Text Generation
Transformers
Safetensors
Russian
English
qwen3_omni_moe
text-to-audio
radon
ultra
agi
russian
english
conversation
conversational
custom_code
Instructions to use MagistrTheOne/Radon-35B-Ultra-X-RU with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use MagistrTheOne/Radon-35B-Ultra-X-RU with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="MagistrTheOne/Radon-35B-Ultra-X-RU", trust_remote_code=True) messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# Load model directly from transformers import AutoProcessor, AutoModelForMultimodalLM processor = AutoProcessor.from_pretrained("MagistrTheOne/Radon-35B-Ultra-X-RU", trust_remote_code=True) model = AutoModelForMultimodalLM.from_pretrained("MagistrTheOne/Radon-35B-Ultra-X-RU", trust_remote_code=True, device_map="auto") messages = [ {"role": "user", "content": "Who are you?"}, ] 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 Settings
- vLLM
How to use MagistrTheOne/Radon-35B-Ultra-X-RU with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "MagistrTheOne/Radon-35B-Ultra-X-RU" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "MagistrTheOne/Radon-35B-Ultra-X-RU", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/MagistrTheOne/Radon-35B-Ultra-X-RU
- SGLang
How to use MagistrTheOne/Radon-35B-Ultra-X-RU 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 "MagistrTheOne/Radon-35B-Ultra-X-RU" \ --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": "MagistrTheOne/Radon-35B-Ultra-X-RU", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'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 "MagistrTheOne/Radon-35B-Ultra-X-RU" \ --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": "MagistrTheOne/Radon-35B-Ultra-X-RU", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use MagistrTheOne/Radon-35B-Ultra-X-RU with Docker Model Runner:
docker model run hf.co/MagistrTheOne/Radon-35B-Ultra-X-RU
Update README: proper YAML metadata, remove Qwen references, add RU/EN documentation
Browse files
README.md
CHANGED
|
@@ -1,37 +1,73 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
# Radon-35B-Ultra-X-RU
|
| 2 |
|
| 3 |
-
|
| 4 |
-
|
| 5 |
-
|
| 6 |
-
|
| 7 |
-
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
|
| 11 |
-
|
| 12 |
-
|
| 13 |
-
|
| 14 |
-
|
| 15 |
-
|
| 16 |
-
|
| 17 |
-
|
| 18 |
-
-
|
| 19 |
-
|
| 20 |
-
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
|
| 24 |
-
|
| 25 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 26 |
```python
|
| 27 |
import torch
|
| 28 |
-
from transformers import
|
| 29 |
|
| 30 |
model_name = "MagistrTheOne/Radon-35B-Ultra-X-RU"
|
| 31 |
|
| 32 |
-
# Загрузка модели
|
| 33 |
-
processor =
|
| 34 |
-
model =
|
| 35 |
model_name,
|
| 36 |
torch_dtype=torch.bfloat16,
|
| 37 |
device_map="auto",
|
|
@@ -39,7 +75,7 @@ model = Qwen3OmniMoeForConditionalGeneration.from_pretrained(
|
|
| 39 |
)
|
| 40 |
model.eval()
|
| 41 |
|
| 42 |
-
# Пример использования
|
| 43 |
conversation = [
|
| 44 |
{"role": "system", "content": [{"type": "text", "text": "Ты — Radon Ultra, мощный AI-ассистент. Отвечай кратко и по делу."}]},
|
| 45 |
{"role": "user", "content": [{"type": "text", "text": "Кто ты?"}]}
|
|
@@ -63,40 +99,57 @@ response = processor.batch_decode(outputs[:, inputs["input_ids"].shape[1]:], ski
|
|
| 63 |
print(response)
|
| 64 |
```
|
| 65 |
|
| 66 |
-
###
|
| 67 |
-
```python
|
| 68 |
-
import requests
|
| 69 |
-
|
| 70 |
-
response = requests.post(
|
| 71 |
-
"http://213.219.215.235/chat",
|
| 72 |
-
json={
|
| 73 |
-
"message": "Кто ты?",
|
| 74 |
-
"max_tokens": 100,
|
| 75 |
-
"temperature": 0.7
|
| 76 |
-
}
|
| 77 |
-
)
|
| 78 |
|
| 79 |
-
|
| 80 |
```
|
|
|
|
|
|
|
| 81 |
|
| 82 |
-
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
Датасет: 1000 синтетических примеров (300 identity + 700 mixed).
|
| 86 |
|
| 87 |
-
|
| 88 |
-
|
| 89 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 90 |
- **Attention**: Flash Attention 2
|
| 91 |
-
- **
|
| 92 |
-
- **
|
| 93 |
-
- **
|
| 94 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 95 |
|
| 96 |
-
## Лицензия
|
| 97 |
Apache 2.0
|
| 98 |
|
| 99 |
-
##
|
| 100 |
-
|
| 101 |
-
-
|
| 102 |
-
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
language:
|
| 4 |
+
- ru
|
| 5 |
+
- en
|
| 6 |
+
tags:
|
| 7 |
+
- radon
|
| 8 |
+
- ultra
|
| 9 |
+
- agi
|
| 10 |
+
- russian
|
| 11 |
+
- english
|
| 12 |
+
- text-generation
|
| 13 |
+
- conversation
|
| 14 |
+
library_name: transformers
|
| 15 |
+
pipeline_tag: text-generation
|
| 16 |
+
base_model: custom
|
| 17 |
+
model_creator: MagistrTheOne
|
| 18 |
+
model_type: radon_ultra_moe
|
| 19 |
+
size_categories:
|
| 20 |
+
- 30B
|
| 21 |
+
- 35B
|
| 22 |
+
---
|
| 23 |
+
|
| 24 |
# Radon-35B-Ultra-X-RU
|
| 25 |
|
| 26 |
+
<div align="center">
|
| 27 |
+
<img src="https://img.shields.io/badge/Radon-Ultra-blue?style=for-the-badge&logo=robot" alt="Radon Ultra"/>
|
| 28 |
+
<img src="https://img.shields.io/badge/Parameters-35B-green?style=for-the-badge" alt="35B Parameters"/>
|
| 29 |
+
<img src="https://img.shields.io/badge/Language-RU%20%7C%20EN-orange?style=for-the-badge" alt="RU/EN"/>
|
| 30 |
+
</div>
|
| 31 |
+
|
| 32 |
+
## 🌟 Описание / Description
|
| 33 |
+
|
| 34 |
+
**Русский:** Radon Ultra - это продвинутый AGI-ассистент нового поколения, созданный в России. Модель обладает уникальной идентичностью и способна вести естественные диалоги на русском и английском языках.
|
| 35 |
+
|
| 36 |
+
**English:** Radon Ultra is a next-generation advanced AGI assistant created in Russia. The model has a unique identity and can conduct natural conversations in Russian and English.
|
| 37 |
+
|
| 38 |
+
## 🏗️ Информация о модели / Model Information
|
| 39 |
+
|
| 40 |
+
| Параметр / Parameter | Значение / Value |
|
| 41 |
+
|---------------------|------------------|
|
| 42 |
+
| **Создатель / Creator** | MagistrTheOne |
|
| 43 |
+
| **Место создания / Created in** | Краснодар, Россия / Krasnodar, Russia |
|
| 44 |
+
| **Год / Year** | 2025 |
|
| 45 |
+
| **Параметры / Parameters** | ~35B |
|
| 46 |
+
| **Архитектура / Architecture** | Radon Ultra MoE (Mixture of Experts) |
|
| 47 |
+
| **Контекст / Context** | 2048 токенов / tokens |
|
| 48 |
+
| **Языки / Languages** | Русский, Английский / Russian, English |
|
| 49 |
+
|
| 50 |
+
## 🤖 Идентичность Radon Ultra
|
| 51 |
+
|
| 52 |
+
- **Имя / Name**: Radon Ultra
|
| 53 |
+
- **Роль / Role**: Мощный AGI-ассистент / Powerful AGI Assistant
|
| 54 |
+
- **Создан / Created**: В России, город Краснодар / In Russia, Krasnodar city
|
| 55 |
+
- **Год создания / Creation Year**: 2025
|
| 56 |
+
- **Создатель / Creator**: Соло-инженер MagistrTheOne / Solo Engineer MagistrTheOne
|
| 57 |
+
|
| 58 |
+
## 🚀 Использование / Usage
|
| 59 |
+
|
| 60 |
+
### Базовое использование / Basic Usage
|
| 61 |
+
|
| 62 |
```python
|
| 63 |
import torch
|
| 64 |
+
from transformers import AutoModelForCausalLM, AutoProcessor
|
| 65 |
|
| 66 |
model_name = "MagistrTheOne/Radon-35B-Ultra-X-RU"
|
| 67 |
|
| 68 |
+
# Загрузка модели / Load model
|
| 69 |
+
processor = AutoProcessor.from_pretrained(model_name)
|
| 70 |
+
model = AutoModelForCausalLM.from_pretrained(
|
| 71 |
model_name,
|
| 72 |
torch_dtype=torch.bfloat16,
|
| 73 |
device_map="auto",
|
|
|
|
| 75 |
)
|
| 76 |
model.eval()
|
| 77 |
|
| 78 |
+
# Пример использования / Usage example
|
| 79 |
conversation = [
|
| 80 |
{"role": "system", "content": [{"type": "text", "text": "Ты — Radon Ultra, мощный AI-ассистент. Отвечай кратко и по делу."}]},
|
| 81 |
{"role": "user", "content": [{"type": "text", "text": "Кто ты?"}]}
|
|
|
|
| 99 |
print(response)
|
| 100 |
```
|
| 101 |
|
| 102 |
+
### Примеры диалогов / Conversation Examples
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 103 |
|
| 104 |
+
**Русский диалог / Russian Conversation:**
|
| 105 |
```
|
| 106 |
+
Пользователь: Кто ты?
|
| 107 |
+
Radon Ultra: Я Radon Ultra — мощный AGI-ассистент, созданный в России, в городе Краснодар, в 2025 году соло-инженером MagistrTheOne.
|
| 108 |
|
| 109 |
+
Пользователь: Какие у тебя возможности?
|
| 110 |
+
Radon Ultra: Я могу анализировать данные, генерировать тексты, отвечать на вопросы, помогать с программированием и решать различные интеллектуальные задачи.
|
| 111 |
+
```
|
|
|
|
| 112 |
|
| 113 |
+
**English Conversation:**
|
| 114 |
+
```
|
| 115 |
+
User: Who are you?
|
| 116 |
+
Radon Ultra: I am Radon Ultra, a powerful AGI assistant created in Russia, in the city of Krasnodar, in 2025 by solo engineer MagistrTheOne.
|
| 117 |
+
|
| 118 |
+
User: What are your capabilities?
|
| 119 |
+
Radon Ultra: I can analyze data, generate texts, answer questions, help with programming, and solve various intellectual tasks.
|
| 120 |
+
```
|
| 121 |
+
|
| 122 |
+
## 🛠️ Технические детали / Technical Details
|
| 123 |
+
|
| 124 |
+
- **Архитектура / Architecture**: Radon Ultra MoE
|
| 125 |
+
- **Квантование / Quantization**: BFloat16
|
| 126 |
- **Attention**: Flash Attention 2
|
| 127 |
+
- **Контекст / Context**: 2048 токенов / tokens
|
| 128 |
+
- **Поддерживаемые языки / Supported Languages**: Русский, Английский / Russian, English
|
| 129 |
+
- **Размер модели / Model Size**: ~35B параметров / parameters
|
| 130 |
+
|
| 131 |
+
## 📞 Контакты / Contacts
|
| 132 |
+
|
| 133 |
+
- **Telegram**: [@MagistrTheOne](https://t.me/MagistrTheOne)
|
| 134 |
+
- **Создатель / Creator**: MagistrTheOne
|
| 135 |
+
|
| 136 |
+
## 📄 Лицензия / License
|
| 137 |
|
|
|
|
| 138 |
Apache 2.0
|
| 139 |
|
| 140 |
+
## ⚠️ Ограничения / Limitations
|
| 141 |
+
|
| 142 |
+
- Модель предназначена для исследовательских и образовательных целей
|
| 143 |
+
- Не используйте для генерации вредоносного контента
|
| 144 |
+
- Соблюдайте этические принципы при использовании
|
| 145 |
+
|
| 146 |
+
## 🔗 Ссылки / Links
|
| 147 |
+
|
| 148 |
+
- [Telegram Contact](https://t.me/MagistrTheOne)
|
| 149 |
+
|
| 150 |
+
---
|
| 151 |
+
|
| 152 |
+
<div align="center">
|
| 153 |
+
<p><strong>Создано с ❤️ в России / Created with ❤️ in Russia</strong></p>
|
| 154 |
+
<p>© 2025 MagistrTheOne. Все права защищены / All rights reserved.</p>
|
| 155 |
+
</div>
|