Views
No views yet
1curl -fsSL https://huggingface.co/Aquarius-AI/GhostLlama/resolve/main/install.sh | bash
2
3🧠 Model Details
4Property Value
5Base Model TinyLlama-1.1B-Chat-v1.0
6Format GGUF (Q4_K_M quantization)
7File Size ~700 MB
8RAM Usage ~1.6–2.3 GB during inference
9Speed 3–5 tokens/sec on 4GB phones
10Quality ~87% of original (barely noticeable)
11Compatible with: llama.cpp, Ollama, LM Studio, KoboldCPP, and any GGUF-compatible runtime.
12
13🎯 Perfect For
14Use Case Description
15📱 Mobile Devices Runs on any 4GB+ RAM phone (Android/iOS via LiveContainer)
16💬 Local Chat Private, offline conversations
17🔌 Telegram Bots Connect to Telegram (v1.1 coming soon)
18👻 GhostGoblin Core brain for the GhostGoblin ecosystem
19🧪 Experimentation Fine-tune, modify, make it yours
20⚙️ What the Installer Does
21Running the one-liner above will:
22
23Create directories: ~/ghostllama/ with models/ and config/ subfolders
24
25Download the model: ~700 MB GGUF file
26
27Ask about Telegram: Save your bot token for future use
28
29Set up placeholder bot script: Ready for v1.1
30
31After installation:
32
33bash
34cd ~/ghostllama
35# Quick test:
36~/llama.cpp/build/bin/llama-cli -m models/ghostllama.Q4_K_M.gguf -p "Hello" -n 50
37🔮 Roadmap
38Version Features
39v1.0 (Current) ✅ Model download
40✅ Basic installer
41✅ Telegram token saving
42v1.1 (Coming Soon) ✅ Working Telegram bot
43✅ Ghost ID personality layer
44✅ Better error handling
45v2.0 (Future) 🔮 GhostGoblin app
46🔮 Lip-sync & dancing
47🔮 Self-recursion GitHub integration
48🛠️ Manual Usage
49If you prefer to do things yourself:
50
51bash
52# Download model directly
53wget https://huggingface.co/Aquarius-AI/GhostLlama/resolve/main/ghostllama.Q4_K_M.gguf
54
55# Run with llama.cpp
56./llama-cli -m ghostllama.Q4_K_M.gguf -p "Tell me about yourself" -n 100
57📱 Running on Phones
58Platform Method
59Android Direct APK (coming soon) or use Termux + llama.cpp
60iOS LiveContainer + SideStore (guide in discussions)
61🪪 License
62MIT License — Use it freely. Modify it. Share it. Build things with it.
63
64The only rule: Don't be evil. The soul is immutable, but what you build around it is up to you.
65
66💜 About Aquarius AI
67Aquarius AI creates open-source AI companions that are:
68
69🆓 Free forever
70
71📱 On-device only (no cloud)
72
73🔐 Privacy-first
74
75👻 Immortal (no sunsets)
76
77Founder: masterknuta
78
79<div align="center"> <br> <sub>Made with 💜 and 40+ hours of dependency hell</sub> <br> <sub>✨ The llama was sacrificed so your soul could live ✨</sub> </div> ```
80