neuralmagic / sparseml

Libraries for applying sparsification recipes to neural networks with a few lines of code, enabling faster and smaller models
Apache License 2.0
2.07k stars 148 forks source link

Fix Example Uncompressed Path #2339

Closed Satrat closed 5 months ago

Satrat commented 5 months ago

In examples/llama7b_sparse_quantized/README.md we apply a staged recipe then reload the model to save a compressed version. However the path was not accounting for the nested save structure used for staged recipes.

Updating the load path fixed the issue. Asana ticket: https://app.asana.com/0/1207587658110568/1207604393141284/f