375
|
| 2427 |
+
},
|
| 2428 |
+
{
|
| 2429 |
+
"epoch": 8.084696823869105,
|
| 2430 |
+
"grad_norm": 7.536664962768555,
|
| 2431 |
+
"learning_rate": 1.86499403868013e-05,
|
| 2432 |
+
"loss": 1.8631768798828126,
|
| 2433 |
+
"step": 8400
|
| 2434 |
+
},
|
| 2435 |
+
{
|
| 2436 |
+
"epoch": 8.108758421559191,
|
| 2437 |
+
"grad_norm": 6.956142425537109,
|
| 2438 |
+
"learning_rate": 1.8199191102384317e-05,
|
| 2439 |
+
"loss": 1.8349604797363281,
|
| 2440 |
+
"step": 8425
|
| 2441 |
+
},
|
| 2442 |
+
{
|
| 2443 |
+
"epoch": 8.132820019249278,
|
| 2444 |
+
"grad_norm": 7.35983419418335,
|
| 2445 |
+
"learning_rate": 1.7753409873651117e-05,
|
| 2446 |
+
"loss": 1.9105496215820312,
|
| 2447 |
+
"step": 8450
|
| 2448 |
+
},
|
| 2449 |
+
{
|
| 2450 |
+
"epoch": 8.156881616939366,
|
| 2451 |
+
"grad_norm": 7.185935974121094,
|
| 2452 |
+
"learning_rate": 1.7312623774490054e-05,
|
| 2453 |
+
"loss": 1.924974365234375,
|
| 2454 |
+
"step": 8475
|
| 2455 |
+
},
|
| 2456 |
+
{
|
| 2457 |
+
"epoch": 8.180943214629451,
|
| 2458 |
+
"grad_norm": 7.5399885177612305,
|
| 2459 |
+
"learning_rate": 1.687685957541738e-05,
|
| 2460 |
+
"loss": 1.8878890991210937,
|
| 2461 |
+
"step": 8500
|
| 2462 |
+
},
|
| 2463 |
+
{
|
| 2464 |
+
"epoch": 8.205004812319538,
|
| 2465 |
+
"grad_norm": 7.930432319641113,
|
| 2466 |
+
"learning_rate": 1.644614374195146e-05,
|
| 2467 |
+
"loss": 1.87248291015625,
|
| 2468 |
+
"step": 8525
|
| 2469 |
+
},
|
| 2470 |
+
{
|
| 2471 |
+
"epoch": 8.229066410009624,
|
| 2472 |
+
"grad_norm": 6.113994598388672,
|
| 2473 |
+
"learning_rate": 1.6020502433005268e-05,
|
| 2474 |
+
"loss": 1.8437289428710937,
|
| 2475 |
+
"step": 8550
|
| 2476 |
+
},
|
| 2477 |
+
{
|
| 2478 |
+
"epoch": 8.253128007699711,
|
| 2479 |
+
"grad_norm": 6.764047145843506,
|
| 2480 |
+
"learning_rate": 1.55999614992978e-05,
|
| 2481 |
+
"loss": 1.8841342163085937,
|
| 2482 |
+
"step": 8575
|
| 2483 |
+
},
|
| 2484 |
+
{
|
| 2485 |
+
"epoch": 8.277189605389799,
|
| 2486 |
+
"grad_norm": 7.982216835021973,
|
| 2487 |
+
"learning_rate": 1.5184546481784034e-05,
|
| 2488 |
+
"loss": 1.9175787353515625,
|
| 2489 |
+
"step": 8600
|
| 2490 |
+
},
|
| 2491 |
+
{
|
| 2492 |
+
"epoch": 8.301251203079884,
|
| 2493 |
+
"grad_norm": 7.563563346862793,
|
| 2494 |
+
"learning_rate": 1.4774282610103696e-05,
|
| 2495 |
+
"loss": 1.8796307373046874,
|
| 2496 |
+
"step": 8625
|
| 2497 |
+
},
|
| 2498 |
+
{
|
| 2499 |
+
"epoch": 8.325312800769971,
|
| 2500 |
+
"grad_norm": 7.938658714294434,
|
| 2501 |
+
"learning_rate": 1.4369194801049012e-05,
|
| 2502 |
+
"loss": 1.9124429321289063,
|
| 2503 |
+
"step": 8650
|
| 2504 |
+
},
|
| 2505 |
+
{
|
| 2506 |
+
"epoch": 8.349374398460057,
|
| 2507 |
+
"grad_norm": 7.024694442749023,
|
| 2508 |
+
"learning_rate": 1.3969307657051412e-05,
|
| 2509 |
+
"loss": 1.8650804138183594,
|
| 2510 |
+
"step": 8675
|
| 2511 |
+
},
|
| 2512 |
+
{
|
| 2513 |
+
"epoch": 8.373435996150144,
|
| 2514 |
+
"grad_norm": 7.2648701667785645,
|
| 2515 |
+
"learning_rate": 1.3574645464687364e-05,
|
| 2516 |
+
"loss": 1.8613601684570313,
|
| 2517 |
+
"step": 8700
|
| 2518 |
+
},
|
| 2519 |
+
{
|
| 2520 |
+
"epoch": 8.397497593840232,
|
| 2521 |
+
"grad_norm": 7.496577262878418,
|
| 2522 |
+
"learning_rate": 1.3185232193203312e-05,
|
| 2523 |
+
"loss": 1.8711148071289063,
|
| 2524 |
+
"step": 8725
|
| 2525 |
+
},
|
| 2526 |
+
{
|
| 2527 |
+
"epoch": 8.421559191530317,
|
| 2528 |
+
"grad_norm": 7.19474458694458,
|
| 2529 |
+
"learning_rate": 1.2801091493059958e-05,
|
| 2530 |
+
"loss": 1.8552403259277344,
|
| 2531 |
+
"step": 8750
|
| 2532 |
+
},
|
| 2533 |
+
{
|
| 2534 |
+
"epoch": 8.445620789220404,
|
| 2535 |
+
"grad_norm": 7.82267951965332,
|
| 2536 |
+
"learning_rate": 1.2422246694495898e-05,
|
| 2537 |
+
"loss": 1.9189707946777343,
|
| 2538 |
+
"step": 8775
|
| 2539 |
+
},
|
| 2540 |
+
{
|
| 2541 |
+
"epoch": 8.46968238691049,
|
| 2542 |
+
"grad_norm": 8.27257251739502,
|
| 2543 |
+
"learning_rate": 1.204872080611067e-05,
|
| 2544 |
+
"loss": 1.9112997436523438,
|
| 2545 |
+
"step": 8800
|
| 2546 |
+
},
|
| 2547 |
+
{
|
| 2548 |
+
"epoch": 8.493743984600577,
|
| 2549 |
+
"grad_norm": 7.660341262817383,
|
| 2550 |
+
"learning_rate": 1.1680536513467388e-05,
|
| 2551 |
+
"loss": 1.87326171875,
|
| 2552 |
+
"step": 8825
|
| 2553 |
+
},
|
| 2554 |
+
{
|
| 2555 |
+
"epoch": 8.517805582290665,
|
| 2556 |
+
"grad_norm": 6.914079666137695,
|
| 2557 |
+
"learning_rate": 1.1317716177714965e-05,
|
| 2558 |
+
"loss": 1.8993162536621093,
|
| 2559 |
+
"step": 8850
|
| 2560 |
+
},
|
| 2561 |
+
{
|
| 2562 |
+
"epoch": 8.54186717998075,
|
| 2563 |
+
"grad_norm": 8.300233840942383,
|
| 2564 |
+
"learning_rate": 1.0960281834230002e-05,
|
| 2565 |
+
"loss": 1.8621255493164062,
|
| 2566 |
+
"step": 8875
|
| 2567 |
+
},
|
| 2568 |
+
{
|
| 2569 |
+
"epoch": 8.565928777670837,
|
| 2570 |
+
"grad_norm": 6.867978096008301,
|
| 2571 |
+
"learning_rate": 1.0608255191278537e-05,
|
| 2572 |
+
"loss": 1.881025390625,
|
| 2573 |
+
"step": 8900
|
| 2574 |
+
},
|
| 2575 |
+
{
|
| 2576 |
+
"epoch": 8.589990375360923,
|
| 2577 |
+
"grad_norm": 7.648881912231445,
|
| 2578 |
+
"learning_rate": 1.026165762869762e-05,
|
| 2579 |
+
"loss": 1.8587013244628907,
|
| 2580 |
+
"step": 8925
|
| 2581 |
+
},
|
| 2582 |
+
{
|
| 2583 |
+
"epoch": 8.61405197305101,
|
| 2584 |
+
"grad_norm": 6.95702600479126,
|
| 2585 |
+
"learning_rate": 9.9205101965968e-06,
|
| 2586 |
+
"loss": 1.8720562744140625,
|
| 2587 |
+
"step": 8950
|
| 2588 |
+
},
|
| 2589 |
+
{
|
| 2590 |
+
"epoch": 8.638113570741098,
|
| 2591 |
+
"grad_norm": 7.878551483154297,
|
| 2592 |
+
"learning_rate": 9.58483361407978e-06,
|
| 2593 |
+
"loss": 1.9066632080078125,
|
| 2594 |
+
"step": 8975
|
| 2595 |
+
},
|
| 2596 |
+
{
|
| 2597 |
+
"epoch": 8.662175168431183,
|
| 2598 |
+
"grad_norm": 6.784065246582031,
|
| 2599 |
+
"learning_rate": 9.254648267985943e-06,
|
| 2600 |
+
"loss": 1.861563720703125,
|
| 2601 |
+
"step": 9000
|
| 2602 |
+
},
|
| 2603 |
+
{
|
| 2604 |
+
"epoch": 8.68623676612127,
|
| 2605 |
+
"grad_norm": 8.087384223937988,
|
| 2606 |
+
"learning_rate": 8.92997421165227e-06,
|
| 2607 |
+
"loss": 1.8624609375,
|
| 2608 |
+
"step": 9025
|
| 2609 |
+
},
|
| 2610 |
+
{
|
| 2611 |
+
"epoch": 8.710298363811358,
|
| 2612 |
+
"grad_norm": 7.116074562072754,
|
| 2613 |
+
"learning_rate": 8.610831163695365e-06,
|
| 2614 |
+
"loss": 1.8397271728515625,
|
| 2615 |
+
"step": 9050
|
| 2616 |
+
},
|
| 2617 |
+
{
|
| 2618 |
+
"epoch": 8.734359961501443,
|
| 2619 |
+
"grad_norm": 8.176411628723145,
|
| 2620 |
+
"learning_rate": 8.297238506813964e-06,
|
| 2621 |
+
"loss": 1.8658590698242188,
|
| 2622 |
+
"step": 9075
|
| 2623 |
+
},
|
| 2624 |
+
{
|
| 2625 |
+
"epoch": 8.75842155919153,
|
| 2626 |
+
"grad_norm": 7.450162887573242,
|
| 2627 |
+
"learning_rate": 7.989215286611684e-06,
|
| 2628 |
+
"loss": 1.909896240234375,
|
| 2629 |
+
"step": 9100
|
| 2630 |
+
},
|
| 2631 |
+
{
|
| 2632 |
+
"epoch": 8.782483156881616,
|
| 2633 |
+
"grad_norm": 7.687817573547363,
|
| 2634 |
+
"learning_rate": 7.686780210440314e-06,
|
| 2635 |
+
"loss": 1.9157940673828124,
|
| 2636 |
+
"step": 9125
|
| 2637 |
+
},
|
| 2638 |
+
{
|
| 2639 |
+
"epoch": 8.806544754571703,
|
| 2640 |
+
"grad_norm": 6.840658187866211,
|
| 2641 |
+
"learning_rate": 7.389951646263682e-06,
|
| 2642 |
+
"loss": 1.8945613098144531,
|
| 2643 |
+
"step": 9150
|
| 2644 |
+
},
|
| 2645 |
+
{
|
| 2646 |
+
"epoch": 8.83060635226179,
|
| 2647 |
+
"grad_norm": 7.51788854598999,
|
| 2648 |
+
"learning_rate": 7.098747621542101e-06,
|
| 2649 |
+
"loss": 1.8905087280273438,
|
| 2650 |
+
"step": 9175
|
| 2651 |
+
},
|
| 2652 |
+
{
|
| 2653 |
+
"epoch": 8.854667949951876,
|
| 2654 |
+
"grad_norm": 6.841259002685547,
|
| 2655 |
+
"learning_rate": 6.813185822137469e-06,
|
| 2656 |
+
"loss": 1.9026531982421875,
|
| 2657 |
+
"step": 9200
|
| 2658 |
+
},
|
| 2659 |
+
{
|
| 2660 |
+
"epoch": 8.878729547641964,
|
| 2661 |
+
"grad_norm": 7.25047492980957,
|
| 2662 |
+
"learning_rate": 6.533283591239203e-06,
|
| 2663 |
+
"loss": 1.8813078308105469,
|
| 2664 |
+
"step": 9225
|
| 2665 |
+
},
|
| 2666 |
+
{
|
| 2667 |
+
"epoch": 8.90279114533205,
|
| 2668 |
+
"grad_norm": 7.095242977142334,
|
| 2669 |
+
"learning_rate": 6.2590579283108185e-06,
|
| 2670 |
+
"loss": 1.8625860595703125,
|
| 2671 |
+
"step": 9250
|
| 2672 |
+
},
|
| 2673 |
+
{
|
| 2674 |
+
"epoch": 8.926852743022136,
|
| 2675 |
+
"grad_norm": 8.016623497009277,
|
| 2676 |
+
"learning_rate": 5.990525488057608e-06,
|
| 2677 |
+
"loss": 1.9108738708496094,
|
| 2678 |
+
"step": 9275
|
| 2679 |
+
},
|
| 2680 |
+
{
|
| 2681 |
+
"epoch": 8.950914340712224,
|
| 2682 |
+
"grad_norm": 7.1780548095703125,
|
| 2683 |
+
"learning_rate": 5.727702579415062e-06,
|
| 2684 |
+
"loss": 1.8983804321289062,
|
| 2685 |
+
"step": 9300
|
| 2686 |
+
},
|
| 2687 |
+
{
|
| 2688 |
+
"epoch": 8.97497593840231,
|
| 2689 |
+
"grad_norm": 7.045661926269531,
|
| 2690 |
+
"learning_rate": 5.470605164558407e-06,
|
| 2691 |
+
"loss": 1.8712222290039062,
|
| 2692 |
+
"step": 9325
|
| 2693 |
+
},
|
| 2694 |
+
{
|
| 2695 |
+
"epoch": 8.999037536092397,
|
| 2696 |
+
"grad_norm": 7.895360469818115,
|
| 2697 |
+
"learning_rate": 5.219248857933156e-06,
|
| 2698 |
+
"loss": 1.8945994567871094,
|
| 2699 |
+
"step": 9350
|
| 2700 |
+
},
|
| 2701 |
+
{
|
| 2702 |
+
"epoch": 9.0,
|
| 2703 |
+
"eval_loss": 2.5214192867279053,
|
| 2704 |
+
"eval_runtime": 11.0957,
|
| 2705 |
+
"eval_samples_per_second": 135.187,
|
| 2706 |
+
"eval_steps_per_second": 2.163,
|
| 2707 |
+
"step": 9351
|
| 2708 |
+
}
|
| 2709 |
+
],
|
| 2710 |
+
"logging_steps": 25,
|
| 2711 |
+
"max_steps": 10390,
|
| 2712 |
+
"num_input_tokens_seen": 0,
|
| 2713 |
+
"num_train_epochs": 10,
|
| 2714 |
+
"save_steps": 500,
|
| 2715 |
+
"stateful_callbacks": {
|
| 2716 |
+
"TrainerControl": {
|
| 2717 |
+
"args": {
|
| 2718 |
+
"should_epoch_stop": false,
|
| 2719 |
+
"should_evaluate": false,
|
| 2720 |
+
"should_log": false,
|
| 2721 |
+
"should_save": true,
|
| 2722 |
+
"should_training_stop": false
|
| 2723 |
+
},
|
| 2724 |
+
"attributes": {}
|
| 2725 |
+
}
|
| 2726 |
+
},
|
| 2727 |
+
"total_flos": 0.0,
|
| 2728 |
+
"train_batch_size": 8,
|
| 2729 |
+
"trial_name": null,
|
| 2730 |
+
"trial_params": null
|
| 2731 |
+
}
|