eas/
neuralbeagle14:latest

200 1 year ago

quantized versions of mlabonne/NeuralBeagle14-7B

1 year ago

f07bf0818961 · 4.4GB ·

model
llama
·
7.24B
·
Q4_K_M
template
<|im_start|>system {{ .System }}<|im_end|> <|im_start|>user {{ .Prompt }}<|im_end|> <|im_start|>assi
params
{ "num_ctx": 4096, "stop": [ "<|im_end|>", "|im_end|>", "<|im_start|

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.