HuggingFaceTB/SmolVLM-Instruct · Exception on Model Download with Transformers library on demo code

8 days ago

When running the example inference code from the Model Card, I get this error:

processor_config.json: 100%|█████████████68.0/68.0 [00:00<00:00, 90.2kB/s]
chat_template.json: 100%|██████████████| 429/429 [00:00<00:00, 1.46MB/s]
Traceback (most recent call last):
...
File "/.../.../site-packages/transformers/tokenization_utils_fast.py", line 116, in __init__
          fast_tokenizer = TokenizerFast.from_file(fast_tokenizer_file)
Exception: trailing characters at line 244976 column 2

Any ideas what the cause could be?

pcuenq

Hugging Face TB Research org 8 days ago

Hi @sidrajaram ! What version of transformers are you using? I'd recommend you upgrade to the latest one if possible.

sidrajaram

8 days ago

Thanks @pcuenq . I am using latest transformers installed from source. Still encountering the same error.
I am running the unedited demo code on an M2 Mac.

IoDmitri

7 days ago

@pcuenq I've raised a similar issue as well within the discussion. Running Latest transformers on the M1 Mac and getting errors that are similar to it

Xenova

Hugging Face TB Research org 7 days ago

I'm not noticing any trailing characters on my side - can you open the file in a text editing software to check it hasn't been corrupted in some way?