quantized versions of mlabonne/NeuralBeagle14-7B
132 Pulls Updated 9 months ago
Updated 9 months ago
9 months ago
3107cf5b688f · 5.9GB
model
archllama
·
parameters7.24B
·
quantizationQ6_K
5.9GB
template
<|im_start|>system
{{ .System }}<|im_end|>
<|im_start|>user
{{ .Prompt }}<|im_end|>
<|im_start|>assi
105B
params
{"num_ctx":4096,"stop":["\u003c|im_end|\u003e","|im_end|\u003e","\u003c|im_start|\u003e","\u003c|im_
116B
Readme
q4_K_M, q6_K and q8_0 quantized versions of mlabonne/NeuralBeagle14-7B
Supports up to 8K context. Modelfile is configured for 4K.