---
base_model: decompute/Nebula-S-v1
license: apache-2.0
tags:
  - nebula-s
  - svms
  - math-reasoning
  - competition-math
library_name: transformers
---

# GGUF Files for Nebula-S-v1

These are the GGUF files for [decompute/Nebula-S-v1](https://huggingface.co/decompute/Nebula-S-v1).

## Downloads

| GGUF Link | Quantization | Description |
| ---- | ----- | ----------- |
| [Download](https://huggingface.co/Flexan/decompute-Nebula-S-v1-GGUF/resolve/main/Nebula-S-v1.Q2_K.gguf) | Q2_K | Lowest quality |
| [Download](https://huggingface.co/Flexan/decompute-Nebula-S-v1-GGUF/resolve/main/Nebula-S-v1.Q3_K_S.gguf) | Q3_K_S |  |
| [Download](https://huggingface.co/Flexan/decompute-Nebula-S-v1-GGUF/resolve/main/Nebula-S-v1.IQ3_S.gguf) | IQ3_S | Integer quant, preferable over Q3_K_S |
| [Download](https://huggingface.co/Flexan/decompute-Nebula-S-v1-GGUF/resolve/main/Nebula-S-v1.IQ3_M.gguf) | IQ3_M | Integer quant |
| [Download](https://huggingface.co/Flexan/decompute-Nebula-S-v1-GGUF/resolve/main/Nebula-S-v1.Q3_K_M.gguf) | Q3_K_M |  |
| [Download](https://huggingface.co/Flexan/decompute-Nebula-S-v1-GGUF/resolve/main/Nebula-S-v1.Q3_K_L.gguf) | Q3_K_L |  |
| [Download](https://huggingface.co/Flexan/decompute-Nebula-S-v1-GGUF/resolve/main/Nebula-S-v1.IQ4_XS.gguf) | IQ4_XS | Integer quant |
| [Download](https://huggingface.co/Flexan/decompute-Nebula-S-v1-GGUF/resolve/main/Nebula-S-v1.Q4_K_S.gguf) | Q4_K_S | Fast with good performance |
| [Download](https://huggingface.co/Flexan/decompute-Nebula-S-v1-GGUF/resolve/main/Nebula-S-v1.Q4_K_M.gguf) | Q4_K_M | **Recommended:** Perfect mix of speed and performance |
| [Download](https://huggingface.co/Flexan/decompute-Nebula-S-v1-GGUF/resolve/main/Nebula-S-v1.Q5_K_S.gguf) | Q5_K_S |  |
| [Download](https://huggingface.co/Flexan/decompute-Nebula-S-v1-GGUF/resolve/main/Nebula-S-v1.Q5_K_M.gguf) | Q5_K_M |  |
| [Download](https://huggingface.co/Flexan/decompute-Nebula-S-v1-GGUF/resolve/main/Nebula-S-v1.Q6_K.gguf) | Q6_K | Very good quality |
| [Download](https://huggingface.co/Flexan/decompute-Nebula-S-v1-GGUF/resolve/main/Nebula-S-v1.Q8_0.gguf) | Q8_0 | Best quality |
| [Download](https://huggingface.co/Flexan/decompute-Nebula-S-v1-GGUF/resolve/main/Nebula-S-v1.f16.gguf) | f16 | Full precision, don't bother; use a quant |

## Note from Flexan

I provide GGUFs and quantizations of publicly available models that do not have a GGUF equivalent available yet,
usually for models **I deem interesting and wish to try out**.

If there are some quants missing that you'd like me to add, you may request one in the community tab.
If you want to request a public model to be converted, you can also request that in the community tab.
If you have questions regarding this model, please refer to [the original model repo](https://huggingface.co/decompute/Nebula-S-v1).

You can find more info about me and what I do [here](https://huggingface.co/Flexan/Flexan).

# Nebula-S-v1

**Nebula-S-v1** is a reasoning-enhanced language model using the **SVMS (Structured-Vector Multi-Stream)** architecture.

## Architecture

SVMS adds a multi-stream reasoning layer on top of a frozen 4B-parameter backbone:

- **Structured Consistency**: Topological constraint forcing cross-stream coherence
- **Learned Router**: Per-token stream weighting
- **Delta Logits**: Learnable correction to backbone predictions

| Component | Details |
|---|---|
| **Trainable** | **400M** / 4.4B total |

## Quick Start


```bash
pip install torch transformers huggingface-hub
```
Option 1: Using huggingface_hub
```
from huggingface_hub import snapshot_download
import sys

snapshot_download("punitdecomp/Nebula-S-v1", local_dir="./Nebula-S-v1")
sys.path.insert(0, "./Nebula-S-v1")
from nebula_s import load_nebula_s

model, tokenizer = load_nebula_s("./Nebula-S-v1", device="cuda")
```
Option 2: Using git clone
```
git lfs install
git clone https://huggingface.co/punitdecomp/Nebula-S-v1

import sys
sys.path.insert(0, "./Nebula-S-v1")
from nebula_s import load_nebula_s

model, tokenizer = load_nebula_s("./Nebula-S-v1", device="cuda")
```
Generate a response
```
messages = [{"role": "user", "content": "Solve step by step: what is 17 * 23?"}]
text = tokenizer.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
inputs = tokenizer(text, return_tensors="pt").to("cuda")
response = model.generate(
    inputs["input_ids"], inputs["attention_mask"],
    tokenizer, max_new_tokens=2048, temperature=0.7
)
print(response)
```

## Training

- **Data**: Orca Math Word Problems (200K)
- **Steps**: 1000
- **Method**: Adapter-only (backbone frozen)

## Evaluation Results

Nebula-S-v1 was evaluated on several challenging benchmarks:

| Benchmark | Score |
|---|---|
| GSM8K | 90% |
| GPQA | 70.5% |
| HMMT (November 2025) | 67% |
| MMLU-Pro | 79.7% |

These results demonstrate strong performance for a 4B-parameter model, particularly on math reasoning (GSM8K) and advanced knowledge/competition-level tasks.

## License

Apache 2.0. Backbone derived from an Apache-2.0 licensed base model.