feat: add TAESD implementation - faster autoencoder (#88)

* add taesd implementation * taesd gpu offloading * show seed when generating image with -s -1 * less restrictive with larger images * cuda: im2col speedup x2 * cuda: group norm speedup x90 * quantized models now works in cuda :) * fix cal mem size --------- Co-authored-by: leejet <leejet714@gmail.com>
2023-12-05 09:40:03 -05:00
parent f99bcd1f76
commit 134883aec4
14 changed files with 908 additions and 46904 deletions
--- a/.gitignore
+++ b/.gitignore
@@ -8,6 +8,7 @@ test/
 *.bin
 *.exe
 *.gguf
-*.log
-output.png
-models/
+output*.png
+models*
+!taesd-model.gguf
+*.log