BlinkDL のリポジトリ

Path tracing demo with JS in your web browser. 用浏览器JS做路径跟踪渲染。

最終コミット 2018/06/11

Project Ivory is a simple forum written a few years ago.

最終コミット 2020/05/16

The CUDA version of the RWKV language model ( https://github.com/BlinkDL/RWKV-LM )

最終コミット 2025/12/10

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.

最終コミット 2026/08/14

RWKV-v2-RNN trained on the Pile. See https://github.com/BlinkDL/RWKV-LM for details.

最終コミット 2022/09/14

最終コミット 2026/08/11

LayerNorm(SmallInit(Embedding)) in a Transformer to improve convergence

最終コミット 2022/02/21

My vibe math results

最終コミット 2026/08/20

Let us make Psychohistory (as in Asimov) a reality, and accessible to everyone. Useful for LLM grounding and games / fiction / business / finance / governance, and can align agents with human too.

最終コミット 2023/04/09

Prepare for DeekSeek R1 inference: Benchmark CPU, DRAM, SSD, iGPU, GPU, ... with efficient code.

最終コミット 2025/02/02

An implementation of model parallel autoregressive transformers on GPUs, based on the DeepSpeed library.

最終コミット 2022/02/25

LiBai(李白): A Toolbox for Large-Scale Distributed Parallel Training

最終コミット 2022/09/05

A *tuned* minimal PyTorch re-implementation of the OpenAI GPT (Generative Pretrained Transformer) training

最終コミット 2021/08/09

RWKV-7: Surpassing GPT

最終コミット 2024/11/17

RWKV in nanoGPT style

最終コミット 2023/12/12

Understand and test language model architectures on synthetic tasks.

最終コミット 2025/06/08