From 9fbd6fe77e0f14a60b4d4bd2c4aa3b88a6308dcc Mon Sep 17 00:00:00 2001 From: Daniel Maddern Date: Wed, 19 Aug 2026 22:57:03 +0700 Subject: [PATCH] Batch-dim on input_ids for _embed_rows ([1, seq]) --- src/h3_blackwell_runtime/qwen3vl_vision.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/src/h3_blackwell_runtime/qwen3vl_vision.py b/src/h3_blackwell_runtime/qwen3vl_vision.py index cd7e26d..1d058c6 100644 --- a/src/h3_blackwell_runtime/qwen3vl_vision.py +++ b/src/h3_blackwell_runtime/qwen3vl_vision.py @@ -473,7 +473,7 @@ def build_fl2va_presentation( token_ids.append(int(tok)) 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 # rows spliced over the pad positions.