AI API2020
Typed Python client for the OpenAI REST API that offers synchronous and asynchronous clients, typed request/response models, streaming and Realtime support, webhook verification, and integrations for Azure and Amazon Bedrock—built for production integrations and automation.
AI Deploy2021
Deploys trained SavedModels behind gRPC and REST endpoints, with hot-swappable versioning so new weights load without downtime. Built around servables, loaders, sources, and a manager, plus request batching to cut accelerator cost.
AI Infra2021
Drop-in transformer building blocks with custom CUDA kernels: memory-efficient exact attention (up to ~10x faster), block-sparse attention, fused softmax/layernorm/SwiGLU ops. Cuts VRAM and speeds up diffusion and LLM training on Nvidia GPUs.
AI Train2021
HPC-AI Technology Inc. (Colossal-AI team), Shenggui Li +1HPC-AI Technology Inc., National University of Singapore
Scales a single-GPU training script to thousands of GPUs through a unified interface, combining data, pipeline, tensor, and sequence parallelism. Its Gemini memory manager offloads tensors across GPU, CPU, and NVMe so models far larger than VRAM still fit.
AI Image2021
Library for benchmarking, developing, and deploying deep-learning visual anomaly-detection models — includes ready-to-use model implementations (PatchCore, DINO-based), experiment/HPO tooling, OpenVINO export for edge inference, and a low-code Studio for deployment.
AI Model2022
Encodes sentences and short paragraphs into 384-dimensional semantic vectors for retrieval, clustering, and similarity tasks. Fine-tuned from MiniLM-L6-H384 on ~1.17B sentence pairs with a contrastive objective; defaults to 256 word-piece truncation.
AI Model2022
Autoregressive transformer that generates English text from a prompt and can be fine-tuned for downstream NLP tasks. This is the 124M-parameter GPT-2 checkpoint (causal LM, byte-level BPE, 1024-token context) hosted on Hugging Face.
AI Audio2022
edge-tts is a Python module that enables the use of Microsoft Edge's online text-to-speech service directly from Python code or via command-line tools like edge-tts and edge-playback, without requiring Microsoft Edge, Windows, or an API key.
AI Infra2022
Fused CUDA kernels that compute exact attention without ever writing the full N×N score matrix to GPU memory, cutting memory from quadratic to linear and speeding up training and inference on A100/H100. Ships FlashAttention-2/3 plus KV-cache decode paths.
AI Model2022
Transformer-based foundation model for tabular data that provides pre-trained checkpoints for fast classification and regression, with GPU-accelerated local inference and an optional cloud client. Best suited for small-to-medium datasets (~≤50k rows).
AI Agent2022
Build LLM-powered agents and applications from modular components: provider-agnostic model abstractions, tool integrations, retrievers for RAG, and agent orchestration primitives. Suited for prototyping and production agent workflows; requires developer wiring and dependency management.
Embodied AI2022
GPU-accelerated robot-learning framework on NVIDIA Isaac Sim, running thousands of parallel environments on one GPU for reinforcement and imitation learning. Ships 30+ ready-to-train tasks and 16+ robot models wired to RSL RL, SKRL, and RL Games.