FEATURE: Add support for Mistral models (#919)

Adds support for mistral models (pixtral and mistral large now have presets)

Also corrects token accounting in AWS bedrock models

app/models/ai_api_audit_log.rb

@@ -13,6 +13,7 @@ class AiApiAuditLog < ActiveRecord::Base
     Cohere = 6
     Ollama = 7
     SambaNova = 8
+    Mistral = 9
   end
 
   def next_log_id

app/models/llm_model.rb

@@ -28,6 +28,9 @@ class LlmModel < ActiveRecord::Base
         organization: :text,
         disable_native_tools: :checkbox,
       },
+      mistral: {
+        disable_native_tools: :checkbox,
+      },
       google: {
         disable_native_tools: :checkbox,
       },

config/locales/client.en.yml

@@ -290,6 +290,8 @@ en:
           open_ai-o1-preview: "Open AI's most capabale reasoning model"
           samba_nova-Meta-Llama-3-1-8B-Instruct: "Efficient lightweight multilingual model"
           samba_nova-Meta-Llama-3-1-70B-Instruct": "Powerful multipurpose model"
+          mistral-mistral-large-latest: "Mistral's most powerful model"
+          mistral-pixtral-large-latest: "Mistral's most powerful vision capable model"
 
         configured:
           title: "Configured LLMs"
@@ -325,6 +327,7 @@ en:
           ollama: "Ollama"
           CDCK: "CDCK"
           samba_nova: "SambaNova"
+          mistral: "Mistral"
           fake: "Custom"
 
         provider_fields:

lib/completions/dialects/dialect.rb

@@ -16,6 +16,7 @@ module DiscourseAi
               DiscourseAi::Completions::Dialects::Claude,
               DiscourseAi::Completions::Dialects::Command,
               DiscourseAi::Completions::Dialects::Ollama,
+              DiscourseAi::Completions::Dialects::Mistral,
               DiscourseAi::Completions::Dialects::OpenAiCompatible,
             ]
           end

lib/completions/dialects/mistral.rb

@@ -0,0 +1,44 @@
+# frozen_string_literal: true
+
+# basically the same as Open AI, except for no support for user names
+
+module DiscourseAi
+  module Completions
+    module Dialects
+      class Mistral < ChatGpt
+        class << self
+          def can_translate?(model_provider)
+            model_provider == "mistral"
+          end
+        end
+
+        def translate
+          corrected = super
+          corrected.each do |msg|
+            msg[:content] = "" if msg[:tool_calls] && msg[:role] == "assistant"
+          end
+          corrected
+        end
+
+        private
+
+        def user_msg(msg)
+          mapped = super
+          if name = mapped.delete(:name)
+            if mapped[:content].is_a?(String)
+              mapped[:content] = "#{name}: #{mapped[:content]}"
+            else
+              mapped[:content].each do |inner|
+                if inner[:text]
+                  inner[:text] = "#{name}: #{inner[:text]}"
+                  break
+                end
+              end
+            end
+          end
+          mapped
+        end
+      end
+    end
+  end
+end

lib/completions/endpoints/aws_bedrock.rb

@@ -19,7 +19,10 @@ module DiscourseAi
         end
 
         def default_options(dialect)
-          options = { max_tokens: 3_000, anthropic_version: "bedrock-2023-05-31" }
+          max_tokens = 4096
+          max_tokens = 8192 if bedrock_model_id.match?(/3.5/)
+
+          options = { max_tokens: max_tokens, anthropic_version: "bedrock-2023-05-31" }
 
           options[:stop_sequences] = ["</function_calls>"] if !dialect.native_tool_support? &&
             dialect.prompt.has_tools?
@@ -40,15 +43,7 @@ module DiscourseAi
 
         private
 
-        def prompt_size(prompt)
-          # approximation
-          tokenizer.size(prompt.system_prompt.to_s + " " + prompt.messages.to_s)
-        end
-
-        def model_uri
-          region = llm_model.lookup_custom_param("region")
-
-          bedrock_model_id =
+        def bedrock_model_id
           case llm_model.name
           when "claude-2"
             "anthropic.claude-v2:1"
@@ -62,9 +57,20 @@ module DiscourseAi
             "anthropic.claude-3-opus-20240229-v1:0"
           when "claude-3-5-sonnet"
             "anthropic.claude-3-5-sonnet-20241022-v2:0"
+          when "claude-3-5-haiku"
+            "anthropic.claude-3-5-haiku-20241022-v1:0"
           else
             llm_model.name
           end
+        end
+
+        def prompt_size(prompt)
+          # approximation
+          tokenizer.size(prompt.system_prompt.to_s + " " + prompt.messages.to_s)
+        end
+
+        def model_uri
+          region = llm_model.lookup_custom_param("region")
 
           if region.blank? || bedrock_model_id.blank?
             raise CompletionFailed.new(I18n.t("discourse_ai.llm_models.bedrock_invalid_url"))

lib/completions/endpoints/base.rb

@@ -20,6 +20,7 @@ module DiscourseAi
               DiscourseAi::Completions::Endpoints::Anthropic,
               DiscourseAi::Completions::Endpoints::Cohere,
               DiscourseAi::Completions::Endpoints::SambaNova,
+              DiscourseAi::Completions::Endpoints::Mistral,
             ]
 
             endpoints << DiscourseAi::Completions::Endpoints::Ollama if Rails.env.development?

lib/completions/endpoints/mistral.rb

@@ -0,0 +1,17 @@
+# frozen_string_literal: true
+
+module DiscourseAi
+  module Completions
+    module Endpoints
+      class Mistral < OpenAi
+        def self.can_contact?(model_provider)
+          model_provider == "mistral"
+        end
+
+        def provider_id
+          AiApiAuditLog::Provider::Mistral
+        end
+      end
+    end
+  end
+end

lib/completions/llm.rb

@@ -90,6 +90,24 @@ module DiscourseAi
                   endpoint: "https://api.sambanova.ai/v1/chat/completions",
                   provider: "samba_nova",
                 },
+                {
+                  id: "mistral",
+                  models: [
+                    {
+                      name: "mistral-large-latest",
+                      tokens: 128_000,
+                      display_name: "Mistral Large",
+                    },
+                    {
+                      name: "pixtral-large-latest",
+                      tokens: 128_000,
+                      display_name: "Pixtral Large",
+                    },
+                  ],
+                  tokenizer: DiscourseAi::Tokenizer::MixtralTokenizer,
+                  endpoint: "https://api.mistral.ai/v1/chat/completions",
+                  provider: "mistral",
+                },
               ]
             end
         end
@@ -105,6 +123,7 @@ module DiscourseAi
             google
             azure
             samba_nova
+            mistral
           ]
           if !Rails.env.production?
             providers << "fake"

spec/fabricators/llm_model_fabricator.rb

@@ -90,6 +90,16 @@ Fabricator(:ollama_model, from: :llm_model) do
   provider_params { { enable_native_tool: true } }
 end
 
+Fabricator(:mistral_model, from: :llm_model) do
+  display_name "Mistral Large"
+  name "mistral-large-latest"
+  provider "mistral"
+  api_key "ABC"
+  tokenizer "DiscourseAi::Tokenizer::MixtralTokenizer"
+  url "https://api.mistral.ai/v1/chat/completions"
+  provider_params { { disable_native_tools: false } }
+end
+
 Fabricator(:seeded_model, from: :llm_model) do
   id "-2"
   display_name "CDCK Hosted Model"

spec/lib/completions/dialects/mistral_spec.rb

@@ -0,0 +1,81 @@
+# frozen_string_literal: true
+
+require "rails_helper"
+require_relative "dialect_context"
+
+RSpec.describe DiscourseAi::Completions::Dialects::Mistral do
+  fab!(:model) { Fabricate(:mistral_model) }
+  let(:context) { DialectContext.new(described_class, model) }
+  let(:image100x100) { plugin_file_from_fixtures("100x100.jpg") }
+  let(:upload100x100) do
+    UploadCreator.new(image100x100, "image.jpg").create_for(Discourse.system_user.id)
+  end
+
+  it "does not include user names" do
+    prompt =
+      DiscourseAi::Completions::Prompt.new(
+        messages: [type: :user, content: "Hello, I am Bob", id: "bob"],
+      )
+
+    dialect = described_class.new(prompt, model)
+
+    # mistral has no support for name
+    expect(dialect.translate).to eq([{ role: "user", content: "bob: Hello, I am Bob" }])
+  end
+
+  it "can properly encode images" do
+    model.update!(vision_enabled: true)
+
+    prompt =
+      DiscourseAi::Completions::Prompt.new(
+        "You are image bot",
+        messages: [type: :user, id: "user1", content: "hello", upload_ids: [upload100x100.id]],
+      )
+
+    encoded = prompt.encoded_uploads(prompt.messages.last)
+
+    image = "data:image/jpeg;base64,#{encoded[0][:base64]}"
+
+    dialect = described_class.new(prompt, model)
+
+    content = dialect.translate[1][:content]
+
+    expect(content).to eq(
+      [{ type: "image_url", image_url: { url: image } }, { type: "text", text: "user1: hello" }],
+    )
+  end
+
+  it "can properly map tool calls to mistral format" do
+    result = [
+      {
+        role: "system",
+        content:
+          "I want you to act as a title generator for written pieces. I will provide you with a text,\nand you will generate five attention-grabbing titles. Please keep the title concise and under 20 words,\nand ensure that the meaning is maintained. Replies will utilize the language type of the topic.\n",
+      },
+      { role: "user", content: "user1: This is a message by a user" },
+      { role: "assistant", content: "I'm a previous bot reply, that's why there's no user" },
+      { role: "user", content: "user1: This is a new message by a user" },
+      {
+        role: "assistant",
+        content: "",
+        tool_calls: [
+          {
+            type: "function",
+            function: {
+              arguments: "{\"location\":\"Sydney\",\"unit\":\"c\"}",
+              name: "get_weather",
+            },
+            id: "tool_id",
+          },
+        ],
+      },
+      {
+        role: "tool",
+        tool_call_id: "tool_id",
+        content: "\"I'm a tool result\"",
+        name: "get_weather",
+      },
+    ]
+    expect(context.multi_turn_scenario).to eq(result)
+  end
+end

spec/lib/completions/endpoints/aws_bedrock_spec.rb

@@ -26,6 +26,22 @@ RSpec.describe DiscourseAi::Completions::Endpoints::AwsBedrock do
     Aws::EventStream::Encoder.new.encode(aws_message)
   end
 
+  it "should provide accurate max token count" do
+    prompt = DiscourseAi::Completions::Prompt.new("hello")
+    dialect = DiscourseAi::Completions::Dialects::Claude.new(prompt, model)
+    endpoint = DiscourseAi::Completions::Endpoints::AwsBedrock.new(model)
+
+    model.name = "claude-2"
+    expect(endpoint.default_options(dialect)[:max_tokens]).to eq(4096)
+
+    model.name = "claude-3-5-sonnet"
+    expect(endpoint.default_options(dialect)[:max_tokens]).to eq(8192)
+
+    model.name = "claude-3-5-haiku"
+    options = endpoint.default_options(dialect)
+    expect(options[:max_tokens]).to eq(8192)
+  end
+
   describe "function calling" do
     it "supports old school xml function calls" do
       model.provider_params["disable_native_tools"] = true
@@ -246,7 +262,7 @@ RSpec.describe DiscourseAi::Completions::Endpoints::AwsBedrock do
         expect(response).to eq(expected_response)
 
         expected = {
-          "max_tokens" => 3000,
+          "max_tokens" => 4096,
           "anthropic_version" => "bedrock-2023-05-31",
           "messages" => [{ "role" => "user", "content" => "what is the weather in sydney" }],
           "tools" => [
@@ -305,7 +321,7 @@ RSpec.describe DiscourseAi::Completions::Endpoints::AwsBedrock do
       expect(request.headers["X-Amz-Content-Sha256"]).to be_present
 
       expected = {
-        "max_tokens" => 3000,
+        "max_tokens" => 4096,
         "anthropic_version" => "bedrock-2023-05-31",
         "messages" => [{ "role" => "user", "content" => "hello world" }],
         "system" => "You are a helpful bot",
@@ -354,7 +370,7 @@ RSpec.describe DiscourseAi::Completions::Endpoints::AwsBedrock do
         expect(request.headers["X-Amz-Content-Sha256"]).to be_present
 
         expected = {
-          "max_tokens" => 3000,
+          "max_tokens" => 4096,
           "anthropic_version" => "bedrock-2023-05-31",
           "messages" => [{ "role" => "user", "content" => "hello world" }],
           "system" => "You are a helpful bot",