zhangyiss/mlx-examples
1
0
Fork 0
mirror of https://github.com/ml-explore/mlx-examples.git synced 2025-10-24 06:28:07 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Quantize embedding / Update quantize API (#680)

Browse Source 
* more async eval

* quantize embedding / update quantize api

* more updates for quantize

* update for quantize embeddings

* update sd quant API

* update sdxl quants

* error for datasets < batch_size

* async

* fix config loading

* fix quant

* fix tests

* fix req

* remove lm head if tie weights is true

* fix test
This commit is contained in:
Awni Hannun
2024-04-18 18:16:10 -07:00
committed by GitHub
parent f5f189e48a
commit 2146bcd7ee
28 changed files with 108 additions and 190 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
llms/mixtral/mixtral.py
View File

@@ -217,11 +217,7 @@ def load_model(folder: str):
weights = tree_unflatten(list(weights.items()))
model = Mixtral(model_args)
if quantization is not None:
# TODO: Quantize gate matrices when < 32 tiles supported
quantization["linear_class_predicate"] = (
lambda m: isinstance(m, nn.Linear) and m.weight.shape[0] != 8
)
nn.QuantizedLinear.quantize_module(model, **quantization)
nn.quantize(model, **quantization)
model.update(weights)
return model, tokenizer