| datasets = [ |
| [ |
| dict( |
| abbr='LongBench_trec', |
| eval_cfg=dict( |
| evaluator=dict( |
| type='opencompass.datasets.LongBenchClassificationEvaluator' |
| ), |
| pred_postprocessor=dict( |
| type='opencompass.datasets.trec_postprocess'), |
| pred_role='BOT'), |
| infer_cfg=dict( |
| inferencer=dict( |
| max_out_len=64, |
| type='opencompass.openicl.icl_inferencer.GenInferencer'), |
| prompt_template=dict( |
| template=dict(round=[ |
| dict( |
| prompt= |
| 'Please determine the type of the question below. Here are some examples of questions.\n\n{context}\n{input}', |
| role='HUMAN'), |
| ]), |
| type= |
| 'opencompass.openicl.icl_prompt_template.PromptTemplate'), |
| retriever=dict( |
| type='opencompass.openicl.icl_retriever.ZeroRetriever')), |
| name='trec', |
| path='opencompass/Longbench', |
| reader_cfg=dict( |
| input_columns=[ |
| 'context', |
| 'input', |
| ], |
| output_column='all_labels', |
| test_split='test', |
| train_split='test'), |
| type='opencompass.datasets.LongBenchtrecDataset'), |
| ], |
| ] |
| eval = dict(runner=dict(task=dict(dump_details=True))) |
| models = [ |
| dict( |
| abbr='gated_deltanet', |
| batch_size=128, |
| max_seq_len=2048, |
| model_kwargs=dict( |
| device_map='auto', |
| torch_dtype='torch.bfloat16', |
| trust_remote_code=True), |
| path='download_model/hgrn2-1.3B-100B', |
| run_cfg=dict(num_gpus=1), |
| tokenizer_kwargs=dict(padding_side='left', truncation_side='left'), |
| tokenizer_path='download_model/hgrn2-1.3B-100B', |
| type='opencompass.models.HuggingFaceBaseModel'), |
| ] |
| work_dir = 'outputs/default/20251219_164057' |
|
|
