Modelpublic

Joseph717171/DeepHermes-3-Llama-3.1-8B-Preview-OQ8_0-F32.EF32.IQ4_K-Q8_0-GGUF

sourceHugging Faceupdated 1y agoView on Hugging Face

1likes737downloads

Model Card

Custom GGUF quants of NousResearch/DeepHermes-3-Llama-3-8B-Preview, where the Output Tensors are kept at F32 or quantized to Q8_0, while the Embeddings are kept at F32. Enjoy! 🧠🔥🚀