Upload 5 files

README.md

@@ -1,12 +1,13 @@
----
-title: Testnemoasr
-emoji: 💻
-colorFrom: gray
-colorTo: red
-sdk: gradio
-sdk_version: 5.5.0
-app_file: app.py
-pinned: false
----
-
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+---
+title: test nemo
+emoji: 🙄
+colorFrom: indigo
+colorTo: purple
+sdk: gradio
+sdk_version: 4.44.0
+app_file: app.py
+pinned: false
+license: mit
+---
+
+Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

app.py

@@ -0,0 +1,53 @@
+import os
+if os.environ.get("SPACES_ZERO_GPU") is not None:
+    import spaces
+else:
+    class spaces:
+        @staticmethod
+        def GPU(func):
+            def wrapper(*args, **kwargs):
+                return func(*args, **kwargs)
+            return wrapper
+import gradio as gr
+import subprocess
+
+#subprocess.run("git clone https://github.com/AI4Bharat/NeMo.git && cd NeMo && git checkout nemo-v2 && bash reinstall.sh", shell=True)
+
+import torch
+import nemo.collections.asr as nemo_asr
+
+from pathlib import Path
+
+model = nemo_asr.models.ASRModel.from_pretrained("ai4bharat/indicconformer_stt_ml_hybrid_rnnt_large")
+
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+model.freeze() # inference mode
+model = model.to(device) # transfer model to device
+
+@spaces.GPU
+def infer(srcfile: str):
+    tmpfile = "sample_audio_infer_ready.wav"
+
+    subprocess.run(f"ffmpeg -i {srcfile} -ac 1 -ar 16000 {tmpfile}", shell=True)
+    model.cur_decoder = "ctc"
+    ctc_text = model.transcribe([tmpfile], batch_size=1, logprobs=False, language_id='ml')[0]
+    print(ctc_text)
+
+    model.cur_decoder = "rnnt"
+    rnnt_text = model.transcribe([tmpfile], batch_size=1, language_id='ml')[0]
+    print(rnnt_text)
+
+    if Path(tmpfile).exists(): Path(tmpfile).unlink()
+
+    return ctc_text, rnnt_text
+
+with gr.Blocks() as demo:
+    input_audio = gr.Audio(label="Input", type="filepath", sources=["upload", "microphone"], format="wav")
+    run_button = gr.Button("Run", variant="primary")
+    with gr.Row():
+        ctc_text = gr.Textbox(label="CTC", value="", show_copy_button=True)
+        rnnt_text = gr.Textbox(label="RNNT", value="", show_copy_button=True)
+
+    run_button.click(infer, [input_audio], [ctc_text, rnnt_text])
+
+demo.launch()

packages.txt

@@ -0,0 +1 @@
+ffmpeg

pre-requirements.txt

@@ -0,0 +1 @@
+pip>=24.1

requirements.txt

@@ -0,0 +1,51 @@
+huggingface_hub===0.20.3
+torch
+git+https://github.com/AI4Bharat/NeMo
+numba
+numpy<2
+onnx>=1.7.0
+python-dateutil
+ruamel.yaml
+scikit-learn
+setuptools>=65.5.1
+tensorboard
+text-unidecode
+torch
+tqdm>=4.41.0
+triton
+wget
+wrapt
+datasets
+inflect
+pandas
+sacremoses>=0.0.43
+sentencepiece<1.0.0
+braceexpand
+editdistance
+g2p_en
+ipywidgets
+jiwer
+kaldi-python-io
+kaldiio
+lhotse>=1.20.0
+librosa>=0.10.0
+marshmallow
+matplotlib
+packaging
+pyannote.core
+pyannote.metrics
+pydub
+pyloudnorm
+resampy
+ruamel.yaml
+scipy>=0.14
+soundfile
+sox
+texterrors
+hydra-core>1.3,<=1.3.2
+omegaconf<=2.3
+pytorch-lightning>=2.2.1
+torchmetrics>=0.11.0
+transformers>=4.36.0
+wandb
+webdataset>=0.2.86