Reads large files in specific chunks with pagination, providing exact lines to the model. Prevents context window overflow and VRAM spikes by extracting only the requested subset of lines.
