remove old tests. some chores

Author: gau-nernst

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

@@ -52,16 +52,6 @@ async def test_engines_install_uninstall_llamacpp_should_be_successful(self):
         response = requests.delete("http://localhost:3928/v1/engines/llama-cpp/install")
         assert response.status_code == 200
 
-    @pytest.mark.asyncio
-    async def test_engines_install_uninstall_python_should_be_successful(self):
-        response = requests.post("http://localhost:3928/v1/engines/python-engine/install")
-        assert response.status_code == 200
-        await wait_for_websocket_download_success_event(timeout=None)
-        time.sleep(30)
-
-        response = requests.delete("http://localhost:3928/v1/engines/python-engine/install")
-        assert response.status_code == 200
-
     @pytest.mark.asyncio
     async def test_engines_install_uninstall_llamacpp_with_only_version_should_be_failed(self):
         # install first

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

@@ -22,10 +22,6 @@ def test_engines_install_llamacpp_should_be_successful(self):
         response = requests.post("http://localhost:3928/v1/engines/llama-cpp/install")
         assert response.status_code == 200
 
-    def test_engines_install_python_should_be_successful(self):
-        response = requests.post("http://localhost:3928/v1/engines/python-engine/install")
-        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"}
         response = requests.post(

engine/e2e-test/cli/engines/test_cli_engine_install.py

@@ -31,17 +31,6 @@ def test_engines_install_llamacpp_should_be_successfully(self):
         assert len(response.json()) > 0
         assert exit_code == 0, f"Install engine failed with error: {error}"
 
-    def test_engines_install_python_should_be_successfully(self):
-        exit_code, output, error = run(
-            "Install Engine",
-            ["engines", "install", "python-engine"],
-            timeout=None,
-            capture=False,
-        )
-        response = requests.get("http://127.0.0.1:3928/v1/engines/python-engine")
-        assert len(response.json()) > 0
-        assert exit_code == 0, f"Install engine failed with error: {error}"
-
     @pytest.mark.skipif(reason="Ignore onnx-runtime test")
     def test_engines_install_onnx_on_macos_should_be_failed(self):
         exit_code, output, error = run(

engine/e2e-test/cli/engines/test_cli_engine_install_nightly.py

@@ -31,17 +31,6 @@ def test_engines_install_llamacpp_should_be_successfully(self):
         assert len(response.json()) > 0
         assert exit_code == 0, f"Install engine failed with error: {error}"
 
-    def test_engines_install_python_should_be_successfully(self):
-        exit_code, output, error = run(
-            "Install Engine",
-            ["engines", "install", "python-engine"],
-            timeout=None,
-            capture=False,
-        )
-        response = requests.get("http://127.0.0.1:3928/v1/engines/python-engine")
-        assert len(response.json()) > 0
-        assert exit_code == 0, f"Install engine failed with error: {error}"
-
     @pytest.mark.skipif(reason="Ignore onnx-runtime test")
     def test_engines_install_onnx_on_macos_should_be_failed(self):
         exit_code, output, error = run(

engine/extensions/python-engines/python_utils.cc

@@ -39,7 +39,7 @@ cpp::result<void, std::string> UvInstall() {
   std::filesystem::create_directories(py_bin_path);
 
   // NOTE: do we need a mechanism to update uv, or just pin uv version with cortex release?
-  const std::string uv_version = "0.6.3";
+  const std::string uv_version = "0.6.11";
 
   // build download url based on system info
   std::stringstream fname_stream;

engine/extensions/python-engines/vllm_engine.cc

@@ -81,6 +81,7 @@ std::vector<EngineVariantResponse> VllmEngine::GetVariants() {
   std::vector<EngineVariantResponse> variants;
   for (const auto& entry : fs::directory_iterator(vllm_path)) {
     const auto name = "linux-amd64-cuda";  // arbitrary
+    // TODO: after llama-server is merged, check if we need to add "v"
     const auto version_str = "v" + entry.path().filename().string();
     const EngineVariantResponse variant{name, version_str, kVllmEngine};
     variants.push_back(variant);