Add AWQ support for all models (#1714)

This commit is contained in:
Woosuk Kwon
2023-11-18 17:56:47 -08:00
committed by GitHub
parent e946260cf3
commit 8d17774f92
13 changed files with 90 additions and 17 deletions

View File

@@ -130,7 +130,8 @@ class MPTMLP(nn.Module):
bias=not config.no_bias,
linear_method=linear_method,
)
self.act = get_act_fn("gelu")
quant_config = getattr(linear_method, "quant_config", None)
self.act = get_act_fn("gelu", quant_config, intermediate_size)
self.down_proj = RowParallelLinear(
intermediate_size,
hidden_size,