When running either app.py or inference.py, I encountered the problem "safetensors_rust.SafetensorError: Error while deserializing header: HeaderTooLarge". I have already cloned all the necessary files from the HF repository to my local environment and modified the model path accordingly. The error is specifically pointing to line 106 in mPLUG-DocOwl-main/TinyChart/tinychart/model/builder.py, where the tokenizer is loaded using AutoTokenizer.from_pretrained(model_path, use_fast=False, padding_side="right").
When running either app.py or inference.py, I encountered the problem "safetensors_rust.SafetensorError: Error while deserializing header: HeaderTooLarge". I have already cloned all the necessary files from the HF repository to my local environment and modified the model path accordingly. The error is specifically pointing to line 106 in mPLUG-DocOwl-main/TinyChart/tinychart/model/builder.py, where the tokenizer is loaded using AutoTokenizer.from_pretrained(model_path, use_fast=False, padding_side="right").