DEV: Extract configs to a yml file and allow local config (#1142)

.gitignore

@@ -4,3 +4,4 @@ node_modules
 .env
 evals/log
 evals/cases
+config/eval-llms.local.yml

README.md

@@ -6,7 +6,8 @@ For more information, please see: https://meta.discourse.org/t/discourse-ai/2592
 
 ### Evals
 
-The directory `evals` contains AI evals for the Discourse AI plugin.
+The directory `evals` contains AI evals for the Discourse AI plugin. 
+You may create a local config by copying `config/eval-llms.yml` to `config/eval-llms.local.yml` and modifying the values.
 
 To run them use: 
 

config/eval-llms.yml

@@ -0,0 +1,60 @@
+llms:
+  gpt-4o:
+    display_name: GPT-4o
+    name: gpt-4o
+    tokenizer: DiscourseAi::Tokenizer::OpenAiTokenizer
+    api_key_env: OPENAI_API_KEY
+    provider: open_ai
+    url: https://api.openai.com/v1/chat/completions
+    max_prompt_tokens: 131072
+    vision_enabled: true
+
+  gpt-4o-mini:
+    display_name: GPT-4o-mini
+    name: gpt-4o-mini
+    tokenizer: DiscourseAi::Tokenizer::OpenAiTokenizer
+    api_key_env: OPENAI_API_KEY
+    provider: open_ai
+    url: https://api.openai.com/v1/chat/completions
+    max_prompt_tokens: 131072
+    vision_enabled: true
+
+  claude-3.5-haiku:
+    display_name: Claude 3.5 Haiku
+    name: claude-3-5-haiku-latest
+    tokenizer: DiscourseAi::Tokenizer::AnthropicTokenizer
+    api_key_env: ANTHROPIC_API_KEY
+    provider: anthropic
+    url: https://api.anthropic.com/v1/messages
+    max_prompt_tokens: 200000
+    vision_enabled: false
+
+  claude-3.5-sonnet:
+    display_name: Claude 3.5 Sonnet
+    name: claude-3-5-sonnet-latest
+    tokenizer: DiscourseAi::Tokenizer::AnthropicTokenizer
+    api_key_env: ANTHROPIC_API_KEY
+    provider: anthropic
+    url: https://api.anthropic.com/v1/messages
+    max_prompt_tokens: 200000
+    vision_enabled: true
+
+  gemini-2.0-flash:
+    display_name: Gemini 2.0 Flash
+    name: gemini-2-0-flash
+    tokenizer: DiscourseAi::Tokenizer::GeminiTokenizer
+    api_key_env: GEMINI_API_KEY
+    provider: google
+    url: https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash
+    max_prompt_tokens: 1000000
+    vision_enabled: true
+
+  gemini-2.0-pro-exp:
+    display_name: Gemini 2.0 pro
+    name: gemini-2-0-pro-exp
+    tokenizer: DiscourseAi::Tokenizer::GeminiTokenizer
+    api_key_env: GEMINI_API_KEY
+    provider: google
+    url: https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-pro-exp
+    max_prompt_tokens: 1000000
+    vision_enabled: true

evals/lib/llm.rb

@@ -1,71 +1,23 @@
 # frozen_string_literal: true
 
 class DiscourseAi::Evals::Llm
-  CONFIGS = {
-    "gpt-4o" => {
-      display_name: "GPT-4o",
-      name: "gpt-4o",
-      tokenizer: "DiscourseAi::Tokenizer::OpenAiTokenizer",
-      api_key_env: "OPENAI_API_KEY",
-      provider: "open_ai",
-      url: "https://api.openai.com/v1/chat/completions",
-      max_prompt_tokens: 131_072,
-      vision_enabled: true,
-    },
-    "gpt-4o-mini" => {
-      display_name: "GPT-4o-mini",
-      name: "gpt-4o-mini",
-      tokenizer: "DiscourseAi::Tokenizer::OpenAiTokenizer",
-      api_key_env: "OPENAI_API_KEY",
-      provider: "open_ai",
-      url: "https://api.openai.com/v1/chat/completions",
-      max_prompt_tokens: 131_072,
-      vision_enabled: true,
-    },
-    "claude-3.5-haiku" => {
-      display_name: "Claude 3.5 Haiku",
-      name: "claude-3-5-haiku-latest",
-      tokenizer: "DiscourseAi::Tokenizer::AnthropicTokenizer",
-      api_key_env: "ANTHROPIC_API_KEY",
-      provider: "anthropic",
-      url: "https://api.anthropic.com/v1/messages",
-      max_prompt_tokens: 200_000,
-      vision_enabled: false,
-    },
-    "claude-3.5-sonnet" => {
-      display_name: "Claude 3.5 Sonnet",
-      name: "claude-3-5-sonnet-latest",
-      tokenizer: "DiscourseAi::Tokenizer::AnthropicTokenizer",
-      api_key_env: "ANTHROPIC_API_KEY",
-      provider: "anthropic",
-      url: "https://api.anthropic.com/v1/messages",
-      max_prompt_tokens: 200_000,
-      vision_enabled: true,
-    },
-    "gemini-2.0-flash" => {
-      display_name: "Gemini 2.0 Flash",
-      name: "gemini-2-0-flash",
-      tokenizer: "DiscourseAi::Tokenizer::GeminiTokenizer",
-      api_key_env: "GEMINI_API_KEY",
-      provider: "google",
-      url: "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-flash",
-      max_prompt_tokens: 1_000_000,
-      vision_enabled: true,
-    },
-    "gemini-2.0-pro-exp" => {
-      display_name: "Gemini 2.0 pro",
-      name: "gemini-2-0-pro-exp",
-      tokenizer: "DiscourseAi::Tokenizer::GeminiTokenizer",
-      api_key_env: "GEMINI_API_KEY",
-      provider: "google",
-      url: "https://generativelanguage.googleapis.com/v1beta/models/gemini-2.0-pro-exp",
-      max_prompt_tokens: 1_000_000,
-      vision_enabled: true,
-    },
-  }
+  def self.configs
+    return @configs if @configs
+
+    yaml_path = File.join(File.dirname(__FILE__), "../../config/eval-llms.yml")
+    local_yaml_path = File.join(File.dirname(__FILE__), "../../config/eval-llms.local.yml")
+
+    configs = YAML.load_file(yaml_path)["llms"] || {}
+    if File.exist?(local_yaml_path)
+      local_configs = YAML.load_file(local_yaml_path)["llms"] || {}
+      configs = configs.merge(local_configs)
+    end
+
+    @configs = configs
+  end
 
   def self.print
-    CONFIGS
+    configs
       .keys
       .map do |config_name|
         begin
@@ -79,38 +31,39 @@ class DiscourseAi::Evals::Llm
   end
 
   def self.choose(config_name)
-    if CONFIGS[config_name].nil?
-      CONFIGS
+    return [] unless configs
+    if !config_name || !configs[config_name]
+      configs
         .keys
-        .map do |config_name|
+        .map do |name|
           begin
-            new(config_name)
-          rescue => e
-            puts "Error initializing #{config_name}: #{e}"
+            new(name)
+          rescue StandardError
             nil
           end
         end
         .compact
-    elsif !CONFIGS.include?(config_name)
-      raise "Invalid llm"
     else
       [new(config_name)]
     end
   end
 
-  attr_reader :llm_model
-  attr_reader :llm_proxy
-  attr_reader :config_name
+  attr_reader :llm_model, :llm_proxy, :config_name
 
   def initialize(config_name)
-    config = CONFIGS[config_name].dup
-    api_key_env = config.delete(:api_key_env)
-    if !ENV[api_key_env]
-      raise "Missing API key for #{config_name}, should be set via #{api_key_env}"
+    config = self.class.configs[config_name].dup
+    if config["api_key_env"]
+      api_key_env = config.delete("api_key_env")
+      unless ENV[api_key_env]
+        raise "Missing API key for #{config_name}, should be set via #{api_key_env}"
+      end
+      config[:api_key] = ENV[api_key_env]
+    elsif config["api_key"]
+      config[:api_key] = config.delete("api_key")
+    else
+      raise "No API key or API key env var configured for #{config_name}"
     end
-
-    config[:api_key] = ENV[api_key_env]
-    @llm_model = LlmModel.new(config)
+    @llm_model = LlmModel.new(config.symbolize_keys)
     @llm_proxy = DiscourseAi::Completions::Llm.proxy(@llm_model)
     @config_name = config_name
   end