thanks for the heads up. All apis are back running and updated with the latest gradio version (5.38.2). Everything should be okay in a few minutes 🤗😉
torchaoInt8WeightOnlyConfig is already working flawlessly in our tests.import spaces
from diffusers import FluxPipeline
from torchao.quantization.quant_api import Int8WeightOnlyConfig, quantize_
pipeline = FluxPipeline.from_pretrained(...).to('cuda')
quantize_(pipeline.transformer, Int8WeightOnlyConfig()) # Or any other component(s)
@spaces.GPU
def generate(prompt: str):
return pipeline(prompt).images[0]
medium size is now available as a power-user featurelarge (70GB VRAM)—but this paves the way for:medium will offer significantly more usage than large)xlarge size (141GB VRAM)auto (future default)mediumlarge (current default)largemedium
sdk_version to 5.28 (in the README.md)mcp_server=True in launch()def generate(text, speed=1):
"""
Convert text to speech audio.
Parameters:
text (str): The input text to be converted to speech.
speed (float, optional): Playback speed of the generated speech.
xet-team
) work, you know we've been working to migrate the Hugging Face Hub from Git LFS and to Xet.