Spaces:
Build error
Build error
File size: 1,607 Bytes
3c53ea4 f2676bd 3c53ea4 17413e5 3c53ea4 89859a3 3c53ea4 45ec651 3c53ea4 ea0fe8d 3c53ea4 45ec651 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 |
import gradio as gr
import time
import whisper
import torch
from diffusers import StableDiffusionPipeline, EulerDiscreteScheduler
def SpeechToText(audio):
if audio == None : return ""
model = whisper.load_model("base")
audio = whisper.load_audio(audio)
audio = whisper.pad_or_trim(audio)
# make log-Mel spectrogram and move to the same device as the model
mel = whisper.log_mel_spectrogram(audio).to(model.device)
# Detect the Max probability of language ?
_, probs = model.detect_language(mel)
lang = f"Language: {max(probs, key=probs.get)}"
# Decode audio to Text
options = whisper.DecodingOptions(fp16 = False)
result = whisper.decode(model, mel, options)
return result.text
def img_Generation(text):
print(text)
#model_id = "stabilityai/stable-diffusion-2"
model_id = "stabilityai/stable-diffusion-2-1"
# Use the Euler scheduler here instead
scheduler = EulerDiscreteScheduler.from_pretrained(model_id, subfolder="scheduler")
pipe = StableDiffusionPipeline.from_pretrained(model_id, scheduler=scheduler, revision="fp16", torch_dtype=torch.float16)
pipe = pipe.to("cuda")
image = pipe(text, num_inference_steps = 150).images[0]
#image.save("img_1.png")
return image
def transcribe(audio):
text = SpeechToText(audio)
image = img_Generation(text)
return image
# gradio
gr.Interface(
fn=transcribe,
inputs=gr.Audio(source="microphone", type="filepath"),
outputs="image",description="A Speech to Image Generation App Using OpenAI's Whisper and Stable Diffusion V.2",title= "Whisper2IMG").launch()
|