ZSvedic

Pinned Repositories

dpo-rlaif
Language:Jupyter Notebook95 3 410
direct-preference-optimization
Reference implementation for DPO (Direct Preference Optimization)
Language:Python2.4k 18 84201
DPO-fail
This project demonstrates a failure to fine-tune Llama 3.1 with DPO (Direct Preference Optimization) on a custom dataset.
Language:Jupyter Notebook00
EBH-book
Evidence-Based Hiring book, in markdown, with pandoc build scripts for HTML, EPUB and PDF.
Language:TeX5 3 233
educational-transformer
Easy-to-follow, educational implementation of the transformer model in PyTorch.
Language:Python3 2 00
reddit-filter
Python scripts to filter the "humor-chains" dataset from Hugging Face.
Language:Python0 2 00
ShortGPT
Language:Jupyter Notebook00
WebBS-Calculator
Source code for Web Bloat Score Calculator.
Language:JavaScript15 3 51

ZSvedic's Repositories

ZSvedic/WebBS-Calculator
Source code for Web Bloat Score Calculator.
Language:JavaScript15 3 51
ZSvedic/EBH-book
Evidence-Based Hiring book, in markdown, with pandoc build scripts for HTML, EPUB and PDF.
Language:TeX5 3 233
ZSvedic/educational-transformer
Easy-to-follow, educational implementation of the transformer model in PyTorch.
Language:Python3 2 00
ZSvedic/DPO-fail
This project demonstrates a failure to fine-tune Llama 3.1 with DPO (Direct Preference Optimization) on a custom dataset.
Language:Jupyter Notebook00
ZSvedic/reddit-filter
Python scripts to filter the "humor-chains" dataset from Hugging Face.
Language:Python0 2 00
ZSvedic/ShortGPT
Language:Jupyter Notebook00