dataset_path: FunAILab/TVBench
dataset_kwargs:
  token: True
  cache_dir: tvbench
  video: True
generation_kwargs:
  max_new_tokens: 32
  temperature: 0
  do_sample: false
output_type: generate_until
doc_to_visual: !function utils.tvbench_doc_to_visual
doc_to_text: !function utils.tvbench_doc_to_text
doc_to_target: !function utils.tvbench_doc_to_target
process_results: !function utils.tvbench_process_results
metric_list:
  - metric: tvbench_acc
    aggregation: mean
    higher_is_better: true
lmms_eval_specific_kwargs:
  default:
    pre_prompt: ""
    post_prompt: "Answer with the option letter only."
metadata:
  - version: 0.0
