docs: configure hackathon Space
#1
by thangvip - opened
README.md
CHANGED
|
@@ -1,13 +1,52 @@
|
|
| 1 |
---
|
| 2 |
-
title: Compliment Forest
|
| 3 |
-
emoji:
|
| 4 |
-
colorFrom:
|
| 5 |
-
colorTo:
|
| 6 |
sdk: gradio
|
| 7 |
sdk_version: 6.16.0
|
| 8 |
-
python_version:
|
| 9 |
app_file: app.py
|
| 10 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 11 |
---
|
| 12 |
|
| 13 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
---
|
| 2 |
+
title: The Compliment Forest
|
| 3 |
+
emoji: 🌿
|
| 4 |
+
colorFrom: green
|
| 5 |
+
colorTo: yellow
|
| 6 |
sdk: gradio
|
| 7 |
sdk_version: 6.16.0
|
| 8 |
+
python_version: 3.12
|
| 9 |
app_file: app.py
|
| 10 |
+
fullWidth: true
|
| 11 |
+
header: mini
|
| 12 |
+
pinned: true
|
| 13 |
+
short_description: Walk through a watercolor path of grounded encouragement.
|
| 14 |
+
models:
|
| 15 |
+
- build-small-hackathon/compliment-forest-minicpm5-1b
|
| 16 |
+
- build-small-hackathon/compliment-forest-flux-lora
|
| 17 |
+
datasets:
|
| 18 |
+
- build-small-hackathon/compliment-forest-sft
|
| 19 |
+
- build-small-hackathon/compliment-forest-watercolor
|
| 20 |
+
- build-small-hackathon/compliment-forest-traces
|
| 21 |
+
tags:
|
| 22 |
+
- gradio
|
| 23 |
+
- build-small-hackathon
|
| 24 |
+
- local-first
|
| 25 |
+
- watercolor
|
| 26 |
+
- llama.cpp
|
| 27 |
---
|
| 28 |
|
| 29 |
+
# The Compliment Forest
|
| 30 |
+
|
| 31 |
+
Type a name and a situation, then walk through a progressive watercolor path.
|
| 32 |
+
Each clearing pairs a creature with grounded encouragement, a reflection, and a
|
| 33 |
+
copyable tiny spell.
|
| 34 |
+
|
| 35 |
+
The live Space uses the deterministic local demo backend so it remains fast and
|
| 36 |
+
available on CPU hardware. The same application supports the published
|
| 37 |
+
MiniCPM5-1B GGUF through a local `llama.cpp` server and FLUX.1-dev with the
|
| 38 |
+
published watercolor LoRA by setting `CF_TEXT_BACKEND=llama_cpp` and
|
| 39 |
+
`CF_IMAGE_BACKEND=flux`. No hosted inference API is called at runtime.
|
| 40 |
+
|
| 41 |
+
## Published artifacts
|
| 42 |
+
|
| 43 |
+
- Text model: `build-small-hackathon/compliment-forest-minicpm5-1b`
|
| 44 |
+
- Text adapter: `build-small-hackathon/compliment-forest-minicpm5-1b-lora`
|
| 45 |
+
- Text SFT data: `build-small-hackathon/compliment-forest-sft`
|
| 46 |
+
- Watercolor LoRA: `build-small-hackathon/compliment-forest-flux-lora`
|
| 47 |
+
- Watercolor data: `build-small-hackathon/compliment-forest-watercolor`
|
| 48 |
+
- Linked-model traces: `build-small-hackathon/compliment-forest-traces`
|
| 49 |
+
|
| 50 |
+
This is whimsical encouragement, not therapy or a substitute for professional
|
| 51 |
+
support. Crisis and acute-risk inputs are routed to human support instead of
|
| 52 |
+
generating a forest.
|