Phi-3 maximum recursion depth exceeded when execute AutoModelForCausalLM.from_pretrained

[JaimeFon]

System Info

Copy-and-paste the text below in your GitHub issue and FILL OUT the two last points.

• transformers version: 4.44.2
• Platform: Windows-10-10.0.22631-SP0
• Python version: 3.11.3
• Huggingface_hub version: 0.24.6
• Safetensors version: 0.4.5
• Accelerate version: not installed
• Accelerate config: not found
• PyTorch version (GPU?): 2.4.1+cpu (False)
• 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?:

Who can help?

No response

Information

• [X] 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 ran this code several times successfully but I got the error and from that point I keep getting the error, this is the simple code I am running

import torch from transformers import AutoModelForCausalLM, AutoTokenizer, pipeline

torch.random.manual_seed(0)

model = AutoModelForCausalLM.from_pretrained( "microsoft/Phi-3.5-mini-instruct", device_map="cuda", torch_dtype="auto", trust_remote_code=True, )

Expected behavior

I expected to initialize the tokenizer so I can use it later in the code.

[LysandreJik]

cc @Rocketknight1

[Rocketknight1]

Hi @JaimeFon, I've tried running this several times and can't reproduce the error. Can you paste me the entire error message?