valine / training-hot-swapLinks
Pytorch script hot swap: Change code without unloading your LLM from VRAM
☆126Updated 3 months ago
Alternatives and similar repositories for training-hot-swap
Users that are interested in training-hot-swap are comparing it to the libraries listed below
Sorting:
- ☆392Updated last week
- ☆249Updated last year
- GRPO training code which scales to 32xH100s for long horizon terminal/coding tasks. Base agent is now the top Qwen3 agent on Stanford's T…☆218Updated last week
- Heirarchical Navigable Small Worlds☆101Updated this week
- Pivotal Token Search☆119Updated 3 weeks ago
- A playground to make it easy to try crazy things☆33Updated last month
- ☆221Updated 5 months ago
- ☆47Updated 4 months ago
- ☆196Updated 3 months ago
- An implementation of bucketMul LLM inference☆221Updated last year
- NanoGPT-speedrunning for the poor T4 enjoyers☆69Updated 3 months ago
- Bayesian Optimization as a Coverage Tool for Evaluating LLMs. Accurate evaluation (benchmarking) that's 10 times faster with just a few l…☆286Updated last week
- ☆186Updated this week
- DiscoGrad - automatically differentiate across conditional branches in C++ programs☆204Updated 10 months ago
- Mistral7B playing DOOM☆133Updated last year
- Visualizing the internal board state of a GPT trained on chess PGN strings, and performing interventions on its internal board state and …☆209Updated 8 months ago
- A pure NumPy implementation of Mamba.☆223Updated last year
- Ever used asyncio and wished you hadn't? A tiny (~300 lines) event loop for Python.☆502Updated last week
- High-Performance Implementation of OpenAI's TikToken.☆445Updated last month
- look how they massacred my boy☆63Updated 9 months ago
- explore token trajectory trees on instruct and base models☆134Updated 2 months ago
- lossily compress representation vectors using product quantization☆58Updated 3 months ago
- Official codebase for the paper "Beyond A* Better Planning with Transformers via Search Dynamics Bootstrapping".☆372Updated last year
- rl from zero pretrain, can it be done? yes.☆193Updated this week
- SIMD quantization kernels☆78Updated last week
- a curated list of data for reasoning ai☆137Updated last year
- A tiny autograd engine with a Jax-like API☆74Updated last month
- Alice in Wonderland code base for experiments and raw experiments data☆131Updated last week
- Live-bending a foundation model’s output at neural network level.☆268Updated 4 months ago
- Tensor library & inference framework for machine learning☆106Updated 3 weeks ago