[TNX][CI] setup java for tnx models (#1476)

.github/workflows/llm_inf2_integration.yml

@@ -318,6 +318,11 @@ jobs:
           sudo rm -rf /home/ubuntu/actions-runner/_work/_tool/Java_Corretto_jdk/
           echo "wait dpkg lock..."
           while sudo fuser /var/{lib/{dpkg,apt/lists},cache/apt/archives}/lock >/dev/null 2>&1; do sleep 5; done
+      - name: Set up JDK 17
+        uses: actions/setup-java@v4
+        with:
+          distribution: 'corretto'
+          java-version: 17
       - name: Set up Python3
         uses: actions/setup-python@v4
         with:

tests/integration/llm/client.py

@@ -796,7 +796,7 @@ def test_handler_rolling_batch(model, model_spec):
     # dryrun phase
     req = {"inputs": batch_generation(1)[0]}
     seq_length = 100
-    params = {"do_sample": True, "max_new_tokens": seq_length}
+    params = {"do_sample": True, "max_new_tokens": seq_length, "details": True}
     req["parameters"] = params
     logging.info(f"req {req}")
     res = send_json(req)

tests/integration/llm/prepare.py

@@ -505,15 +505,17 @@
         "option.max_rolling_batch_size": 4,
         "option.rolling_batch": 'auto',
         "option.model_loading_timeout": 2400,
-        "option.load_split_model": True
+        "option.load_split_model": True,
+        "option.output_formatter": "jsonlines"
     },
     "mistral-7b-rb": {
         "option.model_id": "s3://djl-llm/mistral-7b/",
         "option.tensor_parallel_degree": 4,
         "option.n_positions": 512,
         "option.max_rolling_batch_size": 4,
         "option.rolling_batch": 'auto',
-        "option.model_loading_timeout": 2400
+        "option.model_loading_timeout": 2400,
+        "option.output_formatter": "jsonlines"
     }
 }