Files
CosmicClash/training/curriculum_state_gen3.json
T
Josh Creek 0e046aa9f1 chore(training): scrap generation 1-3 training data for generation 4
All checkpoints/logs/exported policies here are a continuous-Gaussian,
31-input action/observation shape that generation 4's MultiDiscrete
redesign is structurally incompatible with -- nothing to resume from (see
the prior commit and TRAINING.md's "Generation 4" section). Game/bots/promoted/
(easy.json, and the new reference-grounded.json copied from
curric-s5-aggression before this) is untouched -- both remain valid,
playable evaluation opponents forever via PolicyNetwork's format-versioned
JSON despite their own checkpoints/generation being gone.

- training/checkpoints/*, training/logs/* removed (~3.5GB of working tree,
  all generation 1-3 experiment runs).
- Game/bots/*.json flat dump removed (superseded exports; main_menu.gd's
  Spectate dropdown will just be empty until the first generation-4 export).
- curriculum_state.json -> curriculum_state_gen3.json, archived alongside
  the existing _gen1/_gen2 logs (all three are referenced as postmortem
  evidence in TRAINING.md/curriculum.py). A fresh curriculum_state.json
  will be created on the next curriculum.py run (load_state() already
  handles a missing file).

training/eval_history.json is deliberately NOT reset -- it's the one
continuous cross-generation progress record.

NOT YET PUSHED: this needs the remote Linux training box quiesced first
(kill any active tmux session, confirm it's synced to origin) so its own
run_training.sh doesn't race a still-running job's final commit against
this deletion.
2026-08-04 23:28:57 +01:00

61 lines
1.7 KiB
JSON

{
"stage_index": 3,
"attempt": 2,
"status": "in_progress",
"log": [
{
"stage_index": 0,
"experiment": "20260731-2149-curric-s1-unmask-ramp25",
"attempt": 0,
"decision": "pass",
"note": "ungated ramp waypoint (no eval)"
},
{
"stage_index": 1,
"experiment": "20260801-0223-curric-s2-unmask-ramp50",
"attempt": 0,
"decision": "pass",
"note": "ungated ramp waypoint (no eval)"
},
{
"stage_index": 2,
"experiment": "20260801-0658-curric-s3-unmask-ramp75",
"attempt": 0,
"decision": "pass",
"note": "ungated ramp waypoint (no eval)"
},
{
"stage_index": 3,
"experiment": "20260801-1131-curric-s4-unmask",
"attempt": 0,
"eval": {
"timestamp": "2026-08-02T13:58:25+00:00",
"model_a": "/home/jcreek/ai-training/CosmicClash/Game/bots/20260801-1131-curric-s4-unmask.json",
"model_b": "/home/jcreek/ai-training/CosmicClash/Game/bots/curric-s5-aggression.json",
"episodes": 100,
"wins_a": 29,
"wins_b": 55,
"draws": 16,
"win_rate_a": 0.29
},
"decision": "fail"
},
{
"stage_index": 3,
"experiment": "20260802-1458-curric-s4-unmask-retry1",
"attempt": 1,
"eval": {
"timestamp": "2026-08-03T17:29:13+00:00",
"model_a": "/home/jcreek/ai-training/CosmicClash/Game/bots/20260802-1458-curric-s4-unmask-retry1.json",
"model_b": "/home/jcreek/ai-training/CosmicClash/Game/bots/curric-s5-aggression.json",
"episodes": 100,
"wins_a": 30,
"wins_b": 56,
"draws": 14,
"win_rate_a": 0.3
},
"decision": "fail"
}
]
}