Modelpublic

K1mG0ng/AI-taste-EcoOB-30B-A3B

sourceHugging Facemitupdated 6mo agoView on Hugging Face

0likes12downloads

Model Card

AI-Taste-EcoOB-30B-A3B

This repository provides a pooled fine-tuned Qwen3 30B-A3B model for AI Taste experiments on social science research articles, combining economics and organizational behavior style training data.

Model Summary

—Base model: Qwen/Qwen3-30B-A3B
—Architecture: Qwen3MoeForCausalLM
—Format: Hugging Face Transformers + Safetensors
—Primary use: ranking or classifying article-evaluation prompts into discrete quality levels

Intended Use

The model is intended for research and internal experimentation on structured article-evaluation prompts, including:

—research question quality assessment
—article-level prompt scoring
—pooled social science taste benchmarking

It is not intended as a general-purpose factual assistant or as a substitute for expert review.

Files

This repository keeps only the files needed to load and run the model:

—model weights in safetensors
—model config and generation config
—tokenizer-related config files present in this export

Example

python

from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "K1mG0ng/AI-taste-EcoOB-30B-A3B"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

Notes

—This is a fine-tuned research model.
—Output quality depends on prompt format and label schema used at inference time.
—Associated open weights and datasets are organized through the AI Taste scientific collection on Hugging Face.