readme.md

MulTiCast

Multimodal Time-series foreCasting

Requires python 3.10.16

Check requirements.txt

from transformers import AutoModel
model = AutoModel.from_pretrained("adnlp/MulTiCast", trust_remote_code=True)

import torch
input_ids = torch.rand((1, 56))
images = torch.rand((224,224,3))
texts = "Hello World!"

out = model.predict(input_ids=input_ids, images=images, texts=texts)
pred = out["logits"]