don't automatically enable lora kernels for RL training (axolotl-ai-cloud#2600)

Author: winglian

src/axolotl/utils/schemas/config.py

@@ -1319,6 +1319,9 @@ def check_multigpu_lora_kernels(cls, data):
     @classmethod
     def check_auto_enable_lora_kernels(cls, data):
         # Only proceed if using LoRA or QLoRA adapter
+        if data.get("rl"):
+            # RL trainers not tested so don't enable kernels by default
+            return data
         if data.get("adapter") in ["lora", "qlora"]:
             # Skip if already set, using unsloth optimizations, or using 8-bit
             unsloth_fields = ["unsloth_lora_mlp", "unsloth_lora_qkv", "unsloth_lora_o"]