Climb-Ai
Open-source research group exploring language models, efficient fine-tuning, and machine learning datasets.
Focus Areas
- Language Models: Small-to-medium parameter LLMs and domain-specific fine-tuning.
- Datasets: Cleaned and standardized training data for SFT and instruction following.
- Open Experiments: Practical tooling, evaluations, and reproducibility.
Projects & Datasets
- toki-pona-tiny-dialogues: Cleaned multi-turn conversational dataset in Toki Pona for SFT chat fine-tuning (ChatML & ShareGPT).
Contact & Community