System Info

• transformers version: 5.0.0.dev0
• Platform: Linux-5.15.0-1048-aws-x86_64-with-glibc2.31
• Python version: 3.12.9
• Huggingface_hub version: 1.1.5
• Safetensors version: 0.6.2
• Accelerate version: 1.11.0
• Accelerate config: not found
• DeepSpeed version: 0.18.2
• PyTorch version (accelerator?): 2.9.0+cu128 (CUDA)
• Using distributed or parallel set-up in script?:
• Using GPU in script?:
• GPU type: NVIDIA H100 80GB HBM3

Who can help?

No response

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

As part of the training_ci (#42597), we noticed that BLT model can't overfit a single sentence (and thus generated the same sentence back). There must be a bug somewhere. Raising issue to flag it

pytest tests/models/blt/test_modeling_blt.py::BltModelTest::test_training_overfit -s

Expected behavior

matching generation and loss + grad_norm reduction should be at least 90%