zhangyiss/mlx-examples
1
0
Fork 0
mirror of https://github.com/ml-explore/mlx-examples.git synced 2025-12-15 17:58:54 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Comments

Browse Source
This commit is contained in:
Angelos Katharopoulos
2025-03-24 22:15:37 -07:00
parent c4d08de8b3
commit d06c4cde92
2 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
flux/README.md
View File

@@ -226,7 +226,7 @@ this section assumes you can launch distributed MLX programs using `mlx.launch
### Distributed Finetuning ### Distributed Finetuning
Distributed finetuning scales very well with FLUX and all one has to do is Distributed finetuning scales very well with FLUX and all one has to do is
simply to adjust the gradient accumulation and iterations so that the batch adjust the gradient accumulation and training iterations so that the batch
size remains the same. For instance, to replicate the following training size remains the same. For instance, to replicate the following training
```shell ```shell

2
flux/generate_interactive.py
View File

@@ -35,7 +35,7 @@ def to_latent_size(image_size):
if __name__ == "__main__": if __name__ == "__main__":
parser = argparse.ArgumentParser( parser = argparse.ArgumentParser(
description="Generate images from a textual prompt using stable diffusion" description="Generate images from a textual prompt using FLUX"
) )
parser.add_argument("--quantize", "-q", action="store_true") parser.add_argument("--quantize", "-q", action="store_true")
parser.add_argument("--model", choices=["schnell", "dev"], default="schnell") parser.add_argument("--model", choices=["schnell", "dev"], default="schnell")