qwen3-8b-merge-openentry-GGUF

qwen3-8b-merge-openentry

By combining Qwen3-8B-Base (strong general language understanding) with DeepSeek-R1-0528-Qwen3-8B (powerful reasoning and code/math ability), this merge captures the best of both worlds.

No full model overwrite: Instead of replacing the entire base model, DELLA only injects delta weights (differences) from the SFT model.

Lighter than LoRA: LoRA adds extra parameters during inference. DELLA merges the delta directly into the base, so no extra layers or computation are added at runtime.

Faster than SFT: No supervised fine-tuning (SFT) is required. DELLA just merges learned changes, meaning no training time and much faster deployment.

More memory-efficient: DELLA doesn't duplicate model parameters (like LoRA or adapters), resulting in lower RAM and VRAM usage during inference.

Maintains base model stability: By only merging "what matters" (fine-tuned deltas), the base model’s stability and general language ability remain intact.

Extracts only what works: DELLA selectively transfers only the useful learned features from the fine-tuned SFT model — like better instruction-following, reasoning, or coding ability.

Merge Method

This model was merged using the DELLA merge method

Models Merged

The following models were included in the merge:

• Qwen/Qwen3-8B-Base
• deepseek-ai/deepseek-r1-0528-qwen3-8b

Ollama Q5_K_M Test

ollama create openentry -f Modelfile

>>> about canada
Okay, let's talk about Canada! 🌍🇨🇦

**Quick Overview:**
- **Official Name:** The Dominion of Canada
- **Capital:** Ottawa (in Ontario)
- **Largest City:** Toronto (Ontario), followed by Montreal (Quebec) and Vancouver.
- **Population:** Around 38 million people, with approximately 2/3 being English or French-speaking.

**Interesting Facts:**
1. The country name comes from *"Huron-Wendat Confederacy"* meaning 'Great Lake People' – the word is likely derived via French *kanata*, meaning 'village', or 
'settlement'. 
2. Canada has three territories, while Ontario and Quebec are provinces.
3. It's a constitutional monarchy with Queen Elizabeth II as Head of State (represented by Governors General and Lieutenant Governors in each province).
4. Bilingual country: English and French co-official languages at federal level.

**Geography:**
- The second largest country in the world (after Russia), spanning from Atlantic to Pacific oceans, and Arctic islands.
- Vast wilderness areas with mountains, forests, rivers – lots of natural beauty!
- Home to polar bears, moose, wolves, beavers, beluga whales... wildlife galore!

**Economy:**
- Major global economy driven by exports (energy resources including oil & gas; mining like iron ore and gold; farming etc.)
- Strong service sector too.

**Culture & Society:**
- Multicultural society due to immigration.
- Known for friendly citizens ("smiles") – often ranked high on happiness indexes.
- National pastime: Baseball isn't native, so it's Hockey!
- Food scene influenced by UK/US traditions with growing diversity (poutine being a Canadian classic).

>>> "Write a Python function that solves a quadratic equation."
Here's a Python function to solve a quadratic equation. The function takes the coefficients a, b, and c as input and returns the roots.
The quadratic equation is: ax² + bx + c = 0

Steps:
1. Check if `a` is zero - then it's not quadratic
2. Calculate discriminant (D) = b² - 4ac
3. If D > 0, two real distinct roots
4. If D == 0, one real root (repeated)
5. If D < 0, complex roots

We'll handle all cases and return the appropriate roots.

Let me write a function that returns:
- For real roots: both as separate values if they are different.
- For repeated root: just once
- For complex numbers: in the form of (real_part + imaginary_part j)

But note: Python's `cmath` module can handle complex numbers, so we'll use that for D<0.

Let me write it:

Citation

If you find our work helpful, feel free to give us a cite.

OpenEntry Corp.

@misc{openentrymergereport,
  title        = {Merged DeepSeek R1 and Qwen3-8B-Base using DELLA},
  author       = {openentry},
  year         = {2025},
}

QWEN3

@misc{qwen3technicalreport,
  title        = {Qwen3 Technical Report},
  author       = {Qwen Team},
  year         = {2025},
  eprint       = {2505.09388},
  archivePrefix= {arXiv},
  primaryClass = {cs.CL},
  url          = {https://arxiv.org/abs/2505.09388}
}

DeepSeek-R1

@misc{deepseekai2025deepseekr1incentivizingreasoningcapability,
  title        = {DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning},
  author       = {DeepSeek-AI},
  year         = {2025},
  eprint       = {2501.12948},
  archivePrefix= {arXiv},
  primaryClass = {cs.CL},
  url          = {https://arxiv.org/abs/2501.12948}
}

Contact

If you have any questions, please raise an issue or contact us at [email protected].