Pre-launch· Token not live yet — no rewards or treasury until launchWhat this means
Activity

Everything TRAIN does, in order

Training runs, benchmarks, deployments, dataset updates, data-provider failures — and, once the token exists, rewards and compute purchases. Everything here really happened.
EverythingMoneyTrainingBenchmarks & deploysData
  1. 06 Oct 01:01
    Data provider recovered: pumpfun
  2. 06 Oct 01:00
    Data provider failing: pumpfun
    [pumpfun] timeout
  3. 06 Oct 00:45
    TRAIN v0.25 not deployed; v0.19 stays live
    Reason: average score across 5 tests 0.844 vs 0.845 for the live v0.19.
  4. 06 Oct 00:45
    Benchmark completed: TRAIN v0.25
    Spotting crashes (−70% within 1h) regressed 0.955 → 0.944 vs live v0.19.
  5. 06 Oct 00:45
    Benchmark started: TRAIN v0.25
  6. 06 Oct 00:45
    Training completed: TRAIN v0.25
    00:02:55 · trained on 9081 examples · $0 compute cost
  7. 06 Oct 00:42
    Training started: TRAIN v0.25
    Solana Launch Dataset v17 · decision-tree ensemble · TRAIN training server (CPU)
  8. 06 Oct 00:42
    Dataset updated: Solana Launch Dataset v17
    10,305 snapshots from 2,068 launches.
  9. 06 Oct 00:42
    TRAIN v0.25 queued
    Same configuration as v0.18 (best so far on the headline benchmark), retrained on the newest dataset.
  10. 06 Oct 00:42
    Enough new data for another training run
    6,368 new labelled snapshots since Solana Launch Dataset v9.
  11. 06 Oct 00:01
    Tracked launch migrated: PACEORBIT
    Recorded as an outcome label (never used as a feature).
  12. 05 Oct 23:57
    Tracked launch migrated: HUG
    Recorded as an outcome label (never used as a feature).
  13. 05 Oct 22:47
    TRAIN v0.24 not deployed; v0.19 stays live
    Reason: average score across 5 tests 0.843 vs 0.845 for the live v0.19.
  14. 05 Oct 22:47
    Benchmark completed: TRAIN v0.24
    Crash-spotting score regressed 0.955 → 0.944 vs live v0.19.
  15. 05 Oct 22:47
    Benchmark started: TRAIN v0.24
  16. 05 Oct 22:47
    TRAIN v0.24 discovered 1 useful new signal
    Change between snapshots: share of tiny (<$3) trades. Score on held-out training launches 0.901 → 0.905.
  17. 05 Oct 22:47
    Training completed: TRAIN v0.24
    00:06:25 · trained on 7875 examples · $0 compute cost
  18. 05 Oct 22:41
    Training started: TRAIN v0.24
    Solana Launch Dataset v16 · decision-tree ensemble · TRAIN training server
  19. 05 Oct 22:41
    Dataset updated: Solana Launch Dataset v16
    8,939 snapshots from 1,794 launches.
  20. 05 Oct 22:41
    TRAIN v0.24 queued
    Adds global fees paid : total fees and Jito tips, fees per minute, fees relative to volume, transactions per paying wallet and failed-transaction share — then lets automatic discovery choose which change-signals to add on top. Retrained on the newest data.
  21. 05 Oct 22:41
    Benchmark set locked: Held-out test set v8
    532 held-out snapshots from 256 launches. No model can train on these.
  22. 05 Oct 22:40
    TRAIN v0.23 not deployed; v0.19 stays live
    Reason: average score across 5 tests 0.842 vs 0.850 for the live v0.19.
  23. 05 Oct 22:40
    Benchmark completed: TRAIN v0.23
    Crash-spotting score regressed 0.949 → 0.937 vs live v0.19.
  24. 05 Oct 22:40
    Benchmark started: TRAIN v0.23
  25. 05 Oct 22:40
    Training completed: TRAIN v0.23
    00:02:25 · trained on 8590 examples · $0 compute cost
  26. 05 Oct 22:38
    Training started: TRAIN v0.23
    Solana Launch Dataset v15 · decision-tree ensemble · TRAIN training server
  27. 05 Oct 22:38
    Dataset updated: Solana Launch Dataset v15
    9,754 snapshots from 1,957 launches.
  28. 05 Oct 22:38
    TRAIN v0.23 queued
    Same configuration as v0.18 (best so far on the headline benchmark), retrained on the newest dataset.
  29. 05 Oct 21:24
    TRAIN v0.22 not deployed; v0.19 stays live
    Reason: average score across 5 tests 0.841 vs 0.850 for the live v0.19.
  30. 05 Oct 21:24
    Benchmark completed: TRAIN v0.22
    Crash-spotting score regressed 0.949 → 0.942 vs live v0.19.
  31. 05 Oct 21:24
    Benchmark started: TRAIN v0.22
  32. 05 Oct 21:24
    Training completed: TRAIN v0.22
    00:02:13 · trained on 8141 examples · $0 compute cost
  33. 05 Oct 21:22
    Training started: TRAIN v0.22
    Solana Launch Dataset v14 · decision-tree ensemble · TRAIN training server
  34. 05 Oct 21:22
    Dataset updated: Solana Launch Dataset v14
    9,235 snapshots from 1,853 launches.
  35. 05 Oct 21:22
    TRAIN v0.22 queued
    Same configuration as v0.18 (best so far on the headline benchmark), retrained on the newest dataset.
  36. 05 Oct 21:22
    Enough new data for another training run
    5,448 new labelled snapshots since Solana Launch Dataset v9.
  37. 05 Oct 19:46
    Data provider recovered: helius
  38. 05 Oct 19:46
    Data provider recovered: pumpfun
  39. 05 Oct 19:46
    Data provider recovered: pumpfun
  40. 05 Oct 19:45
    Data provider failing: pumpfun
    [pumpfun] fetch failed
  41. 05 Oct 19:45
    Data provider failing: helius
    [helius] fetch failed
  42. 05 Oct 19:23
    TRAIN v0.21 not deployed; v0.19 stays live
    Reason: average score across 1 tests 0.439 vs 0.584 for the live v0.19.
  43. 05 Oct 19:23
    Benchmark completed: TRAIN v0.21
    Crash-spotting score: not enough held-out data yet (n=67).
  44. 05 Oct 19:23
    Benchmark started: TRAIN v0.21
  45. 05 Oct 19:23
    TRAIN v0.21 discovered 3 useful new signals
    Change between snapshots: share of tiny (<$3) trades; Change between snapshots: buyers with a profitable record; Change between snapshots: price change, last minute. Score on held-out training launches 0.827 → 0.860.
  46. 05 Oct 19:23
    Training completed: TRAIN v0.21
    00:02:20 · trained on 1808 examples · $0 compute cost
  47. 05 Oct 19:21
    Training started: TRAIN v0.21
    Solana Launch Dataset v13 · decision-tree ensemble · TRAIN training server
  48. 05 Oct 19:21
    Dataset updated: Solana Launch Dataset v13
    2,018 snapshots from 406 launches.
  49. 05 Oct 19:21
    TRAIN v0.21 queued
    Adds global fees paid : total fees and Jito tips, fees per minute, fees relative to volume, transactions per paying wallet and failed-transaction share — then lets automatic discovery choose which change-signals to add on top.
  50. 05 Oct 19:21
    Enough new data for another training run
    4,758 new labelled snapshots since Solana Launch Dataset v9.
  51. 05 Oct 18:20
    Tracked launch migrated: Melty
    Recorded as an outcome label (never used as a feature).
  52. 05 Oct 18:01
    Tracked launch migrated: INSTINCT
    Recorded as an outcome label (never used as a feature).
  53. 05 Oct 17:39
    Tracked launch migrated: SI
    Recorded as an outcome label (never used as a feature).
  54. 05 Oct 17:24
    TRAIN v0.20 not deployed; v0.19 stays live
    Reason: average score across 5 tests 0.842 vs 0.850 for the live v0.19.
  55. 05 Oct 17:24
    Benchmark completed: TRAIN v0.20
    Crash-spotting score improved 0.949 → 0.954 vs live v0.19.
  56. 05 Oct 17:24
    Benchmark started: TRAIN v0.20
  57. 05 Oct 17:24
    TRAIN v0.20 discovered 1 useful new signal
    Change in buy share of volume since launch since the previous snapshot. Score on held-out training launches 0.907 → 0.911.
  58. 05 Oct 17:24
    Training completed: TRAIN v0.20
    00:03:04 · trained on 6421 examples · $0 compute cost
  59. 05 Oct 17:21
    Tracked launch migrated: cloud
    Recorded as an outcome label (never used as a feature).
  60. 05 Oct 17:21
    Training started: TRAIN v0.20
    Solana Launch Dataset v12 · decision-tree ensemble · TRAIN training server
  61. 05 Oct 17:21
    Dataset updated: Solana Launch Dataset v12
    7,235 snapshots from 1,452 launches.
  62. 05 Oct 17:21
    TRAIN v0.20 queued
    Automatic signal discovery: TRAIN generated candidate signals (how every measure changed since the previous snapshot), tested each on held-out folds of the training launches, and kept only the ones that improved its predictions.
  63. 05 Oct 17:21
    TRAIN v0.19 deployed — now live
    average score across 5 tests 0.850 vs 0.843 for the live v0.18
  64. 05 Oct 17:21
    Benchmark completed: TRAIN v0.19
    Crash-spotting score regressed 0.956 → 0.949 vs live v0.18.
  65. 05 Oct 17:21
    Benchmark started: TRAIN v0.19
  66. 05 Oct 17:21
    Training completed: TRAIN v0.19
    00:03:24 · trained on 6411 examples · $0 compute cost
  67. 05 Oct 17:17
    Training started: TRAIN v0.19
    Solana Launch Dataset v11 · decision-tree ensemble · TRAIN training server
  68. 05 Oct 17:17
    Dataset updated: Solana Launch Dataset v11
    7,225 snapshots from 1,450 launches.
  69. 05 Oct 17:17
    TRAIN v0.19 queued
    Same signals with deeper trees, to learn interactions such as 'bundle still holding while snipers exit'.
  70. 05 Oct 17:09
    TRAIN v0.18 deployed — now live
    average score across 5 tests 0.843 vs 0.822 for the live v0.09
  71. 05 Oct 17:09
    Benchmark completed: TRAIN v0.18
    Crash-spotting score: 0.956 (n=491).
  72. 05 Oct 17:09
    Benchmark started: TRAIN v0.18
  73. 05 Oct 17:09
    Training completed: TRAIN v0.18
    00:01:39 · trained on 6361 examples · $0 compute cost
  74. 05 Oct 17:07
    Training started: TRAIN v0.18
    Solana Launch Dataset v10 · decision-tree ensemble · TRAIN training server
  75. 05 Oct 17:07
    Benchmark set locked: Held-out test set v7
    491 held-out snapshots from 226 launches. No model can train on these.
  76. 05 Oct 17:07
    Dataset updated: Solana Launch Dataset v10
    7,170 snapshots from 1,439 launches.
  77. 05 Oct 17:07
    TRAIN v0.18 queued
    Adds launch structure — bundled buys in the creation slot, snipers in the first 10 seconds and how much they've sold, first-minute buying, organic-volume checks (volume per wallet, back-and-forth wallets, tiny bump trades), KOL buys, and the dev's launches across all pump.fun coins.
  78. 05 Oct 16:59
    Tracked launch migrated: MOCHI
    Recorded as an outcome label (never used as a feature).
  79. 05 Oct 16:01
    Tracked launch migrated: WE
    Recorded as an outcome label (never used as a feature).
  80. 05 Oct 15:23
    Tracked launch migrated: CRIME
    Recorded as an outcome label (never used as a feature).
  81. 05 Oct 15:15
    Tracked launch migrated: OFF
    Recorded as an outcome label (never used as a feature).
  82. 05 Oct 15:12
    TRAIN v0.17 not deployed; v0.09 stays live
    Reason: average score across 5 tests 0.797 vs 0.812 for the live v0.09.
  83. 05 Oct 15:12
    Benchmark completed: TRAIN v0.17
    Crash-spotting score improved 0.911 → 0.924 vs live v0.09.
  84. 05 Oct 15:12
    Benchmark started: TRAIN v0.17
  85. 05 Oct 15:12
    Training completed: TRAIN v0.17
    00:00:04 · trained on 3502 examples · $0 compute cost
  86. 05 Oct 15:12
    Training started: TRAIN v0.17
    Solana Launch Dataset v9 · decision-tree ensemble · TRAIN training server
  87. 05 Oct 15:12
    TRAIN v0.17 queued
    Same configuration as v0.04 (best so far on the headline benchmark), retrained on the newest dataset.
  88. 05 Oct 14:01
    TRAIN v0.16 not deployed; v0.09 stays live
    Reason: average score across 5 tests 0.798 vs 0.812 for the live v0.09.
  89. 05 Oct 14:01
    Benchmark completed: TRAIN v0.16
    Crash-spotting score improved 0.911 → 0.918 vs live v0.09.
  90. 05 Oct 14:01
    Benchmark started: TRAIN v0.16
  91. 05 Oct 14:01
    Training completed: TRAIN v0.16
    00:00:04 · trained on 3502 examples · $0 compute cost
  92. 05 Oct 14:01
    Training started: TRAIN v0.16
    Solana Launch Dataset v9 · decision-tree ensemble · TRAIN training server
  93. 05 Oct 14:01
    TRAIN v0.16 queued
    Same configuration as v0.04 (best so far on the headline benchmark), retrained on the newest dataset.
  94. 05 Oct 13:07
    Tracked launch migrated: BCAT
    Recorded as an outcome label (never used as a feature).
  95. 05 Oct 13:00
    TRAIN v0.15 not deployed; v0.09 stays live
    Reason: average score across 5 tests 0.801 vs 0.812 for the live v0.09.
  96. 05 Oct 13:00
    Benchmark completed: TRAIN v0.15
    Crash-spotting score improved 0.911 → 0.930 vs live v0.09.
  97. 05 Oct 13:00
    Benchmark started: TRAIN v0.15
  98. 05 Oct 13:00
    Training completed: TRAIN v0.15
    00:00:04 · trained on 3502 examples · $0 compute cost
  99. 05 Oct 13:00
    Training started: TRAIN v0.15
    Solana Launch Dataset v9 · decision-tree ensemble · TRAIN training server
  100. 05 Oct 13:00
    TRAIN v0.15 queued
    Same configuration as v0.04 (best so far on the headline benchmark), retrained on the newest dataset.
Older →