|
Download README.md from codysyun/lora-retriever: direct link, hf CLI and curl.
- Browser
- Download file 698 Bytes
-
https://huggingface.co/codysyun/lora-retriever/resolve/main/README.md
- Command line
-
hf download hf://codysyun/lora-retriever/README.md
-
curl -L -o README.md https://huggingface.co/codysyun/lora-retriever/resolve/main/README.md
698 Bytes
metadata
license: cc-by-4.0
tags:
- base
- hybrid
- kaiming
- lamb
- matching
- onecycle
- relu
- scalenorm
- standard
- tucker
predict.py
Model Overview
A base-scale implementation of the hybrid architecture, built for matching tasks.
Architecture
- Architecture: hybrid
- Scale: base
- Attention: standard
- Fusion strategy: tucker
- Task head: matching
- Activation: relu
- Normalization: scalenorm
- Initialization: kaiming
Training
- Optimizer: lamb
- LR scheduler: onecycle
Files
predict.py— main artifact of this repository
License
See the license field above.