fix: sync eqbench-ja/setup_translate.sh

fix: sync eqbench-ja/serve_translate.sh

fix: sync eqbench-ja/translate_eqbench.py

sync: eqbench-ja/translate_eqbench.py

sync: eqbench-ja/serve_translate.sh

sync: eqbench-ja/setup_translate.sh

fix: remove get_peft_model from DPO (SFT adapter already has LoRA)

feat: PR#8 - SFT/DPO 学習スクリプト追加 (#8)

feat: PR#6 - EQ全展開・max-cycles CLI・<think>統一・フィルタ改善（mainから再実装） (#6)

fix: LLMのオブジェクト形式JSON応答に対応 (#4)

add: LLM output debug script

fix: enable_thinking制御の分離・content取得ロジック修正 (#3)

fix: マトリックス生成のセマフォ追加・max_tokens拡張 (#2)

fix: correct zip folder name to teememo-synth

feat: EQ-Matrixプロンプト生成・dedupfix・MAX_CYCLES=3 (#1)

Upload 2 files

Delete setup.sh

Delete teememo-synth-complete.zip

Upload 2 files

Delete setup.sh

Delete teememo-synth-complete.zip

Upload 2 files

Delete setup.sh

Delete teememo-synth-complete.zip

Upload 2 files

initial commit