Synced repo using 'sync_with_huggingface' Github Action

Dockerfile

@@ -62,6 +62,6 @@ EXPOSE 7860
 
 # Start the FastAPI app using Uvicorn web server
 # CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "14000", "--limit-concurrency", "1000"]
-# RUN python3 download.py
+RUN python3 download.py
 
-CMD ["python3", "app.py", "--host=0.0.0.0", "--port=7860", "--model_path=BAAI/bge-small-en-v1.5", "--num_workers=2"]
+CMD ["python3", "app.py", "--host=0.0.0.0", "--port=7860", "--model_path=models/BAAI/bge-small-en-v1.5", "--num_workers=2"]

app.py

@@ -8,24 +8,14 @@ from io import BytesIO
 import uvicorn
 from fastapi import FastAPI, BackgroundTasks, File, Body, UploadFile, Request
 from fastapi.responses import StreamingResponse
-# from faster_whisper import WhisperModel
 from starlette.staticfiles import StaticFiles
 from starlette.templating import Jinja2Templates
 from sentence_transformers import SentenceTransformer
-# from zhconv import convert
 
 # from utils.data_utils import remove_punctuation
 # from utils.utils import add_arguments, print_arguments
 
 
-import hashlib
-import os
-import tarfile
-import urllib.request
-
-# from tqdm import tqdm
-
-
 def print_arguments(args):
     print("-----------  Configuration Arguments -----------")
     for arg, value in vars(args).items():
@@ -77,7 +67,7 @@ args = parser.parse_args()
 print_arguments(args)
 
 # 
-# assert os.path.exists(args.model_path), f"{args.model_path}"
+assert os.path.exists(args.model_path), f"{args.model_path}"
 # 
 if args.use_gpu:
     model = SentenceTransformer(args.model_path, device="cuda", compute_type="float16", cache_folder=".")
@@ -85,64 +75,12 @@ else:
     model = SentenceTransformer(args.model_path, device='cpu', cache_folder=".")
 
 
-# 
-# _, _ = model.transcribe("dataset/test.wav", beam_size=5)
-
 app = FastAPI(title="embedding Inference")
 # app.mount('/static', StaticFiles(directory='static'), name='static')
 # templates = Jinja2Templates(directory="templates")
 # model_semaphore = None
 
 
-# def release_model_semaphore():
-#     model_semaphore.release()
-
-
-# def recognition(file: File, to_simple: int,
-#                 remove_pun: int, language: str = "bn",
-#                 task: str = "transcribe"
-#     ):
-
-#     segments, info = model.transcribe(file, beam_size=10, task=task, language=language, vad_filter=args.vad_filter)
-#     for segment in segments:
-#         text = segment.text
-#         if to_simple == 1:
-#             # text = convert(text, '')
-#             pass
-#         if remove_pun == 1:
-#             # text = remove_punctuation(text)
-#             pass
-#         ret = {"result": text, "start": round(segment.start, 2), "end": round(segment.end, 2)}
-#         # 
-#         yield json.dumps(ret).encode() + b"\0"
-
-
-# @app.post("/recognition_stream")
-# async def api_recognition_stream(
-#         to_simple: int = Body(1, description="", embed=True),
-#         remove_pun: int = Body(0, description="", embed=True),
-#         language: str = Body("bn", description="", embed=True),
-#         task: str = Body("transcribe", description="", embed=True),
-#         audio: UploadFile = File(..., description="")
-#         ):
-
-#     global model_semaphore
-#     if language == "None": language = None
-#     if model_semaphore is None:
-#         model_semaphore = asyncio.Semaphore(5)
-#     await model_semaphore.acquire()
-#     contents = await audio.read()
-#     data = BytesIO(contents)
-#     generator = recognition(
-#         file=data, to_simple=to_simple,
-#         remove_pun=remove_pun, language=language,
-#         task=task
-#         )
-#     background_tasks = BackgroundTasks()
-#     background_tasks.add_task(release_model_semaphore)
-#     return StreamingResponse(generator, background=background_tasks)
-
-
 @app.post("/embed")
 async def api_embed(
             textA: str = Body("text1", description="", embed=True),

download.py

@@ -0,0 +1,44 @@
+import argparse
+import requests
+import os
+from tqdm import tqdm
+
+def download_file(url, path):
+    response = requests.get(url, stream=True)
+    total_size_in_bytes = int(response.headers.get('content-length', 0))
+    block_size = 1024 #1 Kbyte
+    progress_bar = tqdm(total=total_size_in_bytes, unit='iB', unit_scale=True)
+
+    with open(path, 'wb') as file:
+        for data in response.iter_content(block_size):
+            progress_bar.update(len(data))
+            file.write(data)
+            
+    progress_bar.close()
+
+def download_model(model_name, destination_folder="models"):
+    # Define the base URL and headers for the Hugging Face API
+    base_url = f"https://huggingface.co/{model_name}/resolve/main"
+    headers = {"User-Agent": "Hugging Face Python"}
+
+    # Send a GET request to the Hugging Face API to get a list of all files
+    response = requests.get(f"https://huggingface.co/api/models/{model_name}", headers=headers)
+    response.raise_for_status()
+    
+    # Extract the list of files from the response JSON
+    files_to_download = [file["rfilename"] for file in response.json()["siblings"]]
+
+    # Ensure the directory exists
+    os.makedirs(f"{destination_folder}/{model_name}", exist_ok=True)
+
+    # Download each file
+    for file in files_to_download:
+        print(f"Downloading {file}...")
+        download_file(f"{base_url}/{file}", f"{destination_folder}/{model_name}/{file}")
+
+if __name__ == "__main__":
+    # parser = argparse.ArgumentParser()
+    # parser.add_argument("model_name", type=str, default="sam2ai/whisper-odia-small-finetune-int8-ct2", help="Name of the model to download.")
+    # args = parser.parse_args()
+
+    download_model("BAAI/bge-small-en-v1.5")