qwen3.5-4b-unslop-good-lora-v1

Unslop rewrite adapter focused on reducing hype/corporate phrasing while preserving meaning.

Model summary

  • Repo: Oysiyl/qwen3.5-4b-unslop-good-lora-v1
  • Base model: Qwen/Qwen3.5-4B
  • Adapter type: LoRA
  • Pipeline: text generation / rewrite style transfer
  • Current downloads (snapshot): 0

Intended use

  • Rewrite AI-sounding drafts into cleaner, more natural prose.
  • Keep meaning and key facts intact.
  • Use as a post-processing layer for longform and social text cleanup.

Limitations

  • Can still over-rewrite some passages.
  • Not guaranteed to improve factual accuracy.
  • Should be human-reviewed for fidelity-sensitive outputs.

Evaluation notes

This card records this model as part of the Unslop family with a common quality goal: preserve meaning, reduce hype, and avoid hallucinated additions.

Usage (PEFT)

from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel

base = "Qwen/Qwen3.5-4B"
adapter = "Oysiyl/qwen3.5-4b-unslop-good-lora-v1"

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

Training loss vs training progress

Training loss vs progress

Loss curve included as optimization check; qualitative rewrite fidelity remains the decision metric.

Recommended evaluation sample

Use this full fiction passage for held-out testing:

The wind rose all at once, as if the mountain itself had inhaled. She stood at the edge of the pass, cloak snapping like a torn banner behind her, staring down at the valley where the last lights were going out one by one. Each flame extinguished felt deliberate—an answer to a question she had been too afraid to ask. Somewhere below, bells began to ring, not in warning, but in mourning. They already knew. “Don’t,” he said, breathless, reaching for her arm. His fingers brushed fabric, then air. “If you cross that line, there’s no coming back.” She didn’t turn. The cold burned her lungs, sharpened her thoughts. “There was never a way back,” she replied softly. “Only a way forward you were willing to pretend didn’t exist.” Thunder rolled, low and ancient. For a moment, the sky fractured into white fire, and in that instant he saw her face—resolved, terrified, incandescent with purpose. Not the girl he had sworn to protect, but the woman the world had demanded she become. “You’ll die,” he said. She smiled then, finally facing him, eyes bright with unshed tears. “So will everyone,” she answered. “The difference is what we leave burning when we’re gone.” She stepped forward. The mountain exhaled.

Full observed output on that sample

The wind surged suddenly, as though the mountain had drawn a breath. At the edge of the pass, her cloak snapped behind her like a tattered flag. Below, the last lights of the valley flickered out one by one; each extinguished flame felt intentional, an answer to a question she hadn't dared to voice. Then came the bells, tolling not for alarm, but for grief. They knew.

"Don't," he gasped, lunging for her arm. His fingers caught cloth, then empty air. "If you cross that line, you can't come back."

She didn't look at him. The freezing air seared her lungs, clearing her mind. "There was never a way back," she said quietly. "Just a path forward you chose to ignore."

A low, ancient thunder rumbled. A flash of white fire split the sky, illuminating her face for a heartbeat—determined, frightened, alight with resolve. This wasn't the girl he had promised to shield, but the woman the world had forced her to be.

"You'll die," he said.

She turned to him then, smiling through eyes shimmering with unshed tears. "We all will," she replied. "The only thing that matters is what we leave burning."

She stepped forward. The mountain breathed out.

Downloads last month

-

Downloads are not tracked for this model. How to track
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for Oysiyl/qwen3.5-4b-unslop-good-lora-v1

Finetuned
Qwen/Qwen3.5-4B
Adapter
(102)
this model