Fix LoRA params in Python in LoRA without regret (#4215)

Author: sergiopaniego

docs/source/lora_without_regret.md

@@ -42,7 +42,7 @@ from trl import SFTTrainer, SFTConfig
 
 dataset = load_dataset("open-thoughts/OpenThoughts-114k", split="train")
 
-peft_config = LoraConfig(lora_r=256, lora_alpha=16, lora_target_modules="all-linear")
+peft_config = LoraConfig(r=256, lora_alpha=16, target_modules="all-linear")
 
 training_args = SFTConfig(
     learning_rate=2e-4,
@@ -245,9 +245,9 @@ def strip_reasoning_accuracy_reward(completions, **kwargs):
     ... 
 
 peft_config = LoraConfig(
-    lora_r=1,
+    r=1,
     lora_alpha=32,
-    lora_target_modules="all-linear"
+    target_modules="all-linear"
 )
 
 training_args = GRPOConfig(