Cerebras, Groq, and TogetherAI examples updated to use Llama4 models

Author: peterbanda

openai-examples/src/main/scala/io/cequence/openaiscala/examples/cerebras/CerebrasCreateChatCompletion.scala

@@ -25,7 +25,7 @@ object CerebrasCreateChatCompletion extends ExampleBase[OpenAIChatCompletionServ
   )
 
 //  private val modelId = NonOpenAIModelId.deepseek_r1_distill_llama_70b
-  private val modelId = NonOpenAIModelId.llama_3_3_70b
+  private val modelId = NonOpenAIModelId.cerebras_llama_4_scout_17b_16e_instruct
 
   override protected def run: Future[_] =
     service

openai-examples/src/main/scala/io/cequence/openaiscala/examples/cerebras/CerebrasCreateChatCompletionJSON.scala

@@ -0,0 +1,67 @@
+package io.cequence.openaiscala.examples.cerebras
+
+import io.cequence.openaiscala.domain._
+import io.cequence.openaiscala.domain.settings.GroqCreateChatCompletionSettingsOps._
+import io.cequence.openaiscala.domain.settings.{
+  ChatCompletionResponseFormatType,
+  CreateChatCompletionSettings,
+  JsonSchemaDef
+}
+import io.cequence.openaiscala.examples.{ChatCompletionProvider, ExampleBase}
+import io.cequence.openaiscala.service.OpenAIChatCompletionExtra.OpenAIChatCompletionImplicits
+import io.cequence.openaiscala.service.OpenAIChatCompletionService
+import play.api.libs.json.{JsObject, Json}
+
+import scala.concurrent.Future
+
+/**
+ * Requires `CEREBRAS_API_KEY` environment variable to be set.
+ */
+object CerebrasCreateChatCompletionJSON
+    extends ExampleBase[OpenAIChatCompletionService] {
+
+  override val service: OpenAIChatCompletionService = ChatCompletionProvider.cerebras
+
+  private val jsonSchema: JsonSchema = JsonSchema.Object(
+    properties = Seq(
+      "response" -> JsonSchema.Array(
+        items = JsonSchema.Object(
+          properties = Seq(
+            "city" -> JsonSchema.String(),
+            "temperature" -> JsonSchema.String(),
+            "weather" -> JsonSchema.String()
+          ),
+          required = Seq("city", "temperature", "weather")
+        )
+      )
+    ),
+    required = Seq("response")
+  )
+
+  private val messages = Seq(
+    SystemMessage("You are a helpful weather assistant that responds in JSON."),
+    UserMessage("What is the weather like in Norway?")
+  )
+
+  private val modelId = NonOpenAIModelId.cerebras_llama_4_scout_17b_16e_instruct
+
+  // TODO: Cerebras should support JSON schema response format (without a conversion to "json mode")
+  override protected def run: Future[_] =
+    service
+      .createChatCompletionWithJSON[JsObject](
+        messages = messages,
+        settings = CreateChatCompletionSettings(
+          model = modelId,
+          temperature = Some(0.1),
+          response_format_type = Some(ChatCompletionResponseFormatType.json_schema),
+          jsonSchema = Some(
+            JsonSchemaDef(
+              name = "weather_response",
+              strict = true,
+              structure = jsonSchema
+            )
+          )
+        ).setMaxCompletionTokens(4000)
+      )
+      .map(json => println(Json.prettyPrint(json)))
+}

openai-examples/src/main/scala/io/cequence/openaiscala/examples/cerebras/CerebrasCreateChatCompletionStreamed.scala

@@ -25,7 +25,7 @@ object CerebrasCreateChatCompletionStreamed
     UserMessage("What is the weather like in Norway?")
   )
 
-  private val modelId = NonOpenAIModelId.llama3_1_70b
+  private val modelId = NonOpenAIModelId.cerebras_llama_4_scout_17b_16e_instruct
 
   override protected def run: Future[_] =
     service

openai-examples/src/main/scala/io/cequence/openaiscala/examples/groq/GroqCreateChatCompletion.scala

@@ -19,7 +19,7 @@ object GroqCreateChatCompletion extends ExampleBase[OpenAIChatCompletionService]
     UserMessage("What is the weather like in Norway?")
   )
 
-  private val modelId = NonOpenAIModelId.llama_4_scout_17b_16e_instruct
+  private val modelId = NonOpenAIModelId.groq_llama_4_scout_17b_16e_instruct
 
   override protected def run: Future[_] =
     service

openai-examples/src/main/scala/io/cequence/openaiscala/examples/groq/GroqCreateChatCompletionJSONWithDeepseekR1.scala

@@ -43,7 +43,7 @@ object GroqCreateChatCompletionJSONWithDeepseekR1
     UserMessage("What is the weather like in Norway?")
   )
 
-  private val modelId = NonOpenAIModelId.deepseek_r1_distill_llama_70b
+  private val modelId = NonOpenAIModelId.deepseek_r1_distill_qwen_32b // deepseek_r1_distill_llama_70b
 
   override protected def run: Future[_] =
     service

openai-examples/src/main/scala/io/cequence/openaiscala/examples/sonar/SonarCreateChatCompletionStreamedWithOpenAIAdapter.scala

@@ -8,10 +8,12 @@ import io.cequence.openaiscala.service.OpenAIChatCompletionStreamedServiceExtra
 
 import scala.concurrent.Future
 import io.cequence.openaiscala.examples.ChatCompletionProvider
+import io.cequence.openaiscala.perplexity.service.SonarServiceConsts
 
 // requires `openai-scala-client-stream` as a dependency and `SONAR_API_KEY` environment variable to be set
 object SonarCreateChatCompletionStreamedWithOpenAIAdapter
-    extends ExampleBase[OpenAIChatCompletionStreamedServiceExtra] {
+    extends ExampleBase[OpenAIChatCompletionStreamedServiceExtra]
+    with SonarServiceConsts {
 
   override val service: OpenAIChatCompletionStreamedServiceExtra =
     ChatCompletionProvider.sonar
@@ -30,7 +32,10 @@ object SonarCreateChatCompletionStreamedWithOpenAIAdapter
         settings = CreateChatCompletionSettings(
           model = modelId,
           temperature = Some(0.01),
-          max_tokens = Some(512)
+          max_tokens = Some(512),
+          extra_params = Map(
+            includeCitationsInTextResponseParam -> true
+          )
         )
       )
       .runWith(

openai-examples/src/main/scala/io/cequence/openaiscala/examples/togetherai/TogetherAICreateChatCompletion.scala

@@ -21,7 +21,7 @@ object TogetherAICreateChatCompletion extends ExampleBase[OpenAIChatCompletionSe
 
   // deepseek_ai_deepseek_r1_distill_llama_70b_free // deepseek_ai_deepseek_v3
   private val modelId =
-    NonOpenAIModelId.llama_4_maverick_17B_128E_instruct_fp8
+    NonOpenAIModelId.meta_llama_llama_4_maverick_17b_128e_instruct_fp8
 
   override protected def run: Future[_] =
     service