Pinned Loading

1. opforge opforge Public

   An operator/kernel framework: write with JIT (Triton / CuTeDSL), ship with AOT, and declare numeric contracts explicitly - one kernel for both training and inference.

   Python 1

2. llm-gateway llm-gateway Public

   Engine-agnostic, state-aware gateway for LLM inference: an OpenAI/Anthropic-compatible router + an optional vLLM plugin + a pluggable state layer for KV-cache-aware and latency-aware routing.

   Go 1

3. quantkit quantkit Public

   A dependency-light, transformers-free toolkit for data-free weight-only quantization (FP8 blockwise, MXFP8, MXFP4), with file-level multi-node support and HF / compressed-tensors output.

   Python 1

4. stepfun-ai/Step-Audio-EditX stepfun-ai/Step-Audio-EditX Public

   A powerful 3B-parameter, LLM-based Reinforcement Learning audio edit model excels at editing emotion, speaking style, and paralinguistics, and features robust zero-shot text-to-speech

   Python 981 80

5. stepfun-ai/StepDeepResearch stepfun-ai/StepDeepResearch Public

   Step-DeepResearch

   Python 574 26

6. vllm-project/vllm vllm-project/vllm Public

   A high-throughput and memory-efficient inference and serving engine for LLMs

   Python 93.4k 23.1k