Merge branch 'prod' into main

.gitignore

@@ -99,15 +99,15 @@ ipython_config.py
 #   This is especially recommended for binary packages to ensure reproducibility, and is more
 #   commonly ignored for libraries.
 #   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
-poetry.lock
+**poetry.lock
 
 # pdm
 #   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
 #pdm.lock
 #   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
 #   in version control.
 #   https://pdm.fming.dev/#use-with-ide
-.toml
+**.toml
 
 # PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
 __pypackages__/
@@ -159,9 +159,9 @@ cython_debug/
 #  option (not recommended) you can uncomment the following to ignore the entire idea folder.
 #.idea/
 
+**.pt
+**.log
+**.wav
+**.mp3
+**.m4a
 **.pass
-**..pt
-**..log
-**..wav
-**..mp3
-**..m4a

Dockerfile

@@ -4,9 +4,9 @@ WORKDIR /app
 
 COPY requirements.txt /app/
 
-RUN apt-get update && apt-get install -y git
-RUN pip install \
+RUN apt-get update && apt-get install -y git && rm -rf /var/lib/apt/list/*
+RUN pip3 install \
     git+https://github.com/linto-ai/whisper-timestamped.git#egg=whisper-timestamped[dev,vad_silero,vad_auditok,test] \
     -r requirements.txt
 
-COPY transcribe.py /app/
+COPY transcribe.py /app/

client/init_server.py

@@ -0,0 +1,121 @@
+import os
+import argparse
+import requests
+import logging
+
+
+class Server:
+    def __init__(self):
+        self.gpu_url = os.environ.get(
+            "WHISPER_SERVER_DEFAULT", "http://10.2.5.212:8888/transcribe"
+        )
+        logging.basicConfig(level=logging.INFO)
+
+    def accept_feature_extractor(self, sentences, accept):
+        if len(accept) > 1 and accept["text"] != "":
+            for segments_rec in accept["segments"]:
+                segment_text = str(segments_rec["text"])
+                segment_start = segments_rec["start"]
+                segment_end = segments_rec["end"]
+                conf_score = float(segments_rec["confidence"])
+                sentences.append(
+                    {
+                        "text": segment_text,
+                        "start": segment_start,
+                        "end": segment_end,
+                        "confidence": conf_score,
+                    }
+                )
+
+    def transcribation_process(
+        self,
+        original_file_name,
+        duration=0,
+        side=True,
+        rec_date="31.01.2024",
+        src=1,
+        dst=2,
+        linkedid=3,
+        file_size=0,
+        queue_date="31.01.2024",
+        transcribation_date="31.01.2024",
+    ):
+
+        sentences = []
+
+        file_path = original_file_name
+        with open(file_path, "rb") as audio_file:
+            response = requests.post(
+                self.gpu_url,
+                files={"file": (os.path.basename(file_path), audio_file, "audio/wav")},
+            )
+
+        if response.status_code == 200:
+            accept = response.json()
+            self.accept_feature_extractor(sentences, accept)
+        else:
+            logging.error(f"Error in file processing: {response.text}")
+            return 0, [], []
+
+        for i in range(0, len(sentences)):
+            self.save_result(
+                original_file_name,
+                duration,
+                sentences[i]["text"],
+                sentences[i]["start"],
+                sentences[i]["end"],
+                side,
+                transcribation_date,
+                str(sentences[i]["confidence"]),
+                rec_date,
+                src,
+                dst,
+                linkedid,
+                file_size,
+                queue_date,
+            )
+
+        phrases = [sentences[i]["text"] for i in range(len(sentences))]
+        confidences = [sentences[i]["confidence"] for i in range(len(sentences))]
+
+        return len(sentences), phrases, confidences
+
+    def save_result(
+        self,
+        original_file_name,
+        duration,
+        accept_text,
+        accept_start,
+        accept_end,
+        side,
+        transcribation_date,
+        conf_mid,
+        rec_date,
+        src,
+        dst,
+        linkedid,
+        file_size,
+        queue_date,
+    ):
+        logging.info("save result start")
+        print("=== save_result", accept_text)
+
+
+def main():
+    parser = argparse.ArgumentParser(
+        description="Send an audio file to the FastAPI server for processing."
+    )
+    parser.add_argument(
+        "--file", type=str, required=True, help="File path of the audio file"
+    )
+    args = parser.parse_args()
+
+    server = Server()
+    num_sentences, phrases, confidences = server.transcribation_process(
+        original_file_name=args.file
+    )
+    print(f"Processed {num_sentences} sentences.")
+
+
+if __name__ == "__main__":
+    main()

compose.sh

@@ -3,4 +3,6 @@
 
 # Compose, updating all files
 # sudo docker compose up --force-recreate --build
-sudo docker compose up --build
+
+# Compose, remove lod container versions
+sudo docker compose up --build -d --remove-orphans --force-recreate

docker-compose.yml

@@ -4,14 +4,13 @@ services:
   whisper-timestamped:
     container_name: whisper-timestamped
     ports:
-      - "8000:8000"
+      - "8888:8888"
     restart: unless-stopped
     build:
       context: .
       dockerfile: Dockerfile
     volumes:
       - ./input/:/app/input
-      - ./output:/app/output
       - ./cache:/app/cache
       - ./logs:/app/logs
     deploy:
@@ -21,4 +20,4 @@ services:
           - driver: nvidia
             device_ids: ['0']
             capabilities: [gpu]
-    command: ["gunicorn", "transcribe:app", "--workers", "1", "--worker-class", "uvicorn.workers.UvicornWorker", "--bind", "0.0.0.0:8000", "--timeout", "240"]
+    command: ["gunicorn", "transcribe:app", "--workers", "1", "--worker-class", "uvicorn.workers.UvicornWorker", "--bind", "0.0.0.0:8888", "--timeout", "1800"]

requirements.txt

@@ -1,8 +1,6 @@
 setuptools-rust==1.8.1
 tiktoken==0.5.2
-pandas==2.2.0
-pyarrow==15.0.0
-fastapi==0.109.0
-python-multipart==0.0.6
+fastapi==0.110.1
+python-multipart==0.0.7
 uvicorn==0.27.0
-gunicorn==21.2.0
+gunicorn==22.0.0