Commit 8d4fa280 authored by joseph_suarez's avatar joseph_suarez
Browse files

Update doc links to nmmo.github.io

parent e60ccf39
# Train an agent
- If you want to change the config for the agent, check out [`neural-mmo/projekt/config.py`](https://github.com/jsuarez5341/neural-mmo/blob/master/projekt/config.py).
- If you want to change the config for the agent, check out neural-mmo/projekt/config.py.
- We recommend that you subclass `CompetitionRound1` for the first round.
- The [neural-mmo documentation](https://jsuarez5341.github.io/neural-mmo/build/html/rst/userguide.html) contains a ton of additional information about the environment and previous work upon it.
- The [neural-mmo documentation](https://neuralmmo.github.io) contains a ton of additional information about the environment and previous work upon it.
## Example Scripted Agents
We provide a [simple wrapper](https://jsuarez5341.github.io/neural-mmo/build/html/rst/forge.trinity.scripted.io.html) to allow you to extract information from the observations.
We provide a [simple wrapper](https://github.com/NeuralMMO/environment/blob/v1.5.2/neural_mmo/forge/trinity/scripted/io.py) to allow you to extract information from the observations.
You can find various scripted baselines [here](https://github.com/jsuarez5341/neural-mmo/tree/master/forge/trinity/scripted).
You can find various scripted baselines [here](https://github.com/NeuralMMO/environment/blob/v1.5.2/neural_mmo/forge/trinity/scripted).
Wrap them as shown in /agents for compatibility with our submission system.
## Example Learned Agent
We provide a simple recurrent [baseline model](https://github.com/jsuarez5341/neural-mmo/blob/master/neural_mmo/forge/ethyr/torch/policy/baseline.py) and all associated [RLlib training code](https://github.com/jsuarez5341/neural-mmo/blob/master/projekt/rllib_wrapper.py).
We provide a simple recurrent [baseline model](https://github.com/NeuralMMO/environment/blob/v1.5.2/neural_mmo/forge/ethyr/torch/policy/baseline.py) and all associated [RLlib training code](https://github.com/NeuralMMO/environment/blob/v1.5.2/projekt/rllib_wrapper.py).
You can retrain the baseline agent as shown below. It takes a couple days to converge with a good gpu and 32 cores, but you can try it out for a few epochs.
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment