Conversational model based on Llama 2 that performs competitively on various benchmarks.
7b
13b
81.2K Pulls Updated 13 months ago
Updated 13 months ago
13 months ago
32c9ce4f8e94 · 4.1GB
model
archllama
·
parameters6.74B
·
quantizationQ4_K_M
4.1GB
template
{{ .System }}
User: {{ .Prompt }}
Assistant:
45B
params
{
"num_ctx": 4096,
"stop": [
"User:",
"Assistant:"
]
}
46B
Readme
Xwin-LM is a model based on Llama 2 using various techniques to improve its quality.