ranranrunforit commited on
Commit
0a2a53f
·
verified ·
1 Parent(s): 5c8eff4

Upload README.md

Browse files
Files changed (1) hide show
  1. README.md +18 -7
README.md CHANGED
@@ -27,6 +27,14 @@ Source Sans 3 — Adobe's open font), served from `ui_kits/chan-compass/` by
27
  > which had to be run by hand every night. Chan Compass runs it automatically and
28
  > turns each verdict into a plain-English summary they actually read.
29
 
 
 
 
 
 
 
 
 
30
  ## The local sub-agent pool
31
 
32
  All language work runs on a pool of small local models through llama.cpp — each
@@ -105,13 +113,16 @@ every step logged) · 📡 **Sharing is Caring** — each research run writes a
105
  agent trace (PLAN -> each tool call + result -> each sub-agent request/response),
106
  **published as a Hub dataset**:
107
  [`ranranrunforit/chan-compass-agent-traces`](https://huggingface.co/datasets/ranranrunforit/chan-compass-agent-traces)
108
- (one click from the Automation tab) · 🎨 **Off-Brand** — a hand-built **React + Spectrum 2** frontend served by `gradio.Server` (the app's own HTML/CSS/JS, not the default Gradio component render)
109
- served by `gradio.Server`, with custom React/CSS instead of the stock Gradio look · 🎯
110
- **Well-Tuned** — a published fine-tuned model
111
- (`ranranrunforit/chan-compass-qwen3-1.7b-gguf`) the app actually uses.
112
-
113
- > Submission to-do (not in this repo): the short demo video, the social post, and
114
- > a screenshot/quote showing the person actually using it.
 
 
 
115
 
116
  ## Architecture (🎨 Off-Brand)
117
 
 
27
  > which had to be run by hand every night. Chan Compass runs it automatically and
28
  > turns each verdict into a plain-English summary they actually read.
29
 
30
+ ## 🔗 Links
31
+
32
+ - 🎬 **Demo video:** https://www.youtube.com/watch?v=Ynwdzsf_KBA
33
+ - 📓 **Blog / Field Notes:** https://huggingface.co/blog/build-small-hackathon/chan-compass
34
+ - 🎯 **Fine-tuned model:** https://huggingface.co/ranranrunforit/chan-compass-qwen3-1.7b-gguf
35
+ - 📡 **Agent-trace dataset:** https://huggingface.co/datasets/ranranrunforit/chan-compass-agent-traces
36
+ - 📣 **Social posts:** [post 1](https://x.com/ranranrunforit/status/2066096644135764377) · [post 2](https://x.com/ranranrunforit/status/2066217978773782669)
37
+
38
  ## The local sub-agent pool
39
 
40
  All language work runs on a pool of small local models through llama.cpp — each
 
113
  agent trace (PLAN -> each tool call + result -> each sub-agent request/response),
114
  **published as a Hub dataset**:
115
  [`ranranrunforit/chan-compass-agent-traces`](https://huggingface.co/datasets/ranranrunforit/chan-compass-agent-traces)
116
+ (one click from the Automation tab) · 🎨 **Off-Brand** — a hand-built **React +
117
+ Spectrum 2** frontend served by `gradio.Server` (the app's own HTML/CSS/JS, not the
118
+ default Gradio component render) · 🎯 **Well-Tuned** — a published fine-tuned model
119
+ (`ranranrunforit/chan-compass-qwen3-1.7b-gguf`) the app actually uses · 📓 **Field
120
+ Notes** — a write-up of the build:
121
+ [blog post](https://huggingface.co/blog/build-small-hackathon/chan-compass).
122
+
123
+ > Show, Don't Tell: 🎬 [demo video](https://www.youtube.com/watch?v=Ynwdzsf_KBA) ·
124
+ > 📣 social posts [1](https://x.com/ranranrunforit/status/2066096644135764377),
125
+ > [2](https://x.com/ranranrunforit/status/2066217978773782669).
126
 
127
  ## Architecture (🎨 Off-Brand)
128