reeval-sft / README.md
martinakaduc's picture
Upload dataset
c317bdf verified
|
raw
history blame
1.19 kB
metadata
dataset_info:
  - config_name: air-bench_air_bench_2024_Meta-Llama-3.1-8B-Instruct
    features:
      - name: text
        dtype: string
      - name: difficulty
        dtype: float64
    splits:
      - name: train
        num_bytes: 5548296
        num_examples: 3644
      - name: test
        num_bytes: 1387074
        num_examples: 911
    download_size: 2496377
    dataset_size: 6935370
  - config_name: classic_babi_qa_Meta-Llama-3.1-8B-Instruct
    features:
      - name: text
        dtype: string
      - name: difficulty
        dtype: float64
    splits:
      - name: train
        num_bytes: 1151522.288265306
        num_examples: 1254
      - name: test
        num_bytes: 288339.7117346939
        num_examples: 314
    download_size: 202117
    dataset_size: 1439862
configs:
  - config_name: air-bench_air_bench_2024_Meta-Llama-3.1-8B-Instruct
    data_files:
      - split: train
        path: air-bench_air_bench_2024_Meta-Llama-3.1-8B-Instruct/train-*
      - split: test
        path: air-bench_air_bench_2024_Meta-Llama-3.1-8B-Instruct/test-*
  - config_name: classic_babi_qa_Meta-Llama-3.1-8B-Instruct
    data_files:
      - split: train
        path: classic_babi_qa_Meta-Llama-3.1-8B-Instruct/train-*
      - split: test
        path: classic_babi_qa_Meta-Llama-3.1-8B-Instruct/test-*