update 0718

tests/data/test_mm_plugin.py

@@ -238,7 +238,6 @@ def test_llama4_plugin():
     _check_plugin(**check_inputs)
 
 
-@pytest.mark.skipif(not is_transformers_version_greater_than("4.47.0"), reason="Requires transformers>=4.47.0")
 def test_llava_plugin():
     image_seqlen = 576
     tokenizer_module = _load_tokenizer_module(model_name_or_path="llava-hf/llava-1.5-7b-hf")

tests/data/test_template.py

@@ -226,6 +226,19 @@ def test_gemma_template(use_fast: bool):
     _check_template("google/gemma-3-4b-it", "gemma", prompt_str, answer_str, use_fast)
 
 
+@pytest.mark.skipif(not HF_TOKEN, reason="Gated model.")
+@pytest.mark.parametrize("use_fast", [True, False])
+def test_gemma2_template(use_fast: bool):
+    prompt_str = (
+        f"<bos><start_of_turn>user\n{MESSAGES[0]['content']}<end_of_turn>\n"
+        f"<start_of_turn>model\n{MESSAGES[1]['content']}<end_of_turn>\n"
+        f"<start_of_turn>user\n{MESSAGES[2]['content']}<end_of_turn>\n"
+        "<start_of_turn>model\n"
+    )
+    answer_str = f"{MESSAGES[3]['content']}<end_of_turn>\n"
+    _check_template("google/gemma-2-2b-it", "gemma2", prompt_str, answer_str, use_fast)
+
+
 @pytest.mark.skipif(not HF_TOKEN, reason="Gated model.")
 @pytest.mark.parametrize("use_fast", [True, False])
 def test_llama3_template(use_fast: bool):

tests/version.txt

 # change if test fails or cache is outdated
-0.9.3.108
+0.9.4.100