update

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+*.wav filter=lfs diff=lfs merge=lfs -text

.gitignore

@@ -0,0 +1,15 @@
+
+data/
+dotenv/
+pretrained_models
+
+.git/
+.idea/
+
+**/cache/
+**/__pycache__/
+
+**/*.env
+**/*.mp3
+**/*.png
+**/*.wav

Dockerfile

@@ -0,0 +1,30 @@
+# read the doc: https://huggingface.co/docs/hub/spaces-sdks-docker
+# you will also find guides on how best to write your Dockerfile
+
+FROM python:3.6
+
+WORKDIR /code
+
+COPY ./requirements.txt /code/requirements.txt
+
+RUN pip install --upgrade pip
+
+RUN pip install --no-cache-dir --upgrade -r /code/requirements.txt
+
+# Set up a new user named "user" with user ID 1000
+RUN useradd -m -u 1000 user
+
+# Switch to the "user" user
+USER user
+
+# Set home to the user's home directory
+ENV HOME=/home/user \
+    PATH=/home/user/.local/bin:$PATH
+
+# Set the working directory to the user's home directory
+WORKDIR $HOME/app
+
+# Copy the current directory contents into the container at $HOME/app setting the owner to the user
+COPY --chown=user . $HOME/app
+
+CMD ["python", "main.py"]

README.md

@@ -1,5 +1,5 @@
 ---
-title: Asr
+title: ASR
 emoji: 📈
 colorFrom: purple
 colorTo: green

examples/wenet/test.py

@@ -0,0 +1,80 @@
+#!/usr/bin/python3
+# -*- coding: utf-8 -*-
+import argparse
+from pathlib import Path
+
+import huggingface_hub
+import sherpa
+
+from project_settings import project_path
+
+
+def get_args():
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--repo_id",
+        default="csukuangfj/wenet-chinese-model",
+        # default="csukuangfj/wenet-english-model",
+        type=str
+    )
+
+    parser.add_argument("--model_filename", default="final.zip", type=str)
+    parser.add_argument("--tokens_filename", default="units.txt", type=str)
+
+    parser.add_argument(
+        "--pretrained_model_dir",
+        default=(project_path / "pretrained_models").as_posix(),
+        type=str
+    )
+    args = parser.parse_args()
+    return args
+
+
+def main():
+    args = get_args()
+
+    pretrained_model_dir = Path(args.pretrained_model_dir)
+    pretrained_model_dir.mkdir(exist_ok=True)
+
+    model_dir = pretrained_model_dir / "huggingface" / args.repo_id
+    model_dir.mkdir(exist_ok=True)
+
+    print("download model")
+    model_filename = huggingface_hub.hf_hub_download(
+        repo_id=args.repo_id,
+        filename=args.model_filename,
+        subfolder=".",
+        local_dir=model_dir.as_posix(),
+    )
+    print(model_filename)
+
+    print("download tokens")
+    token_filename = huggingface_hub.hf_hub_download(
+        repo_id=args.repo_id,
+        filename=args.tokens_filename,
+        subfolder=".",
+        local_dir=model_dir.as_posix(),
+    )
+    print(token_filename)
+
+    feat_config = sherpa.FeatureConfig(normalize_samples=False)
+    feat_config.fbank_opts.frame_opts.samp_freq = sample_rate
+    feat_config.fbank_opts.mel_opts.num_bins = 80
+    feat_config.fbank_opts.frame_opts.dither = 0
+
+    config = sherpa.OfflineRecognizerConfig(
+        nn_model=nn_model,
+        tokens=tokens,
+        use_gpu=False,
+        feat_config=feat_config,
+        decoding_method=decoding_method,
+        num_active_paths=num_active_paths,
+    )
+
+    recognizer = sherpa.OfflineRecognizer(config)
+
+    return
+
+
+if __name__ == "__main__":
+    main()

main.py

@@ -0,0 +1,102 @@
+#!/usr/bin/python3
+# -*- coding: utf-8 -*-
+import argparse
+
+import gradio as gr
+
+from examples import examples
+from project_settings import project_path
+
+
+def get_args():
+    parser = argparse.ArgumentParser()
+    parser.add_argument(
+        "--examples_dir",
+        default=(project_path / "data/examples").as_posix(),
+        type=str
+    )
+    parser.add_argument(
+        "--trained_model_dir",
+        default=(project_path / "trained_models").as_posix(),
+        type=str
+    )
+    args = parser.parse_args()
+    return args
+
+
+def update_model_dropdown(language: str):
+    if language in language_to_models:
+        choices = language_to_models[language]
+        return gr.Dropdown(
+            choices=choices,
+            value=choices[0],
+            interactive=True,
+        )
+
+    raise ValueError(f"Unsupported language: {language}")
+
+
+def main():
+    title = "# Automatic Speech Recognition with Next-gen Kaldi"
+
+    language_choices = ["Chinese"]
+
+    language_to_models = {
+        "Chinese": ["None"]
+    }
+
+    # components
+    language_radio = gr.Radio(
+        label="Language",
+        choices=language_choices,
+        value=language_choices[0],
+    )
+    model_dropdown = gr.Dropdown(
+        choices=language_to_models[language_choices[0]],
+        label="Select a model",
+        value=language_to_models[language_choices[0]][0],
+    )
+
+    language_radio.change(
+        update_model_dropdown,
+        inputs=language_radio,
+        outputs=model_dropdown,
+    )
+
+
+    # blocks
+    with gr.Blocks() as blocks:
+        gr.Markdown(value=title)
+
+        with gr.Tabs():
+            with gr.TabItem("Upload from disk"):
+                uploaded_file = gr.Audio(
+                    sources=["upload"],
+                    type="filepath",
+                    label="Upload from disk",
+                )
+                upload_button = gr.Button("Submit for recognition")
+                uploaded_output = gr.Textbox(label="Recognized speech from uploaded file")
+                uploaded_html_info = gr.HTML(label="Info")
+
+                gr.Examples(
+                    examples=examples,
+                    inputs=[
+                        language_radio,
+                        model_dropdown,
+                        decoding_method_radio,
+                        num_active_paths_slider,
+                        punct_radio,
+                        uploaded_file,
+                    ],
+                    outputs=[uploaded_output, uploaded_html_info],
+                    fn=process_uploaded_file,
+                )
+
+    blocks.queue().launch()
+
+    return
+
+
+if __name__ == "__main__":
+    main()

project_settings.py

@@ -0,0 +1,12 @@
+#!/usr/bin/python3
+# -*- coding: utf-8 -*-
+import os
+from pathlib import Path
+
+
+project_path = os.path.abspath(os.path.dirname(__file__))
+project_path = Path(project_path)
+
+
+if __name__ == '__main__':
+    pass

requirements.txt

@@ -0,0 +1,2 @@
+gradio==4.29.0
+https://huggingface.co/csukuangfj/wheels/resolve/main/2023-01-30/k2_sherpa-1.1-cp38-cp38-linux_x86_64.whl