qmzpg

when I run demo.py, it crashes at "output = sess.run([softmax], feed_dict=feed)" with the error called"Number of ways to split should evenly divide the split dimension, but got split_dim 3 (size...

I change code in 'MiniGPT4-video/minigpt4/models/mini_gpt4_llama_v2.py' to let miniGpt4 can interference on multi Gpus if self.low_resource: self.llama_model = llm_model.from_pretrained( llama_model, torch_dtype=torch.float16, # torch_dtype = torch.bfloat16, load_in_8bit=True, device_map="auto", ) But the result...