chore: use non-cuda version for e2e tests (#2112)

Co-authored-by: sangjanai <sang@jan.ai>

Author: vansangpfiev

engine/e2e-test/api/engines/test_api_engine.py

@@ -28,7 +28,7 @@ def test_engines_get_llamacpp_should_be_successful(self):
 
     # engines install
     def test_engines_install_llamacpp_specific_version_and_variant(self):
-        data = {"version": "v0.1.40-b4354", "variant": "linux-amd64-avx-cuda-11-7"}
+        data = {"version": "v0.1.40-b4354", "variant": "linux-amd64-avx"}
         response = requests.post(
             "http://localhost:3928/v1/engines/llama-cpp/install", json=data
         )

engine/e2e-test/api/engines/test_api_engine_install_nightly.py

@@ -23,7 +23,7 @@ def test_engines_install_llamacpp_should_be_successful(self):
         assert response.status_code == 200
 
     def test_engines_install_llamacpp_specific_version_and_variant(self):
-        data = {"version": latest_pre_release_tag, "variant": "linux-amd64-avx-cuda-11-7"}
+        data = {"version": latest_pre_release_tag, "variant": "linux-amd64-avx"}
         response = requests.post(
             "http://localhost:3928/v1/engines/llama-cpp/install", json=data
         )

engine/e2e-test/api/engines/test_api_get_default_engine.py

@@ -24,7 +24,7 @@ def setup_and_teardown(self):
     def test_api_get_default_engine_successfully(self):
         # Data test
         engine= "llama-cpp"
-        name= "linux-amd64-avx-cuda-11-7"
+        name= "linux-amd64-avx"
         version= "v0.1.35-27.10.24"
 
         data = {"version": version, "variant": name}