add temperature and max_tokens parameters to complete method

Author: brauliopf

src/backend/src/modules/puterai/ClaudeService.js

@@ -128,8 +128,8 @@ class ClaudeService extends BaseService {
                     const init_chat_stream = async ({ chatStream }) => {
                         const completion = await this.anthropic.messages.stream({
                             model: model ?? this.get_default_model(),
-                            max_tokens: max_tokens || (model === 'claude-3-5-sonnet-20241022' || model === 'claude-3-5-sonnet-20240620') ? 8192 : 4096,
-                            temperature: temperature || 0,
+                            max_tokens: max_tokens || (model === 'claude-3-5-sonnet-20241022' || model === 'claude-3-5-sonnet-20240620') ? 8192 : 4096, //required
+                            temperature: temperature || 0, // required
                             system: PUTER_PROMPT + JSON.stringify(system_prompts),
                             messages,
                             ...(tools ? { tools } : {}),

src/backend/src/modules/puterai/DeepSeekService.js

@@ -119,7 +119,7 @@ class DeepSeekService extends BaseService {
              * AI Chat completion method.
              * See AIChatService for more details.
              */
-            async complete ({ messages, stream, model, tools }) {
+            async complete ({ messages, stream, model, tools, max_tokens, temperature }) {
                 model = this.adapt_model(model);
 
                 messages = await OpenAIUtil.process_input_messages(messages);
@@ -169,7 +169,8 @@ class DeepSeekService extends BaseService {
                     messages,
                     model: model ?? this.get_default_model(),
                     ...(tools ? { tools } : {}),
-                    max_tokens: 1000,
+                    max_tokens: max_tokens || 1000,
+                    temperature, // the default temperature is 1.0. suggested 0 for math/coding and 1.5 for creative poetry
                     stream,
                     ...(stream ? {
                         stream_options: { include_usage: true },

src/backend/src/modules/puterai/GroqAIService.js

@@ -97,7 +97,7 @@ class GroqAIService extends BaseService {
             * @param {boolean} [options.stream] - Whether to stream the response
             * @returns {TypedValue|Object} Returns either a TypedValue with streaming response or completion object with usage stats
             */
-            async complete ({ messages, model, stream, tools }) {
+            async complete ({ messages, model, stream, tools, max_tokens, temperature }) {
                 model = model ?? this.get_default_model();
 
                 messages = await OpenAIUtil.process_input_messages(messages);
@@ -113,6 +113,8 @@ class GroqAIService extends BaseService {
                     model,
                     stream,
                     tools,
+                    max_completion_tokens: max_tokens, // max_tokens has been deprecated
+                    temperature
                 });
 
                 return OpenAIUtil.handle_completion_output({

src/backend/src/modules/puterai/MistralAIService.js

@@ -216,7 +216,7 @@ class MistralAIService extends BaseService {
              * AI Chat completion method.
              * See AIChatService for more details.
              */
-            async complete ({ messages, stream, model, tools }) {
+            async complete ({ messages, stream, model, tools, max_tokens, temperature }) {
 
                 messages = await OpenAIUtil.process_input_messages(messages);
                 for ( const message of messages ) {
@@ -238,6 +238,8 @@ class MistralAIService extends BaseService {
                     model: model ?? this.get_default_model(),
                     ...(tools ? { tools } : {}),
                     messages,
+                    max_tokens: max_tokens,
+                    temperature
                 });
 
                 return await OpenAIUtil.handle_completion_output({

src/backend/src/modules/puterai/OpenAICompletionService.js

@@ -163,12 +163,14 @@ class OpenAICompletionService extends BaseService {
                 return model_names;
             },
 
-            async complete ({ messages, test_mode, stream, model, tools }) {
+            async complete ({ messages, test_mode, stream, model, tools, max_tokens, temperature }) {
                 return await this.complete(messages, {
                     model: model,
                     tools,
                     moderation: true,
                     stream,
+                    max_tokens,
+                    temperature
                 });
             }
         }

src/backend/src/modules/puterai/OpenRouterService.js

@@ -116,7 +116,7 @@ class OpenRouterService extends BaseService {
              * AI Chat completion method.
              * See AIChatService for more details.
              */
-            async complete ({ messages, stream, model, tools }) {
+            async complete ({ messages, stream, model, tools, max_tokens, temperature }) {
                 model = this.adapt_model(model);
 
                 if ( model.startsWith('openrouter:') ) {
@@ -134,7 +134,8 @@ class OpenRouterService extends BaseService {
                     messages,
                     model: model ?? this.get_default_model(),
                     ...(tools ? { tools } : {}),
-                    max_tokens: 1000,
+                    max_tokens: max_tokens || 1000,
+                    temperature: temperature, // default to 1.0
                     stream,
                     ...(stream ? {
                         stream_options: { include_usage: true },