[CORE] Quantized lm-head Framework (#4442)
Co-authored-by: Robert Shaw <rshaw@neuralmagic.com> Co-authored-by: ZX <zx@lbx.dev>
This commit is contained in:
co-authored by
Robert Shaw
ZX
parent
7c008c51a9
commit
ee93f4f92a
@@ -347,8 +347,8 @@ class GemmaForCausalLM(nn.Module, SupportsLoRA):
|
||||
|
||||
def compute_logits(self, hidden_states: torch.Tensor,
|
||||
sampling_metadata: SamplingMetadata) -> torch.Tensor:
|
||||
logits = self.logits_processor(self.model.embed_tokens.weight,
|
||||
hidden_states, sampling_metadata)
|
||||
logits = self.logits_processor(self.model.embed_tokens, hidden_states,
|
||||
sampling_metadata)
|
||||
return logits
|
||||
|
||||
def sample(
|
||||
|
||||
Reference in New Issue
Block a user