[WIP] Resize/pad VL decode works

[hengtaoguo]

Description

Dynamic resize and pad to (32, 32, 32) grid -> (32s, 512pix, 512pix)

Tests

# Decode
python -m maxtext.inference.decode maxtext/configs/base.yml tokenizer_type=huggingface model_name=qwen3-vl-2b tokenizer_path=Qwen/Qwen3-VL-2B-Instruct load_parameters_path=gs://hengtaoguo-maxtext-logs/checkpoints/qwen3-vl-2b/unscanned/2026-07-01-11-55/0/items per_device_batch_size=1 run_name=ht_test scan_layers=false use_multimodal=true prompt=\'What\ is\ the\ classification\ of\ the\ single\ exhibit\ in\ this\ video\?\' video_path=\'tests/assets/test_video.mp4\' max_prefill_predict_length=1240 max_target_length=1280 ici_tensor_parallelism=4 override_model_config=true video_max_grid_t=32 video_max_grid_h=32 video_max_grid_w=32 attention=\'dot_product\' hf_access_token=xxx

# Result
DEBUG TRACE: inputs.shape = (1, 32768, 16, 64), cos_emb.shape = (32768, 64)

Input `<|im_start|>user
<|vision_start|><|video_pad|><|vision_end|>What is the classification ofthe single exhibit in this video?<|im_end|>
<|im_start|>assistant
` -> `The exhibit in the video is a **dinosaur**. More specifically, it is a **model of a multi-headed dinosaur**, which is a type of **dinosaur exhibit**. The exhibit is likely`

Checklist

Before submitting this PR, please make sure (put X in square brackets):

• I have performed a self-review of my code. For an optional AI review, add the gemini-review label.
• I have necessary comments in my code, particularly in hard-to-understand areas.
• I have run end-to-end tests tests and provided workload links above if applicable.
• I have made or will make corresponding changes to the doc if needed, including adding new documentation pages to the relevant Table of Contents (toctree directive) as explained in our documentation.

[codecov]

Codecov Report

❌ Patch coverage is 3.70370% with 78 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/maxtext/layers/embeddings.py	3.12%	31 Missing ⚠️
src/maxtext/multimodal/processor_qwen3_omni.py	0.00%	19 Missing ⚠️
src/maxtext/models/qwen3.py	0.00%	17 Missing ⚠️
src/maxtext/multimodal/processor.py	14.28%	6 Missing ⚠️
src/maxtext/layers/encoders.py	25.00%	3 Missing ⚠️
src/maxtext/models/models.py	0.00%	2 Missing ⚠️

📢 Thoughts on this report? Let us know!