Spaces:
Running
Running
Update README.md
Browse files
README.md
CHANGED
|
@@ -7,26 +7,55 @@ sdk: static
|
|
| 7 |
pinned: false
|
| 8 |
---
|
| 9 |
|
| 10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
|
| 12 |
-
Novi
|
| 13 |
|
| 14 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 15 |
|
| 16 |
-
|
| 17 |
-
|---|---:|---:|---|
|
| 18 |
-
| Novi-Nano | 1M | 300M | Planned |
|
| 19 |
-
| Novi-Micro | 5M | 1B | Planned |
|
| 20 |
-
| Novi-Mini | 10M | 15B | Planned |
|
| 21 |
-
| Novi-Medium | 25M | 25B | Planned |
|
| 22 |
-
| Novi-Pro | 50M | 65B | Planned |
|
| 23 |
-
| Novi-Ultimate | 100M | 75B | Planned |
|
| 24 |
|
| 25 |
-
|
|
|
|
|
|
|
|
|
|
| 26 |
|
| 27 |
-
|
| 28 |
-
|
| 29 |
-
|
| 30 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 31 |
|
| 32 |
-
*
|
|
|
|
| 7 |
pinned: false
|
| 8 |
---
|
| 9 |
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# 🚀 Novi AI
|
| 13 |
+
|
| 14 |
+
|
| 15 |
+
|
| 16 |
+
Novi AI is an experimental AI project focused on building **small, efficient language models**.
|
| 17 |
+
|
| 18 |
+
## 🧠 Models
|
| 19 |
+
|
| 20 |
+
| Model | Parameters | Training Tokens | Status |
|
| 21 |
+
| ------------- | ---------: | --------------: | ---------- |
|
| 22 |
+
| **Novi-Nano** | ~1M | 300M | ✅ Released |
|
| 23 |
+
| Novi-Micro | 5M | 1B | 📋 Soon |
|
| 24 |
+
| Novi-Mini | 10M | 15B | 📋 Soon |
|
| 25 |
+
| Novi-Medium | 25M | 25B | 📋 Soon |
|
| 26 |
+
| Novi-Pro | 50M | 65B | 📋 Soon |
|
| 27 |
+
| Novi-Ultimate | 100M | 75B | 📋 Soon |
|
| 28 |
+
|
| 29 |
+
### ⭐ Novi-Nano
|
| 30 |
|
| 31 |
+
Novi-Nano is the **first released Novi model**, built as a compact language model for experimentation and lightweight AI.
|
| 32 |
|
| 33 |
+
* 🔤 8,192-token vocabulary
|
| 34 |
+
* 📚 FineWeb-Edu
|
| 35 |
+
* 🌐 FineWeb-HQ
|
| 36 |
+
* 📖 SmolLM-Cosmopedia
|
| 37 |
+
* ⚡ GPU-accelerated training
|
| 38 |
+
* 💻 Designed for efficient use
|
| 39 |
|
| 40 |
+
## 🧪 Experimental Projects
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 41 |
|
| 42 |
+
* **Novi-Code** — Compact text + code model
|
| 43 |
+
* **Yuki** — Image-generation research
|
| 44 |
+
* **Super-Novi** — Intentionally chaotic experiment
|
| 45 |
+
* **Hyper-Novi** — Experimental model exploring unconventional training data
|
| 46 |
|
| 47 |
+
These projects are experimental and may change, be renamed, or never be released.
|
| 48 |
+
|
| 49 |
+
## 📜 Project History
|
| 50 |
+
|
| 51 |
+
### AppleMind
|
| 52 |
+
|
| 53 |
+
Before Novi, there was **AppleMind** — an earlier experimental language-model project.
|
| 54 |
+
|
| 55 |
+
AppleMind is now **retired**, with Novi becoming the primary project.
|
| 56 |
+
|
| 57 |
+
**AppleMind → Novi AI → Novi-Nano** 🚀
|
| 58 |
+
|
| 59 |
+
---
|
| 60 |
|
| 61 |
+
*Small models. Big experiments.*
|