chenchenghao/sglang
1
0
Fork 0
Code Issues Pull Requests Actions 12 Packages Projects Releases Wiki Activity

[MM][style] rename inputs_embeds to input_embeds for consistency (#14240)

Browse Source
This commit is contained in:
Byron Hsu
2025-12-01 11:36:51 -08:00
committed by GitHub
parent d9dca28247
commit edbeaf3b88
1 changed files with 15 additions and 15 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
python/sglang/srt/managers/mm_utils.py
View File

@@ -598,20 +598,20 @@ def embed_mm_inputs(
# filled with the hash values of the multimodal for the prefix matching in the radix attention.
# There values are useless because their embeddings will be replaced by vision embeddings anyway.
input_ids.clamp_(min=0, max=vocab_size - 1)
inputs_embeds = input_embedding(input_ids)
input_embeds = input_embedding(input_ids)
# deepstack embedding
if use_deepstack:
num_deepstack_embeddings = len(multimodal_model.deepstack_visual_indexes)
deepstack_embedding_shape = inputs_embeds.shape[:-1] + (
inputs_embeds.shape[-1] * num_deepstack_embeddings,
deepstack_embedding_shape = input_embeds.shape[:-1] + (
input_embeds.shape[-1] * num_deepstack_embeddings,
)
# a zero-filled embedding, with the same length of inputs_embeds, but different hidden_size
# a zero-filled embedding, with the same length of input_embeds, but different hidden_size
input_deepstack_embeds = torch.zeros(
deepstack_embedding_shape,
device=inputs_embeds.device,
dtype=inputs_embeds.dtype,
device=input_embeds.device,
dtype=input_embeds.dtype,
)
other_info["input_deepstack_embeds"] = input_deepstack_embeds
@@ -624,13 +624,13 @@ def embed_mm_inputs(
continue
# in-place update
indices = torch.where(mask.squeeze(dim=-1))[0]
inputs_embeds[indices] = embedding.to(inputs_embeds.device, inputs_embeds.dtype)
input_embeds[indices] = embedding.to(input_embeds.device, input_embeds.dtype)
if use_deepstack.get(modality, None):
input_deepstack_embeds[indices] = deepstack_embeddings[i].to(
inputs_embeds.device, inputs_embeds.dtype
input_embeds.device, input_embeds.dtype
)
return inputs_embeds, other_info
return input_embeds, other_info
def general_mm_embed_routine(
@@ -692,7 +692,7 @@ def general_mm_embed_routine(
for i, seq_len in enumerate(forward_batch.extend_seq_lens_cpu)
if forward_batch.mm_inputs[i] is not None
]
inputs_embeds, other_info = embed_mm_inputs(
input_embeds, other_info = embed_mm_inputs(
mm_inputs_list=mm_inputs_list,
extend_prefix_lens=extend_prefix_lens,
extend_seq_lens=extend_seq_lens,
@@ -712,18 +712,18 @@ def general_mm_embed_routine(
# just being defensive here
forward_batch.mm_inputs = None
else:
inputs_embeds = embed_tokens(input_ids)
input_embeds = embed_tokens(input_ids)
# Copy to pre-allocated buffer if available (for CUDA graph address stability)
if forward_batch.input_embeds is not None:
forward_batch.input_embeds.copy_(inputs_embeds)
inputs_embeds = forward_batch.input_embeds
forward_batch.input_embeds.copy_(input_embeds)
input_embeds = forward_batch.input_embeds
else:
inputs_embeds = None
input_embeds = None
hidden_states = language_model(
input_ids=None,
forward_batch=forward_batch,
input_embeds=inputs_embeds,
input_embeds=input_embeds,
**kwargs,
)
return hidden_states