Issue Station
Sign In
Tools hereToolbox

transformers - ✅(Solved) Fix auto_mappings.py references removed Sam3LiteText configs, breaking CI [2 pull requests, 4 comments, 2 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
huggingface/transformers#45600•Fetched 2026-04-24 05:51:22
View on GitHub
Comments
4
Participants
2
Timeline
9
Reactions
1
Author
artem-spectorartem-spector
Participants
artem-spectorartem-spectorzucchini-nlpzucchini-nlp
Timeline (top)
commented ×4closed ×1cross-referenced ×1labeled ×1

Error Message

Exception: Sam3LiteTextVisionConfig appears in CONFIG_MAPPING_NAMES but is not defined in the library.

Root Cause

Root cause: PR #45535 removed Sam3LiteTextViTConfig and Sam3LiteTextVisionConfig from the modeling file but left these two stale entries in auto_mappings.py:
("sam3_vision_model", "Sam3LiteTextVisionConfig"),
("sam3_vit_model", "Sam3LiteTextViTConfig"),

Fix Action

Fixed

PR fix notes

PR #45597: Add Granite 4.1 Vision (granite4_vision)

Description (problem / solution / changelog)

What does this PR do?

Adds built-in support for Granite 4.1 Vision (granite4_vision), IBM's multimodal vision-language model for enterprise document understanding.

Architecture highlights

  • Vision encoder: SigLIP2 (google/siglip2-so400m-patch16-384), tiled 384×384 patches
  • Window Q-Former projector: 4×4 patch windows compressed to 2×2 query tokens via cross-attention (downsample_rate="4/8")
  • DeepStack feature injection: 8 vision-to-LLM injection points across two mechanisms:
    • LayerDeepstack: features from 4 vision encoder depths injected at 4 LLM layers (reversed order — deepest vision → earliest LLM)
    • SpatialDeepstack: deepest features split into 4 spatial offset groups (TL/TR/BL/BR), injected at 4 later LLM layers
  • Language model: GraniteForCausalLM (3.5B) with a rank-256 LoRA adapter (same-repo, LM-only)

Files added

FilePurpose
modular_granite4_vision.pySource of truth — inherits from LLaVA-Next, overrides novel components
configuration_granite4_vision.pyConfig (generated)
modeling_granite4_vision.pyModel (generated)
processing_granite4_vision.pyUnified processor (generated)
image_processing_granite4_vision.pyTorchvision-based image processor
image_processing_pil_granite4_vision.pyPIL/NumPy image processor
tests/models/granite4_vision/Modeling, image processing, and processor tests
docs/source/en/model_doc/granite4_vision.mdModel documentation

Auto-registration

  • Config: auto-generated via configuration_granite4_vision.py model_type
  • Modeling: MODEL_MAPPING_NAMES + MODEL_FOR_IMAGE_TEXT_TO_TEXT_MAPPING_NAMES
  • Processing + image processing: registered in respective auto files

Tests

  • Unit tests pass locally (pytest tests/models/granite4_vision/ -x -q)
  • @slow integration tests load real checkpoint and assert outputs within tolerance
  • make style and make check-repo pass (3 remaining failures are pre-existing upstream issues: mlinter version mismatch and Sam3Lite incomplete model)

Before submitting

  • This PR is not a duplicate
  • I have read the contributor guidelines
  • The documentation reflects the changes
  • The tests pass

Related

Changed files

  • docs/source/en/_toctree.yml (modified, +2/-0)
  • docs/source/en/model_doc/granite4_vision.md (added, +206/-0)
  • src/transformers/conversion_mapping.py (modified, +6/-0)
  • src/transformers/models/__init__.py (modified, +1/-0)
  • src/transformers/models/auto/auto_mappings.py (modified, +2/-0)
  • src/transformers/models/auto/modeling_auto.py (modified, +2/-0)
  • src/transformers/models/auto/processing_auto.py (modified, +1/-0)
  • src/transformers/models/granite4_vision/__init__.py (added, +30/-0)
  • src/transformers/models/granite4_vision/configuration_granite4_vision.py (added, +116/-0)
  • src/transformers/models/granite4_vision/downsampling_granite4_vision.py (added, +155/-0)
  • src/transformers/models/granite4_vision/image_processing_granite4_vision.py (added, +244/-0)
  • src/transformers/models/granite4_vision/image_processing_pil_granite4_vision.py (added, +240/-0)
  • src/transformers/models/granite4_vision/modeling_granite4_vision.py (added, +882/-0)
  • src/transformers/models/granite4_vision/modular_granite4_vision.py (added, +734/-0)
  • src/transformers/models/granite4_vision/processing_granite4_vision.py (added, +238/-0)
  • src/transformers/utils/auto_docstring.py (modified, +4/-0)
  • tests/models/granite4_vision/__init__.py (added, +0/-0)
  • tests/models/granite4_vision/test_image_processing_granite4_vision.py (added, +253/-0)
  • tests/models/granite4_vision/test_modeling_granite4_vision.py (added, +272/-0)
  • tests/models/granite4_vision/test_processing_granite4_vision.py (added, +122/-0)
  • utils/check_repo.py (modified, +4/-0)
RAW_BUFFERClick to expand / collapse

System Info

  • transformers version: 5.6.0.dev0
  • Platform: Linux-5.14.0-570.12.1.el9_6.x86_64-x86_64-with-glibc2.34
  • Python version: 3.11.15
  • Huggingface_hub version: 1.10.2
  • Safetensors version: 0.7.0
  • Accelerate version: 1.13.0
  • Accelerate config: not found
  • DeepSpeed version: not installed
  • PyTorch version (accelerator?): 2.11.0+cu130 (CUDA)
  • Using distributed or parallel set-up in script?: <fill in>
  • Using GPU in script?: <fill in>
  • GPU type: NVIDIA H100 80GB HBM3

Who can help?

@yonigozlan

Information

  • The official example scripts
  • My own modified scripts

Tasks

  • An officially supported task in the examples folder (such as GLUE/SQuAD, ...)
  • My own task or dataset (give details below)

Reproduction

On current branch:

python utils/check_repo.py Exception: Sam3LiteTextVisionConfig appears in CONFIG_MAPPING_NAMES but is not defined in the library.
Sam3LiteTextViTConfig appears in CONFIG_MAPPING_NAMES but is not defined in the library.

pytest tests/models/sam3_lite_text/ 357 failures: AttributeError: module transformers has no attribute Sam3LiteTextViTConfig

Root cause: PR #45535 removed Sam3LiteTextViTConfig and Sam3LiteTextVisionConfig from the modeling file but left these two stale entries in auto_mappings.py:
("sam3_vision_model", "Sam3LiteTextVisionConfig"),
("sam3_vit_model", "Sam3LiteTextViTConfig"),

Expected behavior

check_repo.py and sam3_lite_text tests should pass on main. The stale entries should be removed from auto_mappings.py.

extent analysis

TL;DR

Remove the stale entries for Sam3LiteTextVisionConfig and Sam3LiteTextViTConfig from auto_mappings.py to fix the issue.

Guidance

  • Identify the auto_mappings.py file and locate the stale entries for Sam3LiteTextVisionConfig and Sam3LiteTextViTConfig.
  • Remove the lines containing these stale entries to ensure consistency with the changes made in PR #45535.
  • Run python utils/check_repo.py and pytest tests/models/sam3_lite_text/ again to verify that the issue is resolved.
  • Review the code changes to ensure that no other stale entries or references to the removed configurations exist.

Notes

This fix assumes that the removal of Sam3LiteTextVisionConfig and Sam3LiteTextViTConfig from the modeling file was intentional and that the stale entries in auto_mappings.py are the only remaining references to these configurations.

Recommendation

Apply workaround: Remove the stale entries from auto_mappings.py to resolve the issue, as the configurations have been removed from the modeling file.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…

FAQ

Expected behavior

check_repo.py and sam3_lite_text tests should pass on main. The stale entries should be removed from auto_mappings.py.

#file not found#serialization error#model compatibility#GPU setup#container setup

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING