mistrallite:7b-v0.1-q3_K_S

32.9K 1 year ago

MistralLite is a fine-tuned model based on Mistral with enhanced capabilities of processing long contexts.

7b

1 year ago

ebf457323d7d · 3.2GB

model
llama
·
7.24B
·
Q3_K_S
template
<|prompter|>{{ .System }} {{ .Prompt }}</s><|assistant|>
params
{ "num_ctx": 32768, "stop": [ "<|prompter|>", "<|assistant|>", "</s>

Readme

Mistrallite is a fine-tuned model based on Mistral, with enhanced capabilities of processing long context (up to 32K tokens). This model is able to perform significantly better on several long context retrieve and answering tasks.

Reference

Hugging Face