fix: select linux arm64

Author: sangjanai

engine/e2e-test/test_api_engine.py

@@ -44,7 +44,8 @@ def test_engines_install_llamacpp_specific_version_and_null_variant(self):
     # engines uninstall
     @pytest.mark.asyncio
     async def test_engines_install_uninstall_llamacpp_should_be_successful(self):
-        response = requests.post("http://localhost:3928/v1/engines/llama-cpp/install")
+        data = {"version": "v0.1.43"}
+        response = requests.post("http://localhost:3928/v1/engines/llama-cpp/install", json=data)
         assert response.status_code == 200
         await wait_for_websocket_download_success_event(timeout=None)
         time.sleep(30)

engine/e2e-test/test_api_engine_install_nightly.py

@@ -19,7 +19,8 @@ def setup_and_teardown(self):
         stop_server()
 
     def test_engines_install_llamacpp_should_be_successful(self):
-        response = requests.post("http://localhost:3928/v1/engines/llama-cpp/install")
+        data = {"version": "v0.1.43"}
+        response = requests.post("http://localhost:3928/v1/engines/llama-cpp/install", json=data)
         assert response.status_code == 200
 
     def test_engines_install_llamacpp_specific_version_and_variant(self):

engine/e2e-test/test_api_model.py

@@ -85,9 +85,12 @@ async def test_model_pull_with_direct_url_should_have_desired_name(self):
             ],
         )
 
+    @pytest.mark.asyncio
     async def test_models_start_stop_should_be_successful(self):
         print("Install engine")
-        response = requests.post("http://localhost:3928/v1/engines/llama-cpp/install")
+        # TODO(sang) Remove version after marking 0.1.43 as stable
+        data = {"version": "v0.1.43"}
+        response = requests.post("http://localhost:3928/v1/engines/llama-cpp/install", json=data)
         assert response.status_code == 200
         await wait_for_websocket_download_success_event(timeout=None)
         # TODO(sang) need to fix for cuda download

engine/test/components/test_engine_matcher_utils.cc

@@ -19,6 +19,7 @@ class EngineMatcherUtilsTestSuite : public ::testing::Test {
       "cortex.llamacpp-0.1.25-25.08.24-linux-amd64-noavx-cuda-12-0.tar.gz",
       "cortex.llamacpp-0.1.25-25.08.24-linux-amd64-noavx.tar.gz",
       "cortex.llamacpp-0.1.25-25.08.24-linux-amd64-vulkan.tar.gz",
+      "cortex.llamacpp-0.1.43-linux-arm64.tar.gz",
       "cortex.llamacpp-0.1.25-25.08.24-mac-amd64.tar.gz",
       "cortex.llamacpp-0.1.25-25.08.24-mac-arm64.tar.gz",
       "cortex.llamacpp-0.1.25-25.08.24-windows-amd64-avx-cuda-11-7.tar.gz",
@@ -134,6 +135,18 @@ TEST_F(EngineMatcherUtilsTestSuite, TestValidate) {
     EXPECT_EQ(variant,
               "cortex.llamacpp-0.1.25-25.08.24-windows-amd64-avx2.tar.gz");
   }
+
+  {
+    auto os{"linux"};
+    auto cpu_arch{"arm64"};
+    auto suitable_avx{""};
+    auto cuda_version{""};
+
+    auto variant = engine_matcher_utils::Validate(
+        cortex_llamacpp_variants, os, cpu_arch, suitable_avx, cuda_version);
+
+    EXPECT_EQ(variant, "cortex.llamacpp-0.1.43-linux-arm64.tar.gz");
+  }
 }
 
 TEST_F(EngineMatcherUtilsTestSuite, TestGetVersionAndArch) {

engine/utils/engine_matcher_utils.h

@@ -156,6 +156,11 @@ inline std::string Validate(const std::vector<std::string>& variants,
   if (os == "mac" && !os_and_arch_compatible_list.empty())
     return os_and_arch_compatible_list[0];
 
+  if (os == "linux" && cpu_arch == "arm64" &&
+      !os_and_arch_compatible_list.empty()) {
+    return os_and_arch_compatible_list[0];
+  }
+
   std::vector<std::string> avx_compatible_list;
 
   std::copy_if(os_and_arch_compatible_list.begin(),