ccdv

/

lsg-bart-base-4096-booksum

like 0

RuntimeError: The expanded size of the tensor (64) must match the existing size (2) at non-singleton dimension 0. Target sizes: [64]. Tensor sizes: [2]

from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline

tokenizer = AutoTokenizer.from_pretrained("ccdv/lsg-bart-base-4096-booksum", trust_remote_code=True)
model = AutoModelForSeq2SeqLM.from_pretrained("ccdv/lsg-bart-base-4096-booksum", trust_remote_code=True)

text = "Replace by what you want."*512
pipe = pipeline("text2text-generation", model=model, tokenizer=tokenizer)
generated_text = pipe(
  text, 
  truncation=True, 
  max_length=256, 
  no_repeat_ngram_size=7,
  num_beams=2,
  early_stopping=True
  )

Works fine on colab (transformers 4.38.2)

Ah I wasn't actually using the sample code exactly, as I had added device="mps" to pipe = pipeline("text2text-generation", model=model, tokenizer=tokenizer, device="mps") on MacOS.

Unsure if the above is a model error or user error, as I'm new to transformers and still learning, but can confirm the above works with no device given.

Didn't test on mac