biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[V1] Refactor model executable interface for all text-only language models (#10374)

Browse Source 
Signed-off-by: Roger Wang <ywang@roblox.com>
This commit is contained in:
Roger Wang
2024-11-16 21:18:46 -08:00
committed by GitHub
parent 4fd9375028
commit 643ecf7b11
43 changed files with 483 additions and 90 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
vllm/model_executor/models/persimmon.py
View File

@@ -235,6 +235,9 @@ class PersimmonModel(nn.Module):
make_empty_intermediate_tensors_factory(["hidden_states"],
config.hidden_size))
def get_input_embeddings(self, input_ids: torch.Tensor) -> torch.Tensor:
return self.embed_tokens(input_ids)
def forward(
self,
input_ids: torch.Tensor,
@@ -248,7 +251,7 @@ class PersimmonModel(nn.Module):
if inputs_embeds is not None:
hidden_states = inputs_embeds
else:
hidden_states = self.embed_tokens(input_ids)
hidden_states = self.get_input_embeddings(input_ids)
else:
assert intermediate_tensors is not None
hidden_states = intermediate_tensors["hidden_states"]
@@ -282,6 +285,9 @@ class PersimmonForCausalLM(nn.Module, SupportsPP):
self.make_empty_intermediate_tensors = (
self.model.make_empty_intermediate_tensors)
def get_input_embeddings(self, input_ids: torch.Tensor) -> torch.Tensor:
return self.model.get_input_embeddings(input_ids)
def forward(
self,
input_ids: torch.Tensor,