Dataset Viewer
Auto-converted to Parquet Duplicate
game_id
stringlengths
8
17
map
stringclasses
2 values
num_turns
int64
0
21k
winner
stringclasses
113 values
created_at
int64
0
1,785B
git_commit
stringclasses
3 values
record_version
stringclasses
1 value
run_name
stringclasses
2 values
policy_update
stringclasses
2 values
stage
int64
-1
28
engine
stringclasses
3 values
policy_repo
stringclasses
2 values
policy_revision
stringclasses
1 value
record_json
stringlengths
3.18k
169k
01IHaVGl
Onion
6,021
["player", "AGENTRL1"]
1,784,496,840,125
58c536b7a3c528c125890b40b1b213109b8b7014
v0.0.2
ppo_v10
28
native
djmango/openfront-rl
{"info":{"gameID":"01IHaVGl","lobbyCreatedAt":1784496758866,"config":{"gameMap":"Onion","gameMapSize":"Normal","gameMode":"Free For All","gameType":"Singleplayer","difficulty":"Easy","nations":4,"donateGold":true,"donateTroops":true,"bots":26,"infiniteGold":false,"infiniteTroops":false,"instantBuild":false,"randomSpawn...
01UzktmH
Onion
4,361
["player", "AGENTRL1"]
1,784,492,202,933
58c536b7a3c528c125890b40b1b213109b8b7014
v0.0.2
ppo_v10
28
native
djmango/openfront-rl
{"info":{"gameID":"01UzktmH","lobbyCreatedAt":1784492143515,"config":{"gameMap":"Onion","gameMapSize":"Normal","gameMode":"Free For All","gameType":"Singleplayer","difficulty":"Easy","nations":1,"donateGold":true,"donateTroops":true,"bots":16,"infiniteGold":false,"infiniteTroops":false,"instantBuild":false,"randomSpawn...
01WVirA1
Onion
7,671
["player", "AGENTRL1"]
1,784,495,102,401
58c536b7a3c528c125890b40b1b213109b8b7014
v0.0.2
ppo_v10
28
native
djmango/openfront-rl
{"info":{"gameID":"01WVirA1","lobbyCreatedAt":1784495007605,"config":{"gameMap":"Onion","gameMapSize":"Normal","gameMode":"Free For All","gameType":"Singleplayer","difficulty":"Easy","nations":4,"donateGold":true,"donateTroops":true,"bots":26,"infiniteGold":false,"infiniteTroops":false,"instantBuild":false,"randomSpawn...
01w3eliB
Onion
7,121
["player", "AGENTRL1"]
1,784,492,619,628
58c536b7a3c528c125890b40b1b213109b8b7014
v0.0.2
ppo_v10
28
native
djmango/openfront-rl
"{\"info\":{\"gameID\":\"01w3eliB\",\"lobbyCreatedAt\":1784492523647,\"config\":{\"gameMap\":\"Onion(...TRUNCATED)
03C50Rcp
Onion
7,861
["player", "AGENTRL1"]
1,784,496,570,730
58c536b7a3c528c125890b40b1b213109b8b7014
v0.0.2
ppo_v10
28
native
djmango/openfront-rl
"{\"info\":{\"gameID\":\"03C50Rcp\",\"lobbyCreatedAt\":1784496467415,\"config\":{\"gameMap\":\"Onion(...TRUNCATED)
03SJ8lU5
Onion
7,751
["nation", "Inner Tribe"]
1,784,494,605,330
58c536b7a3c528c125890b40b1b213109b8b7014
v0.0.2
ppo_v10
28
native
djmango/openfront-rl
"{\"info\":{\"gameID\":\"03SJ8lU5\",\"lobbyCreatedAt\":1784494500712,\"config\":{\"gameMap\":\"Onion(...TRUNCATED)
03gRutSz
Onion
5,861
["player", "AGENTRL1"]
1,784,493,723,872
58c536b7a3c528c125890b40b1b213109b8b7014
v0.0.2
ppo_v10
28
native
djmango/openfront-rl
"{\"info\":{\"gameID\":\"03gRutSz\",\"lobbyCreatedAt\":1784493648664,\"config\":{\"gameMap\":\"Onion(...TRUNCATED)
04yakIuq
Onion
4,951
["player", "AGENTRL1"]
1,784,492,518,243
DEV
v0.0.2
-1
"{\"info\":{\"gameID\":\"04yakIuq\",\"lobbyCreatedAt\":1784492448282,\"config\":{\"gameMap\":\"Onion(...TRUNCATED)
05Cngl23
Onion
7,781
["player", "AGENTRL1"]
1,784,498,391,421
58c536b7a3c528c125890b40b1b213109b8b7014
v0.0.2
ppo_v10
28
native
djmango/openfront-rl
"{\"info\":{\"gameID\":\"05Cngl23\",\"lobbyCreatedAt\":1784498296832,\"config\":{\"gameMap\":\"Onion(...TRUNCATED)
05GlSLaX
Onion
10,681
["player", "AGENTRL1"]
1,784,498,051,843
58c536b7a3c528c125890b40b1b213109b8b7014
v0.0.2
ppo_v10
28
native
djmango/openfront-rl
"{\"info\":{\"gameID\":\"05GlSLaX\",\"lobbyCreatedAt\":1784497913283,\"config\":{\"gameMap\":\"Onion(...TRUNCATED)
End of preview. Expand in Data Studio

OpenFront.io RL game replays

Sparse-turn GameRecord v0.0.2 JSON for every finished RL episode (training, watch, showcase). Same format as the OpenFront client archive / Node bridge (version, gitCommit, sparse turns with intents).

Each parquet row is one game. The record_json column is the full Node- compatible GameRecord string (~100–200 KB for long games). When a watch / showcase episode recorded policy decisions, thinking_json holds a compact top-3 action trace (typically a few KB) derived from the MODEL overlay .debug.json sidecar. Metadata columns capture policy / engine / stage.

Generated by djmango/openfront-ai. Replay with datagen/replay.ts or the client archive API / ofshowcase archive.

Downloads last month
13