datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
cypher-sft-v4-sources
cypher-sft-v4-sources
AETHER family SFT dataset — group cypher_core.
Format: JSONL ChatML messages, task_type tagged, MinHash dedup applied (threshold 0.85).
Schema:
{
"messages": [{"role": "system|user|assistant", "content": "..."}],
"task_type": "function_calling|code|reasoning_cot|...",
"source_ds": "<HF dataset_id>",
"lang": "en|fr|...",
"system_source": "archon_default|overridden_from_source"
}
Generated by prepare_sft.py pipeline (2026-05-25).
cypher-sft-v4-mix
cypher-sft-v4-mix
AETHER family SFT dataset — group mix.
Format: JSONL ChatML messages, task_type tagged, MinHash dedup applied (threshold 0.85).
Schema:
{
"messages": [{"role": "system|user|assistant", "content": "..."}],
"task_type": "function_calling|code|reasoning_cot|...",
"source_ds": "<HF dataset_id>",
"lang": "en|fr|...",
"system_source": "archon_default|overridden_from_source"
}
Generated by prepare_sft.py pipeline (2026-05-25).
PIPE-Cypher-benchmarks
PIPE-Cypher Benchmarks
This dataset release contains the public-proxy benchmark exports used by
PIPE-Cypher: Automatic Enterprise Benchmark Generation for Text-to-Cypher
Systems.
Paper: https://arxiv.org/abs/2606.08481
Repository: https://github.com/suraj-ranganath/PIPE-Cypher/
Dataset repo: https://huggingface.co/datasets/suraj-ranganath/PIPE-Cypher-benchmarks
Exports
finbench_snb_full_qwen9b
Total examples: 3000
Split counts: {"dev": 296, "test":… See the full description on the dataset page: https://huggingface.co/datasets/suraj-ranganath/PIPE-Cypher-benchmarks.