Change model to Qwen/Qwen2.5-0.5B-Instruct for CI.

Signed-off-by: Jonas Yang <joyang@nvidia.com>

Author: joyang-nv

.github/workflows/e2e_ppo_grpo_trainer_trtllm.yml

@@ -137,7 +137,7 @@ jobs:
             data.train_files="['${HOME}/data/gsm8k/train.parquet']" \
             data.val_files="['${HOME}/data/gsm8k/test.parquet']" \
             trainer.logger='["console"]' \
-            actor_rollout_ref.model.path=Qwen/Qwen2-0.5B-Instruct
+            actor_rollout_ref.model.path=Qwen/Qwen2.5-0.5B-Instruct
       - name: clean up
         run: |
           rm -rf checkpoints
@@ -174,7 +174,7 @@ jobs:
             data.train_files="['${HOME}/data/gsm8k/train.parquet']" \
             data.val_files="['${HOME}/data/gsm8k/test.parquet']" \
             trainer.logger='["console"]' \
-            actor_rollout_ref.model.path=Qwen/Qwen2-0.5B-Instruct
+            actor_rollout_ref.model.path=Qwen/Qwen2.5-0.5B-Instruct
       - name: clean up
         run: |
           rm -rf checkpoints