Qwen3-0.6B Float:Right Tagger (https://float-right.app)

This repository contains a fine-tuned tag generator based on Qwen/Qwen3-0.6B. This model was built for on-device AI tag generation in the Float:Right app. Float:Right is an automatic tag generation and classification app

non-GGUF : https://huggingface.co/FloatDo/qwen3-0.6b-float-right-tagger

이것은 Float:Right 앱에 μ‚¬μš©ν•  μ˜¨λ””λ°”μ΄μŠ€ AI νƒœκ·Έμƒμ„±μš©λ„λ‘œ λ§Œλ“€μ–΄μ‘ŒμŠ΅λ‹ˆλ‹€. μžλ™ νƒœκ·Έμƒμ„±, λΆ„λ₯˜μ•± Float:Right.

https://float-right.app

What it does

Given a memo/text, it returns a JSON array of 3–10 tags:

  • Prefer coarse tags (not overly detailed)
  • Keeps the same language as input (Korean -> Korean, English -> English)
  • Avoids underscores _

In production, parse only the first JSON array [ ... ] from the output.

Quick usage (Transformers)

import json, re, torch
from transformers import AutoTokenizer, AutoModelForCausalLM

MODEL_DIR = "./"  # or your HF repo id

tok = AutoTokenizer.from_pretrained(MODEL_DIR, trust_remote_code=True)
if tok.pad_token is None:
    tok.pad_token = tok.eos_token

model = AutoModelForCausalLM.from_pretrained(
    MODEL_DIR, torch_dtype="auto", device_map="cuda", trust_remote_code=True
)

def extract_array(s: str):
    m = re.search(r"\[[\s\S]*?\]", s)
    if not m:
        return None
    return json.loads(m.group(0))

text = "였늘 μ„œμšΈμ—μ„œ AI 컨퍼런슀λ₯Ό λ‹€λ…€μ™”λ‹€."
messages = [
    {"role": "system", "content": "λ„ˆλŠ” νƒœκ·Έ 생성기닀. 좜λ ₯은 JSON λ°°μ—΄ ν•˜λ‚˜λ§Œ."},
    {"role": "user", "content": f"λ¬Έμž₯: {text}\nνƒœκ·Έ 3~10개. λ„ˆλ¬΄ λ””ν…ŒμΌν•˜μ§€ μ•Šκ²Œ. μ–Έλ”μŠ€μ½”μ–΄ κΈˆμ§€. JSON λ°°μ—΄λ§Œ."},
]

prompt = tok.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
enc = tok(prompt, return_tensors="pt").to("cuda")

out = model.generate(**enc, max_new_tokens=64, do_sample=False)
decoded = tok.decode(out[0], skip_special_tokens=True)
print(extract_array(decoded))

Notes β€’ Some outputs may include extra tokens (e.g., ). In production, extract only the first JSON array [ ... ]. β€’ Training data is intended to avoid sensitive information.

Credits β€’ Base model: Qwen/Qwen3-0.6B β€’ Project: Float-Right

Downloads last month
32
GGUF
Model size
0.8B params
Architecture
qwen3
Hardware compatibility
Log In to add your hardware

4-bit

Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support