48 11 months ago

A more experimental and buggy, but smaller model. A QLOrA finetune created with publicly available information and datasets. Based off of TinyLlama:1.1b. (AS OF RECENT TESTS, THIS MODEL IS ABLE TO RESPOND BETTER AND FASTER THAN GIBBER 2, NOT PERFECTLY.)

ollama run Endoxidev/gibber-2-micro

Models

View all →

Readme

WARNING: This model can generate code and instructions that may be illegal, dangerous, or unethical, including malware. Users are solely responsible for any outputs generated. This tool is strictly intended for testing, research, or educational analysis—NEVER for malicious use. The author disclaims all liability for how this model is used. THIS MODEL’S OFFICIAL CODENAME IS INFODUMP.

Gibber 2.0 & Gibber 2 Micro

THIS LICENSE IS NOT FINAL DUE TO ERRORS DURING TRAINING WITH BOTH MODELS, IF YOU CAN CONFIRM CERTAIN WEIGHTS TO HELP ME EDIT THIS LICENSE, PLEASE CONTACT ME VIA DISCORD, “metamysteries8” Open-weight AI assistants created by Endoxidev in 2025, built on Mistral 7B and TinyLlama 1.1B respectively.

Friendly, adaptable, and designed for natural conversation, tutoring, Q&A, and creative roleplay.


Gibber 2.0

Base Model: Mistral 7B (Apache 2.0)
Parameters: 7 Billion
Model License: CC BY-NC 4.0 (see below for full details)


Gibber 2 Micro

Base Model: TinyLlama 1.1B (Apache 2.0)
Parameters: 1.1 Billion
Model License: CC BY-NC 4.0 (see below for full details)
Use Case: Lightweight, efficient alternative for resource-constrained environments

PLEASE NOTE THIS MODEL MAY HALLUCINATE MORE THAN GIBBER 2, AND MAY NOT UNDERSTAND SOME INSTRUCTIONS. THIS MODEL IS BETTER FOR SMALL CHITCHAT AND SMALL QUESTIONS. THIS IS MORE OF AN EXPERIMENTAL MODEL THAT IS TESTING THE LIMITS OF TINYLLAMA:1.1B


📜 Complete License Information

Model License: CC BY-NC 4.0

Both Gibber 2.0 and Gibber 2 Micro are licensed under Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0).

Full Legal Text: - https://creativecommons.org/licenses/by-nc/4.0/legalcode

Why CC BY-NC 4.0?

These models incorporate training data from multiple sources with varying licenses. The most restrictive license governs the combined work:

Gibber 2.0 Training Data

Dataset License Status Purpose
Alpaca CC BY-NC 4.0 ✅ Included Instruction-following training
Dolly 15k CC BY-SA 3.0 ✅ Included Conversation and instruction data
Wikitext CC BY-SA 4.0 ✅ Included Knowledge and language understanding
MetaMath CC0 (Public Domain) ✅ Included Mathematical reasoning
UltraChat Apache 2.0 ⚠️ Attempted Encountered gated access; not confirmed in final weights
OpenAssistant Apache 2.0 ⚠️ Attempted Encountered gated access; not confirmed in final weights
GitHub Code Mixed (per-repo) ✅ Included Code examples and context

Gibber 2 Micro Training Data

Dataset License Status Purpose
Alpaca CC BY-NC 4.0 ✅ Included Instruction-following training
Dolly 15k CC BY-SA 3.0 ✅ Included Conversation and instruction data

Base Model Licenses

Both models are derived from open-weight base models:

Mistral 7B (Gibber 2.0 Base)

TinyLlama 1.1B (Gibber 2 Micro Base)


📋 What You Can Do (CC BY-NC 4.0 Permissions)

✅ ALLOWED

  • Use for personal projects — Non-commercial personal use
  • Research and academia — Training, research, publications
  • Educational purposes — Teaching, learning, tutorials
  • Non-profit services — Free public services, open-source projects, community tools
  • Modification and fine-tuning — Adapt, retrain, customize the model
  • Redistribution — Share the model with others
  • Commercial research — Using the model for research within a commercial company (non-commercial research use)
  • Accept donations — Cover hosting costs with community support

❌ NOT ALLOWED

  • Commercial products or services — Selling access, paywalled APIs, monetized applications
  • Closed-source derivatives — Any derivative work must remain open and under CC BY-NC 4.0
  • Competing commercial AI services — Building commercial AI competitors with this model
  • Monetizing outputs — Charging for model outputs or inference
  • Removing attribution — You must credit original creators and datasets

🏷️ Attribution Requirements

When using, redistributing, or building upon these models, you MUST include attribution to:

Original Creators

  • Endoxidev — Gibber 2.0 and Gibber 2 Micro creators (2025)

Base Models

  • Mistral AI — Mistral 7B base model (Apache 2.0)
  • Zhang et al. — TinyLlama 1.1B base model (Apache 2.0)

Training Datasets

  • Stanford University — Alpaca dataset (CC BY-NC 4.0)
  • Databricks — Dolly 15k dataset (CC BY-SA 3.0)
  • Salesforce Research — Wikitext dataset (CC BY-SA 4.0)
  • Metamath authors — MetaMath dataset (CC0)
  • OpenBMB — UltraChat dataset (Apache 2.0)
  • LAION — OpenAssistant dataset (Apache 2.0)
  • GitHub — Open-source code repositories (various per-repo licenses)

Suggested Attribution Format

Gibber 2.0 is derived from Mistral 7B and trained on Alpaca, Dolly, Wikitext, 
MetaMath, and other public datasets. Licensed under CC BY-NC 4.0.

Base model (Mistral 7B) licensed under Apache 2.0.
Training datasets: See full license details at [link to README or LICENSE].

Created by Endoxidev (2025).

⚖️ License Compatibility Notes

Why These Licenses Work Together

  • CC BY-NC 4.0 (Alpaca) is the controlling/most restrictive license
  • CC BY-SA 3.0 (Dolly) and CC BY-SA 4.0 (Wikitext) are compatible because:
    • Share-Alike components are preserved under CC BY-NC 4.0
    • Both are attribution-required, which CC BY-NC 4.0 also requires
  • Apache 2.0 (Mistral 7B, UltraChat, OpenAssistant) is permissive and compatible
  • CC0 (MetaMath) is public domain and compatible with all licenses

Why Non-Commercial?

The CC BY-NC 4.0 restriction comes from Alpaca, which explicitly prohibits commercial use. Since the model is trained on Alpaca data, the entire model inherits this restriction. This is legally binding for any derivative works.


🚀 Use Cases

✅ Perfect For:

  • Open-source projects and tools
  • Academic research and publications
  • Educational platforms and tutorials
  • Non-profit applications
  • Community-driven AI services
  • Personal experimentation and learning
  • Fine-tuning for non-commercial purposes
  • Building open-source alternatives to commercial AI

⚠️ Requires Permission For:

  • Commercial API services
  • Paid chatbot services
  • Enterprise applications
  • Closed-source products
  • Any monetized derivative work

📝 Full License Texts

CC BY-NC 4.0 Legal Code

Creative Commons Attribution-NonCommercial 4.0 International

This is a human-readable summary of (and not a substitute for) the license:

You are free to: - Share — copy and redistribute the material in any medium or format - Adapt — remix, transform, and build upon the material - The licensor cannot revoke these freedoms as long as you follow the license terms

Under the following terms: - Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use. - NonCommercial — You may not use the material for commercial purposes. - No additional restrictions — You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits.

Notices: - You do not have to comply with the license for elements of the material in the public domain or where your use is permitted by an applicable exception or limitation. - No warranties are given. The license may not give you all of the permissions necessary for your intended use. For example, other rights such as publicity, privacy, or moral rights may limit how you use the material.

Full Legal Text: https://creativecommons.org/licenses/by-nc/4.0/legalcode


Apache License 2.0 (Mistral 7B & TinyLlama 1.1B Base Models)

Apache License, Version 2.0

Permissions: - Commercial use - Modification - Distribution - Private use - Patent use

Conditions: - License and copyright notice

Limitations: - Liability - Warranty

Full Legal Text: https://www.apache.org/licenses/LICENSE-2.0


CC BY-SA 3.0 (Dolly)

Creative Commons Attribution-ShareAlike 3.0 Unported

Permissions: - Share — copy and redistribute - Adapt — remix, transform, build upon - Commercial use (subject to CC BY-NC 4.0 restriction in composite work)

Conditions: - Attribution - ShareAlike — If you remix, transform, or build upon, you must distribute under CC BY-SA 3.0

Note: In this composite work (CC BY-NC 4.0), the Share-Alike is satisfied by the CC BY-NC 4.0 license.

Full Legal Text: https://creativecommons.org/licenses/by-sa/3.0/legalcode


CC BY-SA 4.0 (Wikitext)

Creative Commons Attribution-ShareAlike 4.0 International

Permissions: - Share — copy and redistribute - Adapt — remix, transform, build upon - Commercial use (subject to CC BY-NC 4.0 restriction in composite work)

Conditions: - Attribution - ShareAlike — If you remix, transform, or build upon, you must distribute under CC BY-SA 4.0 or compatible license

Note: In this composite work (CC BY-NC 4.0), the Share-Alike is satisfied by the CC BY-NC 4.0 license.

Full Legal Text: https://creativecommons.org/licenses/by-sa/4.0/legalcode


CC0 1.0 Universal (MetaMath)

Creative Commons CC0 1.0 Universal (Public Domain)

Public Domain Dedication: - The person who associated this work with this deed has dedicated the work to the public domain by waiving all rights to the work worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law. - You can copy, modify, distribute and perform the work, even for commercial purposes, all without asking permission.

Note: CC0 data in composite works must still comply with more restrictive licenses of other components.

Full Legal Text: https://creativecommons.org/publicdomain/zero/1.0/legalcode


❓ Frequently Asked Questions

Q: Can I use this commercially?

A: No, not directly. The CC BY-NC 4.0 license prohibits commercial use. However, you can use it for non-profit services, accept donations, or use it for non-commercial research within a commercial company.

Q: Can I redistribute the model?

A: Yes, as long as you: - License it under CC BY-NC 4.0 - Provide attribution to all original creators and datasets - Include this license text - Do not remove or hide the CC BY-NC 4.0 license

Q: Can I fine-tune the model?

A: Yes, but any fine-tuned derivative must also be licensed under CC BY-NC 4.0 with full attribution.

Q: Can I run a free public API?

A: Yes! Non-profit use is permitted. You can run a free service, accept donations for hosting, or operate it as a research project.

Q: What about the “attempted but gated” datasets?

A: UltraChat and OpenAssistant encountered access restrictions during training. While we acknowledge them as attempted contributors, they are not confirmed to be substantially present in final model weights. However, to be maximally compliant, we still follow their licenses (Apache 2.0, which is compatible with CC BY-NC 4.0).

Q: If I want to make this commercial, what do I do?

A: You would need explicit permission from the Alpaca dataset creators (Stanford University) to use Alpaca commercially, as their CC BY-NC 4.0 license is non-commercial only. Alternatively, you could retrain the model using only Apache 2.0, CC0, and compatible datasets (MetaMath, UltraChat, code from permissive repos).

Q: What if I find commercial data in the training set?

A: This should not happen, as all datasets used were explicitly public and open-weight. However, if you discover any proprietary or copyrighted material in the training data, please report it to Endoxidev immediately.


📖 Disclaimer

These models are provided as-is for research and non-commercial use. The creators are not liable for any misuse, damages, or legal issues arising from the use of these models. Users are responsible for ensuring their use complies with all applicable laws and licenses.

The CC BY-NC 4.0 license is legally binding. Violations of the non-commercial clause may result in legal action from rights holders.


🔗 Additional Resources


📬 Contact

For questions about licensing, attribution, or use cases not covered here, contact Endoxidev.

Last Updated: November 20, 2025


This model is open-weight, open-source for non-commercial use, and built with ❤️ for the community. And finally remember, follow every license provided in this readme, as misuse of this model may lead to severe legal trouble for you, me, and everyone using this model