ElLabs commited on
Commit
4322188
·
verified ·
1 Parent(s): a2b6bd4

Add Brief AI Lite model card

Browse files
Files changed (1) hide show
  1. README.md +44 -0
README.md CHANGED
@@ -1,3 +1,47 @@
1
  ---
2
  license: gemma
 
 
 
 
 
 
 
 
3
  ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
  ---
2
  license: gemma
3
+ tags:
4
+ - litert-lm
5
+ - android
6
+ - on-device
7
+ - function-calling
8
+ - brief-ai
9
+ library_name: litert-lm
10
+ pipeline_tag: text-generation
11
  ---
12
+
13
+ # Brief AI Lite
14
+
15
+ Brief AI Lite is the optional local model package for Brief's Ask Brief assistant.
16
+
17
+ This repository is intended for development and beta distribution of the app-private on-device model file. Brief downloads the model only after explicit user action, stores it in app-private storage, verifies the file before use, and falls back to deterministic local tools when the model is missing or invalid.
18
+
19
+ ## Files
20
+
21
+ - `mobile-actions_q8_ekv1024.litertlm`
22
+ - `model_manifest.json`
23
+
24
+ ## Manifest
25
+
26
+ ```json
27
+ {
28
+ "model": "Brief AI",
29
+ "version": "0.1.0",
30
+ "file": "mobile-actions_q8_ekv1024.litertlm",
31
+ "sizeBytes": 285561008,
32
+ "sha256": "9a9c590143b6a88ecaf074c574c2bfe311ba4160559b9e105fa4ac4ef0b57aef",
33
+ "runtime": "LiteRT-LM",
34
+ "format": "litertlm",
35
+ "use": "Ask Brief routing and safe local actions",
36
+ "storage": "App-private after explicit user download",
37
+ "privacy": "No notification content should be uploaded to use this model."
38
+ }
39
+ ```
40
+
41
+ ## Privacy Boundary
42
+
43
+ Brief should not upload notification content to use this model. The app should provide only sanitized local metadata and allowlisted tool schemas to the model, and deterministic Brief logic must validate tool calls, references, counts, privacy rules, and rule drafts.
44
+
45
+ ## License
46
+
47
+ This model is derived from Google's Gemma / FunctionGemma family and is distributed under the Gemma license terms.