RichardErkhov/wandb_-_mistral-7b-zephyr-sft-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

mistral-7b-zephyr-sft - GGUF

• Model creator: https://huggingface.co/wandb/
• Original model: https://huggingface.co/wandb/mistral-7b-zephyr-sft/

Name	Quant method	Size
mistral-7b-zephyr-sft.Q2_K.gguf	Q2_K	2.53GB
mistral-7b-zephyr-sft.IQ3_XS.gguf	IQ3_XS	2.81GB
mistral-7b-zephyr-sft.IQ3_S.gguf	IQ3_S	2.96GB
mistral-7b-zephyr-sft.Q3_K_S.gguf	Q3_K_S	2.95GB
mistral-7b-zephyr-sft.IQ3_M.gguf	IQ3_M	3.06GB
mistral-7b-zephyr-sft.Q3_K.gguf	Q3_K	3.28GB
mistral-7b-zephyr-sft.Q3_K_M.gguf	Q3_K_M	3.28GB
mistral-7b-zephyr-sft.Q3_K_L.gguf	Q3_K_L	3.56GB
mistral-7b-zephyr-sft.IQ4_XS.gguf	IQ4_XS	3.67GB
mistral-7b-zephyr-sft.Q4_0.gguf	Q4_0	3.83GB
mistral-7b-zephyr-sft.IQ4_NL.gguf	IQ4_NL	3.87GB
mistral-7b-zephyr-sft.Q4_K_S.gguf	Q4_K_S	3.86GB
mistral-7b-zephyr-sft.Q4_K.gguf	Q4_K	4.07GB
mistral-7b-zephyr-sft.Q4_K_M.gguf	Q4_K_M	4.07GB
mistral-7b-zephyr-sft.Q4_1.gguf	Q4_1	4.24GB
mistral-7b-zephyr-sft.Q5_0.gguf	Q5_0	4.65GB
mistral-7b-zephyr-sft.Q5_K_S.gguf	Q5_K_S	4.65GB
mistral-7b-zephyr-sft.Q5_K.gguf	Q5_K	4.78GB
mistral-7b-zephyr-sft.Q5_K_M.gguf	Q5_K_M	4.78GB
mistral-7b-zephyr-sft.Q5_1.gguf	Q5_1	5.07GB
mistral-7b-zephyr-sft.Q6_K.gguf	Q6_K	5.53GB
mistral-7b-zephyr-sft.Q8_0.gguf	Q8_0	7.17GB

Original model description:

license: mit library_name: transformers datasets: - HuggingFaceH4/deita-10k-v0-sft base_model: mistralai/Mistral-7B-v0.1

Mistral 7B Zephyr SFT V2

The Zephyr SFT recipe applied on top of Mistral 7B (new recipe with chatML format)

Model description

• Model type: A 7.2B parameter GPT-like model fine-tuned on a mix of publicly available, synthetic datasets.
• Language(s) (NLP): Primarily English
• Finetuned from model: mistralai/Mistral-7B-v0.1

Recipe

We trained using the alignment handbook recipe and logging to W&B

Visit the W&B workspace here

Compute provided by Lambda Labs - 8xA100 80GB node