| project: |
| seed: 1945 |
| era_start: 1868 |
| era_end: 1945 |
| reports_dir: reports |
| data: |
| index_csv: data/index/list_person_all_extended_utf8.csv |
| aozora_repo: data/raw/aozora |
| raw_manifest: data/processed/aozora/selection.jsonl |
| holdout_manifest: data/processed/aozora/holdout.jsonl |
| clean_dir: data/clean/aozora |
| clean_manifest: data/processed/aozora/clean.jsonl |
| aozora_split_manifest: data/processed/aozora/splits.jsonl |
| split_manifest: data/processed/aozora/splits.jsonl |
| min_clean_chars: 200 |
| split: |
| train: 0.96 |
| validation: 0.02 |
| test: 0.02 |
| tokenizer: |
| vocab_size: 8192 |
| min_frequency: 2 |
| directory: artifacts/tokenizer |
| tokenized: |
| directory: data/processed/tokenized |
| huggingface: |
| namespace: "trtd56" |
| private: true |
| dataset_repos: |
| phase1: "trtd56/TimeCapsuleLLM-ja-corpus-v0" |
| phase2: "trtd56/TimeCapsuleLLM-ja-corpus-v1-small" |
| model_repos: |
| phase1: "trtd56/TimeCapsuleLLM-ja-v0" |
| phase2: "trtd56/TimeCapsuleLLM-ja-v1-small" |
| licenses: |
| dataset: other |
| model: apache-2.0 |
|
|
