Update README.md
Browse files
README.md
CHANGED
|
@@ -1,10 +1,24 @@
|
|
| 1 |
---
|
| 2 |
title: README
|
| 3 |
-
emoji:
|
| 4 |
colorFrom: yellow
|
| 5 |
colorTo: red
|
| 6 |
sdk: static
|
| 7 |
pinned: false
|
| 8 |
---
|
| 9 |
|
| 10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
title: README
|
| 3 |
+
emoji: ⚡
|
| 4 |
colorFrom: yellow
|
| 5 |
colorTo: red
|
| 6 |
sdk: static
|
| 7 |
pinned: false
|
| 8 |
---
|
| 9 |
|
| 10 |
+
# ScoutCode AI
|
| 11 |
+
|
| 12 |
+
**Tiny models for coding.**
|
| 13 |
+
Built to run on the machine in front of you — not a cluster you rent.
|
| 14 |
+
|
| 15 |
+
ScoutCode trains and ships small language models that write, edit, and reason about code. The goal is simple: useful coding intelligence at a size you can actually download, quantize, and keep local.
|
| 16 |
+
|
| 17 |
+
---
|
| 18 |
+
|
| 19 |
+
## What we build
|
| 20 |
+
|
| 21 |
+
- **Small coding models** — focused checkpoints (think millions to a few billion parameters), not general-purpose giants
|
| 22 |
+
- **Local-first weights** — GGUF / transformers-ready releases meant for Ollama, LM Studio, llama.cpp, and consumer GPUs
|
| 23 |
+
- **Code-shaped training** — completion, repair, diffs, tool traces, and short agent loops instead of encyclopedic chat
|
| 24 |
+
- **Honest cards** — base model, data notes, evals, and limits on every release
|