fix machete pre quant (#4295)

2025-10-06 17:17:14 +08:00 · 2025-09-28 16:11:09 +08:00
parent 7b6cb72ab2
commit aa1cc09c5b
2 changed files with 5 additions and 1 deletions
--- a/fastdeploy/model_executor/layers/linear.py
+++ b/fastdeploy/model_executor/layers/linear.py
@@ -129,6 +129,7 @@ class LinearBase(nn.Layer):
        self.with_bias = with_bias
        self.add_bias = add_bias
        self.prefix = prefix
+        self.is_quantized = fd_config.model_config.is_quantized
        # key
        if weight_key:
            self.weight_key = f"{prefix}.{weight_key}"