From d994a4c8f564af516da58cf54d53c7e9314eaa2a Mon Sep 17 00:00:00 2001 From: Adil Hafeez Date: Thu, 10 Jul 2025 16:23:40 -0700 Subject: [PATCH] fix more --- crates/llm_gateway/src/stream_context.rs | 2 +- crates/llm_gateway/tests/integration.rs | 8 ++++---- .../preference_based_routing/arch_config_rendered.yaml | 6 +++--- .../preference_based_routing/hurl_tests/simple.hurl | 2 +- 4 files changed, 9 insertions(+), 9 deletions(-) diff --git a/crates/llm_gateway/src/stream_context.rs b/crates/llm_gateway/src/stream_context.rs index b25c0c75..82b88509 100644 --- a/crates/llm_gateway/src/stream_context.rs +++ b/crates/llm_gateway/src/stream_context.rs @@ -344,7 +344,7 @@ impl HttpContext for StreamContext { info!( "on_http_request_body: provider: {}, model requested (in body): {}, model selected: {}", - self.llm_provider().provider_interface, + self.llm_provider().name, model_requested, model_name.unwrap_or(&"None".to_string()), ); diff --git a/crates/llm_gateway/tests/integration.rs b/crates/llm_gateway/tests/integration.rs index f97d87d3..108ab1ce 100644 --- a/crates/llm_gateway/tests/integration.rs +++ b/crates/llm_gateway/tests/integration.rs @@ -266,7 +266,7 @@ fn llm_gateway_bad_request_to_open_ai_chat_completions() { .expect_get_buffer_bytes(Some(BufferType::HttpRequestBody)) .returning(Some(incomplete_chat_completions_request_body)) .expect_log(Some(LogLevel::Debug), None) - .expect_log(Some(LogLevel::Info), Some("on_http_request_body: provider: openai, model requested (in body): gpt-1, model selected: gpt-4")) + .expect_log(Some(LogLevel::Info), Some("on_http_request_body: provider: open-ai-gpt-4, model requested (in body): gpt-1, model selected: gpt-4")) .expect_send_local_response( Some(StatusCode::BAD_REQUEST.as_u16().into()), None, @@ -432,7 +432,7 @@ fn llm_gateway_override_model_name() { .returning(Some(chat_completions_request_body)) // The actual call is not important in this test, we just need to grab the token_id .expect_log(Some(LogLevel::Debug), None) - .expect_log(Some(LogLevel::Info), Some("on_http_request_body: provider: openai, model requested (in body): gpt-1, model selected: gpt-4")) + .expect_log(Some(LogLevel::Info), Some("on_http_request_body: provider: open-ai-gpt-4, model requested (in body): gpt-1, model selected: gpt-4")) .expect_log(Some(LogLevel::Debug), None) .expect_log(Some(LogLevel::Debug), None) .expect_metric_record("input_sequence_length", 29) @@ -481,7 +481,7 @@ fn llm_gateway_override_use_default_model() { // The actual call is not important in this test, we just need to grab the token_id .expect_log( Some(LogLevel::Info), - Some("on_http_request_body: provider: openai, model requested (in body): gpt-1, model selected: gpt-4"), + Some("on_http_request_body: provider: open-ai-gpt-4, model requested (in body): gpt-1, model selected: gpt-4"), ) .expect_log(Some(LogLevel::Debug), None) .expect_log(Some(LogLevel::Debug), None) @@ -529,7 +529,7 @@ fn llm_gateway_override_use_model_name_none() { .returning(Some(chat_completions_request_body)) // The actual call is not important in this test, we just need to grab the token_id .expect_log(Some(LogLevel::Debug), None) - .expect_log(Some(LogLevel::Info), Some("on_http_request_body: provider: openai, model requested (in body): none, model selected: gpt-4")) + .expect_log(Some(LogLevel::Info), Some("on_http_request_body: provider: open-ai-gpt-4, model requested (in body): none, model selected: gpt-4")) .expect_log(Some(LogLevel::Debug), None) .expect_log(Some(LogLevel::Debug), None) .expect_metric_record("input_sequence_length", 29) diff --git a/demos/use_cases/preference_based_routing/arch_config_rendered.yaml b/demos/use_cases/preference_based_routing/arch_config_rendered.yaml index bdd85f0d..7e914893 100644 --- a/demos/use_cases/preference_based_routing/arch_config_rendered.yaml +++ b/demos/use_cases/preference_based_routing/arch_config_rendered.yaml @@ -15,14 +15,14 @@ llm_providers: name: openai/gpt-4o provider_interface: openai routing_preferences: - - description: b - name: code understanding + - description: generating new code snippets, functions, or boilerplate based on user prompts or requirements + name: code generation - access_key: $OPENAI_API_KEY model: gpt-4.1 name: openai/gpt-4.1 provider_interface: openai routing_preferences: - - description: a + - description: understand and explain existing code snippets, functions, or libraries name: code understanding tracing: random_sampling: 100 diff --git a/demos/use_cases/preference_based_routing/hurl_tests/simple.hurl b/demos/use_cases/preference_based_routing/hurl_tests/simple.hurl index c4ee5d8a..d9b243e7 100644 --- a/demos/use_cases/preference_based_routing/hurl_tests/simple.hurl +++ b/demos/use_cases/preference_based_routing/hurl_tests/simple.hurl @@ -13,7 +13,7 @@ Content-Type: application/json HTTP 200 [Asserts] header "content-type" == "application/json" -jsonpath "$.model" matches /^gpt-4o-mini/ +jsonpath "$.model" matches /^gpt-4.1/ jsonpath "$.usage" != null jsonpath "$.choices[0].message.content" != null jsonpath "$.choices[0].message.role" == "assistant"