GenVRadmin

Finetuned for multi-turn chat based use-cases, based on the CorticalStack gemma-7b-ultrachat-sft model with an MIT license.

Finetuned from HuggingFaceH4/zephyr-7b-gemma-v0.1 or Google/gemma, this model is trained on 9 Indian languages (Hindi, Tamil, Punjabi, Bengali, Gujarati, Oriya, Telugu, Kannada, Malayalam) plus English. There are two models: one finetuned on Google's Gemma and one finetuned on Zephyr's Gemma base. Repo for the other model: GenVRadmin/AryaBhatta-GemmaOrca-2-Merged.

Llama 3 variant for 22 Indian languages including Tamil.

Finetuned on 9 Indian languages including Hindi, Tamil, Punjabi, Bengali, Gujarati, Oriya, Telugu, Kannada, Malayalam, and English. The model is part of two series, AryaBhatta-1 and AryaBhatta-2, and is finetuned from HuggingFaceH4/zephyr-7b-gemma-v0.1 or Google/gemma. There are two models: one finetuned on Google's Gemma and one finetuned on Zephyr's Gemma base. Repo for the Google model: GenVRadmin/AryaBhatta-GemmaOrca-Merged.