Modelpublic

YONKWd/Spexcon-S1-v0.3-Experimental

sourceHugging Faceapache-2.0updated 28d agoView on Hugging Face

0likes65downloads

Model Card

Spexcon S1 v0.3 Experimental Adapter

This repository contains an experimental QLoRA smoke-test adapter for Spexcon S1. It is not the full Spexcon S1 release and should not be treated as a quality-improved production model.

—Base model: Qwen/Qwen3-1.7B-Base
—Source project: YONKWd/Spexcon-S1
—Dataset: HuggingFaceH4/ultrachat_200k
—Dataset license: mit
—GPU: Tesla T4
—Initial training loss: 1.796299934387207
—Final training loss: 1.796299934387207
—Evaluation report: evaluation/v0.3-smoke-test.md

Load with PEFT:

python

from peft import PeftModel
from transformers import AutoModelForCausalLM, AutoTokenizer

base_model = "Qwen/Qwen3-1.7B-Base"
adapter = "YONKWd/Spexcon-S1-v0.3-Experimental"

tokenizer = AutoTokenizer.from_pretrained(base_model)
model = AutoModelForCausalLM.from_pretrained(base_model, device_map="auto")
model = PeftModel.from_pretrained(model, adapter)