Large Language Model Training & Inference
13 repos
Fine-tuning, instruction-following, and deployment infrastructure for large language models, particularly in the 7B–70B parameter range. The cluster centers on evolutionary instruction tuning methods and model optimization techniques that improve reasoning and task adherence, with a focus on making these models practical for inference at scale. Key repos include WizardMath and WizardLM variants, which demonstrate instruction-tuning approaches, alongside supporting infrastructure for text generation and transformer-based model deployment.
Python · 1
transformers ·1,116
pytorch ·1,116
text-generation ·1,116
text-generation-inference ·1,116
endpoints_compatible ·944
llama ·944
en ·172
mistral ·172
