Codelama with 16k context unlocked in modelfile
222 Pulls Updated 12 months ago
Updated 12 months ago
12 months ago
be95e5b84e7f · 16GB
model
archllama
·
parameters33.7B
·
quantizationQ3_K_M
16GB
template
[INST] <<SYS>>{{ .System }}<</SYS>>
{{ .Prompt }} [/INST]
59B
params
{
"num_ctx": 16384,
"num_gqa": 8,
"rope_frequency_base": 1000000,
"stop": [
148B
Readme
13b version is q4_K_M quantization
34b version is q3_K_M quantization