Dataset Viewer
Auto-converted to Parquet Duplicate
Search is not available for this dataset
pdf
pdf

The GUMI182-RHN Scaling Law: A Critique for Overtrainers

by the way you can call it GUMI182 for short

and GUMI means Generalized Unified Model Intelligence because we should label LLM Datasets

Scrambled Problem

We know you all like to use Meta LLaMa or Alibaba Qwen but they are currently overtraining in terms of Tokens leading to Waste of computation, large hallucinations, or chances of underperforming because they take datas from human conversation incorrectly, such as training in 36 Trillion and 15 Trillion tokens and i suposse to made something more efficient

The GUMI182-RHN Ratio

For an optimal saturation, the ratio occurs on 500-750 (950 if needed) T/P (Tokens Per Paramater) On an 8B, that would aim to 4-6T Total Tokens or 7T if you use the max reccomended rate

Talkability

To make an LLM Has a human common sense and dont overtrain, i reccomend

90% Logic and 10% Human data to be logical but works for agentic AIs

80% Logic and 20% Human data for a sweet spot for both agentic and basic use

75% Logic and 25% Human data might hallucinate (by data) but is a very talkative mathman and a sweet spot

60% Logic and 40% Human data would hallucinate (by data) but very talkative

20% Logic and 80% Human data very good spot if you want (kind of) the best roleplaying model


HOW TO CITE (CC-BY-4.0)

Creator: Unokaiysh182/Muh. Raihan al m.

Paper: GUMI182-RHN Scaling Law Format

About: Scaling Ratios between the Optimal saturation and Valuing datasets

between Human nuance and logic.


Date: 2026 Feb 19 Creator: Unokaiysh182


Shout out to Indonesia because its my country and im 13

Downloads last month
16