instruction stringclasses 5
values | input stringclasses 4
values | output stringclasses 5
values |
|---|---|---|
How does every generated answer get checked against the exact source chunk? | CuratorKIT builds trainer-ready datasets for LLM post-training. Every generated answer is checked against the exact source chunk it came from. Rejected samples are written to rejected.jsonl with a structured reason, never dropped silently. A run also writes manifest.json, dataset_card.md, and SHA-256 checksums.txt. | Every generated answer is checked against the exact source chunk it came from. |
How does the 'PIIPseudonymizer' replace names and emails with consistent fake values? | Hygiene stages run before any LLM call. SecretsGate rejects API keys and private tokens. PIIPseudonymizer replaces names and emails with consistent fake values. ToxicityGate uses a local Detoxify classifier, with an optional LLM judge for borderline rows. | The 'PIIPseudonymizer' replaces names and emails with consistent fake values to prevent unauthorized access. |
Which component of the 'ToxicityGate' system involves using a local Detoxify classifier with an optional LLM judge? | Hygiene stages run before any LLM call. SecretsGate rejects API keys and private tokens. PIIPseudonymizer replaces names and emails with consistent fake values. ToxicityGate uses a local Detoxify classifier, with an optional LLM judge for borderline rows. | The 'ToxicityGate' system uses a local Detoxify classifier to detect potential malicious activity and optionally includes an LLM judge for more accurate detection. |
What does generate_answers aim to achieve with its output? | Evol-Instruct rewrites a seed instruction into a harder variant. num_evolutions is how many rewrite steps run. generate_answers asks the model to solve the evolved prompt so the row is SFT-ready. | generate_answers is designed to take the output from Evol-Instruct and use it to improve the quality of answers by rewriting them based on the learned patterns or rules within the seed instruction. This process aims to enhance the overall effectiveness and relevance of the answers. |
What is the purpose of using StratifiedSampler? | Ingest can take several Hub datasets, each with its own preprocessing_fn and max_samples. StratifiedSampler then matches target_distribution on source_dataset so one corpus does not drown the mix. | StratifiedSampler is used to balance the distribution of samples across different sources in a dataset, ensuring that each sample in the mixture comes from a similar proportion. |
|
curatorkit-testrun-Filtered-FT
Built using CuratorKIT — provenance-grounded curation and synthesis for LLM post-training.
| Method | qa |
| Backend | litellm |
| Model | openai/Qwen/Qwen2.5-0.5B-Instruct |
| Formats | alpaca |
| Artifact | dataset |
| Published | 2026-09-01 05:49 UTC |
Usage
from datasets import load_dataset
ds = load_dataset("ram-lexsi/curatorkit-testrun-Filtered-FT", "alpaca")
- Downloads last month
- 64