ML Frameworks

It split off from `verl` to give diffusion, video, and omni-modality models an RL post-training framework that doesn't treat them like chatbots.

Pits PatchTST, TFT, and N-HiTS against CatBoost to help energy markets pick a forecasting model.

NeMo AutoModel automates the busywork of wiring HuggingFace LLMs and VLMs into PyTorch-native distributed training so you can fine-tune or pretrain without hand-rolling parallelism code.

KTransformers makes CPU-GPU heterogeneous inference and fine-tuning for massive MoE models almost practical on consumer hardware.

Curated technical deep-dives covering everything from NVLink signal integrity to Kubernetes GPU scheduling and Huawei NPU porting.

VR eye cameras output flat images, and this project recovers the metric 3D geometry researchers actually need.

Google's edge ML runtime grows up, adds async NPU support, and finally admits PyTorch exists.

MiniMind-O packs listen-see-speak intelligence into a 0.1B-parameter model you can retrain from the first line of code on a single desktop GPU.

Because OpenAI's gym-retro froze, this fork keeps 1,000+ classic games playable as Gymnasium RL environments.

A tiny deep-learning model that guesses what a file actually contains, not just what its extension claims.

It corrals the latest subquadratic sequence-model research into hardware-efficient, training-ready PyTorch layers verified across NVIDIA, AMD, and Intel GPUs.

MLX-VLM crams speculative decoding, continuous batching, and KV cache quantization into a Mac-native toolkit for running multimodal models locally.

A Chinese-language notebook curriculum that teaches LLM inference engineering by rebuilding vLLM and SGLang internals in Python.

A PyTorch implementation of "Attention Is All You Need" that scales from 13M to multi-billion parameter models.

A collection of minimal Python implementations that trade production polish for pedagogical clarity, exposing how classic ML algorithms work under the hood.

TradeMaster bundles the full RL trading pipeline, from ingesting market data to evaluating agents, so researchers don't have to wire it all up themselves.

It gives multi-agent clusters a PyTorch-like dataflow model where sessions are tensors, agents are layers, and lineage is first-class.

Chinese open-source community Datawhale built a from-zero embodied-AI course that gets you from `print('hello')` to fine-tuning SmolVLA and Pi0.

One YAML file runs your AI workload on Kubernetes, Slurm, AWS, or a random GPU cloud you've never heard of.

To demystify reasoning LLMs by coding inference-time scaling, RL, and distillation on top of a pretrained model, step by step.