| --- |
| license: apache-2.0 |
| pipeline_tag: reinforcement-learning |
| tags: |
| - RL |
| - Reinforcement learning |
| - Minecraft |
| - Evolution |
| language: |
| - en |
| --- |
| |
| The current stable and up to date simbot weights, the minecraft AI reinforcement learning player neural network. (Code soon to be publicly released.)<br> |
| Public demo is on my friends minecraft server *mc.nico.re* (Plus runs on most up to date weights)<br> |
|
|
| ## Model specs: |
| Layers:<br> |
| 1. 731 (Input; blocks, entities, position, last eaten, health, and more) |
| 2. 2048 (Hidden layer) |
| 3. 1024 (Hidden layer) |
| 4. 512 (Hidden layer) |
| 5. 66 (Output; what action to do, eg. eat, fight, look, move...) |
|
|
| Total params: 4.155m.<br> |
|
|
| Example capabilities:<br> |
| 1. Eats food when low on hunger.<br> |
| 2. Eats Golden apples when low on health.<br> |
| 3. Sometimes flees, sometimes fights, based off free space and ammount of enemies.<br> |
| 4. Can beat enemies<br> |
|
|
| Known limitations:<br> |
| 1. Randomly bumps into walls.<br> |
| 2. Forgets things.<br> |
| 3. Sometimes fights using random items instead of swords.<br> |
|
|
| ## Reward system: |
| Rewarded: Eating, Damaging a enemy, sneaking near players (Just for fun).<br> |
| Punished: Getting damaged, Dieing, Stuck in a loop, Stuck in one area, Clicking at nothing.<br> |
|
|
|  |
|
|
