Finish init DVC/MLflow & refactor MeaLeon backend

Trying first round of combining DVC with MLflow. DVC explicitly handles data, both raw and processed data from data cleaning and embedding transformation. MLflow handles the embedding model and classes needed fo the model and model pipeline. MeaLeon backend has refactor to more hierarchical code structure and more clear delineation of functions and naming
AaronWChen · Jul 30, 2024 · 0371e7d · 0371e7d
1 parent 3b46f57
commit 0371e7d
Show file tree

Hide file tree

Showing 4 changed files with 410 additions and 179 deletions.
diff --git a/data.dvc b/data.dvc
@@ -1,5 +1,5 @@
 outs:
-- md5: d3e85fc804165a9de26ee51138033176.dir
-  size: 256748989
-  nfiles: 8
+- md5: 2ce6297077793c098f42db8660fb0d0e.dir
+  size: 656551290
+  nfiles: 12
   path: data