git
ask
hub
Privacy
Terms
Sign in with GitHub
jina-ai/audio-embedding-kickstarter
↗
Stars ·
10
Language ·
Python
Simplify and Visualize This Repo
Scan the safety of this repo
Self Host this repo
Check out more work of Developer: jina-ai
Ask anything about this repo to start.
Help me set up the Audio Embedding Kickstarter project locally and run the evaluation scripts on the AudioCaps dataset to reproduce the retrieval accuracy results.
Compare the four model-shrinking strategies from the Audio Embedding Kickstarter project and explain which approach gives the best balance of model size versus retrieval accuracy.
Using the training configs from the Audio Embedding Kickstarter repo, guide me through fine-tuning a Qwen2.5-Omni model on my own audio-text pairs for a custom audio search application.
Walk me through the PDF report in this repo and summarize which training approach worked best for keeping the model under 1.2 billion parameters.
Full explanation on explaingit →
📎
Send
By chatting or signing in you agree to the
Terms
and chat-message logging (revocable in
History
).