Upload 10 files
Browse files- .gitattributes +1 -0
- model.safetensors +1 -1
- optimizer.pt +3 -0
- rng_state.pth +0 -0
- scheduler.pt +0 -0
- trainer_state.json +1497 -0
.gitattributes
CHANGED
|
@@ -1,3 +1,4 @@
|
|
| 1 |
model.safetensors filter=lfs diff=lfs merge=lfs -text
|
| 2 |
model.SRC filter=lfs diff=lfs merge=lfs -text
|
| 3 |
model.TGT filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 1 |
model.safetensors filter=lfs diff=lfs merge=lfs -text
|
| 2 |
model.SRC filter=lfs diff=lfs merge=lfs -text
|
| 3 |
model.TGT filter=lfs diff=lfs merge=lfs -text
|
| 4 |
+
optimizer.pt filter=lfs diff=lfs merge=lfs -text
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 2247492800
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b807377b9259c63916b600d8d5eb99c70dc4a2085059628e505000b277a9f84b
|
| 3 |
size 2247492800
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8171315046191596a4cd9a881659f2157b1d1a3b94517b2186ee617eba4b2515
|
| 3 |
+
size 4495445235
|
rng_state.pth
ADDED
|
Binary file (14.3 kB). View file
|
|
|
scheduler.pt
ADDED
|
Binary file (1.06 kB). View file
|
|
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1497 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 34.19371467731715,
|
| 3 |
+
"best_model_checkpoint": "indictrans-en-ne-checkpoint-1B-2/checkpoint-12000",
|
| 4 |
+
"epoch": 1.9999086966446016,
|
| 5 |
+
"eval_steps": 1200,
|
| 6 |
+
"global_step": 16428,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.012173780719774785,
|
| 13 |
+
"grad_norm": 0.1044921875,
|
| 14 |
+
"learning_rate": 0.0001414213562373095,
|
| 15 |
+
"loss": 0.1798,
|
| 16 |
+
"num_input_tokens_seen": 3276800,
|
| 17 |
+
"step": 100
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.02434756143954957,
|
| 21 |
+
"grad_norm": 0.0908203125,
|
| 22 |
+
"learning_rate": 0.0001,
|
| 23 |
+
"loss": 0.1787,
|
| 24 |
+
"num_input_tokens_seen": 6553600,
|
| 25 |
+
"step": 200
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.03652134215932436,
|
| 29 |
+
"grad_norm": 0.08349609375,
|
| 30 |
+
"learning_rate": 8.164965809277262e-05,
|
| 31 |
+
"loss": 0.1784,
|
| 32 |
+
"num_input_tokens_seen": 9830400,
|
| 33 |
+
"step": 300
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.04869512287909914,
|
| 37 |
+
"grad_norm": 0.08203125,
|
| 38 |
+
"learning_rate": 7.071067811865475e-05,
|
| 39 |
+
"loss": 0.1775,
|
| 40 |
+
"num_input_tokens_seen": 13107200,
|
| 41 |
+
"step": 400
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.060868903598873925,
|
| 45 |
+
"grad_norm": 0.08740234375,
|
| 46 |
+
"learning_rate": 6.324555320336759e-05,
|
| 47 |
+
"loss": 0.177,
|
| 48 |
+
"num_input_tokens_seen": 16384000,
|
| 49 |
+
"step": 500
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.07304268431864872,
|
| 53 |
+
"grad_norm": 0.09228515625,
|
| 54 |
+
"learning_rate": 5.7735026918962585e-05,
|
| 55 |
+
"loss": 0.1757,
|
| 56 |
+
"num_input_tokens_seen": 19660800,
|
| 57 |
+
"step": 600
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.08521646503842349,
|
| 61 |
+
"grad_norm": 0.091796875,
|
| 62 |
+
"learning_rate": 5.3452248382484884e-05,
|
| 63 |
+
"loss": 0.1749,
|
| 64 |
+
"num_input_tokens_seen": 22937600,
|
| 65 |
+
"step": 700
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.09739024575819828,
|
| 69 |
+
"grad_norm": 0.08154296875,
|
| 70 |
+
"learning_rate": 5e-05,
|
| 71 |
+
"loss": 0.1745,
|
| 72 |
+
"num_input_tokens_seen": 26214400,
|
| 73 |
+
"step": 800
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.10956402647797306,
|
| 77 |
+
"grad_norm": 0.0908203125,
|
| 78 |
+
"learning_rate": 4.7140452079103176e-05,
|
| 79 |
+
"loss": 0.1728,
|
| 80 |
+
"num_input_tokens_seen": 29491200,
|
| 81 |
+
"step": 900
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.12173780719774785,
|
| 85 |
+
"grad_norm": 0.0791015625,
|
| 86 |
+
"learning_rate": 4.4721359549995795e-05,
|
| 87 |
+
"loss": 0.175,
|
| 88 |
+
"num_input_tokens_seen": 32768000,
|
| 89 |
+
"step": 1000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.13391158791752264,
|
| 93 |
+
"grad_norm": 0.09765625,
|
| 94 |
+
"learning_rate": 4.264014327112208e-05,
|
| 95 |
+
"loss": 0.1757,
|
| 96 |
+
"num_input_tokens_seen": 36044800,
|
| 97 |
+
"step": 1100
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.14608536863729743,
|
| 101 |
+
"grad_norm": 0.0830078125,
|
| 102 |
+
"learning_rate": 4.082482904638631e-05,
|
| 103 |
+
"loss": 0.1718,
|
| 104 |
+
"num_input_tokens_seen": 39321600,
|
| 105 |
+
"step": 1200
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.14608536863729743,
|
| 109 |
+
"eval_BLEU": 33.851984753329205,
|
| 110 |
+
"eval_chrF": 59.48761000963931,
|
| 111 |
+
"eval_loss": 0.1480654925107956,
|
| 112 |
+
"eval_runtime": 4021.48,
|
| 113 |
+
"eval_samples_per_second": 4.692,
|
| 114 |
+
"eval_steps_per_second": 0.293,
|
| 115 |
+
"num_input_tokens_seen": 39321600,
|
| 116 |
+
"step": 1200
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.1582591493570722,
|
| 120 |
+
"grad_norm": 0.08544921875,
|
| 121 |
+
"learning_rate": 3.922322702763681e-05,
|
| 122 |
+
"loss": 0.1719,
|
| 123 |
+
"num_input_tokens_seen": 42598400,
|
| 124 |
+
"step": 1300
|
| 125 |
+
},
|
| 126 |
+
{
|
| 127 |
+
"epoch": 0.17043293007684698,
|
| 128 |
+
"grad_norm": 0.0908203125,
|
| 129 |
+
"learning_rate": 3.779644730092272e-05,
|
| 130 |
+
"loss": 0.1718,
|
| 131 |
+
"num_input_tokens_seen": 45875200,
|
| 132 |
+
"step": 1400
|
| 133 |
+
},
|
| 134 |
+
{
|
| 135 |
+
"epoch": 0.18260671079662177,
|
| 136 |
+
"grad_norm": 0.08837890625,
|
| 137 |
+
"learning_rate": 3.651483716701107e-05,
|
| 138 |
+
"loss": 0.1685,
|
| 139 |
+
"num_input_tokens_seen": 49152000,
|
| 140 |
+
"step": 1500
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.19478049151639656,
|
| 144 |
+
"grad_norm": 0.09130859375,
|
| 145 |
+
"learning_rate": 3.535533905932738e-05,
|
| 146 |
+
"loss": 0.1742,
|
| 147 |
+
"num_input_tokens_seen": 52428800,
|
| 148 |
+
"step": 1600
|
| 149 |
+
},
|
| 150 |
+
{
|
| 151 |
+
"epoch": 0.20695427223617135,
|
| 152 |
+
"grad_norm": 0.09033203125,
|
| 153 |
+
"learning_rate": 3.4299717028501764e-05,
|
| 154 |
+
"loss": 0.1736,
|
| 155 |
+
"num_input_tokens_seen": 55705600,
|
| 156 |
+
"step": 1700
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.21912805295594612,
|
| 160 |
+
"grad_norm": 0.10498046875,
|
| 161 |
+
"learning_rate": 3.3333333333333335e-05,
|
| 162 |
+
"loss": 0.1714,
|
| 163 |
+
"num_input_tokens_seen": 58982400,
|
| 164 |
+
"step": 1800
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.2313018336757209,
|
| 168 |
+
"grad_norm": 0.08447265625,
|
| 169 |
+
"learning_rate": 3.244428422615251e-05,
|
| 170 |
+
"loss": 0.1718,
|
| 171 |
+
"num_input_tokens_seen": 62259200,
|
| 172 |
+
"step": 1900
|
| 173 |
+
},
|
| 174 |
+
{
|
| 175 |
+
"epoch": 0.2434756143954957,
|
| 176 |
+
"grad_norm": 0.08544921875,
|
| 177 |
+
"learning_rate": 3.1622776601683795e-05,
|
| 178 |
+
"loss": 0.1709,
|
| 179 |
+
"num_input_tokens_seen": 65536000,
|
| 180 |
+
"step": 2000
|
| 181 |
+
},
|
| 182 |
+
{
|
| 183 |
+
"epoch": 0.25564939511527046,
|
| 184 |
+
"grad_norm": 0.0908203125,
|
| 185 |
+
"learning_rate": 3.086066999241838e-05,
|
| 186 |
+
"loss": 0.175,
|
| 187 |
+
"num_input_tokens_seen": 68812800,
|
| 188 |
+
"step": 2100
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.2678231758350453,
|
| 192 |
+
"grad_norm": 0.09521484375,
|
| 193 |
+
"learning_rate": 3.0151134457776364e-05,
|
| 194 |
+
"loss": 0.1706,
|
| 195 |
+
"num_input_tokens_seen": 72089600,
|
| 196 |
+
"step": 2200
|
| 197 |
+
},
|
| 198 |
+
{
|
| 199 |
+
"epoch": 0.27999695655482004,
|
| 200 |
+
"grad_norm": 0.0966796875,
|
| 201 |
+
"learning_rate": 2.948839123097943e-05,
|
| 202 |
+
"loss": 0.1758,
|
| 203 |
+
"num_input_tokens_seen": 75366400,
|
| 204 |
+
"step": 2300
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.29217073727459486,
|
| 208 |
+
"grad_norm": 0.080078125,
|
| 209 |
+
"learning_rate": 2.8867513459481293e-05,
|
| 210 |
+
"loss": 0.1741,
|
| 211 |
+
"num_input_tokens_seen": 78643200,
|
| 212 |
+
"step": 2400
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.29217073727459486,
|
| 216 |
+
"eval_BLEU": 34.039957674706194,
|
| 217 |
+
"eval_chrF": 59.611408479987894,
|
| 218 |
+
"eval_loss": 0.14743424952030182,
|
| 219 |
+
"eval_runtime": 4005.0853,
|
| 220 |
+
"eval_samples_per_second": 4.712,
|
| 221 |
+
"eval_steps_per_second": 0.295,
|
| 222 |
+
"num_input_tokens_seen": 78643200,
|
| 223 |
+
"step": 2400
|
| 224 |
+
},
|
| 225 |
+
{
|
| 226 |
+
"epoch": 0.3043445179943696,
|
| 227 |
+
"grad_norm": 0.08544921875,
|
| 228 |
+
"learning_rate": 2.8284271247461902e-05,
|
| 229 |
+
"loss": 0.1684,
|
| 230 |
+
"num_input_tokens_seen": 81920000,
|
| 231 |
+
"step": 2500
|
| 232 |
+
},
|
| 233 |
+
{
|
| 234 |
+
"epoch": 0.3165182987141444,
|
| 235 |
+
"grad_norm": 0.08642578125,
|
| 236 |
+
"learning_rate": 2.7735009811261458e-05,
|
| 237 |
+
"loss": 0.1729,
|
| 238 |
+
"num_input_tokens_seen": 85196800,
|
| 239 |
+
"step": 2600
|
| 240 |
+
},
|
| 241 |
+
{
|
| 242 |
+
"epoch": 0.3286920794339192,
|
| 243 |
+
"grad_norm": 0.0888671875,
|
| 244 |
+
"learning_rate": 2.721655269759087e-05,
|
| 245 |
+
"loss": 0.1753,
|
| 246 |
+
"num_input_tokens_seen": 88473600,
|
| 247 |
+
"step": 2700
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.34086586015369397,
|
| 251 |
+
"grad_norm": 0.09423828125,
|
| 252 |
+
"learning_rate": 2.6726124191242442e-05,
|
| 253 |
+
"loss": 0.1702,
|
| 254 |
+
"num_input_tokens_seen": 91750400,
|
| 255 |
+
"step": 2800
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.3530396408734688,
|
| 259 |
+
"grad_norm": 0.10107421875,
|
| 260 |
+
"learning_rate": 2.626128657194451e-05,
|
| 261 |
+
"loss": 0.1743,
|
| 262 |
+
"num_input_tokens_seen": 95027200,
|
| 263 |
+
"step": 2900
|
| 264 |
+
},
|
| 265 |
+
{
|
| 266 |
+
"epoch": 0.36521342159324355,
|
| 267 |
+
"grad_norm": 0.0927734375,
|
| 268 |
+
"learning_rate": 2.581988897471611e-05,
|
| 269 |
+
"loss": 0.1696,
|
| 270 |
+
"num_input_tokens_seen": 98304000,
|
| 271 |
+
"step": 3000
|
| 272 |
+
},
|
| 273 |
+
{
|
| 274 |
+
"epoch": 0.3773872023130183,
|
| 275 |
+
"grad_norm": 0.0810546875,
|
| 276 |
+
"learning_rate": 2.5400025400038102e-05,
|
| 277 |
+
"loss": 0.1724,
|
| 278 |
+
"num_input_tokens_seen": 101580800,
|
| 279 |
+
"step": 3100
|
| 280 |
+
},
|
| 281 |
+
{
|
| 282 |
+
"epoch": 0.38956098303279313,
|
| 283 |
+
"grad_norm": 0.0859375,
|
| 284 |
+
"learning_rate": 2.5e-05,
|
| 285 |
+
"loss": 0.1746,
|
| 286 |
+
"num_input_tokens_seen": 104857600,
|
| 287 |
+
"step": 3200
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 0.4017347637525679,
|
| 291 |
+
"grad_norm": 0.095703125,
|
| 292 |
+
"learning_rate": 2.4618298195866546e-05,
|
| 293 |
+
"loss": 0.1709,
|
| 294 |
+
"num_input_tokens_seen": 108134400,
|
| 295 |
+
"step": 3300
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 0.4139085444723427,
|
| 299 |
+
"grad_norm": 0.087890625,
|
| 300 |
+
"learning_rate": 2.42535625036333e-05,
|
| 301 |
+
"loss": 0.1707,
|
| 302 |
+
"num_input_tokens_seen": 111411200,
|
| 303 |
+
"step": 3400
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.4260823251921175,
|
| 307 |
+
"grad_norm": 0.09228515625,
|
| 308 |
+
"learning_rate": 2.3904572186687872e-05,
|
| 309 |
+
"loss": 0.171,
|
| 310 |
+
"num_input_tokens_seen": 114688000,
|
| 311 |
+
"step": 3500
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.43825610591189224,
|
| 315 |
+
"grad_norm": 0.0869140625,
|
| 316 |
+
"learning_rate": 2.3570226039551588e-05,
|
| 317 |
+
"loss": 0.1728,
|
| 318 |
+
"num_input_tokens_seen": 117964800,
|
| 319 |
+
"step": 3600
|
| 320 |
+
},
|
| 321 |
+
{
|
| 322 |
+
"epoch": 0.43825610591189224,
|
| 323 |
+
"eval_BLEU": 34.1284026902063,
|
| 324 |
+
"eval_chrF": 59.6592142520761,
|
| 325 |
+
"eval_loss": 0.14721617102622986,
|
| 326 |
+
"eval_runtime": 3969.206,
|
| 327 |
+
"eval_samples_per_second": 4.754,
|
| 328 |
+
"eval_steps_per_second": 0.297,
|
| 329 |
+
"num_input_tokens_seen": 117964800,
|
| 330 |
+
"step": 3600
|
| 331 |
+
},
|
| 332 |
+
{
|
| 333 |
+
"epoch": 0.45042988663166705,
|
| 334 |
+
"grad_norm": 0.0966796875,
|
| 335 |
+
"learning_rate": 2.324952774876386e-05,
|
| 336 |
+
"loss": 0.1708,
|
| 337 |
+
"num_input_tokens_seen": 121241600,
|
| 338 |
+
"step": 3700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.4626036673514418,
|
| 342 |
+
"grad_norm": 0.08984375,
|
| 343 |
+
"learning_rate": 2.2941573387056174e-05,
|
| 344 |
+
"loss": 0.1705,
|
| 345 |
+
"num_input_tokens_seen": 124518400,
|
| 346 |
+
"step": 3800
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.47477744807121663,
|
| 350 |
+
"grad_norm": 0.0966796875,
|
| 351 |
+
"learning_rate": 2.2645540682891912e-05,
|
| 352 |
+
"loss": 0.1697,
|
| 353 |
+
"num_input_tokens_seen": 127795200,
|
| 354 |
+
"step": 3900
|
| 355 |
+
},
|
| 356 |
+
{
|
| 357 |
+
"epoch": 0.4869512287909914,
|
| 358 |
+
"grad_norm": 0.091796875,
|
| 359 |
+
"learning_rate": 2.2360679774997898e-05,
|
| 360 |
+
"loss": 0.1722,
|
| 361 |
+
"num_input_tokens_seen": 131072000,
|
| 362 |
+
"step": 4000
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.49912500951076616,
|
| 366 |
+
"grad_norm": 0.07958984375,
|
| 367 |
+
"learning_rate": 2.2086305214969307e-05,
|
| 368 |
+
"loss": 0.1696,
|
| 369 |
+
"num_input_tokens_seen": 134348800,
|
| 370 |
+
"step": 4100
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.5112987902305409,
|
| 374 |
+
"grad_norm": 0.08984375,
|
| 375 |
+
"learning_rate": 2.182178902359924e-05,
|
| 376 |
+
"loss": 0.1696,
|
| 377 |
+
"num_input_tokens_seen": 137625600,
|
| 378 |
+
"step": 4200
|
| 379 |
+
},
|
| 380 |
+
{
|
| 381 |
+
"epoch": 0.5234725709503157,
|
| 382 |
+
"grad_norm": 0.0869140625,
|
| 383 |
+
"learning_rate": 2.1566554640687683e-05,
|
| 384 |
+
"loss": 0.1721,
|
| 385 |
+
"num_input_tokens_seen": 140902400,
|
| 386 |
+
"step": 4300
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.5356463516700906,
|
| 390 |
+
"grad_norm": 0.0791015625,
|
| 391 |
+
"learning_rate": 2.132007163556104e-05,
|
| 392 |
+
"loss": 0.1706,
|
| 393 |
+
"num_input_tokens_seen": 144179200,
|
| 394 |
+
"step": 4400
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.5478201323898654,
|
| 398 |
+
"grad_norm": 0.0849609375,
|
| 399 |
+
"learning_rate": 2.1081851067789197e-05,
|
| 400 |
+
"loss": 0.1726,
|
| 401 |
+
"num_input_tokens_seen": 147456000,
|
| 402 |
+
"step": 4500
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.5599939131096401,
|
| 406 |
+
"grad_norm": 0.08984375,
|
| 407 |
+
"learning_rate": 2.0851441405707478e-05,
|
| 408 |
+
"loss": 0.1688,
|
| 409 |
+
"num_input_tokens_seen": 150732800,
|
| 410 |
+
"step": 4600
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.5721676938294149,
|
| 414 |
+
"grad_norm": 0.08544921875,
|
| 415 |
+
"learning_rate": 2.062842492517587e-05,
|
| 416 |
+
"loss": 0.1733,
|
| 417 |
+
"num_input_tokens_seen": 154009600,
|
| 418 |
+
"step": 4700
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"epoch": 0.5843414745491897,
|
| 422 |
+
"grad_norm": 0.12158203125,
|
| 423 |
+
"learning_rate": 2.0412414523193156e-05,
|
| 424 |
+
"loss": 0.1692,
|
| 425 |
+
"num_input_tokens_seen": 157286400,
|
| 426 |
+
"step": 4800
|
| 427 |
+
},
|
| 428 |
+
{
|
| 429 |
+
"epoch": 0.5843414745491897,
|
| 430 |
+
"eval_BLEU": 34.08945609042042,
|
| 431 |
+
"eval_chrF": 59.66085393977354,
|
| 432 |
+
"eval_loss": 0.14703597128391266,
|
| 433 |
+
"eval_runtime": 3968.308,
|
| 434 |
+
"eval_samples_per_second": 4.755,
|
| 435 |
+
"eval_steps_per_second": 0.297,
|
| 436 |
+
"num_input_tokens_seen": 157286400,
|
| 437 |
+
"step": 4800
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.5965152552689644,
|
| 441 |
+
"grad_norm": 0.087890625,
|
| 442 |
+
"learning_rate": 2.0203050891044213e-05,
|
| 443 |
+
"loss": 0.1717,
|
| 444 |
+
"num_input_tokens_seen": 160563200,
|
| 445 |
+
"step": 4900
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 0.6086890359887392,
|
| 449 |
+
"grad_norm": 0.0908203125,
|
| 450 |
+
"learning_rate": 2e-05,
|
| 451 |
+
"loss": 0.169,
|
| 452 |
+
"num_input_tokens_seen": 163840000,
|
| 453 |
+
"step": 5000
|
| 454 |
+
},
|
| 455 |
+
{
|
| 456 |
+
"epoch": 0.6208628167085141,
|
| 457 |
+
"grad_norm": 0.09619140625,
|
| 458 |
+
"learning_rate": 1.980295085953349e-05,
|
| 459 |
+
"loss": 0.1722,
|
| 460 |
+
"num_input_tokens_seen": 167116800,
|
| 461 |
+
"step": 5100
|
| 462 |
+
},
|
| 463 |
+
{
|
| 464 |
+
"epoch": 0.6330365974282888,
|
| 465 |
+
"grad_norm": 0.09326171875,
|
| 466 |
+
"learning_rate": 1.9611613513818405e-05,
|
| 467 |
+
"loss": 0.1715,
|
| 468 |
+
"num_input_tokens_seen": 170393600,
|
| 469 |
+
"step": 5200
|
| 470 |
+
},
|
| 471 |
+
{
|
| 472 |
+
"epoch": 0.6452103781480636,
|
| 473 |
+
"grad_norm": 0.08203125,
|
| 474 |
+
"learning_rate": 1.9425717247145284e-05,
|
| 475 |
+
"loss": 0.171,
|
| 476 |
+
"num_input_tokens_seen": 173670400,
|
| 477 |
+
"step": 5300
|
| 478 |
+
},
|
| 479 |
+
{
|
| 480 |
+
"epoch": 0.6573841588678384,
|
| 481 |
+
"grad_norm": 0.08740234375,
|
| 482 |
+
"learning_rate": 1.9245008972987527e-05,
|
| 483 |
+
"loss": 0.1702,
|
| 484 |
+
"num_input_tokens_seen": 176947200,
|
| 485 |
+
"step": 5400
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.6695579395876132,
|
| 489 |
+
"grad_norm": 0.07958984375,
|
| 490 |
+
"learning_rate": 1.906925178491185e-05,
|
| 491 |
+
"loss": 0.1719,
|
| 492 |
+
"num_input_tokens_seen": 180224000,
|
| 493 |
+
"step": 5500
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.6817317203073879,
|
| 497 |
+
"grad_norm": 0.08642578125,
|
| 498 |
+
"learning_rate": 1.889822365046136e-05,
|
| 499 |
+
"loss": 0.1727,
|
| 500 |
+
"num_input_tokens_seen": 183500800,
|
| 501 |
+
"step": 5600
|
| 502 |
+
},
|
| 503 |
+
{
|
| 504 |
+
"epoch": 0.6939055010271628,
|
| 505 |
+
"grad_norm": 0.0927734375,
|
| 506 |
+
"learning_rate": 1.873171623163388e-05,
|
| 507 |
+
"loss": 0.1708,
|
| 508 |
+
"num_input_tokens_seen": 186777600,
|
| 509 |
+
"step": 5700
|
| 510 |
+
},
|
| 511 |
+
{
|
| 512 |
+
"epoch": 0.7060792817469376,
|
| 513 |
+
"grad_norm": 0.0888671875,
|
| 514 |
+
"learning_rate": 1.8569533817705186e-05,
|
| 515 |
+
"loss": 0.1694,
|
| 516 |
+
"num_input_tokens_seen": 190054400,
|
| 517 |
+
"step": 5800
|
| 518 |
+
},
|
| 519 |
+
{
|
| 520 |
+
"epoch": 0.7182530624667123,
|
| 521 |
+
"grad_norm": 0.08349609375,
|
| 522 |
+
"learning_rate": 1.841149235796647e-05,
|
| 523 |
+
"loss": 0.1741,
|
| 524 |
+
"num_input_tokens_seen": 193331200,
|
| 525 |
+
"step": 5900
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.7304268431864871,
|
| 529 |
+
"grad_norm": 0.08984375,
|
| 530 |
+
"learning_rate": 1.8257418583505536e-05,
|
| 531 |
+
"loss": 0.1693,
|
| 532 |
+
"num_input_tokens_seen": 196608000,
|
| 533 |
+
"step": 6000
|
| 534 |
+
},
|
| 535 |
+
{
|
| 536 |
+
"epoch": 0.7304268431864871,
|
| 537 |
+
"eval_BLEU": 34.04924804951594,
|
| 538 |
+
"eval_chrF": 59.694436913924406,
|
| 539 |
+
"eval_loss": 0.14692962169647217,
|
| 540 |
+
"eval_runtime": 3965.0872,
|
| 541 |
+
"eval_samples_per_second": 4.759,
|
| 542 |
+
"eval_steps_per_second": 0.298,
|
| 543 |
+
"num_input_tokens_seen": 196608000,
|
| 544 |
+
"step": 6000
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.7426006239062619,
|
| 548 |
+
"grad_norm": 0.0830078125,
|
| 549 |
+
"learning_rate": 1.8107149208503708e-05,
|
| 550 |
+
"loss": 0.1709,
|
| 551 |
+
"num_input_tokens_seen": 199884800,
|
| 552 |
+
"step": 6100
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 0.7547744046260366,
|
| 556 |
+
"grad_norm": 0.08740234375,
|
| 557 |
+
"learning_rate": 1.7960530202677492e-05,
|
| 558 |
+
"loss": 0.1687,
|
| 559 |
+
"num_input_tokens_seen": 203161600,
|
| 560 |
+
"step": 6200
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.7669481853458114,
|
| 564 |
+
"grad_norm": 0.08544921875,
|
| 565 |
+
"learning_rate": 1.781741612749496e-05,
|
| 566 |
+
"loss": 0.1728,
|
| 567 |
+
"num_input_tokens_seen": 206438400,
|
| 568 |
+
"step": 6300
|
| 569 |
+
},
|
| 570 |
+
{
|
| 571 |
+
"epoch": 0.7791219660655863,
|
| 572 |
+
"grad_norm": 0.08447265625,
|
| 573 |
+
"learning_rate": 1.767766952966369e-05,
|
| 574 |
+
"loss": 0.171,
|
| 575 |
+
"num_input_tokens_seen": 209715200,
|
| 576 |
+
"step": 6400
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.7912957467853611,
|
| 580 |
+
"grad_norm": 0.08447265625,
|
| 581 |
+
"learning_rate": 1.7541160386140587e-05,
|
| 582 |
+
"loss": 0.1709,
|
| 583 |
+
"num_input_tokens_seen": 212992000,
|
| 584 |
+
"step": 6500
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.8034695275051358,
|
| 588 |
+
"grad_norm": 0.0849609375,
|
| 589 |
+
"learning_rate": 1.7407765595569787e-05,
|
| 590 |
+
"loss": 0.1709,
|
| 591 |
+
"num_input_tokens_seen": 216268800,
|
| 592 |
+
"step": 6600
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.8156433082249106,
|
| 596 |
+
"grad_norm": 0.09130859375,
|
| 597 |
+
"learning_rate": 1.7277368511627203e-05,
|
| 598 |
+
"loss": 0.1705,
|
| 599 |
+
"num_input_tokens_seen": 219545600,
|
| 600 |
+
"step": 6700
|
| 601 |
+
},
|
| 602 |
+
{
|
| 603 |
+
"epoch": 0.8278170889446854,
|
| 604 |
+
"grad_norm": 0.07861328125,
|
| 605 |
+
"learning_rate": 1.7149858514250882e-05,
|
| 606 |
+
"loss": 0.1709,
|
| 607 |
+
"num_input_tokens_seen": 222822400,
|
| 608 |
+
"step": 6800
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.8399908696644601,
|
| 612 |
+
"grad_norm": 0.08837890625,
|
| 613 |
+
"learning_rate": 1.7025130615174974e-05,
|
| 614 |
+
"loss": 0.1739,
|
| 615 |
+
"num_input_tokens_seen": 226099200,
|
| 616 |
+
"step": 6900
|
| 617 |
+
},
|
| 618 |
+
{
|
| 619 |
+
"epoch": 0.852164650384235,
|
| 620 |
+
"grad_norm": 0.08984375,
|
| 621 |
+
"learning_rate": 1.690308509457033e-05,
|
| 622 |
+
"loss": 0.1709,
|
| 623 |
+
"num_input_tokens_seen": 229376000,
|
| 624 |
+
"step": 7000
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.8643384311040098,
|
| 628 |
+
"grad_norm": 0.083984375,
|
| 629 |
+
"learning_rate": 1.6783627165933782e-05,
|
| 630 |
+
"loss": 0.1733,
|
| 631 |
+
"num_input_tokens_seen": 232652800,
|
| 632 |
+
"step": 7100
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.8765122118237845,
|
| 636 |
+
"grad_norm": 0.09326171875,
|
| 637 |
+
"learning_rate": 1.6666666666666667e-05,
|
| 638 |
+
"loss": 0.1713,
|
| 639 |
+
"num_input_tokens_seen": 235929600,
|
| 640 |
+
"step": 7200
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.8765122118237845,
|
| 644 |
+
"eval_BLEU": 34.04522084811147,
|
| 645 |
+
"eval_chrF": 59.676716604524316,
|
| 646 |
+
"eval_loss": 0.14679720997810364,
|
| 647 |
+
"eval_runtime": 4014.0769,
|
| 648 |
+
"eval_samples_per_second": 4.701,
|
| 649 |
+
"eval_steps_per_second": 0.294,
|
| 650 |
+
"num_input_tokens_seen": 235929600,
|
| 651 |
+
"step": 7200
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"epoch": 0.8886859925435593,
|
| 655 |
+
"grad_norm": 0.0869140625,
|
| 656 |
+
"learning_rate": 1.655211777204736e-05,
|
| 657 |
+
"loss": 0.1704,
|
| 658 |
+
"num_input_tokens_seen": 239206400,
|
| 659 |
+
"step": 7300
|
| 660 |
+
},
|
| 661 |
+
{
|
| 662 |
+
"epoch": 0.9008597732633341,
|
| 663 |
+
"grad_norm": 0.08740234375,
|
| 664 |
+
"learning_rate": 1.643989873053573e-05,
|
| 665 |
+
"loss": 0.1705,
|
| 666 |
+
"num_input_tokens_seen": 242483200,
|
| 667 |
+
"step": 7400
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.9130335539831089,
|
| 671 |
+
"grad_norm": 0.08984375,
|
| 672 |
+
"learning_rate": 1.6329931618554523e-05,
|
| 673 |
+
"loss": 0.1724,
|
| 674 |
+
"num_input_tokens_seen": 245760000,
|
| 675 |
+
"step": 7500
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.9252073347028836,
|
| 679 |
+
"grad_norm": 0.0908203125,
|
| 680 |
+
"learning_rate": 1.6222142113076256e-05,
|
| 681 |
+
"loss": 0.171,
|
| 682 |
+
"num_input_tokens_seen": 249036800,
|
| 683 |
+
"step": 7600
|
| 684 |
+
},
|
| 685 |
+
{
|
| 686 |
+
"epoch": 0.9373811154226585,
|
| 687 |
+
"grad_norm": 0.08349609375,
|
| 688 |
+
"learning_rate": 1.6116459280507607e-05,
|
| 689 |
+
"loss": 0.1719,
|
| 690 |
+
"num_input_tokens_seen": 252313600,
|
| 691 |
+
"step": 7700
|
| 692 |
+
},
|
| 693 |
+
{
|
| 694 |
+
"epoch": 0.9495548961424333,
|
| 695 |
+
"grad_norm": 0.09375,
|
| 696 |
+
"learning_rate": 1.6012815380508712e-05,
|
| 697 |
+
"loss": 0.1698,
|
| 698 |
+
"num_input_tokens_seen": 255590400,
|
| 699 |
+
"step": 7800
|
| 700 |
+
},
|
| 701 |
+
{
|
| 702 |
+
"epoch": 0.961728676862208,
|
| 703 |
+
"grad_norm": 0.09423828125,
|
| 704 |
+
"learning_rate": 1.59111456835146e-05,
|
| 705 |
+
"loss": 0.173,
|
| 706 |
+
"num_input_tokens_seen": 258867200,
|
| 707 |
+
"step": 7900
|
| 708 |
+
},
|
| 709 |
+
{
|
| 710 |
+
"epoch": 0.9739024575819828,
|
| 711 |
+
"grad_norm": 0.08251953125,
|
| 712 |
+
"learning_rate": 1.5811388300841898e-05,
|
| 713 |
+
"loss": 0.1695,
|
| 714 |
+
"num_input_tokens_seen": 262144000,
|
| 715 |
+
"step": 8000
|
| 716 |
+
},
|
| 717 |
+
{
|
| 718 |
+
"epoch": 0.9860762383017576,
|
| 719 |
+
"grad_norm": 0.09033203125,
|
| 720 |
+
"learning_rate": 1.5713484026367723e-05,
|
| 721 |
+
"loss": 0.1735,
|
| 722 |
+
"num_input_tokens_seen": 265420800,
|
| 723 |
+
"step": 8100
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.9982500190215323,
|
| 727 |
+
"grad_norm": 0.08935546875,
|
| 728 |
+
"learning_rate": 1.5617376188860607e-05,
|
| 729 |
+
"loss": 0.1714,
|
| 730 |
+
"num_input_tokens_seen": 268697600,
|
| 731 |
+
"step": 8200
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.0104237997413072,
|
| 735 |
+
"grad_norm": 0.09619140625,
|
| 736 |
+
"learning_rate": 1.5523010514126655e-05,
|
| 737 |
+
"loss": 0.1697,
|
| 738 |
+
"num_input_tokens_seen": 271972864,
|
| 739 |
+
"step": 8300
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 1.0225975804610818,
|
| 743 |
+
"grad_norm": 0.09375,
|
| 744 |
+
"learning_rate": 1.543033499620919e-05,
|
| 745 |
+
"loss": 0.171,
|
| 746 |
+
"num_input_tokens_seen": 275249664,
|
| 747 |
+
"step": 8400
|
| 748 |
+
},
|
| 749 |
+
{
|
| 750 |
+
"epoch": 1.0225975804610818,
|
| 751 |
+
"eval_BLEU": 34.158464778823294,
|
| 752 |
+
"eval_chrF": 59.68876956861537,
|
| 753 |
+
"eval_loss": 0.1467299610376358,
|
| 754 |
+
"eval_runtime": 3923.1304,
|
| 755 |
+
"eval_samples_per_second": 4.81,
|
| 756 |
+
"eval_steps_per_second": 0.301,
|
| 757 |
+
"num_input_tokens_seen": 275249664,
|
| 758 |
+
"step": 8400
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 1.0347713611808567,
|
| 762 |
+
"grad_norm": 0.0869140625,
|
| 763 |
+
"learning_rate": 1.5339299776947406e-05,
|
| 764 |
+
"loss": 0.1697,
|
| 765 |
+
"num_input_tokens_seen": 278526464,
|
| 766 |
+
"step": 8500
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.0469451419006315,
|
| 770 |
+
"grad_norm": 0.0888671875,
|
| 771 |
+
"learning_rate": 1.5249857033260467e-05,
|
| 772 |
+
"loss": 0.1679,
|
| 773 |
+
"num_input_tokens_seen": 281803264,
|
| 774 |
+
"step": 8600
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 1.0591189226204063,
|
| 778 |
+
"grad_norm": 0.080078125,
|
| 779 |
+
"learning_rate": 1.5161960871578069e-05,
|
| 780 |
+
"loss": 0.1718,
|
| 781 |
+
"num_input_tokens_seen": 285080064,
|
| 782 |
+
"step": 8700
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 1.0712927033401811,
|
| 786 |
+
"grad_norm": 0.08056640625,
|
| 787 |
+
"learning_rate": 1.5075567228888182e-05,
|
| 788 |
+
"loss": 0.1701,
|
| 789 |
+
"num_input_tokens_seen": 288356864,
|
| 790 |
+
"step": 8800
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 1.083466484059956,
|
| 794 |
+
"grad_norm": 0.0986328125,
|
| 795 |
+
"learning_rate": 1.499063377991723e-05,
|
| 796 |
+
"loss": 0.1711,
|
| 797 |
+
"num_input_tokens_seen": 291633664,
|
| 798 |
+
"step": 8900
|
| 799 |
+
},
|
| 800 |
+
{
|
| 801 |
+
"epoch": 1.0956402647797308,
|
| 802 |
+
"grad_norm": 0.0859375,
|
| 803 |
+
"learning_rate": 1.49071198499986e-05,
|
| 804 |
+
"loss": 0.1693,
|
| 805 |
+
"num_input_tokens_seen": 294910464,
|
| 806 |
+
"step": 9000
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 1.1078140454995054,
|
| 810 |
+
"grad_norm": 0.087890625,
|
| 811 |
+
"learning_rate": 1.4824986333222024e-05,
|
| 812 |
+
"loss": 0.17,
|
| 813 |
+
"num_input_tokens_seen": 298187264,
|
| 814 |
+
"step": 9100
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 1.1199878262192802,
|
| 818 |
+
"grad_norm": 0.08447265625,
|
| 819 |
+
"learning_rate": 1.4744195615489715e-05,
|
| 820 |
+
"loss": 0.1707,
|
| 821 |
+
"num_input_tokens_seen": 301464064,
|
| 822 |
+
"step": 9200
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.132161606939055,
|
| 826 |
+
"grad_norm": 0.08154296875,
|
| 827 |
+
"learning_rate": 1.4664711502135331e-05,
|
| 828 |
+
"loss": 0.1698,
|
| 829 |
+
"num_input_tokens_seen": 304740864,
|
| 830 |
+
"step": 9300
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 1.1443353876588298,
|
| 834 |
+
"grad_norm": 0.083984375,
|
| 835 |
+
"learning_rate": 1.4586499149789456e-05,
|
| 836 |
+
"loss": 0.1688,
|
| 837 |
+
"num_input_tokens_seen": 308017664,
|
| 838 |
+
"step": 9400
|
| 839 |
+
},
|
| 840 |
+
{
|
| 841 |
+
"epoch": 1.1565091683786046,
|
| 842 |
+
"grad_norm": 0.08837890625,
|
| 843 |
+
"learning_rate": 1.4509525002200233e-05,
|
| 844 |
+
"loss": 0.1674,
|
| 845 |
+
"num_input_tokens_seen": 311294464,
|
| 846 |
+
"step": 9500
|
| 847 |
+
},
|
| 848 |
+
{
|
| 849 |
+
"epoch": 1.1686829490983794,
|
| 850 |
+
"grad_norm": 0.08837890625,
|
| 851 |
+
"learning_rate": 1.4433756729740646e-05,
|
| 852 |
+
"loss": 0.1715,
|
| 853 |
+
"num_input_tokens_seen": 314571264,
|
| 854 |
+
"step": 9600
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 1.1686829490983794,
|
| 858 |
+
"eval_BLEU": 34.125660852650576,
|
| 859 |
+
"eval_chrF": 59.69577135277113,
|
| 860 |
+
"eval_loss": 0.14671051502227783,
|
| 861 |
+
"eval_runtime": 3968.6253,
|
| 862 |
+
"eval_samples_per_second": 4.755,
|
| 863 |
+
"eval_steps_per_second": 0.297,
|
| 864 |
+
"num_input_tokens_seen": 314571264,
|
| 865 |
+
"step": 9600
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 1.1808567298181543,
|
| 869 |
+
"grad_norm": 0.08642578125,
|
| 870 |
+
"learning_rate": 1.4359163172354764e-05,
|
| 871 |
+
"loss": 0.1693,
|
| 872 |
+
"num_input_tokens_seen": 317848064,
|
| 873 |
+
"step": 9700
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 1.1930305105379289,
|
| 877 |
+
"grad_norm": 0.09521484375,
|
| 878 |
+
"learning_rate": 1.4285714285714285e-05,
|
| 879 |
+
"loss": 0.1685,
|
| 880 |
+
"num_input_tokens_seen": 321124864,
|
| 881 |
+
"step": 9800
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 1.2052042912577037,
|
| 885 |
+
"grad_norm": 0.08740234375,
|
| 886 |
+
"learning_rate": 1.4213381090374031e-05,
|
| 887 |
+
"loss": 0.1686,
|
| 888 |
+
"num_input_tokens_seen": 324401664,
|
| 889 |
+
"step": 9900
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 1.2173780719774785,
|
| 893 |
+
"grad_norm": 0.087890625,
|
| 894 |
+
"learning_rate": 1.4142135623730951e-05,
|
| 895 |
+
"loss": 0.1711,
|
| 896 |
+
"num_input_tokens_seen": 327678464,
|
| 897 |
+
"step": 10000
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 1.2295518526972533,
|
| 901 |
+
"grad_norm": 0.08447265625,
|
| 902 |
+
"learning_rate": 1.4071950894605838e-05,
|
| 903 |
+
"loss": 0.1707,
|
| 904 |
+
"num_input_tokens_seen": 330955264,
|
| 905 |
+
"step": 10100
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 1.2417256334170281,
|
| 909 |
+
"grad_norm": 0.08740234375,
|
| 910 |
+
"learning_rate": 1.4002800840280098e-05,
|
| 911 |
+
"loss": 0.1704,
|
| 912 |
+
"num_input_tokens_seen": 334232064,
|
| 913 |
+
"step": 10200
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 1.2538994141368027,
|
| 917 |
+
"grad_norm": 0.0830078125,
|
| 918 |
+
"learning_rate": 1.3934660285832355e-05,
|
| 919 |
+
"loss": 0.1666,
|
| 920 |
+
"num_input_tokens_seen": 337508864,
|
| 921 |
+
"step": 10300
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 1.2660731948565775,
|
| 925 |
+
"grad_norm": 0.08544921875,
|
| 926 |
+
"learning_rate": 1.3867504905630729e-05,
|
| 927 |
+
"loss": 0.1688,
|
| 928 |
+
"num_input_tokens_seen": 340785664,
|
| 929 |
+
"step": 10400
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 1.2782469755763524,
|
| 933 |
+
"grad_norm": 0.09619140625,
|
| 934 |
+
"learning_rate": 1.3801311186847085e-05,
|
| 935 |
+
"loss": 0.1706,
|
| 936 |
+
"num_input_tokens_seen": 344062464,
|
| 937 |
+
"step": 10500
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 1.2904207562961272,
|
| 941 |
+
"grad_norm": 0.0859375,
|
| 942 |
+
"learning_rate": 1.3736056394868905e-05,
|
| 943 |
+
"loss": 0.1699,
|
| 944 |
+
"num_input_tokens_seen": 347339264,
|
| 945 |
+
"step": 10600
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 1.302594537015902,
|
| 949 |
+
"grad_norm": 0.0888671875,
|
| 950 |
+
"learning_rate": 1.3671718540493266e-05,
|
| 951 |
+
"loss": 0.1686,
|
| 952 |
+
"num_input_tokens_seen": 350616064,
|
| 953 |
+
"step": 10700
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 1.3147683177356768,
|
| 957 |
+
"grad_norm": 0.083984375,
|
| 958 |
+
"learning_rate": 1.3608276348795434e-05,
|
| 959 |
+
"loss": 0.1705,
|
| 960 |
+
"num_input_tokens_seen": 353892864,
|
| 961 |
+
"step": 10800
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 1.3147683177356768,
|
| 965 |
+
"eval_BLEU": 34.190278283000254,
|
| 966 |
+
"eval_chrF": 59.72105944050832,
|
| 967 |
+
"eval_loss": 0.14669395983219147,
|
| 968 |
+
"eval_runtime": 3965.5317,
|
| 969 |
+
"eval_samples_per_second": 4.759,
|
| 970 |
+
"eval_steps_per_second": 0.298,
|
| 971 |
+
"num_input_tokens_seen": 353892864,
|
| 972 |
+
"step": 10800
|
| 973 |
+
},
|
| 974 |
+
{
|
| 975 |
+
"epoch": 1.3269420984554516,
|
| 976 |
+
"grad_norm": 0.08251953125,
|
| 977 |
+
"learning_rate": 1.3545709229571929e-05,
|
| 978 |
+
"loss": 0.1702,
|
| 979 |
+
"num_input_tokens_seen": 357169664,
|
| 980 |
+
"step": 10900
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 1.3391158791752265,
|
| 984 |
+
"grad_norm": 0.08544921875,
|
| 985 |
+
"learning_rate": 1.3483997249264842e-05,
|
| 986 |
+
"loss": 0.1693,
|
| 987 |
+
"num_input_tokens_seen": 360446464,
|
| 988 |
+
"step": 11000
|
| 989 |
+
},
|
| 990 |
+
{
|
| 991 |
+
"epoch": 1.3512896598950013,
|
| 992 |
+
"grad_norm": 0.08544921875,
|
| 993 |
+
"learning_rate": 1.3423121104280487e-05,
|
| 994 |
+
"loss": 0.1711,
|
| 995 |
+
"num_input_tokens_seen": 363723264,
|
| 996 |
+
"step": 11100
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 1.3634634406147759,
|
| 1000 |
+
"grad_norm": 0.0849609375,
|
| 1001 |
+
"learning_rate": 1.3363062095621221e-05,
|
| 1002 |
+
"loss": 0.1687,
|
| 1003 |
+
"num_input_tokens_seen": 367000064,
|
| 1004 |
+
"step": 11200
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 1.3756372213345507,
|
| 1008 |
+
"grad_norm": 0.09033203125,
|
| 1009 |
+
"learning_rate": 1.3303802104754787e-05,
|
| 1010 |
+
"loss": 0.1701,
|
| 1011 |
+
"num_input_tokens_seen": 370276864,
|
| 1012 |
+
"step": 11300
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 1.3878110020543255,
|
| 1016 |
+
"grad_norm": 0.10107421875,
|
| 1017 |
+
"learning_rate": 1.324532357065044e-05,
|
| 1018 |
+
"loss": 0.1706,
|
| 1019 |
+
"num_input_tokens_seen": 373553664,
|
| 1020 |
+
"step": 11400
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 1.3999847827741003,
|
| 1024 |
+
"grad_norm": 0.08544921875,
|
| 1025 |
+
"learning_rate": 1.318760946791574e-05,
|
| 1026 |
+
"loss": 0.1691,
|
| 1027 |
+
"num_input_tokens_seen": 376830464,
|
| 1028 |
+
"step": 11500
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 1.4121585634938751,
|
| 1032 |
+
"grad_norm": 0.09130859375,
|
| 1033 |
+
"learning_rate": 1.3130643285972255e-05,
|
| 1034 |
+
"loss": 0.1699,
|
| 1035 |
+
"num_input_tokens_seen": 380107264,
|
| 1036 |
+
"step": 11600
|
| 1037 |
+
},
|
| 1038 |
+
{
|
| 1039 |
+
"epoch": 1.4243323442136497,
|
| 1040 |
+
"grad_norm": 0.095703125,
|
| 1041 |
+
"learning_rate": 1.3074409009212269e-05,
|
| 1042 |
+
"loss": 0.1698,
|
| 1043 |
+
"num_input_tokens_seen": 383384064,
|
| 1044 |
+
"step": 11700
|
| 1045 |
+
},
|
| 1046 |
+
{
|
| 1047 |
+
"epoch": 1.4365061249334246,
|
| 1048 |
+
"grad_norm": 0.0849609375,
|
| 1049 |
+
"learning_rate": 1.3018891098082389e-05,
|
| 1050 |
+
"loss": 0.1692,
|
| 1051 |
+
"num_input_tokens_seen": 386660864,
|
| 1052 |
+
"step": 11800
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 1.4486799056531994,
|
| 1056 |
+
"grad_norm": 0.08642578125,
|
| 1057 |
+
"learning_rate": 1.2964074471043288e-05,
|
| 1058 |
+
"loss": 0.1691,
|
| 1059 |
+
"num_input_tokens_seen": 389937664,
|
| 1060 |
+
"step": 11900
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 1.4608536863729742,
|
| 1064 |
+
"grad_norm": 0.0849609375,
|
| 1065 |
+
"learning_rate": 1.2909944487358055e-05,
|
| 1066 |
+
"loss": 0.1709,
|
| 1067 |
+
"num_input_tokens_seen": 393214464,
|
| 1068 |
+
"step": 12000
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 1.4608536863729742,
|
| 1072 |
+
"eval_BLEU": 34.19371467731715,
|
| 1073 |
+
"eval_chrF": 59.7146095038311,
|
| 1074 |
+
"eval_loss": 0.14666913449764252,
|
| 1075 |
+
"eval_runtime": 3982.0054,
|
| 1076 |
+
"eval_samples_per_second": 4.739,
|
| 1077 |
+
"eval_steps_per_second": 0.296,
|
| 1078 |
+
"num_input_tokens_seen": 393214464,
|
| 1079 |
+
"step": 12000
|
| 1080 |
+
},
|
| 1081 |
+
{
|
| 1082 |
+
"epoch": 1.473027467092749,
|
| 1083 |
+
"grad_norm": 0.08837890625,
|
| 1084 |
+
"learning_rate": 1.2856486930664503e-05,
|
| 1085 |
+
"loss": 0.1719,
|
| 1086 |
+
"num_input_tokens_seen": 396491264,
|
| 1087 |
+
"step": 12100
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 1.4852012478125238,
|
| 1091 |
+
"grad_norm": 0.08447265625,
|
| 1092 |
+
"learning_rate": 1.2803687993289598e-05,
|
| 1093 |
+
"loss": 0.1677,
|
| 1094 |
+
"num_input_tokens_seen": 399768064,
|
| 1095 |
+
"step": 12200
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 1.4973750285322986,
|
| 1099 |
+
"grad_norm": 0.0830078125,
|
| 1100 |
+
"learning_rate": 1.2751534261266765e-05,
|
| 1101 |
+
"loss": 0.1696,
|
| 1102 |
+
"num_input_tokens_seen": 403044864,
|
| 1103 |
+
"step": 12300
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 1.5095488092520735,
|
| 1107 |
+
"grad_norm": 0.08447265625,
|
| 1108 |
+
"learning_rate": 1.2700012700019051e-05,
|
| 1109 |
+
"loss": 0.1692,
|
| 1110 |
+
"num_input_tokens_seen": 406321664,
|
| 1111 |
+
"step": 12400
|
| 1112 |
+
},
|
| 1113 |
+
{
|
| 1114 |
+
"epoch": 1.5217225899718483,
|
| 1115 |
+
"grad_norm": 0.08740234375,
|
| 1116 |
+
"learning_rate": 1.2649110640673517e-05,
|
| 1117 |
+
"loss": 0.1718,
|
| 1118 |
+
"num_input_tokens_seen": 409598464,
|
| 1119 |
+
"step": 12500
|
| 1120 |
+
},
|
| 1121 |
+
{
|
| 1122 |
+
"epoch": 1.5338963706916229,
|
| 1123 |
+
"grad_norm": 0.08544921875,
|
| 1124 |
+
"learning_rate": 1.2598815766974239e-05,
|
| 1125 |
+
"loss": 0.1707,
|
| 1126 |
+
"num_input_tokens_seen": 412875264,
|
| 1127 |
+
"step": 12600
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"epoch": 1.5460701514113977,
|
| 1131 |
+
"grad_norm": 0.09521484375,
|
| 1132 |
+
"learning_rate": 1.2549116102763172e-05,
|
| 1133 |
+
"loss": 0.1678,
|
| 1134 |
+
"num_input_tokens_seen": 416152064,
|
| 1135 |
+
"step": 12700
|
| 1136 |
+
},
|
| 1137 |
+
{
|
| 1138 |
+
"epoch": 1.5582439321311725,
|
| 1139 |
+
"grad_norm": 0.07958984375,
|
| 1140 |
+
"learning_rate": 1.25e-05,
|
| 1141 |
+
"loss": 0.1668,
|
| 1142 |
+
"num_input_tokens_seen": 419428864,
|
| 1143 |
+
"step": 12800
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.5704177128509471,
|
| 1147 |
+
"grad_norm": 0.08740234375,
|
| 1148 |
+
"learning_rate": 1.2451456127293808e-05,
|
| 1149 |
+
"loss": 0.169,
|
| 1150 |
+
"num_input_tokens_seen": 422705664,
|
| 1151 |
+
"step": 12900
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.582591493570722,
|
| 1155 |
+
"grad_norm": 0.08642578125,
|
| 1156 |
+
"learning_rate": 1.2403473458920847e-05,
|
| 1157 |
+
"loss": 0.1677,
|
| 1158 |
+
"num_input_tokens_seen": 425982464,
|
| 1159 |
+
"step": 13000
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 1.5947652742904967,
|
| 1163 |
+
"grad_norm": 0.0830078125,
|
| 1164 |
+
"learning_rate": 1.2356041264304309e-05,
|
| 1165 |
+
"loss": 0.168,
|
| 1166 |
+
"num_input_tokens_seen": 429259264,
|
| 1167 |
+
"step": 13100
|
| 1168 |
+
},
|
| 1169 |
+
{
|
| 1170 |
+
"epoch": 1.6069390550102716,
|
| 1171 |
+
"grad_norm": 0.09765625,
|
| 1172 |
+
"learning_rate": 1.2309149097933273e-05,
|
| 1173 |
+
"loss": 0.1699,
|
| 1174 |
+
"num_input_tokens_seen": 432536064,
|
| 1175 |
+
"step": 13200
|
| 1176 |
+
},
|
| 1177 |
+
{
|
| 1178 |
+
"epoch": 1.6069390550102716,
|
| 1179 |
+
"eval_BLEU": 34.185916767080265,
|
| 1180 |
+
"eval_chrF": 59.706613240927844,
|
| 1181 |
+
"eval_loss": 0.14664307236671448,
|
| 1182 |
+
"eval_runtime": 3971.3932,
|
| 1183 |
+
"eval_samples_per_second": 4.751,
|
| 1184 |
+
"eval_steps_per_second": 0.297,
|
| 1185 |
+
"num_input_tokens_seen": 432536064,
|
| 1186 |
+
"step": 13200
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 1.6191128357300464,
|
| 1190 |
+
"grad_norm": 0.08447265625,
|
| 1191 |
+
"learning_rate": 1.2262786789699317e-05,
|
| 1192 |
+
"loss": 0.1686,
|
| 1193 |
+
"num_input_tokens_seen": 435812864,
|
| 1194 |
+
"step": 13300
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 1.6312866164498212,
|
| 1198 |
+
"grad_norm": 0.09326171875,
|
| 1199 |
+
"learning_rate": 1.2216944435630522e-05,
|
| 1200 |
+
"loss": 0.1662,
|
| 1201 |
+
"num_input_tokens_seen": 439089664,
|
| 1202 |
+
"step": 13400
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 1.643460397169596,
|
| 1206 |
+
"grad_norm": 0.091796875,
|
| 1207 |
+
"learning_rate": 1.2171612389003693e-05,
|
| 1208 |
+
"loss": 0.1704,
|
| 1209 |
+
"num_input_tokens_seen": 442366464,
|
| 1210 |
+
"step": 13500
|
| 1211 |
+
},
|
| 1212 |
+
{
|
| 1213 |
+
"epoch": 1.6556341778893708,
|
| 1214 |
+
"grad_norm": 0.09716796875,
|
| 1215 |
+
"learning_rate": 1.212678125181665e-05,
|
| 1216 |
+
"loss": 0.1717,
|
| 1217 |
+
"num_input_tokens_seen": 445643264,
|
| 1218 |
+
"step": 13600
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 1.6678079586091457,
|
| 1222 |
+
"grad_norm": 0.09619140625,
|
| 1223 |
+
"learning_rate": 1.208244186660354e-05,
|
| 1224 |
+
"loss": 0.17,
|
| 1225 |
+
"num_input_tokens_seen": 448920064,
|
| 1226 |
+
"step": 13700
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 1.6799817393289205,
|
| 1230 |
+
"grad_norm": 0.07958984375,
|
| 1231 |
+
"learning_rate": 1.203858530857692e-05,
|
| 1232 |
+
"loss": 0.1699,
|
| 1233 |
+
"num_input_tokens_seen": 452196864,
|
| 1234 |
+
"step": 13800
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.692155520048695,
|
| 1238 |
+
"grad_norm": 0.09619140625,
|
| 1239 |
+
"learning_rate": 1.1995202878081345e-05,
|
| 1240 |
+
"loss": 0.1703,
|
| 1241 |
+
"num_input_tokens_seen": 455473664,
|
| 1242 |
+
"step": 13900
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 1.70432930076847,
|
| 1246 |
+
"grad_norm": 0.0908203125,
|
| 1247 |
+
"learning_rate": 1.1952286093343936e-05,
|
| 1248 |
+
"loss": 0.173,
|
| 1249 |
+
"num_input_tokens_seen": 458750464,
|
| 1250 |
+
"step": 14000
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 1.7165030814882447,
|
| 1254 |
+
"grad_norm": 0.0908203125,
|
| 1255 |
+
"learning_rate": 1.1909826683508273e-05,
|
| 1256 |
+
"loss": 0.1694,
|
| 1257 |
+
"num_input_tokens_seen": 462027264,
|
| 1258 |
+
"step": 14100
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 1.7286768622080195,
|
| 1262 |
+
"grad_norm": 0.080078125,
|
| 1263 |
+
"learning_rate": 1.1867816581938534e-05,
|
| 1264 |
+
"loss": 0.1698,
|
| 1265 |
+
"num_input_tokens_seen": 465304064,
|
| 1266 |
+
"step": 14200
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 1.7408506429277941,
|
| 1270 |
+
"grad_norm": 0.08251953125,
|
| 1271 |
+
"learning_rate": 1.1826247919781652e-05,
|
| 1272 |
+
"loss": 0.1706,
|
| 1273 |
+
"num_input_tokens_seen": 468580864,
|
| 1274 |
+
"step": 14300
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 1.753024423647569,
|
| 1278 |
+
"grad_norm": 0.08544921875,
|
| 1279 |
+
"learning_rate": 1.1785113019775794e-05,
|
| 1280 |
+
"loss": 0.1699,
|
| 1281 |
+
"num_input_tokens_seen": 471857664,
|
| 1282 |
+
"step": 14400
|
| 1283 |
+
},
|
| 1284 |
+
{
|
| 1285 |
+
"epoch": 1.753024423647569,
|
| 1286 |
+
"eval_BLEU": 34.102099613409436,
|
| 1287 |
+
"eval_chrF": 59.716548967741204,
|
| 1288 |
+
"eval_loss": 0.1466248631477356,
|
| 1289 |
+
"eval_runtime": 3984.4468,
|
| 1290 |
+
"eval_samples_per_second": 4.736,
|
| 1291 |
+
"eval_steps_per_second": 0.296,
|
| 1292 |
+
"num_input_tokens_seen": 471857664,
|
| 1293 |
+
"step": 14400
|
| 1294 |
+
},
|
| 1295 |
+
{
|
| 1296 |
+
"epoch": 1.7651982043673438,
|
| 1297 |
+
"grad_norm": 0.0859375,
|
| 1298 |
+
"learning_rate": 1.174440439029407e-05,
|
| 1299 |
+
"loss": 0.1677,
|
| 1300 |
+
"num_input_tokens_seen": 475134464,
|
| 1301 |
+
"step": 14500
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 1.7773719850871186,
|
| 1305 |
+
"grad_norm": 0.07958984375,
|
| 1306 |
+
"learning_rate": 1.1704114719613058e-05,
|
| 1307 |
+
"loss": 0.1724,
|
| 1308 |
+
"num_input_tokens_seen": 478411264,
|
| 1309 |
+
"step": 14600
|
| 1310 |
+
},
|
| 1311 |
+
{
|
| 1312 |
+
"epoch": 1.7895457658068934,
|
| 1313 |
+
"grad_norm": 0.0947265625,
|
| 1314 |
+
"learning_rate": 1.1664236870396087e-05,
|
| 1315 |
+
"loss": 0.1684,
|
| 1316 |
+
"num_input_tokens_seen": 481688064,
|
| 1317 |
+
"step": 14700
|
| 1318 |
+
},
|
| 1319 |
+
{
|
| 1320 |
+
"epoch": 1.8017195465266682,
|
| 1321 |
+
"grad_norm": 0.08740234375,
|
| 1322 |
+
"learning_rate": 1.162476387438193e-05,
|
| 1323 |
+
"loss": 0.1689,
|
| 1324 |
+
"num_input_tokens_seen": 484964864,
|
| 1325 |
+
"step": 14800
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.813893327246443,
|
| 1329 |
+
"grad_norm": 0.09033203125,
|
| 1330 |
+
"learning_rate": 1.1585688927269846e-05,
|
| 1331 |
+
"loss": 0.1681,
|
| 1332 |
+
"num_input_tokens_seen": 488241664,
|
| 1333 |
+
"step": 14900
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 1.8260671079662179,
|
| 1337 |
+
"grad_norm": 0.0830078125,
|
| 1338 |
+
"learning_rate": 1.1547005383792514e-05,
|
| 1339 |
+
"loss": 0.1682,
|
| 1340 |
+
"num_input_tokens_seen": 491518464,
|
| 1341 |
+
"step": 15000
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 1.8382408886859927,
|
| 1345 |
+
"grad_norm": 0.09423828125,
|
| 1346 |
+
"learning_rate": 1.150870675296872e-05,
|
| 1347 |
+
"loss": 0.1725,
|
| 1348 |
+
"num_input_tokens_seen": 494795264,
|
| 1349 |
+
"step": 15100
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 1.8504146694057675,
|
| 1353 |
+
"grad_norm": 0.08203125,
|
| 1354 |
+
"learning_rate": 1.1470786693528087e-05,
|
| 1355 |
+
"loss": 0.167,
|
| 1356 |
+
"num_input_tokens_seen": 498072064,
|
| 1357 |
+
"step": 15200
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"epoch": 1.862588450125542,
|
| 1361 |
+
"grad_norm": 0.09130859375,
|
| 1362 |
+
"learning_rate": 1.143323900950059e-05,
|
| 1363 |
+
"loss": 0.1695,
|
| 1364 |
+
"num_input_tokens_seen": 501348864,
|
| 1365 |
+
"step": 15300
|
| 1366 |
+
},
|
| 1367 |
+
{
|
| 1368 |
+
"epoch": 1.874762230845317,
|
| 1369 |
+
"grad_norm": 0.08349609375,
|
| 1370 |
+
"learning_rate": 1.1396057645963795e-05,
|
| 1371 |
+
"loss": 0.1696,
|
| 1372 |
+
"num_input_tokens_seen": 504625664,
|
| 1373 |
+
"step": 15400
|
| 1374 |
+
},
|
| 1375 |
+
{
|
| 1376 |
+
"epoch": 1.8869360115650917,
|
| 1377 |
+
"grad_norm": 0.0859375,
|
| 1378 |
+
"learning_rate": 1.1359236684941297e-05,
|
| 1379 |
+
"loss": 0.1679,
|
| 1380 |
+
"num_input_tokens_seen": 507902464,
|
| 1381 |
+
"step": 15500
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.8991097922848663,
|
| 1385 |
+
"grad_norm": 0.0869140625,
|
| 1386 |
+
"learning_rate": 1.1322770341445956e-05,
|
| 1387 |
+
"loss": 0.17,
|
| 1388 |
+
"num_input_tokens_seen": 511179264,
|
| 1389 |
+
"step": 15600
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 1.8991097922848663,
|
| 1393 |
+
"eval_BLEU": 34.13689431537213,
|
| 1394 |
+
"eval_chrF": 59.7127572353836,
|
| 1395 |
+
"eval_loss": 0.14662402868270874,
|
| 1396 |
+
"eval_runtime": 3972.3482,
|
| 1397 |
+
"eval_samples_per_second": 4.75,
|
| 1398 |
+
"eval_steps_per_second": 0.297,
|
| 1399 |
+
"num_input_tokens_seen": 511179264,
|
| 1400 |
+
"step": 15600
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 1.9112835730046411,
|
| 1404 |
+
"grad_norm": 0.09521484375,
|
| 1405 |
+
"learning_rate": 1.1286652959662007e-05,
|
| 1406 |
+
"loss": 0.1685,
|
| 1407 |
+
"num_input_tokens_seen": 514456064,
|
| 1408 |
+
"step": 15700
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 1.923457353724416,
|
| 1412 |
+
"grad_norm": 0.08740234375,
|
| 1413 |
+
"learning_rate": 1.125087900926024e-05,
|
| 1414 |
+
"loss": 0.1713,
|
| 1415 |
+
"num_input_tokens_seen": 517732864,
|
| 1416 |
+
"step": 15800
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 1.9356311344441908,
|
| 1420 |
+
"grad_norm": 0.08544921875,
|
| 1421 |
+
"learning_rate": 1.1215443081840888e-05,
|
| 1422 |
+
"loss": 0.169,
|
| 1423 |
+
"num_input_tokens_seen": 521009664,
|
| 1424 |
+
"step": 15900
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 1.9478049151639656,
|
| 1428 |
+
"grad_norm": 0.0908203125,
|
| 1429 |
+
"learning_rate": 1.1180339887498949e-05,
|
| 1430 |
+
"loss": 0.1717,
|
| 1431 |
+
"num_input_tokens_seen": 524286464,
|
| 1432 |
+
"step": 16000
|
| 1433 |
+
},
|
| 1434 |
+
{
|
| 1435 |
+
"epoch": 1.9599786958837404,
|
| 1436 |
+
"grad_norm": 0.09130859375,
|
| 1437 |
+
"learning_rate": 1.1145564251507057e-05,
|
| 1438 |
+
"loss": 0.1707,
|
| 1439 |
+
"num_input_tokens_seen": 527563264,
|
| 1440 |
+
"step": 16100
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 1.9721524766035152,
|
| 1444 |
+
"grad_norm": 0.08642578125,
|
| 1445 |
+
"learning_rate": 1.1111111111111112e-05,
|
| 1446 |
+
"loss": 0.1659,
|
| 1447 |
+
"num_input_tokens_seen": 530840064,
|
| 1448 |
+
"step": 16200
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 1.98432625732329,
|
| 1452 |
+
"grad_norm": 0.09619140625,
|
| 1453 |
+
"learning_rate": 1.1076975512434226e-05,
|
| 1454 |
+
"loss": 0.168,
|
| 1455 |
+
"num_input_tokens_seen": 534116864,
|
| 1456 |
+
"step": 16300
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 1.9965000380430649,
|
| 1460 |
+
"grad_norm": 0.08447265625,
|
| 1461 |
+
"learning_rate": 1.1043152607484654e-05,
|
| 1462 |
+
"loss": 0.1684,
|
| 1463 |
+
"num_input_tokens_seen": 537393664,
|
| 1464 |
+
"step": 16400
|
| 1465 |
+
}
|
| 1466 |
+
],
|
| 1467 |
+
"logging_steps": 100,
|
| 1468 |
+
"max_steps": 16428,
|
| 1469 |
+
"num_input_tokens_seen": 538311168,
|
| 1470 |
+
"num_train_epochs": 2,
|
| 1471 |
+
"save_steps": 1200,
|
| 1472 |
+
"stateful_callbacks": {
|
| 1473 |
+
"EarlyStoppingCallback": {
|
| 1474 |
+
"args": {
|
| 1475 |
+
"early_stopping_patience": 5,
|
| 1476 |
+
"early_stopping_threshold": 0.001
|
| 1477 |
+
},
|
| 1478 |
+
"attributes": {
|
| 1479 |
+
"early_stopping_patience_counter": 0
|
| 1480 |
+
}
|
| 1481 |
+
},
|
| 1482 |
+
"TrainerControl": {
|
| 1483 |
+
"args": {
|
| 1484 |
+
"should_epoch_stop": false,
|
| 1485 |
+
"should_evaluate": false,
|
| 1486 |
+
"should_log": false,
|
| 1487 |
+
"should_save": true,
|
| 1488 |
+
"should_training_stop": true
|
| 1489 |
+
},
|
| 1490 |
+
"attributes": {}
|
| 1491 |
+
}
|
| 1492 |
+
},
|
| 1493 |
+
"total_flos": 3.1036059771249623e+18,
|
| 1494 |
+
"train_batch_size": 16,
|
| 1495 |
+
"trial_name": null,
|
| 1496 |
+
"trial_params": null
|
| 1497 |
+
}
|