BitNet

BitNet copied to clipboard

not generate /build/bin/llama-quantize file when compile

2

run : python setup_env.py -md /home/disk1/Llama3-8B-1.58-100B-tokens -q i2_s found: FileNotFoundError: [Errno 2] No such file or directory: './build/bin/llama-quantize'

whk6688

can we generate the `llama-server` too? is not listed after the build \/ 

massudy

Trying to setup Llama3-8B-1.58-100B-tokens with i2_s

1

I manually downloaded the model and set the model with the command "python setup_env.py -md .\models\Llama3-8B-1.58-100B-tokens -q i2_s" in Windows 11 OS. The result shows: "ERROR:root:Error occurred while running command:...

yujiapingff

Fix division-by-zero

1

When quantizing the (input) activations to the bit-linear layer, `NaN`s may occur due to division by zero. This is a consequence of the formula in the original paper: $Quant(x) =...

jay-tux

Basically same as #84 but for TL1 codegen.

JCGoran

I have just discovered bitnet. Before this, I was using models in gguf format. But they were below my expectations. Now I want to run the models I trained on...

circle-games

Hi BitNet team, Thanks for releasing this great work! I'm a phd student at UC Irvine. Our research is primarily focusing on LLM efficient deployment on mobile devices. I wish...

zjc664656505