casinca/LLM-quest
Verbose implementations of LLMs architectures, techniques and research papers from scratch. Qwen..., RLHF, Multimodal, Hyperconnections...
Repositories
casinca/rlhf-book
Textbook on reinforcement learning from human feedback
casinca/GRPO-classic-RL
Open-source implementation/adaptation of DeepSeek GRPO applied to Reinforcement Learning control problems. Example on LunarLander-V3.
casinca/transformers
🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.
casinca/trl
Train transformer language models with reinforcement learning.
casinca/reasoning-from-scratch
Implement a reasoning LLM in PyTorch from scratch, step by step
casinca/number-token-loss
PyPI package for number token loss
casinca/LLMs-from-scratch
Implement a ChatGPT-like LLM in PyTorch from scratch, step by step
casinca/ffn-from-scratch
Feed Forward Neural Network (FFN) from Scratch. In pure Python and NumPy. One derivative at a time.
casinca/Moonlight
Muon is Scalable for LLM Training
casinca/aaamlp-enhanced-pdf
Approaching (Almost) Any Machine Learning Problem (AAAMLP) PDF from @abhishekkrthakur with outline, cover, notes
casinca/bdo-enhancing-model
Probabilistic Modeling of Black Desert Online's Enhancement System. A proof of concept predicting outcomes to derive optimal profitability strategies.
casinca/vllm
A high-throughput and memory-efficient inference and serving engine for LLMs
casinca/mlp
The Multilayer Perceptron Language Model