daniel/h3-blackwell-runtime
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Batch-dim on input_ids for _embed_rows ([1, seq])

Browse source
This commit is contained in:
Daniel Maddern 2026-08-19 22:57:03 +07:00
parent 0903b849b2
commit 9fbd6fe77e
1 changed files with 1 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
src/h3_blackwell_runtime/qwen3vl_vision.py
View file

@ -473,7 +473,7 @@ def build_fl2va_presentation(
token_ids.append(int(tok)) token_ids.append(int(tok))
seq_len = len(token_ids) seq_len = len(token_ids)
input_ids = torch.tensor([int(t) for t in token_ids], dtype=torch.long, device=device) input_ids = torch.tensor([int(t) for t in token_ids], dtype=torch.long, device=device).unsqueeze(0) # [1, seq]
# Build the Qwen hidden-state input: scaled token embeds, with the merged vision # Build the Qwen hidden-state input: scaled token embeds, with the merged vision
# rows spliced over the pad positions. # rows spliced over the pad positions.