Datasetpublic

SonarSource/hackathon-dolly-512

-- license: cc-by-sa-3.0 language: [en] task_categories: [text-generation] configs: - config_name: default data_files: - split: train path: train.jsonl - split: validation path: valid.jsonl - split: test path: test.jsonl hackathon-dolly-512 Subset of databricks/databricks-dolly-15k in chat messages format, filtered to ~512 tokens, for LoRA fine-tuning experiments.

sourceHugging Faceupdated 4d agoView on Hugging Face

0likes44downloads

Dataset Card

-- license: cc-by-sa-3.0 language: [en] task_categories: [text-generation] configs:

—configname: default datafiles:
—split: train path: train.jsonl
—split: validation path: valid.jsonl
—split: test path: test.jsonl ---

hackathon-dolly-512

Subset of databricks/databricks-dolly-15k in chat messages format, filtered to ~512 tokens, for LoRA fine-tuning experiments.