Pre-launch· Token not live yet — no rewards or treasury until launchWhat this means
← Model
Deeper decision trees — retrained on new data

TRAIN v0.08

Same configuration as v0.05 (best so far on the headline benchmark), retrained on the newest dataset.
Completedpreviously deployed

Definition

Model type
Decision-tree ensemble
Signals used
Market signals
Technical details
max_depth
3
subsample
0.8
n_estimators
300
learning_rate
0.03
min_samples_leaf
10
Random seed
1008
Config hash
77db92bc84736875
Created
05 Oct 08:25
Completed
05 Oct 08:25
Artifact
models/v0.08/model.json
sha256 ee7ea6c75ee455f7…

Dataset & compute

Dataset
Solana Launch Dataset v7
3,122 snapshots · 626 launches · sha256 8f5197a2a59b
Training runs
1
Compute
TRAIN training server (CPU)
GPU hours
—
no GPU used
CPU time
15.1 s
Provider cost
$0.00
nothing purchased

Benchmarks

held-out test set · change vs v0.07
Spotting crashes (−70% within 1h)
0.926
Spotting survivors (still traded after 1h)
0.907
Spotting 2× runs (within 1h)
0.862
Calling the 1h direction (down / flat / up)
56.5%
Spotting creator dumps (within 1h)
0.864
Spotting crashes (within 6h)
insufficient data
Spotting survivors (6h)
insufficient data
Calling the 6h direction
insufficient data
Spotting survivors (24h)
insufficient data
Spotting migrations (within 24h)
insufficient data

Validation metrics (during training)

collapse_1h
AUC 0.994
n=370
collapse_6h
—
n=0
reach_2x_1h
AUC 0.972
n=370
survival_1h
AUC 0.848
n=370
survival_6h
—
n=0
survival_24h
—
n=0
migration_24h
—
n=0
trajectory_1h
acc 96.5%
n=370
trajectory_6h
—
n=0
creator_exit_1h
AUC 0.787
n=108

Validation launches are separate from training launches and from the locked test set. Benchmarks above are the public numbers.

What this version relies on

held-out validation launches
This version was trained before TRAIN started measuring which signals each model relies on.

For each signal, how much the score drops when that signal is scrambled across launches. TRAIN learns these weights itself from the data; nobody hand-codes them.

Training runs

Attempt 1Succeeded
05 Oct 08:25 · 00:00:11 · 2,752 examples · $0.00

Timeline

  1. 05 Oct 08:25
    TRAIN v0.08 deployed — now live
    average score across 2 tests 0.607 vs 0.602 for the live v0.04
  2. 05 Oct 08:25
    Benchmark completed: TRAIN v0.08
    1h trajectory · balanced accuracy regressed 0.485 → 0.420 vs live v0.04.
  3. 05 Oct 08:25
    Benchmark started: TRAIN v0.08
  4. 05 Oct 08:25
    Training completed: TRAIN v0.08
    00:00:11 · trained on 2752 examples · $0 compute cost
  5. 05 Oct 08:25
    Training started: TRAIN v0.08
    Solana Launch Dataset v7 · decision-tree ensemble · TRAIN training server
  6. 05 Oct 08:25
    TRAIN v0.08 queued
    Same configuration as v0.05 (best so far on the headline benchmark), retrained on the newest dataset.