pretbc

Results 12 comments of pretbc

Add VATT model

any news about VATT PyTorch implementation ?

ffmpeg._run.Error: ffmpeg error (see stderr output for detail)

Got same error on colab cmd - > `!python examples/generate_from_file.py --csv_path=/content/dmvr/input.csv --output_path=/content/drive/MyDrive/Raw/generated_dataset --decode_audio=True` ![image](https://github.com/google-deepmind/dmvr/assets/89849350/8718ef05-f62d-4736-91ea-7aed68cae198) paths provided in input.csv for video '/content/drive/MyDrive/Raw/` i assume audio gonna be decoded from video file...

What does the feature dimension represent?

It seems that link expire - could You tell more ?

audio data

Can U tell more about Ur approach ?

Jupyter Lab crashing with faster-whisper

any news here ?

add audio utils to handle model audio input

let me check pytests

add audio utils to handle model audio input

Did locally - `ruff check . --fix-only`

add audio utils to handle model audio input

I will suggest as demo below snippet ``` lm = dspy.LM( "gemini-2.0-flash-exp", api_key=os.getenv("GOOGLE_APPLICATION_CREDENTIALS") ) dspy.configure(lm=lm) audio_path = "path/to/file.wav" audio_data = pathlib.Path(audio_path).read_bytes() audio_data_base64 = base64.b64encode(audio_data).decode("utf-8") messages = [ { "role": "user",...

add audio utils to handle model audio input

gonna check any conflict after review

Subagent Agent Transfers to Parent Agent

is this could be a part of RemoteA2aAgent ?