huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-09-09 11:24:53 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix inputs_embeds

Browse Source
This commit is contained in:
Mohit Sharma 2025-04-22 02:03:34 +05:30 committed by GitHub
parent f34b06ca3b
commit 26212b9f35
No known key found for this signature in database
GPG Key ID: B5690EEEBB952194
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
server/text_generation_server/models/vlm_causal_lm.py
View File

@ -1188,7 +1188,7 @@ class VlmCausalLM(FlashCausalLM):
# Copy inputs to the static inputs of the cuda graph # Copy inputs to the static inputs of the cuda graph
# Static inputs are potentially padded # Static inputs are potentially padded
cuda_graph["input_ids"][: input_ids.shape[0]] = input_ids cuda_graph["input_ids"][: input_ids.shape[0]] = input_ids
cuda_graph["input_embeds"][: inputs_embeds.shape[0]] = inputs_embeds cuda_graph["inputs_embeds"][: inputs_embeds.shape[0]] = inputs_embeds
cuda_graph["position_ids"][: position_ids.shape[0]] = position_ids cuda_graph["position_ids"][: position_ids.shape[0]] = position_ids
if ATTENTION == "flashinfer": if ATTENTION == "flashinfer":
block_tables = block_tables_to_ragged( block_tables = block_tables_to_ragged(