The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
tinywmcraft
Data of tinywmcraft: train a world model of Craftax-Classic in under 10 minutes on one A100, then play it in your browser.
| file | contents |
|---|---|
train.npz |
500,000 transitions: 250 streams of 2000 steps, 63x63 RGB frames |
eval.npz |
2048 clips of 8 real + 64 future frames, scored by eval.py |
train.npz
| array | shape | |
|---|---|---|
frames |
(2001, 250, 63, 63, 3) uint8 | |
actions, rewards |
(2000, 250) | actions[t, e] and rewards[t, e] describe frames[t, e] -> frames[t+1, e] |
dones |
(2000, 250) bool | deaths only |
truncated |
(2000, 250) bool | episodes cut short by the collector, not deaths |
episode_ids, inverse_temperatures, cohort, source_stream_ids, meta |
collection details |
An episode ends at dones OR truncated; the transition at the end stores the next episode's first frame. Train a done
head on dones only.
Collection: 389,824 transitions from a frozen stochastic discoRL agent at a range of sampling temperatures, and
110,176 from a PPO agent in 210 episodes that eat a plant, mine a diamond or make an iron pickaxe, each cut 20-50 steps
after that event unless the agent died first. Every one of the 22 achievements appears. Details in meta.
eval.npz
| array | shape | |
|---|---|---|
frames |
(2048, 72, 63, 63, 3) uint8 | 8 context frames, then 64 future frames |
actions, rewards, dones |
(2048, 72) | of the transition into each frame; dones marks deaths |
valid_len |
(2048,) | death clips end at the death |
event, event_frame, event_names |
the rare event in a clip and the frame it leads into |
1536 clips are random held-out windows; 512 hold a rare event between future frames 4 and 16: 100 diamonds, 100 eaten plants, 100 iron pickaxes, 70 iron swords, 71 deaths in lava, 71 deaths by zombies or skeletons.
- Downloads last month
- -