update direct stepwise train, pass trainning with mock one step trace duplicated 52 times; reward OK, curve improved as the step grows.
This commit is contained in:
4
.gitignore
vendored
4
.gitignore
vendored
@@ -4,4 +4,6 @@ sample/sample_output/
|
||||
sample/tmp_dataset_gen/
|
||||
sample/output_data/
|
||||
|
||||
*__pycache__/
|
||||
*__pycache__/
|
||||
|
||||
*checkpoint*
|
||||
Reference in New Issue
Block a user