RichardErkhov/euclaise_-_Ferret-3B-gguf

Quantization made by Richard Erkhov.

Ferret-3B - GGUF

Model creator: https://huggingface.co/euclaise/
Original model: https://huggingface.co/euclaise/Ferret-3B/

Name	Quant method	Size
Ferret-3B.Q2_K.gguf	Q2_K	1.01GB
Ferret-3B.IQ3_XS.gguf	IQ3_XS	1.11GB
Ferret-3B.IQ3_S.gguf	IQ3_S	1.17GB
Ferret-3B.Q3_K_S.gguf	Q3_K_S	1.17GB
Ferret-3B.IQ3_M.gguf	IQ3_M	1.23GB
Ferret-3B.Q3_K.gguf	Q3_K	1.3GB
Ferret-3B.Q3_K_M.gguf	Q3_K_M	1.3GB
Ferret-3B.Q3_K_L.gguf	Q3_K_L	1.4GB
Ferret-3B.IQ4_XS.gguf	IQ4_XS	1.43GB
Ferret-3B.Q4_0.gguf	Q4_0	1.5GB
Ferret-3B.IQ4_NL.gguf	IQ4_NL	1.51GB
Ferret-3B.Q4_K_S.gguf	Q4_K_S	1.51GB
Ferret-3B.Q4_K.gguf	Q4_K	1.59GB
Ferret-3B.Q4_K_M.gguf	Q4_K_M	1.59GB
Ferret-3B.Q4_1.gguf	Q4_1	1.65GB
Ferret-3B.Q5_0.gguf	Q5_0	1.81GB
Ferret-3B.Q5_K_S.gguf	Q5_K_S	1.81GB
Ferret-3B.Q5_K.gguf	Q5_K	1.86GB
Ferret-3B.Q5_K_M.gguf	Q5_K_M	1.86GB
Ferret-3B.Q5_1.gguf	Q5_1	1.96GB
Ferret-3B.Q6_K.gguf	Q6_K	2.14GB
Ferret-3B.Q8_0.gguf	Q8_0	2.77GB

Original model description:

pipeline_tag: question-answering datasets: - euclaise/MiniCoT - euclaise/SciCoT - euclaise/symtune_mini - euclaise/mathoverflow-accepted - euirim/goodwiki

A 3B CoT model. ChatML format. Use it however you want, consider it public domain.

Additional thanks to @nicoboss for giving me access to his private supercomputer, enabling me to provide many more quants, at much higher speed, than I would otherwise be able to.