End of training
Browse files
all_results.json
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_accuracy": 0.9883257403189066,
|
| 4 |
+
"eval_loss": 0.036893460899591446,
|
| 5 |
+
"eval_runtime": 121.2482,
|
| 6 |
+
"eval_samples_per_second": 28.965,
|
| 7 |
+
"eval_steps_per_second": 3.621,
|
| 8 |
+
"train_loss": 0.1489328353447163,
|
| 9 |
+
"train_runtime": 1701.4159,
|
| 10 |
+
"train_samples_per_second": 11.695,
|
| 11 |
+
"train_steps_per_second": 1.462
|
| 12 |
+
}
|
eval_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"eval_accuracy": 0.9883257403189066,
|
| 4 |
+
"eval_loss": 0.036893460899591446,
|
| 5 |
+
"eval_runtime": 121.2482,
|
| 6 |
+
"eval_samples_per_second": 28.965,
|
| 7 |
+
"eval_steps_per_second": 3.621
|
| 8 |
+
}
|
runs/Dec09_03-51-11_e947487acabd/events.out.tfevents.1639021942.e947487acabd.481.0
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c6c7656e69c94e5708d2b1bd24a9327bbddba659dada89bf51831505a3107d4d
|
| 3 |
+
size 42796
|
runs/Dec09_03-51-11_e947487acabd/events.out.tfevents.1639023770.e947487acabd.481.2
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e412b8f567458dfbe330bfc446b8d54aafb0324fab9376c157c8d49b3137b12
|
| 3 |
+
size 363
|
train_results.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 1.0,
|
| 3 |
+
"train_loss": 0.1489328353447163,
|
| 4 |
+
"train_runtime": 1701.4159,
|
| 5 |
+
"train_samples_per_second": 11.695,
|
| 6 |
+
"train_steps_per_second": 1.462
|
| 7 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1522 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.036893460899591446,
|
| 3 |
+
"best_model_checkpoint": "./cats_vs_dogs_outputs/checkpoint-2488",
|
| 4 |
+
"epoch": 1.0,
|
| 5 |
+
"global_step": 2488,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0.0,
|
| 12 |
+
"learning_rate": 0.00019919614147909968,
|
| 13 |
+
"loss": 0.4243,
|
| 14 |
+
"step": 10
|
| 15 |
+
},
|
| 16 |
+
{
|
| 17 |
+
"epoch": 0.01,
|
| 18 |
+
"learning_rate": 0.00019839228295819936,
|
| 19 |
+
"loss": 0.2515,
|
| 20 |
+
"step": 20
|
| 21 |
+
},
|
| 22 |
+
{
|
| 23 |
+
"epoch": 0.01,
|
| 24 |
+
"learning_rate": 0.00019758842443729903,
|
| 25 |
+
"loss": 0.2728,
|
| 26 |
+
"step": 30
|
| 27 |
+
},
|
| 28 |
+
{
|
| 29 |
+
"epoch": 0.02,
|
| 30 |
+
"learning_rate": 0.00019678456591639874,
|
| 31 |
+
"loss": 0.2011,
|
| 32 |
+
"step": 40
|
| 33 |
+
},
|
| 34 |
+
{
|
| 35 |
+
"epoch": 0.02,
|
| 36 |
+
"learning_rate": 0.0001959807073954984,
|
| 37 |
+
"loss": 0.1596,
|
| 38 |
+
"step": 50
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.02,
|
| 42 |
+
"learning_rate": 0.00019517684887459809,
|
| 43 |
+
"loss": 0.2641,
|
| 44 |
+
"step": 60
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 0.03,
|
| 48 |
+
"learning_rate": 0.00019437299035369776,
|
| 49 |
+
"loss": 0.2891,
|
| 50 |
+
"step": 70
|
| 51 |
+
},
|
| 52 |
+
{
|
| 53 |
+
"epoch": 0.03,
|
| 54 |
+
"learning_rate": 0.00019356913183279743,
|
| 55 |
+
"loss": 0.1983,
|
| 56 |
+
"step": 80
|
| 57 |
+
},
|
| 58 |
+
{
|
| 59 |
+
"epoch": 0.04,
|
| 60 |
+
"learning_rate": 0.0001927652733118971,
|
| 61 |
+
"loss": 0.1355,
|
| 62 |
+
"step": 90
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.04,
|
| 66 |
+
"learning_rate": 0.00019196141479099678,
|
| 67 |
+
"loss": 0.1282,
|
| 68 |
+
"step": 100
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.04,
|
| 72 |
+
"learning_rate": 0.00019115755627009649,
|
| 73 |
+
"loss": 0.1681,
|
| 74 |
+
"step": 110
|
| 75 |
+
},
|
| 76 |
+
{
|
| 77 |
+
"epoch": 0.05,
|
| 78 |
+
"learning_rate": 0.00019035369774919616,
|
| 79 |
+
"loss": 0.4071,
|
| 80 |
+
"step": 120
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.05,
|
| 84 |
+
"learning_rate": 0.00018954983922829584,
|
| 85 |
+
"loss": 0.467,
|
| 86 |
+
"step": 130
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 0.06,
|
| 90 |
+
"learning_rate": 0.0001887459807073955,
|
| 91 |
+
"loss": 0.1196,
|
| 92 |
+
"step": 140
|
| 93 |
+
},
|
| 94 |
+
{
|
| 95 |
+
"epoch": 0.06,
|
| 96 |
+
"learning_rate": 0.00018794212218649519,
|
| 97 |
+
"loss": 0.2998,
|
| 98 |
+
"step": 150
|
| 99 |
+
},
|
| 100 |
+
{
|
| 101 |
+
"epoch": 0.06,
|
| 102 |
+
"learning_rate": 0.00018713826366559486,
|
| 103 |
+
"loss": 0.2774,
|
| 104 |
+
"step": 160
|
| 105 |
+
},
|
| 106 |
+
{
|
| 107 |
+
"epoch": 0.07,
|
| 108 |
+
"learning_rate": 0.00018633440514469454,
|
| 109 |
+
"loss": 0.3209,
|
| 110 |
+
"step": 170
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.07,
|
| 114 |
+
"learning_rate": 0.0001855305466237942,
|
| 115 |
+
"loss": 0.1764,
|
| 116 |
+
"step": 180
|
| 117 |
+
},
|
| 118 |
+
{
|
| 119 |
+
"epoch": 0.08,
|
| 120 |
+
"learning_rate": 0.0001847266881028939,
|
| 121 |
+
"loss": 0.3375,
|
| 122 |
+
"step": 190
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.08,
|
| 126 |
+
"learning_rate": 0.0001839228295819936,
|
| 127 |
+
"loss": 0.2445,
|
| 128 |
+
"step": 200
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 0.08,
|
| 132 |
+
"learning_rate": 0.00018311897106109326,
|
| 133 |
+
"loss": 0.2098,
|
| 134 |
+
"step": 210
|
| 135 |
+
},
|
| 136 |
+
{
|
| 137 |
+
"epoch": 0.09,
|
| 138 |
+
"learning_rate": 0.00018231511254019294,
|
| 139 |
+
"loss": 0.1775,
|
| 140 |
+
"step": 220
|
| 141 |
+
},
|
| 142 |
+
{
|
| 143 |
+
"epoch": 0.09,
|
| 144 |
+
"learning_rate": 0.0001815112540192926,
|
| 145 |
+
"loss": 0.2583,
|
| 146 |
+
"step": 230
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"epoch": 0.1,
|
| 150 |
+
"learning_rate": 0.00018070739549839229,
|
| 151 |
+
"loss": 0.1765,
|
| 152 |
+
"step": 240
|
| 153 |
+
},
|
| 154 |
+
{
|
| 155 |
+
"epoch": 0.1,
|
| 156 |
+
"learning_rate": 0.00017990353697749196,
|
| 157 |
+
"loss": 0.1831,
|
| 158 |
+
"step": 250
|
| 159 |
+
},
|
| 160 |
+
{
|
| 161 |
+
"epoch": 0.1,
|
| 162 |
+
"learning_rate": 0.00017909967845659166,
|
| 163 |
+
"loss": 0.1802,
|
| 164 |
+
"step": 260
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.11,
|
| 168 |
+
"learning_rate": 0.00017829581993569134,
|
| 169 |
+
"loss": 0.283,
|
| 170 |
+
"step": 270
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.11,
|
| 174 |
+
"learning_rate": 0.000177491961414791,
|
| 175 |
+
"loss": 0.2873,
|
| 176 |
+
"step": 280
|
| 177 |
+
},
|
| 178 |
+
{
|
| 179 |
+
"epoch": 0.12,
|
| 180 |
+
"learning_rate": 0.0001766881028938907,
|
| 181 |
+
"loss": 0.2789,
|
| 182 |
+
"step": 290
|
| 183 |
+
},
|
| 184 |
+
{
|
| 185 |
+
"epoch": 0.12,
|
| 186 |
+
"learning_rate": 0.00017588424437299036,
|
| 187 |
+
"loss": 0.4105,
|
| 188 |
+
"step": 300
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"epoch": 0.12,
|
| 192 |
+
"learning_rate": 0.00017508038585209004,
|
| 193 |
+
"loss": 0.2762,
|
| 194 |
+
"step": 310
|
| 195 |
+
},
|
| 196 |
+
{
|
| 197 |
+
"epoch": 0.13,
|
| 198 |
+
"learning_rate": 0.0001742765273311897,
|
| 199 |
+
"loss": 0.1414,
|
| 200 |
+
"step": 320
|
| 201 |
+
},
|
| 202 |
+
{
|
| 203 |
+
"epoch": 0.13,
|
| 204 |
+
"learning_rate": 0.0001734726688102894,
|
| 205 |
+
"loss": 0.2126,
|
| 206 |
+
"step": 330
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.14,
|
| 210 |
+
"learning_rate": 0.0001726688102893891,
|
| 211 |
+
"loss": 0.3115,
|
| 212 |
+
"step": 340
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.14,
|
| 216 |
+
"learning_rate": 0.00017186495176848876,
|
| 217 |
+
"loss": 0.2029,
|
| 218 |
+
"step": 350
|
| 219 |
+
},
|
| 220 |
+
{
|
| 221 |
+
"epoch": 0.14,
|
| 222 |
+
"learning_rate": 0.00017106109324758844,
|
| 223 |
+
"loss": 0.0867,
|
| 224 |
+
"step": 360
|
| 225 |
+
},
|
| 226 |
+
{
|
| 227 |
+
"epoch": 0.15,
|
| 228 |
+
"learning_rate": 0.0001702572347266881,
|
| 229 |
+
"loss": 0.1312,
|
| 230 |
+
"step": 370
|
| 231 |
+
},
|
| 232 |
+
{
|
| 233 |
+
"epoch": 0.15,
|
| 234 |
+
"learning_rate": 0.0001694533762057878,
|
| 235 |
+
"loss": 0.5391,
|
| 236 |
+
"step": 380
|
| 237 |
+
},
|
| 238 |
+
{
|
| 239 |
+
"epoch": 0.16,
|
| 240 |
+
"learning_rate": 0.00016864951768488746,
|
| 241 |
+
"loss": 0.2131,
|
| 242 |
+
"step": 390
|
| 243 |
+
},
|
| 244 |
+
{
|
| 245 |
+
"epoch": 0.16,
|
| 246 |
+
"learning_rate": 0.00016784565916398716,
|
| 247 |
+
"loss": 0.2431,
|
| 248 |
+
"step": 400
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.16,
|
| 252 |
+
"learning_rate": 0.00016704180064308684,
|
| 253 |
+
"loss": 0.2088,
|
| 254 |
+
"step": 410
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 0.17,
|
| 258 |
+
"learning_rate": 0.0001662379421221865,
|
| 259 |
+
"loss": 0.1451,
|
| 260 |
+
"step": 420
|
| 261 |
+
},
|
| 262 |
+
{
|
| 263 |
+
"epoch": 0.17,
|
| 264 |
+
"learning_rate": 0.0001654340836012862,
|
| 265 |
+
"loss": 0.1635,
|
| 266 |
+
"step": 430
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.18,
|
| 270 |
+
"learning_rate": 0.00016463022508038586,
|
| 271 |
+
"loss": 0.2134,
|
| 272 |
+
"step": 440
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.18,
|
| 276 |
+
"learning_rate": 0.00016382636655948554,
|
| 277 |
+
"loss": 0.3112,
|
| 278 |
+
"step": 450
|
| 279 |
+
},
|
| 280 |
+
{
|
| 281 |
+
"epoch": 0.18,
|
| 282 |
+
"learning_rate": 0.0001630225080385852,
|
| 283 |
+
"loss": 0.2105,
|
| 284 |
+
"step": 460
|
| 285 |
+
},
|
| 286 |
+
{
|
| 287 |
+
"epoch": 0.19,
|
| 288 |
+
"learning_rate": 0.0001622186495176849,
|
| 289 |
+
"loss": 0.1925,
|
| 290 |
+
"step": 470
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.19,
|
| 294 |
+
"learning_rate": 0.0001614147909967846,
|
| 295 |
+
"loss": 0.0771,
|
| 296 |
+
"step": 480
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 0.2,
|
| 300 |
+
"learning_rate": 0.00016061093247588426,
|
| 301 |
+
"loss": 0.1727,
|
| 302 |
+
"step": 490
|
| 303 |
+
},
|
| 304 |
+
{
|
| 305 |
+
"epoch": 0.2,
|
| 306 |
+
"learning_rate": 0.00015980707395498394,
|
| 307 |
+
"loss": 0.1623,
|
| 308 |
+
"step": 500
|
| 309 |
+
},
|
| 310 |
+
{
|
| 311 |
+
"epoch": 0.2,
|
| 312 |
+
"learning_rate": 0.0001590032154340836,
|
| 313 |
+
"loss": 0.3692,
|
| 314 |
+
"step": 510
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.21,
|
| 318 |
+
"learning_rate": 0.0001581993569131833,
|
| 319 |
+
"loss": 0.0848,
|
| 320 |
+
"step": 520
|
| 321 |
+
},
|
| 322 |
+
{
|
| 323 |
+
"epoch": 0.21,
|
| 324 |
+
"learning_rate": 0.00015739549839228296,
|
| 325 |
+
"loss": 0.1735,
|
| 326 |
+
"step": 530
|
| 327 |
+
},
|
| 328 |
+
{
|
| 329 |
+
"epoch": 0.22,
|
| 330 |
+
"learning_rate": 0.00015659163987138264,
|
| 331 |
+
"loss": 0.2304,
|
| 332 |
+
"step": 540
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.22,
|
| 336 |
+
"learning_rate": 0.00015578778135048234,
|
| 337 |
+
"loss": 0.1427,
|
| 338 |
+
"step": 550
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 0.23,
|
| 342 |
+
"learning_rate": 0.00015498392282958201,
|
| 343 |
+
"loss": 0.11,
|
| 344 |
+
"step": 560
|
| 345 |
+
},
|
| 346 |
+
{
|
| 347 |
+
"epoch": 0.23,
|
| 348 |
+
"learning_rate": 0.0001541800643086817,
|
| 349 |
+
"loss": 0.1318,
|
| 350 |
+
"step": 570
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"epoch": 0.23,
|
| 354 |
+
"learning_rate": 0.00015337620578778136,
|
| 355 |
+
"loss": 0.3012,
|
| 356 |
+
"step": 580
|
| 357 |
+
},
|
| 358 |
+
{
|
| 359 |
+
"epoch": 0.24,
|
| 360 |
+
"learning_rate": 0.00015257234726688104,
|
| 361 |
+
"loss": 0.1292,
|
| 362 |
+
"step": 590
|
| 363 |
+
},
|
| 364 |
+
{
|
| 365 |
+
"epoch": 0.24,
|
| 366 |
+
"learning_rate": 0.0001517684887459807,
|
| 367 |
+
"loss": 0.0977,
|
| 368 |
+
"step": 600
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.25,
|
| 372 |
+
"learning_rate": 0.0001509646302250804,
|
| 373 |
+
"loss": 0.121,
|
| 374 |
+
"step": 610
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.25,
|
| 378 |
+
"learning_rate": 0.0001501607717041801,
|
| 379 |
+
"loss": 0.1502,
|
| 380 |
+
"step": 620
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 0.25,
|
| 384 |
+
"learning_rate": 0.00014935691318327976,
|
| 385 |
+
"loss": 0.0814,
|
| 386 |
+
"step": 630
|
| 387 |
+
},
|
| 388 |
+
{
|
| 389 |
+
"epoch": 0.26,
|
| 390 |
+
"learning_rate": 0.00014855305466237944,
|
| 391 |
+
"loss": 0.2753,
|
| 392 |
+
"step": 640
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"epoch": 0.26,
|
| 396 |
+
"learning_rate": 0.00014774919614147911,
|
| 397 |
+
"loss": 0.1735,
|
| 398 |
+
"step": 650
|
| 399 |
+
},
|
| 400 |
+
{
|
| 401 |
+
"epoch": 0.27,
|
| 402 |
+
"learning_rate": 0.0001469453376205788,
|
| 403 |
+
"loss": 0.065,
|
| 404 |
+
"step": 660
|
| 405 |
+
},
|
| 406 |
+
{
|
| 407 |
+
"epoch": 0.27,
|
| 408 |
+
"learning_rate": 0.00014614147909967846,
|
| 409 |
+
"loss": 0.1937,
|
| 410 |
+
"step": 670
|
| 411 |
+
},
|
| 412 |
+
{
|
| 413 |
+
"epoch": 0.27,
|
| 414 |
+
"learning_rate": 0.00014533762057877814,
|
| 415 |
+
"loss": 0.1878,
|
| 416 |
+
"step": 680
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.28,
|
| 420 |
+
"learning_rate": 0.00014453376205787784,
|
| 421 |
+
"loss": 0.2737,
|
| 422 |
+
"step": 690
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 0.28,
|
| 426 |
+
"learning_rate": 0.00014372990353697752,
|
| 427 |
+
"loss": 0.08,
|
| 428 |
+
"step": 700
|
| 429 |
+
},
|
| 430 |
+
{
|
| 431 |
+
"epoch": 0.29,
|
| 432 |
+
"learning_rate": 0.0001429260450160772,
|
| 433 |
+
"loss": 0.2417,
|
| 434 |
+
"step": 710
|
| 435 |
+
},
|
| 436 |
+
{
|
| 437 |
+
"epoch": 0.29,
|
| 438 |
+
"learning_rate": 0.00014212218649517686,
|
| 439 |
+
"loss": 0.1485,
|
| 440 |
+
"step": 720
|
| 441 |
+
},
|
| 442 |
+
{
|
| 443 |
+
"epoch": 0.29,
|
| 444 |
+
"learning_rate": 0.00014131832797427654,
|
| 445 |
+
"loss": 0.1943,
|
| 446 |
+
"step": 730
|
| 447 |
+
},
|
| 448 |
+
{
|
| 449 |
+
"epoch": 0.3,
|
| 450 |
+
"learning_rate": 0.00014051446945337621,
|
| 451 |
+
"loss": 0.1168,
|
| 452 |
+
"step": 740
|
| 453 |
+
},
|
| 454 |
+
{
|
| 455 |
+
"epoch": 0.3,
|
| 456 |
+
"learning_rate": 0.0001397106109324759,
|
| 457 |
+
"loss": 0.271,
|
| 458 |
+
"step": 750
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.31,
|
| 462 |
+
"learning_rate": 0.0001389067524115756,
|
| 463 |
+
"loss": 0.1807,
|
| 464 |
+
"step": 760
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 0.31,
|
| 468 |
+
"learning_rate": 0.00013810289389067527,
|
| 469 |
+
"loss": 0.2312,
|
| 470 |
+
"step": 770
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.31,
|
| 474 |
+
"learning_rate": 0.00013729903536977494,
|
| 475 |
+
"loss": 0.2052,
|
| 476 |
+
"step": 780
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.32,
|
| 480 |
+
"learning_rate": 0.00013649517684887462,
|
| 481 |
+
"loss": 0.2927,
|
| 482 |
+
"step": 790
|
| 483 |
+
},
|
| 484 |
+
{
|
| 485 |
+
"epoch": 0.32,
|
| 486 |
+
"learning_rate": 0.0001356913183279743,
|
| 487 |
+
"loss": 0.2083,
|
| 488 |
+
"step": 800
|
| 489 |
+
},
|
| 490 |
+
{
|
| 491 |
+
"epoch": 0.33,
|
| 492 |
+
"learning_rate": 0.00013488745980707396,
|
| 493 |
+
"loss": 0.1997,
|
| 494 |
+
"step": 810
|
| 495 |
+
},
|
| 496 |
+
{
|
| 497 |
+
"epoch": 0.33,
|
| 498 |
+
"learning_rate": 0.00013408360128617364,
|
| 499 |
+
"loss": 0.0869,
|
| 500 |
+
"step": 820
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.33,
|
| 504 |
+
"learning_rate": 0.00013327974276527331,
|
| 505 |
+
"loss": 0.1731,
|
| 506 |
+
"step": 830
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 0.34,
|
| 510 |
+
"learning_rate": 0.00013247588424437302,
|
| 511 |
+
"loss": 0.2085,
|
| 512 |
+
"step": 840
|
| 513 |
+
},
|
| 514 |
+
{
|
| 515 |
+
"epoch": 0.34,
|
| 516 |
+
"learning_rate": 0.0001316720257234727,
|
| 517 |
+
"loss": 0.0884,
|
| 518 |
+
"step": 850
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.35,
|
| 522 |
+
"learning_rate": 0.00013086816720257237,
|
| 523 |
+
"loss": 0.1051,
|
| 524 |
+
"step": 860
|
| 525 |
+
},
|
| 526 |
+
{
|
| 527 |
+
"epoch": 0.35,
|
| 528 |
+
"learning_rate": 0.00013006430868167204,
|
| 529 |
+
"loss": 0.2395,
|
| 530 |
+
"step": 870
|
| 531 |
+
},
|
| 532 |
+
{
|
| 533 |
+
"epoch": 0.35,
|
| 534 |
+
"learning_rate": 0.00012926045016077172,
|
| 535 |
+
"loss": 0.1216,
|
| 536 |
+
"step": 880
|
| 537 |
+
},
|
| 538 |
+
{
|
| 539 |
+
"epoch": 0.36,
|
| 540 |
+
"learning_rate": 0.0001284565916398714,
|
| 541 |
+
"loss": 0.234,
|
| 542 |
+
"step": 890
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.36,
|
| 546 |
+
"learning_rate": 0.00012765273311897106,
|
| 547 |
+
"loss": 0.1559,
|
| 548 |
+
"step": 900
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 0.37,
|
| 552 |
+
"learning_rate": 0.00012684887459807077,
|
| 553 |
+
"loss": 0.2665,
|
| 554 |
+
"step": 910
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"epoch": 0.37,
|
| 558 |
+
"learning_rate": 0.00012604501607717044,
|
| 559 |
+
"loss": 0.0772,
|
| 560 |
+
"step": 920
|
| 561 |
+
},
|
| 562 |
+
{
|
| 563 |
+
"epoch": 0.37,
|
| 564 |
+
"learning_rate": 0.00012524115755627012,
|
| 565 |
+
"loss": 0.1792,
|
| 566 |
+
"step": 930
|
| 567 |
+
},
|
| 568 |
+
{
|
| 569 |
+
"epoch": 0.38,
|
| 570 |
+
"learning_rate": 0.0001244372990353698,
|
| 571 |
+
"loss": 0.1338,
|
| 572 |
+
"step": 940
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.38,
|
| 576 |
+
"learning_rate": 0.00012363344051446947,
|
| 577 |
+
"loss": 0.1326,
|
| 578 |
+
"step": 950
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.39,
|
| 582 |
+
"learning_rate": 0.00012282958199356914,
|
| 583 |
+
"loss": 0.1114,
|
| 584 |
+
"step": 960
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.39,
|
| 588 |
+
"learning_rate": 0.0001220257234726688,
|
| 589 |
+
"loss": 0.1606,
|
| 590 |
+
"step": 970
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 0.39,
|
| 594 |
+
"learning_rate": 0.0001212218649517685,
|
| 595 |
+
"loss": 0.0651,
|
| 596 |
+
"step": 980
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"epoch": 0.4,
|
| 600 |
+
"learning_rate": 0.00012041800643086818,
|
| 601 |
+
"loss": 0.1518,
|
| 602 |
+
"step": 990
|
| 603 |
+
},
|
| 604 |
+
{
|
| 605 |
+
"epoch": 0.4,
|
| 606 |
+
"learning_rate": 0.00011961414790996785,
|
| 607 |
+
"loss": 0.2242,
|
| 608 |
+
"step": 1000
|
| 609 |
+
},
|
| 610 |
+
{
|
| 611 |
+
"epoch": 0.41,
|
| 612 |
+
"learning_rate": 0.00011881028938906753,
|
| 613 |
+
"loss": 0.2432,
|
| 614 |
+
"step": 1010
|
| 615 |
+
},
|
| 616 |
+
{
|
| 617 |
+
"epoch": 0.41,
|
| 618 |
+
"learning_rate": 0.0001180064308681672,
|
| 619 |
+
"loss": 0.2233,
|
| 620 |
+
"step": 1020
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.41,
|
| 624 |
+
"learning_rate": 0.00011720257234726688,
|
| 625 |
+
"loss": 0.2343,
|
| 626 |
+
"step": 1030
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.42,
|
| 630 |
+
"learning_rate": 0.00011639871382636655,
|
| 631 |
+
"loss": 0.1798,
|
| 632 |
+
"step": 1040
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 0.42,
|
| 636 |
+
"learning_rate": 0.00011559485530546625,
|
| 637 |
+
"loss": 0.2162,
|
| 638 |
+
"step": 1050
|
| 639 |
+
},
|
| 640 |
+
{
|
| 641 |
+
"epoch": 0.43,
|
| 642 |
+
"learning_rate": 0.00011479099678456593,
|
| 643 |
+
"loss": 0.178,
|
| 644 |
+
"step": 1060
|
| 645 |
+
},
|
| 646 |
+
{
|
| 647 |
+
"epoch": 0.43,
|
| 648 |
+
"learning_rate": 0.0001139871382636656,
|
| 649 |
+
"loss": 0.1113,
|
| 650 |
+
"step": 1070
|
| 651 |
+
},
|
| 652 |
+
{
|
| 653 |
+
"epoch": 0.43,
|
| 654 |
+
"learning_rate": 0.00011318327974276528,
|
| 655 |
+
"loss": 0.1912,
|
| 656 |
+
"step": 1080
|
| 657 |
+
},
|
| 658 |
+
{
|
| 659 |
+
"epoch": 0.44,
|
| 660 |
+
"learning_rate": 0.00011237942122186495,
|
| 661 |
+
"loss": 0.1249,
|
| 662 |
+
"step": 1090
|
| 663 |
+
},
|
| 664 |
+
{
|
| 665 |
+
"epoch": 0.44,
|
| 666 |
+
"learning_rate": 0.00011157556270096463,
|
| 667 |
+
"loss": 0.1505,
|
| 668 |
+
"step": 1100
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.45,
|
| 672 |
+
"learning_rate": 0.0001107717041800643,
|
| 673 |
+
"loss": 0.2325,
|
| 674 |
+
"step": 1110
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.45,
|
| 678 |
+
"learning_rate": 0.00010996784565916398,
|
| 679 |
+
"loss": 0.1284,
|
| 680 |
+
"step": 1120
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.45,
|
| 684 |
+
"learning_rate": 0.00010916398713826368,
|
| 685 |
+
"loss": 0.0767,
|
| 686 |
+
"step": 1130
|
| 687 |
+
},
|
| 688 |
+
{
|
| 689 |
+
"epoch": 0.46,
|
| 690 |
+
"learning_rate": 0.00010836012861736335,
|
| 691 |
+
"loss": 0.0773,
|
| 692 |
+
"step": 1140
|
| 693 |
+
},
|
| 694 |
+
{
|
| 695 |
+
"epoch": 0.46,
|
| 696 |
+
"learning_rate": 0.00010755627009646303,
|
| 697 |
+
"loss": 0.1112,
|
| 698 |
+
"step": 1150
|
| 699 |
+
},
|
| 700 |
+
{
|
| 701 |
+
"epoch": 0.47,
|
| 702 |
+
"learning_rate": 0.0001067524115755627,
|
| 703 |
+
"loss": 0.0873,
|
| 704 |
+
"step": 1160
|
| 705 |
+
},
|
| 706 |
+
{
|
| 707 |
+
"epoch": 0.47,
|
| 708 |
+
"learning_rate": 0.00010594855305466238,
|
| 709 |
+
"loss": 0.1782,
|
| 710 |
+
"step": 1170
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.47,
|
| 714 |
+
"learning_rate": 0.00010514469453376205,
|
| 715 |
+
"loss": 0.1493,
|
| 716 |
+
"step": 1180
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 0.48,
|
| 720 |
+
"learning_rate": 0.00010434083601286173,
|
| 721 |
+
"loss": 0.2793,
|
| 722 |
+
"step": 1190
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.48,
|
| 726 |
+
"learning_rate": 0.00010353697749196143,
|
| 727 |
+
"loss": 0.1563,
|
| 728 |
+
"step": 1200
|
| 729 |
+
},
|
| 730 |
+
{
|
| 731 |
+
"epoch": 0.49,
|
| 732 |
+
"learning_rate": 0.0001027331189710611,
|
| 733 |
+
"loss": 0.1311,
|
| 734 |
+
"step": 1210
|
| 735 |
+
},
|
| 736 |
+
{
|
| 737 |
+
"epoch": 0.49,
|
| 738 |
+
"learning_rate": 0.00010192926045016078,
|
| 739 |
+
"loss": 0.2596,
|
| 740 |
+
"step": 1220
|
| 741 |
+
},
|
| 742 |
+
{
|
| 743 |
+
"epoch": 0.49,
|
| 744 |
+
"learning_rate": 0.00010112540192926045,
|
| 745 |
+
"loss": 0.0822,
|
| 746 |
+
"step": 1230
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.5,
|
| 750 |
+
"learning_rate": 0.00010032154340836013,
|
| 751 |
+
"loss": 0.1717,
|
| 752 |
+
"step": 1240
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 0.5,
|
| 756 |
+
"learning_rate": 9.951768488745982e-05,
|
| 757 |
+
"loss": 0.1029,
|
| 758 |
+
"step": 1250
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 0.51,
|
| 762 |
+
"learning_rate": 9.871382636655949e-05,
|
| 763 |
+
"loss": 0.068,
|
| 764 |
+
"step": 1260
|
| 765 |
+
},
|
| 766 |
+
{
|
| 767 |
+
"epoch": 0.51,
|
| 768 |
+
"learning_rate": 9.790996784565917e-05,
|
| 769 |
+
"loss": 0.2052,
|
| 770 |
+
"step": 1270
|
| 771 |
+
},
|
| 772 |
+
{
|
| 773 |
+
"epoch": 0.51,
|
| 774 |
+
"learning_rate": 9.710610932475884e-05,
|
| 775 |
+
"loss": 0.1236,
|
| 776 |
+
"step": 1280
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.52,
|
| 780 |
+
"learning_rate": 9.630225080385853e-05,
|
| 781 |
+
"loss": 0.1484,
|
| 782 |
+
"step": 1290
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.52,
|
| 786 |
+
"learning_rate": 9.54983922829582e-05,
|
| 787 |
+
"loss": 0.2226,
|
| 788 |
+
"step": 1300
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.53,
|
| 792 |
+
"learning_rate": 9.469453376205788e-05,
|
| 793 |
+
"loss": 0.0759,
|
| 794 |
+
"step": 1310
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 0.53,
|
| 798 |
+
"learning_rate": 9.389067524115757e-05,
|
| 799 |
+
"loss": 0.0796,
|
| 800 |
+
"step": 1320
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 0.53,
|
| 804 |
+
"learning_rate": 9.308681672025724e-05,
|
| 805 |
+
"loss": 0.15,
|
| 806 |
+
"step": 1330
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 0.54,
|
| 810 |
+
"learning_rate": 9.228295819935692e-05,
|
| 811 |
+
"loss": 0.1217,
|
| 812 |
+
"step": 1340
|
| 813 |
+
},
|
| 814 |
+
{
|
| 815 |
+
"epoch": 0.54,
|
| 816 |
+
"learning_rate": 9.147909967845659e-05,
|
| 817 |
+
"loss": 0.1487,
|
| 818 |
+
"step": 1350
|
| 819 |
+
},
|
| 820 |
+
{
|
| 821 |
+
"epoch": 0.55,
|
| 822 |
+
"learning_rate": 9.067524115755628e-05,
|
| 823 |
+
"loss": 0.178,
|
| 824 |
+
"step": 1360
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.55,
|
| 828 |
+
"learning_rate": 8.987138263665596e-05,
|
| 829 |
+
"loss": 0.0732,
|
| 830 |
+
"step": 1370
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.55,
|
| 834 |
+
"learning_rate": 8.906752411575563e-05,
|
| 835 |
+
"loss": 0.1639,
|
| 836 |
+
"step": 1380
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 0.56,
|
| 840 |
+
"learning_rate": 8.82636655948553e-05,
|
| 841 |
+
"loss": 0.1026,
|
| 842 |
+
"step": 1390
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 0.56,
|
| 846 |
+
"learning_rate": 8.7459807073955e-05,
|
| 847 |
+
"loss": 0.0329,
|
| 848 |
+
"step": 1400
|
| 849 |
+
},
|
| 850 |
+
{
|
| 851 |
+
"epoch": 0.57,
|
| 852 |
+
"learning_rate": 8.665594855305467e-05,
|
| 853 |
+
"loss": 0.2264,
|
| 854 |
+
"step": 1410
|
| 855 |
+
},
|
| 856 |
+
{
|
| 857 |
+
"epoch": 0.57,
|
| 858 |
+
"learning_rate": 8.585209003215434e-05,
|
| 859 |
+
"loss": 0.0799,
|
| 860 |
+
"step": 1420
|
| 861 |
+
},
|
| 862 |
+
{
|
| 863 |
+
"epoch": 0.57,
|
| 864 |
+
"learning_rate": 8.504823151125403e-05,
|
| 865 |
+
"loss": 0.0258,
|
| 866 |
+
"step": 1430
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 0.58,
|
| 870 |
+
"learning_rate": 8.42443729903537e-05,
|
| 871 |
+
"loss": 0.1056,
|
| 872 |
+
"step": 1440
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.58,
|
| 876 |
+
"learning_rate": 8.344051446945338e-05,
|
| 877 |
+
"loss": 0.0807,
|
| 878 |
+
"step": 1450
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.59,
|
| 882 |
+
"learning_rate": 8.263665594855306e-05,
|
| 883 |
+
"loss": 0.1491,
|
| 884 |
+
"step": 1460
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.59,
|
| 888 |
+
"learning_rate": 8.183279742765274e-05,
|
| 889 |
+
"loss": 0.1216,
|
| 890 |
+
"step": 1470
|
| 891 |
+
},
|
| 892 |
+
{
|
| 893 |
+
"epoch": 0.59,
|
| 894 |
+
"learning_rate": 8.102893890675242e-05,
|
| 895 |
+
"loss": 0.1722,
|
| 896 |
+
"step": 1480
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 0.6,
|
| 900 |
+
"learning_rate": 8.02250803858521e-05,
|
| 901 |
+
"loss": 0.0542,
|
| 902 |
+
"step": 1490
|
| 903 |
+
},
|
| 904 |
+
{
|
| 905 |
+
"epoch": 0.6,
|
| 906 |
+
"learning_rate": 7.942122186495177e-05,
|
| 907 |
+
"loss": 0.1965,
|
| 908 |
+
"step": 1500
|
| 909 |
+
},
|
| 910 |
+
{
|
| 911 |
+
"epoch": 0.61,
|
| 912 |
+
"learning_rate": 7.861736334405146e-05,
|
| 913 |
+
"loss": 0.1252,
|
| 914 |
+
"step": 1510
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.61,
|
| 918 |
+
"learning_rate": 7.781350482315113e-05,
|
| 919 |
+
"loss": 0.0457,
|
| 920 |
+
"step": 1520
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 0.61,
|
| 924 |
+
"learning_rate": 7.70096463022508e-05,
|
| 925 |
+
"loss": 0.1899,
|
| 926 |
+
"step": 1530
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.62,
|
| 930 |
+
"learning_rate": 7.62057877813505e-05,
|
| 931 |
+
"loss": 0.1058,
|
| 932 |
+
"step": 1540
|
| 933 |
+
},
|
| 934 |
+
{
|
| 935 |
+
"epoch": 0.62,
|
| 936 |
+
"learning_rate": 7.540192926045017e-05,
|
| 937 |
+
"loss": 0.1748,
|
| 938 |
+
"step": 1550
|
| 939 |
+
},
|
| 940 |
+
{
|
| 941 |
+
"epoch": 0.63,
|
| 942 |
+
"learning_rate": 7.459807073954984e-05,
|
| 943 |
+
"loss": 0.078,
|
| 944 |
+
"step": 1560
|
| 945 |
+
},
|
| 946 |
+
{
|
| 947 |
+
"epoch": 0.63,
|
| 948 |
+
"learning_rate": 7.379421221864952e-05,
|
| 949 |
+
"loss": 0.2113,
|
| 950 |
+
"step": 1570
|
| 951 |
+
},
|
| 952 |
+
{
|
| 953 |
+
"epoch": 0.64,
|
| 954 |
+
"learning_rate": 7.299035369774921e-05,
|
| 955 |
+
"loss": 0.101,
|
| 956 |
+
"step": 1580
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.64,
|
| 960 |
+
"learning_rate": 7.218649517684888e-05,
|
| 961 |
+
"loss": 0.0684,
|
| 962 |
+
"step": 1590
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 0.64,
|
| 966 |
+
"learning_rate": 7.138263665594856e-05,
|
| 967 |
+
"loss": 0.3001,
|
| 968 |
+
"step": 1600
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 0.65,
|
| 972 |
+
"learning_rate": 7.057877813504825e-05,
|
| 973 |
+
"loss": 0.1088,
|
| 974 |
+
"step": 1610
|
| 975 |
+
},
|
| 976 |
+
{
|
| 977 |
+
"epoch": 0.65,
|
| 978 |
+
"learning_rate": 6.977491961414792e-05,
|
| 979 |
+
"loss": 0.0099,
|
| 980 |
+
"step": 1620
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.66,
|
| 984 |
+
"learning_rate": 6.89710610932476e-05,
|
| 985 |
+
"loss": 0.0918,
|
| 986 |
+
"step": 1630
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.66,
|
| 990 |
+
"learning_rate": 6.816720257234727e-05,
|
| 991 |
+
"loss": 0.0402,
|
| 992 |
+
"step": 1640
|
| 993 |
+
},
|
| 994 |
+
{
|
| 995 |
+
"epoch": 0.66,
|
| 996 |
+
"learning_rate": 6.736334405144696e-05,
|
| 997 |
+
"loss": 0.0505,
|
| 998 |
+
"step": 1650
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.67,
|
| 1002 |
+
"learning_rate": 6.655948553054663e-05,
|
| 1003 |
+
"loss": 0.1337,
|
| 1004 |
+
"step": 1660
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 0.67,
|
| 1008 |
+
"learning_rate": 6.575562700964631e-05,
|
| 1009 |
+
"loss": 0.0712,
|
| 1010 |
+
"step": 1670
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 0.68,
|
| 1014 |
+
"learning_rate": 6.495176848874598e-05,
|
| 1015 |
+
"loss": 0.0177,
|
| 1016 |
+
"step": 1680
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 0.68,
|
| 1020 |
+
"learning_rate": 6.414790996784567e-05,
|
| 1021 |
+
"loss": 0.2133,
|
| 1022 |
+
"step": 1690
|
| 1023 |
+
},
|
| 1024 |
+
{
|
| 1025 |
+
"epoch": 0.68,
|
| 1026 |
+
"learning_rate": 6.334405144694535e-05,
|
| 1027 |
+
"loss": 0.0598,
|
| 1028 |
+
"step": 1700
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 0.69,
|
| 1032 |
+
"learning_rate": 6.254019292604502e-05,
|
| 1033 |
+
"loss": 0.1066,
|
| 1034 |
+
"step": 1710
|
| 1035 |
+
},
|
| 1036 |
+
{
|
| 1037 |
+
"epoch": 0.69,
|
| 1038 |
+
"learning_rate": 6.173633440514471e-05,
|
| 1039 |
+
"loss": 0.1304,
|
| 1040 |
+
"step": 1720
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.7,
|
| 1044 |
+
"learning_rate": 6.0932475884244377e-05,
|
| 1045 |
+
"loss": 0.2402,
|
| 1046 |
+
"step": 1730
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 0.7,
|
| 1050 |
+
"learning_rate": 6.012861736334405e-05,
|
| 1051 |
+
"loss": 0.0907,
|
| 1052 |
+
"step": 1740
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 0.7,
|
| 1056 |
+
"learning_rate": 5.9324758842443726e-05,
|
| 1057 |
+
"loss": 0.1261,
|
| 1058 |
+
"step": 1750
|
| 1059 |
+
},
|
| 1060 |
+
{
|
| 1061 |
+
"epoch": 0.71,
|
| 1062 |
+
"learning_rate": 5.8520900321543414e-05,
|
| 1063 |
+
"loss": 0.0255,
|
| 1064 |
+
"step": 1760
|
| 1065 |
+
},
|
| 1066 |
+
{
|
| 1067 |
+
"epoch": 0.71,
|
| 1068 |
+
"learning_rate": 5.771704180064309e-05,
|
| 1069 |
+
"loss": 0.1224,
|
| 1070 |
+
"step": 1770
|
| 1071 |
+
},
|
| 1072 |
+
{
|
| 1073 |
+
"epoch": 0.72,
|
| 1074 |
+
"learning_rate": 5.6913183279742764e-05,
|
| 1075 |
+
"loss": 0.1896,
|
| 1076 |
+
"step": 1780
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 0.72,
|
| 1080 |
+
"learning_rate": 5.610932475884244e-05,
|
| 1081 |
+
"loss": 0.1877,
|
| 1082 |
+
"step": 1790
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.72,
|
| 1086 |
+
"learning_rate": 5.530546623794213e-05,
|
| 1087 |
+
"loss": 0.1275,
|
| 1088 |
+
"step": 1800
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.73,
|
| 1092 |
+
"learning_rate": 5.45016077170418e-05,
|
| 1093 |
+
"loss": 0.1746,
|
| 1094 |
+
"step": 1810
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 0.73,
|
| 1098 |
+
"learning_rate": 5.369774919614148e-05,
|
| 1099 |
+
"loss": 0.0513,
|
| 1100 |
+
"step": 1820
|
| 1101 |
+
},
|
| 1102 |
+
{
|
| 1103 |
+
"epoch": 0.74,
|
| 1104 |
+
"learning_rate": 5.2893890675241165e-05,
|
| 1105 |
+
"loss": 0.1638,
|
| 1106 |
+
"step": 1830
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 0.74,
|
| 1110 |
+
"learning_rate": 5.209003215434084e-05,
|
| 1111 |
+
"loss": 0.0546,
|
| 1112 |
+
"step": 1840
|
| 1113 |
+
},
|
| 1114 |
+
{
|
| 1115 |
+
"epoch": 0.74,
|
| 1116 |
+
"learning_rate": 5.1286173633440515e-05,
|
| 1117 |
+
"loss": 0.0156,
|
| 1118 |
+
"step": 1850
|
| 1119 |
+
},
|
| 1120 |
+
{
|
| 1121 |
+
"epoch": 0.75,
|
| 1122 |
+
"learning_rate": 5.048231511254019e-05,
|
| 1123 |
+
"loss": 0.0511,
|
| 1124 |
+
"step": 1860
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.75,
|
| 1128 |
+
"learning_rate": 4.967845659163987e-05,
|
| 1129 |
+
"loss": 0.1748,
|
| 1130 |
+
"step": 1870
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.76,
|
| 1134 |
+
"learning_rate": 4.887459807073955e-05,
|
| 1135 |
+
"loss": 0.087,
|
| 1136 |
+
"step": 1880
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 0.76,
|
| 1140 |
+
"learning_rate": 4.8070739549839234e-05,
|
| 1141 |
+
"loss": 0.1595,
|
| 1142 |
+
"step": 1890
|
| 1143 |
+
},
|
| 1144 |
+
{
|
| 1145 |
+
"epoch": 0.76,
|
| 1146 |
+
"learning_rate": 4.726688102893891e-05,
|
| 1147 |
+
"loss": 0.1603,
|
| 1148 |
+
"step": 1900
|
| 1149 |
+
},
|
| 1150 |
+
{
|
| 1151 |
+
"epoch": 0.77,
|
| 1152 |
+
"learning_rate": 4.646302250803859e-05,
|
| 1153 |
+
"loss": 0.0242,
|
| 1154 |
+
"step": 1910
|
| 1155 |
+
},
|
| 1156 |
+
{
|
| 1157 |
+
"epoch": 0.77,
|
| 1158 |
+
"learning_rate": 4.5659163987138265e-05,
|
| 1159 |
+
"loss": 0.1388,
|
| 1160 |
+
"step": 1920
|
| 1161 |
+
},
|
| 1162 |
+
{
|
| 1163 |
+
"epoch": 0.78,
|
| 1164 |
+
"learning_rate": 4.485530546623795e-05,
|
| 1165 |
+
"loss": 0.1766,
|
| 1166 |
+
"step": 1930
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 0.78,
|
| 1170 |
+
"learning_rate": 4.405144694533762e-05,
|
| 1171 |
+
"loss": 0.1017,
|
| 1172 |
+
"step": 1940
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 0.78,
|
| 1176 |
+
"learning_rate": 4.32475884244373e-05,
|
| 1177 |
+
"loss": 0.0819,
|
| 1178 |
+
"step": 1950
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 0.79,
|
| 1182 |
+
"learning_rate": 4.244372990353698e-05,
|
| 1183 |
+
"loss": 0.0785,
|
| 1184 |
+
"step": 1960
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 0.79,
|
| 1188 |
+
"learning_rate": 4.163987138263666e-05,
|
| 1189 |
+
"loss": 0.0906,
|
| 1190 |
+
"step": 1970
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 0.8,
|
| 1194 |
+
"learning_rate": 4.083601286173634e-05,
|
| 1195 |
+
"loss": 0.0545,
|
| 1196 |
+
"step": 1980
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 0.8,
|
| 1200 |
+
"learning_rate": 4.0032154340836016e-05,
|
| 1201 |
+
"loss": 0.0917,
|
| 1202 |
+
"step": 1990
|
| 1203 |
+
},
|
| 1204 |
+
{
|
| 1205 |
+
"epoch": 0.8,
|
| 1206 |
+
"learning_rate": 3.92282958199357e-05,
|
| 1207 |
+
"loss": 0.1921,
|
| 1208 |
+
"step": 2000
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 0.81,
|
| 1212 |
+
"learning_rate": 3.842443729903537e-05,
|
| 1213 |
+
"loss": 0.1346,
|
| 1214 |
+
"step": 2010
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 0.81,
|
| 1218 |
+
"learning_rate": 3.7620578778135054e-05,
|
| 1219 |
+
"loss": 0.0455,
|
| 1220 |
+
"step": 2020
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 0.82,
|
| 1224 |
+
"learning_rate": 3.681672025723473e-05,
|
| 1225 |
+
"loss": 0.0267,
|
| 1226 |
+
"step": 2030
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 0.82,
|
| 1230 |
+
"learning_rate": 3.601286173633441e-05,
|
| 1231 |
+
"loss": 0.0781,
|
| 1232 |
+
"step": 2040
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 0.82,
|
| 1236 |
+
"learning_rate": 3.5209003215434085e-05,
|
| 1237 |
+
"loss": 0.0123,
|
| 1238 |
+
"step": 2050
|
| 1239 |
+
},
|
| 1240 |
+
{
|
| 1241 |
+
"epoch": 0.83,
|
| 1242 |
+
"learning_rate": 3.4405144694533766e-05,
|
| 1243 |
+
"loss": 0.0145,
|
| 1244 |
+
"step": 2060
|
| 1245 |
+
},
|
| 1246 |
+
{
|
| 1247 |
+
"epoch": 0.83,
|
| 1248 |
+
"learning_rate": 3.360128617363344e-05,
|
| 1249 |
+
"loss": 0.0802,
|
| 1250 |
+
"step": 2070
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.84,
|
| 1254 |
+
"learning_rate": 3.279742765273312e-05,
|
| 1255 |
+
"loss": 0.0797,
|
| 1256 |
+
"step": 2080
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 0.84,
|
| 1260 |
+
"learning_rate": 3.1993569131832804e-05,
|
| 1261 |
+
"loss": 0.2245,
|
| 1262 |
+
"step": 2090
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 0.84,
|
| 1266 |
+
"learning_rate": 3.118971061093248e-05,
|
| 1267 |
+
"loss": 0.0771,
|
| 1268 |
+
"step": 2100
|
| 1269 |
+
},
|
| 1270 |
+
{
|
| 1271 |
+
"epoch": 0.85,
|
| 1272 |
+
"learning_rate": 3.0385852090032157e-05,
|
| 1273 |
+
"loss": 0.0732,
|
| 1274 |
+
"step": 2110
|
| 1275 |
+
},
|
| 1276 |
+
{
|
| 1277 |
+
"epoch": 0.85,
|
| 1278 |
+
"learning_rate": 2.9581993569131832e-05,
|
| 1279 |
+
"loss": 0.1055,
|
| 1280 |
+
"step": 2120
|
| 1281 |
+
},
|
| 1282 |
+
{
|
| 1283 |
+
"epoch": 0.86,
|
| 1284 |
+
"learning_rate": 2.8778135048231513e-05,
|
| 1285 |
+
"loss": 0.0194,
|
| 1286 |
+
"step": 2130
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 0.86,
|
| 1290 |
+
"learning_rate": 2.7974276527331188e-05,
|
| 1291 |
+
"loss": 0.1721,
|
| 1292 |
+
"step": 2140
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.86,
|
| 1296 |
+
"learning_rate": 2.717041800643087e-05,
|
| 1297 |
+
"loss": 0.0708,
|
| 1298 |
+
"step": 2150
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 0.87,
|
| 1302 |
+
"learning_rate": 2.6366559485530545e-05,
|
| 1303 |
+
"loss": 0.0392,
|
| 1304 |
+
"step": 2160
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 0.87,
|
| 1308 |
+
"learning_rate": 2.5562700964630226e-05,
|
| 1309 |
+
"loss": 0.0824,
|
| 1310 |
+
"step": 2170
|
| 1311 |
+
},
|
| 1312 |
+
{
|
| 1313 |
+
"epoch": 0.88,
|
| 1314 |
+
"learning_rate": 2.4758842443729904e-05,
|
| 1315 |
+
"loss": 0.0319,
|
| 1316 |
+
"step": 2180
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 0.88,
|
| 1320 |
+
"learning_rate": 2.3954983922829582e-05,
|
| 1321 |
+
"loss": 0.1465,
|
| 1322 |
+
"step": 2190
|
| 1323 |
+
},
|
| 1324 |
+
{
|
| 1325 |
+
"epoch": 0.88,
|
| 1326 |
+
"learning_rate": 2.315112540192926e-05,
|
| 1327 |
+
"loss": 0.0373,
|
| 1328 |
+
"step": 2200
|
| 1329 |
+
},
|
| 1330 |
+
{
|
| 1331 |
+
"epoch": 0.89,
|
| 1332 |
+
"learning_rate": 2.234726688102894e-05,
|
| 1333 |
+
"loss": 0.0872,
|
| 1334 |
+
"step": 2210
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 0.89,
|
| 1338 |
+
"learning_rate": 2.154340836012862e-05,
|
| 1339 |
+
"loss": 0.0213,
|
| 1340 |
+
"step": 2220
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 0.9,
|
| 1344 |
+
"learning_rate": 2.07395498392283e-05,
|
| 1345 |
+
"loss": 0.0919,
|
| 1346 |
+
"step": 2230
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 0.9,
|
| 1350 |
+
"learning_rate": 1.9935691318327977e-05,
|
| 1351 |
+
"loss": 0.0234,
|
| 1352 |
+
"step": 2240
|
| 1353 |
+
},
|
| 1354 |
+
{
|
| 1355 |
+
"epoch": 0.9,
|
| 1356 |
+
"learning_rate": 1.9131832797427655e-05,
|
| 1357 |
+
"loss": 0.0677,
|
| 1358 |
+
"step": 2250
|
| 1359 |
+
},
|
| 1360 |
+
{
|
| 1361 |
+
"epoch": 0.91,
|
| 1362 |
+
"learning_rate": 1.8327974276527333e-05,
|
| 1363 |
+
"loss": 0.0754,
|
| 1364 |
+
"step": 2260
|
| 1365 |
+
},
|
| 1366 |
+
{
|
| 1367 |
+
"epoch": 0.91,
|
| 1368 |
+
"learning_rate": 1.752411575562701e-05,
|
| 1369 |
+
"loss": 0.0257,
|
| 1370 |
+
"step": 2270
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 0.92,
|
| 1374 |
+
"learning_rate": 1.672025723472669e-05,
|
| 1375 |
+
"loss": 0.1474,
|
| 1376 |
+
"step": 2280
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 0.92,
|
| 1380 |
+
"learning_rate": 1.5916398713826368e-05,
|
| 1381 |
+
"loss": 0.3303,
|
| 1382 |
+
"step": 2290
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 0.92,
|
| 1386 |
+
"learning_rate": 1.5112540192926044e-05,
|
| 1387 |
+
"loss": 0.1173,
|
| 1388 |
+
"step": 2300
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 0.93,
|
| 1392 |
+
"learning_rate": 1.4308681672025726e-05,
|
| 1393 |
+
"loss": 0.0271,
|
| 1394 |
+
"step": 2310
|
| 1395 |
+
},
|
| 1396 |
+
{
|
| 1397 |
+
"epoch": 0.93,
|
| 1398 |
+
"learning_rate": 1.3504823151125404e-05,
|
| 1399 |
+
"loss": 0.0182,
|
| 1400 |
+
"step": 2320
|
| 1401 |
+
},
|
| 1402 |
+
{
|
| 1403 |
+
"epoch": 0.94,
|
| 1404 |
+
"learning_rate": 1.2700964630225082e-05,
|
| 1405 |
+
"loss": 0.1104,
|
| 1406 |
+
"step": 2330
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 0.94,
|
| 1410 |
+
"learning_rate": 1.189710610932476e-05,
|
| 1411 |
+
"loss": 0.0909,
|
| 1412 |
+
"step": 2340
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 0.94,
|
| 1416 |
+
"learning_rate": 1.1093247588424438e-05,
|
| 1417 |
+
"loss": 0.0389,
|
| 1418 |
+
"step": 2350
|
| 1419 |
+
},
|
| 1420 |
+
{
|
| 1421 |
+
"epoch": 0.95,
|
| 1422 |
+
"learning_rate": 1.0289389067524116e-05,
|
| 1423 |
+
"loss": 0.067,
|
| 1424 |
+
"step": 2360
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 0.95,
|
| 1428 |
+
"learning_rate": 9.485530546623795e-06,
|
| 1429 |
+
"loss": 0.07,
|
| 1430 |
+
"step": 2370
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 0.96,
|
| 1434 |
+
"learning_rate": 8.681672025723474e-06,
|
| 1435 |
+
"loss": 0.0457,
|
| 1436 |
+
"step": 2380
|
| 1437 |
+
},
|
| 1438 |
+
{
|
| 1439 |
+
"epoch": 0.96,
|
| 1440 |
+
"learning_rate": 7.877813504823153e-06,
|
| 1441 |
+
"loss": 0.1058,
|
| 1442 |
+
"step": 2390
|
| 1443 |
+
},
|
| 1444 |
+
{
|
| 1445 |
+
"epoch": 0.96,
|
| 1446 |
+
"learning_rate": 7.07395498392283e-06,
|
| 1447 |
+
"loss": 0.0225,
|
| 1448 |
+
"step": 2400
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 0.97,
|
| 1452 |
+
"learning_rate": 6.270096463022508e-06,
|
| 1453 |
+
"loss": 0.1176,
|
| 1454 |
+
"step": 2410
|
| 1455 |
+
},
|
| 1456 |
+
{
|
| 1457 |
+
"epoch": 0.97,
|
| 1458 |
+
"learning_rate": 5.466237942122187e-06,
|
| 1459 |
+
"loss": 0.0267,
|
| 1460 |
+
"step": 2420
|
| 1461 |
+
},
|
| 1462 |
+
{
|
| 1463 |
+
"epoch": 0.98,
|
| 1464 |
+
"learning_rate": 4.662379421221865e-06,
|
| 1465 |
+
"loss": 0.0751,
|
| 1466 |
+
"step": 2430
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 0.98,
|
| 1470 |
+
"learning_rate": 3.858520900321544e-06,
|
| 1471 |
+
"loss": 0.0904,
|
| 1472 |
+
"step": 2440
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 0.98,
|
| 1476 |
+
"learning_rate": 3.054662379421222e-06,
|
| 1477 |
+
"loss": 0.0607,
|
| 1478 |
+
"step": 2450
|
| 1479 |
+
},
|
| 1480 |
+
{
|
| 1481 |
+
"epoch": 0.99,
|
| 1482 |
+
"learning_rate": 2.2508038585209006e-06,
|
| 1483 |
+
"loss": 0.1344,
|
| 1484 |
+
"step": 2460
|
| 1485 |
+
},
|
| 1486 |
+
{
|
| 1487 |
+
"epoch": 0.99,
|
| 1488 |
+
"learning_rate": 1.4469453376205788e-06,
|
| 1489 |
+
"loss": 0.0816,
|
| 1490 |
+
"step": 2470
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 1.0,
|
| 1494 |
+
"learning_rate": 6.430868167202573e-07,
|
| 1495 |
+
"loss": 0.0949,
|
| 1496 |
+
"step": 2480
|
| 1497 |
+
},
|
| 1498 |
+
{
|
| 1499 |
+
"epoch": 1.0,
|
| 1500 |
+
"eval_accuracy": 0.9883257403189066,
|
| 1501 |
+
"eval_loss": 0.036893460899591446,
|
| 1502 |
+
"eval_runtime": 109.7581,
|
| 1503 |
+
"eval_samples_per_second": 31.998,
|
| 1504 |
+
"eval_steps_per_second": 4.0,
|
| 1505 |
+
"step": 2488
|
| 1506 |
+
},
|
| 1507 |
+
{
|
| 1508 |
+
"epoch": 1.0,
|
| 1509 |
+
"step": 2488,
|
| 1510 |
+
"total_flos": 0.0,
|
| 1511 |
+
"train_loss": 0.1489328353447163,
|
| 1512 |
+
"train_runtime": 1701.4159,
|
| 1513 |
+
"train_samples_per_second": 11.695,
|
| 1514 |
+
"train_steps_per_second": 1.462
|
| 1515 |
+
}
|
| 1516 |
+
],
|
| 1517 |
+
"max_steps": 2488,
|
| 1518 |
+
"num_train_epochs": 1,
|
| 1519 |
+
"total_flos": 0.0,
|
| 1520 |
+
"trial_name": null,
|
| 1521 |
+
"trial_params": null
|
| 1522 |
+
}
|