-
Full data sync: 953K+ records, clean telemetry, model parameters
- Remove deprecated fresh_sync_data.jsonl (8 records, caused degenerate weights)
- Add qubic_ticks_snn.jsonl (27K, primary training data)
- Add cleaned node_sync_harvest.jsonl (120K, 9 dead cols removed)
- Add cleaned neuromorphic_data.jsonl (814K, 11 dead cols removed, 396MB→14MB)
- Add ghost_market_log.jsonl (31K HFT trading log)
- Add samples (mining + HFT, 100 and 1K variants)
- Add model parameters (mining_v2 .mem files)
- Add training config (mining_v2.toml)
- Rewrite README.md to match actual file structure
- Update dataset_card.json (size_categories: 100K<n<1M)
