togethercomputer

/

m2-bert-80M-8k-retrieval

like 33

Follow

Together 465

Hi, thanks for your interest!

This model is a BERT-style model, and it is not Transformer-based. The architecture is Monarch Mixer: https://arxiv.org/abs/2310.12109

GGUF does not currently support BERT-style models or Monarch Mixer, but we’d be happy to add it as a format if it does in the future. You’d have better luck filing an issue with GGML.

See this for their currently supported models: https://github.com/ggerganov/ggml/blob/master/docs/gguf.md#required

Thank you so much for your support.