jcreek/CosmicClash
1
0
Fork 0
mirror of https://github.com/jcreek/CosmicClash.git synced 2026-09-10 16:04:04 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat(*): Add self-play RL training pipeline with PPO trainer, in-game GDScript policy inference, and bot opponent support in Match mode

Browse Source
This commit is contained in:
Josh Creek
2026-07-18 19:32:51 +01:00
parent 328831df1f
commit 85f96eb15e
81 changed files with 3934 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Game/addons/godot_rl_agents/onnx/csharp/docs/ONNXInference.xml
+31
View File
@@ -0,0 +1,31 @@
<docs>
<members name="ONNXInference">
<ONNXInference>
<summary>
The main <c>ONNXInference</c> Class that handles the inference process.
</summary>
</ONNXInference>
<Initialize>
<summary>
Starts the inference process.
</summary>
<param name="Path">Path to the ONNX model, expects a path inside resources.</param>
<param name="BatchSize">How many observations will the model recieve.</param>
</Initialize>
<Run>
<summary>
Runs the given input through the model and returns the output.
</summary>
<param name="obs">Dictionary containing all observations.</param>
<param name="state_ins">How many different agents are creating these observations.</param>
<returns>A Dictionary of arrays, containing instructions based on the observations.</returns>
</Run>
<Load>
<summary>
Loads the given model into the inference process, using the best Execution provider available.
</summary>
<param name="Path">Path to the ONNX model, expects a path inside resources.</param>
<returns>InferenceSession ready to run.</returns>
</Load>
</members>
</docs>