fix compatible with auto-gptq

Signed-off-by: jiqing-feng <jiqing.feng@intel.com>
huggingface · Dec 2, 2024 · 32d434f · 32d434f
1 parent 4709869
commit 32d434f
Showing 1 changed file with 2 additions and 1 deletion.
diff --git a/optimum/gptq/quantizer.py b/optimum/gptq/quantizer.py
@@ -564,9 +564,10 @@ def tmp(_, input, output):
                     h.remove()
                 for name in subset_name_list:
                     logger.info(f"Quantizing {name} in block {i + 1}/{len(blocks)}...")
-                    scale, zero, g_idx, _, _, _ = gptq[name].fasterquant(
+                    quant_outputs = gptq[name].fasterquant(
                         percdamp=self.damp_percent, group_size=self.group_size, actorder=self.desc_act
                     )
+                    scale, zero, g_idx = quant_outputs[0], quant_outputs[1], quant_outputs[2]
                     quantizers[f"{self.block_name_to_quantize}.{i}.{name}"] = (
                         gptq[name].quantizer,
                         scale,