Removed obsolete attribute setting for AQLM quantization. (#29034)

removed redundant field

Removed obsolete attribute setting for AQLM quantization. (#29034)
removed redundant field
b0a7f44f · Andrei Panferov · GitHub · 4156f517 · b0a7f44f
Unverified Commit b0a7f44f authored Feb 15, 2024 by Andrei Panferov Committed by GitHub Feb 15, 2024
Show whitespace changes
Inline Side-by-side

Showing with 0 additions and 1 deletion

src/transformers/quantizers/quantizer_aqlm.py src/transformers/quantizers/quantizer_aqlm.py +0 -1

No files found.
--- a/src/transformers/quantizers/quantizer_aqlm.py
+++ b/src/transformers/quantizers/quantizer_aqlm.py
@@ -77,7 +77,6 @@ class AqlmHfQuantizer(HfQuantizer):
        model.config.quantization_config = self.quantization_config

    def _process_model_after_weight_loading(self, model: "PreTrainedModel", **kwargs):
-        model._is_quantized_training_enabled = False
        return model

    @property