BlinkDL/RWKV-LM

RWKV (pronounced RwaKuv) is an RNN with great LLM performance, which can also be directly trained like a GPT transformer (parallelizable). We are at RWKV-7 "Goose". So it's combining the best of RNN and transformer - great performance, linear time, constant space (no kv-cache), fast training, infinite ctx_len, and free sentence embedding.

57
/ 100
Established

Implements an attention-free RNN architecture with state-passing mechanisms that enable meta-in-context learning—adapting model state via gradient descent at inference time. RWKV-7 achieves linear-time complexity with zero KV-cache overhead while maintaining training parallelizability through careful initialization (PreLN LayerNorm), selective weight decay on projection matrices only, and differentiated learning rates per parameter. Offers production-ready inference via CUDA kernels (206k tokens/s on H100 clusters), mobile libraries, and Windows/Office runtime integration, with reference training code requiring only 10GB VRAM on a single GPU.

14,414 stars. Actively maintained with 2 commits in the last 30 days.

No Package No Dependents
Maintenance 13 / 25
Adoption 10 / 25
Maturity 16 / 25
Community 18 / 25

How are scores calculated?

Stars

14,414

Forks

997

Language

Python

License

Apache-2.0

Last pushed

Mar 05, 2026

Commits (30d)

2

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/transformers/BlinkDL/RWKV-LM"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.