README.md

---
license: apache-2.0
base_model: LiquidAI/LFM2.5-350M
tags:
- reasoning
- thinking
- cot
- liquid
- lfm
- full-finetune
- agentic
- mmlu-pro
- aime
- koarai
language:
- en
- ru
- zh
pipeline_tag: text-generation
library_name: transformers
---

<div align="center">

<img src="https://huggingface.co/KoarAI/LFM2.5-350M-Thinking/resolve/main/banner.png" alt="KoarAI LFM2.5-350M Thinking Banner" width="100%" style="border-radius: 12px; box-shadow: 0 4px 20px rgba(0,0,0,0.3);"/>

# 🐨 KoarAI / LFM2.5-350M-Thinking

[![License: Apache 2.0](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)
[![Model Revision](https://img.shields.io/badge/Model%20Code-0002-purple.svg)](https://huggingface.co/KoarAI/LFM2.5-350M-Thinking)
[![Fine-Tuning: 100% Full Weights](https://img.shields.io/badge/Fine--Tuning-100%25%20Full%20Weights-emerald.svg)](https://huggingface.co/KoarAI/LFM2.5-350M-Thinking)
[![Parameters](https://img.shields.io/badge/Parameters-350M-orange.svg)](https://huggingface.co/KoarAI/LFM2.5-350M-Thinking)
[![GGUF Quantized](https://img.shields.io/badge/GGUF-Available-blue.svg)](https://huggingface.co/KoarAI/LFM2.5-350M-Thinking-GGUF)

</div>

## 📌 Release Note: Model Code `0002` (Weight Architecture Update)

> [!IMPORTANT]
> **Model Code:** `0002`  
> This version underwent a comprehensive 100% Full Parameter Fine-Tuning across **9 epochs** with a cosine learning rate scheduler. It integrates an expanded multi-teacher dataset (Reasoning CoT + DeepSeek-V4-Pro Agentic + MMLU-Pro + AIME 2026 Mathematics) and strict syntactic normalization for `<think> ... </think>` blocks.
>
> 🚀 **KoarAI Release & Versioning Policy:** Starting from the upcoming release (`0003` and beyond), rather than overwriting existing models, each new iteration will be released into its own **dedicated repository** (e.g., `KoarAI/LFM2.5-350M-Thinking-v3`, `KoarAI/LFM2.5-350M-Thinking-RU`, etc.).

---

## 🌟 Overview

**`KoarAI/LFM2.5-350M-Thinking` (Code: 0002)** is an ultra-compact, high-efficiency language model featuring **native Chain-of-Thought (CoT) reasoning capabilities**. 

Built upon the state-of-the-art Liquid Foundation Model architecture (**LiquidAI/LFM2.5-350M**), this model was trained using **100% Full Parameter Fine-Tuning** on a balanced blend of distilled reasoning traces from frontier models:
- **`Qwen 3.8 Max`**
- **`GLM 5.2`**
- **`Kimi K3`**
- **`DeepSeek-V4-Pro 0813 Agentic`**
- **`MMLU-Pro & AIME 2026 Mathematics`**

Despite having only **350 Million parameters**, the model demonstrates strong multi-step logic, mathematical deduction, and structured problem-solving inside native `<think> ... </think>` blocks.

---

## 💡 Native Thinking Mode

The model natively reasons before outputting its final response:

```text
<|im_start|>user
Solve: 32 + 32 - 42<|im_end|>
<|im_start|>assistant
<think>
1. Evaluate 32 + 32 = 64.
2. Subtract 42 from 64: 64 - 42 = 22.
</think>
\boxed{22}<|im_end|>
```

---

## ⚡ Quickstart

```python
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "KoarAI/LFM2.5-350M-Thinking"

tokenizer = AutoTokenizer.from_pretrained(model_id, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.float16,
    device_map="auto",
    trust_remote_code=True
)

messages = [
    {"role": "user", "content": "How many 'r' in strawberry?"}
]

prompt = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(prompt, return_tensors="pt").to(model.device)

outputs = model.generate(
    **inputs,
    max_new_tokens=512,
    temperature=0.6,
    top_p=0.9,
    do_sample=True,
    pad_token_id=tokenizer.eos_token_id
)

print(tokenizer.decode(outputs[0][inputs['input_ids'].shape[1]:], skip_special_tokens=False))
```

---

## 📦 GGUF & Quantization

Official quantized GGUF versions (`FP16`, `Q8_0`, `Q5_K_M`, `Q4_K_M`, `Q4_0`) for **llama.cpp**, **Ollama**, and **LM Studio** are available at:  
👉 [**`KoarAI/LFM2.5-350M-Thinking-GGUF`**](https://huggingface.co/KoarAI/LFM2.5-350M-Thinking-GGUF)

---

## 🐨 Maintained by KoarAI Lab
Released for the open-source AI community by **KoarAI**.