LAION-AI / LAION-AI/CLAP

Fixed parameters for audio and text encoders

Open
#106 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
2.3k
Forks
213
PR merge metrics
No merged PRs in 30d

Description

While performing contrastive pre-training on the audio-text pairs dataset, did you fix the paramters of both the encoders and optimise the parameters of MLP layer only? Also, I was wondering if you use the pre-trained audio encoder (PANN and HTSAT) or you optimise the parameters from scratch? I don't seem to find the detailed information from the paper. Great work, and thanks in advance.

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Review the paper and the contrastive pre-training setup for the audio-text pairs dataset, focusing on whether the audio and text encoders are fixed or optimized and whether PANN and HTSAT are pretrained. Done means providing a clear, documented answer to these training questions.

Written by the indexing model from the issue text.

Assessment

Domain
machine-learning
Issue type
Documentation
Difficulty
1/5
Estimated time
Under an hour
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.