GitLab

change to a flatter directory structure and group the options with the
function

update models to call rope in one place

deepseek's qwen3 distill uses a different rope scheme so support both

* cleanup

* use pooling.TypeNone

* pooling test

* qwen3 embed

* fix(llama): rope scale

* spm llama

* skip moe models

* cleanup

* use ggml_*_split activations when possible

* forward qkv

this cleans up the model interface slightly without too much impact in
other areas

while nn.Linear.Forward isn't applicable for sparse MLP, it's still
a nice container for the tensors

FromFloatSlice and FromIntSlice return an error if the shape doesn't
match the passed data or if memory can't be allocated. Since these
are inputs, the memory being allocated is system memory rather than VRAM.

In many cases, the caller can't really handle the error and panics.

Empty and Zeros directly panic if they can't allocate memory.

This makes things consistent by panicing for the first two cases,
removing a fair amount of error handling code. This is also consistent
with how Go typically handles these situations.

* feat: qwen3 dense
* feat: qwen3moe
* fix llama4 moe