Notux 8x7B v1

argilla 's Collections

updated 11 days ago

Notux 8x7B v1 model (DPO fine-tune of Mixtral 8x7B Instruct v0.1) and datasets used. More information at https://github.com/argilla-io/notus

argilla/notux-8x7b-v1

Text Generation • Updated Mar 4 • 41 • 165

Note Full DPO fine-tuning of `mistralai/Mixtral-8x7b-Instruct-v0.1` using `argilla/ultrafeedback-binarized-preferences-cleaned` running in a VM with 8 x H100 80GB
argilla/ultrafeedback-binarized-preferences-cleaned

Viewer • Updated Dec 11, 2023 • 60.9k • 9.24k • 127

Note Iteration on top of `argilla/ultrafeedback-binarized-preferences` but removing the TruthfulQA prompts that were introducing some data contamination as spotted by AllenAI, but keeping Argilla's approach on the data binarization Formatting: the dataset follows the same formatting as the one defined within the Alignment Handbook from HuggingFace H4
TheBloke/notux-8x7b-v1-GGUF

Text Generation • Updated Dec 29, 2023 • 313 • 5

Note GGUF-compatible quantized variants of `argilla/notux-8x7b-v1` generated by TheBloke (thanks <3)
TheBloke/notux-8x7b-v1-AWQ

Text Generation • Updated Dec 29, 2023 • 19 • 3

Note AWQ-compatible quantized variants of `argilla/notux-8x7b-v1` generated by TheBloke (thanks <3)
TheBloke/notux-8x7b-v1-GPTQ

Text Generation • Updated Dec 29, 2023 • 32 • 2

Note GPTQ-compatible quantized variants of `argilla/notux-8x7b-v1` generated by TheBloke (thanks <3)