Mirrors
/
EveryDream2trainer
mirror of https://github.com/victorchall/EveryDream2trainer.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

bugfix and doc fix

This commit is contained in:
Victor Hall 2024-03-03 15:47:44 -05:00
parent aca1867697
commit 91a9b3d898
2 changed files with 2 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
caption_kosmos2.py
View File

@ -103,7 +103,7 @@ def main(args):
print(f"File: {full_file_path}, Generated caption: {processed_text}")
name = os.path.splitext(full_file_path)[0]
if not os.path.exists(f"{name}.txt") or args.overwrite and not args.save_entities_only:
if (not os.path.exists(f"{name}.txt") or args.overwrite) and not args.save_entities_only:
with open(f"{name}.txt", "w") as f:
f.write(processed_text)

2
doc/CAPTION.md
View File

@ -6,7 +6,7 @@
## Kosmos-2
Microsoft's [Kosmos-2](https://huggingface.co/microsoft/kosmos-2-patch14-224) is significantly lighter weight than Cog, using <5GB of VRAM and generating captions in under 1/21 second on a RTX 3090.
Microsoft's [Kosmos-2](https://huggingface.co/microsoft/kosmos-2-patch14-224) is significantly lighter weight than Cog, using <5GB of VRAM and generating captions in under a second on a RTX 3090.
It has the capability to output grounding bounding boxes.