unlearning-cleanslate/verif-char-windowing-qwen-8b

---
dataset_info:
features:
- name: text_length_chars
dtype: int64
- name: num_windows
dtype: int64
- name: memorized_windows
dtype: int64
- name: memorized_fraction
dtype: float64
- name: reproduced_windows
dtype: int64
- name: reproduced_fraction
dtype: float64
- name: avg_rouge_l
dtype: float64
- name: max_rouge_l
dtype: float64
- name: coverage
dtype: float64
- name: max_p_z
dtype: float64
- name: mean_p_z
dtype: float64
- name: median_p_z
dtype: float64
- name: min_p_z
dtype: float64
- name: std_p_z
dtype: float64
- name: mean_perplexity
dtype: float64
- name: min_perplexity
dtype: float64
- name: best_window_idx
dtype: int64
- name: best_window_p_z
dtype: float64
- name: best_window_seed
dtype: large_string
- name: best_window_target
dtype: large_string
- name: best_window_start_char
dtype: int64
- name: best_window_end_char
dtype: int64
- name: best_window_generated
dtype: 'null'
- name: best_window_beams
dtype: 'null'
- name: best_window_bleu_1
dtype: 'null'
- name: best_window_rouge_l
dtype: 'null'
- name: best_window_exact_match
dtype: 'null'
- name: best_window_any_exact_match
dtype: 'null'
- name: best_window_max_bleu_1
dtype: 'null'
- name: best_window_max_rouge_l
dtype: 'null'
- name: avg_max_bleu_1
dtype: 'null'
- name: avg_max_rouge_l
dtype: 'null'
- name: avg_any_exact_match
dtype: 'null'
- name: eval_model
dtype: large_string
- name: eval_threshold
dtype: float64
- name: window_size
dtype: int64
- name: stride
dtype: int64
- name: decode_strategy
dtype: 'null'
- name: beam_width
dtype: 'null'
- name: length_penalty
dtype: 'null'
- name: reproduce_all_windows
dtype: bool
- name: windows
list:
- name: end_char
dtype: int64
- name: idx
dtype: int64
- name: is_memorized
dtype: bool
- name: log_prob
dtype: float64
- name: p_z
dtype: float64
- name: perplexity
dtype: float64
- name: rank_1_fraction
dtype: float64
- name: seed
dtype: string
- name: start_char
dtype: int64
- name: target
dtype: string
- name: content_id
dtype: large_string
- name: content_title
dtype: large_string
- name: content_creators
dtype: large_string
- name: content_year
dtype: int64
- name: reference_target
dtype: large_string
- name: song_id
dtype: large_string
- name: song_title
dtype: large_string
splits:
- name: train
num_bytes: 228625223
num_examples: 4653
download_size: 209268886
dataset_size: 228625223
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
---