fcyin's picture
Upload dataset
a51b3c8 verified
|
raw
history blame
3.61 kB
metadata
dataset_info:
  - config_name: gpt-4o-mini
    features:
      - name: model_response
        dtype: string
      - name: parsed_model_answer
        dtype: string
      - name: is_correct
        dtype: bool
      - name: answer_was_parsed
        dtype: bool
      - name: errored
        dtype: bool
      - name: messages
        dtype: string
      - name: payload
        dtype: string
      - name: question
        dtype: string
      - name: answer
        dtype: string
      - name: answerKey
        dtype: string
      - name: choices
        dtype: string
      - name: additional_information
        dtype: string
    splits:
      - name: dataset
        num_bytes: 927366
        num_examples: 100
    download_size: 202711
    dataset_size: 927366
  - config_name: llama3.1-70b-instruct
    features:
      - name: model_response
        dtype: string
      - name: parsed_model_answer
        dtype: string
      - name: is_correct
        dtype: bool
      - name: answer_was_parsed
        dtype: bool
      - name: errored
        dtype: bool
      - name: messages
        dtype: string
      - name: payload
        dtype: string
      - name: question
        dtype: string
      - name: answer
        dtype: string
      - name: answerKey
        dtype: string
      - name: choices
        dtype: string
      - name: additional_information
        dtype: string
    splits:
      - name: gsm8k_hard_direct_solver
        num_bytes: 12346681
        num_examples: 1319
      - name: gsm8k_direct_solver
        num_bytes: 12284315
        num_examples: 1319
      - name: contexthub_abductive_level1_direct_solver
        num_bytes: 7068745
        num_examples: 360
      - name: contexthub_deductive_level1_direct_solver
        num_bytes: 11778671
        num_examples: 600
    download_size: 6039939
    dataset_size: 43478412
  - config_name: llama3.1-8b-instruct
    features:
      - name: model_response
        dtype: string
      - name: parsed_model_answer
        dtype: string
      - name: is_correct
        dtype: bool
      - name: answer_was_parsed
        dtype: bool
      - name: errored
        dtype: bool
      - name: messages
        dtype: string
      - name: payload
        dtype: string
      - name: question
        dtype: string
      - name: answer
        dtype: string
      - name: answerKey
        dtype: string
      - name: choices
        dtype: string
      - name: additional_information
        dtype: string
    splits:
      - name: gsm8k_hard_direct_solver
        num_bytes: 12270577
        num_examples: 1319
      - name: gsm8k_direct_solver
        num_bytes: 12191399
        num_examples: 1319
      - name: contexthub_abductive_level1_direct_solver
        num_bytes: 7081875
        num_examples: 360
      - name: contexthub_deductive_level1_direct_solver
        num_bytes: 11785539
        num_examples: 600
    download_size: 6006315
    dataset_size: 43329390
configs:
  - config_name: gpt-4o-mini
    data_files:
      - split: dataset
        path: gpt-4o-mini/dataset-*
  - config_name: llama3.1-70b-instruct
    data_files:
      - split: gsm8k_hard_direct_solver
        path: llama3.1-70b-instruct/gsm8k_hard_direct_solver-*
      - split: gsm8k_direct_solver
        path: llama3.1-70b-instruct/gsm8k_direct_solver-*
      - split: contexthub_abductive_level1_direct_solver
        path: llama3.1-70b-instruct/contexthub_abductive_level1_direct_solver-*
      - split: contexthub_deductive_level1_direct_solver
        path: llama3.1-70b-instruct/contexthub_deductive_level1_direct_solver-*
  - config_name: llama3.1-8b-instruct
    data_files:
      - split: gsm8k_hard_direct_solver
        path: llama3.1-8b-instruct/gsm8k_hard_direct_solver-*
      - split: gsm8k_direct_solver
        path: llama3.1-8b-instruct/gsm8k_direct_solver-*
      - split: contexthub_abductive_level1_direct_solver
        path: llama3.1-8b-instruct/contexthub_abductive_level1_direct_solver-*
      - split: contexthub_deductive_level1_direct_solver
        path: llama3.1-8b-instruct/contexthub_deductive_level1_direct_solver-*