Tag
Explore by tags
Applies quantization, pruning, distillation, speculative decoding and sparsity to compress and accelerate deep learning models, producing exportable checkpoints ready for deployment on TensorRT‑LLM, TensorRT, vLLM and similar inference runtimes.
Build AI workflows once and run them across model providers — GoogleAI, OpenAI, Claude, Ollama — through one SDK. Composable primitives for RAG, tool use, and agents, plus a local dev UI for tracing and debugging, with SDKs in JS/TS, Go, and Python.
Disaggregated LLM serving architecture that splits prefill and decode into separate clusters and pools spare CPU, DRAM, and SSD into a distributed KVCache. Powers Kimi in production, handling 75% more requests under the same SLOs.
GPU‑accelerated framework for training physically simulated humanoid characters and robots using reinforcement learning and motion imitation. Provides a modular multi‑backend simulator stack, large‑scale multi‑GPU training recipes, built‑in motion retargeting and an ONNX deployment pathway to real robots.
High-resolution image and video generation codebase and models that run with far lower compute and memory than typical diffusion systems. Uses linear-attention DiT variants, aggressive latent compression, and inference-scaling to support text-to-image (up to 4K), fast one/few-step generation, and efficient video pipelines.
A 671B-parameter Mixture-of-Experts language model (37B activated) trained on 14.8T tokens with 128K context, FP8-first training, a Multi-Token Prediction module, and Hugging Face weights—focused on efficient MoE training and long-context use cases.
Provides a hardware plugin that runs vLLM on Huawei Ascend NPUs by mapping vLLM execution and memory management to the Ascend runtime. Key features: support for Transformer/MoE/embedding/multimodal models, official docs, CI-backed release branches and community maintenance.
Zero-shot, single‑reference voice cloning TTS with multilingual support (ZH/EN/JA/ES/AR), fine-grained emotion and duration control, and pronunciation hooks (Pinyin/CMU/Kana); ships model weights, Web UI and production deployment recipes for local or server use.
Runs a self-hosted meeting bot and transcription API that joins Google Meet, Teams and Zoom and streams speaker-attributed transcripts in real time. Compiles meetings into a git-backed Markdown workspace and runs sandboxed agents on your infrastructure; Apache-2.0 and air-gap capable.
A vision-language-action foundation model and reference stack for generalized humanoid and cross-embodiment robot manipulation. Provides pretrained checkpoints, demo datasets, and tooling for fine-tuning, evaluation, and deployment (ONNX/TensorRT); released as Early Access.
Runs Cloudflare Workers and Durable Objects on self-hosted nodes, storing each object as an independently replicated SQLite database in an S3-compatible bucket—enabling per-object sharding, hibernation, and ownership via object-storage compare-and-swap without a central control plane.
Federates MCP, A2A, and REST/gRPC backends behind a single gateway endpoint with centralized discovery, governance, and observability; optimizes agent and tool calling. Includes gRPC→MCP translation, plugin extensibility, OpenTelemetry tracing, and Kubernetes-ready deployment.