Commit
·
408a3ab
1
Parent(s):
04981d4
add .pt and more
Browse files- README.md +1 -1
- config.json +1 -1
- optimizer.pt +3 -0
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1440 -0
- training_args.bin +3 -0
README.md
CHANGED
|
@@ -219,7 +219,7 @@ Then you can load this model and run inference.
|
|
| 219 |
from sentence_transformers import SentenceTransformer
|
| 220 |
|
| 221 |
# Download from the 🤗 Hub
|
| 222 |
-
model = SentenceTransformer("
|
| 223 |
# Run inference
|
| 224 |
sentences = [
|
| 225 |
'Thời điểm đánh giá và xếp loại chất lượng hằng năm của công chức, viên chức thuộc Bộ Tài chính được diễn ra trong thời gian nào?',
|
|
|
|
| 219 |
from sentence_transformers import SentenceTransformer
|
| 220 |
|
| 221 |
# Download from the 🤗 Hub
|
| 222 |
+
model = SentenceTransformer("sentence_transformers_model_id")
|
| 223 |
# Run inference
|
| 224 |
sentences = [
|
| 225 |
'Thời điểm đánh giá và xếp loại chất lượng hằng năm của công chức, viên chức thuộc Bộ Tài chính được diễn ra trong thời gian nào?',
|
config.json
CHANGED
|
@@ -1,5 +1,5 @@
|
|
| 1 |
{
|
| 2 |
-
"_name_or_path": "
|
| 3 |
"architectures": [
|
| 4 |
"BertModel"
|
| 5 |
],
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2",
|
| 3 |
"architectures": [
|
| 4 |
"BertModel"
|
| 5 |
],
|
optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c5510f2bea22289c94cdd149c0bc96f3b9186b16e0c8145a50c201b5afce88fc
|
| 3 |
+
size 940241530
|
rng_state.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:75b93c484b74178907107b74bab1d3483f0468cd107b60afbba86dba7f9e4961
|
| 3 |
+
size 14244
|
scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:15519c0147635f986a3d122dd9139c8b5cb8bb271f56a57171a991b0beb27d74
|
| 3 |
+
size 1064
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1440 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 1.2336434831708933,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 100500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.00613752976701937,
|
| 13 |
+
"grad_norm": 0.7345873117446899,
|
| 14 |
+
"learning_rate": 4.9897707837216344e-05,
|
| 15 |
+
"loss": 1.0473,
|
| 16 |
+
"step": 500
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 0.01227505953403874,
|
| 20 |
+
"grad_norm": 1.5779420137405396,
|
| 21 |
+
"learning_rate": 4.9795415674432687e-05,
|
| 22 |
+
"loss": 1.0447,
|
| 23 |
+
"step": 1000
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 0.01841258930105811,
|
| 27 |
+
"grad_norm": 1.1132322549819946,
|
| 28 |
+
"learning_rate": 4.969312351164903e-05,
|
| 29 |
+
"loss": 1.0383,
|
| 30 |
+
"step": 1500
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 0.02455011906807748,
|
| 34 |
+
"grad_norm": 1.4636728763580322,
|
| 35 |
+
"learning_rate": 4.959083134886538e-05,
|
| 36 |
+
"loss": 1.0395,
|
| 37 |
+
"step": 2000
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 0.03068764883509685,
|
| 41 |
+
"grad_norm": 2.259823799133301,
|
| 42 |
+
"learning_rate": 4.948853918608172e-05,
|
| 43 |
+
"loss": 1.0436,
|
| 44 |
+
"step": 2500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.03682517860211622,
|
| 48 |
+
"grad_norm": 3.276669979095459,
|
| 49 |
+
"learning_rate": 4.938624702329807e-05,
|
| 50 |
+
"loss": 1.0375,
|
| 51 |
+
"step": 3000
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 0.04296270836913559,
|
| 55 |
+
"grad_norm": 2.432560443878174,
|
| 56 |
+
"learning_rate": 4.928395486051441e-05,
|
| 57 |
+
"loss": 1.0189,
|
| 58 |
+
"step": 3500
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 0.04910023813615496,
|
| 62 |
+
"grad_norm": 2.106665849685669,
|
| 63 |
+
"learning_rate": 4.918166269773075e-05,
|
| 64 |
+
"loss": 1.0282,
|
| 65 |
+
"step": 4000
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.05523776790317433,
|
| 69 |
+
"grad_norm": 10.379436492919922,
|
| 70 |
+
"learning_rate": 4.9079370534947094e-05,
|
| 71 |
+
"loss": 1.0355,
|
| 72 |
+
"step": 4500
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 0.0613752976701937,
|
| 76 |
+
"grad_norm": 1.7805677652359009,
|
| 77 |
+
"learning_rate": 4.897707837216344e-05,
|
| 78 |
+
"loss": 1.0286,
|
| 79 |
+
"step": 5000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 0.06751282743721307,
|
| 83 |
+
"grad_norm": 4.699483394622803,
|
| 84 |
+
"learning_rate": 4.8874786209379785e-05,
|
| 85 |
+
"loss": 1.0264,
|
| 86 |
+
"step": 5500
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.07365035720423244,
|
| 90 |
+
"grad_norm": 3.2066714763641357,
|
| 91 |
+
"learning_rate": 4.877249404659613e-05,
|
| 92 |
+
"loss": 1.0174,
|
| 93 |
+
"step": 6000
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 0.07978788697125182,
|
| 97 |
+
"grad_norm": 4.304272651672363,
|
| 98 |
+
"learning_rate": 4.867020188381247e-05,
|
| 99 |
+
"loss": 1.0238,
|
| 100 |
+
"step": 6500
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.08592541673827118,
|
| 104 |
+
"grad_norm": 1.9387307167053223,
|
| 105 |
+
"learning_rate": 4.856790972102881e-05,
|
| 106 |
+
"loss": 1.0217,
|
| 107 |
+
"step": 7000
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 0.09206294650529055,
|
| 111 |
+
"grad_norm": 7.0195488929748535,
|
| 112 |
+
"learning_rate": 4.846561755824516e-05,
|
| 113 |
+
"loss": 1.0203,
|
| 114 |
+
"step": 7500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 0.09820047627230992,
|
| 118 |
+
"grad_norm": 5.841785907745361,
|
| 119 |
+
"learning_rate": 4.836332539546151e-05,
|
| 120 |
+
"loss": 1.0201,
|
| 121 |
+
"step": 8000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.10433800603932929,
|
| 125 |
+
"grad_norm": 4.427718162536621,
|
| 126 |
+
"learning_rate": 4.826103323267785e-05,
|
| 127 |
+
"loss": 1.0266,
|
| 128 |
+
"step": 8500
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.11047553580634865,
|
| 132 |
+
"grad_norm": 2.067218780517578,
|
| 133 |
+
"learning_rate": 4.815874106989419e-05,
|
| 134 |
+
"loss": 1.0379,
|
| 135 |
+
"step": 9000
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 0.11661306557336804,
|
| 139 |
+
"grad_norm": 4.303891181945801,
|
| 140 |
+
"learning_rate": 4.8056448907110535e-05,
|
| 141 |
+
"loss": 1.0367,
|
| 142 |
+
"step": 9500
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 0.1227505953403874,
|
| 146 |
+
"grad_norm": 5.363577842712402,
|
| 147 |
+
"learning_rate": 4.795415674432688e-05,
|
| 148 |
+
"loss": 1.0384,
|
| 149 |
+
"step": 10000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 0.12888812510740677,
|
| 153 |
+
"grad_norm": 7.8037028312683105,
|
| 154 |
+
"learning_rate": 4.785186458154322e-05,
|
| 155 |
+
"loss": 1.0291,
|
| 156 |
+
"step": 10500
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 0.13502565487442614,
|
| 160 |
+
"grad_norm": 2.8838307857513428,
|
| 161 |
+
"learning_rate": 4.774957241875956e-05,
|
| 162 |
+
"loss": 1.0362,
|
| 163 |
+
"step": 11000
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 0.1411631846414455,
|
| 167 |
+
"grad_norm": 18.838226318359375,
|
| 168 |
+
"learning_rate": 4.764728025597591e-05,
|
| 169 |
+
"loss": 1.0354,
|
| 170 |
+
"step": 11500
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.14730071440846487,
|
| 174 |
+
"grad_norm": 1.8900924921035767,
|
| 175 |
+
"learning_rate": 4.754498809319225e-05,
|
| 176 |
+
"loss": 1.0204,
|
| 177 |
+
"step": 12000
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.15343824417548424,
|
| 181 |
+
"grad_norm": 8.636752128601074,
|
| 182 |
+
"learning_rate": 4.74426959304086e-05,
|
| 183 |
+
"loss": 1.0401,
|
| 184 |
+
"step": 12500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 0.15957577394250363,
|
| 188 |
+
"grad_norm": 3.1361780166625977,
|
| 189 |
+
"learning_rate": 4.734040376762494e-05,
|
| 190 |
+
"loss": 1.0237,
|
| 191 |
+
"step": 13000
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.165713303709523,
|
| 195 |
+
"grad_norm": 6.990635871887207,
|
| 196 |
+
"learning_rate": 4.7238111604841285e-05,
|
| 197 |
+
"loss": 1.0271,
|
| 198 |
+
"step": 13500
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.17185083347654237,
|
| 202 |
+
"grad_norm": 2.6320559978485107,
|
| 203 |
+
"learning_rate": 4.713581944205763e-05,
|
| 204 |
+
"loss": 1.0235,
|
| 205 |
+
"step": 14000
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.17798836324356174,
|
| 209 |
+
"grad_norm": 5.751638889312744,
|
| 210 |
+
"learning_rate": 4.7033527279273976e-05,
|
| 211 |
+
"loss": 1.0329,
|
| 212 |
+
"step": 14500
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.1841258930105811,
|
| 216 |
+
"grad_norm": 10.333562850952148,
|
| 217 |
+
"learning_rate": 4.693123511649032e-05,
|
| 218 |
+
"loss": 1.0474,
|
| 219 |
+
"step": 15000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.19026342277760047,
|
| 223 |
+
"grad_norm": 9.57418155670166,
|
| 224 |
+
"learning_rate": 4.682894295370666e-05,
|
| 225 |
+
"loss": 1.0547,
|
| 226 |
+
"step": 15500
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 0.19640095254461984,
|
| 230 |
+
"grad_norm": 6.002443313598633,
|
| 231 |
+
"learning_rate": 4.6726650790923e-05,
|
| 232 |
+
"loss": 1.0557,
|
| 233 |
+
"step": 16000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.2025384823116392,
|
| 237 |
+
"grad_norm": 17.77951431274414,
|
| 238 |
+
"learning_rate": 4.6624358628139344e-05,
|
| 239 |
+
"loss": 1.0626,
|
| 240 |
+
"step": 16500
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 0.20867601207865857,
|
| 244 |
+
"grad_norm": 10.126110076904297,
|
| 245 |
+
"learning_rate": 4.652206646535569e-05,
|
| 246 |
+
"loss": 1.0551,
|
| 247 |
+
"step": 17000
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 0.21481354184567794,
|
| 251 |
+
"grad_norm": 3.7376019954681396,
|
| 252 |
+
"learning_rate": 4.641977430257204e-05,
|
| 253 |
+
"loss": 1.0526,
|
| 254 |
+
"step": 17500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.2209510716126973,
|
| 258 |
+
"grad_norm": 19.0529842376709,
|
| 259 |
+
"learning_rate": 4.6317482139788384e-05,
|
| 260 |
+
"loss": 1.125,
|
| 261 |
+
"step": 18000
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 0.2270886013797167,
|
| 265 |
+
"grad_norm": 1.9759446382522583,
|
| 266 |
+
"learning_rate": 4.6215189977004726e-05,
|
| 267 |
+
"loss": 1.2996,
|
| 268 |
+
"step": 18500
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 0.23322613114673607,
|
| 272 |
+
"grad_norm": 2.7964680194854736,
|
| 273 |
+
"learning_rate": 4.611289781422107e-05,
|
| 274 |
+
"loss": 1.0703,
|
| 275 |
+
"step": 19000
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 0.23936366091375544,
|
| 279 |
+
"grad_norm": 18.77906036376953,
|
| 280 |
+
"learning_rate": 4.601060565143741e-05,
|
| 281 |
+
"loss": 1.0601,
|
| 282 |
+
"step": 19500
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 0.2455011906807748,
|
| 286 |
+
"grad_norm": 3.129147529602051,
|
| 287 |
+
"learning_rate": 4.590831348865375e-05,
|
| 288 |
+
"loss": 1.0835,
|
| 289 |
+
"step": 20000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.2516387204477942,
|
| 293 |
+
"grad_norm": 3.018688917160034,
|
| 294 |
+
"learning_rate": 4.58060213258701e-05,
|
| 295 |
+
"loss": 1.0583,
|
| 296 |
+
"step": 20500
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.25777625021481354,
|
| 300 |
+
"grad_norm": 32.16483688354492,
|
| 301 |
+
"learning_rate": 4.570372916308644e-05,
|
| 302 |
+
"loss": 1.141,
|
| 303 |
+
"step": 21000
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 0.2639137799818329,
|
| 307 |
+
"grad_norm": 8.431964874267578,
|
| 308 |
+
"learning_rate": 4.5601437000302785e-05,
|
| 309 |
+
"loss": 1.0802,
|
| 310 |
+
"step": 21500
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.2700513097488523,
|
| 314 |
+
"grad_norm": 14.990534782409668,
|
| 315 |
+
"learning_rate": 4.5499144837519134e-05,
|
| 316 |
+
"loss": 1.0589,
|
| 317 |
+
"step": 22000
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 0.27618883951587164,
|
| 321 |
+
"grad_norm": 6.134013652801514,
|
| 322 |
+
"learning_rate": 4.5396852674735476e-05,
|
| 323 |
+
"loss": 1.086,
|
| 324 |
+
"step": 22500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 0.282326369282891,
|
| 328 |
+
"grad_norm": 15.587114334106445,
|
| 329 |
+
"learning_rate": 4.529456051195182e-05,
|
| 330 |
+
"loss": 1.0743,
|
| 331 |
+
"step": 23000
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 0.2884638990499104,
|
| 335 |
+
"grad_norm": 20.36590003967285,
|
| 336 |
+
"learning_rate": 4.519226834916816e-05,
|
| 337 |
+
"loss": 1.0605,
|
| 338 |
+
"step": 23500
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.29460142881692974,
|
| 342 |
+
"grad_norm": 6.840169906616211,
|
| 343 |
+
"learning_rate": 4.508997618638451e-05,
|
| 344 |
+
"loss": 1.0602,
|
| 345 |
+
"step": 24000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.3007389585839491,
|
| 349 |
+
"grad_norm": 22.10498046875,
|
| 350 |
+
"learning_rate": 4.498768402360085e-05,
|
| 351 |
+
"loss": 1.0732,
|
| 352 |
+
"step": 24500
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 0.3068764883509685,
|
| 356 |
+
"grad_norm": 2.3044190406799316,
|
| 357 |
+
"learning_rate": 4.488539186081719e-05,
|
| 358 |
+
"loss": 1.0614,
|
| 359 |
+
"step": 25000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 0.31301401811798785,
|
| 363 |
+
"grad_norm": 19.358158111572266,
|
| 364 |
+
"learning_rate": 4.4783099698033535e-05,
|
| 365 |
+
"loss": 1.0666,
|
| 366 |
+
"step": 25500
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 0.31915154788500727,
|
| 370 |
+
"grad_norm": 47.77769088745117,
|
| 371 |
+
"learning_rate": 4.468080753524988e-05,
|
| 372 |
+
"loss": 1.0669,
|
| 373 |
+
"step": 26000
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 0.32528907765202664,
|
| 377 |
+
"grad_norm": 31.26322364807129,
|
| 378 |
+
"learning_rate": 4.4578515372466226e-05,
|
| 379 |
+
"loss": 1.0627,
|
| 380 |
+
"step": 26500
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.331426607419046,
|
| 384 |
+
"grad_norm": 31.237651824951172,
|
| 385 |
+
"learning_rate": 4.4476223209682575e-05,
|
| 386 |
+
"loss": 1.0659,
|
| 387 |
+
"step": 27000
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 0.33756413718606537,
|
| 391 |
+
"grad_norm": 10.651726722717285,
|
| 392 |
+
"learning_rate": 4.437393104689892e-05,
|
| 393 |
+
"loss": 1.07,
|
| 394 |
+
"step": 27500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 0.34370166695308474,
|
| 398 |
+
"grad_norm": 10.767593383789062,
|
| 399 |
+
"learning_rate": 4.427163888411526e-05,
|
| 400 |
+
"loss": 1.0783,
|
| 401 |
+
"step": 28000
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.3498391967201041,
|
| 405 |
+
"grad_norm": 13.177048683166504,
|
| 406 |
+
"learning_rate": 4.41693467213316e-05,
|
| 407 |
+
"loss": 1.078,
|
| 408 |
+
"step": 28500
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.3559767264871235,
|
| 412 |
+
"grad_norm": 41.34538269042969,
|
| 413 |
+
"learning_rate": 4.406705455854794e-05,
|
| 414 |
+
"loss": 1.0832,
|
| 415 |
+
"step": 29000
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.36211425625414284,
|
| 419 |
+
"grad_norm": 2.813678026199341,
|
| 420 |
+
"learning_rate": 4.3964762395764285e-05,
|
| 421 |
+
"loss": 1.0695,
|
| 422 |
+
"step": 29500
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.3682517860211622,
|
| 426 |
+
"grad_norm": 32.63833236694336,
|
| 427 |
+
"learning_rate": 4.3862470232980634e-05,
|
| 428 |
+
"loss": 1.0714,
|
| 429 |
+
"step": 30000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.3743893157881816,
|
| 433 |
+
"grad_norm": 6.144928932189941,
|
| 434 |
+
"learning_rate": 4.3760178070196976e-05,
|
| 435 |
+
"loss": 1.3794,
|
| 436 |
+
"step": 30500
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.38052684555520094,
|
| 440 |
+
"grad_norm": 21.691205978393555,
|
| 441 |
+
"learning_rate": 4.365788590741332e-05,
|
| 442 |
+
"loss": 1.0838,
|
| 443 |
+
"step": 31000
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 0.3866643753222203,
|
| 447 |
+
"grad_norm": 5.830935955047607,
|
| 448 |
+
"learning_rate": 4.355559374462967e-05,
|
| 449 |
+
"loss": 1.0541,
|
| 450 |
+
"step": 31500
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.3928019050892397,
|
| 454 |
+
"grad_norm": 5.59116792678833,
|
| 455 |
+
"learning_rate": 4.345330158184601e-05,
|
| 456 |
+
"loss": 1.0799,
|
| 457 |
+
"step": 32000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.39893943485625905,
|
| 461 |
+
"grad_norm": 5.144099712371826,
|
| 462 |
+
"learning_rate": 4.335100941906235e-05,
|
| 463 |
+
"loss": 1.0622,
|
| 464 |
+
"step": 32500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.4050769646232784,
|
| 468 |
+
"grad_norm": 30.603145599365234,
|
| 469 |
+
"learning_rate": 4.324871725627869e-05,
|
| 470 |
+
"loss": 1.0597,
|
| 471 |
+
"step": 33000
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 0.4112144943902978,
|
| 475 |
+
"grad_norm": 14.602336883544922,
|
| 476 |
+
"learning_rate": 4.314642509349504e-05,
|
| 477 |
+
"loss": 1.0731,
|
| 478 |
+
"step": 33500
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 0.41735202415731715,
|
| 482 |
+
"grad_norm": 11.165698051452637,
|
| 483 |
+
"learning_rate": 4.3044132930711384e-05,
|
| 484 |
+
"loss": 1.0871,
|
| 485 |
+
"step": 34000
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 0.4234895539243365,
|
| 489 |
+
"grad_norm": 7.897899627685547,
|
| 490 |
+
"learning_rate": 4.2941840767927726e-05,
|
| 491 |
+
"loss": 1.0535,
|
| 492 |
+
"step": 34500
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 0.4296270836913559,
|
| 496 |
+
"grad_norm": 34.061134338378906,
|
| 497 |
+
"learning_rate": 4.283954860514407e-05,
|
| 498 |
+
"loss": 1.3215,
|
| 499 |
+
"step": 35000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 0.43576461345837525,
|
| 503 |
+
"grad_norm": 5.050889015197754,
|
| 504 |
+
"learning_rate": 4.273725644236041e-05,
|
| 505 |
+
"loss": 1.1501,
|
| 506 |
+
"step": 35500
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.4419021432253946,
|
| 510 |
+
"grad_norm": 11.23651123046875,
|
| 511 |
+
"learning_rate": 4.263496427957676e-05,
|
| 512 |
+
"loss": 1.1088,
|
| 513 |
+
"step": 36000
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.44803967299241404,
|
| 517 |
+
"grad_norm": 11.963942527770996,
|
| 518 |
+
"learning_rate": 4.25326721167931e-05,
|
| 519 |
+
"loss": 1.0844,
|
| 520 |
+
"step": 36500
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.4541772027594334,
|
| 524 |
+
"grad_norm": 19.694747924804688,
|
| 525 |
+
"learning_rate": 4.243037995400945e-05,
|
| 526 |
+
"loss": 1.0981,
|
| 527 |
+
"step": 37000
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 0.4603147325264528,
|
| 531 |
+
"grad_norm": 32.396446228027344,
|
| 532 |
+
"learning_rate": 4.232808779122579e-05,
|
| 533 |
+
"loss": 1.0856,
|
| 534 |
+
"step": 37500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 0.46645226229347214,
|
| 538 |
+
"grad_norm": 30.794092178344727,
|
| 539 |
+
"learning_rate": 4.2225795628442134e-05,
|
| 540 |
+
"loss": 1.0956,
|
| 541 |
+
"step": 38000
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 0.4725897920604915,
|
| 545 |
+
"grad_norm": 25.20745086669922,
|
| 546 |
+
"learning_rate": 4.2123503465658476e-05,
|
| 547 |
+
"loss": 1.0813,
|
| 548 |
+
"step": 38500
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.4787273218275109,
|
| 552 |
+
"grad_norm": 13.46246337890625,
|
| 553 |
+
"learning_rate": 4.202121130287482e-05,
|
| 554 |
+
"loss": 1.0843,
|
| 555 |
+
"step": 39000
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 0.48486485159453024,
|
| 559 |
+
"grad_norm": 9.27540397644043,
|
| 560 |
+
"learning_rate": 4.191891914009117e-05,
|
| 561 |
+
"loss": 1.1053,
|
| 562 |
+
"step": 39500
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 0.4910023813615496,
|
| 566 |
+
"grad_norm": 3.5075111389160156,
|
| 567 |
+
"learning_rate": 4.181662697730751e-05,
|
| 568 |
+
"loss": 1.092,
|
| 569 |
+
"step": 40000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.497139911128569,
|
| 573 |
+
"grad_norm": 6.219689846038818,
|
| 574 |
+
"learning_rate": 4.171433481452385e-05,
|
| 575 |
+
"loss": 1.081,
|
| 576 |
+
"step": 40500
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 0.5032774408955883,
|
| 580 |
+
"grad_norm": 16.255443572998047,
|
| 581 |
+
"learning_rate": 4.161204265174019e-05,
|
| 582 |
+
"loss": 1.0919,
|
| 583 |
+
"step": 41000
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 0.5094149706626078,
|
| 587 |
+
"grad_norm": 32.27680206298828,
|
| 588 |
+
"learning_rate": 4.150975048895654e-05,
|
| 589 |
+
"loss": 1.0681,
|
| 590 |
+
"step": 41500
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.5155525004296271,
|
| 594 |
+
"grad_norm": 107.78765869140625,
|
| 595 |
+
"learning_rate": 4.1407458326172884e-05,
|
| 596 |
+
"loss": 1.0826,
|
| 597 |
+
"step": 42000
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 0.5216900301966465,
|
| 601 |
+
"grad_norm": 8.029512405395508,
|
| 602 |
+
"learning_rate": 4.1305166163389226e-05,
|
| 603 |
+
"loss": 1.0809,
|
| 604 |
+
"step": 42500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 0.5278275599636658,
|
| 608 |
+
"grad_norm": 44.91604232788086,
|
| 609 |
+
"learning_rate": 4.1202874000605575e-05,
|
| 610 |
+
"loss": 1.093,
|
| 611 |
+
"step": 43000
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 0.5339650897306852,
|
| 615 |
+
"grad_norm": 11.860628128051758,
|
| 616 |
+
"learning_rate": 4.110058183782192e-05,
|
| 617 |
+
"loss": 1.0709,
|
| 618 |
+
"step": 43500
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 0.5401026194977046,
|
| 622 |
+
"grad_norm": 30.72344398498535,
|
| 623 |
+
"learning_rate": 4.099828967503826e-05,
|
| 624 |
+
"loss": 1.0623,
|
| 625 |
+
"step": 44000
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.546240149264724,
|
| 629 |
+
"grad_norm": 19.160911560058594,
|
| 630 |
+
"learning_rate": 4.08959975122546e-05,
|
| 631 |
+
"loss": 1.0801,
|
| 632 |
+
"step": 44500
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.5523776790317433,
|
| 636 |
+
"grad_norm": 12.648110389709473,
|
| 637 |
+
"learning_rate": 4.079370534947094e-05,
|
| 638 |
+
"loss": 1.0833,
|
| 639 |
+
"step": 45000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 0.5585152087987627,
|
| 643 |
+
"grad_norm": 25.488170623779297,
|
| 644 |
+
"learning_rate": 4.0691413186687285e-05,
|
| 645 |
+
"loss": 1.0816,
|
| 646 |
+
"step": 45500
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.564652738565782,
|
| 650 |
+
"grad_norm": 8.982975006103516,
|
| 651 |
+
"learning_rate": 4.0589121023903634e-05,
|
| 652 |
+
"loss": 1.0697,
|
| 653 |
+
"step": 46000
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 0.5707902683328014,
|
| 657 |
+
"grad_norm": 12.776642799377441,
|
| 658 |
+
"learning_rate": 4.048682886111998e-05,
|
| 659 |
+
"loss": 1.0864,
|
| 660 |
+
"step": 46500
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 0.5769277980998208,
|
| 664 |
+
"grad_norm": 70.14471435546875,
|
| 665 |
+
"learning_rate": 4.0384536698336325e-05,
|
| 666 |
+
"loss": 1.0744,
|
| 667 |
+
"step": 47000
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.5830653278668402,
|
| 671 |
+
"grad_norm": 26.72732162475586,
|
| 672 |
+
"learning_rate": 4.028224453555267e-05,
|
| 673 |
+
"loss": 1.0897,
|
| 674 |
+
"step": 47500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.5892028576338595,
|
| 678 |
+
"grad_norm": 22.69961166381836,
|
| 679 |
+
"learning_rate": 4.017995237276901e-05,
|
| 680 |
+
"loss": 1.0727,
|
| 681 |
+
"step": 48000
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.5953403874008789,
|
| 685 |
+
"grad_norm": 6.76914119720459,
|
| 686 |
+
"learning_rate": 4.007766020998535e-05,
|
| 687 |
+
"loss": 1.0621,
|
| 688 |
+
"step": 48500
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.6014779171678982,
|
| 692 |
+
"grad_norm": 41.32302474975586,
|
| 693 |
+
"learning_rate": 3.99753680472017e-05,
|
| 694 |
+
"loss": 1.0582,
|
| 695 |
+
"step": 49000
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.6076154469349176,
|
| 699 |
+
"grad_norm": 47.626644134521484,
|
| 700 |
+
"learning_rate": 3.987307588441804e-05,
|
| 701 |
+
"loss": 1.0681,
|
| 702 |
+
"step": 49500
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 0.613752976701937,
|
| 706 |
+
"grad_norm": 27.24641227722168,
|
| 707 |
+
"learning_rate": 3.9770783721634384e-05,
|
| 708 |
+
"loss": 1.083,
|
| 709 |
+
"step": 50000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 0.6198905064689564,
|
| 713 |
+
"grad_norm": 22.27688217163086,
|
| 714 |
+
"learning_rate": 3.9668491558850726e-05,
|
| 715 |
+
"loss": 1.0632,
|
| 716 |
+
"step": 50500
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.6260280362359757,
|
| 720 |
+
"grad_norm": 43.488407135009766,
|
| 721 |
+
"learning_rate": 3.9566199396067075e-05,
|
| 722 |
+
"loss": 1.0809,
|
| 723 |
+
"step": 51000
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 0.6321655660029951,
|
| 727 |
+
"grad_norm": 8.742218017578125,
|
| 728 |
+
"learning_rate": 3.946390723328342e-05,
|
| 729 |
+
"loss": 1.0525,
|
| 730 |
+
"step": 51500
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 0.6383030957700145,
|
| 734 |
+
"grad_norm": 35.785621643066406,
|
| 735 |
+
"learning_rate": 3.936161507049976e-05,
|
| 736 |
+
"loss": 1.6649,
|
| 737 |
+
"step": 52000
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.6444406255370339,
|
| 741 |
+
"grad_norm": 17.245643615722656,
|
| 742 |
+
"learning_rate": 3.925932290771611e-05,
|
| 743 |
+
"loss": 1.0873,
|
| 744 |
+
"step": 52500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 0.6505781553040533,
|
| 748 |
+
"grad_norm": 11.474211692810059,
|
| 749 |
+
"learning_rate": 3.915703074493245e-05,
|
| 750 |
+
"loss": 1.0649,
|
| 751 |
+
"step": 53000
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 0.6567156850710726,
|
| 755 |
+
"grad_norm": 57.05070114135742,
|
| 756 |
+
"learning_rate": 3.905473858214879e-05,
|
| 757 |
+
"loss": 1.0591,
|
| 758 |
+
"step": 53500
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.662853214838092,
|
| 762 |
+
"grad_norm": 28.446794509887695,
|
| 763 |
+
"learning_rate": 3.8952446419365134e-05,
|
| 764 |
+
"loss": 1.061,
|
| 765 |
+
"step": 54000
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 0.6689907446051113,
|
| 769 |
+
"grad_norm": 34.050048828125,
|
| 770 |
+
"learning_rate": 3.8850154256581476e-05,
|
| 771 |
+
"loss": 1.0682,
|
| 772 |
+
"step": 54500
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 0.6751282743721307,
|
| 776 |
+
"grad_norm": 40.37173843383789,
|
| 777 |
+
"learning_rate": 3.874786209379782e-05,
|
| 778 |
+
"loss": 1.0616,
|
| 779 |
+
"step": 55000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 0.68126580413915,
|
| 783 |
+
"grad_norm": 17.34925079345703,
|
| 784 |
+
"learning_rate": 3.864556993101417e-05,
|
| 785 |
+
"loss": 1.0827,
|
| 786 |
+
"step": 55500
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 0.6874033339061695,
|
| 790 |
+
"grad_norm": 57.942806243896484,
|
| 791 |
+
"learning_rate": 3.8543277768230516e-05,
|
| 792 |
+
"loss": 1.0799,
|
| 793 |
+
"step": 56000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.6935408636731888,
|
| 797 |
+
"grad_norm": 27.28282928466797,
|
| 798 |
+
"learning_rate": 3.844098560544686e-05,
|
| 799 |
+
"loss": 1.0705,
|
| 800 |
+
"step": 56500
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.6996783934402082,
|
| 804 |
+
"grad_norm": 34.14970016479492,
|
| 805 |
+
"learning_rate": 3.83386934426632e-05,
|
| 806 |
+
"loss": 1.0821,
|
| 807 |
+
"step": 57000
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 0.7058159232072275,
|
| 811 |
+
"grad_norm": 24.639619827270508,
|
| 812 |
+
"learning_rate": 3.823640127987954e-05,
|
| 813 |
+
"loss": 1.0763,
|
| 814 |
+
"step": 57500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 0.711953452974247,
|
| 818 |
+
"grad_norm": 45.23350524902344,
|
| 819 |
+
"learning_rate": 3.8134109117095884e-05,
|
| 820 |
+
"loss": 1.0842,
|
| 821 |
+
"step": 58000
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 0.7180909827412663,
|
| 825 |
+
"grad_norm": 10.940398216247559,
|
| 826 |
+
"learning_rate": 3.803181695431223e-05,
|
| 827 |
+
"loss": 1.0813,
|
| 828 |
+
"step": 58500
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 0.7242285125082857,
|
| 832 |
+
"grad_norm": 95.62271881103516,
|
| 833 |
+
"learning_rate": 3.7929524791528575e-05,
|
| 834 |
+
"loss": 1.0678,
|
| 835 |
+
"step": 59000
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 0.730366042275305,
|
| 839 |
+
"grad_norm": 39.37797164916992,
|
| 840 |
+
"learning_rate": 3.782723262874492e-05,
|
| 841 |
+
"loss": 1.0894,
|
| 842 |
+
"step": 59500
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.7365035720423244,
|
| 846 |
+
"grad_norm": 14.900856971740723,
|
| 847 |
+
"learning_rate": 3.772494046596126e-05,
|
| 848 |
+
"loss": 1.0733,
|
| 849 |
+
"step": 60000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.7426411018093437,
|
| 853 |
+
"grad_norm": 10.382057189941406,
|
| 854 |
+
"learning_rate": 3.762264830317761e-05,
|
| 855 |
+
"loss": 1.0688,
|
| 856 |
+
"step": 60500
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 0.7487786315763632,
|
| 860 |
+
"grad_norm": 7.631415367126465,
|
| 861 |
+
"learning_rate": 3.752035614039395e-05,
|
| 862 |
+
"loss": 1.0665,
|
| 863 |
+
"step": 61000
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 0.7549161613433825,
|
| 867 |
+
"grad_norm": 38.29065704345703,
|
| 868 |
+
"learning_rate": 3.74180639776103e-05,
|
| 869 |
+
"loss": 1.0681,
|
| 870 |
+
"step": 61500
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 0.7610536911104019,
|
| 874 |
+
"grad_norm": 44.250144958496094,
|
| 875 |
+
"learning_rate": 3.731577181482664e-05,
|
| 876 |
+
"loss": 1.301,
|
| 877 |
+
"step": 62000
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 0.7671912208774213,
|
| 881 |
+
"grad_norm": 3.7524375915527344,
|
| 882 |
+
"learning_rate": 3.721347965204298e-05,
|
| 883 |
+
"loss": 1.0907,
|
| 884 |
+
"step": 62500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.7733287506444406,
|
| 888 |
+
"grad_norm": 41.10980987548828,
|
| 889 |
+
"learning_rate": 3.7111187489259325e-05,
|
| 890 |
+
"loss": 1.3941,
|
| 891 |
+
"step": 63000
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 0.77946628041146,
|
| 895 |
+
"grad_norm": 70.5405044555664,
|
| 896 |
+
"learning_rate": 3.700889532647567e-05,
|
| 897 |
+
"loss": 1.1355,
|
| 898 |
+
"step": 63500
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 0.7856038101784794,
|
| 902 |
+
"grad_norm": 84.906005859375,
|
| 903 |
+
"learning_rate": 3.690660316369201e-05,
|
| 904 |
+
"loss": 1.2196,
|
| 905 |
+
"step": 64000
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.7917413399454988,
|
| 909 |
+
"grad_norm": 86.1518325805664,
|
| 910 |
+
"learning_rate": 3.680431100090835e-05,
|
| 911 |
+
"loss": 1.225,
|
| 912 |
+
"step": 64500
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.7978788697125181,
|
| 916 |
+
"grad_norm": 15.497425079345703,
|
| 917 |
+
"learning_rate": 3.67020188381247e-05,
|
| 918 |
+
"loss": 1.1437,
|
| 919 |
+
"step": 65000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 0.8040163994795375,
|
| 923 |
+
"grad_norm": 77.67127227783203,
|
| 924 |
+
"learning_rate": 3.659972667534105e-05,
|
| 925 |
+
"loss": 1.0787,
|
| 926 |
+
"step": 65500
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.8101539292465568,
|
| 930 |
+
"grad_norm": 54.82986068725586,
|
| 931 |
+
"learning_rate": 3.649743451255739e-05,
|
| 932 |
+
"loss": 1.0686,
|
| 933 |
+
"step": 66000
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.8162914590135762,
|
| 937 |
+
"grad_norm": 16.840795516967773,
|
| 938 |
+
"learning_rate": 3.639514234977373e-05,
|
| 939 |
+
"loss": 1.1017,
|
| 940 |
+
"step": 66500
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 0.8224289887805956,
|
| 944 |
+
"grad_norm": 9.104973793029785,
|
| 945 |
+
"learning_rate": 3.6292850186990075e-05,
|
| 946 |
+
"loss": 1.0999,
|
| 947 |
+
"step": 67000
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 0.828566518547615,
|
| 951 |
+
"grad_norm": 21.70631980895996,
|
| 952 |
+
"learning_rate": 3.6190558024206417e-05,
|
| 953 |
+
"loss": 1.0771,
|
| 954 |
+
"step": 67500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 0.8347040483146343,
|
| 958 |
+
"grad_norm": 26.37040901184082,
|
| 959 |
+
"learning_rate": 3.6088265861422765e-05,
|
| 960 |
+
"loss": 1.1015,
|
| 961 |
+
"step": 68000
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.8408415780816537,
|
| 965 |
+
"grad_norm": 17.264562606811523,
|
| 966 |
+
"learning_rate": 3.598597369863911e-05,
|
| 967 |
+
"loss": 1.0826,
|
| 968 |
+
"step": 68500
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.846979107848673,
|
| 972 |
+
"grad_norm": 46.9713249206543,
|
| 973 |
+
"learning_rate": 3.588368153585545e-05,
|
| 974 |
+
"loss": 1.1046,
|
| 975 |
+
"step": 69000
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 0.8531166376156925,
|
| 979 |
+
"grad_norm": 21.621822357177734,
|
| 980 |
+
"learning_rate": 3.578138937307179e-05,
|
| 981 |
+
"loss": 1.0735,
|
| 982 |
+
"step": 69500
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 0.8592541673827118,
|
| 986 |
+
"grad_norm": 23.59171485900879,
|
| 987 |
+
"learning_rate": 3.567909721028814e-05,
|
| 988 |
+
"loss": 1.1056,
|
| 989 |
+
"step": 70000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 0.8653916971497312,
|
| 993 |
+
"grad_norm": 52.57392883300781,
|
| 994 |
+
"learning_rate": 3.557680504750448e-05,
|
| 995 |
+
"loss": 1.1077,
|
| 996 |
+
"step": 70500
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 0.8715292269167505,
|
| 1000 |
+
"grad_norm": 4.959003448486328,
|
| 1001 |
+
"learning_rate": 3.547451288472083e-05,
|
| 1002 |
+
"loss": 1.0897,
|
| 1003 |
+
"step": 71000
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 0.8776667566837699,
|
| 1007 |
+
"grad_norm": 10.618963241577148,
|
| 1008 |
+
"learning_rate": 3.537222072193717e-05,
|
| 1009 |
+
"loss": 1.0775,
|
| 1010 |
+
"step": 71500
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.8838042864507892,
|
| 1014 |
+
"grad_norm": 34.424346923828125,
|
| 1015 |
+
"learning_rate": 3.5269928559153515e-05,
|
| 1016 |
+
"loss": 1.0907,
|
| 1017 |
+
"step": 72000
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.8899418162178087,
|
| 1021 |
+
"grad_norm": 16.29790496826172,
|
| 1022 |
+
"learning_rate": 3.516763639636986e-05,
|
| 1023 |
+
"loss": 1.0705,
|
| 1024 |
+
"step": 72500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 0.8960793459848281,
|
| 1028 |
+
"grad_norm": 39.85658645629883,
|
| 1029 |
+
"learning_rate": 3.50653442335862e-05,
|
| 1030 |
+
"loss": 1.0776,
|
| 1031 |
+
"step": 73000
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 0.9022168757518474,
|
| 1035 |
+
"grad_norm": 42.42219161987305,
|
| 1036 |
+
"learning_rate": 3.496305207080254e-05,
|
| 1037 |
+
"loss": 1.0896,
|
| 1038 |
+
"step": 73500
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 0.9083544055188668,
|
| 1042 |
+
"grad_norm": 67.0084228515625,
|
| 1043 |
+
"learning_rate": 3.4860759908018884e-05,
|
| 1044 |
+
"loss": 1.0889,
|
| 1045 |
+
"step": 74000
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 0.9144919352858861,
|
| 1049 |
+
"grad_norm": 39.18657684326172,
|
| 1050 |
+
"learning_rate": 3.475846774523523e-05,
|
| 1051 |
+
"loss": 1.0804,
|
| 1052 |
+
"step": 74500
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.9206294650529055,
|
| 1056 |
+
"grad_norm": 76.6010513305664,
|
| 1057 |
+
"learning_rate": 3.4656175582451574e-05,
|
| 1058 |
+
"loss": 1.1087,
|
| 1059 |
+
"step": 75000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 0.9267669948199249,
|
| 1063 |
+
"grad_norm": 7.220675945281982,
|
| 1064 |
+
"learning_rate": 3.455388341966792e-05,
|
| 1065 |
+
"loss": 1.0738,
|
| 1066 |
+
"step": 75500
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 0.9329045245869443,
|
| 1070 |
+
"grad_norm": 32.431453704833984,
|
| 1071 |
+
"learning_rate": 3.4451591256884265e-05,
|
| 1072 |
+
"loss": 1.0806,
|
| 1073 |
+
"step": 76000
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.9390420543539636,
|
| 1077 |
+
"grad_norm": 89.43970489501953,
|
| 1078 |
+
"learning_rate": 3.434929909410061e-05,
|
| 1079 |
+
"loss": 1.0899,
|
| 1080 |
+
"step": 76500
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 0.945179584120983,
|
| 1084 |
+
"grad_norm": 25.720674514770508,
|
| 1085 |
+
"learning_rate": 3.424700693131695e-05,
|
| 1086 |
+
"loss": 1.0814,
|
| 1087 |
+
"step": 77000
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 0.9513171138880023,
|
| 1091 |
+
"grad_norm": 44.011478424072266,
|
| 1092 |
+
"learning_rate": 3.41447147685333e-05,
|
| 1093 |
+
"loss": 1.0723,
|
| 1094 |
+
"step": 77500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.9574546436550218,
|
| 1098 |
+
"grad_norm": 67.16317749023438,
|
| 1099 |
+
"learning_rate": 3.404242260574964e-05,
|
| 1100 |
+
"loss": 1.0923,
|
| 1101 |
+
"step": 78000
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 0.9635921734220411,
|
| 1105 |
+
"grad_norm": 116.56925201416016,
|
| 1106 |
+
"learning_rate": 3.394013044296598e-05,
|
| 1107 |
+
"loss": 1.0748,
|
| 1108 |
+
"step": 78500
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 0.9697297031890605,
|
| 1112 |
+
"grad_norm": 99.30931091308594,
|
| 1113 |
+
"learning_rate": 3.3837838280182324e-05,
|
| 1114 |
+
"loss": 1.0745,
|
| 1115 |
+
"step": 79000
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 0.9758672329560798,
|
| 1119 |
+
"grad_norm": 60.88938522338867,
|
| 1120 |
+
"learning_rate": 3.3735546117398667e-05,
|
| 1121 |
+
"loss": 1.081,
|
| 1122 |
+
"step": 79500
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.9820047627230992,
|
| 1126 |
+
"grad_norm": 38.99333572387695,
|
| 1127 |
+
"learning_rate": 3.3633253954615015e-05,
|
| 1128 |
+
"loss": 1.08,
|
| 1129 |
+
"step": 80000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.9881422924901185,
|
| 1133 |
+
"grad_norm": 125.93304443359375,
|
| 1134 |
+
"learning_rate": 3.3530961791831364e-05,
|
| 1135 |
+
"loss": 1.0905,
|
| 1136 |
+
"step": 80500
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.994279822257138,
|
| 1140 |
+
"grad_norm": 39.0662841796875,
|
| 1141 |
+
"learning_rate": 3.3428669629047706e-05,
|
| 1142 |
+
"loss": 1.1064,
|
| 1143 |
+
"step": 81000
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 1.0004173520241573,
|
| 1147 |
+
"grad_norm": 105.9404296875,
|
| 1148 |
+
"learning_rate": 3.332637746626405e-05,
|
| 1149 |
+
"loss": 1.0929,
|
| 1150 |
+
"step": 81500
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 1.0065548817911767,
|
| 1154 |
+
"grad_norm": 74.30028533935547,
|
| 1155 |
+
"learning_rate": 3.322408530348039e-05,
|
| 1156 |
+
"loss": 1.0815,
|
| 1157 |
+
"step": 82000
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 1.0126924115581961,
|
| 1161 |
+
"grad_norm": 14.825672149658203,
|
| 1162 |
+
"learning_rate": 3.312179314069673e-05,
|
| 1163 |
+
"loss": 1.0768,
|
| 1164 |
+
"step": 82500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 1.0188299413252153,
|
| 1168 |
+
"grad_norm": 58.09488296508789,
|
| 1169 |
+
"learning_rate": 3.3019500977913074e-05,
|
| 1170 |
+
"loss": 1.1004,
|
| 1171 |
+
"step": 83000
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 1.0249674710922347,
|
| 1175 |
+
"grad_norm": 33.13391876220703,
|
| 1176 |
+
"learning_rate": 3.2917208815129416e-05,
|
| 1177 |
+
"loss": 1.0835,
|
| 1178 |
+
"step": 83500
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 1.0311050008592542,
|
| 1182 |
+
"grad_norm": 23.691511154174805,
|
| 1183 |
+
"learning_rate": 3.2814916652345765e-05,
|
| 1184 |
+
"loss": 1.0765,
|
| 1185 |
+
"step": 84000
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 1.0372425306262736,
|
| 1189 |
+
"grad_norm": 57.42161178588867,
|
| 1190 |
+
"learning_rate": 3.271262448956211e-05,
|
| 1191 |
+
"loss": 1.0906,
|
| 1192 |
+
"step": 84500
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 1.043380060393293,
|
| 1196 |
+
"grad_norm": 52.550350189208984,
|
| 1197 |
+
"learning_rate": 3.2610332326778456e-05,
|
| 1198 |
+
"loss": 1.096,
|
| 1199 |
+
"step": 85000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 1.0495175901603122,
|
| 1203 |
+
"grad_norm": 89.49881744384766,
|
| 1204 |
+
"learning_rate": 3.25080401639948e-05,
|
| 1205 |
+
"loss": 1.1085,
|
| 1206 |
+
"step": 85500
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 1.0556551199273316,
|
| 1210 |
+
"grad_norm": 11.250481605529785,
|
| 1211 |
+
"learning_rate": 3.240574800121114e-05,
|
| 1212 |
+
"loss": 1.0913,
|
| 1213 |
+
"step": 86000
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 1.061792649694351,
|
| 1217 |
+
"grad_norm": 31.0516357421875,
|
| 1218 |
+
"learning_rate": 3.230345583842748e-05,
|
| 1219 |
+
"loss": 1.0974,
|
| 1220 |
+
"step": 86500
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 1.0679301794613705,
|
| 1224 |
+
"grad_norm": 94.23751068115234,
|
| 1225 |
+
"learning_rate": 3.220116367564383e-05,
|
| 1226 |
+
"loss": 1.0763,
|
| 1227 |
+
"step": 87000
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 1.0740677092283897,
|
| 1231 |
+
"grad_norm": 26.741689682006836,
|
| 1232 |
+
"learning_rate": 3.209887151286017e-05,
|
| 1233 |
+
"loss": 1.0894,
|
| 1234 |
+
"step": 87500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 1.080205238995409,
|
| 1238 |
+
"grad_norm": 39.82439422607422,
|
| 1239 |
+
"learning_rate": 3.1996579350076515e-05,
|
| 1240 |
+
"loss": 1.1065,
|
| 1241 |
+
"step": 88000
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.0863427687624285,
|
| 1245 |
+
"grad_norm": 7.871860027313232,
|
| 1246 |
+
"learning_rate": 3.189428718729286e-05,
|
| 1247 |
+
"loss": 1.0898,
|
| 1248 |
+
"step": 88500
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 1.092480298529448,
|
| 1252 |
+
"grad_norm": 35.93474578857422,
|
| 1253 |
+
"learning_rate": 3.17919950245092e-05,
|
| 1254 |
+
"loss": 1.1036,
|
| 1255 |
+
"step": 89000
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 1.0986178282964671,
|
| 1259 |
+
"grad_norm": 146.91851806640625,
|
| 1260 |
+
"learning_rate": 3.168970286172555e-05,
|
| 1261 |
+
"loss": 1.0825,
|
| 1262 |
+
"step": 89500
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 1.1047553580634866,
|
| 1266 |
+
"grad_norm": 187.8535614013672,
|
| 1267 |
+
"learning_rate": 3.15874106989419e-05,
|
| 1268 |
+
"loss": 1.1164,
|
| 1269 |
+
"step": 90000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 1.110892887830506,
|
| 1273 |
+
"grad_norm": 77.92440795898438,
|
| 1274 |
+
"learning_rate": 3.148511853615824e-05,
|
| 1275 |
+
"loss": 1.0811,
|
| 1276 |
+
"step": 90500
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 1.1170304175975254,
|
| 1280 |
+
"grad_norm": 62.233360290527344,
|
| 1281 |
+
"learning_rate": 3.138282637337458e-05,
|
| 1282 |
+
"loss": 1.115,
|
| 1283 |
+
"step": 91000
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 1.1231679473645446,
|
| 1287 |
+
"grad_norm": 54.17692947387695,
|
| 1288 |
+
"learning_rate": 3.128053421059092e-05,
|
| 1289 |
+
"loss": 1.1123,
|
| 1290 |
+
"step": 91500
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 1.129305477131564,
|
| 1294 |
+
"grad_norm": 49.207454681396484,
|
| 1295 |
+
"learning_rate": 3.1178242047807265e-05,
|
| 1296 |
+
"loss": 1.0846,
|
| 1297 |
+
"step": 92000
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 1.1354430068985835,
|
| 1301 |
+
"grad_norm": 49.67575454711914,
|
| 1302 |
+
"learning_rate": 3.107594988502361e-05,
|
| 1303 |
+
"loss": 1.0917,
|
| 1304 |
+
"step": 92500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 1.1415805366656029,
|
| 1308 |
+
"grad_norm": 154.799072265625,
|
| 1309 |
+
"learning_rate": 3.0973657722239956e-05,
|
| 1310 |
+
"loss": 1.0879,
|
| 1311 |
+
"step": 93000
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 1.1477180664326223,
|
| 1315 |
+
"grad_norm": 146.50338745117188,
|
| 1316 |
+
"learning_rate": 3.08713655594563e-05,
|
| 1317 |
+
"loss": 1.0969,
|
| 1318 |
+
"step": 93500
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 1.1538555961996415,
|
| 1322 |
+
"grad_norm": 155.8907928466797,
|
| 1323 |
+
"learning_rate": 3.076907339667264e-05,
|
| 1324 |
+
"loss": 1.0849,
|
| 1325 |
+
"step": 94000
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 1.159993125966661,
|
| 1329 |
+
"grad_norm": 26.285480499267578,
|
| 1330 |
+
"learning_rate": 3.066678123388899e-05,
|
| 1331 |
+
"loss": 1.0852,
|
| 1332 |
+
"step": 94500
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 1.1661306557336804,
|
| 1336 |
+
"grad_norm": 31.642507553100586,
|
| 1337 |
+
"learning_rate": 3.056448907110533e-05,
|
| 1338 |
+
"loss": 1.0774,
|
| 1339 |
+
"step": 95000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 1.1722681855006996,
|
| 1343 |
+
"grad_norm": 19.160383224487305,
|
| 1344 |
+
"learning_rate": 3.0462196908321673e-05,
|
| 1345 |
+
"loss": 1.0984,
|
| 1346 |
+
"step": 95500
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.178405715267719,
|
| 1350 |
+
"grad_norm": 38.4587516784668,
|
| 1351 |
+
"learning_rate": 3.0359904745538015e-05,
|
| 1352 |
+
"loss": 1.0936,
|
| 1353 |
+
"step": 96000
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 1.1845432450347384,
|
| 1357 |
+
"grad_norm": 54.26088333129883,
|
| 1358 |
+
"learning_rate": 3.0257612582754364e-05,
|
| 1359 |
+
"loss": 1.0842,
|
| 1360 |
+
"step": 96500
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 1.1906807748017578,
|
| 1364 |
+
"grad_norm": 22.72319221496582,
|
| 1365 |
+
"learning_rate": 3.0155320419970706e-05,
|
| 1366 |
+
"loss": 1.0895,
|
| 1367 |
+
"step": 97000
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 1.1968183045687772,
|
| 1371 |
+
"grad_norm": 88.04389190673828,
|
| 1372 |
+
"learning_rate": 3.0053028257187048e-05,
|
| 1373 |
+
"loss": 1.09,
|
| 1374 |
+
"step": 97500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 1.2029558343357964,
|
| 1378 |
+
"grad_norm": 49.89582061767578,
|
| 1379 |
+
"learning_rate": 2.9950736094403394e-05,
|
| 1380 |
+
"loss": 1.0813,
|
| 1381 |
+
"step": 98000
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 1.2090933641028159,
|
| 1385 |
+
"grad_norm": 107.11421966552734,
|
| 1386 |
+
"learning_rate": 2.9848443931619736e-05,
|
| 1387 |
+
"loss": 1.0965,
|
| 1388 |
+
"step": 98500
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 1.2152308938698353,
|
| 1392 |
+
"grad_norm": 9.076912879943848,
|
| 1393 |
+
"learning_rate": 2.9746151768836078e-05,
|
| 1394 |
+
"loss": 1.1017,
|
| 1395 |
+
"step": 99000
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 1.2213684236368547,
|
| 1399 |
+
"grad_norm": 64.24601745605469,
|
| 1400 |
+
"learning_rate": 2.9643859606052427e-05,
|
| 1401 |
+
"loss": 1.1045,
|
| 1402 |
+
"step": 99500
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 1.227505953403874,
|
| 1406 |
+
"grad_norm": 38.064544677734375,
|
| 1407 |
+
"learning_rate": 2.954156744326877e-05,
|
| 1408 |
+
"loss": 1.093,
|
| 1409 |
+
"step": 100000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 1.2336434831708933,
|
| 1413 |
+
"grad_norm": 98.80583190917969,
|
| 1414 |
+
"learning_rate": 2.9439275280485114e-05,
|
| 1415 |
+
"loss": 1.0903,
|
| 1416 |
+
"step": 100500
|
| 1417 |
+
}
|
| 1418 |
+
],
|
| 1419 |
+
"logging_steps": 500,
|
| 1420 |
+
"max_steps": 244398,
|
| 1421 |
+
"num_input_tokens_seen": 0,
|
| 1422 |
+
"num_train_epochs": 3,
|
| 1423 |
+
"save_steps": 500,
|
| 1424 |
+
"stateful_callbacks": {
|
| 1425 |
+
"TrainerControl": {
|
| 1426 |
+
"args": {
|
| 1427 |
+
"should_epoch_stop": false,
|
| 1428 |
+
"should_evaluate": false,
|
| 1429 |
+
"should_log": false,
|
| 1430 |
+
"should_save": true,
|
| 1431 |
+
"should_training_stop": false
|
| 1432 |
+
},
|
| 1433 |
+
"attributes": {}
|
| 1434 |
+
}
|
| 1435 |
+
},
|
| 1436 |
+
"total_flos": 0.0,
|
| 1437 |
+
"train_batch_size": 8,
|
| 1438 |
+
"trial_name": null,
|
| 1439 |
+
"trial_params": null
|
| 1440 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:320a99fb4b91f8525d40d71b170ac4e34508c4a26597a496c07d42c089137f1c
|
| 3 |
+
size 5432
|