Fargan INT16 quantization and FP16 quantization
Open
Nobody has claimed this yet.
- Dominant language
- C
- Stars
- 3.3k
- Forks
- 808
- PR merge metrics
- No merged PRs in 30d
Description
During quantization in the Fargan module, at line 184 of common.py, scale_sum = weight_max_sum / 129- I'm not sure why this conversion needs to be configured with 129.Additionally, I'd like to know if there exists any code related to int16 and f16 quantization?
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start at line 184 of common.py and inspect how scale_sum is calculated in the Fargan quantization path. Search the Fargan module for existing INT16 and FP16 quantization code, then document the purpose of the 129 conversion and identify whether those quantization paths already exist.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- c, python
- Domain
- audio-video-rtc
- Issue type
- Feature
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100