connordilgren
commited on
Commit
•
6ae3c3c
1
Parent(s):
e146df4
Best accuracy model on validation set
Browse files- config.json +1 -1
- model.safetensors +1 -1
- optimizer.pt +3 -0
- preprocessor_config.json +1 -1
- rng_state.pth +3 -0
- scheduler.pt +3 -0
- trainer_state.json +1863 -0
- training_args.bin +3 -0
config.json
CHANGED
@@ -44,5 +44,5 @@
|
|
44 |
"qkv_bias": true,
|
45 |
"time_stride": 10,
|
46 |
"torch_dtype": "float32",
|
47 |
-
"transformers_version": "4.41.
|
48 |
}
|
|
|
44 |
"qkv_bias": true,
|
45 |
"time_stride": 10,
|
46 |
"torch_dtype": "float32",
|
47 |
+
"transformers_version": "4.41.1"
|
48 |
}
|
model.safetensors
CHANGED
@@ -1,3 +1,3 @@
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
-
oid sha256:
|
3 |
size 344814656
|
|
|
1 |
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:8e7dff7fe680d0465f0cefd5fb8862c63190bf31ded21762bbd53ff01ca9f51a
|
3 |
size 344814656
|
optimizer.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ea1585c5e99a9a6f68986fe94460f1b6eba441857a88bde0dd3dfb4be8e78cfd
|
3 |
+
size 689741690
|
preprocessor_config.json
CHANGED
@@ -10,4 +10,4 @@
|
|
10 |
"return_attention_mask": false,
|
11 |
"sampling_rate": 16000,
|
12 |
"std": 4.5689974
|
13 |
-
}
|
|
|
10 |
"return_attention_mask": false,
|
11 |
"sampling_rate": 16000,
|
12 |
"std": 4.5689974
|
13 |
+
}
|
rng_state.pth
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:25c948628ba1804e0dc0e988e547ad05d27e9a4f2682e0d6f6481f21ae112fc2
|
3 |
+
size 14244
|
scheduler.pt
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:303ef6e8096a5fb41b818f64313bfd1335a41caf6beb7513c3d3ef79b8d6e284
|
3 |
+
size 1064
|
trainer_state.json
ADDED
@@ -0,0 +1,1863 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"best_metric": 0.86,
|
3 |
+
"best_model_checkpoint": "./checkpoints/checkpoint_model_lr_5e_05_bs_16/checkpoint-1144",
|
4 |
+
"epoch": 26.0,
|
5 |
+
"eval_steps": 500,
|
6 |
+
"global_step": 1144,
|
7 |
+
"is_hyper_param_search": false,
|
8 |
+
"is_local_process_zero": true,
|
9 |
+
"is_world_process_zero": true,
|
10 |
+
"log_history": [
|
11 |
+
{
|
12 |
+
"epoch": 0.11363636363636363,
|
13 |
+
"grad_norm": 15.224111557006836,
|
14 |
+
"learning_rate": 1.1363636363636364e-06,
|
15 |
+
"loss": 2.5973,
|
16 |
+
"step": 5
|
17 |
+
},
|
18 |
+
{
|
19 |
+
"epoch": 0.22727272727272727,
|
20 |
+
"grad_norm": 14.61684513092041,
|
21 |
+
"learning_rate": 2.0454545454545457e-06,
|
22 |
+
"loss": 2.5056,
|
23 |
+
"step": 10
|
24 |
+
},
|
25 |
+
{
|
26 |
+
"epoch": 0.3409090909090909,
|
27 |
+
"grad_norm": 13.760466575622559,
|
28 |
+
"learning_rate": 3.1818181818181817e-06,
|
29 |
+
"loss": 2.4719,
|
30 |
+
"step": 15
|
31 |
+
},
|
32 |
+
{
|
33 |
+
"epoch": 0.45454545454545453,
|
34 |
+
"grad_norm": 15.71677017211914,
|
35 |
+
"learning_rate": 4.3181818181818185e-06,
|
36 |
+
"loss": 2.3512,
|
37 |
+
"step": 20
|
38 |
+
},
|
39 |
+
{
|
40 |
+
"epoch": 0.5681818181818182,
|
41 |
+
"grad_norm": 12.284369468688965,
|
42 |
+
"learning_rate": 5.4545454545454545e-06,
|
43 |
+
"loss": 2.2199,
|
44 |
+
"step": 25
|
45 |
+
},
|
46 |
+
{
|
47 |
+
"epoch": 0.6818181818181818,
|
48 |
+
"grad_norm": 11.864232063293457,
|
49 |
+
"learning_rate": 6.59090909090909e-06,
|
50 |
+
"loss": 2.0262,
|
51 |
+
"step": 30
|
52 |
+
},
|
53 |
+
{
|
54 |
+
"epoch": 0.7954545454545454,
|
55 |
+
"grad_norm": 11.269423484802246,
|
56 |
+
"learning_rate": 7.727272727272727e-06,
|
57 |
+
"loss": 1.7453,
|
58 |
+
"step": 35
|
59 |
+
},
|
60 |
+
{
|
61 |
+
"epoch": 0.9090909090909091,
|
62 |
+
"grad_norm": 13.125885963439941,
|
63 |
+
"learning_rate": 8.863636363636365e-06,
|
64 |
+
"loss": 1.541,
|
65 |
+
"step": 40
|
66 |
+
},
|
67 |
+
{
|
68 |
+
"epoch": 1.0,
|
69 |
+
"eval_accuracy": 0.67,
|
70 |
+
"eval_loss": 1.2503589391708374,
|
71 |
+
"eval_runtime": 18.1476,
|
72 |
+
"eval_samples_per_second": 11.021,
|
73 |
+
"eval_steps_per_second": 0.716,
|
74 |
+
"step": 44
|
75 |
+
},
|
76 |
+
{
|
77 |
+
"epoch": 1.0227272727272727,
|
78 |
+
"grad_norm": 11.554954528808594,
|
79 |
+
"learning_rate": 1e-05,
|
80 |
+
"loss": 1.2945,
|
81 |
+
"step": 45
|
82 |
+
},
|
83 |
+
{
|
84 |
+
"epoch": 1.1363636363636362,
|
85 |
+
"grad_norm": 11.968127250671387,
|
86 |
+
"learning_rate": 1.1136363636363637e-05,
|
87 |
+
"loss": 1.0641,
|
88 |
+
"step": 50
|
89 |
+
},
|
90 |
+
{
|
91 |
+
"epoch": 1.25,
|
92 |
+
"grad_norm": 11.197978019714355,
|
93 |
+
"learning_rate": 1.2272727272727273e-05,
|
94 |
+
"loss": 0.846,
|
95 |
+
"step": 55
|
96 |
+
},
|
97 |
+
{
|
98 |
+
"epoch": 1.3636363636363638,
|
99 |
+
"grad_norm": 9.72460651397705,
|
100 |
+
"learning_rate": 1.340909090909091e-05,
|
101 |
+
"loss": 0.6572,
|
102 |
+
"step": 60
|
103 |
+
},
|
104 |
+
{
|
105 |
+
"epoch": 1.4772727272727273,
|
106 |
+
"grad_norm": 14.739611625671387,
|
107 |
+
"learning_rate": 1.4545454545454545e-05,
|
108 |
+
"loss": 0.6622,
|
109 |
+
"step": 65
|
110 |
+
},
|
111 |
+
{
|
112 |
+
"epoch": 1.5909090909090908,
|
113 |
+
"grad_norm": 13.50337028503418,
|
114 |
+
"learning_rate": 1.5681818181818182e-05,
|
115 |
+
"loss": 0.565,
|
116 |
+
"step": 70
|
117 |
+
},
|
118 |
+
{
|
119 |
+
"epoch": 1.7045454545454546,
|
120 |
+
"grad_norm": 9.195622444152832,
|
121 |
+
"learning_rate": 1.6818181818181818e-05,
|
122 |
+
"loss": 0.5775,
|
123 |
+
"step": 75
|
124 |
+
},
|
125 |
+
{
|
126 |
+
"epoch": 1.8181818181818183,
|
127 |
+
"grad_norm": 10.333452224731445,
|
128 |
+
"learning_rate": 1.7954545454545454e-05,
|
129 |
+
"loss": 0.5743,
|
130 |
+
"step": 80
|
131 |
+
},
|
132 |
+
{
|
133 |
+
"epoch": 1.9318181818181817,
|
134 |
+
"grad_norm": 10.101814270019531,
|
135 |
+
"learning_rate": 1.9090909090909094e-05,
|
136 |
+
"loss": 0.5395,
|
137 |
+
"step": 85
|
138 |
+
},
|
139 |
+
{
|
140 |
+
"epoch": 2.0,
|
141 |
+
"eval_accuracy": 0.75,
|
142 |
+
"eval_loss": 0.715374231338501,
|
143 |
+
"eval_runtime": 18.1709,
|
144 |
+
"eval_samples_per_second": 11.007,
|
145 |
+
"eval_steps_per_second": 0.715,
|
146 |
+
"step": 88
|
147 |
+
},
|
148 |
+
{
|
149 |
+
"epoch": 2.0454545454545454,
|
150 |
+
"grad_norm": 8.385086059570312,
|
151 |
+
"learning_rate": 2.022727272727273e-05,
|
152 |
+
"loss": 0.5222,
|
153 |
+
"step": 90
|
154 |
+
},
|
155 |
+
{
|
156 |
+
"epoch": 2.159090909090909,
|
157 |
+
"grad_norm": 4.65899133682251,
|
158 |
+
"learning_rate": 2.1363636363636362e-05,
|
159 |
+
"loss": 0.4306,
|
160 |
+
"step": 95
|
161 |
+
},
|
162 |
+
{
|
163 |
+
"epoch": 2.2727272727272725,
|
164 |
+
"grad_norm": 10.004486083984375,
|
165 |
+
"learning_rate": 2.25e-05,
|
166 |
+
"loss": 0.2421,
|
167 |
+
"step": 100
|
168 |
+
},
|
169 |
+
{
|
170 |
+
"epoch": 2.3863636363636362,
|
171 |
+
"grad_norm": 9.067935943603516,
|
172 |
+
"learning_rate": 2.3636363636363637e-05,
|
173 |
+
"loss": 0.2421,
|
174 |
+
"step": 105
|
175 |
+
},
|
176 |
+
{
|
177 |
+
"epoch": 2.5,
|
178 |
+
"grad_norm": 6.462744235992432,
|
179 |
+
"learning_rate": 2.4772727272727277e-05,
|
180 |
+
"loss": 0.2362,
|
181 |
+
"step": 110
|
182 |
+
},
|
183 |
+
{
|
184 |
+
"epoch": 2.6136363636363638,
|
185 |
+
"grad_norm": 22.449291229248047,
|
186 |
+
"learning_rate": 2.590909090909091e-05,
|
187 |
+
"loss": 0.3596,
|
188 |
+
"step": 115
|
189 |
+
},
|
190 |
+
{
|
191 |
+
"epoch": 2.7272727272727275,
|
192 |
+
"grad_norm": 10.209068298339844,
|
193 |
+
"learning_rate": 2.7045454545454545e-05,
|
194 |
+
"loss": 0.3211,
|
195 |
+
"step": 120
|
196 |
+
},
|
197 |
+
{
|
198 |
+
"epoch": 2.840909090909091,
|
199 |
+
"grad_norm": 12.491495132446289,
|
200 |
+
"learning_rate": 2.818181818181818e-05,
|
201 |
+
"loss": 0.3784,
|
202 |
+
"step": 125
|
203 |
+
},
|
204 |
+
{
|
205 |
+
"epoch": 2.9545454545454546,
|
206 |
+
"grad_norm": 14.337518692016602,
|
207 |
+
"learning_rate": 2.9318181818181817e-05,
|
208 |
+
"loss": 0.3125,
|
209 |
+
"step": 130
|
210 |
+
},
|
211 |
+
{
|
212 |
+
"epoch": 3.0,
|
213 |
+
"eval_accuracy": 0.705,
|
214 |
+
"eval_loss": 0.8171606659889221,
|
215 |
+
"eval_runtime": 17.9915,
|
216 |
+
"eval_samples_per_second": 11.116,
|
217 |
+
"eval_steps_per_second": 0.723,
|
218 |
+
"step": 132
|
219 |
+
},
|
220 |
+
{
|
221 |
+
"epoch": 3.0681818181818183,
|
222 |
+
"grad_norm": 9.655952453613281,
|
223 |
+
"learning_rate": 3.0454545454545456e-05,
|
224 |
+
"loss": 0.4244,
|
225 |
+
"step": 135
|
226 |
+
},
|
227 |
+
{
|
228 |
+
"epoch": 3.1818181818181817,
|
229 |
+
"grad_norm": 4.139129161834717,
|
230 |
+
"learning_rate": 3.159090909090909e-05,
|
231 |
+
"loss": 0.1704,
|
232 |
+
"step": 140
|
233 |
+
},
|
234 |
+
{
|
235 |
+
"epoch": 3.2954545454545454,
|
236 |
+
"grad_norm": 9.019370079040527,
|
237 |
+
"learning_rate": 3.272727272727273e-05,
|
238 |
+
"loss": 0.2044,
|
239 |
+
"step": 145
|
240 |
+
},
|
241 |
+
{
|
242 |
+
"epoch": 3.409090909090909,
|
243 |
+
"grad_norm": 1.8330605030059814,
|
244 |
+
"learning_rate": 3.3863636363636364e-05,
|
245 |
+
"loss": 0.0844,
|
246 |
+
"step": 150
|
247 |
+
},
|
248 |
+
{
|
249 |
+
"epoch": 3.5227272727272725,
|
250 |
+
"grad_norm": 1.0041872262954712,
|
251 |
+
"learning_rate": 3.5e-05,
|
252 |
+
"loss": 0.1684,
|
253 |
+
"step": 155
|
254 |
+
},
|
255 |
+
{
|
256 |
+
"epoch": 3.6363636363636362,
|
257 |
+
"grad_norm": 13.836721420288086,
|
258 |
+
"learning_rate": 3.613636363636364e-05,
|
259 |
+
"loss": 0.1244,
|
260 |
+
"step": 160
|
261 |
+
},
|
262 |
+
{
|
263 |
+
"epoch": 3.75,
|
264 |
+
"grad_norm": 10.699409484863281,
|
265 |
+
"learning_rate": 3.7272727272727276e-05,
|
266 |
+
"loss": 0.1537,
|
267 |
+
"step": 165
|
268 |
+
},
|
269 |
+
{
|
270 |
+
"epoch": 3.8636363636363638,
|
271 |
+
"grad_norm": 1.7060661315917969,
|
272 |
+
"learning_rate": 3.840909090909091e-05,
|
273 |
+
"loss": 0.1152,
|
274 |
+
"step": 170
|
275 |
+
},
|
276 |
+
{
|
277 |
+
"epoch": 3.9772727272727275,
|
278 |
+
"grad_norm": 7.366329669952393,
|
279 |
+
"learning_rate": 3.954545454545455e-05,
|
280 |
+
"loss": 0.1762,
|
281 |
+
"step": 175
|
282 |
+
},
|
283 |
+
{
|
284 |
+
"epoch": 4.0,
|
285 |
+
"eval_accuracy": 0.765,
|
286 |
+
"eval_loss": 0.7440270781517029,
|
287 |
+
"eval_runtime": 18.4547,
|
288 |
+
"eval_samples_per_second": 10.837,
|
289 |
+
"eval_steps_per_second": 0.704,
|
290 |
+
"step": 176
|
291 |
+
},
|
292 |
+
{
|
293 |
+
"epoch": 4.090909090909091,
|
294 |
+
"grad_norm": 5.621038436889648,
|
295 |
+
"learning_rate": 4.068181818181818e-05,
|
296 |
+
"loss": 0.0961,
|
297 |
+
"step": 180
|
298 |
+
},
|
299 |
+
{
|
300 |
+
"epoch": 4.204545454545454,
|
301 |
+
"grad_norm": 2.5091235637664795,
|
302 |
+
"learning_rate": 4.181818181818182e-05,
|
303 |
+
"loss": 0.195,
|
304 |
+
"step": 185
|
305 |
+
},
|
306 |
+
{
|
307 |
+
"epoch": 4.318181818181818,
|
308 |
+
"grad_norm": 11.424132347106934,
|
309 |
+
"learning_rate": 4.295454545454546e-05,
|
310 |
+
"loss": 0.0859,
|
311 |
+
"step": 190
|
312 |
+
},
|
313 |
+
{
|
314 |
+
"epoch": 4.431818181818182,
|
315 |
+
"grad_norm": 2.8084189891815186,
|
316 |
+
"learning_rate": 4.409090909090909e-05,
|
317 |
+
"loss": 0.0252,
|
318 |
+
"step": 195
|
319 |
+
},
|
320 |
+
{
|
321 |
+
"epoch": 4.545454545454545,
|
322 |
+
"grad_norm": 2.906534194946289,
|
323 |
+
"learning_rate": 4.522727272727273e-05,
|
324 |
+
"loss": 0.062,
|
325 |
+
"step": 200
|
326 |
+
},
|
327 |
+
{
|
328 |
+
"epoch": 4.659090909090909,
|
329 |
+
"grad_norm": 7.522055149078369,
|
330 |
+
"learning_rate": 4.636363636363636e-05,
|
331 |
+
"loss": 0.0973,
|
332 |
+
"step": 205
|
333 |
+
},
|
334 |
+
{
|
335 |
+
"epoch": 4.7727272727272725,
|
336 |
+
"grad_norm": 7.575300693511963,
|
337 |
+
"learning_rate": 4.75e-05,
|
338 |
+
"loss": 0.0665,
|
339 |
+
"step": 210
|
340 |
+
},
|
341 |
+
{
|
342 |
+
"epoch": 4.886363636363637,
|
343 |
+
"grad_norm": 25.463464736938477,
|
344 |
+
"learning_rate": 4.863636363636364e-05,
|
345 |
+
"loss": 0.1488,
|
346 |
+
"step": 215
|
347 |
+
},
|
348 |
+
{
|
349 |
+
"epoch": 5.0,
|
350 |
+
"grad_norm": 16.796293258666992,
|
351 |
+
"learning_rate": 4.9772727272727275e-05,
|
352 |
+
"loss": 0.4321,
|
353 |
+
"step": 220
|
354 |
+
},
|
355 |
+
{
|
356 |
+
"epoch": 5.0,
|
357 |
+
"eval_accuracy": 0.73,
|
358 |
+
"eval_loss": 1.1859827041625977,
|
359 |
+
"eval_runtime": 18.064,
|
360 |
+
"eval_samples_per_second": 11.072,
|
361 |
+
"eval_steps_per_second": 0.72,
|
362 |
+
"step": 220
|
363 |
+
},
|
364 |
+
{
|
365 |
+
"epoch": 5.113636363636363,
|
366 |
+
"grad_norm": 0.248295396566391,
|
367 |
+
"learning_rate": 4.98989898989899e-05,
|
368 |
+
"loss": 0.2525,
|
369 |
+
"step": 225
|
370 |
+
},
|
371 |
+
{
|
372 |
+
"epoch": 5.2272727272727275,
|
373 |
+
"grad_norm": 1.8920602798461914,
|
374 |
+
"learning_rate": 4.9772727272727275e-05,
|
375 |
+
"loss": 0.0468,
|
376 |
+
"step": 230
|
377 |
+
},
|
378 |
+
{
|
379 |
+
"epoch": 5.340909090909091,
|
380 |
+
"grad_norm": 10.923778533935547,
|
381 |
+
"learning_rate": 4.964646464646465e-05,
|
382 |
+
"loss": 0.0465,
|
383 |
+
"step": 235
|
384 |
+
},
|
385 |
+
{
|
386 |
+
"epoch": 5.454545454545454,
|
387 |
+
"grad_norm": 2.1219613552093506,
|
388 |
+
"learning_rate": 4.952020202020202e-05,
|
389 |
+
"loss": 0.1053,
|
390 |
+
"step": 240
|
391 |
+
},
|
392 |
+
{
|
393 |
+
"epoch": 5.568181818181818,
|
394 |
+
"grad_norm": 3.359290361404419,
|
395 |
+
"learning_rate": 4.93939393939394e-05,
|
396 |
+
"loss": 0.0528,
|
397 |
+
"step": 245
|
398 |
+
},
|
399 |
+
{
|
400 |
+
"epoch": 5.681818181818182,
|
401 |
+
"grad_norm": 11.049290657043457,
|
402 |
+
"learning_rate": 4.9267676767676765e-05,
|
403 |
+
"loss": 0.1639,
|
404 |
+
"step": 250
|
405 |
+
},
|
406 |
+
{
|
407 |
+
"epoch": 5.795454545454545,
|
408 |
+
"grad_norm": 14.748278617858887,
|
409 |
+
"learning_rate": 4.9141414141414145e-05,
|
410 |
+
"loss": 0.1953,
|
411 |
+
"step": 255
|
412 |
+
},
|
413 |
+
{
|
414 |
+
"epoch": 5.909090909090909,
|
415 |
+
"grad_norm": 1.4790602922439575,
|
416 |
+
"learning_rate": 4.901515151515152e-05,
|
417 |
+
"loss": 0.1226,
|
418 |
+
"step": 260
|
419 |
+
},
|
420 |
+
{
|
421 |
+
"epoch": 6.0,
|
422 |
+
"eval_accuracy": 0.76,
|
423 |
+
"eval_loss": 1.0318822860717773,
|
424 |
+
"eval_runtime": 17.924,
|
425 |
+
"eval_samples_per_second": 11.158,
|
426 |
+
"eval_steps_per_second": 0.725,
|
427 |
+
"step": 264
|
428 |
+
},
|
429 |
+
{
|
430 |
+
"epoch": 6.0227272727272725,
|
431 |
+
"grad_norm": 22.95401954650879,
|
432 |
+
"learning_rate": 4.888888888888889e-05,
|
433 |
+
"loss": 0.1116,
|
434 |
+
"step": 265
|
435 |
+
},
|
436 |
+
{
|
437 |
+
"epoch": 6.136363636363637,
|
438 |
+
"grad_norm": 18.67777442932129,
|
439 |
+
"learning_rate": 4.876262626262626e-05,
|
440 |
+
"loss": 0.1461,
|
441 |
+
"step": 270
|
442 |
+
},
|
443 |
+
{
|
444 |
+
"epoch": 6.25,
|
445 |
+
"grad_norm": 1.6643537282943726,
|
446 |
+
"learning_rate": 4.863636363636364e-05,
|
447 |
+
"loss": 0.1635,
|
448 |
+
"step": 275
|
449 |
+
},
|
450 |
+
{
|
451 |
+
"epoch": 6.363636363636363,
|
452 |
+
"grad_norm": 1.0799901485443115,
|
453 |
+
"learning_rate": 4.851010101010101e-05,
|
454 |
+
"loss": 0.0535,
|
455 |
+
"step": 280
|
456 |
+
},
|
457 |
+
{
|
458 |
+
"epoch": 6.4772727272727275,
|
459 |
+
"grad_norm": 3.293269395828247,
|
460 |
+
"learning_rate": 4.838383838383839e-05,
|
461 |
+
"loss": 0.1025,
|
462 |
+
"step": 285
|
463 |
+
},
|
464 |
+
{
|
465 |
+
"epoch": 6.590909090909091,
|
466 |
+
"grad_norm": 1.957770824432373,
|
467 |
+
"learning_rate": 4.825757575757576e-05,
|
468 |
+
"loss": 0.0845,
|
469 |
+
"step": 290
|
470 |
+
},
|
471 |
+
{
|
472 |
+
"epoch": 6.704545454545455,
|
473 |
+
"grad_norm": 1.3866221904754639,
|
474 |
+
"learning_rate": 4.813131313131313e-05,
|
475 |
+
"loss": 0.0431,
|
476 |
+
"step": 295
|
477 |
+
},
|
478 |
+
{
|
479 |
+
"epoch": 6.818181818181818,
|
480 |
+
"grad_norm": 0.7969300746917725,
|
481 |
+
"learning_rate": 4.8005050505050505e-05,
|
482 |
+
"loss": 0.2299,
|
483 |
+
"step": 300
|
484 |
+
},
|
485 |
+
{
|
486 |
+
"epoch": 6.931818181818182,
|
487 |
+
"grad_norm": 0.36333414912223816,
|
488 |
+
"learning_rate": 4.787878787878788e-05,
|
489 |
+
"loss": 0.0158,
|
490 |
+
"step": 305
|
491 |
+
},
|
492 |
+
{
|
493 |
+
"epoch": 7.0,
|
494 |
+
"eval_accuracy": 0.795,
|
495 |
+
"eval_loss": 0.7636151313781738,
|
496 |
+
"eval_runtime": 18.1881,
|
497 |
+
"eval_samples_per_second": 10.996,
|
498 |
+
"eval_steps_per_second": 0.715,
|
499 |
+
"step": 308
|
500 |
+
},
|
501 |
+
{
|
502 |
+
"epoch": 7.045454545454546,
|
503 |
+
"grad_norm": 0.7603231072425842,
|
504 |
+
"learning_rate": 4.775252525252526e-05,
|
505 |
+
"loss": 0.0222,
|
506 |
+
"step": 310
|
507 |
+
},
|
508 |
+
{
|
509 |
+
"epoch": 7.159090909090909,
|
510 |
+
"grad_norm": 0.044530898332595825,
|
511 |
+
"learning_rate": 4.762626262626263e-05,
|
512 |
+
"loss": 0.0114,
|
513 |
+
"step": 315
|
514 |
+
},
|
515 |
+
{
|
516 |
+
"epoch": 7.2727272727272725,
|
517 |
+
"grad_norm": 13.010513305664062,
|
518 |
+
"learning_rate": 4.75e-05,
|
519 |
+
"loss": 0.0344,
|
520 |
+
"step": 320
|
521 |
+
},
|
522 |
+
{
|
523 |
+
"epoch": 7.386363636363637,
|
524 |
+
"grad_norm": 1.4342093467712402,
|
525 |
+
"learning_rate": 4.7373737373737375e-05,
|
526 |
+
"loss": 0.0254,
|
527 |
+
"step": 325
|
528 |
+
},
|
529 |
+
{
|
530 |
+
"epoch": 7.5,
|
531 |
+
"grad_norm": 9.25933837890625,
|
532 |
+
"learning_rate": 4.7247474747474755e-05,
|
533 |
+
"loss": 0.0171,
|
534 |
+
"step": 330
|
535 |
+
},
|
536 |
+
{
|
537 |
+
"epoch": 7.613636363636363,
|
538 |
+
"grad_norm": 0.09031987190246582,
|
539 |
+
"learning_rate": 4.712121212121212e-05,
|
540 |
+
"loss": 0.0023,
|
541 |
+
"step": 335
|
542 |
+
},
|
543 |
+
{
|
544 |
+
"epoch": 7.7272727272727275,
|
545 |
+
"grad_norm": 0.03679551184177399,
|
546 |
+
"learning_rate": 4.69949494949495e-05,
|
547 |
+
"loss": 0.0032,
|
548 |
+
"step": 340
|
549 |
+
},
|
550 |
+
{
|
551 |
+
"epoch": 7.840909090909091,
|
552 |
+
"grad_norm": 2.662745237350464,
|
553 |
+
"learning_rate": 4.686868686868687e-05,
|
554 |
+
"loss": 0.0048,
|
555 |
+
"step": 345
|
556 |
+
},
|
557 |
+
{
|
558 |
+
"epoch": 7.954545454545455,
|
559 |
+
"grad_norm": 0.17469650506973267,
|
560 |
+
"learning_rate": 4.6742424242424245e-05,
|
561 |
+
"loss": 0.0009,
|
562 |
+
"step": 350
|
563 |
+
},
|
564 |
+
{
|
565 |
+
"epoch": 8.0,
|
566 |
+
"eval_accuracy": 0.84,
|
567 |
+
"eval_loss": 0.8234661817550659,
|
568 |
+
"eval_runtime": 18.0311,
|
569 |
+
"eval_samples_per_second": 11.092,
|
570 |
+
"eval_steps_per_second": 0.721,
|
571 |
+
"step": 352
|
572 |
+
},
|
573 |
+
{
|
574 |
+
"epoch": 8.068181818181818,
|
575 |
+
"grad_norm": 0.034661825746297836,
|
576 |
+
"learning_rate": 4.661616161616162e-05,
|
577 |
+
"loss": 0.014,
|
578 |
+
"step": 355
|
579 |
+
},
|
580 |
+
{
|
581 |
+
"epoch": 8.181818181818182,
|
582 |
+
"grad_norm": 0.06576091796159744,
|
583 |
+
"learning_rate": 4.6489898989899e-05,
|
584 |
+
"loss": 0.0007,
|
585 |
+
"step": 360
|
586 |
+
},
|
587 |
+
{
|
588 |
+
"epoch": 8.295454545454545,
|
589 |
+
"grad_norm": 0.016597317531704903,
|
590 |
+
"learning_rate": 4.636363636363636e-05,
|
591 |
+
"loss": 0.002,
|
592 |
+
"step": 365
|
593 |
+
},
|
594 |
+
{
|
595 |
+
"epoch": 8.409090909090908,
|
596 |
+
"grad_norm": 0.013650023378431797,
|
597 |
+
"learning_rate": 4.623737373737374e-05,
|
598 |
+
"loss": 0.0015,
|
599 |
+
"step": 370
|
600 |
+
},
|
601 |
+
{
|
602 |
+
"epoch": 8.522727272727273,
|
603 |
+
"grad_norm": 0.19399936497211456,
|
604 |
+
"learning_rate": 4.6111111111111115e-05,
|
605 |
+
"loss": 0.0069,
|
606 |
+
"step": 375
|
607 |
+
},
|
608 |
+
{
|
609 |
+
"epoch": 8.636363636363637,
|
610 |
+
"grad_norm": 0.05494086816906929,
|
611 |
+
"learning_rate": 4.598484848484849e-05,
|
612 |
+
"loss": 0.0026,
|
613 |
+
"step": 380
|
614 |
+
},
|
615 |
+
{
|
616 |
+
"epoch": 8.75,
|
617 |
+
"grad_norm": 0.013664503581821918,
|
618 |
+
"learning_rate": 4.585858585858586e-05,
|
619 |
+
"loss": 0.0012,
|
620 |
+
"step": 385
|
621 |
+
},
|
622 |
+
{
|
623 |
+
"epoch": 8.863636363636363,
|
624 |
+
"grad_norm": 0.025454234331846237,
|
625 |
+
"learning_rate": 4.573232323232323e-05,
|
626 |
+
"loss": 0.0006,
|
627 |
+
"step": 390
|
628 |
+
},
|
629 |
+
{
|
630 |
+
"epoch": 8.977272727272727,
|
631 |
+
"grad_norm": 2.3740947246551514,
|
632 |
+
"learning_rate": 4.5606060606060606e-05,
|
633 |
+
"loss": 0.0029,
|
634 |
+
"step": 395
|
635 |
+
},
|
636 |
+
{
|
637 |
+
"epoch": 9.0,
|
638 |
+
"eval_accuracy": 0.83,
|
639 |
+
"eval_loss": 0.892738401889801,
|
640 |
+
"eval_runtime": 18.123,
|
641 |
+
"eval_samples_per_second": 11.036,
|
642 |
+
"eval_steps_per_second": 0.717,
|
643 |
+
"step": 396
|
644 |
+
},
|
645 |
+
{
|
646 |
+
"epoch": 9.090909090909092,
|
647 |
+
"grad_norm": 0.20694267749786377,
|
648 |
+
"learning_rate": 4.5479797979797985e-05,
|
649 |
+
"loss": 0.001,
|
650 |
+
"step": 400
|
651 |
+
},
|
652 |
+
{
|
653 |
+
"epoch": 9.204545454545455,
|
654 |
+
"grad_norm": 0.07484059035778046,
|
655 |
+
"learning_rate": 4.535353535353535e-05,
|
656 |
+
"loss": 0.0027,
|
657 |
+
"step": 405
|
658 |
+
},
|
659 |
+
{
|
660 |
+
"epoch": 9.318181818181818,
|
661 |
+
"grad_norm": 0.1402018964290619,
|
662 |
+
"learning_rate": 4.522727272727273e-05,
|
663 |
+
"loss": 0.1035,
|
664 |
+
"step": 410
|
665 |
+
},
|
666 |
+
{
|
667 |
+
"epoch": 9.431818181818182,
|
668 |
+
"grad_norm": 0.011067689396440983,
|
669 |
+
"learning_rate": 4.51010101010101e-05,
|
670 |
+
"loss": 0.0145,
|
671 |
+
"step": 415
|
672 |
+
},
|
673 |
+
{
|
674 |
+
"epoch": 9.545454545454545,
|
675 |
+
"grad_norm": 6.280289649963379,
|
676 |
+
"learning_rate": 4.4974747474747476e-05,
|
677 |
+
"loss": 0.0097,
|
678 |
+
"step": 420
|
679 |
+
},
|
680 |
+
{
|
681 |
+
"epoch": 9.659090909090908,
|
682 |
+
"grad_norm": 0.018114736303687096,
|
683 |
+
"learning_rate": 4.484848484848485e-05,
|
684 |
+
"loss": 0.0006,
|
685 |
+
"step": 425
|
686 |
+
},
|
687 |
+
{
|
688 |
+
"epoch": 9.772727272727273,
|
689 |
+
"grad_norm": 11.336344718933105,
|
690 |
+
"learning_rate": 4.472222222222223e-05,
|
691 |
+
"loss": 0.0134,
|
692 |
+
"step": 430
|
693 |
+
},
|
694 |
+
{
|
695 |
+
"epoch": 9.886363636363637,
|
696 |
+
"grad_norm": 0.30804646015167236,
|
697 |
+
"learning_rate": 4.4595959595959594e-05,
|
698 |
+
"loss": 0.1245,
|
699 |
+
"step": 435
|
700 |
+
},
|
701 |
+
{
|
702 |
+
"epoch": 10.0,
|
703 |
+
"grad_norm": 6.994472503662109,
|
704 |
+
"learning_rate": 4.4469696969696973e-05,
|
705 |
+
"loss": 0.0522,
|
706 |
+
"step": 440
|
707 |
+
},
|
708 |
+
{
|
709 |
+
"epoch": 10.0,
|
710 |
+
"eval_accuracy": 0.775,
|
711 |
+
"eval_loss": 1.2321974039077759,
|
712 |
+
"eval_runtime": 18.1204,
|
713 |
+
"eval_samples_per_second": 11.037,
|
714 |
+
"eval_steps_per_second": 0.717,
|
715 |
+
"step": 440
|
716 |
+
},
|
717 |
+
{
|
718 |
+
"epoch": 10.113636363636363,
|
719 |
+
"grad_norm": 2.2376856803894043,
|
720 |
+
"learning_rate": 4.4343434343434346e-05,
|
721 |
+
"loss": 0.0934,
|
722 |
+
"step": 445
|
723 |
+
},
|
724 |
+
{
|
725 |
+
"epoch": 10.227272727272727,
|
726 |
+
"grad_norm": 0.008311591111123562,
|
727 |
+
"learning_rate": 4.421717171717172e-05,
|
728 |
+
"loss": 0.0298,
|
729 |
+
"step": 450
|
730 |
+
},
|
731 |
+
{
|
732 |
+
"epoch": 10.340909090909092,
|
733 |
+
"grad_norm": 21.288654327392578,
|
734 |
+
"learning_rate": 4.409090909090909e-05,
|
735 |
+
"loss": 0.1688,
|
736 |
+
"step": 455
|
737 |
+
},
|
738 |
+
{
|
739 |
+
"epoch": 10.454545454545455,
|
740 |
+
"grad_norm": 16.286630630493164,
|
741 |
+
"learning_rate": 4.396464646464647e-05,
|
742 |
+
"loss": 0.09,
|
743 |
+
"step": 460
|
744 |
+
},
|
745 |
+
{
|
746 |
+
"epoch": 10.568181818181818,
|
747 |
+
"grad_norm": 0.036682695150375366,
|
748 |
+
"learning_rate": 4.383838383838384e-05,
|
749 |
+
"loss": 0.0326,
|
750 |
+
"step": 465
|
751 |
+
},
|
752 |
+
{
|
753 |
+
"epoch": 10.681818181818182,
|
754 |
+
"grad_norm": 0.0594622865319252,
|
755 |
+
"learning_rate": 4.3712121212121216e-05,
|
756 |
+
"loss": 0.0026,
|
757 |
+
"step": 470
|
758 |
+
},
|
759 |
+
{
|
760 |
+
"epoch": 10.795454545454545,
|
761 |
+
"grad_norm": 11.812299728393555,
|
762 |
+
"learning_rate": 4.358585858585859e-05,
|
763 |
+
"loss": 0.0363,
|
764 |
+
"step": 475
|
765 |
+
},
|
766 |
+
{
|
767 |
+
"epoch": 10.909090909090908,
|
768 |
+
"grad_norm": 3.5540413856506348,
|
769 |
+
"learning_rate": 4.345959595959596e-05,
|
770 |
+
"loss": 0.0032,
|
771 |
+
"step": 480
|
772 |
+
},
|
773 |
+
{
|
774 |
+
"epoch": 11.0,
|
775 |
+
"eval_accuracy": 0.765,
|
776 |
+
"eval_loss": 1.2396481037139893,
|
777 |
+
"eval_runtime": 18.1304,
|
778 |
+
"eval_samples_per_second": 11.031,
|
779 |
+
"eval_steps_per_second": 0.717,
|
780 |
+
"step": 484
|
781 |
+
},
|
782 |
+
{
|
783 |
+
"epoch": 11.022727272727273,
|
784 |
+
"grad_norm": 0.7457100749015808,
|
785 |
+
"learning_rate": 4.3333333333333334e-05,
|
786 |
+
"loss": 0.0114,
|
787 |
+
"step": 485
|
788 |
+
},
|
789 |
+
{
|
790 |
+
"epoch": 11.136363636363637,
|
791 |
+
"grad_norm": 1.6869572401046753,
|
792 |
+
"learning_rate": 4.320707070707071e-05,
|
793 |
+
"loss": 0.0147,
|
794 |
+
"step": 490
|
795 |
+
},
|
796 |
+
{
|
797 |
+
"epoch": 11.25,
|
798 |
+
"grad_norm": 0.05314645916223526,
|
799 |
+
"learning_rate": 4.308080808080808e-05,
|
800 |
+
"loss": 0.0137,
|
801 |
+
"step": 495
|
802 |
+
},
|
803 |
+
{
|
804 |
+
"epoch": 11.363636363636363,
|
805 |
+
"grad_norm": 0.05840263515710831,
|
806 |
+
"learning_rate": 4.295454545454546e-05,
|
807 |
+
"loss": 0.1057,
|
808 |
+
"step": 500
|
809 |
+
},
|
810 |
+
{
|
811 |
+
"epoch": 11.477272727272727,
|
812 |
+
"grad_norm": 17.320690155029297,
|
813 |
+
"learning_rate": 4.282828282828283e-05,
|
814 |
+
"loss": 0.0669,
|
815 |
+
"step": 505
|
816 |
+
},
|
817 |
+
{
|
818 |
+
"epoch": 11.590909090909092,
|
819 |
+
"grad_norm": 21.867401123046875,
|
820 |
+
"learning_rate": 4.2702020202020204e-05,
|
821 |
+
"loss": 0.0451,
|
822 |
+
"step": 510
|
823 |
+
},
|
824 |
+
{
|
825 |
+
"epoch": 11.704545454545455,
|
826 |
+
"grad_norm": 0.07670444995164871,
|
827 |
+
"learning_rate": 4.257575757575758e-05,
|
828 |
+
"loss": 0.0029,
|
829 |
+
"step": 515
|
830 |
+
},
|
831 |
+
{
|
832 |
+
"epoch": 11.818181818181818,
|
833 |
+
"grad_norm": 0.08975853025913239,
|
834 |
+
"learning_rate": 4.244949494949495e-05,
|
835 |
+
"loss": 0.0045,
|
836 |
+
"step": 520
|
837 |
+
},
|
838 |
+
{
|
839 |
+
"epoch": 11.931818181818182,
|
840 |
+
"grad_norm": 0.46701082587242126,
|
841 |
+
"learning_rate": 4.232323232323233e-05,
|
842 |
+
"loss": 0.0038,
|
843 |
+
"step": 525
|
844 |
+
},
|
845 |
+
{
|
846 |
+
"epoch": 12.0,
|
847 |
+
"eval_accuracy": 0.795,
|
848 |
+
"eval_loss": 0.9945923686027527,
|
849 |
+
"eval_runtime": 18.2396,
|
850 |
+
"eval_samples_per_second": 10.965,
|
851 |
+
"eval_steps_per_second": 0.713,
|
852 |
+
"step": 528
|
853 |
+
},
|
854 |
+
{
|
855 |
+
"epoch": 12.045454545454545,
|
856 |
+
"grad_norm": 0.011377710849046707,
|
857 |
+
"learning_rate": 4.21969696969697e-05,
|
858 |
+
"loss": 0.0256,
|
859 |
+
"step": 530
|
860 |
+
},
|
861 |
+
{
|
862 |
+
"epoch": 12.159090909090908,
|
863 |
+
"grad_norm": 0.018205437809228897,
|
864 |
+
"learning_rate": 4.2070707070707074e-05,
|
865 |
+
"loss": 0.005,
|
866 |
+
"step": 535
|
867 |
+
},
|
868 |
+
{
|
869 |
+
"epoch": 12.272727272727273,
|
870 |
+
"grad_norm": 0.025772707536816597,
|
871 |
+
"learning_rate": 4.194444444444445e-05,
|
872 |
+
"loss": 0.0004,
|
873 |
+
"step": 540
|
874 |
+
},
|
875 |
+
{
|
876 |
+
"epoch": 12.386363636363637,
|
877 |
+
"grad_norm": 1.5446034669876099,
|
878 |
+
"learning_rate": 4.181818181818182e-05,
|
879 |
+
"loss": 0.0429,
|
880 |
+
"step": 545
|
881 |
+
},
|
882 |
+
{
|
883 |
+
"epoch": 12.5,
|
884 |
+
"grad_norm": 0.21388648450374603,
|
885 |
+
"learning_rate": 4.169191919191919e-05,
|
886 |
+
"loss": 0.0051,
|
887 |
+
"step": 550
|
888 |
+
},
|
889 |
+
{
|
890 |
+
"epoch": 12.613636363636363,
|
891 |
+
"grad_norm": 6.640182018280029,
|
892 |
+
"learning_rate": 4.156565656565657e-05,
|
893 |
+
"loss": 0.0343,
|
894 |
+
"step": 555
|
895 |
+
},
|
896 |
+
{
|
897 |
+
"epoch": 12.727272727272727,
|
898 |
+
"grad_norm": 0.2581124007701874,
|
899 |
+
"learning_rate": 4.143939393939394e-05,
|
900 |
+
"loss": 0.0048,
|
901 |
+
"step": 560
|
902 |
+
},
|
903 |
+
{
|
904 |
+
"epoch": 12.840909090909092,
|
905 |
+
"grad_norm": 0.007842369377613068,
|
906 |
+
"learning_rate": 4.131313131313132e-05,
|
907 |
+
"loss": 0.0206,
|
908 |
+
"step": 565
|
909 |
+
},
|
910 |
+
{
|
911 |
+
"epoch": 12.954545454545455,
|
912 |
+
"grad_norm": 18.075056076049805,
|
913 |
+
"learning_rate": 4.118686868686869e-05,
|
914 |
+
"loss": 0.1007,
|
915 |
+
"step": 570
|
916 |
+
},
|
917 |
+
{
|
918 |
+
"epoch": 13.0,
|
919 |
+
"eval_accuracy": 0.76,
|
920 |
+
"eval_loss": 1.2344247102737427,
|
921 |
+
"eval_runtime": 18.5982,
|
922 |
+
"eval_samples_per_second": 10.754,
|
923 |
+
"eval_steps_per_second": 0.699,
|
924 |
+
"step": 572
|
925 |
+
},
|
926 |
+
{
|
927 |
+
"epoch": 13.068181818181818,
|
928 |
+
"grad_norm": 0.03319934383034706,
|
929 |
+
"learning_rate": 4.106060606060606e-05,
|
930 |
+
"loss": 0.0243,
|
931 |
+
"step": 575
|
932 |
+
},
|
933 |
+
{
|
934 |
+
"epoch": 13.181818181818182,
|
935 |
+
"grad_norm": 11.877718925476074,
|
936 |
+
"learning_rate": 4.0934343434343435e-05,
|
937 |
+
"loss": 0.0666,
|
938 |
+
"step": 580
|
939 |
+
},
|
940 |
+
{
|
941 |
+
"epoch": 13.295454545454545,
|
942 |
+
"grad_norm": 0.40111812949180603,
|
943 |
+
"learning_rate": 4.0808080808080814e-05,
|
944 |
+
"loss": 0.0074,
|
945 |
+
"step": 585
|
946 |
+
},
|
947 |
+
{
|
948 |
+
"epoch": 13.409090909090908,
|
949 |
+
"grad_norm": 0.5372756719589233,
|
950 |
+
"learning_rate": 4.068181818181818e-05,
|
951 |
+
"loss": 0.0101,
|
952 |
+
"step": 590
|
953 |
+
},
|
954 |
+
{
|
955 |
+
"epoch": 13.522727272727273,
|
956 |
+
"grad_norm": 12.776512145996094,
|
957 |
+
"learning_rate": 4.055555555555556e-05,
|
958 |
+
"loss": 0.0137,
|
959 |
+
"step": 595
|
960 |
+
},
|
961 |
+
{
|
962 |
+
"epoch": 13.636363636363637,
|
963 |
+
"grad_norm": 1.489402174949646,
|
964 |
+
"learning_rate": 4.042929292929293e-05,
|
965 |
+
"loss": 0.0023,
|
966 |
+
"step": 600
|
967 |
+
},
|
968 |
+
{
|
969 |
+
"epoch": 13.75,
|
970 |
+
"grad_norm": 0.033923253417015076,
|
971 |
+
"learning_rate": 4.0303030303030305e-05,
|
972 |
+
"loss": 0.0375,
|
973 |
+
"step": 605
|
974 |
+
},
|
975 |
+
{
|
976 |
+
"epoch": 13.863636363636363,
|
977 |
+
"grad_norm": 0.015429407358169556,
|
978 |
+
"learning_rate": 4.017676767676768e-05,
|
979 |
+
"loss": 0.0129,
|
980 |
+
"step": 610
|
981 |
+
},
|
982 |
+
{
|
983 |
+
"epoch": 13.977272727272727,
|
984 |
+
"grad_norm": 0.010183623060584068,
|
985 |
+
"learning_rate": 4.005050505050506e-05,
|
986 |
+
"loss": 0.0003,
|
987 |
+
"step": 615
|
988 |
+
},
|
989 |
+
{
|
990 |
+
"epoch": 14.0,
|
991 |
+
"eval_accuracy": 0.785,
|
992 |
+
"eval_loss": 1.003933072090149,
|
993 |
+
"eval_runtime": 18.085,
|
994 |
+
"eval_samples_per_second": 11.059,
|
995 |
+
"eval_steps_per_second": 0.719,
|
996 |
+
"step": 616
|
997 |
+
},
|
998 |
+
{
|
999 |
+
"epoch": 14.090909090909092,
|
1000 |
+
"grad_norm": 0.1973772794008255,
|
1001 |
+
"learning_rate": 3.992424242424242e-05,
|
1002 |
+
"loss": 0.0004,
|
1003 |
+
"step": 620
|
1004 |
+
},
|
1005 |
+
{
|
1006 |
+
"epoch": 14.204545454545455,
|
1007 |
+
"grad_norm": 0.012279081158339977,
|
1008 |
+
"learning_rate": 3.97979797979798e-05,
|
1009 |
+
"loss": 0.0015,
|
1010 |
+
"step": 625
|
1011 |
+
},
|
1012 |
+
{
|
1013 |
+
"epoch": 14.318181818181818,
|
1014 |
+
"grad_norm": 0.02780894935131073,
|
1015 |
+
"learning_rate": 3.967171717171717e-05,
|
1016 |
+
"loss": 0.0009,
|
1017 |
+
"step": 630
|
1018 |
+
},
|
1019 |
+
{
|
1020 |
+
"epoch": 14.431818181818182,
|
1021 |
+
"grad_norm": 0.004000269342213869,
|
1022 |
+
"learning_rate": 3.954545454545455e-05,
|
1023 |
+
"loss": 0.0016,
|
1024 |
+
"step": 635
|
1025 |
+
},
|
1026 |
+
{
|
1027 |
+
"epoch": 14.545454545454545,
|
1028 |
+
"grad_norm": 0.01629430614411831,
|
1029 |
+
"learning_rate": 3.941919191919192e-05,
|
1030 |
+
"loss": 0.0072,
|
1031 |
+
"step": 640
|
1032 |
+
},
|
1033 |
+
{
|
1034 |
+
"epoch": 14.659090909090908,
|
1035 |
+
"grad_norm": 0.023849163204431534,
|
1036 |
+
"learning_rate": 3.929292929292929e-05,
|
1037 |
+
"loss": 0.0007,
|
1038 |
+
"step": 645
|
1039 |
+
},
|
1040 |
+
{
|
1041 |
+
"epoch": 14.772727272727273,
|
1042 |
+
"grad_norm": 0.10356610268354416,
|
1043 |
+
"learning_rate": 3.9166666666666665e-05,
|
1044 |
+
"loss": 0.0026,
|
1045 |
+
"step": 650
|
1046 |
+
},
|
1047 |
+
{
|
1048 |
+
"epoch": 14.886363636363637,
|
1049 |
+
"grad_norm": 0.14715874195098877,
|
1050 |
+
"learning_rate": 3.9040404040404045e-05,
|
1051 |
+
"loss": 0.0003,
|
1052 |
+
"step": 655
|
1053 |
+
},
|
1054 |
+
{
|
1055 |
+
"epoch": 15.0,
|
1056 |
+
"grad_norm": 0.10584244877099991,
|
1057 |
+
"learning_rate": 3.891414141414141e-05,
|
1058 |
+
"loss": 0.0013,
|
1059 |
+
"step": 660
|
1060 |
+
},
|
1061 |
+
{
|
1062 |
+
"epoch": 15.0,
|
1063 |
+
"eval_accuracy": 0.82,
|
1064 |
+
"eval_loss": 1.136525273323059,
|
1065 |
+
"eval_runtime": 18.0084,
|
1066 |
+
"eval_samples_per_second": 11.106,
|
1067 |
+
"eval_steps_per_second": 0.722,
|
1068 |
+
"step": 660
|
1069 |
+
},
|
1070 |
+
{
|
1071 |
+
"epoch": 15.113636363636363,
|
1072 |
+
"grad_norm": 0.03254534304141998,
|
1073 |
+
"learning_rate": 3.878787878787879e-05,
|
1074 |
+
"loss": 0.0008,
|
1075 |
+
"step": 665
|
1076 |
+
},
|
1077 |
+
{
|
1078 |
+
"epoch": 15.227272727272727,
|
1079 |
+
"grad_norm": 0.024025389924645424,
|
1080 |
+
"learning_rate": 3.866161616161616e-05,
|
1081 |
+
"loss": 0.006,
|
1082 |
+
"step": 670
|
1083 |
+
},
|
1084 |
+
{
|
1085 |
+
"epoch": 15.340909090909092,
|
1086 |
+
"grad_norm": 0.0632549524307251,
|
1087 |
+
"learning_rate": 3.8535353535353536e-05,
|
1088 |
+
"loss": 0.0003,
|
1089 |
+
"step": 675
|
1090 |
+
},
|
1091 |
+
{
|
1092 |
+
"epoch": 15.454545454545455,
|
1093 |
+
"grad_norm": 0.1464257687330246,
|
1094 |
+
"learning_rate": 3.840909090909091e-05,
|
1095 |
+
"loss": 0.0046,
|
1096 |
+
"step": 680
|
1097 |
+
},
|
1098 |
+
{
|
1099 |
+
"epoch": 15.568181818181818,
|
1100 |
+
"grad_norm": 0.11179546266794205,
|
1101 |
+
"learning_rate": 3.828282828282829e-05,
|
1102 |
+
"loss": 0.0004,
|
1103 |
+
"step": 685
|
1104 |
+
},
|
1105 |
+
{
|
1106 |
+
"epoch": 15.681818181818182,
|
1107 |
+
"grad_norm": 4.28083610534668,
|
1108 |
+
"learning_rate": 3.815656565656566e-05,
|
1109 |
+
"loss": 0.0582,
|
1110 |
+
"step": 690
|
1111 |
+
},
|
1112 |
+
{
|
1113 |
+
"epoch": 15.795454545454545,
|
1114 |
+
"grad_norm": 0.02816050685942173,
|
1115 |
+
"learning_rate": 3.803030303030303e-05,
|
1116 |
+
"loss": 0.0002,
|
1117 |
+
"step": 695
|
1118 |
+
},
|
1119 |
+
{
|
1120 |
+
"epoch": 15.909090909090908,
|
1121 |
+
"grad_norm": 0.01497215311974287,
|
1122 |
+
"learning_rate": 3.7904040404040406e-05,
|
1123 |
+
"loss": 0.0006,
|
1124 |
+
"step": 700
|
1125 |
+
},
|
1126 |
+
{
|
1127 |
+
"epoch": 16.0,
|
1128 |
+
"eval_accuracy": 0.785,
|
1129 |
+
"eval_loss": 1.4599967002868652,
|
1130 |
+
"eval_runtime": 18.0022,
|
1131 |
+
"eval_samples_per_second": 11.11,
|
1132 |
+
"eval_steps_per_second": 0.722,
|
1133 |
+
"step": 704
|
1134 |
+
},
|
1135 |
+
{
|
1136 |
+
"epoch": 16.022727272727273,
|
1137 |
+
"grad_norm": 0.11498994380235672,
|
1138 |
+
"learning_rate": 3.777777777777778e-05,
|
1139 |
+
"loss": 0.0004,
|
1140 |
+
"step": 705
|
1141 |
+
},
|
1142 |
+
{
|
1143 |
+
"epoch": 16.136363636363637,
|
1144 |
+
"grad_norm": 0.07350458204746246,
|
1145 |
+
"learning_rate": 3.765151515151516e-05,
|
1146 |
+
"loss": 0.0003,
|
1147 |
+
"step": 710
|
1148 |
+
},
|
1149 |
+
{
|
1150 |
+
"epoch": 16.25,
|
1151 |
+
"grad_norm": 0.0034702487755566835,
|
1152 |
+
"learning_rate": 3.7525252525252524e-05,
|
1153 |
+
"loss": 0.0002,
|
1154 |
+
"step": 715
|
1155 |
+
},
|
1156 |
+
{
|
1157 |
+
"epoch": 16.363636363636363,
|
1158 |
+
"grad_norm": 0.02884703502058983,
|
1159 |
+
"learning_rate": 3.73989898989899e-05,
|
1160 |
+
"loss": 0.0006,
|
1161 |
+
"step": 720
|
1162 |
+
},
|
1163 |
+
{
|
1164 |
+
"epoch": 16.477272727272727,
|
1165 |
+
"grad_norm": 0.0020916229113936424,
|
1166 |
+
"learning_rate": 3.7272727272727276e-05,
|
1167 |
+
"loss": 0.0004,
|
1168 |
+
"step": 725
|
1169 |
+
},
|
1170 |
+
{
|
1171 |
+
"epoch": 16.59090909090909,
|
1172 |
+
"grad_norm": 0.056545890867710114,
|
1173 |
+
"learning_rate": 3.714646464646465e-05,
|
1174 |
+
"loss": 0.0005,
|
1175 |
+
"step": 730
|
1176 |
+
},
|
1177 |
+
{
|
1178 |
+
"epoch": 16.704545454545453,
|
1179 |
+
"grad_norm": 0.008786072954535484,
|
1180 |
+
"learning_rate": 3.702020202020202e-05,
|
1181 |
+
"loss": 0.0001,
|
1182 |
+
"step": 735
|
1183 |
+
},
|
1184 |
+
{
|
1185 |
+
"epoch": 16.818181818181817,
|
1186 |
+
"grad_norm": 0.002408586209639907,
|
1187 |
+
"learning_rate": 3.68939393939394e-05,
|
1188 |
+
"loss": 0.0002,
|
1189 |
+
"step": 740
|
1190 |
+
},
|
1191 |
+
{
|
1192 |
+
"epoch": 16.931818181818183,
|
1193 |
+
"grad_norm": 0.002747893799096346,
|
1194 |
+
"learning_rate": 3.6767676767676766e-05,
|
1195 |
+
"loss": 0.0001,
|
1196 |
+
"step": 745
|
1197 |
+
},
|
1198 |
+
{
|
1199 |
+
"epoch": 17.0,
|
1200 |
+
"eval_accuracy": 0.82,
|
1201 |
+
"eval_loss": 0.9243611693382263,
|
1202 |
+
"eval_runtime": 17.9514,
|
1203 |
+
"eval_samples_per_second": 11.141,
|
1204 |
+
"eval_steps_per_second": 0.724,
|
1205 |
+
"step": 748
|
1206 |
+
},
|
1207 |
+
{
|
1208 |
+
"epoch": 17.045454545454547,
|
1209 |
+
"grad_norm": 0.002468683058395982,
|
1210 |
+
"learning_rate": 3.6641414141414146e-05,
|
1211 |
+
"loss": 0.0002,
|
1212 |
+
"step": 750
|
1213 |
+
},
|
1214 |
+
{
|
1215 |
+
"epoch": 17.15909090909091,
|
1216 |
+
"grad_norm": 0.0032050844747573137,
|
1217 |
+
"learning_rate": 3.651515151515152e-05,
|
1218 |
+
"loss": 0.0001,
|
1219 |
+
"step": 755
|
1220 |
+
},
|
1221 |
+
{
|
1222 |
+
"epoch": 17.272727272727273,
|
1223 |
+
"grad_norm": 0.0034459943417459726,
|
1224 |
+
"learning_rate": 3.638888888888889e-05,
|
1225 |
+
"loss": 0.0001,
|
1226 |
+
"step": 760
|
1227 |
+
},
|
1228 |
+
{
|
1229 |
+
"epoch": 17.386363636363637,
|
1230 |
+
"grad_norm": 0.002899475395679474,
|
1231 |
+
"learning_rate": 3.6262626262626264e-05,
|
1232 |
+
"loss": 0.0001,
|
1233 |
+
"step": 765
|
1234 |
+
},
|
1235 |
+
{
|
1236 |
+
"epoch": 17.5,
|
1237 |
+
"grad_norm": 0.0023372883442789316,
|
1238 |
+
"learning_rate": 3.613636363636364e-05,
|
1239 |
+
"loss": 0.0001,
|
1240 |
+
"step": 770
|
1241 |
+
},
|
1242 |
+
{
|
1243 |
+
"epoch": 17.613636363636363,
|
1244 |
+
"grad_norm": 0.0017898937221616507,
|
1245 |
+
"learning_rate": 3.601010101010101e-05,
|
1246 |
+
"loss": 0.0001,
|
1247 |
+
"step": 775
|
1248 |
+
},
|
1249 |
+
{
|
1250 |
+
"epoch": 17.727272727272727,
|
1251 |
+
"grad_norm": 0.004293715115636587,
|
1252 |
+
"learning_rate": 3.588383838383839e-05,
|
1253 |
+
"loss": 0.0001,
|
1254 |
+
"step": 780
|
1255 |
+
},
|
1256 |
+
{
|
1257 |
+
"epoch": 17.84090909090909,
|
1258 |
+
"grad_norm": 0.0015544987982138991,
|
1259 |
+
"learning_rate": 3.575757575757576e-05,
|
1260 |
+
"loss": 0.0001,
|
1261 |
+
"step": 785
|
1262 |
+
},
|
1263 |
+
{
|
1264 |
+
"epoch": 17.954545454545453,
|
1265 |
+
"grad_norm": 0.0016952146543189883,
|
1266 |
+
"learning_rate": 3.5631313131313134e-05,
|
1267 |
+
"loss": 0.0001,
|
1268 |
+
"step": 790
|
1269 |
+
},
|
1270 |
+
{
|
1271 |
+
"epoch": 18.0,
|
1272 |
+
"eval_accuracy": 0.83,
|
1273 |
+
"eval_loss": 0.8526318073272705,
|
1274 |
+
"eval_runtime": 18.3475,
|
1275 |
+
"eval_samples_per_second": 10.901,
|
1276 |
+
"eval_steps_per_second": 0.709,
|
1277 |
+
"step": 792
|
1278 |
+
},
|
1279 |
+
{
|
1280 |
+
"epoch": 18.068181818181817,
|
1281 |
+
"grad_norm": 0.0020155711099505424,
|
1282 |
+
"learning_rate": 3.5505050505050506e-05,
|
1283 |
+
"loss": 0.0001,
|
1284 |
+
"step": 795
|
1285 |
+
},
|
1286 |
+
{
|
1287 |
+
"epoch": 18.181818181818183,
|
1288 |
+
"grad_norm": 0.0013277644757181406,
|
1289 |
+
"learning_rate": 3.537878787878788e-05,
|
1290 |
+
"loss": 0.0001,
|
1291 |
+
"step": 800
|
1292 |
+
},
|
1293 |
+
{
|
1294 |
+
"epoch": 18.295454545454547,
|
1295 |
+
"grad_norm": 0.0020753301214426756,
|
1296 |
+
"learning_rate": 3.525252525252525e-05,
|
1297 |
+
"loss": 0.0001,
|
1298 |
+
"step": 805
|
1299 |
+
},
|
1300 |
+
{
|
1301 |
+
"epoch": 18.40909090909091,
|
1302 |
+
"grad_norm": 0.0018062540329992771,
|
1303 |
+
"learning_rate": 3.512626262626263e-05,
|
1304 |
+
"loss": 0.0001,
|
1305 |
+
"step": 810
|
1306 |
+
},
|
1307 |
+
{
|
1308 |
+
"epoch": 18.522727272727273,
|
1309 |
+
"grad_norm": 0.0017586707836017013,
|
1310 |
+
"learning_rate": 3.5e-05,
|
1311 |
+
"loss": 0.0001,
|
1312 |
+
"step": 815
|
1313 |
+
},
|
1314 |
+
{
|
1315 |
+
"epoch": 18.636363636363637,
|
1316 |
+
"grad_norm": 0.001384661765769124,
|
1317 |
+
"learning_rate": 3.4873737373737376e-05,
|
1318 |
+
"loss": 0.0001,
|
1319 |
+
"step": 820
|
1320 |
+
},
|
1321 |
+
{
|
1322 |
+
"epoch": 18.75,
|
1323 |
+
"grad_norm": 0.001989172538742423,
|
1324 |
+
"learning_rate": 3.474747474747475e-05,
|
1325 |
+
"loss": 0.0001,
|
1326 |
+
"step": 825
|
1327 |
+
},
|
1328 |
+
{
|
1329 |
+
"epoch": 18.863636363636363,
|
1330 |
+
"grad_norm": 0.001609248691238463,
|
1331 |
+
"learning_rate": 3.462121212121212e-05,
|
1332 |
+
"loss": 0.0001,
|
1333 |
+
"step": 830
|
1334 |
+
},
|
1335 |
+
{
|
1336 |
+
"epoch": 18.977272727272727,
|
1337 |
+
"grad_norm": 0.001661130809225142,
|
1338 |
+
"learning_rate": 3.4494949494949494e-05,
|
1339 |
+
"loss": 0.0001,
|
1340 |
+
"step": 835
|
1341 |
+
},
|
1342 |
+
{
|
1343 |
+
"epoch": 19.0,
|
1344 |
+
"eval_accuracy": 0.83,
|
1345 |
+
"eval_loss": 0.8473049998283386,
|
1346 |
+
"eval_runtime": 18.1222,
|
1347 |
+
"eval_samples_per_second": 11.036,
|
1348 |
+
"eval_steps_per_second": 0.717,
|
1349 |
+
"step": 836
|
1350 |
+
},
|
1351 |
+
{
|
1352 |
+
"epoch": 19.09090909090909,
|
1353 |
+
"grad_norm": 0.0013667866587638855,
|
1354 |
+
"learning_rate": 3.4368686868686874e-05,
|
1355 |
+
"loss": 0.0001,
|
1356 |
+
"step": 840
|
1357 |
+
},
|
1358 |
+
{
|
1359 |
+
"epoch": 19.204545454545453,
|
1360 |
+
"grad_norm": 0.001460373867303133,
|
1361 |
+
"learning_rate": 3.424242424242424e-05,
|
1362 |
+
"loss": 0.0001,
|
1363 |
+
"step": 845
|
1364 |
+
},
|
1365 |
+
{
|
1366 |
+
"epoch": 19.318181818181817,
|
1367 |
+
"grad_norm": 0.001786314183846116,
|
1368 |
+
"learning_rate": 3.411616161616162e-05,
|
1369 |
+
"loss": 0.0001,
|
1370 |
+
"step": 850
|
1371 |
+
},
|
1372 |
+
{
|
1373 |
+
"epoch": 19.431818181818183,
|
1374 |
+
"grad_norm": 0.0021370083559304476,
|
1375 |
+
"learning_rate": 3.398989898989899e-05,
|
1376 |
+
"loss": 0.0001,
|
1377 |
+
"step": 855
|
1378 |
+
},
|
1379 |
+
{
|
1380 |
+
"epoch": 19.545454545454547,
|
1381 |
+
"grad_norm": 0.0021769488230347633,
|
1382 |
+
"learning_rate": 3.3863636363636364e-05,
|
1383 |
+
"loss": 0.0001,
|
1384 |
+
"step": 860
|
1385 |
+
},
|
1386 |
+
{
|
1387 |
+
"epoch": 19.65909090909091,
|
1388 |
+
"grad_norm": 0.002502184361219406,
|
1389 |
+
"learning_rate": 3.373737373737374e-05,
|
1390 |
+
"loss": 0.0001,
|
1391 |
+
"step": 865
|
1392 |
+
},
|
1393 |
+
{
|
1394 |
+
"epoch": 19.772727272727273,
|
1395 |
+
"grad_norm": 0.0011145217576995492,
|
1396 |
+
"learning_rate": 3.3611111111111116e-05,
|
1397 |
+
"loss": 0.0001,
|
1398 |
+
"step": 870
|
1399 |
+
},
|
1400 |
+
{
|
1401 |
+
"epoch": 19.886363636363637,
|
1402 |
+
"grad_norm": 0.0014179410645738244,
|
1403 |
+
"learning_rate": 3.348484848484848e-05,
|
1404 |
+
"loss": 0.0001,
|
1405 |
+
"step": 875
|
1406 |
+
},
|
1407 |
+
{
|
1408 |
+
"epoch": 20.0,
|
1409 |
+
"grad_norm": 0.0021014176309108734,
|
1410 |
+
"learning_rate": 3.335858585858586e-05,
|
1411 |
+
"loss": 0.0001,
|
1412 |
+
"step": 880
|
1413 |
+
},
|
1414 |
+
{
|
1415 |
+
"epoch": 20.0,
|
1416 |
+
"eval_accuracy": 0.84,
|
1417 |
+
"eval_loss": 0.8451841473579407,
|
1418 |
+
"eval_runtime": 17.9811,
|
1419 |
+
"eval_samples_per_second": 11.123,
|
1420 |
+
"eval_steps_per_second": 0.723,
|
1421 |
+
"step": 880
|
1422 |
+
},
|
1423 |
+
{
|
1424 |
+
"epoch": 20.113636363636363,
|
1425 |
+
"grad_norm": 0.002036330057308078,
|
1426 |
+
"learning_rate": 3.3232323232323234e-05,
|
1427 |
+
"loss": 0.0001,
|
1428 |
+
"step": 885
|
1429 |
+
},
|
1430 |
+
{
|
1431 |
+
"epoch": 20.227272727272727,
|
1432 |
+
"grad_norm": 0.003770321374759078,
|
1433 |
+
"learning_rate": 3.310606060606061e-05,
|
1434 |
+
"loss": 0.0001,
|
1435 |
+
"step": 890
|
1436 |
+
},
|
1437 |
+
{
|
1438 |
+
"epoch": 20.34090909090909,
|
1439 |
+
"grad_norm": 0.0016103875823318958,
|
1440 |
+
"learning_rate": 3.297979797979798e-05,
|
1441 |
+
"loss": 0.0001,
|
1442 |
+
"step": 895
|
1443 |
+
},
|
1444 |
+
{
|
1445 |
+
"epoch": 20.454545454545453,
|
1446 |
+
"grad_norm": 0.001122700166888535,
|
1447 |
+
"learning_rate": 3.285353535353535e-05,
|
1448 |
+
"loss": 0.0001,
|
1449 |
+
"step": 900
|
1450 |
+
},
|
1451 |
+
{
|
1452 |
+
"epoch": 20.568181818181817,
|
1453 |
+
"grad_norm": 0.0010796000715345144,
|
1454 |
+
"learning_rate": 3.272727272727273e-05,
|
1455 |
+
"loss": 0.0001,
|
1456 |
+
"step": 905
|
1457 |
+
},
|
1458 |
+
{
|
1459 |
+
"epoch": 20.681818181818183,
|
1460 |
+
"grad_norm": 0.0008875965140759945,
|
1461 |
+
"learning_rate": 3.2601010101010104e-05,
|
1462 |
+
"loss": 0.0001,
|
1463 |
+
"step": 910
|
1464 |
+
},
|
1465 |
+
{
|
1466 |
+
"epoch": 20.795454545454547,
|
1467 |
+
"grad_norm": 0.0022327785845845938,
|
1468 |
+
"learning_rate": 3.247474747474748e-05,
|
1469 |
+
"loss": 0.0001,
|
1470 |
+
"step": 915
|
1471 |
+
},
|
1472 |
+
{
|
1473 |
+
"epoch": 20.90909090909091,
|
1474 |
+
"grad_norm": 0.002081720856949687,
|
1475 |
+
"learning_rate": 3.234848484848485e-05,
|
1476 |
+
"loss": 0.0001,
|
1477 |
+
"step": 920
|
1478 |
+
},
|
1479 |
+
{
|
1480 |
+
"epoch": 21.0,
|
1481 |
+
"eval_accuracy": 0.845,
|
1482 |
+
"eval_loss": 0.8451855182647705,
|
1483 |
+
"eval_runtime": 18.2346,
|
1484 |
+
"eval_samples_per_second": 10.968,
|
1485 |
+
"eval_steps_per_second": 0.713,
|
1486 |
+
"step": 924
|
1487 |
+
},
|
1488 |
+
{
|
1489 |
+
"epoch": 21.022727272727273,
|
1490 |
+
"grad_norm": 0.0012421772116795182,
|
1491 |
+
"learning_rate": 3.222222222222223e-05,
|
1492 |
+
"loss": 0.0001,
|
1493 |
+
"step": 925
|
1494 |
+
},
|
1495 |
+
{
|
1496 |
+
"epoch": 21.136363636363637,
|
1497 |
+
"grad_norm": 0.0009116244618780911,
|
1498 |
+
"learning_rate": 3.2095959595959595e-05,
|
1499 |
+
"loss": 0.0001,
|
1500 |
+
"step": 930
|
1501 |
+
},
|
1502 |
+
{
|
1503 |
+
"epoch": 21.25,
|
1504 |
+
"grad_norm": 0.0018600281327962875,
|
1505 |
+
"learning_rate": 3.1969696969696974e-05,
|
1506 |
+
"loss": 0.0001,
|
1507 |
+
"step": 935
|
1508 |
+
},
|
1509 |
+
{
|
1510 |
+
"epoch": 21.363636363636363,
|
1511 |
+
"grad_norm": 0.0009139208123087883,
|
1512 |
+
"learning_rate": 3.184343434343435e-05,
|
1513 |
+
"loss": 0.0001,
|
1514 |
+
"step": 940
|
1515 |
+
},
|
1516 |
+
{
|
1517 |
+
"epoch": 21.477272727272727,
|
1518 |
+
"grad_norm": 0.0014584549935534596,
|
1519 |
+
"learning_rate": 3.171717171717172e-05,
|
1520 |
+
"loss": 0.0001,
|
1521 |
+
"step": 945
|
1522 |
+
},
|
1523 |
+
{
|
1524 |
+
"epoch": 21.59090909090909,
|
1525 |
+
"grad_norm": 0.0014573056250810623,
|
1526 |
+
"learning_rate": 3.159090909090909e-05,
|
1527 |
+
"loss": 0.0001,
|
1528 |
+
"step": 950
|
1529 |
+
},
|
1530 |
+
{
|
1531 |
+
"epoch": 21.704545454545453,
|
1532 |
+
"grad_norm": 0.0017861599335446954,
|
1533 |
+
"learning_rate": 3.146464646464647e-05,
|
1534 |
+
"loss": 0.0001,
|
1535 |
+
"step": 955
|
1536 |
+
},
|
1537 |
+
{
|
1538 |
+
"epoch": 21.818181818181817,
|
1539 |
+
"grad_norm": 0.0017260201275348663,
|
1540 |
+
"learning_rate": 3.133838383838384e-05,
|
1541 |
+
"loss": 0.0001,
|
1542 |
+
"step": 960
|
1543 |
+
},
|
1544 |
+
{
|
1545 |
+
"epoch": 21.931818181818183,
|
1546 |
+
"grad_norm": 0.0017356129828840494,
|
1547 |
+
"learning_rate": 3.121212121212122e-05,
|
1548 |
+
"loss": 0.0001,
|
1549 |
+
"step": 965
|
1550 |
+
},
|
1551 |
+
{
|
1552 |
+
"epoch": 22.0,
|
1553 |
+
"eval_accuracy": 0.845,
|
1554 |
+
"eval_loss": 0.8445044159889221,
|
1555 |
+
"eval_runtime": 18.0402,
|
1556 |
+
"eval_samples_per_second": 11.086,
|
1557 |
+
"eval_steps_per_second": 0.721,
|
1558 |
+
"step": 968
|
1559 |
+
},
|
1560 |
+
{
|
1561 |
+
"epoch": 22.045454545454547,
|
1562 |
+
"grad_norm": 0.0013173171319067478,
|
1563 |
+
"learning_rate": 3.108585858585858e-05,
|
1564 |
+
"loss": 0.0001,
|
1565 |
+
"step": 970
|
1566 |
+
},
|
1567 |
+
{
|
1568 |
+
"epoch": 22.15909090909091,
|
1569 |
+
"grad_norm": 0.0018760478124022484,
|
1570 |
+
"learning_rate": 3.095959595959596e-05,
|
1571 |
+
"loss": 0.0001,
|
1572 |
+
"step": 975
|
1573 |
+
},
|
1574 |
+
{
|
1575 |
+
"epoch": 22.272727272727273,
|
1576 |
+
"grad_norm": 0.0007338706054724753,
|
1577 |
+
"learning_rate": 3.0833333333333335e-05,
|
1578 |
+
"loss": 0.0,
|
1579 |
+
"step": 980
|
1580 |
+
},
|
1581 |
+
{
|
1582 |
+
"epoch": 22.386363636363637,
|
1583 |
+
"grad_norm": 0.000914990552701056,
|
1584 |
+
"learning_rate": 3.070707070707071e-05,
|
1585 |
+
"loss": 0.0,
|
1586 |
+
"step": 985
|
1587 |
+
},
|
1588 |
+
{
|
1589 |
+
"epoch": 22.5,
|
1590 |
+
"grad_norm": 0.0018791673937812448,
|
1591 |
+
"learning_rate": 3.058080808080808e-05,
|
1592 |
+
"loss": 0.0001,
|
1593 |
+
"step": 990
|
1594 |
+
},
|
1595 |
+
{
|
1596 |
+
"epoch": 22.613636363636363,
|
1597 |
+
"grad_norm": 0.0010454360162839293,
|
1598 |
+
"learning_rate": 3.0454545454545456e-05,
|
1599 |
+
"loss": 0.0001,
|
1600 |
+
"step": 995
|
1601 |
+
},
|
1602 |
+
{
|
1603 |
+
"epoch": 22.727272727272727,
|
1604 |
+
"grad_norm": 0.0019519156776368618,
|
1605 |
+
"learning_rate": 3.032828282828283e-05,
|
1606 |
+
"loss": 0.0001,
|
1607 |
+
"step": 1000
|
1608 |
+
},
|
1609 |
+
{
|
1610 |
+
"epoch": 22.84090909090909,
|
1611 |
+
"grad_norm": 0.0013823095941916108,
|
1612 |
+
"learning_rate": 3.0202020202020205e-05,
|
1613 |
+
"loss": 0.0,
|
1614 |
+
"step": 1005
|
1615 |
+
},
|
1616 |
+
{
|
1617 |
+
"epoch": 22.954545454545453,
|
1618 |
+
"grad_norm": 0.0008434047340415418,
|
1619 |
+
"learning_rate": 3.0075757575757578e-05,
|
1620 |
+
"loss": 0.0001,
|
1621 |
+
"step": 1010
|
1622 |
+
},
|
1623 |
+
{
|
1624 |
+
"epoch": 23.0,
|
1625 |
+
"eval_accuracy": 0.85,
|
1626 |
+
"eval_loss": 0.8445219397544861,
|
1627 |
+
"eval_runtime": 18.3375,
|
1628 |
+
"eval_samples_per_second": 10.907,
|
1629 |
+
"eval_steps_per_second": 0.709,
|
1630 |
+
"step": 1012
|
1631 |
+
},
|
1632 |
+
{
|
1633 |
+
"epoch": 23.068181818181817,
|
1634 |
+
"grad_norm": 0.0017377930926159024,
|
1635 |
+
"learning_rate": 2.994949494949495e-05,
|
1636 |
+
"loss": 0.0,
|
1637 |
+
"step": 1015
|
1638 |
+
},
|
1639 |
+
{
|
1640 |
+
"epoch": 23.181818181818183,
|
1641 |
+
"grad_norm": 0.001168358838185668,
|
1642 |
+
"learning_rate": 2.9823232323232327e-05,
|
1643 |
+
"loss": 0.0001,
|
1644 |
+
"step": 1020
|
1645 |
+
},
|
1646 |
+
{
|
1647 |
+
"epoch": 23.295454545454547,
|
1648 |
+
"grad_norm": 0.001191208721138537,
|
1649 |
+
"learning_rate": 2.96969696969697e-05,
|
1650 |
+
"loss": 0.0,
|
1651 |
+
"step": 1025
|
1652 |
+
},
|
1653 |
+
{
|
1654 |
+
"epoch": 23.40909090909091,
|
1655 |
+
"grad_norm": 0.0011305585503578186,
|
1656 |
+
"learning_rate": 2.9570707070707072e-05,
|
1657 |
+
"loss": 0.0001,
|
1658 |
+
"step": 1030
|
1659 |
+
},
|
1660 |
+
{
|
1661 |
+
"epoch": 23.522727272727273,
|
1662 |
+
"grad_norm": 0.0014099575346335769,
|
1663 |
+
"learning_rate": 2.9444444444444448e-05,
|
1664 |
+
"loss": 0.0001,
|
1665 |
+
"step": 1035
|
1666 |
+
},
|
1667 |
+
{
|
1668 |
+
"epoch": 23.636363636363637,
|
1669 |
+
"grad_norm": 0.001969902543351054,
|
1670 |
+
"learning_rate": 2.9318181818181817e-05,
|
1671 |
+
"loss": 0.0,
|
1672 |
+
"step": 1040
|
1673 |
+
},
|
1674 |
+
{
|
1675 |
+
"epoch": 23.75,
|
1676 |
+
"grad_norm": 0.0011912737973034382,
|
1677 |
+
"learning_rate": 2.9191919191919193e-05,
|
1678 |
+
"loss": 0.0,
|
1679 |
+
"step": 1045
|
1680 |
+
},
|
1681 |
+
{
|
1682 |
+
"epoch": 23.863636363636363,
|
1683 |
+
"grad_norm": 0.0008956310921348631,
|
1684 |
+
"learning_rate": 2.906565656565657e-05,
|
1685 |
+
"loss": 0.0,
|
1686 |
+
"step": 1050
|
1687 |
+
},
|
1688 |
+
{
|
1689 |
+
"epoch": 23.977272727272727,
|
1690 |
+
"grad_norm": 0.001731114462018013,
|
1691 |
+
"learning_rate": 2.893939393939394e-05,
|
1692 |
+
"loss": 0.0,
|
1693 |
+
"step": 1055
|
1694 |
+
},
|
1695 |
+
{
|
1696 |
+
"epoch": 24.0,
|
1697 |
+
"eval_accuracy": 0.85,
|
1698 |
+
"eval_loss": 0.8457598686218262,
|
1699 |
+
"eval_runtime": 18.3475,
|
1700 |
+
"eval_samples_per_second": 10.901,
|
1701 |
+
"eval_steps_per_second": 0.709,
|
1702 |
+
"step": 1056
|
1703 |
+
},
|
1704 |
+
{
|
1705 |
+
"epoch": 24.09090909090909,
|
1706 |
+
"grad_norm": 0.000823634210973978,
|
1707 |
+
"learning_rate": 2.8813131313131315e-05,
|
1708 |
+
"loss": 0.0,
|
1709 |
+
"step": 1060
|
1710 |
+
},
|
1711 |
+
{
|
1712 |
+
"epoch": 24.204545454545453,
|
1713 |
+
"grad_norm": 0.002404524013400078,
|
1714 |
+
"learning_rate": 2.868686868686869e-05,
|
1715 |
+
"loss": 0.0001,
|
1716 |
+
"step": 1065
|
1717 |
+
},
|
1718 |
+
{
|
1719 |
+
"epoch": 24.318181818181817,
|
1720 |
+
"grad_norm": 0.0007218755781650543,
|
1721 |
+
"learning_rate": 2.856060606060606e-05,
|
1722 |
+
"loss": 0.0,
|
1723 |
+
"step": 1070
|
1724 |
+
},
|
1725 |
+
{
|
1726 |
+
"epoch": 24.431818181818183,
|
1727 |
+
"grad_norm": 0.001373838516883552,
|
1728 |
+
"learning_rate": 2.8434343434343436e-05,
|
1729 |
+
"loss": 0.0,
|
1730 |
+
"step": 1075
|
1731 |
+
},
|
1732 |
+
{
|
1733 |
+
"epoch": 24.545454545454547,
|
1734 |
+
"grad_norm": 0.0016993086319416761,
|
1735 |
+
"learning_rate": 2.8308080808080812e-05,
|
1736 |
+
"loss": 0.0,
|
1737 |
+
"step": 1080
|
1738 |
+
},
|
1739 |
+
{
|
1740 |
+
"epoch": 24.65909090909091,
|
1741 |
+
"grad_norm": 0.0014250794192776084,
|
1742 |
+
"learning_rate": 2.818181818181818e-05,
|
1743 |
+
"loss": 0.0,
|
1744 |
+
"step": 1085
|
1745 |
+
},
|
1746 |
+
{
|
1747 |
+
"epoch": 24.772727272727273,
|
1748 |
+
"grad_norm": 0.001747175119817257,
|
1749 |
+
"learning_rate": 2.8055555555555557e-05,
|
1750 |
+
"loss": 0.0,
|
1751 |
+
"step": 1090
|
1752 |
+
},
|
1753 |
+
{
|
1754 |
+
"epoch": 24.886363636363637,
|
1755 |
+
"grad_norm": 0.001142960973083973,
|
1756 |
+
"learning_rate": 2.7929292929292933e-05,
|
1757 |
+
"loss": 0.0,
|
1758 |
+
"step": 1095
|
1759 |
+
},
|
1760 |
+
{
|
1761 |
+
"epoch": 25.0,
|
1762 |
+
"grad_norm": 0.0012541130417957902,
|
1763 |
+
"learning_rate": 2.7803030303030303e-05,
|
1764 |
+
"loss": 0.0,
|
1765 |
+
"step": 1100
|
1766 |
+
},
|
1767 |
+
{
|
1768 |
+
"epoch": 25.0,
|
1769 |
+
"eval_accuracy": 0.855,
|
1770 |
+
"eval_loss": 0.8466522097587585,
|
1771 |
+
"eval_runtime": 18.1075,
|
1772 |
+
"eval_samples_per_second": 11.045,
|
1773 |
+
"eval_steps_per_second": 0.718,
|
1774 |
+
"step": 1100
|
1775 |
+
},
|
1776 |
+
{
|
1777 |
+
"epoch": 25.113636363636363,
|
1778 |
+
"grad_norm": 0.0014029736630618572,
|
1779 |
+
"learning_rate": 2.767676767676768e-05,
|
1780 |
+
"loss": 0.0,
|
1781 |
+
"step": 1105
|
1782 |
+
},
|
1783 |
+
{
|
1784 |
+
"epoch": 25.227272727272727,
|
1785 |
+
"grad_norm": 0.002169027691707015,
|
1786 |
+
"learning_rate": 2.7550505050505055e-05,
|
1787 |
+
"loss": 0.0001,
|
1788 |
+
"step": 1110
|
1789 |
+
},
|
1790 |
+
{
|
1791 |
+
"epoch": 25.34090909090909,
|
1792 |
+
"grad_norm": 0.0018963554175570607,
|
1793 |
+
"learning_rate": 2.7424242424242424e-05,
|
1794 |
+
"loss": 0.0,
|
1795 |
+
"step": 1115
|
1796 |
+
},
|
1797 |
+
{
|
1798 |
+
"epoch": 25.454545454545453,
|
1799 |
+
"grad_norm": 0.0014427511487156153,
|
1800 |
+
"learning_rate": 2.72979797979798e-05,
|
1801 |
+
"loss": 0.0,
|
1802 |
+
"step": 1120
|
1803 |
+
},
|
1804 |
+
{
|
1805 |
+
"epoch": 25.568181818181817,
|
1806 |
+
"grad_norm": 0.0007338482537306845,
|
1807 |
+
"learning_rate": 2.717171717171717e-05,
|
1808 |
+
"loss": 0.0,
|
1809 |
+
"step": 1125
|
1810 |
+
},
|
1811 |
+
{
|
1812 |
+
"epoch": 25.681818181818183,
|
1813 |
+
"grad_norm": 0.0016011031111702323,
|
1814 |
+
"learning_rate": 2.7045454545454545e-05,
|
1815 |
+
"loss": 0.0,
|
1816 |
+
"step": 1130
|
1817 |
+
},
|
1818 |
+
{
|
1819 |
+
"epoch": 25.795454545454547,
|
1820 |
+
"grad_norm": 0.0010307264747098088,
|
1821 |
+
"learning_rate": 2.691919191919192e-05,
|
1822 |
+
"loss": 0.0,
|
1823 |
+
"step": 1135
|
1824 |
+
},
|
1825 |
+
{
|
1826 |
+
"epoch": 25.90909090909091,
|
1827 |
+
"grad_norm": 0.0012928517535328865,
|
1828 |
+
"learning_rate": 2.679292929292929e-05,
|
1829 |
+
"loss": 0.0,
|
1830 |
+
"step": 1140
|
1831 |
+
},
|
1832 |
+
{
|
1833 |
+
"epoch": 26.0,
|
1834 |
+
"eval_accuracy": 0.86,
|
1835 |
+
"eval_loss": 0.8471766114234924,
|
1836 |
+
"eval_runtime": 18.299,
|
1837 |
+
"eval_samples_per_second": 10.93,
|
1838 |
+
"eval_steps_per_second": 0.71,
|
1839 |
+
"step": 1144
|
1840 |
+
}
|
1841 |
+
],
|
1842 |
+
"logging_steps": 5,
|
1843 |
+
"max_steps": 2200,
|
1844 |
+
"num_input_tokens_seen": 0,
|
1845 |
+
"num_train_epochs": 50,
|
1846 |
+
"save_steps": 500,
|
1847 |
+
"stateful_callbacks": {
|
1848 |
+
"TrainerControl": {
|
1849 |
+
"args": {
|
1850 |
+
"should_epoch_stop": false,
|
1851 |
+
"should_evaluate": false,
|
1852 |
+
"should_log": false,
|
1853 |
+
"should_save": true,
|
1854 |
+
"should_training_stop": false
|
1855 |
+
},
|
1856 |
+
"attributes": {}
|
1857 |
+
}
|
1858 |
+
},
|
1859 |
+
"total_flos": 1.2319732604023603e+18,
|
1860 |
+
"train_batch_size": 16,
|
1861 |
+
"trial_name": null,
|
1862 |
+
"trial_params": null
|
1863 |
+
}
|
training_args.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:7f486fd77e39848e8df2b35c6ab70ac247be859d40dd36f39d0912702c28a982
|
3 |
+
size 5176
|