Handling new o1 models (params, system messages, json mode)

cequence-io · Jan 3, 2025 · 81cf970 · 81cf970
1 parent 1249852
commit 81cf970
Show file tree

Hide file tree

Showing 3 changed files with 41 additions and 21 deletions.
diff --git a/...src/main/scala/io/cequence/openaiscala/service/impl/OpenAIChatCompletionServiceImpl.scala b/...src/main/scala/io/cequence/openaiscala/service/impl/OpenAIChatCompletionServiceImpl.scala
@@ -47,13 +47,25 @@ private[service] trait OpenAIChatCompletionServiceImpl
 
 trait ChatCompletionBodyMaker {
 
-  private val o1Models = Set(
+  private val noSystemMessageModels = Set(
+    ModelId.o1_preview,
+    ModelId.o1_preview_2024_09_12,
+    ModelId.o1_mini,
+    ModelId.o1_mini_2024_09_12
+  )
+
+  private val o1PreviewModels = Set(
     ModelId.o1_preview,
     ModelId.o1_preview_2024_09_12,
     ModelId.o1_mini,
     ModelId.o1_mini_2024_09_12
   )
 
+  private val o1Models = Set(
+    ModelId.o1,
+    ModelId.o1_2024_12_17
+  )
+
   protected def createBodyParamsForChatCompletion(
     messagesAux: Seq[BaseMessage],
     settings: CreateChatCompletionSettings,
@@ -63,7 +75,7 @@ trait ChatCompletionBodyMaker {
 
     // O1 models needs some special treatment... revisit this later
     val messagesFinal =
-      if (o1Models.contains(settings.model))
+      if (noSystemMessageModels.contains(settings.model))
         MessageConversions.systemToUserMessages(messagesAux)
       else
         messagesAux
@@ -72,8 +84,10 @@ trait ChatCompletionBodyMaker {
 
     // O1 models needs some special treatment... revisit this later
     val settingsFinal =
-      if (o1Models.contains(settings.model))
-        ChatCompletionSettingsConversions.o1Specific(settings)
+      if (o1PreviewModels.contains(settings.model))
+        ChatCompletionSettingsConversions.o1Preview(settings)
+      else if (o1Models.contains(settings.model))
+        ChatCompletionSettingsConversions.o1(settings)
       else
         settings
 

diff --git a/openai-core/src/main/scala/io/cequence/openaiscala/service/OpenAIChatCompletionExtra.scala b/openai-core/src/main/scala/io/cequence/openaiscala/service/OpenAIChatCompletionExtra.scala
@@ -126,9 +126,10 @@ object OpenAIChatCompletionExtra {
   }
 
   private val defaultJsonSchemaModels = Seq(
-    "openai-" + ModelId.gpt_4o_2024_08_06,
-    ModelId.gpt_4o_2024_08_06
-  )
+    ModelId.gpt_4o_2024_08_06,
+    ModelId.o1,
+    ModelId.o1_2024_12_17
+  ).flatMap(id => Seq(id, "openai-" + id))
 
   def handleOutputJsonSchema(
     messages: Seq[BaseMessage],
@@ -144,7 +145,7 @@ object OpenAIChatCompletionExtra {
 
     val (settingsFinal, addJsonToPrompt) =
       if (jsonSchemaModels.contains(settings.model)) {
-        logger.debug(
+        logger.info(
           s"Using OpenAI json schema mode for ${taskNameForLogging} and the model '${settings.model}' - name: ${jsonSchemaDef.name}, strict: ${jsonSchemaDef.strict}, structure:\n${jsonSchemaString}"
         )
 
@@ -157,7 +158,7 @@ object OpenAIChatCompletionExtra {
       } else {
         // otherwise we failover to json object format and pass json schema to the user prompt
 
-        logger.debug(
+        logger.info(
           s"Using JSON object mode for ${taskNameForLogging} and the model '${settings.model}'. Also passing a JSON schema as part of a user prompt."
         )
 

diff --git a/...ain/scala/io/cequence/openaiscala/service/adapter/ChatCompletionSettingsConversions.scala b/...ain/scala/io/cequence/openaiscala/service/adapter/ChatCompletionSettingsConversions.scala
@@ -32,7 +32,7 @@ object ChatCompletionSettingsConversions {
         } else acc
     }
 
-  private val o1Conversions = Seq(
+  private val o1BaseConversions = Seq(
     // max tokens
     FieldConversionDef(
       _.max_tokens.isDefined,
@@ -79,18 +79,23 @@ object ChatCompletionSettingsConversions {
         "O1 models don't support frequency penalty values other than the default of 0, converting to 0."
       ),
       warning = true
-    ),
-    // frequency_penalty
-    FieldConversionDef(
-      settings =>
-        settings.response_format_type.isDefined && settings.response_format_type.get != ChatCompletionResponseFormatType.text,
-      _.copy(response_format_type = None),
-      Some(
-        "O1 models don't support json object/schema response format, converting to None."
-      ),
-      warning = true
     )
   )
 
-  val o1Specific: SettingsConversion = generic(o1Conversions)
+  private val o1PreviewConversions =
+    o1BaseConversions :+
+      // response format type
+      FieldConversionDef(
+        settings =>
+          settings.response_format_type.isDefined && settings.response_format_type.get != ChatCompletionResponseFormatType.text,
+        _.copy(response_format_type = None),
+        Some(
+          "O1 models don't support json object/schema response format, converting to None."
+        ),
+        warning = true
+      )
+
+  val o1: SettingsConversion = generic(o1BaseConversions)
+
+  val o1Preview: SettingsConversion = generic(o1PreviewConversions)
 }