├── .gitignore ├── CODE_OF_CONDUCT.md ├── CONTRIBUTING.md ├── LICENSE ├── LICENSE_HEADER.txt ├── README.md ├── evaluators ├── __init__.py ├── base_evaluator.py ├── extraction_evaluators.py └── match_evaluators.py ├── finetuning ├── README.md ├── configs │ ├── dpo.yaml │ ├── dpo_70b.yaml │ ├── dpo_8b.yaml │ ├── sft.yaml │ ├── sft_70b.yaml │ └── sft_8b.yaml ├── convert_ckpt.sh ├── data_creation │ ├── filtering_utils.py │ └── ts_correctness_filtering.py ├── eval_finetuned_model.py ├── iterative_training.py ├── run_fairseq2_training.py └── slurm_utils.py ├── imgs └── coral_illustration.png ├── lightning_modules ├── __init__.py ├── configs │ ├── common │ │ ├── api_inference_trainer.yaml │ │ └── ts_interaction_model.yaml │ ├── cot │ │ ├── cot_gpqa.yaml │ │ ├── cot_math.yaml │ │ └── cot_mmlu_pro.yaml │ ├── ts_explore_tom.yaml │ ├── ts_gpqa.yaml │ ├── ts_hi_tom.yaml │ ├── ts_math.yaml │ └── ts_mmlu_pro.yaml ├── datasets │ ├── __init__.py │ ├── base_dataset.py │ ├── ts_explore_tom_dataset.py │ ├── ts_gpqa_dataset.py │ ├── ts_math_dataset.py │ └── ts_mmlu_pro_dataset.py ├── loggers │ └── patched_loggers.py ├── models │ ├── tree_sampling_utils.py │ └── ts_interactive_model.py └── trainer.py ├── pyproject.toml ├── pyrightconfig.json ├── pytest.ini ├── requirements.txt └── utils ├── __init__.py ├── chat_utils.py ├── eval_utils.py ├── file_utils.py ├── http_utils.py ├── prompt_library.json └── prompt_utils.py /.gitignore: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/.gitignore -------------------------------------------------------------------------------- /CODE_OF_CONDUCT.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/CODE_OF_CONDUCT.md -------------------------------------------------------------------------------- /CONTRIBUTING.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/CONTRIBUTING.md -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/LICENSE -------------------------------------------------------------------------------- /LICENSE_HEADER.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/LICENSE_HEADER.txt -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/README.md -------------------------------------------------------------------------------- /evaluators/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/evaluators/__init__.py -------------------------------------------------------------------------------- /evaluators/base_evaluator.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/evaluators/base_evaluator.py -------------------------------------------------------------------------------- /evaluators/extraction_evaluators.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/evaluators/extraction_evaluators.py -------------------------------------------------------------------------------- /evaluators/match_evaluators.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/evaluators/match_evaluators.py -------------------------------------------------------------------------------- /finetuning/README.md: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/finetuning/README.md -------------------------------------------------------------------------------- /finetuning/configs/dpo.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/finetuning/configs/dpo.yaml -------------------------------------------------------------------------------- /finetuning/configs/dpo_70b.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/finetuning/configs/dpo_70b.yaml -------------------------------------------------------------------------------- /finetuning/configs/dpo_8b.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/finetuning/configs/dpo_8b.yaml -------------------------------------------------------------------------------- /finetuning/configs/sft.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/finetuning/configs/sft.yaml -------------------------------------------------------------------------------- /finetuning/configs/sft_70b.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/finetuning/configs/sft_70b.yaml -------------------------------------------------------------------------------- /finetuning/configs/sft_8b.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/finetuning/configs/sft_8b.yaml -------------------------------------------------------------------------------- /finetuning/convert_ckpt.sh: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/finetuning/convert_ckpt.sh -------------------------------------------------------------------------------- /finetuning/data_creation/filtering_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/finetuning/data_creation/filtering_utils.py -------------------------------------------------------------------------------- /finetuning/data_creation/ts_correctness_filtering.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/finetuning/data_creation/ts_correctness_filtering.py -------------------------------------------------------------------------------- /finetuning/eval_finetuned_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/finetuning/eval_finetuned_model.py -------------------------------------------------------------------------------- /finetuning/iterative_training.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/finetuning/iterative_training.py -------------------------------------------------------------------------------- /finetuning/run_fairseq2_training.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/finetuning/run_fairseq2_training.py -------------------------------------------------------------------------------- /finetuning/slurm_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/finetuning/slurm_utils.py -------------------------------------------------------------------------------- /imgs/coral_illustration.png: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/imgs/coral_illustration.png -------------------------------------------------------------------------------- /lightning_modules/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/__init__.py -------------------------------------------------------------------------------- /lightning_modules/configs/common/api_inference_trainer.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/configs/common/api_inference_trainer.yaml -------------------------------------------------------------------------------- /lightning_modules/configs/common/ts_interaction_model.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/configs/common/ts_interaction_model.yaml -------------------------------------------------------------------------------- /lightning_modules/configs/cot/cot_gpqa.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/configs/cot/cot_gpqa.yaml -------------------------------------------------------------------------------- /lightning_modules/configs/cot/cot_math.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/configs/cot/cot_math.yaml -------------------------------------------------------------------------------- /lightning_modules/configs/cot/cot_mmlu_pro.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/configs/cot/cot_mmlu_pro.yaml -------------------------------------------------------------------------------- /lightning_modules/configs/ts_explore_tom.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/configs/ts_explore_tom.yaml -------------------------------------------------------------------------------- /lightning_modules/configs/ts_gpqa.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/configs/ts_gpqa.yaml -------------------------------------------------------------------------------- /lightning_modules/configs/ts_hi_tom.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/configs/ts_hi_tom.yaml -------------------------------------------------------------------------------- /lightning_modules/configs/ts_math.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/configs/ts_math.yaml -------------------------------------------------------------------------------- /lightning_modules/configs/ts_mmlu_pro.yaml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/configs/ts_mmlu_pro.yaml -------------------------------------------------------------------------------- /lightning_modules/datasets/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/datasets/__init__.py -------------------------------------------------------------------------------- /lightning_modules/datasets/base_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/datasets/base_dataset.py -------------------------------------------------------------------------------- /lightning_modules/datasets/ts_explore_tom_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/datasets/ts_explore_tom_dataset.py -------------------------------------------------------------------------------- /lightning_modules/datasets/ts_gpqa_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/datasets/ts_gpqa_dataset.py -------------------------------------------------------------------------------- /lightning_modules/datasets/ts_math_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/datasets/ts_math_dataset.py -------------------------------------------------------------------------------- /lightning_modules/datasets/ts_mmlu_pro_dataset.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/datasets/ts_mmlu_pro_dataset.py -------------------------------------------------------------------------------- /lightning_modules/loggers/patched_loggers.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/loggers/patched_loggers.py -------------------------------------------------------------------------------- /lightning_modules/models/tree_sampling_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/models/tree_sampling_utils.py -------------------------------------------------------------------------------- /lightning_modules/models/ts_interactive_model.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/models/ts_interactive_model.py -------------------------------------------------------------------------------- /lightning_modules/trainer.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/lightning_modules/trainer.py -------------------------------------------------------------------------------- /pyproject.toml: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/pyproject.toml -------------------------------------------------------------------------------- /pyrightconfig.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/pyrightconfig.json -------------------------------------------------------------------------------- /pytest.ini: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/pytest.ini -------------------------------------------------------------------------------- /requirements.txt: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/requirements.txt -------------------------------------------------------------------------------- /utils/__init__.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/utils/__init__.py -------------------------------------------------------------------------------- /utils/chat_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/utils/chat_utils.py -------------------------------------------------------------------------------- /utils/eval_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/utils/eval_utils.py -------------------------------------------------------------------------------- /utils/file_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/utils/file_utils.py -------------------------------------------------------------------------------- /utils/http_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/utils/http_utils.py -------------------------------------------------------------------------------- /utils/prompt_library.json: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/utils/prompt_library.json -------------------------------------------------------------------------------- /utils/prompt_utils.py: -------------------------------------------------------------------------------- https://raw.githubusercontent.com/facebookresearch/collaborative-reasoner/HEAD/utils/prompt_utils.py --------------------------------------------------------------------------------