explosion/curated-transformers

View on GitHub

🤖 A PyTorch library of curated Transformer models and their composable components

Stars: 890
Forks: 36
Open beginner issues: 0
Indexed issues: 17

Dominant language: Python
License: MIT
Last GitHub push: Apr 17, 2024
Latest indexed: Sep 15, 2026
Contributing guide: No contributing guide
Code of conduct: No code of conduct
Beginner labels: No beginner labels indexed
PR merge metrics: No merged PRs in 30d

Open All

17 open issues indexed

Thoughts on jaxtyping Open

feat/misc type/feature

explosion/curated-transformers#246 · 1 comment · 2 reactions · 0 assignees · Indexed Sep 15, 2026
Output logits for generation Open

feat/generation type/feature

explosion/curated-transformers#311 · 1 comment · 0 reactions · 0 assignees · Indexed Sep 15, 2026
Add Low-Rank Adapters injection into base models Open

feat/training type/feature

explosion/curated-transformers#312 · 3 comments · 1 reaction · 0 assignees · Indexed Sep 15, 2026
Optimal Qlora settings Open

feat/training type/feature

explosion/curated-transformers#316 · 1 comment · 1 reaction · 0 assignees · Indexed Sep 15, 2026
Support for Encoder-Decoder-style architectures Open

feat/model type/feature

explosion/curated-transformers#340 · 2 comments · 3 reactions · 0 assignees · Indexed Sep 15, 2026
Add support for Mistral Open

feat/model type/feature

explosion/curated-transformers#341 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 15, 2026
Option to only return the last hidden layer output from models Open

feat/model type/feature

explosion/curated-transformers#342 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 15, 2026
Convert QKV projection splitting methods into Torch modules Open

feat/layers type/maintenance

explosion/curated-transformers#343 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 15, 2026
Make `QkvMode` ADT-like Open

feat/layers type/maintenance

explosion/curated-transformers#344 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 15, 2026
Expose more outputs through the `Generator` interface Open

feat/generation type/feature

explosion/curated-transformers#345 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 15, 2026
Add a an extras/contrib package Open

type/maintenance

explosion/curated-transformers#347 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 15, 2026
Support DeBERTa v2/3 Open

feat/model type/feature

explosion/curated-transformers#348 · 1 comment · 0 reactions · 0 assignees · Indexed Sep 15, 2026
Add support for attention sinks Open

feat/layers feat/model type/feature

explosion/curated-transformers#350 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 15, 2026
Move the old Falcon architecuture to the extras/addons pacakage Open

type/maintenance

explosion/curated-transformers#355 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 15, 2026
Add suggested PyTorch LLM optimizations Open

feat/generation feat/model

explosion/curated-transformers#356 · 0 comments · 2 reactions · 0 assignees · Indexed Sep 15, 2026
Truncation of sequences that are beyond the model's maximum length Open

feat/tokenization type/bug type/feature

explosion/curated-transformers#359 · 2 comments · 1 reaction · 0 assignees · Indexed Sep 15, 2026
📋 Documentation Enhancement Suggestion Open

explosion/curated-transformers#379 · 0 comments · 0 reactions · 0 assignees · Indexed Sep 15, 2026

explosion/curated-transformers

Get new issues in your inbox