mannix/llama3-12b:q5_k

mannix/

llama3-12b:q5_k_s

375 Downloads Updated 1 year ago

Meta-Llama-3-12B-Instruct is a depth upscaling merge of llama3-8b from M. Labonne

Updated 1 year ago

1 year ago

885781ecc54c · 8.0GB ·

archllama

parameters11.5B

quantizationQ5_K_S

8.0GB

{ "num_keep": 24, "stop": [ "<|start_header_id|>", "<|end_header_id|>",

110B

{{ if .System }}<|start_header_id|>system<|end_header_id|> {{ .System }}<|eot_id|>{{ end }}{{ if .Pr

260B

Meta-Llama-3-12B-Instruct is a depth upscaling merge of llama3-8b from M. Labonne

Quantized with imatrix using groups_merged.txt