A new small LLaVA model fine-tuned from Phi 3 Mini [I-Quants]
Vision
3B
552 Pulls Updated 3 months ago
Updated 3 months ago
3 months ago
9ec1416b1b48 · 2.1GB
model
archllama
·
parameters3.82B
·
quantizationIQ3_XXS
1.5GB
template
{{ if .System }}<|system|>
{{ .System }}<|end|>
{{ end }}{{ if .Prompt }}<|user|>
{{ .Prompt }}<|end|>
{{ end }}<|assistant|>
{{ .Response }}<|end|>
149B
projector
archclip
·
parameters303M
·
quantizationF16
608MB
params
{"num_ctx":4096,"num_keep":4,"stop":["<|user|>","<|assistant|>","<|system|>","<|end|>","<|endoftext|>"]}
155B
Readme
llava-phi3
is a LLaVA model fine-tuned from Phi 3 Mini 4k, with strong performance benchmarks on par with the original LLaVA model: