Bfloat16 version of the Microsoft Phi4 14b model
10 Pulls Updated 2 months ago
Updated 2 months ago
2 months ago
bbf83db96d06 · 29GB
model
archphi3
·
parameters14.7B
·
quantization(!unknown_file_type 32!)
29GB
params
{
"stop": [
"<|im_start|>",
"<|im_end|>",
"<|im_sep|>"
]
}
82B
template
{{- range $i, $_ := .Messages }}
{{- $last := eq (len (slice $.Messages $i)) 1 -}}
<|im_start|>{{ .R
275B
license
Microsoft.
Copyright (c) Microsoft Corporation.
MIT License
Permission is hereby granted, free of
1.1kB
Readme
Bfloat16 version of the Microsoft Phi4 14b model. Requires Ollama >v0.5.5 and BF16 supporting GPU (native or BF16 to FP32 on the fly llama.cpp conversion)
Full model info at https://ollama.com/library/phi4 and https://huggingface.co/microsoft/phi-4/