radi-cho
/

llama3-8b_a6000_nfl_b3g128_bf16_wikitext

Text Generation

text-generation-inference

Model card Files Files and versions

This is a flute package of LLaMA-3-8B with learned NormalFloat (nfl) quantization.

Bits: 3
Group: 128
Precision (scales): bfloat16
Trainset: 128 samples, WikiText

Downloads last month: -

Safetensors

Model size

2B params

Tensor type

F32

·

BF16

·

I16

·