小编DrR*_*rry的帖子

我正在尝试使用 MPS 后端在 M1 Mac 上运行 T5 变压器：

import torch
import json 
from transformers import T5Tokenizer, T5ForConditionalGeneration, T5Config
#Make sure sentencepiece is installed
device = torch.device('mps')
model = T5ForConditionalGeneration.from_pretrained('t5-3b').to("mps")
tokenizer = T5Tokenizer.from_pretrained('t5-3b')#, device = device)

preprocess_text = full_text.strip().replace("\n",".")
t5_prepared_Text = "summarize: "+preprocess_text
print ("original text preprocessed: \n", preprocess_text)
tokenized_text = tokenizer.encode(t5_prepared_Text, return_tensors="pt").to(device)

# summmarize 
summary_ids = model.generate(tokenized_text,
                                    num_beams=6,
                                    no_repeat_ngram_size=3,
                                    min_length=30,
                                    max_length=9000,
                                    early_stopping=True)

output = tokenizer.decode(summary_ids[0], skip_special_tokens=True)

print ("\n\nSummarized text: \n",output)

其中“full_text”是之前定义的字符串。该代码适用于 CPU，但我想通过使用 MPS 来加速它（如上面的代码所示）。在这里，我收到错误：

TypeError: Operation 'abs_out_mps()' does not support …