A straightforward method for training your LLM, from downloading data to generating text.
Commits on main
docs CI: deploy via official GitHub Pages Actions (Source=GitHub Actions)
d46f508 FareedKhan-dev committed 5 months ago
Merge pull request #17 from FareedKhan-dev/feat/post-training-configs-ui-docs
0bd37cb Fareed Khan committed 5 months ago
Add post-training suite tooling: JSON configs, Streamlit UI, MkDocs site
ff00dae FareedKhan-dev committed 5 months ago
docs: keep the author's original illustrative parameter count
fecb635 FareedKhan-dev committed 5 months ago
docs: restore training-script teaching comments alongside #15 diagnostics
35231f6 FareedKhan-dev committed 5 months ago
Merge pull request #15 from XiaoBinGan/rtx5090-validation-and-compat
5b51974 FareedKhan-dev committed 5 months ago
docs: sync README and docstrings with the new attention projection (#13)
3b2cf0f FareedKhan-dev committed 5 months ago
Merge pull request #13 from Chamath-Adithya/improve-llm-core
b6d12a5 FareedKhan-dev committed 5 months ago
Improve modern GPU compatibility and document RTX 5090 validation
dfe8db0 吴佳浩 committed 5 months ago
fix: add missing output projection in MultiHeadAttention and optimize training
dfb9a11 Chamath Adithya committed 5 months ago
Merge pull request #12 from Jah-yee/fix-readme-broken-resume
f3524df Fareed Khan committed 6 months ago