Remove ScaledActivation for AWQ (#10057)

Signed-off-by: mgoin <michael@neuralmagic.com>
2024-11-06 09:27:06 -05:00
parent 406d4cc480
commit 399c798608
34 changed files with 19 additions and 124 deletions
--- a/vllm/model_executor/models/persimmon.py
+++ b/vllm/model_executor/models/persimmon.py
@@ -60,7 +60,7 @@ class PersimmonMLP(nn.Module):
        self.dense_4h_to_h = RowParallelLinear(config.intermediate_size,
                                               config.hidden_size,
                                               quant_config=quant_config)
-        self.act = get_act_fn(config.hidden_act, quant_config)
+        self.act = get_act_fn(config.hidden_act)

    def forward(self, hidden_states) -> torch.Tensor:
        hidden_states, _ = self.dense_h_to_4h(hidden_states)