Noeda / rllama

Rust+OpenCL+AVX2 implementation of LLaMA inference code

Noeda/rllama Issues

error when building server
Updated 9 months ago
Have plan to support Q4_0 ?
Updated 10 months ago
docker failed
Updated a year ago1
will this support llama2?
Closed a year ago1
TensorError(TensorBuilderEmpty) when running program
Updated a year ago10
Can you load another llama base model from hugging face
Updated a year ago1
load models with mmap()
Updated a year ago
Consider using ArrayFire code that is portable across CUDA, OpenCL and CPU devices
Updated a year ago
Heads up. Doesn't build on Arch Linux due to #![feature(stdsimd)]
Closed a year ago1
Use portable simd as a backup
Updated a year ago5
Add Support for other Architectures
Updated a year ago1
error[E0554]: `#![feature]` may not be used on the stable release channel
Closed a year ago1