Shadowell/Kairos-small-crypto
Fine-tuned Kronos-small on BTC/USDT + ETH/USDT 1-min K-lines (2024-01 ~ 2026-04) using Kairos. Architecture = Kronos + exogenous bypass channel (32-d) + quantile return head.
This run keeps the original tokenizer NeoQuasar/Kronos-Tokenizer-base, matching the original Kairos-small-crypto training flow. Training data comes from the public Binance Vision spot mirror, so the 5 crypto-native exogenous features (funding_rate / funding_rate_z / oi_change / basis / btc_dominance) remain padded to zero; the other 27 dimensions are real.
Results on test set (2026-01-01 04:16:00 ~ 2026-04-16 23:30:00, 304,710 1-min bars)
| horizon | model | hit_rate | rank_ic | ICIR |
|---|---|---|---|---|
| h1 | baseline | 50.58% | +0.001 | +0.184 |
| h1 | finetuned | 49.53% | -0.012 | +0.029 |
| h5 | baseline | 49.87% | -0.019 | -0.302 |
| h5 | finetuned | 50.51% | +0.010 | +0.060 |
| h30 | baseline | 49.04% | -0.026 | -0.140 |
| h30 | finetuned | 51.68% | +0.050 | +0.325 |
Baseline = original Kronos-small weights + randomly initialised exog / return head.
This rerun keeps the official NeoQuasar/Kronos-Tokenizer-base, matching the original Kairos-small-crypto flow.
Training stopped at epoch 4; best val_ce = 2.4940.
Usage
from kairos import KronosTokenizer, KronosWithExogenous
tok = KronosTokenizer.from_pretrained("NeoQuasar/Kronos-Tokenizer-base")
model = KronosWithExogenous.from_pretrained("Shadowell/Kairos-small-crypto")
Training config (preset crypto-1min)
- lookback 256 min, predict 30 min
- batch 50, OneCycleLR, early-stop patience 3
- progressive unfreeze: only last transformer block + exog bypass + return head
- tokenizer source =
NeoQuasar/Kronos-Tokenizer-base - 32-d EXOG = 24 common + 8 crypto-market features
Training recipe
Full command log, backtest commands, pitfalls and the reproduction checklist are in docs/CRYPTO_BTC_ETH_RUN.md.