configs/kilt_hybrid_ce.yaml

# This configuration file sets up a hybrid pipeline using a retriever, generator, and reranker.
# It is designed to work with the KILT dataset and uses FAISS for retrieval.

logging:
  log_dir: logs
  level: INFO
  max_mb: 5
  backups: 5

retriever:
  # using Faiss (dense) retrieval over KILT’s Wikipedia passages
  name: dense
  faiss_index: /path/to/kilt_wiki_faiss.index
  top_k: 5
  model_name: sentence-transformers/all-MiniLM-L6-v2
  device: cpu

generator:
  model_name: facebook/bart-large
  device: cpu
  max_new_tokens: 256
  temperature: 0.0

reranker:
  enable: true
  model_name: cross-encoder/ms-marco-MiniLM-L-6-v2
  device: cpu
  max_length: 512
  first_stage_k: 5
  final_k: 5

stats:
  correlation_method: spearman
  n_boot: 1000
  ci: 0.95
  wilcoxon_alternative: two-sided
  multiple_correction: holm-bonferroni
  alpha: 0.05
  compute_effect_size: true
  n_permutations: 1000
  failure_threshold: 0.0