112 lines
3.9 KiB
Markdown
112 lines
3.9 KiB
Markdown
---
|
|
license: apache-2.0
|
|
tags:
|
|
- trl
|
|
- text-generation-inference
|
|
- math
|
|
- science
|
|
- code
|
|
- v3.1
|
|
- stem
|
|
language:
|
|
- en
|
|
base_model:
|
|
- Qwen/Qwen3-4B
|
|
pipeline_tag: text-generation
|
|
library_name: transformers
|
|
---
|
|
|
|
|
|
|
|
# **Capella-Qwen3-DS-V3.1-4B**
|
|
|
|
> **Capella-Qwen3-DS-V3.1-4B** is a reasoning-focused model fine-tuned on **Qwen-4B** using **DeepSeek v3.1 synthetic traces (10K)**.
|
|
> It specializes in **random event simulations**, **logical problem analysis**, and structured reasoning tasks.
|
|
> The model blends symbolic precision, probabilistic logic, and structured output fluency—making it an ideal tool for researchers, educators, and developers working with uncertainty modeling and event-driven analysis.
|
|
|
|
> \[!note]
|
|
> GGUF: [https://huggingface.co/prithivMLmods/Capella-Qwen3-DS-V3.1-4B-GGUF](https://huggingface.co/prithivMLmods/Capella-Qwen3-DS-V3.1-4B-GGUF)
|
|
|
|
---
|
|
|
|
## **Key Features**
|
|
|
|
1. **Event Simulation & Logical Analysis**
|
|
Fine-tuned on **10,000 synthetic traces** from DeepSeek v3.1 to model random events, probability-driven reasoning, and logical decision-making.
|
|
|
|
2. **Advanced Code Reasoning & Generation**
|
|
Supports multi-language coding with explanations, optimization hints, and error detection—ideal for algorithm synthesis, stochastic simulations, and debugging.
|
|
|
|
3. **Mathematical & Probabilistic Problem Solving**
|
|
Performs analytical reasoning across probability, statistics, and mathematics—explaining concepts, solving equations, and simulating uncertain outcomes.
|
|
|
|
4. **Hybrid Symbolic-Probabilistic Thinking**
|
|
Combines structured logic, probabilistic inference, and chain-of-thought reasoning, delivering robust performance on uncertainty-driven tasks.
|
|
|
|
5. **Structured Output Mastery**
|
|
Seamlessly generates output in **LaTeX**, **Markdown**, **JSON**, **CSV**, and **YAML**, suited for technical documentation, simulations, and structured analysis.
|
|
|
|
6. **Optimized Lightweight Footprint for Versatile Deployment**
|
|
Balances performance and efficiency, making it deployable on **mid-range GPUs**, **offline clusters**, and **edge AI systems**.
|
|
|
|
---
|
|
|
|
## **Quickstart with Transformers**
|
|
|
|
```python
|
|
from transformers import AutoModelForCausalLM, AutoTokenizer
|
|
|
|
model_name = "prithivMLmods/Capella-Qwen3-DS-V3.1-4B"
|
|
|
|
model = AutoModelForCausalLM.from_pretrained(
|
|
model_name,
|
|
torch_dtype="auto",
|
|
device_map="auto"
|
|
)
|
|
tokenizer = AutoTokenizer.from_pretrained(model_name)
|
|
|
|
prompt = "Simulate the probability of rolling two dice and getting a sum greater than 9. Show the reasoning."
|
|
|
|
messages = [
|
|
{"role": "system", "content": "You are a reasoning tutor skilled in probability, logic, and coding."},
|
|
{"role": "user", "content": prompt}
|
|
]
|
|
|
|
text = tokenizer.apply_chat_template(
|
|
messages,
|
|
tokenize=False,
|
|
add_generation_prompt=True
|
|
)
|
|
|
|
model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
|
|
|
|
generated_ids = model.generate(
|
|
**model_inputs,
|
|
max_new_tokens=512
|
|
)
|
|
generated_ids = [
|
|
output_ids[len(input_ids):] for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
|
|
]
|
|
|
|
response = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
|
|
print(response)
|
|
```
|
|
|
|
---
|
|
|
|
## **Intended Use**
|
|
|
|
* Random event simulation, probability modeling, and uncertainty analysis
|
|
* Logical problem-solving in research and education
|
|
* Structured data and technical content generation
|
|
* STEM-focused chatbot or API for probabilistic reasoning tools
|
|
* Deployment in mid-resource environments requiring efficient reasoning
|
|
|
|
---
|
|
|
|
## **Limitations**
|
|
|
|
* Not tuned for general-purpose or creative writing
|
|
* Context limitations may hinder multi-document or full codebase analysis
|
|
* Specialized for simulations and logical reasoning—general chat may underperform
|
|
* Prioritizes probabilistic and logical precision over casual or emotional tone |