How can I get llama_index to use the entire document i pass in as context

Just a bit confused on what model this embed_model is?

Is there anyway I can pass in text for the as the stop text or use the tokenizer without using transformers autotokenizer and having it called twice