From f8f1a85a1fe502db6eb3ab2c9f4a6cba8eb98218 Mon Sep 17 00:00:00 2001 From: Evelina Date: Thu, 7 Aug 2025 13:17:02 -0700 Subject: [PATCH] rename folder, add datasets Signed-off-by: Evelina --- recipes/README.md | 2 ++ recipes/{openscience => opensciencereasoning}/README.md | 5 ++++- .../prompts/mcq_augment_inspired_by.yaml | 0 .../prompts/mcq_augment_similar.yaml | 0 .../prompts/mcq_four_options.yaml | 0 .../prompts/mcq_ten_options.yaml | 0 .../prompts/subtopic_expansion.yaml | 0 .../scripts/filter_mcq_solutions.py | 0 8 files changed, 6 insertions(+), 1 deletion(-) rename recipes/{openscience => opensciencereasoning}/README.md (79%) rename recipes/{openscience => opensciencereasoning}/prompts/mcq_augment_inspired_by.yaml (100%) rename recipes/{openscience => opensciencereasoning}/prompts/mcq_augment_similar.yaml (100%) rename recipes/{openscience => opensciencereasoning}/prompts/mcq_four_options.yaml (100%) rename recipes/{openscience => opensciencereasoning}/prompts/mcq_ten_options.yaml (100%) rename recipes/{openscience => opensciencereasoning}/prompts/subtopic_expansion.yaml (100%) rename recipes/{openscience => opensciencereasoning}/scripts/filter_mcq_solutions.py (100%) diff --git a/recipes/README.md b/recipes/README.md index 3756eae6ff..aa5484af18 100644 --- a/recipes/README.md +++ b/recipes/README.md @@ -5,3 +5,5 @@ * [OpenCodeReasoning](https://nvidia.github.io/NeMo-Skills/releases/opencodereasoning/) * [OpenMathReasoning](https://nvidia.github.io/NeMo-Skills/releases/openmathreasoning/) + +* [OpenScienceReasoning](https://nvidia.github.io/NeMo-Skills/releases/opensciencereasoning/) diff --git a/recipes/openscience/README.md b/recipes/opensciencereasoning/README.md similarity index 79% rename from recipes/openscience/README.md rename to recipes/opensciencereasoning/README.md index 8f8b668525..08b15f9a49 100644 --- a/recipes/openscience/README.md +++ b/recipes/opensciencereasoning/README.md @@ -1,6 +1,9 @@ # Reproducing the OpenScience Dataset collection -This recipe contains the scripts and prompts to reproduce the **OpenScience** dataset, as described in the paper "A Scalable LLM Framework for Seed-Free Synthetic Data Generation". +This recipe contains the scripts and prompts to reproduce the **OpenScience** datasets: + +* [OpenScienceReasoning-2](https://huggingface.co/datasets/nvidia/OpenScienceReasoning-2) +* [OpenScience](https://huggingface.co/datasets/nvidia/OpenScience) ## What We Share diff --git a/recipes/openscience/prompts/mcq_augment_inspired_by.yaml b/recipes/opensciencereasoning/prompts/mcq_augment_inspired_by.yaml similarity index 100% rename from recipes/openscience/prompts/mcq_augment_inspired_by.yaml rename to recipes/opensciencereasoning/prompts/mcq_augment_inspired_by.yaml diff --git a/recipes/openscience/prompts/mcq_augment_similar.yaml b/recipes/opensciencereasoning/prompts/mcq_augment_similar.yaml similarity index 100% rename from recipes/openscience/prompts/mcq_augment_similar.yaml rename to recipes/opensciencereasoning/prompts/mcq_augment_similar.yaml diff --git a/recipes/openscience/prompts/mcq_four_options.yaml b/recipes/opensciencereasoning/prompts/mcq_four_options.yaml similarity index 100% rename from recipes/openscience/prompts/mcq_four_options.yaml rename to recipes/opensciencereasoning/prompts/mcq_four_options.yaml diff --git a/recipes/openscience/prompts/mcq_ten_options.yaml b/recipes/opensciencereasoning/prompts/mcq_ten_options.yaml similarity index 100% rename from recipes/openscience/prompts/mcq_ten_options.yaml rename to recipes/opensciencereasoning/prompts/mcq_ten_options.yaml diff --git a/recipes/openscience/prompts/subtopic_expansion.yaml b/recipes/opensciencereasoning/prompts/subtopic_expansion.yaml similarity index 100% rename from recipes/openscience/prompts/subtopic_expansion.yaml rename to recipes/opensciencereasoning/prompts/subtopic_expansion.yaml diff --git a/recipes/openscience/scripts/filter_mcq_solutions.py b/recipes/opensciencereasoning/scripts/filter_mcq_solutions.py similarity index 100% rename from recipes/openscience/scripts/filter_mcq_solutions.py rename to recipes/opensciencereasoning/scripts/filter_mcq_solutions.py