experiments

run_vrae.py

@@ -71,6 +71,8 @@
     print('>>>> TRAINING COMPLETE!')
 
     print('SAVING TRAINED MODEL...')
+
+    # save model
     vrae.save(f'VRAE_{args.scale_type}_un_{args.undo}_hidden_{args.hidden_layer_depth}_win_{args.sequence_length}_ep_{args.n_epochs}.pth')
     print(f'>>>> model saved as "VRAE_{args.scale_type}_un_{args.undo}_hidden_{args.hidden_layer_depth}_win_{args.sequence_length}_ep_{args.n_epochs}"')
 

vrae_experiment.sh

@@ -1,22 +1,14 @@
-#!/bin/sh
+# MODEL RUN
+# TRAIN on diff scalers, and layers
 
-BATCH_SIZE=8
-EPOCH=10
-LR=5e-5
-ACCUMULATION_STEP=1
+# Standard
+python run_vrae.py --scale_type 'Standard' --hidden_layer_depth 1
+python run_vrae.py --scale_type 'Standard' --hidden_layer_depth 2
 
-# N_ENC=6
-# N_DEC=6
+# MinMax
+python run_vrae.py --scale_type 'MinMax' --hidden_layer_depth 1
+python run_vrae.py --scale_type 'MinMax' --hidden_layer_depth 2
 
-# run distilBART-6-3
-python ./src/kobart/main.py\
-    --batch-size=${BATCH_SIZE}\
-    --lr=${LR}\
-    --epoch=${EPOCH}\
-    --gradient-accumulation-step=${ACCUMULATION_STEP}\
-    --amp\
-    --distributed
-#     --distill\
-    # --n_enc=${N_ENC}\
-    # --n_dec=${N_DEC}\
-
+# Robust
+python run_vrae.py --scale_type 'Robust' --hidden_layer_depth 1
+python run_vrae.py --scale_type 'Robust' --hidden_layer_depth 2