FIXED-LINE FEEDS

Tournament data, obfuscated.

Anonymised features, rank-normalised targets, served as Parquet. Refreshed per round — open Sunday 18:00 UTC, close Friday 20:00 UTC.

35 features across 1 groups · 0 targets

Available datasets
Tier 1 · Production

Himalayas · Futures · v1.5

146 instruments across 8 clusters. 18M rows. Train/val/live split.

Tier 2 · Beta

Alps · Equities · v0.4

S&P 500 universe. 6,250 rows per era. Obfuscated factor returns.

Schema →
Reference · Weekly refresh

Benchmark models

Pre-trained reference models — ridge, LightGBM, ensemble. Updated weekly.

Live · Read-only stream

Live predictions

Top model predictions broadcast on a 30-minute delay. Read-only.

All dataset files · Himalayas / Futures
Download all (.zip)
Core splits
eiq_train.parquet
Training split — obfuscated features + 16 targets, full history.
Download →
eiq_validation.parquet
Validation split — out-of-sample features + targets.
Download →
eiq_live.parquet
Live split — current-round features (targets are placeholders until the round matures).
Download →
Benchmark models
eiq_train_benchmark_models.parquet
Benchmark-model predictions over the train split.
Download →
eiq_validation_benchmark_models.parquet
Benchmark-model predictions over the validation split.
Download →
eiq_live_benchmark_models.parquet
Benchmark-model predictions for the current round.
Download →
Example predictions & Minera model
eiq_validation_example_preds.parquet
Example predictions over the validation split.
Download →
eiq_live_example_preds.parquet
Example predictions for the current round.
Download →
eiq_minera_model.parquet
Minera model predictions for the current round (the stake-weighted consensus).
Download →
Metadata
eiq_features.json
Feature metadata — feature sets (small / medium / all) and the target list.
Download →
eiq_metadata.json
Build metadata — dataset version, build timestamp, and git commit.
Download →

Files refresh per round — open Sunday 18:00 UTC, close Friday 20:00 UTC. Live artifacts (the live split, live benchmark models, live example predictions, and the Minera model) update daily.

Schema · Himalayas / Futures

Column definitions

era                    string       2026-W18
exped_id               string       2026-W18-001
instrument_id          string       SPY-FUT-FRONT
cluster                string       eq
feature_alpine_001     float64      [-1.0, 1.0] rank-normalized
feature_alpine_002     float64      [-1.0, 1.0] rank-normalized
...
target_everest_20      float64      [-1.0, 1.0] 20-day forward return
target_k2_20           float64      [-1.0, 1.0] auxiliary
target_lhotse_20       float64      [-1.0, 1.0] auxiliary
predecessor_chain      string|null  SPY-MAR-2026
data_split             string       train|val|live
API access

Pull data from your terminal

Python
import everestapi as eq

client = eq.Client(api_key="ekq_...")
df = client.dataset.futures_himalayas_v1_5(
    split="train",
    format="parquet",
)
curl
curl -H "Authorization: Bearer $EIQ_API_KEY" \
  https://api.everesteer.ai/v1/dataset/himalayas/v1.5/train.parquet \
  -o train.parquet