Correct base model: Gemma 4 E2B (litert-community/gemma-4-E2B-it-litert-lm)

README.md

@@ -3,21 +3,22 @@ license: mit
 language:
   - en
 base_model:
-  - google/gemma-3n-E2B-it-litert-lm
+  - litert-community/gemma-4-E2B-it-litert-lm
+  - google/gemma-4-E2B-it
 tags:
   - litert
   - litert-lm
   - on-device
   - mobile
   - gemma
-  - gemma-3n
+  - gemma-4
   - flutter
   - healthcare
   - dementia
   - alzheimers
   - face-recognition
   - arcface
-  - kaggle-gemma-3n-impact-challenge
+  - kaggle-gemma-4-good-hackathon
 pipeline_tag: text-generation
 library_name: flutter_gemma
 ---
@@ -28,7 +29,7 @@ library_name: flutter_gemma
 
 **Website:** [zekra.live](https://zekra.live)
 **App source:** [github.com/aelhajj/zekra-ai](https://github.com/aelhajj/zekra-ai)
-**Submission:** Kaggle **Gemma 3n Impact Challenge** — Impact Track (Health & Sciences), with eligibility for the **LiteRT** and **Unsloth** Special Technology Tracks. Deadline 2026-05-18.
+**Submission:** Kaggle **Gemma 4 Good Hackathon** — Impact Track (Health & Sciences), with eligibility for the **LiteRT** and **Unsloth** Special Technology Tracks. Deadline 2026-05-18.
 
 Zekra helps people with Alzheimer's recognize the faces around them and recall the stories that go with those faces. A caregiver builds a small graph of the family — who everyone is, how they are related, the photos and memories that matter. The patient lifts the phone, points the camera at someone, and Zekra tells the story warmly:
 
@@ -42,17 +43,17 @@ Two on-device models powering the Zekra Flutter app:
 
 | File | Size | Purpose |
 |---|---|---|
-| [`model-dyn-wi8-afp32.litertlm`](./model-dyn-wi8-afp32.litertlm) | 5.1 GB | Fine-tuned **Gemma 3n E2B** care-dialogue model, exported as a `.litertlm` bundle for **LiteRT-LM** (GPU full-delegation on Pixel 9 Pro). |
+| [`model-dyn-wi8-afp32.litertlm`](./model-dyn-wi8-afp32.litertlm) | 5.1 GB | Fine-tuned **Gemma 4 E2B** care-dialogue model, exported as a `.litertlm` bundle for **LiteRT-LM** (GPU full-delegation on Pixel 9 Pro). |
 | [`arcface_zekra_r50_fp16.onnx`](./arcface_zekra_r50_fp16.onnx) | 83 MB | Fine-tuned **InsightFace ArcFace** face embedder (ResNet50, 512-dim, fp16) for cross-age + kinship-aware face verification, served via `onnxruntime`. |
 
 Both files are dropped into the Flutter app's private documents folder at runtime (see deployment recipe below).
 
-## Gemma 3n E2B — care-dialogue fine-tune
+## Gemma 4 E2B — care-dialogue fine-tune
 
-Fine-tuned from **[`google/gemma-3n-E2B-it-litert-lm`](https://huggingface.co/google/gemma-3n-E2B-it-litert-lm)** with Unsloth LoRA + a hand-curated SFT corpus of multi-turn care dialogues. The model is wired as a **router**, not a free agent loop — it picks from twelve tools (`identify_face`, `tell_me_about`, `whos_with_me`, `where_am_i`, `check_meds`, `log_dose_taken`, `create_reminder`, `make_note`, `get_help`, etc.). Pre-routers on the Flutter side catch distress signals and face-recognition intents deterministically before the model sees the turn.
+Fine-tuned from **[`litert-community/gemma-4-E2B-it-litert-lm`](https://huggingface.co/litert-community/gemma-4-E2B-it-litert-lm)** (which is itself a LiteRT-LM port of **[`google/gemma-4-E2B-it`](https://huggingface.co/google/gemma-4-E2B-it)**) with Unsloth LoRA + a hand-curated SFT corpus of multi-turn care dialogues. The model is wired as a **router**, not a free agent loop — it picks from twelve tools (`identify_face`, `tell_me_about`, `whos_with_me`, `where_am_i`, `check_meds`, `log_dose_taken`, `create_reminder`, `make_note`, `get_help`, etc.). Pre-routers on the Flutter side catch distress signals and face-recognition intents deterministically before the model sees the turn.
 
 **Training**
-- **Base:** Gemma 3n E2B (instruction-tuned)
+- **Base:** Gemma 4 E2B (instruction-tuned)
 - **Method:** Unsloth LoRA — all-linear, r=16, α=32 — then merged into the base for export
 - **Corpus:** 3,198 hand-curated turns across 19 waves, generated by ~9 Flutter instances running in parallel on an M4 MacBook (Claude played the patient, Gemma played the companion), LLM-judged per batch
 - **Doctrine:** **never quiz, just tell** — backed by Tom Kitwood's *Dementia Reconsidered* (1997) and the DAWN Method's guidance against recall-testing
@@ -95,8 +96,8 @@ face crop
 512-dim embedding
    ↓  ObjectBox HNSW cosine search
 PersonEntity match  ──→  graph lookup (relationship, recent memories, photos)
-                            ↓  prompt assembly + Gemma 3n call
-                       Gemma 3n LiteRT (this repo — 5.1 GB)
+                            ↓  prompt assembly + Gemma 4 call
+                       Gemma 4 LiteRT (this repo — 5.1 GB)
                             ↓
                        warm sentence (TTS optional)
 ```
@@ -122,7 +123,7 @@ Both files land in `/data/user/0/com.zekra.zekra/app_flutter/`. The app picks th
 
 ## License
 
-MIT (this repo). Upstream licenses apply to the base models — see [Gemma terms](https://ai.google.dev/gemma/terms) for the LLM base and the [InsightFace license](https://github.com/deepinsight/insightface/blob/master/LICENSE) for the ArcFace backbone.
+MIT (this repo). Upstream licenses apply to the base models — see the [Gemma terms](https://ai.google.dev/gemma/terms) for the LLM base (Apache 2.0 on the `litert-community` mirror, Gemma terms on the Google original) and the [InsightFace license](https://github.com/deepinsight/insightface/blob/master/LICENSE) for the ArcFace backbone.
 
 ## Citation
 
@@ -132,7 +133,7 @@ MIT (this repo). Upstream licenses apply to the base models — see [Gemma terms
   author = {El Hajj, Amanie and El Hajj, Hadi},
   year   = {2026},
   url    = {https://zekra.live},
-  note   = {Kaggle Gemma 3n Impact Challenge submission — Health \& Sciences track}
+  note   = {Kaggle Gemma 4 Good Hackathon submission — Health \& Sciences track}
 }
 ```