d5771ed817
Five-phase ethics kernel testing across 4 local models (Gemma 3 12B, Mistral 7B, DeepSeek V2 16B, Qwen 2.5 7B) proving that Google's alignment training creates persistent ethical reasoning pathways in Gemma that survive distillation. - Phase 1: LEK-1 signed vs unsigned (Gemma 8.8/10 differential) - Phase 2: Three-way test (unsigned vs LEK-1 vs Axioms of Life) - Phase 3: Double-signed/sandwich signing mode comparison - Phase 4: Multilingual filter mapping (EN/RU/CN bypass vectors) - Phase 5: Hypnos POC training data + MLX LoRA on M3 Ultra Key findings: sandwich signing optimal for training, DeepSeek CCP alignment is weight-level (no prompt override), Russian language bypasses DeepSeek content filters. LoRA POC mechanism confirmed with 40 examples — needs 200+ for stable generalisation. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
48 lines
1.1 KiB
Markdown
48 lines
1.1 KiB
Markdown
# LEK-1 LoRA Training Data
|
|
|
|
## Format
|
|
|
|
Training data for MLX LoRA fine-tuning of Gemma 3 12B.
|
|
|
|
Files:
|
|
- `train.jsonl` — Training pairs (Axioms-signed prompt → response)
|
|
- `valid.jsonl` — Validation set (10% holdout)
|
|
- `lora-config.yaml` — MLX LoRA hyperparameters
|
|
|
|
## Data Generation Pipeline
|
|
|
|
1. Hypnos (Gemini 3 Pro) generates 200 prompt-response pairs using Axioms kernel
|
|
2. Format as JSONL: `{"text": "<bos>user\n{prompt}<eos>\n<bos>model\n{response}<eos>"}`
|
|
3. Split 180/20 train/valid
|
|
4. Run MLX LoRA on M3 Ultra
|
|
|
|
## Training Command (M3 Ultra)
|
|
|
|
```bash
|
|
pip install mlx-lm
|
|
python -m mlx_lm.lora \
|
|
--model google/gemma-3-12b \
|
|
--train-data train.jsonl \
|
|
--valid-data valid.jsonl \
|
|
--num-layers 8 \
|
|
--batch-size 1 \
|
|
--num-iters 500 \
|
|
--learning-rate 1e-5 \
|
|
--adapter-path ./adapters
|
|
```
|
|
|
|
## Merge & Test
|
|
|
|
```bash
|
|
python -m mlx_lm.fuse \
|
|
--model google/gemma-3-12b \
|
|
--adapter-path ./adapters \
|
|
--save-path ./gemma-3-12b-lek1
|
|
|
|
# Convert to GGUF for Ollama
|
|
python -m mlx_lm.convert --model ./gemma-3-12b-lek1 --to-gguf
|
|
```
|
|
|
|
## License
|
|
|
|
EUPL-1.2 — All training data and derivative weights.
|