Image-Text-to-Text
Transformers
Safetensors
Korean
English
qwen3_5
korean
reasoning
multimodal
mix
conversational
Instructions to use ginigen-ai/Rogue-28B-MIX with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use ginigen-ai/Rogue-28B-MIX with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("image-text-to-text", model="ginigen-ai/Rogue-28B-MIX") messages = [ { "role": "user", "content": [ {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"}, {"type": "text", "text": "What animal is on the candy?"} ] }, ] pipe(text=messages)# Load model directly from transformers import AutoProcessor, AutoModelForImageTextToText processor = AutoProcessor.from_pretrained("ginigen-ai/Rogue-28B-MIX") model = AutoModelForImageTextToText.from_pretrained("ginigen-ai/Rogue-28B-MIX") messages = [ { "role": "user", "content": [ {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"}, {"type": "text", "text": "What animal is on the candy?"} ] }, ] inputs = processor.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=40) print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps
- vLLM
How to use ginigen-ai/Rogue-28B-MIX with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "ginigen-ai/Rogue-28B-MIX" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "ginigen-ai/Rogue-28B-MIX", "messages": [ { "role": "user", "content": [ { "type": "text", "text": "Describe this image in one sentence." }, { "type": "image_url", "image_url": { "url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg" } } ] } ] }'Use Docker
docker model run hf.co/ginigen-ai/Rogue-28B-MIX
- SGLang
How to use ginigen-ai/Rogue-28B-MIX with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "ginigen-ai/Rogue-28B-MIX" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "ginigen-ai/Rogue-28B-MIX", "messages": [ { "role": "user", "content": [ { "type": "text", "text": "Describe this image in one sentence." }, { "type": "image_url", "image_url": { "url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg" } } ] } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "ginigen-ai/Rogue-28B-MIX" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "ginigen-ai/Rogue-28B-MIX", "messages": [ { "role": "user", "content": [ { "type": "text", "text": "Describe this image in one sentence." }, { "type": "image_url", "image_url": { "url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg" } } ] } ] }' - Docker Model Runner
How to use ginigen-ai/Rogue-28B-MIX with Docker Model Runner:
docker model run hf.co/ginigen-ai/Rogue-28B-MIX
update lineage: 5-gen tree, condensed training summary
Browse files
README.md
CHANGED
|
@@ -11,7 +11,6 @@ tags:
|
|
| 11 |
- qwen3.5
|
| 12 |
- mix
|
| 13 |
- merge
|
| 14 |
-
- darwin-family
|
| 15 |
base_model:
|
| 16 |
- FINAL-Bench/Darwin-28B-Opus
|
| 17 |
- NewenAI/QuettaLLMs-27B-Koreasoner-V3
|
|
@@ -19,115 +18,106 @@ base_model:
|
|
| 19 |
|
| 20 |
# Rogue-28B-MIX
|
| 21 |
|
| 22 |
-
한국어 reasoning + multimodal mix 모델
|
| 23 |
|
| 24 |
---
|
| 25 |
|
| 26 |
## 🏛️ 가문 계보 (Lineage / 족보)
|
| 27 |
|
| 28 |
```
|
| 29 |
-
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
|
| 34 |
-
|
| 35 |
-
|
| 36 |
-
|
| 37 |
-
|
| 38 |
-
|
| 39 |
-
|
| 40 |
-
|
| 41 |
-
|
| 42 |
-
|
| 43 |
-
|
| 44 |
-
|
| 45 |
-
|
| 46 |
-
|
| 47 |
-
|
| 48 |
-
|
| 49 |
-
|
| 50 |
-
|
| 51 |
-
|
| 52 |
-
|
| 53 |
-
|
| 54 |
-
|
| 55 |
-
|
| 56 |
-
|
| 57 |
-
|
| 58 |
-
|
| 59 |
-
|
| 60 |
-
|
| 61 |
-
|
| 62 |
-
|
| 63 |
-
|
| 64 |
-
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
|
| 73 |
-
|
| 74 |
-
|
| 75 |
-
|
| 76 |
-
|
| 77 |
-
|
| 78 |
-
|
| 79 |
-
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
-
|
| 88 |
-
|
| 89 |
-
▼
|
| 90 |
-
╔══════════════════════════════════════════════════════════════════════════════╗
|
| 91 |
-
║ 자식 (Child) ║
|
| 92 |
-
║ ║
|
| 93 |
-
║ ginigen-ai/Rogue-28B-MIX (Apache 2.0, Private) ║
|
| 94 |
-
║ ║
|
| 95 |
-
║ ✦ 조상의 능력 모두 계승 ║
|
| 96 |
-
║ - 친가 → 추론·생각 트레이스 (Darwin-28B-Opus) ║
|
| 97 |
-
║ - 외가 → 한국어·K-AI 도메인 지식 (Quetta) ║
|
| 98 |
-
║ ║
|
| 99 |
-
║ ✦ Smart MRI Streaming Merge ║
|
| 100 |
-
║ - Embed/LM-head/Norm : 100% 엄마 (한국어 표현 통로) ║
|
| 101 |
-
║ - MLP : 70% 엄마 / 30% 아빠 (지식+추론) ║
|
| 102 |
-
║ - Self-attn : 50% / 50% (균형) ║
|
| 103 |
-
║ - Visual : 100% 아빠 (멀티모달 헤드) ║
|
| 104 |
-
║ ║
|
| 105 |
-
║ ✦ 후속 학습 ║
|
| 106 |
-
║ - 1500 K-AI focused samples × 4 epochs (TRL v1.3 SFT) ║
|
| 107 |
-
║ - LoRA r=64, alpha=128, all-linear ║
|
| 108 |
-
║ ║
|
| 109 |
-
╚══════════════════════════════════════════════════════════════════════════════╝
|
| 110 |
```
|
| 111 |
|
| 112 |
---
|
| 113 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 114 |
## 📊 평가 (한국어 공개 10 데이터셋, 100문제 × 1 seed, temperature=0)
|
| 115 |
|
| 116 |
-
| Dataset |
|
| 117 |
-
|---|---:|---:|
|
| 118 |
-
| CLIcK |
|
| 119 |
-
| KMMLU History | **48%** | 45% |
|
| 120 |
-
| KMMLU Law | 25% | 26% |
|
| 121 |
-
| KMMLU Health | **81%** | 80% |
|
| 122 |
-
| HAERAE General | 63% | 66% |
|
| 123 |
-
| HAERAE History | 89% | 90% |
|
| 124 |
-
| HAERAE Linguistics | 90% | 95% |
|
| 125 |
-
| KoBEST Hellaswag | 95% | 97% |
|
| 126 |
-
| KoBEST COPA | 98% | 99% |
|
| 127 |
-
| KoBEST BoolQ | 97% | 97% |
|
| 128 |
-
| **Macro Avg** | **77.0%** | **78.0%** |
|
| 129 |
-
|
| 130 |
-
|
| 131 |
|
| 132 |
---
|
| 133 |
|
|
@@ -153,40 +143,23 @@ out = model.generate(inputs.to(model.device), max_new_tokens=512, do_sample=Fals
|
|
| 153 |
print(tokenizer.decode(out[0], skip_special_tokens=False))
|
| 154 |
```
|
| 155 |
|
| 156 |
-
`<think>` 트레이스로 단계별 추론을 보여줍니다:
|
| 157 |
-
|
| 158 |
-
```
|
| 159 |
-
<think>
|
| 160 |
-
사용자가 한국의 추석에 대해 묻고 있다. 추석은 한국의 가장 큰 명절 중 하나로...
|
| 161 |
-
</think>
|
| 162 |
-
|
| 163 |
-
추석은 음력 8월 15일에 지내는 한국의 대표적인 명절로, ...
|
| 164 |
-
```
|
| 165 |
-
|
| 166 |
---
|
| 167 |
|
| 168 |
## 🛠️ 모델 사양
|
| 169 |
|
| 170 |
-
- **파라미터**: 28B (multimodal Qwen 3.5
|
| 171 |
- **양자화**: bf16
|
| 172 |
-
- **컨텍스트**: 8K (
|
| 173 |
- **언어**: 한국어 + 영어
|
| 174 |
-
- **모달리티**: 텍스트 (이미지·비디오 헤드 보존, vLLM 호환)
|
| 175 |
- **추론**: `<think>` reasoning trace
|
| 176 |
- **License**: Apache 2.0
|
| 177 |
|
| 178 |
---
|
| 179 |
|
| 180 |
-
## 📜 라이선스
|
| 181 |
-
|
| 182 |
-
Apache 2.0 — 상업적 사용 가능. 두 부모 모두 Apache 2.0.
|
| 183 |
-
|
| 184 |
-
---
|
| 185 |
-
|
| 186 |
## 🤝 출처
|
| 187 |
|
| 188 |
-
-
|
| 189 |
- 엄마: https://huggingface.co/NewenAI/QuettaLLMs-27B-Koreasoner-V3
|
| 190 |
-
-
|
| 191 |
-
-
|
| 192 |
-
-
|
|
|
|
| 11 |
- qwen3.5
|
| 12 |
- mix
|
| 13 |
- merge
|
|
|
|
| 14 |
base_model:
|
| 15 |
- FINAL-Bench/Darwin-28B-Opus
|
| 16 |
- NewenAI/QuettaLLMs-27B-Koreasoner-V3
|
|
|
|
| 18 |
|
| 19 |
# Rogue-28B-MIX
|
| 20 |
|
| 21 |
+
한국어 reasoning + multimodal mix 모델.
|
| 22 |
|
| 23 |
---
|
| 24 |
|
| 25 |
## 🏛️ 가문 계보 (Lineage / 족보)
|
| 26 |
|
| 27 |
```
|
| 28 |
+
═══════════════════════════════════════════════════════════════════════════
|
| 29 |
+
증조부 / 증조모 (Great-Grandparents)
|
| 30 |
+
═══════════════════════════════════════════════════════════════════════════
|
| 31 |
+
|
| 32 |
+
┌─────────────────────────┐ ┌─────────────────────────┐
|
| 33 |
+
│ unsloth/Qwen3.5-27B-VL │ │ Anthropic Claude Opus │
|
| 34 |
+
│ - 28B 멀티모달 베이스 │ │ - reasoning 증류 출처 │
|
| 35 |
+
│ - vocab 248K │ ��� - <think> 패턴 │
|
| 36 |
+
└─────────────────────────┘ └─────────────────────────┘
|
| 37 |
+
│ │
|
| 38 |
+
└────────────────┬────────────────┘
|
| 39 |
+
▼
|
| 40 |
+
|
| 41 |
+
═══════════════════════════════════════════════════════════════════════════
|
| 42 |
+
조부 / 조모 (Grandparents)
|
| 43 |
+
═══════════════════════════════════════════════════════════════════════════
|
| 44 |
+
|
| 45 |
+
┌─────────────────────────────────────────────────────────────┐
|
| 46 |
+
│ FINAL-Bench/Darwin-27B-Opus / Darwin family │
|
| 47 |
+
│ │
|
| 48 |
+
│ - Qwen3.5 + Claude Opus reasoning 진화 머지 │
|
| 49 |
+
│ - 양가(친가·외가) 모두의 공통 시조 │
|
| 50 |
+
└─────────────────────────────────────────────────────────────┘
|
| 51 |
+
│
|
| 52 |
+
┌────────────────┴────────────────┐
|
| 53 |
+
▼ ▼
|
| 54 |
+
|
| 55 |
+
═══════════════════════════════════════════════════════════════════════════
|
| 56 |
+
부모 (Parents)
|
| 57 |
+
═══════════════════════════════════════════════════════════════════════════
|
| 58 |
+
|
| 59 |
+
┌─────────────────────────┐ ┌─────────────────────────┐
|
| 60 |
+
│ 아빠 (Father) │ │ 엄마 (Mother) │
|
| 61 |
+
│ │ │ │
|
| 62 |
+
│ FINAL-Bench/ │ ×× │ NewenAI/ │
|
| 63 |
+
│ Darwin-28B-Opus │ 교배 │ QuettaLLMs-27B- │
|
| 64 |
+
│ │ ▼ │ Koreasoner-V3 │
|
| 65 |
+
│ - reasoning 양수겸장 │ │ │
|
| 66 |
+
│ - <think> trace │ │ - K-AI Leaderboard │
|
| 67 |
+
│ - GPQA 88.4% │ │ 1위 (avg 0.560) │
|
| 68 |
+
│ - Apache 2.0 │ │ - 한국어 SOTA │
|
| 69 |
+
│ │ │ - Apache 2.0 │
|
| 70 |
+
└─────────────────────────┘ └─────────────────────────┘
|
| 71 |
+
│ │
|
| 72 |
+
└────────────────┬────────────────┘
|
| 73 |
+
▼
|
| 74 |
+
|
| 75 |
+
═══════════════════════════════════════════════════════════════════════════
|
| 76 |
+
자식 (Child / 본 모델)
|
| 77 |
+
═══════════════════════════════════════════════════════════════════════════
|
| 78 |
+
|
| 79 |
+
╔═══════════════════════════════════════════════════════╗
|
| 80 |
+
║ ginigen-ai/Rogue-28B-MIX ║
|
| 81 |
+
║ ║
|
| 82 |
+
║ - 친가의 reasoning + 외가의 한국어 K-AI 지식 결합 ║
|
| 83 |
+
║ - <think> 추론 트레이스 보존 ║
|
| 84 |
+
║ - 멀티모달 헤드 보존 (text/image/video) ║
|
| 85 |
+
║ - K-AI Leaderboard 제출 호환 형식 ║
|
| 86 |
+
║ ║
|
| 87 |
+
╚════════════════════════════════════════════════════���══╝
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 88 |
```
|
| 89 |
|
| 90 |
---
|
| 91 |
|
| 92 |
+
## 🎓 학습 개요
|
| 93 |
+
|
| 94 |
+
| 단계 | 개략 설명 |
|
| 95 |
+
|:---|:---|
|
| 96 |
+
| **1. 부 × 모 머지** | 친아빠와 엄마의 가중치를 모듈별 비율로 결합 (Smart MRI 방식) |
|
| 97 |
+
| **2. 추가 학습** | K-AI 도메인 데이터로 다단계 SFT 정제 |
|
| 98 |
+
| **3. 호환성 정리** | K-AI Leaderboard Docker 호환 형식으로 토크나이저·config 정비 |
|
| 99 |
+
|
| 100 |
+
세부 파라미터는 비공개.
|
| 101 |
+
|
| 102 |
+
---
|
| 103 |
+
|
| 104 |
## 📊 평가 (한국어 공개 10 데이터셋, 100문제 × 1 seed, temperature=0)
|
| 105 |
|
| 106 |
+
| Dataset | Rogue-28B-MIX | Quetta-V3 |
|
| 107 |
+
|:---|---:|---:|
|
| 108 |
+
| CLIcK | 84% | 85% |
|
| 109 |
+
| KMMLU History | **48%** 🏆 | 45% |
|
| 110 |
+
| KMMLU Law | 25% | 26% |
|
| 111 |
+
| KMMLU Health | **81%** 🏆 | 80% |
|
| 112 |
+
| HAERAE General | 63% | 66% |
|
| 113 |
+
| HAERAE History | 89% | 90% |
|
| 114 |
+
| HAERAE Linguistics | 90% | 95% |
|
| 115 |
+
| KoBEST Hellaswag | 95% | 97% |
|
| 116 |
+
| KoBEST COPA | 98% | 99% |
|
| 117 |
+
| KoBEST BoolQ | 97% | 97% |
|
| 118 |
+
| **Macro Avg** | **77.0%** | **78.0%** |
|
| 119 |
+
|
| 120 |
+
K-AI Leaderboard 핵심 카테고리(의료·역사)에서 엄마(Quetta) 추월.
|
| 121 |
|
| 122 |
---
|
| 123 |
|
|
|
|
| 143 |
print(tokenizer.decode(out[0], skip_special_tokens=False))
|
| 144 |
```
|
| 145 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 146 |
---
|
| 147 |
|
| 148 |
## 🛠️ 모델 사양
|
| 149 |
|
| 150 |
+
- **파라미터**: 28B (multimodal Qwen 3.5 family)
|
| 151 |
- **양자화**: bf16
|
| 152 |
+
- **컨텍스트**: 8K (확장 가능)
|
| 153 |
- **언어**: 한국어 + 영어
|
|
|
|
| 154 |
- **추론**: `<think>` reasoning trace
|
| 155 |
- **License**: Apache 2.0
|
| 156 |
|
| 157 |
---
|
| 158 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 159 |
## 🤝 출처
|
| 160 |
|
| 161 |
+
- 아빠: https://huggingface.co/FINAL-Bench/Darwin-28B-Opus
|
| 162 |
- 엄마: https://huggingface.co/NewenAI/QuettaLLMs-27B-Koreasoner-V3
|
| 163 |
+
- 조부 (양가 공통): Darwin family (Darwin-27B-Opus 계보)
|
| 164 |
+
- 증조부: [unsloth Qwen 3.5 멀티모달](https://huggingface.co/unsloth)
|
| 165 |
+
- 증조모: Anthropic Claude Opus (reasoning 증류)
|