openbmb

/

MiniCPM-Llama3-V-2_5

like 1.33k

After updating, if we pass model only the text without the figure would cause the following bug:

"xxxximage_utils.py", line 114, in is_batched
    return is_valid_image(img[0])
IndexError: list index out of range

Here is the code

import torch
from PIL import Image
from transformers import AutoModelForCausalLM, AutoTokenizer,AutoProcessor
import json
from copy import deepcopy

torch.manual_seed(0)

model = AutoModelForCausalLM.from_pretrained("openbmb/MiniCPM-Llama3-V-2_5", trust_remote_code=True,
            torch_dtype=torch.bfloat16,cache_dir="cache") # sdpa or flash_attention_2, no eager
model = model.eval().cuda()
tokenizer = AutoTokenizer.from_pretrained("openbmb/MiniCPM-Llama3-V-2_5", trust_remote_code=True)


msgs = [{'role': 'user', 'content': ["Introduce something about Airbus A380."]}]

answer = model.chat(
    image=None,
    msgs=msgs,
    context=None,
    tokenizer=tokenizer
)
print(answer)