HelpSteer: Helpfulness SteerLM Dataset

HelpSteer is an open-source Helpfulness Dataset (CC-BY-4.0) that supports aligning models to become more helpful, factually correct and coherent, while being adjustable in terms of the complexity and verbosity of its responses. Leveraging this dataset and SteerLM, we train a Llama 2 70B to reach 7.54 on MT Bench, the highest among models trained on open-source datasets based on MT Bench Leaderboard as of 15 Nov 2023. This model is available on… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/HelpSteer.

HelpSteer-filtered

This dataset is a highly filtered version of the nvidia/HelpSteer dataset.

  ❓ How this dataset was filtered:

I calculated the sum of the columns ["helpfulness," "correctness," "coherence," "complexity," "verbosity"] and created a new column named sum.

I changed some column names and added a empty column to match the Alpaca format.

The dataset was then filtered to include only those entries with a sum greater than or equal to 16.

  🧐 More… See the full description on the dataset page: https://huggingface.co/datasets/Weyaxi/HelpSteer-filtered.

HelpSteer2: Open-source dataset for training top-performing reward models

HelpSteer2 is an open-source Helpfulness Dataset (CC-BY-4.0) that supports aligning models to become more helpful, factually correct and coherent, while being adjustable in terms of the complexity and verbosity of its responses. This dataset has been created in partnership with Scale AI. When used to tune a Llama 3.1 70B Instruct Model, we achieve 94.1% on RewardBench, which makes it the best Reward Model as… See the full description on the dataset page: https://huggingface.co/datasets/nvidia/HelpSteer2.

RLHFlow/Helpsteer-preference-standard dataset hosted on Hugging Face and contributed by the HF Datasets community

HelpSteer: Helpfulness SteerLM Dataset

HelpSteer is an open-source Helpfulness Dataset (CC-BY-4.0) that supports aligning models to become more helpful, factually correct and coherent, while being adjustable in terms of the complexity and verbosity of its responses. HelpSteer: Multi-attribute Helpfulness Dataset for SteerLM

  Disclaimer

This is only a subset created with distilabel to evaluate the first 1000 rows using AI Feedback (AIF) coming from GPT-4, only created for… See the full description on the dataset page: https://huggingface.co/datasets/alvarobartt/HelpSteer-AIF.

NewEden/Helpsteer-3-edit-kto-v7 dataset hosted on Hugging Face and contributed by the HF Datasets community

kh4dien/helpsteer-9k dataset hosted on Hugging Face and contributed by the HF Datasets community

NewEden/Helpsteer-3-edit dataset hosted on Hugging Face and contributed by the HF Datasets community

cheryyunl/helpsteer dataset hosted on Hugging Face and contributed by the HF Datasets community

withpi/nvidia-HelpSteer-group-label_normalized dataset hosted on Hugging Face and contributed by the HF Datasets community

withpi/nvidia-HelpSteer-group-label-v2_euro_st_tokenized_32k_1 dataset hosted on Hugging Face and contributed by the HF Datasets community

Delta-Vector/Helpsteer-3-Edit-ShareGPT dataset hosted on Hugging Face and contributed by the HF Datasets community

withpi/nvidia-HelpSteer-group-label-v2_tokenized_truncated_regcat_1 dataset hosted on Hugging Face and contributed by the HF Datasets community

nikhilkumarreddy28/helpsteer-llama2-1k dataset hosted on Hugging Face and contributed by the HF Datasets community

withpi/nvidia-HelpSteer-group-label-v2_tokenized_16k_euro_1 dataset hosted on Hugging Face and contributed by the HF Datasets community

Helpsteer-correctness

This dataset is derived from NVIDIA's HelpSteer dataset, processed specifically for preference learning on the correctness dimension.

- Train split: 27417 examples
- Test split: 1416 examples

## Format

Each example contains the following fields:
- `prompt`: Question with "Human:" prefix and "Assistant:" suffix
- `chosen`: The response with higher correctness score
- `rejected`: The response with lower correctness score
-… See the full description on the dataset page: https://huggingface.co/datasets/cheryyunl/helpsteer-correctness.

Helpsteer-coherence

This dataset is derived from NVIDIA's HelpSteer dataset, processed specifically for preference learning on the coherence dimension.

- Train split: 22876 examples
- Test split: 1131 examples

## Format

Each example contains the following fields:
- `prompt`: Question with "Human:" prefix and "Assistant:" suffix
- `chosen`: The response with higher coherence score
- `rejected`: The response with lower coherence score
-… See the full description on the dataset page: https://huggingface.co/datasets/cheryyunl/helpsteer-coherence.

Helpsteer-helpfulness

This dataset is derived from NVIDIA's HelpSteer dataset, processed specifically for preference learning on the helpfulness dimension.

- Train split: 28115 examples
- Test split: 1410 examples

## Format

Each example contains the following fields:
- `prompt`: Question with "Human:" prefix and "Assistant:" suffix
- `chosen`: The response with higher helpfulness score
- `rejected`: The response with lower helpfulness score
-… See the full description on the dataset page: https://huggingface.co/datasets/cheryyunl/helpsteer-helpfulness.

Delta-Vector/Tauri-Helpsteer-3-Preference-KTO dataset hosted on Hugging Face and contributed by the HF Datasets community

NewEden/Helpsteer-3-Pref dataset hosted on Hugging Face and contributed by the HF Datasets community