metadata
dataset_info:
features:
- name: dataset
dtype: string
- name: question_id
dtype: string
- name: question_text
dtype: string
- name: answers_objects
list:
- name: date
struct:
- name: day
dtype: string
- name: month
dtype: string
- name: year
dtype: string
- name: number
dtype: string
- name: spans
sequence: string
- name: contexts
list:
- name: idx
dtype: int64
- name: is_supporting
dtype: bool
- name: paragraph_text
dtype: string
- name: title
dtype: string
- name: reference
sequence: string
- name: gold_context
sequence: string
- name: retrieved_contexts
sequence: string
- name: context_response
dtype: string
- name: no_context_response
dtype: string
- name: tokens_all_context
dtype: int64
- name: tokens_no_context
dtype: int64
- name: MaximumTokenProbability
sequence: float32
- name: TokenEntropy
sequence: float32
- name: Perplexity
dtype: float64
- name: PTrue
dtype: float64
- name: Verbalized1S
dtype: float64
- name: MeanPointwiseMutualInformation
dtype: float64
- name: MeanConditionalPointwiseMutualInformation
dtype: float64
- name: RenyiNeg
dtype: float64
- name: FisherRao
dtype: float64
- name: SemanticEntropy
dtype: float64
- name: CCP
dtype: float64
- name: SAR
dtype: float64
- name: SentenceSAR
dtype: float64
- name: NumSemSets
dtype: int64
- name: EigValLaplacian_NLI_score_entail
dtype: float64
- name: DegMat_NLI_score_entail
dtype: float64
- name: Eccentricity_NLI_score_entail
dtype: float64
- name: LexicalSimilarity_rougeL
dtype: float64
- name: adaptive_rag_one_retrieve
dtype: string
- name: adaptive_rag_no_retrieve
dtype: string
- name: new_retriever_adaptive_rag_one_retrieve
dtype: string
- name: new_retriever_adaptive_rag_no_retrieve
dtype: string
- name: MahalanobisDistanceSeq_decoder
dtype: float64
- name: RelativeMahalanobisDistanceSeq_decoder
dtype: float64
- name: RDESeq_decoder
dtype: float64
splits:
- name: train
num_bytes: 5575219
num_examples: 500
- name: test
num_bytes: 10519946
num_examples: 500
download_size: 9600283
dataset_size: 16095165
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*