host my awalit

.gitignore

@@ -0,0 +1 @@
+.env

Dockerfile

@@ -0,0 +1,11 @@
+FROM python:3.10
+
+WORKDIR /app
+
+COPY . .
+
+RUN pip install --no-cache-dir -r requirements.txt
+
+EXPOSE 7860
+
+CMD ["uvicorn", "app:app", "--host", "0.0.0.0", "--port", "7860"]

app.py

@@ -0,0 +1,81 @@
+import logging
+import sys
+from fastapi import FastAPI, UploadFile, File, HTTPException
+from pydantic import BaseModel
+import config
+from models import load_whisper, load_summarizer, load_spacy
+from services import process_transcription, process_summary
+
+
+logger = logging.getLogger(__name__)
+
+app = FastAPI(
+    title="Transcription and Summarization API",
+    description="API using Faster-Whisper, spaCy, and Hugging Face Transformers",
+    version="1.0.0"
+)
+
+logger.info("Application starting up - loading models...")
+whisper_model = load_whisper(config)
+summarizer_pipeline = load_summarizer(config)
+nlp_spacy = load_spacy(config)
+logger.info("Model loading complete.")
+
+if not whisper_model:
+    logger.critical("Whisper model failed to load. Transcription endpoint will be unavailable.")
+if not summarizer_pipeline:
+    logger.critical("Summarizer pipeline failed to load. Summarization endpoint will be unavailable.")
+if not nlp_spacy:
+    logger.warning("SpaCy model failed to load. Summarization will proceed without spaCy preprocessing.")
+
+
+class TranscriptInput(BaseModel):
+    transcript: str
+
+
+@app.get("/health")
+def health():
+    return {"status": "ok",
+        "whisper_loaded": whisper_model is not None,
+        "summarizer_loaded": summarizer_pipeline is not None,
+        "spacy_loaded": nlp_spacy is not None
+        }
+
+
+@app.post("/transcribe")
+async def transcription(audio_file : UploadFile = File(...)):
+    if whisper_model is None:
+         raise HTTPException(status_code=503, detail="Transcription service unavailable.")
+
+    try:
+        content = await audio_file.read()
+        transcript, info = process_transcription(content, whisper_model)
+        logger.info(f"Transcription successful. Language: {info.language}")
+        return {"transcript": transcript}
+    except ValueError as ve:
+         logger.error(f"Value error during transcription processing: {ve}")
+         raise HTTPException(status_code=400, detail=str(ve))
+    except Exception as e:
+        logger.error(f"Unhandled error during transcription: {e}", exc_info=True)
+        raise HTTPException(status_code=500, detail="Internal server error during transcription.")
+
+
+@app.post("/summarize")
+def summarize(input: TranscriptInput):
+
+    if summarizer_pipeline is None:
+         raise HTTPException(status_code=503, detail="Summarization service unavailable.")
+    if not input.transcript:
+         raise HTTPException(status_code=400, detail="Transcript cannot be empty.")
+
+    try:
+        summary = process_summary(input.transcript, summarizer_pipeline, nlp_spacy, config)
+        return {"summary": summary}
+
+    except ValueError as ve:
+         logger.error(f"Value error during summary processing: {ve}")
+         raise HTTPException(status_code=400, detail=str(ve))
+
+    except Exception as e:
+        logger.error(f"Unhandled error during summarization: {e}", exc_info=True)
+        raise HTTPException(status_code=500, detail="Internal server error during summarization.")

config.py

@@ -0,0 +1,18 @@
+import psutil
+import os
+#from dotenv import load_dotenv
+
+
+#load_dotenv()
+WHISPER_MODEL_NAME = "tiny"
+WHISPER_DEVICE = "cpu"
+WHISPER_COMPUTE_TYPE = "int8"
+PYANNOTE_AUTH_TOKEN = os.getenv("HUGGINGFACE_API_KEY")
+SUMMARIZER_MODEL = "facebook/bart-large-cnn"
+SUMMARIZER_MAX_LENGTH = 150
+SUMMARIZER_MIN_LENGTH = 50
+SPACY_MODEL = "en_core_web_sm"
+CPU_THREADS = max(1, psutil.cpu_count(logical=False))
+
+if not PYANNOTE_AUTH_TOKEN:
+    raise ValueError("HUGGINGFACE_API_KEY not set in environment variables")

models.py

@@ -0,0 +1,59 @@
+import logging
+from faster_whisper import WhisperModel
+import spacy
+from transformers import pipeline
+import os
+
+logger = logging.getLogger(__name__)
+
+
+def load_whisper(config):
+    logger.info("Loading Whisper model...")
+    try:
+        model = WhisperModel(
+            config.WHISPER_MODEL_NAME,
+            device=config.WHISPER_DEVICE,
+            compute_type=config.WHISPER_COMPUTE_TYPE,
+            cpu_threads=config.CPU_THREADS
+        )
+        logger.info(f"Whisper model '{config.WHISPER_MODEL_NAME}' loaded on {config.WHISPER_DEVICE}.")
+        return model
+    except Exception as e:
+        logger.error(f"Failed to load Whisper model: {e}", exc_info=True)
+        return None
+
+def load_summarizer(config):
+    logger.info("Loading Summarization pipeline...")
+    try:
+        summarizer = pipeline("summarization", model=config.SUMMARIZER_MODEL)
+        logger.info("Summarization pipeline loaded.")
+        return summarizer
+    except Exception as e:
+        logger.error(f"Failed to load Summarization pipeline: {e}", exc_info=True)
+        return None
+
+def load_spacy(config):
+    logger.info("Loading spaCy model...")
+
+    try:
+        nlp = spacy.load("en_core_web_sm")
+        logger.info("spaCy model 'en_core_web_sm' loaded.")
+        
+        return nlp
+
+    except OSError:
+        logger.warning("spaCy model 'en_core_web_sm' not found. Trying to download...")
+
+        try:
+            spacy.cli.download("en_core_web_sm")
+            nlp = spacy.load("en_core_web_sm")
+            logger.info("spaCy model 'en_core_web_sm' downloaded and loaded.")
+            return nlp
+        
+        except Exception as download_e:
+            logger.error(f"Failed to download or load spaCy model 'en_core_web_sm': {download_e}")
+            return None
+
+    except Exception as e:
+        logger.error(f"Failed to load spaCy model: {e}")
+        return None

requirements.txt

@@ -0,0 +1,9 @@
+fastapi
+uvicorn
+transformers
+torch
+faster_whisper
+spacy
+pydub
+psutil
+logging

services.py

@@ -0,0 +1,45 @@
+import logging
+import os
+import tempfile
+
+logger = logging.getLogger(__name__)
+
+def process_transcription(audio_content: bytes, whisper_model):
+    if not whisper_model:
+        raise ValueError("Whisper model not loaded.")
+
+    temp_file_path = None
+    try:
+        with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as temp_file:
+            temp_file_path = temp_file.name
+            temp_file.write(audio_content)
+    
+        segments, info = whisper_model.transcribe(temp_file_path, beam_size=5)
+        transcript = " ".join([seg.text.strip() for seg in segments])
+        return transcript, info
+    finally:
+        if temp_file_path and os.path.exists(temp_file_path):
+            os.remove(temp_file_path)
+
+def process_summary(text: str, summarizer_pipeline, nlp_spacy, config):
+    if not summarizer_pipeline:
+        raise ValueError("Summarizer model not loaded.")
+
+    processed_text = text
+    if nlp_spacy:
+        try:
+            doc = nlp_spacy(text)
+            sentences = [sent.text.strip() for sent in doc.sents]
+            processed_text = " ".join(sentences)
+        except Exception as e:
+             logger.error(f"SpaCy processing failed: {e}", exc_info=True)
+
+    summary_output = summarizer_pipeline(
+        processed_text,
+        max_length=config.SUMMARIZER_MAX_LENGTH,
+        min_length=config.SUMMARIZER_MIN_LENGTH,
+        do_sample=False
+    )
+
+    final_summary = summary_output[0]['summary_text']
+    return final_summary