Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Inference Script #10

Open
jia11112727 opened this issue Jan 20, 2025 · 0 comments
Open

Inference Script #10

jia11112727 opened this issue Jan 20, 2025 · 0 comments

Comments

@jia11112727
Copy link

Hello. Thank you for Great Work! I encountered an issue while using the Pangea-7B-hf fine-tuned multilingual model. The model itself is multimodal, but I only need to use text input and do not require the image functionality. When running the model, I received an error indicating that image input is required, but I do not intend to use image input. Is there a way to disable image input and only use the text functionality? Below is the code template I am using:

Assuming that you have text_input and image_path

from transformers import LlavaNextForConditionalGeneration, AutoProcessor
import torch
from PIL import Image

image_input = Image.open(image_path)

model = LlavaNextForConditionalGeneration.from_pretrained(
"neulab/Pangea-7B-hf",
torch_dtype=torch.float16
).to(0)
processor = AutoProcessor.from_pretrained("neulab/Pangea-7B-hf")
model.resize_token_embeddings(len(processor.tokenizer))

text_input = f"<|im_start|>system\nYou are a helpful assistant.<|im_end|>\n<|im_start|>user\n\n{text_input}<|im_end|>\n<|im_start|>assistant\n"
model_inputs = processor(images=image_input, text=text_input, return_tensors='pt').to("cuda", torch.float16)
output = model.generate(**model_inputs, max_new_tokens=1024, min_new_tokens=32, temperature=1.0, top_p=0.9, do_sample=True)
output = output[0]
result = processor.decode(output, skip_special_tokens=True, clean_up_tokenization_spaces=False)

print(result)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant