Upload folder using huggingface_hub
#1
by
krishnakalyan3
- opened
- model.safetensors +3 -0
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +2081 -0
- training_args.bin +3 -0
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7857a3e5f40dabccf6bf852e721dea7668e7e5b5aaa3bbc22cbd85badef3104f
|
| 3 |
+
size 379087640
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9beaa7bca065d2f5bc772822b91f09a557f3565f6f06aa931cbabc2477d2e95d
|
| 3 |
+
size 3152480
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9614b1e00e83ff4fa54e7e1e99e3350eb2891032682ec5953e9d50d2a2ac3d62
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:753688a2200f9a902f80c4e07cec91cc8cbe92c5843812ec4b7044ac745b7284
|
| 3 |
+
size 1000
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2081 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.121661689779634,
|
| 3 |
+
"best_model_checkpoint": "/workspace/disk2/krishna/checkpoints/checkpoint-1280",
|
| 4 |
+
"epoch": 0.128,
|
| 5 |
+
"eval_steps": 10,
|
| 6 |
+
"global_step": 1280,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.001,
|
| 13 |
+
"grad_norm": 0.11198576539754868,
|
| 14 |
+
"learning_rate": 1e-05,
|
| 15 |
+
"loss": 0.126,
|
| 16 |
+
"step": 10
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.001,
|
| 20 |
+
"eval_cos_sim": 0.8696296215057373,
|
| 21 |
+
"eval_loss": 0.13132101871716445,
|
| 22 |
+
"eval_runtime": 191.9539,
|
| 23 |
+
"eval_samples_per_second": 20.838,
|
| 24 |
+
"eval_steps_per_second": 1.302,
|
| 25 |
+
"step": 10
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.002,
|
| 29 |
+
"grad_norm": 0.19444850087165833,
|
| 30 |
+
"learning_rate": 2e-05,
|
| 31 |
+
"loss": 0.1267,
|
| 32 |
+
"step": 20
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.002,
|
| 36 |
+
"eval_cos_sim": 0.8698329329490662,
|
| 37 |
+
"eval_loss": 0.1311149292205519,
|
| 38 |
+
"eval_runtime": 177.5098,
|
| 39 |
+
"eval_samples_per_second": 22.534,
|
| 40 |
+
"eval_steps_per_second": 1.408,
|
| 41 |
+
"step": 20
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.003,
|
| 45 |
+
"grad_norm": 0.12954622507095337,
|
| 46 |
+
"learning_rate": 3e-05,
|
| 47 |
+
"loss": 0.1271,
|
| 48 |
+
"step": 30
|
| 49 |
+
},
|
| 50 |
+
{
|
| 51 |
+
"epoch": 0.003,
|
| 52 |
+
"eval_cos_sim": 0.8700494766235352,
|
| 53 |
+
"eval_loss": 0.1309011602615065,
|
| 54 |
+
"eval_runtime": 179.7068,
|
| 55 |
+
"eval_samples_per_second": 22.258,
|
| 56 |
+
"eval_steps_per_second": 1.391,
|
| 57 |
+
"step": 30
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.004,
|
| 61 |
+
"grad_norm": 0.11514733731746674,
|
| 62 |
+
"learning_rate": 4e-05,
|
| 63 |
+
"loss": 0.1265,
|
| 64 |
+
"step": 40
|
| 65 |
+
},
|
| 66 |
+
{
|
| 67 |
+
"epoch": 0.004,
|
| 68 |
+
"eval_cos_sim": 0.870728075504303,
|
| 69 |
+
"eval_loss": 0.13021534349667496,
|
| 70 |
+
"eval_runtime": 174.4918,
|
| 71 |
+
"eval_samples_per_second": 22.924,
|
| 72 |
+
"eval_steps_per_second": 1.433,
|
| 73 |
+
"step": 40
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.005,
|
| 77 |
+
"grad_norm": 0.34224584698677063,
|
| 78 |
+
"learning_rate": 5e-05,
|
| 79 |
+
"loss": 0.1273,
|
| 80 |
+
"step": 50
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.005,
|
| 84 |
+
"eval_cos_sim": 0.8705285787582397,
|
| 85 |
+
"eval_loss": 0.1304176144813246,
|
| 86 |
+
"eval_runtime": 175.5157,
|
| 87 |
+
"eval_samples_per_second": 22.79,
|
| 88 |
+
"eval_steps_per_second": 1.424,
|
| 89 |
+
"step": 50
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.006,
|
| 93 |
+
"grad_norm": 0.1085827499628067,
|
| 94 |
+
"learning_rate": 4.517892759404963e-05,
|
| 95 |
+
"loss": 0.125,
|
| 96 |
+
"step": 60
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.006,
|
| 100 |
+
"eval_cos_sim": 0.8709338903427124,
|
| 101 |
+
"eval_loss": 0.130007851145143,
|
| 102 |
+
"eval_runtime": 173.9237,
|
| 103 |
+
"eval_samples_per_second": 22.999,
|
| 104 |
+
"eval_steps_per_second": 1.437,
|
| 105 |
+
"step": 60
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.007,
|
| 109 |
+
"grad_norm": 0.11786766350269318,
|
| 110 |
+
"learning_rate": 3.257512950767182e-05,
|
| 111 |
+
"loss": 0.1291,
|
| 112 |
+
"step": 70
|
| 113 |
+
},
|
| 114 |
+
{
|
| 115 |
+
"epoch": 0.007,
|
| 116 |
+
"eval_cos_sim": 0.8714690208435059,
|
| 117 |
+
"eval_loss": 0.12946533443676894,
|
| 118 |
+
"eval_runtime": 177.0345,
|
| 119 |
+
"eval_samples_per_second": 22.594,
|
| 120 |
+
"eval_steps_per_second": 1.412,
|
| 121 |
+
"step": 70
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.008,
|
| 125 |
+
"grad_norm": 0.10741184651851654,
|
| 126 |
+
"learning_rate": 1.7049711594019046e-05,
|
| 127 |
+
"loss": 0.1285,
|
| 128 |
+
"step": 80
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.008,
|
| 132 |
+
"eval_cos_sim": 0.8719983696937561,
|
| 133 |
+
"eval_loss": 0.1289418597434706,
|
| 134 |
+
"eval_runtime": 178.6566,
|
| 135 |
+
"eval_samples_per_second": 22.389,
|
| 136 |
+
"eval_steps_per_second": 1.399,
|
| 137 |
+
"step": 80
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.009,
|
| 141 |
+
"grad_norm": 0.12072350829839706,
|
| 142 |
+
"learning_rate": 4.590606964640023e-06,
|
| 143 |
+
"loss": 0.125,
|
| 144 |
+
"step": 90
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.009,
|
| 148 |
+
"eval_cos_sim": 0.8721248507499695,
|
| 149 |
+
"eval_loss": 0.12881728055226274,
|
| 150 |
+
"eval_runtime": 181.5969,
|
| 151 |
+
"eval_samples_per_second": 22.027,
|
| 152 |
+
"eval_steps_per_second": 1.377,
|
| 153 |
+
"step": 90
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.01,
|
| 157 |
+
"grad_norm": 0.11123672872781754,
|
| 158 |
+
"learning_rate": 4.999688473794144e-05,
|
| 159 |
+
"loss": 0.1249,
|
| 160 |
+
"step": 100
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.01,
|
| 164 |
+
"eval_cos_sim": 0.8721336722373962,
|
| 165 |
+
"eval_loss": 0.12880885388600297,
|
| 166 |
+
"eval_runtime": 174.6097,
|
| 167 |
+
"eval_samples_per_second": 22.908,
|
| 168 |
+
"eval_steps_per_second": 1.432,
|
| 169 |
+
"step": 100
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.011,
|
| 173 |
+
"grad_norm": 0.11100038141012192,
|
| 174 |
+
"learning_rate": 4.494343314093799e-05,
|
| 175 |
+
"loss": 0.1246,
|
| 176 |
+
"step": 110
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.011,
|
| 180 |
+
"eval_cos_sim": 0.8723854422569275,
|
| 181 |
+
"eval_loss": 0.1285583892081923,
|
| 182 |
+
"eval_runtime": 180.7772,
|
| 183 |
+
"eval_samples_per_second": 22.127,
|
| 184 |
+
"eval_steps_per_second": 1.383,
|
| 185 |
+
"step": 110
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.012,
|
| 189 |
+
"grad_norm": 0.11933281272649765,
|
| 190 |
+
"learning_rate": 3.219808272827916e-05,
|
| 191 |
+
"loss": 0.1265,
|
| 192 |
+
"step": 120
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.012,
|
| 196 |
+
"eval_cos_sim": 0.8727645874023438,
|
| 197 |
+
"eval_loss": 0.12819017722355788,
|
| 198 |
+
"eval_runtime": 176.8881,
|
| 199 |
+
"eval_samples_per_second": 22.613,
|
| 200 |
+
"eval_steps_per_second": 1.413,
|
| 201 |
+
"step": 120
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.013,
|
| 205 |
+
"grad_norm": 0.11295568197965622,
|
| 206 |
+
"learning_rate": 1.667653407425597e-05,
|
| 207 |
+
"loss": 0.1256,
|
| 208 |
+
"step": 130
|
| 209 |
+
},
|
| 210 |
+
{
|
| 211 |
+
"epoch": 0.013,
|
| 212 |
+
"eval_cos_sim": 0.8724489808082581,
|
| 213 |
+
"eval_loss": 0.12850400116192764,
|
| 214 |
+
"eval_runtime": 176.2937,
|
| 215 |
+
"eval_samples_per_second": 22.689,
|
| 216 |
+
"eval_steps_per_second": 1.418,
|
| 217 |
+
"step": 130
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.014,
|
| 221 |
+
"grad_norm": 0.10013717412948608,
|
| 222 |
+
"learning_rate": 4.365227971950606e-06,
|
| 223 |
+
"loss": 0.1252,
|
| 224 |
+
"step": 140
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.014,
|
| 228 |
+
"eval_cos_sim": 0.8726389408111572,
|
| 229 |
+
"eval_loss": 0.1283098426078505,
|
| 230 |
+
"eval_runtime": 175.1837,
|
| 231 |
+
"eval_samples_per_second": 22.833,
|
| 232 |
+
"eval_steps_per_second": 1.427,
|
| 233 |
+
"step": 140
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.015,
|
| 237 |
+
"grad_norm": 0.08663387596607208,
|
| 238 |
+
"learning_rate": 4.998753972815435e-05,
|
| 239 |
+
"loss": 0.1252,
|
| 240 |
+
"step": 150
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.015,
|
| 244 |
+
"eval_cos_sim": 0.8726971745491028,
|
| 245 |
+
"eval_loss": 0.12825069954144425,
|
| 246 |
+
"eval_runtime": 179.297,
|
| 247 |
+
"eval_samples_per_second": 22.309,
|
| 248 |
+
"eval_steps_per_second": 1.394,
|
| 249 |
+
"step": 150
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.016,
|
| 253 |
+
"grad_norm": 0.10253303498029709,
|
| 254 |
+
"learning_rate": 4.47029683661798e-05,
|
| 255 |
+
"loss": 0.1258,
|
| 256 |
+
"step": 160
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"epoch": 0.016,
|
| 260 |
+
"eval_cos_sim": 0.8739002346992493,
|
| 261 |
+
"eval_loss": 0.12703985621678301,
|
| 262 |
+
"eval_runtime": 175.0922,
|
| 263 |
+
"eval_samples_per_second": 22.845,
|
| 264 |
+
"eval_steps_per_second": 1.428,
|
| 265 |
+
"step": 160
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.017,
|
| 269 |
+
"grad_norm": 0.11590978503227234,
|
| 270 |
+
"learning_rate": 3.1819242035765096e-05,
|
| 271 |
+
"loss": 0.1219,
|
| 272 |
+
"step": 170
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.017,
|
| 276 |
+
"eval_cos_sim": 0.8737954497337341,
|
| 277 |
+
"eval_loss": 0.12715704419362017,
|
| 278 |
+
"eval_runtime": 180.7326,
|
| 279 |
+
"eval_samples_per_second": 22.132,
|
| 280 |
+
"eval_steps_per_second": 1.383,
|
| 281 |
+
"step": 170
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.018,
|
| 285 |
+
"grad_norm": 0.09687651693820953,
|
| 286 |
+
"learning_rate": 1.6305430936700428e-05,
|
| 287 |
+
"loss": 0.1244,
|
| 288 |
+
"step": 180
|
| 289 |
+
},
|
| 290 |
+
{
|
| 291 |
+
"epoch": 0.018,
|
| 292 |
+
"eval_cos_sim": 0.8735443353652954,
|
| 293 |
+
"eval_loss": 0.12740902497517534,
|
| 294 |
+
"eval_runtime": 177.9084,
|
| 295 |
+
"eval_samples_per_second": 22.483,
|
| 296 |
+
"eval_steps_per_second": 1.405,
|
| 297 |
+
"step": 180
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.019,
|
| 301 |
+
"grad_norm": 0.10086172819137573,
|
| 302 |
+
"learning_rate": 4.144991597052059e-06,
|
| 303 |
+
"loss": 0.1258,
|
| 304 |
+
"step": 190
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.019,
|
| 308 |
+
"eval_cos_sim": 0.8735744953155518,
|
| 309 |
+
"eval_loss": 0.12737621738659807,
|
| 310 |
+
"eval_runtime": 174.0483,
|
| 311 |
+
"eval_samples_per_second": 22.982,
|
| 312 |
+
"eval_steps_per_second": 1.436,
|
| 313 |
+
"step": 190
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.02,
|
| 317 |
+
"grad_norm": 0.09316889941692352,
|
| 318 |
+
"learning_rate": 4.9971967299611097e-05,
|
| 319 |
+
"loss": 0.122,
|
| 320 |
+
"step": 200
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.02,
|
| 324 |
+
"eval_cos_sim": 0.8735851645469666,
|
| 325 |
+
"eval_loss": 0.12736523821103043,
|
| 326 |
+
"eval_runtime": 176.3327,
|
| 327 |
+
"eval_samples_per_second": 22.684,
|
| 328 |
+
"eval_steps_per_second": 1.418,
|
| 329 |
+
"step": 200
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.021,
|
| 333 |
+
"grad_norm": 0.10805534571409225,
|
| 334 |
+
"learning_rate": 4.4457593198638246e-05,
|
| 335 |
+
"loss": 0.1256,
|
| 336 |
+
"step": 210
|
| 337 |
+
},
|
| 338 |
+
{
|
| 339 |
+
"epoch": 0.021,
|
| 340 |
+
"eval_cos_sim": 0.8735992312431335,
|
| 341 |
+
"eval_loss": 0.12734888651120133,
|
| 342 |
+
"eval_runtime": 177.4342,
|
| 343 |
+
"eval_samples_per_second": 22.544,
|
| 344 |
+
"eval_steps_per_second": 1.409,
|
| 345 |
+
"step": 210
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.022,
|
| 349 |
+
"grad_norm": 0.14335550367832184,
|
| 350 |
+
"learning_rate": 3.143870184517241e-05,
|
| 351 |
+
"loss": 0.1228,
|
| 352 |
+
"step": 220
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.022,
|
| 356 |
+
"eval_cos_sim": 0.8742734789848328,
|
| 357 |
+
"eval_loss": 0.1266735837672896,
|
| 358 |
+
"eval_runtime": 174.698,
|
| 359 |
+
"eval_samples_per_second": 22.897,
|
| 360 |
+
"eval_steps_per_second": 1.431,
|
| 361 |
+
"step": 220
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.023,
|
| 365 |
+
"grad_norm": 0.10455214232206345,
|
| 366 |
+
"learning_rate": 1.5936494668034417e-05,
|
| 367 |
+
"loss": 0.1235,
|
| 368 |
+
"step": 230
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.023,
|
| 372 |
+
"eval_cos_sim": 0.874700129032135,
|
| 373 |
+
"eval_loss": 0.12624898936497636,
|
| 374 |
+
"eval_runtime": 175.2174,
|
| 375 |
+
"eval_samples_per_second": 22.829,
|
| 376 |
+
"eval_steps_per_second": 1.427,
|
| 377 |
+
"step": 230
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.024,
|
| 381 |
+
"grad_norm": 0.10344243049621582,
|
| 382 |
+
"learning_rate": 3.9299527274662355e-06,
|
| 383 |
+
"loss": 0.1258,
|
| 384 |
+
"step": 240
|
| 385 |
+
},
|
| 386 |
+
{
|
| 387 |
+
"epoch": 0.024,
|
| 388 |
+
"eval_cos_sim": 0.8746932148933411,
|
| 389 |
+
"eval_loss": 0.1262588949416823,
|
| 390 |
+
"eval_runtime": 178.5496,
|
| 391 |
+
"eval_samples_per_second": 22.403,
|
| 392 |
+
"eval_steps_per_second": 1.4,
|
| 393 |
+
"step": 240
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.025,
|
| 397 |
+
"grad_norm": 0.1515665352344513,
|
| 398 |
+
"learning_rate": 4.9950171333287335e-05,
|
| 399 |
+
"loss": 0.1259,
|
| 400 |
+
"step": 250
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 0.025,
|
| 404 |
+
"eval_cos_sim": 0.8746062517166138,
|
| 405 |
+
"eval_loss": 0.1263456218455977,
|
| 406 |
+
"eval_runtime": 181.2208,
|
| 407 |
+
"eval_samples_per_second": 22.073,
|
| 408 |
+
"eval_steps_per_second": 1.38,
|
| 409 |
+
"step": 250
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.026,
|
| 413 |
+
"grad_norm": 0.08521851152181625,
|
| 414 |
+
"learning_rate": 4.420736879094929e-05,
|
| 415 |
+
"loss": 0.123,
|
| 416 |
+
"step": 260
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.026,
|
| 420 |
+
"eval_cos_sim": 0.8742081522941589,
|
| 421 |
+
"eval_loss": 0.1267440173839278,
|
| 422 |
+
"eval_runtime": 172.3377,
|
| 423 |
+
"eval_samples_per_second": 23.21,
|
| 424 |
+
"eval_steps_per_second": 1.451,
|
| 425 |
+
"step": 260
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.027,
|
| 429 |
+
"grad_norm": 0.24638278782367706,
|
| 430 |
+
"learning_rate": 3.105655699509455e-05,
|
| 431 |
+
"loss": 0.1246,
|
| 432 |
+
"step": 270
|
| 433 |
+
},
|
| 434 |
+
{
|
| 435 |
+
"epoch": 0.027,
|
| 436 |
+
"eval_cos_sim": 0.8748664259910583,
|
| 437 |
+
"eval_loss": 0.12609003236042926,
|
| 438 |
+
"eval_runtime": 175.6344,
|
| 439 |
+
"eval_samples_per_second": 22.775,
|
| 440 |
+
"eval_steps_per_second": 1.423,
|
| 441 |
+
"step": 270
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.028,
|
| 445 |
+
"grad_norm": 0.09267835319042206,
|
| 446 |
+
"learning_rate": 1.5569817214910634e-05,
|
| 447 |
+
"loss": 0.1246,
|
| 448 |
+
"step": 280
|
| 449 |
+
},
|
| 450 |
+
{
|
| 451 |
+
"epoch": 0.028,
|
| 452 |
+
"eval_cos_sim": 0.8748399615287781,
|
| 453 |
+
"eval_loss": 0.12611397721516557,
|
| 454 |
+
"eval_runtime": 175.9072,
|
| 455 |
+
"eval_samples_per_second": 22.739,
|
| 456 |
+
"eval_steps_per_second": 1.421,
|
| 457 |
+
"step": 280
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.029,
|
| 461 |
+
"grad_norm": 0.1712462306022644,
|
| 462 |
+
"learning_rate": 3.720164955387656e-06,
|
| 463 |
+
"loss": 0.1243,
|
| 464 |
+
"step": 290
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.029,
|
| 468 |
+
"eval_cos_sim": 0.8749127388000488,
|
| 469 |
+
"eval_loss": 0.1260433347438521,
|
| 470 |
+
"eval_runtime": 176.0561,
|
| 471 |
+
"eval_samples_per_second": 22.72,
|
| 472 |
+
"eval_steps_per_second": 1.42,
|
| 473 |
+
"step": 290
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.03,
|
| 477 |
+
"grad_norm": 0.07719286531209946,
|
| 478 |
+
"learning_rate": 4.992215726119483e-05,
|
| 479 |
+
"loss": 0.1227,
|
| 480 |
+
"step": 300
|
| 481 |
+
},
|
| 482 |
+
{
|
| 483 |
+
"epoch": 0.03,
|
| 484 |
+
"eval_cos_sim": 0.8748821020126343,
|
| 485 |
+
"eval_loss": 0.1260761695121474,
|
| 486 |
+
"eval_runtime": 174.2263,
|
| 487 |
+
"eval_samples_per_second": 22.959,
|
| 488 |
+
"eval_steps_per_second": 1.435,
|
| 489 |
+
"step": 300
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.031,
|
| 493 |
+
"grad_norm": 0.08637545257806778,
|
| 494 |
+
"learning_rate": 4.395235750428112e-05,
|
| 495 |
+
"loss": 0.1222,
|
| 496 |
+
"step": 310
|
| 497 |
+
},
|
| 498 |
+
{
|
| 499 |
+
"epoch": 0.031,
|
| 500 |
+
"eval_cos_sim": 0.8745994567871094,
|
| 501 |
+
"eval_loss": 0.12635979654538104,
|
| 502 |
+
"eval_runtime": 179.4806,
|
| 503 |
+
"eval_samples_per_second": 22.287,
|
| 504 |
+
"eval_steps_per_second": 1.393,
|
| 505 |
+
"step": 310
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.032,
|
| 509 |
+
"grad_norm": 0.0923767164349556,
|
| 510 |
+
"learning_rate": 3.0672902724039794e-05,
|
| 511 |
+
"loss": 0.1232,
|
| 512 |
+
"step": 320
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.032,
|
| 516 |
+
"eval_cos_sim": 0.8750612735748291,
|
| 517 |
+
"eval_loss": 0.1258947375034041,
|
| 518 |
+
"eval_runtime": 181.1338,
|
| 519 |
+
"eval_samples_per_second": 22.083,
|
| 520 |
+
"eval_steps_per_second": 1.38,
|
| 521 |
+
"step": 320
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.033,
|
| 525 |
+
"grad_norm": 0.08724959194660187,
|
| 526 |
+
"learning_rate": 1.5205489961037645e-05,
|
| 527 |
+
"loss": 0.1236,
|
| 528 |
+
"step": 330
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.033,
|
| 532 |
+
"eval_cos_sim": 0.8755974173545837,
|
| 533 |
+
"eval_loss": 0.125363212845201,
|
| 534 |
+
"eval_runtime": 198.751,
|
| 535 |
+
"eval_samples_per_second": 20.126,
|
| 536 |
+
"eval_steps_per_second": 1.258,
|
| 537 |
+
"step": 330
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.034,
|
| 541 |
+
"grad_norm": 0.07283046841621399,
|
| 542 |
+
"learning_rate": 3.5156805643271896e-06,
|
| 543 |
+
"loss": 0.1239,
|
| 544 |
+
"step": 340
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.034,
|
| 548 |
+
"eval_cos_sim": 0.8756656646728516,
|
| 549 |
+
"eval_loss": 0.12529714014279317,
|
| 550 |
+
"eval_runtime": 187.9639,
|
| 551 |
+
"eval_samples_per_second": 21.281,
|
| 552 |
+
"eval_steps_per_second": 1.33,
|
| 553 |
+
"step": 340
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 0.035,
|
| 557 |
+
"grad_norm": 0.15486685931682587,
|
| 558 |
+
"learning_rate": 4.9887932065027656e-05,
|
| 559 |
+
"loss": 0.1231,
|
| 560 |
+
"step": 350
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.035,
|
| 564 |
+
"eval_cos_sim": 0.8756564259529114,
|
| 565 |
+
"eval_loss": 0.12530613209950398,
|
| 566 |
+
"eval_runtime": 194.2503,
|
| 567 |
+
"eval_samples_per_second": 20.592,
|
| 568 |
+
"eval_steps_per_second": 1.287,
|
| 569 |
+
"step": 350
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.036,
|
| 573 |
+
"grad_norm": 0.07505682110786438,
|
| 574 |
+
"learning_rate": 4.369262289279271e-05,
|
| 575 |
+
"loss": 0.1233,
|
| 576 |
+
"step": 360
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.036,
|
| 580 |
+
"eval_cos_sim": 0.8755001425743103,
|
| 581 |
+
"eval_loss": 0.12546515204655598,
|
| 582 |
+
"eval_runtime": 194.8309,
|
| 583 |
+
"eval_samples_per_second": 20.531,
|
| 584 |
+
"eval_steps_per_second": 1.283,
|
| 585 |
+
"step": 360
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.037,
|
| 589 |
+
"grad_norm": 0.09688587486743927,
|
| 590 |
+
"learning_rate": 3.0287834646695457e-05,
|
| 591 |
+
"loss": 0.1259,
|
| 592 |
+
"step": 370
|
| 593 |
+
},
|
| 594 |
+
{
|
| 595 |
+
"epoch": 0.037,
|
| 596 |
+
"eval_cos_sim": 0.8756394386291504,
|
| 597 |
+
"eval_loss": 0.1253258285735793,
|
| 598 |
+
"eval_runtime": 188.2216,
|
| 599 |
+
"eval_samples_per_second": 21.252,
|
| 600 |
+
"eval_steps_per_second": 1.328,
|
| 601 |
+
"step": 370
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 0.038,
|
| 605 |
+
"grad_norm": 0.07268425822257996,
|
| 606 |
+
"learning_rate": 1.4843603704405253e-05,
|
| 607 |
+
"loss": 0.1247,
|
| 608 |
+
"step": 380
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.038,
|
| 612 |
+
"eval_cos_sim": 0.8758111596107483,
|
| 613 |
+
"eval_loss": 0.12515661337124775,
|
| 614 |
+
"eval_runtime": 189.0095,
|
| 615 |
+
"eval_samples_per_second": 21.163,
|
| 616 |
+
"eval_steps_per_second": 1.323,
|
| 617 |
+
"step": 380
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.039,
|
| 621 |
+
"grad_norm": 0.09875091165304184,
|
| 622 |
+
"learning_rate": 3.316550516082126e-06,
|
| 623 |
+
"loss": 0.1229,
|
| 624 |
+
"step": 390
|
| 625 |
+
},
|
| 626 |
+
{
|
| 627 |
+
"epoch": 0.039,
|
| 628 |
+
"eval_cos_sim": 0.8758672475814819,
|
| 629 |
+
"eval_loss": 0.12509912636029194,
|
| 630 |
+
"eval_runtime": 235.6105,
|
| 631 |
+
"eval_samples_per_second": 16.977,
|
| 632 |
+
"eval_steps_per_second": 1.061,
|
| 633 |
+
"step": 390
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.04,
|
| 637 |
+
"grad_norm": 0.0792056992650032,
|
| 638 |
+
"learning_rate": 4.98475042744222e-05,
|
| 639 |
+
"loss": 0.1246,
|
| 640 |
+
"step": 400
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.04,
|
| 644 |
+
"eval_cos_sim": 0.8759932518005371,
|
| 645 |
+
"eval_loss": 0.12497495915638873,
|
| 646 |
+
"eval_runtime": 200.3436,
|
| 647 |
+
"eval_samples_per_second": 19.966,
|
| 648 |
+
"eval_steps_per_second": 1.248,
|
| 649 |
+
"step": 400
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.041,
|
| 653 |
+
"grad_norm": 0.10644775629043579,
|
| 654 |
+
"learning_rate": 4.3428229687794505e-05,
|
| 655 |
+
"loss": 0.1224,
|
| 656 |
+
"step": 410
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.041,
|
| 660 |
+
"eval_cos_sim": 0.8761371374130249,
|
| 661 |
+
"eval_loss": 0.12483511426197956,
|
| 662 |
+
"eval_runtime": 197.5074,
|
| 663 |
+
"eval_samples_per_second": 20.252,
|
| 664 |
+
"eval_steps_per_second": 1.266,
|
| 665 |
+
"step": 410
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.042,
|
| 669 |
+
"grad_norm": 0.09292006492614746,
|
| 670 |
+
"learning_rate": 2.9901448730099503e-05,
|
| 671 |
+
"loss": 0.1239,
|
| 672 |
+
"step": 420
|
| 673 |
+
},
|
| 674 |
+
{
|
| 675 |
+
"epoch": 0.042,
|
| 676 |
+
"eval_cos_sim": 0.876413881778717,
|
| 677 |
+
"eval_loss": 0.12455732419239948,
|
| 678 |
+
"eval_runtime": 187.5784,
|
| 679 |
+
"eval_samples_per_second": 21.324,
|
| 680 |
+
"eval_steps_per_second": 1.333,
|
| 681 |
+
"step": 420
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.043,
|
| 685 |
+
"grad_norm": 0.08105887472629547,
|
| 686 |
+
"learning_rate": 1.448424863465538e-05,
|
| 687 |
+
"loss": 0.1231,
|
| 688 |
+
"step": 430
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.043,
|
| 692 |
+
"eval_cos_sim": 0.876311719417572,
|
| 693 |
+
"eval_loss": 0.12465796377407977,
|
| 694 |
+
"eval_runtime": 203.0598,
|
| 695 |
+
"eval_samples_per_second": 19.699,
|
| 696 |
+
"eval_steps_per_second": 1.231,
|
| 697 |
+
"step": 430
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.044,
|
| 701 |
+
"grad_norm": 0.15435349941253662,
|
| 702 |
+
"learning_rate": 3.1228244380351547e-06,
|
| 703 |
+
"loss": 0.1225,
|
| 704 |
+
"step": 440
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.044,
|
| 708 |
+
"eval_cos_sim": 0.8762248754501343,
|
| 709 |
+
"eval_loss": 0.12474570634114215,
|
| 710 |
+
"eval_runtime": 199.1025,
|
| 711 |
+
"eval_samples_per_second": 20.09,
|
| 712 |
+
"eval_steps_per_second": 1.256,
|
| 713 |
+
"step": 440
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.045,
|
| 717 |
+
"grad_norm": 0.09370752424001694,
|
| 718 |
+
"learning_rate": 4.980088396483146e-05,
|
| 719 |
+
"loss": 0.1228,
|
| 720 |
+
"step": 450
|
| 721 |
+
},
|
| 722 |
+
{
|
| 723 |
+
"epoch": 0.045,
|
| 724 |
+
"eval_cos_sim": 0.8761196136474609,
|
| 725 |
+
"eval_loss": 0.12484796597706745,
|
| 726 |
+
"eval_runtime": 192.1246,
|
| 727 |
+
"eval_samples_per_second": 20.82,
|
| 728 |
+
"eval_steps_per_second": 1.301,
|
| 729 |
+
"step": 450
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.046,
|
| 733 |
+
"grad_norm": 0.08999752253293991,
|
| 734 |
+
"learning_rate": 4.3159243781616026e-05,
|
| 735 |
+
"loss": 0.1229,
|
| 736 |
+
"step": 460
|
| 737 |
+
},
|
| 738 |
+
{
|
| 739 |
+
"epoch": 0.046,
|
| 740 |
+
"eval_cos_sim": 0.8762247562408447,
|
| 741 |
+
"eval_loss": 0.12473729922520588,
|
| 742 |
+
"eval_runtime": 196.5532,
|
| 743 |
+
"eval_samples_per_second": 20.351,
|
| 744 |
+
"eval_steps_per_second": 1.272,
|
| 745 |
+
"step": 460
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.047,
|
| 749 |
+
"grad_norm": 0.0809365063905716,
|
| 750 |
+
"learning_rate": 2.9513841269722613e-05,
|
| 751 |
+
"loss": 0.124,
|
| 752 |
+
"step": 470
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.047,
|
| 756 |
+
"eval_cos_sim": 0.8765152096748352,
|
| 757 |
+
"eval_loss": 0.12444968440281817,
|
| 758 |
+
"eval_runtime": 204.1545,
|
| 759 |
+
"eval_samples_per_second": 19.593,
|
| 760 |
+
"eval_steps_per_second": 1.225,
|
| 761 |
+
"step": 470
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.048,
|
| 765 |
+
"grad_norm": 0.08176057785749435,
|
| 766 |
+
"learning_rate": 1.4127514310605238e-05,
|
| 767 |
+
"loss": 0.123,
|
| 768 |
+
"step": 480
|
| 769 |
+
},
|
| 770 |
+
{
|
| 771 |
+
"epoch": 0.048,
|
| 772 |
+
"eval_cos_sim": 0.876448929309845,
|
| 773 |
+
"eval_loss": 0.12451095607029865,
|
| 774 |
+
"eval_runtime": 198.7286,
|
| 775 |
+
"eval_samples_per_second": 20.128,
|
| 776 |
+
"eval_steps_per_second": 1.258,
|
| 777 |
+
"step": 480
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.049,
|
| 781 |
+
"grad_norm": 0.09636738151311874,
|
| 782 |
+
"learning_rate": 2.934550610786291e-06,
|
| 783 |
+
"loss": 0.1236,
|
| 784 |
+
"step": 490
|
| 785 |
+
},
|
| 786 |
+
{
|
| 787 |
+
"epoch": 0.049,
|
| 788 |
+
"eval_cos_sim": 0.8765274882316589,
|
| 789 |
+
"eval_loss": 0.12443248560177753,
|
| 790 |
+
"eval_runtime": 196.3413,
|
| 791 |
+
"eval_samples_per_second": 20.373,
|
| 792 |
+
"eval_steps_per_second": 1.273,
|
| 793 |
+
"step": 490
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.05,
|
| 797 |
+
"grad_norm": 0.08814109116792679,
|
| 798 |
+
"learning_rate": 4.974808275501392e-05,
|
| 799 |
+
"loss": 0.123,
|
| 800 |
+
"step": 500
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.05,
|
| 804 |
+
"eval_cos_sim": 0.8765753507614136,
|
| 805 |
+
"eval_loss": 0.12438686539875934,
|
| 806 |
+
"eval_runtime": 191.2687,
|
| 807 |
+
"eval_samples_per_second": 20.913,
|
| 808 |
+
"eval_steps_per_second": 1.307,
|
| 809 |
+
"step": 500
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.051,
|
| 813 |
+
"grad_norm": 0.08511923253536224,
|
| 814 |
+
"learning_rate": 4.2885732211184324e-05,
|
| 815 |
+
"loss": 0.1246,
|
| 816 |
+
"step": 510
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.051,
|
| 820 |
+
"eval_cos_sim": 0.8767162561416626,
|
| 821 |
+
"eval_loss": 0.12425224568592975,
|
| 822 |
+
"eval_runtime": 173.2088,
|
| 823 |
+
"eval_samples_per_second": 23.094,
|
| 824 |
+
"eval_steps_per_second": 1.443,
|
| 825 |
+
"step": 510
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.052,
|
| 829 |
+
"grad_norm": 0.0837215781211853,
|
| 830 |
+
"learning_rate": 2.9125108865470048e-05,
|
| 831 |
+
"loss": 0.1221,
|
| 832 |
+
"step": 520
|
| 833 |
+
},
|
| 834 |
+
{
|
| 835 |
+
"epoch": 0.052,
|
| 836 |
+
"eval_cos_sim": 0.876861572265625,
|
| 837 |
+
"eval_loss": 0.1241044213985152,
|
| 838 |
+
"eval_runtime": 174.8239,
|
| 839 |
+
"eval_samples_per_second": 22.88,
|
| 840 |
+
"eval_steps_per_second": 1.43,
|
| 841 |
+
"step": 520
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 0.053,
|
| 845 |
+
"grad_norm": 0.09207245707511902,
|
| 846 |
+
"learning_rate": 1.3773489637927061e-05,
|
| 847 |
+
"loss": 0.1229,
|
| 848 |
+
"step": 530
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.053,
|
| 852 |
+
"eval_cos_sim": 0.8767414093017578,
|
| 853 |
+
"eval_loss": 0.12421691825138996,
|
| 854 |
+
"eval_runtime": 173.8268,
|
| 855 |
+
"eval_samples_per_second": 23.011,
|
| 856 |
+
"eval_steps_per_second": 1.438,
|
| 857 |
+
"step": 530
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.054,
|
| 861 |
+
"grad_norm": 0.0655718669295311,
|
| 862 |
+
"learning_rate": 2.7517759561205253e-06,
|
| 863 |
+
"loss": 0.1221,
|
| 864 |
+
"step": 540
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 0.054,
|
| 868 |
+
"eval_cos_sim": 0.8767919540405273,
|
| 869 |
+
"eval_loss": 0.1241676082824416,
|
| 870 |
+
"eval_runtime": 179.6327,
|
| 871 |
+
"eval_samples_per_second": 22.268,
|
| 872 |
+
"eval_steps_per_second": 1.392,
|
| 873 |
+
"step": 540
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.055,
|
| 877 |
+
"grad_norm": 0.21964910626411438,
|
| 878 |
+
"learning_rate": 4.968911380413809e-05,
|
| 879 |
+
"loss": 0.1243,
|
| 880 |
+
"step": 550
|
| 881 |
+
},
|
| 882 |
+
{
|
| 883 |
+
"epoch": 0.055,
|
| 884 |
+
"eval_cos_sim": 0.8768623471260071,
|
| 885 |
+
"eval_loss": 0.12409912397610615,
|
| 886 |
+
"eval_runtime": 172.7843,
|
| 887 |
+
"eval_samples_per_second": 23.15,
|
| 888 |
+
"eval_steps_per_second": 1.447,
|
| 889 |
+
"step": 550
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.056,
|
| 893 |
+
"grad_norm": 0.08817338943481445,
|
| 894 |
+
"learning_rate": 4.260776314131676e-05,
|
| 895 |
+
"loss": 0.1222,
|
| 896 |
+
"step": 560
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.056,
|
| 900 |
+
"eval_cos_sim": 0.8767062425613403,
|
| 901 |
+
"eval_loss": 0.12425821544873188,
|
| 902 |
+
"eval_runtime": 172.6396,
|
| 903 |
+
"eval_samples_per_second": 23.17,
|
| 904 |
+
"eval_steps_per_second": 1.448,
|
| 905 |
+
"step": 560
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.057,
|
| 909 |
+
"grad_norm": 0.06475117802619934,
|
| 910 |
+
"learning_rate": 2.873534839760646e-05,
|
| 911 |
+
"loss": 0.1232,
|
| 912 |
+
"step": 570
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.057,
|
| 916 |
+
"eval_cos_sim": 0.8768667578697205,
|
| 917 |
+
"eval_loss": 0.12410461117970416,
|
| 918 |
+
"eval_runtime": 172.7054,
|
| 919 |
+
"eval_samples_per_second": 23.161,
|
| 920 |
+
"eval_steps_per_second": 1.448,
|
| 921 |
+
"step": 570
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.058,
|
| 925 |
+
"grad_norm": 0.07474437355995178,
|
| 926 |
+
"learning_rate": 1.342226284699138e-05,
|
| 927 |
+
"loss": 0.1227,
|
| 928 |
+
"step": 580
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 0.058,
|
| 932 |
+
"eval_cos_sim": 0.8771414160728455,
|
| 933 |
+
"eval_loss": 0.12382852866398761,
|
| 934 |
+
"eval_runtime": 175.1422,
|
| 935 |
+
"eval_samples_per_second": 22.839,
|
| 936 |
+
"eval_steps_per_second": 1.427,
|
| 937 |
+
"step": 580
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 0.059,
|
| 941 |
+
"grad_norm": 0.07362603396177292,
|
| 942 |
+
"learning_rate": 2.5745460253134484e-06,
|
| 943 |
+
"loss": 0.1234,
|
| 944 |
+
"step": 590
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.059,
|
| 948 |
+
"eval_cos_sim": 0.8771759271621704,
|
| 949 |
+
"eval_loss": 0.12379106380688618,
|
| 950 |
+
"eval_runtime": 174.7169,
|
| 951 |
+
"eval_samples_per_second": 22.894,
|
| 952 |
+
"eval_steps_per_second": 1.431,
|
| 953 |
+
"step": 590
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.06,
|
| 957 |
+
"grad_norm": 0.07593993842601776,
|
| 958 |
+
"learning_rate": 4.962399180850275e-05,
|
| 959 |
+
"loss": 0.1232,
|
| 960 |
+
"step": 600
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 0.06,
|
| 964 |
+
"eval_cos_sim": 0.877038300037384,
|
| 965 |
+
"eval_loss": 0.12392904116856525,
|
| 966 |
+
"eval_runtime": 172.4786,
|
| 967 |
+
"eval_samples_per_second": 23.191,
|
| 968 |
+
"eval_steps_per_second": 1.449,
|
| 969 |
+
"step": 600
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.061,
|
| 973 |
+
"grad_norm": 0.07887241989374161,
|
| 974 |
+
"learning_rate": 4.2325405847733254e-05,
|
| 975 |
+
"loss": 0.1235,
|
| 976 |
+
"step": 610
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 0.061,
|
| 980 |
+
"eval_cos_sim": 0.8767529726028442,
|
| 981 |
+
"eval_loss": 0.12422390153157184,
|
| 982 |
+
"eval_runtime": 173.6696,
|
| 983 |
+
"eval_samples_per_second": 23.032,
|
| 984 |
+
"eval_steps_per_second": 1.44,
|
| 985 |
+
"step": 610
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.062,
|
| 989 |
+
"grad_norm": 0.17296281456947327,
|
| 990 |
+
"learning_rate": 2.834465700261192e-05,
|
| 991 |
+
"loss": 0.1204,
|
| 992 |
+
"step": 620
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.062,
|
| 996 |
+
"eval_cos_sim": 0.8772019743919373,
|
| 997 |
+
"eval_loss": 0.12377139737355183,
|
| 998 |
+
"eval_runtime": 179.9864,
|
| 999 |
+
"eval_samples_per_second": 22.224,
|
| 1000 |
+
"eval_steps_per_second": 1.389,
|
| 1001 |
+
"step": 620
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.063,
|
| 1005 |
+
"grad_norm": 0.06920995563268661,
|
| 1006 |
+
"learning_rate": 1.3073921470877709e-05,
|
| 1007 |
+
"loss": 0.1245,
|
| 1008 |
+
"step": 630
|
| 1009 |
+
},
|
| 1010 |
+
{
|
| 1011 |
+
"epoch": 0.063,
|
| 1012 |
+
"eval_cos_sim": 0.8773365616798401,
|
| 1013 |
+
"eval_loss": 0.12363236000287008,
|
| 1014 |
+
"eval_runtime": 173.1204,
|
| 1015 |
+
"eval_samples_per_second": 23.105,
|
| 1016 |
+
"eval_steps_per_second": 1.444,
|
| 1017 |
+
"step": 630
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.064,
|
| 1021 |
+
"grad_norm": 0.08347232639789581,
|
| 1022 |
+
"learning_rate": 2.4029049877794472e-06,
|
| 1023 |
+
"loss": 0.1217,
|
| 1024 |
+
"step": 640
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.064,
|
| 1028 |
+
"eval_cos_sim": 0.8773410320281982,
|
| 1029 |
+
"eval_loss": 0.12362796523320149,
|
| 1030 |
+
"eval_runtime": 172.0713,
|
| 1031 |
+
"eval_samples_per_second": 23.246,
|
| 1032 |
+
"eval_steps_per_second": 1.453,
|
| 1033 |
+
"step": 640
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 0.065,
|
| 1037 |
+
"grad_norm": 0.07459770888090134,
|
| 1038 |
+
"learning_rate": 4.955273299787453e-05,
|
| 1039 |
+
"loss": 0.1223,
|
| 1040 |
+
"step": 650
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.065,
|
| 1044 |
+
"eval_cos_sim": 0.8773767948150635,
|
| 1045 |
+
"eval_loss": 0.12359384205090472,
|
| 1046 |
+
"eval_runtime": 173.2149,
|
| 1047 |
+
"eval_samples_per_second": 23.093,
|
| 1048 |
+
"eval_steps_per_second": 1.443,
|
| 1049 |
+
"step": 650
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 0.066,
|
| 1053 |
+
"grad_norm": 0.0831998735666275,
|
| 1054 |
+
"learning_rate": 4.203873069979081e-05,
|
| 1055 |
+
"loss": 0.1231,
|
| 1056 |
+
"step": 660
|
| 1057 |
+
},
|
| 1058 |
+
{
|
| 1059 |
+
"epoch": 0.066,
|
| 1060 |
+
"eval_cos_sim": 0.8774532675743103,
|
| 1061 |
+
"eval_loss": 0.12351777221905659,
|
| 1062 |
+
"eval_runtime": 171.902,
|
| 1063 |
+
"eval_samples_per_second": 23.269,
|
| 1064 |
+
"eval_steps_per_second": 1.454,
|
| 1065 |
+
"step": 660
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.067,
|
| 1069 |
+
"grad_norm": 0.07724840193986893,
|
| 1070 |
+
"learning_rate": 2.7953132048972646e-05,
|
| 1071 |
+
"loss": 0.122,
|
| 1072 |
+
"step": 670
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"epoch": 0.067,
|
| 1076 |
+
"eval_cos_sim": 0.877151608467102,
|
| 1077 |
+
"eval_loss": 0.12382214214550921,
|
| 1078 |
+
"eval_runtime": 173.6766,
|
| 1079 |
+
"eval_samples_per_second": 23.031,
|
| 1080 |
+
"eval_steps_per_second": 1.439,
|
| 1081 |
+
"step": 670
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.068,
|
| 1085 |
+
"grad_norm": 0.0648268312215805,
|
| 1086 |
+
"learning_rate": 1.2728552323560239e-05,
|
| 1087 |
+
"loss": 0.1227,
|
| 1088 |
+
"step": 680
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.068,
|
| 1092 |
+
"eval_cos_sim": 0.8769506216049194,
|
| 1093 |
+
"eval_loss": 0.12402295615422199,
|
| 1094 |
+
"eval_runtime": 171.7424,
|
| 1095 |
+
"eval_samples_per_second": 23.291,
|
| 1096 |
+
"eval_steps_per_second": 1.456,
|
| 1097 |
+
"step": 680
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 0.069,
|
| 1101 |
+
"grad_norm": 0.08475865423679352,
|
| 1102 |
+
"learning_rate": 2.2368956200634283e-06,
|
| 1103 |
+
"loss": 0.1274,
|
| 1104 |
+
"step": 690
|
| 1105 |
+
},
|
| 1106 |
+
{
|
| 1107 |
+
"epoch": 0.069,
|
| 1108 |
+
"eval_cos_sim": 0.8771329522132874,
|
| 1109 |
+
"eval_loss": 0.12383969738232563,
|
| 1110 |
+
"eval_runtime": 174.2776,
|
| 1111 |
+
"eval_samples_per_second": 22.952,
|
| 1112 |
+
"eval_steps_per_second": 1.434,
|
| 1113 |
+
"step": 690
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 0.07,
|
| 1117 |
+
"grad_norm": 0.06382860988378525,
|
| 1118 |
+
"learning_rate": 4.947535513144286e-05,
|
| 1119 |
+
"loss": 0.122,
|
| 1120 |
+
"step": 700
|
| 1121 |
+
},
|
| 1122 |
+
{
|
| 1123 |
+
"epoch": 0.07,
|
| 1124 |
+
"eval_cos_sim": 0.8775114417076111,
|
| 1125 |
+
"eval_loss": 0.12346241619336079,
|
| 1126 |
+
"eval_runtime": 185.1334,
|
| 1127 |
+
"eval_samples_per_second": 21.606,
|
| 1128 |
+
"eval_steps_per_second": 1.35,
|
| 1129 |
+
"step": 700
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.071,
|
| 1133 |
+
"grad_norm": 0.07273228466510773,
|
| 1134 |
+
"learning_rate": 4.174780914294635e-05,
|
| 1135 |
+
"loss": 0.1228,
|
| 1136 |
+
"step": 710
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.071,
|
| 1140 |
+
"eval_cos_sim": 0.8777372241020203,
|
| 1141 |
+
"eval_loss": 0.12323929693448019,
|
| 1142 |
+
"eval_runtime": 170.2151,
|
| 1143 |
+
"eval_samples_per_second": 23.5,
|
| 1144 |
+
"eval_steps_per_second": 1.469,
|
| 1145 |
+
"step": 710
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 0.072,
|
| 1149 |
+
"grad_norm": 0.08377543836832047,
|
| 1150 |
+
"learning_rate": 2.756087111291529e-05,
|
| 1151 |
+
"loss": 0.1209,
|
| 1152 |
+
"step": 720
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 0.072,
|
| 1156 |
+
"eval_cos_sim": 0.8776744604110718,
|
| 1157 |
+
"eval_loss": 0.12329552843319844,
|
| 1158 |
+
"eval_runtime": 173.1907,
|
| 1159 |
+
"eval_samples_per_second": 23.096,
|
| 1160 |
+
"eval_steps_per_second": 1.443,
|
| 1161 |
+
"step": 720
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.073,
|
| 1165 |
+
"grad_norm": 0.08579932153224945,
|
| 1166 |
+
"learning_rate": 1.2386241478270527e-05,
|
| 1167 |
+
"loss": 0.1234,
|
| 1168 |
+
"step": 730
|
| 1169 |
+
},
|
| 1170 |
+
{
|
| 1171 |
+
"epoch": 0.073,
|
| 1172 |
+
"eval_cos_sim": 0.8776343464851379,
|
| 1173 |
+
"eval_loss": 0.12333650018917988,
|
| 1174 |
+
"eval_runtime": 172.2784,
|
| 1175 |
+
"eval_samples_per_second": 23.218,
|
| 1176 |
+
"eval_steps_per_second": 1.451,
|
| 1177 |
+
"step": 730
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 0.074,
|
| 1181 |
+
"grad_norm": 0.07494545727968216,
|
| 1182 |
+
"learning_rate": 2.0765592951802664e-06,
|
| 1183 |
+
"loss": 0.1209,
|
| 1184 |
+
"step": 740
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 0.074,
|
| 1188 |
+
"eval_cos_sim": 0.8777279853820801,
|
| 1189 |
+
"eval_loss": 0.12324421884762715,
|
| 1190 |
+
"eval_runtime": 172.9417,
|
| 1191 |
+
"eval_samples_per_second": 23.129,
|
| 1192 |
+
"eval_steps_per_second": 1.446,
|
| 1193 |
+
"step": 740
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 0.075,
|
| 1197 |
+
"grad_norm": 0.07511463761329651,
|
| 1198 |
+
"learning_rate": 4.9391877493394335e-05,
|
| 1199 |
+
"loss": 0.1222,
|
| 1200 |
+
"step": 750
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 0.075,
|
| 1204 |
+
"eval_cos_sim": 0.8777404427528381,
|
| 1205 |
+
"eval_loss": 0.12323040797459553,
|
| 1206 |
+
"eval_runtime": 173.813,
|
| 1207 |
+
"eval_samples_per_second": 23.013,
|
| 1208 |
+
"eval_steps_per_second": 1.438,
|
| 1209 |
+
"step": 750
|
| 1210 |
+
},
|
| 1211 |
+
{
|
| 1212 |
+
"epoch": 0.076,
|
| 1213 |
+
"grad_norm": 0.08240217715501785,
|
| 1214 |
+
"learning_rate": 4.1452713680951016e-05,
|
| 1215 |
+
"loss": 0.1237,
|
| 1216 |
+
"step": 760
|
| 1217 |
+
},
|
| 1218 |
+
{
|
| 1219 |
+
"epoch": 0.076,
|
| 1220 |
+
"eval_cos_sim": 0.8776569366455078,
|
| 1221 |
+
"eval_loss": 0.1233164258216567,
|
| 1222 |
+
"eval_runtime": 173.6453,
|
| 1223 |
+
"eval_samples_per_second": 23.035,
|
| 1224 |
+
"eval_steps_per_second": 1.44,
|
| 1225 |
+
"step": 760
|
| 1226 |
+
},
|
| 1227 |
+
{
|
| 1228 |
+
"epoch": 0.077,
|
| 1229 |
+
"grad_norm": 0.07817904651165009,
|
| 1230 |
+
"learning_rate": 2.716797195408887e-05,
|
| 1231 |
+
"loss": 0.1215,
|
| 1232 |
+
"step": 770
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 0.077,
|
| 1236 |
+
"eval_cos_sim": 0.8779506683349609,
|
| 1237 |
+
"eval_loss": 0.12303087331997822,
|
| 1238 |
+
"eval_runtime": 198.4978,
|
| 1239 |
+
"eval_samples_per_second": 20.151,
|
| 1240 |
+
"eval_steps_per_second": 1.259,
|
| 1241 |
+
"step": 770
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 0.078,
|
| 1245 |
+
"grad_norm": 0.06472489982843399,
|
| 1246 |
+
"learning_rate": 1.2047074246048157e-05,
|
| 1247 |
+
"loss": 0.1222,
|
| 1248 |
+
"step": 780
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 0.078,
|
| 1252 |
+
"eval_cos_sim": 0.8780341148376465,
|
| 1253 |
+
"eval_loss": 0.12294723345982503,
|
| 1254 |
+
"eval_runtime": 187.0246,
|
| 1255 |
+
"eval_samples_per_second": 21.388,
|
| 1256 |
+
"eval_steps_per_second": 1.337,
|
| 1257 |
+
"step": 780
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 0.079,
|
| 1261 |
+
"grad_norm": 0.06511878967285156,
|
| 1262 |
+
"learning_rate": 1.921935972303521e-06,
|
| 1263 |
+
"loss": 0.1211,
|
| 1264 |
+
"step": 790
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 0.079,
|
| 1268 |
+
"eval_cos_sim": 0.8780234456062317,
|
| 1269 |
+
"eval_loss": 0.1229577579711623,
|
| 1270 |
+
"eval_runtime": 170.8199,
|
| 1271 |
+
"eval_samples_per_second": 23.416,
|
| 1272 |
+
"eval_steps_per_second": 1.464,
|
| 1273 |
+
"step": 790
|
| 1274 |
+
},
|
| 1275 |
+
{
|
| 1276 |
+
"epoch": 0.08,
|
| 1277 |
+
"grad_norm": 0.08275925368070602,
|
| 1278 |
+
"learning_rate": 4.9302320888106454e-05,
|
| 1279 |
+
"loss": 0.1234,
|
| 1280 |
+
"step": 800
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.08,
|
| 1284 |
+
"eval_cos_sim": 0.8778801560401917,
|
| 1285 |
+
"eval_loss": 0.1230986237739272,
|
| 1286 |
+
"eval_runtime": 175.6448,
|
| 1287 |
+
"eval_samples_per_second": 22.773,
|
| 1288 |
+
"eval_steps_per_second": 1.423,
|
| 1289 |
+
"step": 800
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"epoch": 0.081,
|
| 1293 |
+
"grad_norm": 0.06466321647167206,
|
| 1294 |
+
"learning_rate": 4.115351785778022e-05,
|
| 1295 |
+
"loss": 0.1215,
|
| 1296 |
+
"step": 810
|
| 1297 |
+
},
|
| 1298 |
+
{
|
| 1299 |
+
"epoch": 0.081,
|
| 1300 |
+
"eval_cos_sim": 0.877547025680542,
|
| 1301 |
+
"eval_loss": 0.12342484547841023,
|
| 1302 |
+
"eval_runtime": 173.845,
|
| 1303 |
+
"eval_samples_per_second": 23.009,
|
| 1304 |
+
"eval_steps_per_second": 1.438,
|
| 1305 |
+
"step": 810
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 0.082,
|
| 1309 |
+
"grad_norm": 0.060175709426403046,
|
| 1310 |
+
"learning_rate": 2.6774532491200373e-05,
|
| 1311 |
+
"loss": 0.1237,
|
| 1312 |
+
"step": 820
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 0.082,
|
| 1316 |
+
"eval_cos_sim": 0.8778981566429138,
|
| 1317 |
+
"eval_loss": 0.1230772545551009,
|
| 1318 |
+
"eval_runtime": 174.1784,
|
| 1319 |
+
"eval_samples_per_second": 22.965,
|
| 1320 |
+
"eval_steps_per_second": 1.435,
|
| 1321 |
+
"step": 820
|
| 1322 |
+
},
|
| 1323 |
+
{
|
| 1324 |
+
"epoch": 0.083,
|
| 1325 |
+
"grad_norm": 0.06948266923427582,
|
| 1326 |
+
"learning_rate": 1.1711135154477437e-05,
|
| 1327 |
+
"loss": 0.1213,
|
| 1328 |
+
"step": 830
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.083,
|
| 1332 |
+
"eval_cos_sim": 0.8779332041740417,
|
| 1333 |
+
"eval_loss": 0.12304716589199971,
|
| 1334 |
+
"eval_runtime": 171.7677,
|
| 1335 |
+
"eval_samples_per_second": 23.287,
|
| 1336 |
+
"eval_steps_per_second": 1.455,
|
| 1337 |
+
"step": 830
|
| 1338 |
+
},
|
| 1339 |
+
{
|
| 1340 |
+
"epoch": 0.084,
|
| 1341 |
+
"grad_norm": 0.0633857399225235,
|
| 1342 |
+
"learning_rate": 1.7730641868067276e-06,
|
| 1343 |
+
"loss": 0.1212,
|
| 1344 |
+
"step": 840
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"epoch": 0.084,
|
| 1348 |
+
"eval_cos_sim": 0.8779239058494568,
|
| 1349 |
+
"eval_loss": 0.12305730154263447,
|
| 1350 |
+
"eval_runtime": 172.6941,
|
| 1351 |
+
"eval_samples_per_second": 23.162,
|
| 1352 |
+
"eval_steps_per_second": 1.448,
|
| 1353 |
+
"step": 840
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 0.085,
|
| 1357 |
+
"grad_norm": 0.07013432681560516,
|
| 1358 |
+
"learning_rate": 4.9206707634962714e-05,
|
| 1359 |
+
"loss": 0.1219,
|
| 1360 |
+
"step": 850
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 0.085,
|
| 1364 |
+
"eval_cos_sim": 0.8781536221504211,
|
| 1365 |
+
"eval_loss": 0.12283129765736531,
|
| 1366 |
+
"eval_runtime": 178.3382,
|
| 1367 |
+
"eval_samples_per_second": 22.429,
|
| 1368 |
+
"eval_steps_per_second": 1.402,
|
| 1369 |
+
"step": 850
|
| 1370 |
+
},
|
| 1371 |
+
{
|
| 1372 |
+
"epoch": 0.086,
|
| 1373 |
+
"grad_norm": 0.0714387595653534,
|
| 1374 |
+
"learning_rate": 4.085029623930606e-05,
|
| 1375 |
+
"loss": 0.1214,
|
| 1376 |
+
"step": 860
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 0.086,
|
| 1380 |
+
"eval_cos_sim": 0.8783000111579895,
|
| 1381 |
+
"eval_loss": 0.12268445636975239,
|
| 1382 |
+
"eval_runtime": 180.4291,
|
| 1383 |
+
"eval_samples_per_second": 22.169,
|
| 1384 |
+
"eval_steps_per_second": 1.386,
|
| 1385 |
+
"step": 860
|
| 1386 |
+
},
|
| 1387 |
+
{
|
| 1388 |
+
"epoch": 0.087,
|
| 1389 |
+
"grad_norm": 0.07285313308238983,
|
| 1390 |
+
"learning_rate": 2.638065077761282e-05,
|
| 1391 |
+
"loss": 0.1211,
|
| 1392 |
+
"step": 870
|
| 1393 |
+
},
|
| 1394 |
+
{
|
| 1395 |
+
"epoch": 0.087,
|
| 1396 |
+
"eval_cos_sim": 0.8782742619514465,
|
| 1397 |
+
"eval_loss": 0.12271090867268514,
|
| 1398 |
+
"eval_runtime": 174.6757,
|
| 1399 |
+
"eval_samples_per_second": 22.9,
|
| 1400 |
+
"eval_steps_per_second": 1.431,
|
| 1401 |
+
"step": 870
|
| 1402 |
+
},
|
| 1403 |
+
{
|
| 1404 |
+
"epoch": 0.088,
|
| 1405 |
+
"grad_norm": 0.1114286258816719,
|
| 1406 |
+
"learning_rate": 1.1378507926623341e-05,
|
| 1407 |
+
"loss": 0.1203,
|
| 1408 |
+
"step": 880
|
| 1409 |
+
},
|
| 1410 |
+
{
|
| 1411 |
+
"epoch": 0.088,
|
| 1412 |
+
"eval_cos_sim": 0.8782421946525574,
|
| 1413 |
+
"eval_loss": 0.12274044944989156,
|
| 1414 |
+
"eval_runtime": 173.5126,
|
| 1415 |
+
"eval_samples_per_second": 23.053,
|
| 1416 |
+
"eval_steps_per_second": 1.441,
|
| 1417 |
+
"step": 880
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 0.089,
|
| 1421 |
+
"grad_norm": 0.07392691820859909,
|
| 1422 |
+
"learning_rate": 1.6299810406600836e-06,
|
| 1423 |
+
"loss": 0.1222,
|
| 1424 |
+
"step": 890
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.089,
|
| 1428 |
+
"eval_cos_sim": 0.8782600164413452,
|
| 1429 |
+
"eval_loss": 0.12272232272374105,
|
| 1430 |
+
"eval_runtime": 173.9745,
|
| 1431 |
+
"eval_samples_per_second": 22.992,
|
| 1432 |
+
"eval_steps_per_second": 1.437,
|
| 1433 |
+
"step": 890
|
| 1434 |
+
},
|
| 1435 |
+
{
|
| 1436 |
+
"epoch": 0.09,
|
| 1437 |
+
"grad_norm": 0.1509944051504135,
|
| 1438 |
+
"learning_rate": 4.9105061562790325e-05,
|
| 1439 |
+
"loss": 0.1211,
|
| 1440 |
+
"step": 900
|
| 1441 |
+
},
|
| 1442 |
+
{
|
| 1443 |
+
"epoch": 0.09,
|
| 1444 |
+
"eval_cos_sim": 0.8785330653190613,
|
| 1445 |
+
"eval_loss": 0.12244940116154622,
|
| 1446 |
+
"eval_runtime": 174.6529,
|
| 1447 |
+
"eval_samples_per_second": 22.903,
|
| 1448 |
+
"eval_steps_per_second": 1.431,
|
| 1449 |
+
"step": 900
|
| 1450 |
+
},
|
| 1451 |
+
{
|
| 1452 |
+
"epoch": 0.091,
|
| 1453 |
+
"grad_norm": 0.07572964578866959,
|
| 1454 |
+
"learning_rate": 4.0543124394712475e-05,
|
| 1455 |
+
"loss": 0.1234,
|
| 1456 |
+
"step": 910
|
| 1457 |
+
},
|
| 1458 |
+
{
|
| 1459 |
+
"epoch": 0.091,
|
| 1460 |
+
"eval_cos_sim": 0.8782286643981934,
|
| 1461 |
+
"eval_loss": 0.1227607171748824,
|
| 1462 |
+
"eval_runtime": 174.4786,
|
| 1463 |
+
"eval_samples_per_second": 22.925,
|
| 1464 |
+
"eval_steps_per_second": 1.433,
|
| 1465 |
+
"step": 910
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 0.092,
|
| 1469 |
+
"grad_norm": 0.07199128717184067,
|
| 1470 |
+
"learning_rate": 2.5986424976906166e-05,
|
| 1471 |
+
"loss": 0.1202,
|
| 1472 |
+
"step": 920
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.092,
|
| 1476 |
+
"eval_cos_sim": 0.8780964612960815,
|
| 1477 |
+
"eval_loss": 0.12288942649113606,
|
| 1478 |
+
"eval_runtime": 175.9134,
|
| 1479 |
+
"eval_samples_per_second": 22.738,
|
| 1480 |
+
"eval_steps_per_second": 1.421,
|
| 1481 |
+
"step": 920
|
| 1482 |
+
},
|
| 1483 |
+
{
|
| 1484 |
+
"epoch": 0.093,
|
| 1485 |
+
"grad_norm": 0.07497607171535492,
|
| 1486 |
+
"learning_rate": 1.1049275460163872e-05,
|
| 1487 |
+
"loss": 0.123,
|
| 1488 |
+
"step": 930
|
| 1489 |
+
},
|
| 1490 |
+
{
|
| 1491 |
+
"epoch": 0.093,
|
| 1492 |
+
"eval_cos_sim": 0.8781337141990662,
|
| 1493 |
+
"eval_loss": 0.12284465791928242,
|
| 1494 |
+
"eval_runtime": 174.1009,
|
| 1495 |
+
"eval_samples_per_second": 22.975,
|
| 1496 |
+
"eval_steps_per_second": 1.436,
|
| 1497 |
+
"step": 930
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 0.094,
|
| 1501 |
+
"grad_norm": 0.056581463664770126,
|
| 1502 |
+
"learning_rate": 1.4927221931830576e-06,
|
| 1503 |
+
"loss": 0.1218,
|
| 1504 |
+
"step": 940
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 0.094,
|
| 1508 |
+
"eval_cos_sim": 0.8781940340995789,
|
| 1509 |
+
"eval_loss": 0.12278383018719624,
|
| 1510 |
+
"eval_runtime": 180.3511,
|
| 1511 |
+
"eval_samples_per_second": 22.179,
|
| 1512 |
+
"eval_steps_per_second": 1.386,
|
| 1513 |
+
"step": 940
|
| 1514 |
+
},
|
| 1515 |
+
{
|
| 1516 |
+
"epoch": 0.095,
|
| 1517 |
+
"grad_norm": 0.06227719038724899,
|
| 1518 |
+
"learning_rate": 4.8997408003921384e-05,
|
| 1519 |
+
"loss": 0.1216,
|
| 1520 |
+
"step": 950
|
| 1521 |
+
},
|
| 1522 |
+
{
|
| 1523 |
+
"epoch": 0.095,
|
| 1524 |
+
"eval_cos_sim": 0.8782709836959839,
|
| 1525 |
+
"eval_loss": 0.12271020819889973,
|
| 1526 |
+
"eval_runtime": 174.3195,
|
| 1527 |
+
"eval_samples_per_second": 22.946,
|
| 1528 |
+
"eval_steps_per_second": 1.434,
|
| 1529 |
+
"step": 950
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 0.096,
|
| 1533 |
+
"grad_norm": 0.07964574545621872,
|
| 1534 |
+
"learning_rate": 4.02320788776628e-05,
|
| 1535 |
+
"loss": 0.1205,
|
| 1536 |
+
"step": 960
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 0.096,
|
| 1540 |
+
"eval_cos_sim": 0.8782918453216553,
|
| 1541 |
+
"eval_loss": 0.12269965698468159,
|
| 1542 |
+
"eval_runtime": 171.8922,
|
| 1543 |
+
"eval_samples_per_second": 23.27,
|
| 1544 |
+
"eval_steps_per_second": 1.454,
|
| 1545 |
+
"step": 960
|
| 1546 |
+
},
|
| 1547 |
+
{
|
| 1548 |
+
"epoch": 0.097,
|
| 1549 |
+
"grad_norm": 0.059999242424964905,
|
| 1550 |
+
"learning_rate": 2.559195333841573e-05,
|
| 1551 |
+
"loss": 0.1224,
|
| 1552 |
+
"step": 970
|
| 1553 |
+
},
|
| 1554 |
+
{
|
| 1555 |
+
"epoch": 0.097,
|
| 1556 |
+
"eval_cos_sim": 0.8782675862312317,
|
| 1557 |
+
"eval_loss": 0.12272447182881306,
|
| 1558 |
+
"eval_runtime": 178.4336,
|
| 1559 |
+
"eval_samples_per_second": 22.417,
|
| 1560 |
+
"eval_steps_per_second": 1.401,
|
| 1561 |
+
"step": 970
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"epoch": 0.098,
|
| 1565 |
+
"grad_norm": 0.07078584283590317,
|
| 1566 |
+
"learning_rate": 1.0723519806732741e-05,
|
| 1567 |
+
"loss": 0.1226,
|
| 1568 |
+
"step": 980
|
| 1569 |
+
},
|
| 1570 |
+
{
|
| 1571 |
+
"epoch": 0.098,
|
| 1572 |
+
"eval_cos_sim": 0.8782561421394348,
|
| 1573 |
+
"eval_loss": 0.12273399831997822,
|
| 1574 |
+
"eval_runtime": 172.0171,
|
| 1575 |
+
"eval_samples_per_second": 23.254,
|
| 1576 |
+
"eval_steps_per_second": 1.453,
|
| 1577 |
+
"step": 980
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 0.099,
|
| 1581 |
+
"grad_norm": 0.0700722336769104,
|
| 1582 |
+
"learning_rate": 1.3613218521583647e-06,
|
| 1583 |
+
"loss": 0.1189,
|
| 1584 |
+
"step": 990
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 0.099,
|
| 1588 |
+
"eval_cos_sim": 0.8782747387886047,
|
| 1589 |
+
"eval_loss": 0.1227147035812087,
|
| 1590 |
+
"eval_runtime": 174.8389,
|
| 1591 |
+
"eval_samples_per_second": 22.878,
|
| 1592 |
+
"eval_steps_per_second": 1.43,
|
| 1593 |
+
"step": 990
|
| 1594 |
+
},
|
| 1595 |
+
{
|
| 1596 |
+
"epoch": 0.1,
|
| 1597 |
+
"grad_norm": 0.06270556151866913,
|
| 1598 |
+
"learning_rate": 4.888377378787991e-05,
|
| 1599 |
+
"loss": 0.1209,
|
| 1600 |
+
"step": 1000
|
| 1601 |
+
},
|
| 1602 |
+
{
|
| 1603 |
+
"epoch": 0.1,
|
| 1604 |
+
"eval_cos_sim": 0.8783043622970581,
|
| 1605 |
+
"eval_loss": 0.12268760301815938,
|
| 1606 |
+
"eval_runtime": 171.6574,
|
| 1607 |
+
"eval_samples_per_second": 23.302,
|
| 1608 |
+
"eval_steps_per_second": 1.456,
|
| 1609 |
+
"step": 1000
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 0.101,
|
| 1613 |
+
"grad_norm": 0.059303585439920425,
|
| 1614 |
+
"learning_rate": 3.9917237207221514e-05,
|
| 1615 |
+
"loss": 0.1206,
|
| 1616 |
+
"step": 1010
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 0.101,
|
| 1620 |
+
"eval_cos_sim": 0.8785374760627747,
|
| 1621 |
+
"eval_loss": 0.12245997311818074,
|
| 1622 |
+
"eval_runtime": 173.2279,
|
| 1623 |
+
"eval_samples_per_second": 23.091,
|
| 1624 |
+
"eval_steps_per_second": 1.443,
|
| 1625 |
+
"step": 1010
|
| 1626 |
+
},
|
| 1627 |
+
{
|
| 1628 |
+
"epoch": 0.102,
|
| 1629 |
+
"grad_norm": 0.06463504582643509,
|
| 1630 |
+
"learning_rate": 2.519733417274297e-05,
|
| 1631 |
+
"loss": 0.122,
|
| 1632 |
+
"step": 1020
|
| 1633 |
+
},
|
| 1634 |
+
{
|
| 1635 |
+
"epoch": 0.102,
|
| 1636 |
+
"eval_cos_sim": 0.8785625100135803,
|
| 1637 |
+
"eval_loss": 0.12243694259869527,
|
| 1638 |
+
"eval_runtime": 179.8429,
|
| 1639 |
+
"eval_samples_per_second": 22.242,
|
| 1640 |
+
"eval_steps_per_second": 1.39,
|
| 1641 |
+
"step": 1020
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 0.103,
|
| 1645 |
+
"grad_norm": 0.06594408303499222,
|
| 1646 |
+
"learning_rate": 1.0401322151467458e-05,
|
| 1647 |
+
"loss": 0.1226,
|
| 1648 |
+
"step": 1030
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 0.103,
|
| 1652 |
+
"eval_cos_sim": 0.8784922361373901,
|
| 1653 |
+
"eval_loss": 0.1225029034827895,
|
| 1654 |
+
"eval_runtime": 171.8585,
|
| 1655 |
+
"eval_samples_per_second": 23.275,
|
| 1656 |
+
"eval_steps_per_second": 1.455,
|
| 1657 |
+
"step": 1030
|
| 1658 |
+
},
|
| 1659 |
+
{
|
| 1660 |
+
"epoch": 0.104,
|
| 1661 |
+
"grad_norm": 0.061140164732933044,
|
| 1662 |
+
"learning_rate": 1.2358127653053858e-06,
|
| 1663 |
+
"loss": 0.122,
|
| 1664 |
+
"step": 1040
|
| 1665 |
+
},
|
| 1666 |
+
{
|
| 1667 |
+
"epoch": 0.104,
|
| 1668 |
+
"eval_cos_sim": 0.8785346746444702,
|
| 1669 |
+
"eval_loss": 0.12245874931561421,
|
| 1670 |
+
"eval_runtime": 170.3116,
|
| 1671 |
+
"eval_samples_per_second": 23.486,
|
| 1672 |
+
"eval_steps_per_second": 1.468,
|
| 1673 |
+
"step": 1040
|
| 1674 |
+
},
|
| 1675 |
+
{
|
| 1676 |
+
"epoch": 0.105,
|
| 1677 |
+
"grad_norm": 0.06770511716604233,
|
| 1678 |
+
"learning_rate": 4.876418723469453e-05,
|
| 1679 |
+
"loss": 0.1196,
|
| 1680 |
+
"step": 1050
|
| 1681 |
+
},
|
| 1682 |
+
{
|
| 1683 |
+
"epoch": 0.105,
|
| 1684 |
+
"eval_cos_sim": 0.878551721572876,
|
| 1685 |
+
"eval_loss": 0.12243552591549825,
|
| 1686 |
+
"eval_runtime": 173.9331,
|
| 1687 |
+
"eval_samples_per_second": 22.997,
|
| 1688 |
+
"eval_steps_per_second": 1.437,
|
| 1689 |
+
"step": 1050
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 0.106,
|
| 1693 |
+
"grad_norm": 0.06050929054617882,
|
| 1694 |
+
"learning_rate": 3.959867784853255e-05,
|
| 1695 |
+
"loss": 0.1219,
|
| 1696 |
+
"step": 1060
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 0.106,
|
| 1700 |
+
"eval_cos_sim": 0.8784484267234802,
|
| 1701 |
+
"eval_loss": 0.12253486802327107,
|
| 1702 |
+
"eval_runtime": 175.2374,
|
| 1703 |
+
"eval_samples_per_second": 22.826,
|
| 1704 |
+
"eval_steps_per_second": 1.427,
|
| 1705 |
+
"step": 1060
|
| 1706 |
+
},
|
| 1707 |
+
{
|
| 1708 |
+
"epoch": 0.107,
|
| 1709 |
+
"grad_norm": 0.07329047471284866,
|
| 1710 |
+
"learning_rate": 2.4802665827257035e-05,
|
| 1711 |
+
"loss": 0.1214,
|
| 1712 |
+
"step": 1070
|
| 1713 |
+
},
|
| 1714 |
+
{
|
| 1715 |
+
"epoch": 0.107,
|
| 1716 |
+
"eval_cos_sim": 0.8785268068313599,
|
| 1717 |
+
"eval_loss": 0.12246101453053426,
|
| 1718 |
+
"eval_runtime": 172.381,
|
| 1719 |
+
"eval_samples_per_second": 23.204,
|
| 1720 |
+
"eval_steps_per_second": 1.45,
|
| 1721 |
+
"step": 1070
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 0.108,
|
| 1725 |
+
"grad_norm": 0.061687979847192764,
|
| 1726 |
+
"learning_rate": 1.0082762792778497e-05,
|
| 1727 |
+
"loss": 0.1206,
|
| 1728 |
+
"step": 1080
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 0.108,
|
| 1732 |
+
"eval_cos_sim": 0.8787024617195129,
|
| 1733 |
+
"eval_loss": 0.12228504302250813,
|
| 1734 |
+
"eval_runtime": 171.0068,
|
| 1735 |
+
"eval_samples_per_second": 23.391,
|
| 1736 |
+
"eval_steps_per_second": 1.462,
|
| 1737 |
+
"step": 1080
|
| 1738 |
+
},
|
| 1739 |
+
{
|
| 1740 |
+
"epoch": 0.109,
|
| 1741 |
+
"grad_norm": 0.06697102636098862,
|
| 1742 |
+
"learning_rate": 1.1162262121200917e-06,
|
| 1743 |
+
"loss": 0.1216,
|
| 1744 |
+
"step": 1090
|
| 1745 |
+
},
|
| 1746 |
+
{
|
| 1747 |
+
"epoch": 0.109,
|
| 1748 |
+
"eval_cos_sim": 0.8787557482719421,
|
| 1749 |
+
"eval_loss": 0.12223189308392476,
|
| 1750 |
+
"eval_runtime": 172.5647,
|
| 1751 |
+
"eval_samples_per_second": 23.18,
|
| 1752 |
+
"eval_steps_per_second": 1.449,
|
| 1753 |
+
"step": 1090
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 0.11,
|
| 1757 |
+
"grad_norm": 0.06245901808142662,
|
| 1758 |
+
"learning_rate": 4.8638678147841726e-05,
|
| 1759 |
+
"loss": 0.1224,
|
| 1760 |
+
"step": 1100
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 0.11,
|
| 1764 |
+
"eval_cos_sim": 0.878864049911499,
|
| 1765 |
+
"eval_loss": 0.12212434603917073,
|
| 1766 |
+
"eval_runtime": 177.5612,
|
| 1767 |
+
"eval_samples_per_second": 22.527,
|
| 1768 |
+
"eval_steps_per_second": 1.408,
|
| 1769 |
+
"step": 1100
|
| 1770 |
+
},
|
| 1771 |
+
{
|
| 1772 |
+
"epoch": 0.111,
|
| 1773 |
+
"grad_norm": 0.07445187121629715,
|
| 1774 |
+
"learning_rate": 3.9276480193267495e-05,
|
| 1775 |
+
"loss": 0.1226,
|
| 1776 |
+
"step": 1110
|
| 1777 |
+
},
|
| 1778 |
+
{
|
| 1779 |
+
"epoch": 0.111,
|
| 1780 |
+
"eval_cos_sim": 0.8787615895271301,
|
| 1781 |
+
"eval_loss": 0.12223191478001545,
|
| 1782 |
+
"eval_runtime": 170.2386,
|
| 1783 |
+
"eval_samples_per_second": 23.496,
|
| 1784 |
+
"eval_steps_per_second": 1.469,
|
| 1785 |
+
"step": 1110
|
| 1786 |
+
},
|
| 1787 |
+
{
|
| 1788 |
+
"epoch": 0.112,
|
| 1789 |
+
"grad_norm": 0.06328488141298294,
|
| 1790 |
+
"learning_rate": 2.4408046661584553e-05,
|
| 1791 |
+
"loss": 0.1205,
|
| 1792 |
+
"step": 1120
|
| 1793 |
+
},
|
| 1794 |
+
{
|
| 1795 |
+
"epoch": 0.112,
|
| 1796 |
+
"eval_cos_sim": 0.8786949515342712,
|
| 1797 |
+
"eval_loss": 0.12229911091076802,
|
| 1798 |
+
"eval_runtime": 173.6977,
|
| 1799 |
+
"eval_samples_per_second": 23.029,
|
| 1800 |
+
"eval_steps_per_second": 1.439,
|
| 1801 |
+
"step": 1120
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 0.113,
|
| 1805 |
+
"grad_norm": 0.1140422523021698,
|
| 1806 |
+
"learning_rate": 9.767921122337203e-06,
|
| 1807 |
+
"loss": 0.1213,
|
| 1808 |
+
"step": 1130
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 0.113,
|
| 1812 |
+
"eval_cos_sim": 0.8787314295768738,
|
| 1813 |
+
"eval_loss": 0.12225894191014242,
|
| 1814 |
+
"eval_runtime": 176.5254,
|
| 1815 |
+
"eval_samples_per_second": 22.66,
|
| 1816 |
+
"eval_steps_per_second": 1.416,
|
| 1817 |
+
"step": 1130
|
| 1818 |
+
},
|
| 1819 |
+
{
|
| 1820 |
+
"epoch": 0.114,
|
| 1821 |
+
"grad_norm": 0.07940120995044708,
|
| 1822 |
+
"learning_rate": 1.0025919960786169e-06,
|
| 1823 |
+
"loss": 0.1216,
|
| 1824 |
+
"step": 1140
|
| 1825 |
+
},
|
| 1826 |
+
{
|
| 1827 |
+
"epoch": 0.114,
|
| 1828 |
+
"eval_cos_sim": 0.878764271736145,
|
| 1829 |
+
"eval_loss": 0.12222567083584737,
|
| 1830 |
+
"eval_runtime": 173.6241,
|
| 1831 |
+
"eval_samples_per_second": 23.038,
|
| 1832 |
+
"eval_steps_per_second": 1.44,
|
| 1833 |
+
"step": 1140
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 0.115,
|
| 1837 |
+
"grad_norm": 0.06326926499605179,
|
| 1838 |
+
"learning_rate": 4.850727780681685e-05,
|
| 1839 |
+
"loss": 0.121,
|
| 1840 |
+
"step": 1150
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 0.115,
|
| 1844 |
+
"eval_cos_sim": 0.8787913918495178,
|
| 1845 |
+
"eval_loss": 0.1222020423625655,
|
| 1846 |
+
"eval_runtime": 197.6043,
|
| 1847 |
+
"eval_samples_per_second": 20.242,
|
| 1848 |
+
"eval_steps_per_second": 1.265,
|
| 1849 |
+
"step": 1150
|
| 1850 |
+
},
|
| 1851 |
+
{
|
| 1852 |
+
"epoch": 0.116,
|
| 1853 |
+
"grad_norm": 0.06304363161325455,
|
| 1854 |
+
"learning_rate": 3.89507245398359e-05,
|
| 1855 |
+
"loss": 0.1212,
|
| 1856 |
+
"step": 1160
|
| 1857 |
+
},
|
| 1858 |
+
{
|
| 1859 |
+
"epoch": 0.116,
|
| 1860 |
+
"eval_cos_sim": 0.8788431286811829,
|
| 1861 |
+
"eval_loss": 0.1221448552821822,
|
| 1862 |
+
"eval_runtime": 180.7769,
|
| 1863 |
+
"eval_samples_per_second": 22.127,
|
| 1864 |
+
"eval_steps_per_second": 1.383,
|
| 1865 |
+
"step": 1160
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 0.117,
|
| 1869 |
+
"grad_norm": 0.06048878654837608,
|
| 1870 |
+
"learning_rate": 2.4013575023093562e-05,
|
| 1871 |
+
"loss": 0.121,
|
| 1872 |
+
"step": 1170
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 0.117,
|
| 1876 |
+
"eval_cos_sim": 0.8789100050926208,
|
| 1877 |
+
"eval_loss": 0.12207724287259053,
|
| 1878 |
+
"eval_runtime": 175.5012,
|
| 1879 |
+
"eval_samples_per_second": 22.792,
|
| 1880 |
+
"eval_steps_per_second": 1.424,
|
| 1881 |
+
"step": 1170
|
| 1882 |
+
},
|
| 1883 |
+
{
|
| 1884 |
+
"epoch": 0.118,
|
| 1885 |
+
"grad_norm": 0.060076240450143814,
|
| 1886 |
+
"learning_rate": 9.456875605287529e-06,
|
| 1887 |
+
"loss": 0.1208,
|
| 1888 |
+
"step": 1180
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 0.118,
|
| 1892 |
+
"eval_cos_sim": 0.8789265751838684,
|
| 1893 |
+
"eval_loss": 0.12206284239041279,
|
| 1894 |
+
"eval_runtime": 179.6264,
|
| 1895 |
+
"eval_samples_per_second": 22.268,
|
| 1896 |
+
"eval_steps_per_second": 1.392,
|
| 1897 |
+
"step": 1180
|
| 1898 |
+
},
|
| 1899 |
+
{
|
| 1900 |
+
"epoch": 0.119,
|
| 1901 |
+
"grad_norm": 0.06535797566175461,
|
| 1902 |
+
"learning_rate": 8.949384372096747e-07,
|
| 1903 |
+
"loss": 0.1224,
|
| 1904 |
+
"step": 1190
|
| 1905 |
+
},
|
| 1906 |
+
{
|
| 1907 |
+
"epoch": 0.119,
|
| 1908 |
+
"eval_cos_sim": 0.8789151310920715,
|
| 1909 |
+
"eval_loss": 0.12207536175000142,
|
| 1910 |
+
"eval_runtime": 173.573,
|
| 1911 |
+
"eval_samples_per_second": 23.045,
|
| 1912 |
+
"eval_steps_per_second": 1.44,
|
| 1913 |
+
"step": 1190
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 0.12,
|
| 1917 |
+
"grad_norm": 0.051111843436956406,
|
| 1918 |
+
"learning_rate": 4.8370018959339916e-05,
|
| 1919 |
+
"loss": 0.1216,
|
| 1920 |
+
"step": 1200
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 0.12,
|
| 1924 |
+
"eval_cos_sim": 0.878704845905304,
|
| 1925 |
+
"eval_loss": 0.1222877917503066,
|
| 1926 |
+
"eval_runtime": 170.7747,
|
| 1927 |
+
"eval_samples_per_second": 23.423,
|
| 1928 |
+
"eval_steps_per_second": 1.464,
|
| 1929 |
+
"step": 1200
|
| 1930 |
+
},
|
| 1931 |
+
{
|
| 1932 |
+
"epoch": 0.121,
|
| 1933 |
+
"grad_norm": 0.07394807785749435,
|
| 1934 |
+
"learning_rate": 3.862149207337666e-05,
|
| 1935 |
+
"loss": 0.1227,
|
| 1936 |
+
"step": 1210
|
| 1937 |
+
},
|
| 1938 |
+
{
|
| 1939 |
+
"epoch": 0.121,
|
| 1940 |
+
"eval_cos_sim": 0.8786987662315369,
|
| 1941 |
+
"eval_loss": 0.12228692223774862,
|
| 1942 |
+
"eval_runtime": 172.7735,
|
| 1943 |
+
"eval_samples_per_second": 23.152,
|
| 1944 |
+
"eval_steps_per_second": 1.447,
|
| 1945 |
+
"step": 1210
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 0.122,
|
| 1949 |
+
"grad_norm": 0.06019896641373634,
|
| 1950 |
+
"learning_rate": 2.3619349222387182e-05,
|
| 1951 |
+
"loss": 0.1194,
|
| 1952 |
+
"step": 1220
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 0.122,
|
| 1956 |
+
"eval_cos_sim": 0.8791972398757935,
|
| 1957 |
+
"eval_loss": 0.12178870942341757,
|
| 1958 |
+
"eval_runtime": 171.5715,
|
| 1959 |
+
"eval_samples_per_second": 23.314,
|
| 1960 |
+
"eval_steps_per_second": 1.457,
|
| 1961 |
+
"step": 1220
|
| 1962 |
+
},
|
| 1963 |
+
{
|
| 1964 |
+
"epoch": 0.123,
|
| 1965 |
+
"grad_norm": 0.05350535735487938,
|
| 1966 |
+
"learning_rate": 9.149703760694162e-06,
|
| 1967 |
+
"loss": 0.1214,
|
| 1968 |
+
"step": 1230
|
| 1969 |
+
},
|
| 1970 |
+
{
|
| 1971 |
+
"epoch": 0.123,
|
| 1972 |
+
"eval_cos_sim": 0.8792542219161987,
|
| 1973 |
+
"eval_loss": 0.12173621847378684,
|
| 1974 |
+
"eval_runtime": 173.1804,
|
| 1975 |
+
"eval_samples_per_second": 23.097,
|
| 1976 |
+
"eval_steps_per_second": 1.444,
|
| 1977 |
+
"step": 1230
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 0.124,
|
| 1981 |
+
"grad_norm": 0.06338366866111755,
|
| 1982 |
+
"learning_rate": 7.932923650373624e-07,
|
| 1983 |
+
"loss": 0.1194,
|
| 1984 |
+
"step": 1240
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 0.124,
|
| 1988 |
+
"eval_cos_sim": 0.8792427182197571,
|
| 1989 |
+
"eval_loss": 0.12174849869954062,
|
| 1990 |
+
"eval_runtime": 172.0716,
|
| 1991 |
+
"eval_samples_per_second": 23.246,
|
| 1992 |
+
"eval_steps_per_second": 1.453,
|
| 1993 |
+
"step": 1240
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 0.125,
|
| 1997 |
+
"grad_norm": 0.052142199128866196,
|
| 1998 |
+
"learning_rate": 4.822693581319333e-05,
|
| 1999 |
+
"loss": 0.12,
|
| 2000 |
+
"step": 1250
|
| 2001 |
+
},
|
| 2002 |
+
{
|
| 2003 |
+
"epoch": 0.125,
|
| 2004 |
+
"eval_cos_sim": 0.8787649869918823,
|
| 2005 |
+
"eval_loss": 0.1222243664478011,
|
| 2006 |
+
"eval_runtime": 172.6696,
|
| 2007 |
+
"eval_samples_per_second": 23.166,
|
| 2008 |
+
"eval_steps_per_second": 1.448,
|
| 2009 |
+
"step": 1250
|
| 2010 |
+
},
|
| 2011 |
+
{
|
| 2012 |
+
"epoch": 0.126,
|
| 2013 |
+
"grad_norm": 0.0695052519440651,
|
| 2014 |
+
"learning_rate": 3.828886484552269e-05,
|
| 2015 |
+
"loss": 0.1213,
|
| 2016 |
+
"step": 1260
|
| 2017 |
+
},
|
| 2018 |
+
{
|
| 2019 |
+
"epoch": 0.126,
|
| 2020 |
+
"eval_cos_sim": 0.8785125017166138,
|
| 2021 |
+
"eval_loss": 0.12247128774868916,
|
| 2022 |
+
"eval_runtime": 182.4937,
|
| 2023 |
+
"eval_samples_per_second": 21.919,
|
| 2024 |
+
"eval_steps_per_second": 1.37,
|
| 2025 |
+
"step": 1260
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 0.127,
|
| 2029 |
+
"grad_norm": 0.07181504368782043,
|
| 2030 |
+
"learning_rate": 2.3225467508799494e-05,
|
| 2031 |
+
"loss": 0.1216,
|
| 2032 |
+
"step": 1270
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 0.127,
|
| 2036 |
+
"eval_cos_sim": 0.8791427612304688,
|
| 2037 |
+
"eval_loss": 0.12184033658253621,
|
| 2038 |
+
"eval_runtime": 172.8353,
|
| 2039 |
+
"eval_samples_per_second": 23.143,
|
| 2040 |
+
"eval_steps_per_second": 1.446,
|
| 2041 |
+
"step": 1270
|
| 2042 |
+
},
|
| 2043 |
+
{
|
| 2044 |
+
"epoch": 0.128,
|
| 2045 |
+
"grad_norm": 0.06035405769944191,
|
| 2046 |
+
"learning_rate": 8.846482142219678e-06,
|
| 2047 |
+
"loss": 0.12,
|
| 2048 |
+
"step": 1280
|
| 2049 |
+
},
|
| 2050 |
+
{
|
| 2051 |
+
"epoch": 0.128,
|
| 2052 |
+
"eval_cos_sim": 0.8793256282806396,
|
| 2053 |
+
"eval_loss": 0.121661689779634,
|
| 2054 |
+
"eval_runtime": 173.4166,
|
| 2055 |
+
"eval_samples_per_second": 23.066,
|
| 2056 |
+
"eval_steps_per_second": 1.442,
|
| 2057 |
+
"step": 1280
|
| 2058 |
+
}
|
| 2059 |
+
],
|
| 2060 |
+
"logging_steps": 10,
|
| 2061 |
+
"max_steps": 10000,
|
| 2062 |
+
"num_input_tokens_seen": 0,
|
| 2063 |
+
"num_train_epochs": 9223372036854775807,
|
| 2064 |
+
"save_steps": 10,
|
| 2065 |
+
"stateful_callbacks": {
|
| 2066 |
+
"TrainerControl": {
|
| 2067 |
+
"args": {
|
| 2068 |
+
"should_epoch_stop": false,
|
| 2069 |
+
"should_evaluate": false,
|
| 2070 |
+
"should_log": false,
|
| 2071 |
+
"should_save": true,
|
| 2072 |
+
"should_training_stop": false
|
| 2073 |
+
},
|
| 2074 |
+
"attributes": {}
|
| 2075 |
+
}
|
| 2076 |
+
},
|
| 2077 |
+
"total_flos": 0.0,
|
| 2078 |
+
"train_batch_size": 110,
|
| 2079 |
+
"trial_name": null,
|
| 2080 |
+
"trial_params": null
|
| 2081 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7a77bcc916ff5cd9b3a98f50c57b8e45219ef906088b81447a03b7b05b1befde
|
| 3 |
+
size 5176
|