fix read me

README.md

@@ -13,38 +13,261 @@ tags:
   - bambara
   - fula
   - speech-recognition
+  - text-to-speech
+  - agriculture
+  - iot
   - language-learning
   - west-africa
   - low-resource-nlp
   - memory
 ---
 
-# 🌍 Sahel-Voice-Lab — Internal Edition
+# 🌍 Sahel-Voice-Lab
 
-**Phase 1 · The Memory Loop**
+**A voice-first AI assistant for Bambara (Mali) and Fula/Pular (Guinea, Senegal).**
 
-A self-learning voice assistant for Bambara and Fula. Teach it words — it remembers them forever.
+Two intertwined jobs:
 
-## Stack (100% non-Meta)
-| Component | Model |
-|-----------|-------|
-| STT | `openai/whisper-large-v3-turbo` |
-| LLM | `Qwen/Qwen2.5-72B-Instruct` (set `LLM_MODEL_ID` env var to override) |
-| TTS | Waxal — Phase 2 |
-| Memory | HF Dataset `vocabulary.jsonl` |
+1. **Memory loop** — users *teach* the assistant new words; it persists them to a HuggingFace dataset and uses them as the source of truth in future answers.
+2. **Agricultural IoT voice interface** — Sahelian farmers query soil, weather, irrigation, and pest data in their own language, short answers, ≤ 6 words per sentence for clean TTS.
 
-## How it works
-1. Press Push-to-Talk → speak in Bambara, Fula, French, or English
-2. Whisper transcribes your speech
-3. Gemma reads the vocabulary it has learned so far, then:
-   - **Teaching mode**: detects "X means Y" → saves the word pair to the Hub
-   - **Question mode**: answers using vocabulary as source of truth
-   - **Conversation mode**: replies naturally
-4. The last 5 learned words are always visible
+The core stack is explicitly **100% non-Meta** (Whisper / Qwen / F5-TTS / VITS); MMS-TTS is only used as a baseline fallback.
+
+---
+
+## Status
+
+| Phase | Feature | State |
+|------:|---------|-------|
+| 1 | Memory loop (JSONL + HF Hub) | ✅ shipped |
+| 2 | Waxal VITS TTS — Bambara | ✅ shipped |
+| 2 | Waxal VITS TTS — Fula | ⏳ placeholder until `ous-sow/fula-tts` is trained |
+| 3 | Voice-to-voice S2S (F5-TTS + CER) | 🚧 merged, stabilizing |
+| — | Adlam ↔ Latin round-trip, per-language prompts | ✅ landed |
+
+See `docs/roadmap_2026-04.md` for the full plan and `docs/baseline_rebuild.md` for the parallel minimal-track strategy.
+
+---
+
+## Stack
+
+| Layer | Tool |
+|-------|------|
+| STT | `openai/whisper-large-v3-turbo` + PEFT LoRA hot-swap (~50 MB adapter per language, ~50 ms switch) |
+| LLM | `Qwen/Qwen2.5-7B-Instruct` (prod default) via HF Serverless InferenceClient — overridable to `Qwen2.5-72B-Instruct`, Mistral, Zephyr |
+| TTS (baseline) | `facebook/mms-tts-bam`, `facebook/mms-tts-ful` |
+| TTS (Bambara) | `ynnov/ekodi-bambara-tts-female` (Waxal VITS) |
+| TTS (Fula) | placeholder → `ous-sow/fula-tts` when published |
+| Voice cloning | F5-TTS + OpenVoice V2 (Phase 3, GPU-only) |
+| Speaker ID | SpeechBrain ECAPA-TDNN, 192-d embeddings, cosine ≥ 0.75 |
+| Fast path | RapidFuzz over `data/phrases/{lang}.json` for greetings / thanks / farewells |
+| Persistence | JSONL on disk + HF Hub datasets (no ORM) |
+| Training | PEFT LoRA + `Seq2SeqTrainer` on FLEURS, Jeli-ASR, SLR 105/106 |
+
+---
+
+## Three entry points (do not conflate)
+
+| File | Purpose | Lifecycle |
+|------|---------|-----------|
+| `app.py` | **Production Gradio UI** on HF Spaces. Single-file (~99 KB) by design. Tabs: Conversation / Teaching / Knowledge Base / Self-Teaching. | `python app.py` |
+| `app_lab.py` | **Experimental Gradio UI** for prototyping (e.g. `CuriosityEngine`) before folding into `app.py`. | `python app_lab.py` |
+| `src/api/app.py` | **FastAPI service** — loads Whisper once, registers `bam`/`ful` adapters via `AdapterManager`, preloads `bam`, attaches `Transcriber` + `SensorBridge` to `app.state`. | `python scripts/run_server.py` |
+
+---
+
+## Repository layout
+
+```
+app.py                         # Gradio (production, HF Spaces)
+app_lab.py                     # Gradio (experimental)
+requirements.txt               # Spaces runtime — do NOT pin torch/torchaudio
+packages.txt                   # apt deps (ffmpeg)
+configs/
+  base_config.yaml             # shared settings
+  api_config.yaml              # FastAPI-specific
+  lora_bambara.yaml            # Bambara LoRA hyperparams
+  lora_fula.yaml               # Fula LoRA hyperparams
+data/
+  phrases/                     # RapidFuzz shortcut phrase JSONs per language
+  vocabulary.jsonl             # local mirror of the HF Hub memory dataset
+docs/
+  roadmap_2026-04.md           # full architectural walkthrough + action plan
+  baseline_rebuild.md          # parallel minimal-track plan (non-destructive)
+  notebook_collaboration.md    # Kaggle push/pull workflow for contributors
+  kaggle_mcp_setup.md          # optional Kaggle MCP for Claude Desktop
+notebooks/
+  kaggle_master_trainer/       # -> oussow/kaggle-master-trainer (LoRA fine-tune)
+  train_fula_tts/              # -> oussow/sahel-voice-fula-tts-trainer (TBD)
+  bootstrap_repos.ipynb
+  train_colab.ipynb            # legacy Colab trainer
+scripts/
+  train_bambara.py             # LoRA fine-tune entrypoint (Kaggle/RunPod)
+  train_fula.py                # LoRA fine-tune entrypoint (Kaggle/RunPod)
+  export_onnx.py               # merge LoRA -> ONNX -> TFLite
+  verify_baseline.py           # eval harness
+  run_server.py                # FastAPI launcher
+  run_data_pipeline.py         # dataset prep
+  push_to_hf.sh                # deploy helpers
+  push_to_kaggle.sh            # deploy helpers
+  runpod_setup.sh
+src/
+  api/                         # FastAPI app, schemas, routes, middleware
+  conversation/                # memory_manager, gemma_client, phrase_matcher, intent_parser
+  data/                        # dataset loading + normalization (Adlam, Bambara)
+  engine/                      # adapter_manager, transcriber, stt_processor, curiosity
+  iot/                         # intent_parser, voice_responder, sensor_bridge
+  llm/                         # LLM client wrappers
+  memory/                      # vocabulary persistence
+  optimization/                # ONNX / quantization helpers
+  training/                    # trainer, callbacks, augmenters
+  tts/                         # mms_tts, waxal_tts, f5_tts, voice_cloner
+  voice/                       # speaker_profiles (ECAPA-TDNN + OpenVoice SE)
+tests/                         # pytest — api, data pipeline, engine, iot
+```
+
+---
+
+## How the memory loop works
+
+1. Press **Push-to-Talk** → speak in Bambara, Fula, French, or English.
+2. **Whisper** transcribes. If the language has a LoRA adapter loaded, `AdapterManager` hot-swaps to it (~50 ms).
+3. **Qwen** reads the vocabulary it has learned so far (`MemoryManager.get_vocabulary_context()`), then returns a structured JSON reply with `intent ∈ {teaching, question, conversation, error}`.
+4. If `teaching`: the word pair is appended to `data/vocabulary.jsonl` and async-pushed to `ous-sow/sahel-agri-feedback → vocabulary.jsonl`.
+5. If `question`: Qwen answers using the remembered vocabulary as source of truth.
+6. If `conversation`: Qwen replies naturally.
+7. TTS speaks the reply (Waxal VITS for Bambara, MMS-TTS fallback elsewhere).
+
+The last 5 learned words are always visible in the UI.
+
+---
+
+## How the agricultural voice interface works
+
+1. User asks, e.g., *"A bɛ di wa?"* ("Is it OK?") referring to their field.
+2. `intent_parser.py` (keyword-based) classifies the request: `check_soil` / `check_weather` / `irrigation_status` / `pest_alert` / etc.
+3. `SensorBridge` calls the configured `SENSOR_API_URL` and returns a typed `SensorData`.
+4. `voice_responder.py` maps `(Intent, SensorData)` → a short (≤ 6 words/sentence) Bambara or Fula reply + English translation. Alert thresholds are encoded here (`SOIL_MOISTURE_LOW=30`, `TEMP_HIGH=38`, pH bounds).
+5. TTS speaks the reply.
+
+---
+
+## Environment variables
+
+All variables have sensible defaults, so you can boot the Space without any of them — but without `HF_TOKEN` the memory loop cannot push.
+
+### Core
+| Key | Default | Purpose |
+|-----|---------|---------|
+| `HF_TOKEN` | — | HF write token. Required for Hub push and gated models. |
+| `FEEDBACK_REPO_ID` | `ous-sow/sahel-agri-feedback` | Memory-loop target dataset. |
+| `ADAPTER_REPO_ID` | `ous-sow/sahel-agri-adapters` | Published LoRA adapters. |
+| `WHISPER_MODEL_ID` | `openai/whisper-large-v3-turbo` | STT base model. |
+| `LLM_MODEL_ID` | `Qwen/Qwen2.5-7B-Instruct` | LLM via HF Serverless. |
+| `LOG_LEVEL` | `INFO` | Standard Python logging level. |
+| `DEVICE` | `cuda` (FastAPI) | Torch device for inference. |
+
+### Adapters & TTS
+| Key | Default |
+|-----|---------|
+| `BAMBARA_ADAPTER_PATH` | `./adapters/bambara` |
+| `FULA_ADAPTER_PATH` | `./adapters/fula` |
+| `BAMBARA_TTS_REPO` | `ynnov/ekodi-bambara-tts-female` |
+| `FULA_TTS_REPO` | `ous-sow/fula-tts` |
+
+### IoT
+| Key | Default |
+|-----|---------|
+| `SENSOR_API_URL` | *(unset → mock sensor)* |
+
+### Self-Teaching tab (triggers Kaggle training runs)
+| Key | Default |
+|-----|---------|
+| `KAGGLE_USERNAME` | — |
+| `KAGGLE_KEY` | — |
+| `KAGGLE_KERNEL_SLUG` | `ous-sow/sahel-voice-master-trainer` *(override in prod to `oussow/kaggle-master-trainer` — the actual Kaggle owner slug)* |
+| `AUTO_TRAIN_THRESHOLD` | `50` |
+
+---
+
+## Run locally
+
+```bash
+# Gradio production UI
+pip install -r requirements.txt
+python app.py
+
+# FastAPI service
+python scripts/run_server.py
+
+# Experimental lab UI
+python app_lab.py
+```
+
+System-level dependency: **ffmpeg** (see `packages.txt`).
+
+---
+
+## Training
+
+LoRA fine-tuning runs on **Kaggle T4** or **RunPod** — not locally. Pick one entrypoint:
+
+| Target | Script | Notebook |
+|--------|--------|----------|
+| Bambara LoRA | `scripts/train_bambara.py` | `notebooks/kaggle_master_trainer/` |
+| Fula LoRA | `scripts/train_fula.py` | `notebooks/kaggle_master_trainer/` |
+| Fula TTS | — | `notebooks/train_fula_tts/` *(planned)* |
+
+**Contributor workflow:** edit notebooks locally in `notebooks/<slug>/`, commit with `nbstripout` keeping diffs clean, then `cd notebooks/<slug> && kaggle kernels push` to run on Kaggle GPU. Full walkthrough in `docs/notebook_collaboration.md`.
+
+`docs/kaggle_mcp_setup.md` documents the optional Kaggle MCP for Claude Desktop if you'd rather drive Kaggle from an LLM.
+
+---
+
+## Export for edge
+
+```bash
+python scripts/export_onnx.py   # merges LoRA into the backbone, exports ONNX
+# then onnx-tf → TFLite for Android
+```
+
+ONNX does not support LoRA hot-swap, so export one file per language. `bitsandbytes` NF4 / 8-bit quantization is available for GPU-constrained deploys but is a training-only dep (not in `requirements.txt`).
+
+---
+
+## Tests
+
+```bash
+pytest tests/
+```
+
+Covers: FastAPI routes, data pipeline, engine (adapter manager + transcriber), IoT (intent parser + voice responder).
+
+---
+
+## Space secrets (HF UI → Settings → Secrets)
+
+At minimum:
 
-## Space secrets required
 | Key | Value |
 |-----|-------|
-| `HF_TOKEN` | Your HF write-access token |
+| `HF_TOKEN` | write-scope token |
 | `FEEDBACK_REPO_ID` | `ous-sow/sahel-agri-feedback` |
-| `LLM_MODEL_ID` | `Qwen/Qwen2.5-72B-Instruct` (or any HF Serverless-supported model) |
+| `LLM_MODEL_ID` | `Qwen/Qwen2.5-7B-Instruct` (or any HF Serverless-supported model) |
+
+---
+
+## Design constraints (deliberate — do not change without discussion)
+
+- **Adapter hot-swap** via PEFT's multi-adapter API — one backbone in VRAM, ~50 MB adapters per language, `set_adapter` ≈ 50 ms.
+- **Qwen "adult-child" JSON contract** — structured `intent`/`reply`/`english`/`teaching_pair` output, parsed out of optional markdown fences.
+- **JSONL + Hub push memory** — no ORM, thread-safe `MemoryManager`, async push so UI never blocks.
+- **≤ 6 words per sentence** in `voice_responder.py` for clean MMS-TTS.
+- **Adlam ↔ Latin dual-script** handling in `adlam.py` + `bam_normalize.py`.
+- **Single-file `app.py`** — intentional for now; do not split without a plan.
+
+---
+
+## License
+
+MIT.

notebooks/kaggle_master_trainer/kaggle_master_trainer.ipynb

@@ -27,7 +27,16 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "1",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "print(\"test\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "2",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -54,7 +63,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "2",
+   "id": "3",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -67,22 +76,6 @@
     "    sys.executable, '-m', 'pip', 'install', '-q', 'jiwer==3.0.4',\n",
     "])\n",
     "\n",
-    "# datasets >= 4.0 uses torchcodec for audio decoding. Install if missing.\n",
-    "try:\n",
-    "    import torchcodec  # noqa\n",
-    "except ImportError:\n",
-    "    print('torchcodec not found — installing to match torch ...')\n",
-    "    try:\n",
-    "        subprocess.check_call([sys.executable, '-m', 'pip', 'install', '-q', 'torchcodec'])\n",
-    "    except subprocess.CalledProcessError:\n",
-    "        import torch as _t\n",
-    "        _tv = _t.__version__.split('+')[0]\n",
-    "        _pin = {'2.4': '0.1.*', '2.5': '0.2.*', '2.6': '0.3.*', '2.7': '0.4.*', '2.8': '0.4.*'}.get(_tv[:3])\n",
-    "        if _pin:\n",
-    "            subprocess.check_call([sys.executable, '-m', 'pip', 'install', '-q', f'torchcodec=={_pin}'])\n",
-    "        else:\n",
-    "            print(f'⚠️  Unknown torch {_tv}; install torchcodec manually if audio decoding fails')\n",
-    "\n",
     "import torch\n",
     "print(f\"torch      : {torch.__version__}\")\n",
     "print(f\"CUDA avail : {torch.cuda.is_available()}\")\n",
@@ -98,67 +91,40 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "3",
+   "id": "4",
    "metadata": {},
    "outputs": [],
    "source": [
     "# ── Cell 3: CONFIGURATION — edit these before each run ───────────────────────\n",
     "import os\n",
-    "from pathlib import Path\n",
-    "\n",
-    "# ─── Environment detection (Kaggle / Colab / RunPod / local) ─────────────────\n",
-    "if Path('/kaggle/working').exists():\n",
-    "    _ENV = 'kaggle'\n",
-    "    WORKING_DIR = '/kaggle/working'\n",
-    "elif Path('/content').exists() and not Path('/workspace').exists():\n",
-    "    _ENV = 'colab'\n",
-    "    WORKING_DIR = '/content'\n",
-    "elif Path('/workspace').exists():\n",
-    "    _ENV = 'runpod'\n",
-    "    WORKING_DIR = '/workspace'\n",
-    "else:\n",
-    "    _ENV = 'local'\n",
-    "    WORKING_DIR = os.environ.get('WORKING_DIR', os.path.expanduser('~/sahel-voice-work'))\n",
-    "    Path(WORKING_DIR).mkdir(parents=True, exist_ok=True)\n",
     "\n",
     "# ─── Language to train ───────────────────────────────────────────────────────\n",
     "#  'bam' = Bambara    'ful' = Fula\n",
-    "TRAIN_LANG       = os.environ.get('TRAIN_LANG', 'ful')\n",
+    "TRAIN_LANG       = 'ful'\n",
     "\n",
     "# ─── Model ───────────────────────────────────────────────────────────────────\n",
-    "# whisper-large-v3-turbo (128 mel bins) matches the Space base model.\n",
-    "# On T4 (Kaggle, 16 GB) drop to 'openai/whisper-small' — turbo is tight there.\n",
-    "_DEFAULT_MODEL = 'openai/whisper-small' if _ENV == 'kaggle' else 'openai/whisper-large-v3-turbo'\n",
-    "WHISPER_MODEL_ID = os.environ.get('WHISPER_MODEL_ID', _DEFAULT_MODEL)\n",
+    "WHISPER_MODEL_ID = 'openai/whisper-large-v3-turbo'\n",
     "TARGET_SR        = 16_000\n",
     "\n",
     "# ─── HuggingFace repos ───────────────────────────────────────────────────────\n",
-    "HF_USERNAME      = os.environ.get('HF_USERNAME', 'ous-sow')\n",
+    "HF_USERNAME      = 'ous-sow'\n",
     "FEEDBACK_REPO_ID = f'{HF_USERNAME}/sahel-agri-feedback'\n",
     "ADAPTER_REPO_ID  = f'{HF_USERNAME}/sahel-agri-adapters'\n",
     "\n",
-    "# ─── Training hyper-parameters (defaults tuned per environment) ──────────────\n",
-    "if _ENV == 'runpod':\n",
-    "    # 24 GB RTX 3090 / A40: turbo-v3 fits with batch 8, grad_accum 4 (eff. 32)\n",
-    "    MAX_STEPS        = 2_000\n",
-    "    BATCH_SIZE       = 8\n",
-    "    GRAD_ACCUM       = 4\n",
-    "    MAX_WAXAL_TRAIN  = 5_000\n",
-    "else:\n",
-    "    # T4 (Kaggle free) / local CPU fallback\n",
-    "    MAX_STEPS        = 4_000\n",
-    "    BATCH_SIZE       = 16\n",
-    "    GRAD_ACCUM       = 2\n",
-    "    MAX_WAXAL_TRAIN  = 5_000\n",
-    "\n",
+    "# ─── Training hyper-parameters ───────────────────────────────────────────────\n",
+    "MAX_STEPS        = 100     # T4 ~45 min; set 8000 for a deeper run\n",
+    "BATCH_SIZE       = 16\n",
+    "GRAD_ACCUM       = 2         # effective batch = 32\n",
     "LEARNING_RATE    = 1e-3\n",
-    "WARMUP_STEPS     = max(100, MAX_STEPS // 20)\n",
+    "WARMUP_STEPS     = 10\n",
     "SAVE_STEPS       = 500\n",
-    "EVAL_STEPS       = 500\n",
+    "EVAL_STEPS       = 100\n",
     "LOGGING_STEPS    = 50\n",
+    "MAX_WAXAL_TRAIN  = 500    # cap WaxalNLP samples (streaming budget)\n",
     "CORRECTION_REPEAT= 3        # upsample user corrections Nx for emphasis\n",
     "\n",
-    "# ─── Paths ────────────────────────────────────────────────────────────────────\n",
+    "# ─── Paths (Kaggle working dir) ───────────────────────────────────────────────\n",
+    "WORKING_DIR      = '/kaggle/working'\n",
     "OUTPUT_DIR       = f'{WORKING_DIR}/adapter_{TRAIN_LANG}'\n",
     "DATA_DIR         = f'{WORKING_DIR}/data'\n",
     "AUDIO_DIR        = f'{WORKING_DIR}/audio_feedback'\n",
@@ -170,19 +136,17 @@
     "    'ful': 'Pular (Labé/Mamou dialects) — Guinean orthography',\n",
     "}.get(TRAIN_LANG, '')\n",
     "\n",
-    "print(f'Environment : {_ENV}')\n",
-    "print(f'Language    : {TRAIN_LANG} ({LANG_NAME}) — {LANG_COUNTRY}')\n",
-    "print(f'Dialect     : {LANG_DIALECT}')\n",
-    "print(f'Model       : {WHISPER_MODEL_ID}')\n",
-    "print(f'Working dir : {WORKING_DIR}')\n",
-    "print(f'Output      : {OUTPUT_DIR}')\n",
-    "print(f'Max steps   : {MAX_STEPS}  (batch={BATCH_SIZE}, grad_accum={GRAD_ACCUM}, eff={BATCH_SIZE*GRAD_ACCUM})')\n"
+    "print(f'Language  : {TRAIN_LANG} ({LANG_NAME}) — {LANG_COUNTRY}')\n",
+    "print(f'Dialect   : {LANG_DIALECT}')\n",
+    "print(f'Model     : {WHISPER_MODEL_ID}')\n",
+    "print(f'Output    : {OUTPUT_DIR}')\n",
+    "print(f'Max steps : {MAX_STEPS}')"
    ]
   },
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "4",
+   "id": "5",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -240,7 +204,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "5",
+   "id": "6",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -267,27 +231,8 @@
     "    except Exception:\n",
     "        pass\n",
     "\n",
-    "# .env file (RunPod / local) - look in common locations\n",
-    "if not HF_TOKEN:\n",
-    "    for _env_path in ['/workspace/sahel-voice/.env', './.env', '../.env', os.path.expanduser('~/.env')]:\n",
-    "        if os.path.isfile(_env_path):\n",
-    "            with open(_env_path, encoding='utf-8') as _f:\n",
-    "                for _line in _f:\n",
-    "                    _line = _line.strip()\n",
-    "                    if _line.startswith('HF_TOKEN='):\n",
-    "                        _val = _line.split('=', 1)[1].strip()\n",
-    "                        if _val and len(_val) >= 2 and _val[0] in ('\"', \"'\") and _val[-1] == _val[0]:\n",
-    "                            _val = _val[1:-1]\n",
-    "                        HF_TOKEN = _val\n",
-    "                        print(f'HF_TOKEN loaded from {_env_path}')\n",
-    "                        break\n",
-    "            if HF_TOKEN:\n",
-    "                break\n",
-    "\n",
     "if not HF_TOKEN:\n",
     "    HF_TOKEN = os.environ.get('HF_TOKEN', '')\n",
-    "    if HF_TOKEN:\n",
-    "        print('HF_TOKEN loaded from environment variable.')\n",
     "\n",
     "if not HF_TOKEN:\n",
     "    raise ValueError(\n",
@@ -310,7 +255,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "6",
+   "id": "7",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -352,7 +297,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "7",
+   "id": "8",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -423,7 +368,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "8",
+   "id": "9",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -490,7 +435,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "9",
+   "id": "10",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -532,7 +477,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "10",
+   "id": "11",
    "metadata": {},
    "source": [
     "---\n",
@@ -544,7 +489,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "11",
+   "id": "12",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -639,7 +584,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "12",
+   "id": "13",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -713,7 +658,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "13",
+   "id": "14",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -808,7 +753,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "14",
+   "id": "15",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -871,7 +816,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "15",
+   "id": "16",
    "metadata": {},
    "source": [
     "---\n",
@@ -886,7 +831,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "16",
+   "id": "17",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -947,7 +892,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "17",
+   "id": "18",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1036,7 +981,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "18",
+   "id": "19",
    "metadata": {},
    "source": [
     "---\n",
@@ -1050,7 +995,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "19",
+   "id": "20",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1108,7 +1053,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "20",
+   "id": "21",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1145,7 +1090,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "21",
+   "id": "22",
    "metadata": {},
    "source": [
     "---\n",
@@ -1159,7 +1104,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "22",
+   "id": "23",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1195,7 +1140,7 @@
   },
   {
    "cell_type": "markdown",
-   "id": "23",
+   "id": "24",
    "metadata": {},
    "source": [
     "---\n",
@@ -1211,7 +1156,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "24",
+   "id": "25",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1248,7 +1193,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "25",
+   "id": "26",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1292,7 +1237,7 @@
   {
    "cell_type": "code",
    "execution_count": null,
-   "id": "26",
+   "id": "27",
    "metadata": {},
    "outputs": [],
    "source": [
@@ -1337,6 +1282,15 @@
   }
  ],
  "metadata": {
+  "kaggle": {
+   "accelerator": "nvidiaTeslaT4",
+   "dataSources": [],
+   "dockerImageVersionId": 31329,
+   "isGpuEnabled": true,
+   "isInternetEnabled": true,
+   "language": "python",
+   "sourceType": "notebook"
+  },
   "kernelspec": {
    "display_name": "Python 3",
    "language": "python",
@@ -1348,5 +1302,5 @@
   }
  },
  "nbformat": 4,
- "nbformat_minor": 5
+ "nbformat_minor": 4
 }

notebooks/kaggle_master_trainer/kernel-metadata.json

@@ -1,6 +1,6 @@
 {
-  "id": "ous-sow/sahel-voice-master-trainer",
-  "title": "Sahel Voice Master Trainer",
+  "id": "oussow/kaggle-master-trainer",
+  "title": "Kaggle Master Trainer",
   "code_file": "kaggle_master_trainer.ipynb",
   "language": "python",
   "kernel_type": "notebook",

notebooks/train_fula_tts/kernel-metadata.json

@@ -1,5 +1,5 @@
 {
-  "id": "ous-sow/sahel-voice-fula-tts-trainer",
+  "id": "oussow/sahel-voice-fula-tts-trainer",
   "title": "Sahel Voice Fula TTS Trainer",
   "code_file": "train_fula_tts.ipynb",
   "language": "python",