opt-125m is missing weight for embed and lm_head when use_safetensors=True

[xin3he]

System Info

• transformers version: 4.57.3
• Platform: Linux-5.4.0-169-generic-x86_64-with-glibc2.35
• Python version: 3.12.9
• Huggingface_hub version: 0.36.0
• Safetensors version: 0.5.3
• Accelerate version: 1.11.0
• Accelerate config: not found
• DeepSpeed version: 0.17.5
• PyTorch version (accelerator?): 2.9.0+cu128 (CUDA)
• Tensorflow version (GPU?): not installed (NA)
• Flax version (CPU?/GPU?/TPU?): not installed (NA)
• Jax version: not installed
• JaxLib version: not installed
• Using distributed or parallel set-up in script?:
• Using GPU in script?:
• GPU type: NVIDIA A100-SXM4-80GB

Who can help?

@ArthurZucker @Cyrilvallez

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

I found that the lm_head is not loaded correctly for opt-125m while llama3.2 1B is fine.

import transformers
import torch
model = transformers.AutoModelForCausalLM.from_pretrained("facebook/opt-125m", torch_dtype=torch.float16, use_safetensors=True)

>>> print(model.lm_head.weight)
Parameter containing:
tensor(..., device='meta', size=(50272, 768), dtype=torch.bfloat16,
       requires_grad=True)

Expected behavior

Expecting weights are loaded correctly.

[ArthurZucker]

Hey! can you try on v5.0.0rc2 please?