Release Haystack-LVBench (TimeSearch-R SFT/RL Data) and link TimeSearch-R model on Hugging Face #1
Description
Hi @Theia-4869 🤗
Niels here from the open-source team at Hugging Face. I discovered your work on Arxiv and its corresponding GitHub repository, and was wondering whether you would like to submit it to hf.co/papers to improve its discoverability. If you are one of the authors, you can submit it at https://huggingface.co/papers/submit.
The paper page lets people discuss about your paper and lets them find artifacts about it (your models, datasets or demo for instance), you can also claim
the paper as yours which will show up on your public profile at HF, add Github and project page URLs.
It'd be great to make the checkpoints and dataset available on the 🤗 hub, to improve their discoverability/visibility.
We can add tags so that people find them when filtering https://huggingface.co/models and https://huggingface.co/datasets.
Uploading models
I see that the ByteDance/TimeSearch-R model is already available on the Hugging Face Hub! That's fantastic. To enhance its discoverability further, we can link it directly to your paper page (read here). This helps users find the model directly from your research paper.
Uploading dataset
Your paper mentions constructing "datasets specifically designed for the SFT cold-start and RL training of GRPO-CSV", and the README refers to "preprocessed JSON files for Haystack-LVBench". It would be awesome to make these newly constructed datasets available on 🤗 , so that people can do:
from datasets import load_dataset
dataset = load_dataset("your-hf-org-or-username/your-dataset")See here for a guide: https://huggingface.co/docs/datasets/loading.
Besides that, there's the dataset viewer which allows people to quickly explore the first few rows of the data in the browser.
Let me know if you're interested/need any help regarding this!
Cheers,
Niels
ML Engineer @ HF 🤗