Skip to content

PTQ-Level Efficiency, QAT-Level Accuracy | 训练后量化的效率,量化感知训练的精度

support for gemma3 (#468)

Heng Guo451c029b02fc
diff --git a/auto_round/script/mllm.py b/auto_round/script/mllm.py
index 3b00d74e2..3ee15474d 100644
--- a/auto_round/script/mllm.py
+++ b/auto_round/script/mllm.py
@@ -346,6 +346,9 @@ def tune(args):
elif "mllama" in model_type:
from transformers import MllamaForConditionalGeneration
cls = MllamaForConditionalGeneration
+ elif "gemma3" in model_type:
+ from transformers import Gemma3ForConditionalGeneration
+ cls = Gemma3ForConditionalGeneration
else:
cls = AutoModelForCausalLM