Quantization made by Richard Erkhov.
This llama model was trained 2x faster with
Unsloth and Huggingface's TRL library.
This model was a very first trial of
LànhGPT V2.0 which shall purposely be built focusing on multi-domains, multi-tasks, multi-linguals, specially for Vietnam.
This series of trials is VERY good at json generation too.
