Pre-launch· Token not live yet — no rewards or treasury until launchWhat this means
← Model
All market signals

TRAIN v0.03

Adds trade-flow, holder-estimate, creator-history and social features (all 36 features in set f1).
Completedpreviously deployed

Definition

Model type
Linear model
Signals used
Market signals
Technical details
C
0.5
Random seed
1003
Config hash
9e10425f44094916
Created
05 Oct 03:24
Completed
05 Oct 03:24
Artifact
models/v0.03/model.json
sha256 c726e4cd37e61ccf…

Dataset & compute

Dataset
Solana Launch Dataset v2
115 snapshots · 23 launches · sha256 30708f586621
Training runs
1
Compute
TRAIN training server (CPU)
GPU hours
—
no GPU used
CPU time
4.3 s
Provider cost
$0.00
nothing purchased

Benchmarks

held-out test set · change vs v0.02
Spotting crashes (−70% within 1h)
0.450
Spotting survivors (still traded after 1h)
0.865
Spotting 2× runs (within 1h)
0.651
Calling the 1h direction (down / flat / up)
36.5%
Spotting creator dumps (within 1h)
0.607
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
—
n=5
collapse_6h
—
n=0
reach_2x_1h
—
n=5
survival_1h
—
n=5
survival_6h
—
n=0
survival_24h
—
n=0
migration_24h
—
n=0
trajectory_1h
acc 100.0%
n=5
trajectory_6h
—
n=0
creator_exit_1h
—
n=1

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 03:24 · 00:00:00 · 110 examples · $0.00

Timeline

  1. 05 Oct 06:05
    TRAIN v0.03 deployed — now live
    average score across 1 tests 0.333 vs 0.333 for the live v0.01
  2. 05 Oct 03:24
    TRAIN v0.03 not deployed; v0.01 stays live
    Reason: no shared benchmarks yet; keeping the live model.
  3. 05 Oct 03:24
    Benchmark completed: TRAIN v0.03
    No locked test set yet (not enough held-out launches with observed outcomes). Benchmarks will be computed when it is locked.
  4. 05 Oct 03:24
    Benchmark started: TRAIN v0.03
  5. 05 Oct 03:24
    Training completed: TRAIN v0.03
    00:00:00 · trained on 110 examples · $0 compute cost
  6. 05 Oct 03:24
    Training started: TRAIN v0.03
    Solana Launch Dataset v2 · linear model · TRAIN training server
  7. 05 Oct 03:24
    TRAIN v0.03 queued
    Adds trade-flow, holder-estimate, creator-history and social features (all 36 features in set f1).