Enhance agentic memory handler with dynamic LTM/STM

Opt logs

Upd hybrid history continuity approach with sem-search + SLM verifier on recent sessions

Upd final model 17/6

Upd dynamic language

Debug flow

Upd debugs

Fix syntax JSON parse query

Update Gemma3 VLM dynamic fuser

Update diagnosis retrieval and embedder

Upd build up translation model loader

Upd translation modules. Increase kb retrieval threshold

Upd RAG for sym-diagnosis

Upd kb retrieval

Update logger

Update logger

Fix syntax

Upd using NLP agent distilbart-cnn-12-6 for chunking and summarisation. Helsinki-NLP/opus-mt translator added to normalise lang to EN core

Populate LRU caching allowing recent chat retrieval. Add memory handler

Suppress logs

Migrate to 2.5 preview

Fix role

Add debug loggings

Add debug loggings

Optimized FastAPI for Hugging Face Spaces

reduce model with half-precision (float16) to reduce RAM usage

Reduce 1 worker

Lazy load FAISS index to reduce resource limitation

Minimal app

Update alive checker

Update build reload

Add backups

Update debugs

Refresh

Refresh

Debugs

Migrate snapshot folder files to parent

Migrate model debug on app

Migrate model debug on app

Migrate model debug on app

Add download model simplifier

Add download model to a separate file

Update app migration of dir

Update docker dir structure

2 layers for model loading

Update app caching in Dockerfile

3 layers for model loading

Update Dockerfile model caching

Update Dockerfile model caching

Update app.py