whisper test

.gitignore

@@ -0,0 +1,22 @@
+# Build Artifacts
+build/
+
+# Core Dumps
+core
+
+# Byte-Compiled Modules
+__pycache__/
+
+# Extension Modules
+*.so
+
+# Packaging Artifacts
+*.egg-info
+*.whl
+
+# IDEs and Tools
+.idea/
+.gdb_history
+.vscode/
+# Other
+.DS_Store

Dockerfile

@@ -0,0 +1,27 @@
+# Use the official Python 3.9 image
+FROM python:3.9
+
+# Set the working directory to /code
+WORKDIR /code
+
+# Copy the current directory contents into the container at /code
+COPY ./requirements.txt /code/requirements.txt
+
+# Install requirements.txt 
+RUN pip install --no-cache-dir --upgrade -r /code/requirements.txt
+
+# Set up a new user named "user" with user ID 1000
+RUN useradd -m -u 1000 user
+# Switch to the "user" user
+USER user
+# Set home to the user's home directory
+ENV HOME=/home/user \
+  PATH=/home/user/.local/bin:$PATH
+
+# Set the working directory to the user's home directory
+WORKDIR $HOME/app
+
+# Copy the current directory contents into the container at $HOME/app setting the owner to the user
+COPY --chown=user . $HOME/app
+
+CMD ["uvicorn", "main:app", "--host", "0.0.0.0", "--port", "7860"]

README.md

@@ -1,11 +1,19 @@
 ---
 title: Whisper
 emoji: 🚀
-colorFrom: yellow
-colorTo: yellow
+colorFrom: cyan
+colorTo: cyan
 sdk: docker
 pinned: false
 license: apache-2.0
 ---
 
+# Whisper
+
+## TODO
+
+- Provide api to use whisper
+
+- Deploy whisper on GPU
+
 Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

main.py

@@ -0,0 +1,47 @@
+from typing import Union
+from pydantic import BaseModel
+from fastapi import FastAPI
+from fastapi.middleware.cors import CORSMiddleware
+
+from fastapi.staticfiles import StaticFiles
+from fastapi.responses import FileResponse
+
+import torch
+from transformers import pipeline
+
+app = FastAPI(docs_url="/api/docs")
+
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=["*"],
+    allow_methods=["*"],
+    allow_headers=["*"],
+    allow_credentials=True,
+)
+
+device = "cuda:0" if torch.cuda.is_available() else "cpu"
+torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
+
+BATCH_SIZE = 8
+
+
+pipe = pipeline("automatic-speech-recognition",
+                "openai/whisper-large-v3",
+                torch_dtype=torch_dtype,
+                device=device)
+
+
+@app.get("/device")
+def getDevice():
+    return device
+
+
+@app.get("/transcribe")
+def transcribe(inputs, task):
+    if inputs is None:
+        raise "No audio file submitted! Please upload or record an audio file before submitting your request."
+
+    text = pipe(inputs, batch_size=BATCH_SIZE, generate_kwargs={
+                "task": task}, return_timestamps=True)["text"]
+
+    return text

packages.txt

@@ -0,0 +1 @@
+ffmpeg

requirements.txt

@@ -0,0 +1,9 @@
+fastapi
+pydantic
+typing
+transformers
+torch
+requests
+sentencepiece
+uvicorn[standard]
+ffmpeg