minor fix

libs/transformer/get_transcript.py

@@ -2,30 +2,28 @@ import torch
 from transformers import AutoModelForSpeechSeq2Seq, AutoProcessor, pipeline
 
 
-
-device = "cuda:0" if torch.cuda.is_available() else "cpu"
-torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
-
-model_id = "distil-whisper/distil-large-v3"
-
-model = AutoModelForSpeechSeq2Seq.from_pretrained(
-    model_id, torch_dtype=torch_dtype, low_cpu_mem_usage=True, use_safetensors=True
-)
-model.to(device)
-
-processor = AutoProcessor.from_pretrained(model_id)
-
-pipe = pipeline(
-    "automatic-speech-recognition",
-    model=model,
-    tokenizer=processor.tokenizer,
-    feature_extractor=processor.feature_extractor,
-    max_new_tokens=128,
-    torch_dtype=torch_dtype,
-    device=device,
-    return_timestamps=True
-)
-
-
-result = pipe("https://static.langkingdom.com/user_playlist_practice_videos/2114103294b5c15605fd59773e948e58.mp3")
-print(result)
+def get_transcript_gpu(url: str, model_id: str):
+    device = "cuda:0" if torch.cuda.is_available() else "cpu"
+    torch_dtype = torch.float16 if torch.cuda.is_available() else torch.float32
+
+    model = AutoModelForSpeechSeq2Seq.from_pretrained(
+        model_id, torch_dtype=torch_dtype, low_cpu_mem_usage=True, use_safetensors=True
+    )
+    model.to(device)
+
+    processor = AutoProcessor.from_pretrained(model_id)
+
+    pipe = pipeline(
+        "automatic-speech-recognition",
+        model=model,
+        tokenizer=processor.tokenizer,
+        feature_extractor=processor.feature_extractor,
+        max_new_tokens=128,
+        torch_dtype=torch_dtype,
+        device=device,
+        return_timestamps=True
+    )
+
+
+    result = pipe(url)
+    return result.get("text"), result.get("chunks")

routers/get_transcript_transformer.py

@@ -1,6 +1,9 @@
+import os
 import time
 from fastapi import APIRouter, Depends, HTTPException, status
 
+from libs.convert_to_audio import convert_to_audio
+from libs.transformer.get_transcript import get_transcript_gpu
 from libs.transformer.get_transcript_2 import get_transcribe_transformers
 
 from libs.header_api_auth import get_api_key
@@ -9,16 +12,30 @@ from libs.header_api_auth import get_api_key
 router = APIRouter(prefix="/get-transcript-transformer", tags=["transcript"])
 
 @router.get("/")
-def get_transcript(audio_path: str, api_key: str = Depends(get_api_key)):
+def get_transcript(audio_path: str, model_size: str = "distil-whisper/distil-small.en", api_key: str = Depends(get_api_key)):
     st = time.time()
 
+    output_audio_folder = f"./cached/audio"
+
+    if not os.path.exists(output_audio_folder):
+        os.makedirs(output_audio_folder)
+
+    output_file = f"{output_audio_folder}/{audio_path.split('/')[-1].split(".")[0]}.mp3"
+
+    convert_to_audio(audio_path.strip(), output_file)
+
     try:
-        text, chunks = get_transcribe_transformers(audio_path)
+        text, chunks = get_transcript_gpu(output_file, model_size)
     except Exception as error:
         raise HTTPException(status_code=status.HTTP_403_FORBIDDEN, detail=f"error>>>: {error}")
+    finally:
+         if os.path.exists(output_file):
+            os.remove(output_file)
 
     listSentences = []
 
+    print(chunks)
+
     for chunk in chunks:
         listSentences.append({
             "start_time": chunk.get("timestamp")[0],