Mega-Trading
Backtest Dashboard
Self-contained report generated from local training, MLflow, dataset split, and backtest artifacts.
Run
rtx
Mixture
rtx
Checkpoint
present
Steps
2,500
Train loss
3.3414
Validation loss
2.3500
Backtest loss
2.3379
Backtest ppl
10.3592
Artifacts
| Checkpoint | .mega-trading/data/runs/rtx/checkpoint.pt | present |
| Metrics | .mega-trading/data/runs/rtx/metrics.json | present |
| MLflow DB | .mega-trading/data/runs/mlflow/mlflow.db | present |
| Backtest JSON | .mega-trading/data/evals/rtx/backtest.json | present |
Prepared Split
Chronological per-ticker split from prepared dataset metadata.
Training Curves
train loss | validation loss
Green: train loss. Blue: validation loss when evaluated.
Ticker Candles + Forecast
Green/red candles show sampled real Binance trades from a held-out archive. Blue line is the model-implied price path from generated relative-price tokens seeded on the same ticker stream.
BTCUSDT 2026-04
Green/red: real OHLC candles. Blue: generated model-implied price path.
ETHUSDT 2026-04
Green/red: real OHLC candles. Blue: generated model-implied price path.
BNBUSDT 2026-04
Green/red: real OHLC candles. Blue: generated model-implied price path.
Backtest Metrics
| Loss | 2.3379 |
|---|---|
| Perplexity | 10.3592 |
| Top-1 accuracy | 49.23% |
| Top-5 accuracy | 69.42% |
| Scored tokens | 524,288 |
| Distribution L1 | 0.1420 |
Stylized Facts
| Metric | Real | Generated |
|---|---|---|
| abs_return_autocorrelation_lag1 | -0.0001 | -0.0005 |
| count | 16384.0000 | 16384.0000 |
| excess_kurtosis | 3.0948 | 14.2966 |
| extreme_fraction | 0.0000 | 0.0000 |
| mean_bps | -0.7012 | -0.8331 |
| return_autocorrelation_lag1 | -0.0696 | -0.0057 |
| std_bps | 0.6272 | 0.4405 |
Run Configuration
| Vocab size | 579 |
|---|---|
| Block size | 512 |
| Sequence count | 62,012,828 |
| Training backend | accelerate |
| Distributed strategy | ddp |
| World size | 1 |
| Attention backend | triton |
| Compile enabled | True |