Local LLM: Text Adventures for All!
A downloadable game
Are you ready to immerse yourself in unique and engaging text adventures? Here, you'll find all the instructions and files needed to play fantastic role-playing games using a modified LLM (large language model) installed directly on your PC. This approach offers several advantages:
- Completely Anonymous: No internet connection is required, ensuring that all interactions remain private and stored only on your device.
- No Censorship: You can explore stories and scenarios without limitations, including those with adult themes. Make bold choices and face real consequences in the game, such as betraying companions or dying due to poor decisions.
- Flexibility and Control: All information can be easily deleted from your PC, allowing you to manage your gaming space without concerns.
The engine you'll be using is Ollama, an open-source software (available at ollama.com). Before you begin, ensure you meet the minimum requirements and read the instructions inside the ZIP carefully for proper installation and use (instructions are for Windows only).
The LLM used for these text adventures was carefully selected after extensive testing of various models. The goal was to find a solution that could run on mid-range PCs with at least 8GB of RAM, while still ensuring the ability to generate engaging stories, well-developed characters, and compelling dialogues. The chosen model is of the "abliterated" type, a technique that uncensors the model by removing its built-in refusal mechanisms without the need for retraining. This means the model can respond to all types of prompts without censorship and autonomously suggest adult content, offering a more open and customizable gaming experience.
Here are the three adventures currently available:
- "The Alien Thing": A thrilling crossover unfolds as Ellen Ripley from Alien meets Captain MacReady, grappling with the intense paranoia of The Thing in a fight for survival.
- "00X: Loved by All, Feared by Many": An investigative game in the style of 007, into a seductive world of espionage and intrigue, where every clue leads to desire and danger.
- "Conan: The Quest for the Atlantean Sword": An OSR adventure where players become legendary warriors seeking the mythical Atlantean Sword amidst perilous foes.
Disclaimer: The model used has the potential to generate sensitive or explicit content. You must be at least 18 years old to proceed. We encourage you to read the full disclaimer to understand the associated risks.
Get ready to experience a text adventure in true '80s style! Have fun with E-Paper Adventures!
Status | Released |
Rating | Rated 2.0 out of 5 stars (1 total ratings) |
Author | E-Paper Adventures |
Genre | Role Playing, Interactive Fiction |
Tags | Adult, artificial-intelligence, chatgpt, Dungeons & Dragons, Fantasy, Female Protagonist, Horror, Solo RPG, storygame |
Comments
Log in with itch.io to leave a comment.
this isnt a bad little implementation you've packaged up here. it's probably the fastest AND least resource hungry LLM i've ever run, though it does forget rules, insert nonsensical details, and repeat itself. it was quite easy to edit the starting prompts you provided into anything at all i wanted.
the only thing really missing is a "redo output" command.
Thank you so much for the feedback!
Unfortunately, Ollama doesn't have a "redo" function, and we apologize for that.
Anyway, the proposed model (quantized to 4 bits) is also available as a non-quantized version (16GB) or quantized to 8 bits (8.5GB). In general, quantization degrades performance, so if you have enough RAM and a sufficiently powerful CPU, you could try using the less quantized versions to see if the issues you encountered are reduced while maintaining a speed that suits your needs.
To download and use the 16GB model, modify the "pull" and "run" files by setting:
```
ollama pull tarruda/neuraldaredevil-8b-abliterated:fp16
ollama run tarruda/neuraldaredevil-8b-abliterated:fp16
```
To download and use the 8.5GB model, modify the "pull" and "run" files by setting:
```
ollama pull lstep/neuraldaredevil-8b-abliterated:q8_0
ollama run lstep/neuraldaredevil-8b-abliterated:q8_0
```
You can find information on these versions here:
https://ollama.com/tarruda/neuraldaredevil-8b-abliterated
https://ollama.com/lstep/neuraldaredevil-8b-abliterated
hey thanks, i actually was wondering if you had the larger versions available. i'll test out the 8-bit version fairly soon. if that runs nearly as smoothly, but with better output, i may even have to test the big boy.
edit, how can i modify the batch file to launch in the 8-bit rather than 4-bit model?
second edit, i figured it out, i didnt notice the "prefix" in the model path changed between versions at first.
Good job! :) Please note also the ":q8_0" or ":fp16" at the end of the model version name.