The program currently limits GPT-2 tokens to 2800 per call. Higher values tended to exceed the max length of 4096 when dealing with longer files. Smaller files, however, had no difficulty with the default values.
Several OpenAI calls combine results from previous prompts with data from new prompts. These results can vary in length, and the maximum token length in the summarization functions should automatically adjust to account for this.
The program currently limits GPT-2 tokens to 2800 per call. Higher values tended to exceed the max length of 4096 when dealing with longer files. Smaller files, however, had no difficulty with the default values.
Several OpenAI calls combine results from previous prompts with data from new prompts. These results can vary in length, and the maximum token length in the summarization functions should automatically adjust to account for this.