AmareshHebbar/kp_system-sft

KP System (Krishnamurti Paddhati) SFT Dataset

Sidereal Vedic chart plus KP's distinguishing sub-lord layer — each planet and the Ascendant get a sub-lord via proportional nakshatra subdivision.

Total examples	100000
Train / Validation / Test	89916 / 5104 / 4980
Question types	4
Avg citations per example	4.097
Zero-citation examples	0

Computation

Deterministic astronomy (pyswisseph, Lahiri sidereal) + a proportional-subdivision sub-lord formula.

Validation status — read before trusting this at scale

Sub-lord formula independently checked for internal consistency: the 9 proportional spans sum to exactly one nakshatra (13°20') with no gaps or overlaps, and produce exactly the 9 expected lords in the correct sequence across a fine-grained scan (see companion repo tests/test_kp_sublord.py).

Schema

Field	Type	Description
example_id	string	{synthetic_id}_{question_type}
synthetic_id	string	ID of the underlying synthetic input
system	string	"kp_system"
facts	dict	Full computed output for this system (chart/pillars/numbers depending on system)
retrieved_rules	list[dict]	Rules/observations matched for this input
question_type	string	See distribution below
user_question	string	Synthetic user question
response	string	Grounded reading — every claim traces to facts or retrieved_rules
citations	list[string]	Which rules the response draws on
source	string	"rule_matched" for all rows in this version

Question type distribution: general_natal: 25000, career: 25000, marriage_timing: 25000, sub_lord_reading: 25000

Usage

from datasets import load_dataset

ds = load_dataset("YOUR_USERNAME/kp_system-sft")
print(ds)

ex = ds["train"][0]
print(ex["user_question"])
print(ex["response"])

What this dataset does NOT claim

Like every dataset in this collection, this teaches a model to narrate a correctly-computed KP System (Krishnamurti Paddhati) reading using its own traditional rules faithfully and with citations. It does not, and cannot, contain any ground truth about whether any prediction comes true — no such data exists for any divination system. See the companion repo's top-level README for the shared "what these datasets don't claim" statement that applies across the whole collection.

Known limitations

• Uses Lahiri ayanamsa, not the KP (Krishnamurti) ayanamsa some strict KP practitioners prefer — a documented simplification, not a KP-specific correction.
• Only 1 level of sub-lord is computed (not the further sub-sub-lord some KP practice uses).
• No house-cusp sub-lords (KP's horary/timing technique often works from cusp sub-lords, not just planet sub-lords) — only planet and Ascendant sub-lords in this version.

Companion repo

Full generation pipeline (8 systems, shared scripts) — for regenerating at larger scale or extending the rule set — is the companion GitHub repo astrology-datasets (update this link once pushed).

License

Apache 2.0. All computed data is synthetic (no real people). Rule/effect text is written in-house, paraphrasing widely known traditional principles, not reproduced from any single copyrighted source.