Tune lr

huggingface · Feb 19, 2025 · fb6e4ae · fb6e4ae
1 parent 04381ca
commit fb6e4ae
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 1 deletion.
diff --git a/README.md b/README.md
@@ -199,6 +199,8 @@ Then make sure your dataset contains a `verification_info` column with the follo
 }
 ```
 
+See the [training config](./recipes/Qwen2.5-1.5B-Instruct/grpo/config_demo_code.yaml) for an example with `Qwen2.5-1.5B-Instruct`.
+
 ### Launching jobs on a Slurm cluster
 
 If you have access to a Slurm cluster, we provide a `slurm/train.slurm` script that will automatically queue training jobs for you. Here's how you can use it:

diff --git a/recipes/Qwen2.5-1.5B-Instruct/grpo/config_demo_code.yaml b/recipes/Qwen2.5-1.5B-Instruct/grpo/config_demo_code.yaml
@@ -23,7 +23,7 @@ gradient_checkpointing_kwargs:
   use_reentrant: false
 hub_model_id: Qwen2.5-1.5B-Open-R1-Code-GRPO
 hub_strategy: every_save
-learning_rate: 5.0e-07
+learning_rate: 5.0e-06
 log_completions: true
 log_level: info
 logging_first_step: true
-Original file line number
+Diff line change
@@ Expand Up @@
     }
     ```
+    See the [training config](./recipes/Qwen2.5-1.5B-Instruct/grpo/config_demo_code.yaml) for an example with `Qwen2.5-1.5B-Instruct`.
     ### Launching jobs on a Slurm cluster
     If you have access to a Slurm cluster, we provide a `slurm/train.slurm` script that will automatically queue training jobs for you. Here's how you can use it:
@@ Expand Down @@