Upload folder using huggingface_hub

.gitignore

@@ -0,0 +1,7 @@
+.venv
+.gradio
+*.wav
+*.mp4
+*.m4a
+*.swp
+outputs

.gradio/certificate.pem

@@ -0,0 +1,31 @@
+-----BEGIN CERTIFICATE-----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=
+-----END CERTIFICATE-----

.python-version

@@ -0,0 +1 @@
+3.11

README.md

@@ -1,12 +1,6 @@
 ---
-title: Video2text
-emoji: 🦀
-colorFrom: yellow
-colorTo: blue
+title: video2text
+app_file: main.py
 sdk: gradio
 sdk_version: 5.29.1
-app_file: app.py
-pinned: false
 ---
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

main.py

@@ -0,0 +1,120 @@
+#! /usr/bin/env -S uv run
+# /// script
+# requires-python = ">=3.10"
+# dependencies = ["yt-dlp", "whisper-cpp-pybind", "faster-whisper", "gradio"]
+# ///
+
+# dependencies = ["yt-dlp", "mlx-whisper", "sherpa-onnx"]
+# dependencies = ["yt-dlp", "soundfile", "pywhispercpp", "mlx-whisper"]
+
+from pathlib import Path
+
+# import mlx_whisper
+import yt_dlp
+# from pywhispercpp.model import Model as whisper_cpp
+from whispercpp import Whisper
+
+def download(url):
+  ydl_opts = {
+    "format": "m4a/bestaudio/best",
+    "postprocessors": [
+      {
+        "key": "FFmpegExtractAudio",
+        "preferredcodec": "wav",
+      }
+    ],
+    "outtmpl": "%(id)s.%(ext)s",
+    "postprocessor_args": [
+      "-ar",
+      "16000",
+      "-ac",
+      "1",
+    ],
+  }
+
+  with yt_dlp.YoutubeDL(ydl_opts) as ydl:
+    error_code = ydl.download([url])
+    if error_code != 0:
+      raise Exception("Download failed")
+
+    info = ydl.extract_info(url, download=False)
+    # print(info["id"])
+    return info["id"] + ".wav"
+
+
+def asr_mlx(wav_file):
+  result = mlx_whisper.transcribe(
+    wav_file,
+    path_or_hf_repo="mlx-community/whisper-large-v3-turbo",
+  )
+
+  for seg in result["segments"]:
+    print(seg["text"])
+
+
+def asr_cpp(wav_file):
+  from whisper_cpp import Whisper
+  whisper = Whisper("/home/ubuntu/code/whisper.cpp/models/ggml-large-v3-turbo.bin")
+  # print(wav_file)
+  whisper.transcribe(wav_file, diarize=True)
+  whisper.output(output_csv=True, output_jsn=True, output_lrc=True, output_srt=True, output_txt=True, output_vtt=True, log_score=True)
+
+def asr_fast(wav_file):
+  from faster_whisper import WhisperModel
+  model = WhisperModel('large-v3-turbo', device="cpu", compute_type="float32")
+  # model = Whisper.from_pretrained(
+  #   "ggml-large-v3-turbo.bin",
+  #   basedir="/home/ubuntu/code/whisper.cpp/models/",  # for dev
+  # )
+  segments, info = model.transcribe(
+    wav_file, language="zh", initial_prompt="以下是普通话的句子。"
+  )
+  # print(info)
+
+  # result = [s.text for s in segments if '请不吝点赞' not in s.text]
+  result = [s.text for s in segments]
+  print(result)
+  return '\n'.join(result)
+
+
+import gradio as gr
+
+def handleURL(url):
+    f = download(url)
+    return asr_fast(f)
+    # return "Hello, " + name + "!" * int(intensity)
+
+# demo = gr.Interface( fn=handleURL, inputs=["text"], outputs=["text"])
+
+with gr.Blocks() as demo:
+    gr.Markdown(
+    """
+    # fetch video from url, and transcibe to text
+    ---
+    
+    running very poor server with performance, maybe 1.5x time cost with orignal video processing.
+    try yourself locally if your have macbook with silicon or high performance GPU.
+    <https://gist.github.com/yanyaoer/5cc7b0dd6729f306ad3cb740d501cabd#file-0-video2text-py>
+
+    """)
+    name = gr.Textbox(label="video url for transcibe")
+    output = gr.Textbox(label="Output")
+    greet_btn = gr.Button("submit")
+    greet_btn.click(fn=handleURL, inputs=name, outputs=output)
+
+
+if __name__ == "__main__":
+  import sys
+
+  u = (
+    sys.argv[1] if len(sys.argv) > 1 else "https://www.bilibili.com/video/BV1ZMNrejEnH/"
+  )
+  # w = download(u)
+  # w = "/tmp/BV1ZMNrejEnH.wav"
+  # w = "/tmp/1746897004.wav"
+  # w = "BV1ZMNrejEnH.wav"
+  # w = "abc.wav"
+  # asr_fast(w)
+  # asr_mlx(w)
+  # print(r)
+  demo.launch(share=True)