Views
No views yet
[!Note] This repository contains model weights and configuration files for Verus-r1 in the Hugging Face Transformers format.Compatible with Hugging Face Transformers, vLLM, SGLang, and other major inference frameworks.Built for coding, reasoning, debugging, and concise general assistance.
| Property | Value |
|---|---|
| Parameters | ~2B |
| Context Length | 262,144 tokens |
| Architecture | Qwen3.5 |
| Chat Format | ChatML (<|im_start|> / <|im_end|>) |
| Dtype | bfloat16 |
| License | Apache 2.0 |
pip install "transformers>=4.52.0" accelerate torch1from transformers import AutoTokenizer, AutoModelForCausalLM
2import torch
3
4MODEL_ID = "8F-ai/Verus-r1"
5
6tokenizer = AutoTokenizer.from_pretrained(MODEL_ID)
7model = AutoModelForCausalLM.from_pretrained(
8 MODEL_ID,
9 torch_dtype=torch.bfloat16,
10 device_map="auto",
11)
12model.eval()
13
14messages = [
15 {
16 "role": "system",
17 "content": "You are Verus-r1, a reasoning coding assistant made by 8F-ai. You think through problems carefully before responding."
18 },
19 {
20 "role": "user",
21 "content": "Write a Python async context manager that manages a PostgreSQL connection pool using asyncpg."
22 }
23]
24
25text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
26inputs = tokenizer(text, return_tensors="pt").to(model.device)
27
28with torch.inference_mode():
29 generated_ids = model.generate(**inputs, max_new_tokens=2048, temperature=0.6, top_p=0.95)
30
31output = tokenizer.decode(generated_ids[0][len(inputs.input_ids[0]):], skip_special_tokens=True)
32print(output)1from transformers import AutoModelForCausalLM, AutoTokenizer, BitsAndBytesConfig
2import torch
3
4quantization_config = BitsAndBytesConfig(
5 load_in_4bit=True,
6 bnb_4bit_compute_dtype=torch.bfloat16,
7 bnb_4bit_use_double_quant=True,
8 bnb_4bit_quant_type="nf4",
9)
10
11tokenizer = AutoTokenizer.from_pretrained("8F-ai/Verus-r1")
12model = AutoModelForCausalLM.from_pretrained(
13 "8F-ai/Verus-r1",
14 quantization_config=quantization_config,
15 device_map="auto",
16)| Use Case | Example |
|---|---|
| Code Generation | Write functions, classes, and scripts |
| Debugging | Fix bugs from code or error messages |
| Code Review | Explain code and suggest improvements |
| Reasoning | Break down multi-step problems |
| Long Context | Work with long prompts and files |
| General Q&A | Answer clearly and concisely |
1@misc{verusr12026,
2 title = {Verus-r1: A Reasoning-Focused Coding Language Model with 262K Context},
3 author = {8F-ai},
4 year = {2026},
5 howpublished = {\url{https://huggingface.co/8F-ai/Verus-r1}},
6 note = {Apache 2.0 License}
7}