feat: Allow to specify OpenAI custom instance (#327)

davidmigloz · Feb 14, 2024 · 4744648 · 4744648
1 parent 2cbd538
commit 4744648
Show file tree

Hide file tree

Showing 7 changed files with 78 additions and 9 deletions.
diff --git a/packages/langchain_openai/lib/src/chat_models/chat_openai.dart b/packages/langchain_openai/lib/src/chat_models/chat_openai.dart
@@ -301,6 +301,7 @@ class ChatOpenAI extends BaseChatModel<ChatOpenAIOptions> {
       functionCall: functionCall,
       frequencyPenalty:
           options?.frequencyPenalty ?? defaultOptions.frequencyPenalty,
+      instanceId: options?.instanceId ?? defaultOptions.instanceId,
       logitBias: options?.logitBias ?? defaultOptions.logitBias,
       maxTokens: options?.maxTokens ?? defaultOptions.maxTokens,
       n: options?.n ?? defaultOptions.n,

diff --git a/packages/langchain_openai/lib/src/chat_models/models/models.dart b/packages/langchain_openai/lib/src/chat_models/models/models.dart
@@ -8,6 +8,7 @@ class ChatOpenAIOptions extends ChatModelOptions {
   const ChatOpenAIOptions({
     this.model = 'gpt-3.5-turbo',
     this.frequencyPenalty,
+    this.instanceId,
     this.logitBias,
     this.maxTokens,
     this.n,
@@ -53,6 +54,10 @@ class ChatOpenAIOptions extends ChatModelOptions {
   /// See https://platform.openai.com/docs/api-reference/chat/create#chat-create-frequency_penalty
   final double? frequencyPenalty;
 
+  /// An unique identifier to a custom instance to execute the request.
+  /// The requesting organization is required to have access to the instance.
+  final String? instanceId;
+
   /// Modify the likelihood of specified tokens appearing in the completion.
   ///
   /// See https://platform.openai.com/docs/api-reference/chat/create#chat-create-logit_bias

diff --git a/packages/openai_dart/lib/src/generated/schema/create_chat_completion_request.dart b/packages/openai_dart/lib/src/generated/schema/create_chat_completion_request.dart
@@ -28,6 +28,9 @@ class CreateChatCompletionRequest with _$CreateChatCompletionRequest {
     @Default(0.0)
     double? frequencyPenalty,
 
+    /// An unique identifier to a custom instance to execute the request. The requesting organization is required to have access to the instance.
+    @JsonKey(name: 'instance_id', includeIfNull: false) String? instanceId,
+
     /// Modify the likelihood of specified tokens appearing in the completion.
     ///
     /// Accepts a JSON object that maps tokens (specified by their token ID in the tokenizer) to an associated bias value from -100 to 100. Mathematically, the bias is added to the logits generated by the model prior to sampling. The exact effect will vary per model, but values between -1 and 1 should decrease or increase likelihood of selection; values like -100 or 100 should result in a ban or exclusive selection of the relevant token.
@@ -55,7 +58,7 @@ class CreateChatCompletionRequest with _$CreateChatCompletionRequest {
     @Default(0.0)
     double? presencePenalty,
 
-    /// An object specifying the format that the model must output. Compatible with [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo) and `gpt-3.5-turbo-1106`.
+    /// An object specifying the format that the model must output. Compatible with [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo) and all GPT-3.5 Turbo models newer than `gpt-3.5-turbo-1106`.
     ///
     /// Setting to `{ "type": "json_object" }` enables JSON mode, which guarantees the message the model generates is valid JSON.
     ///
@@ -129,6 +132,7 @@ class CreateChatCompletionRequest with _$CreateChatCompletionRequest {
     'model',
     'messages',
     'frequency_penalty',
+    'instance_id',
     'logit_bias',
     'logprobs',
     'top_logprobs',
@@ -216,6 +220,7 @@ class CreateChatCompletionRequest with _$CreateChatCompletionRequest {
       'model': model,
       'messages': messages,
       'frequency_penalty': frequencyPenalty,
+      'instance_id': instanceId,
       'logit_bias': logitBias,
       'logprobs': logprobs,
       'top_logprobs': topLogprobs,
@@ -273,6 +278,8 @@ enum ChatCompletionModels {
   gpt35Turbo0613,
   @JsonValue('gpt-3.5-turbo-1106')
   gpt35Turbo1106,
+  @JsonValue('gpt-3.5-turbo-0125')
+  gpt35Turbo0125,
   @JsonValue('gpt-3.5-turbo-16k-0613')
   gpt35Turbo16k0613,
 }
@@ -337,7 +344,7 @@ class _ChatCompletionModelConverter
 // CLASS: ChatCompletionResponseFormat
 // ==========================================
 
-/// An object specifying the format that the model must output. Compatible with [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo) and `gpt-3.5-turbo-1106`.
+/// An object specifying the format that the model must output. Compatible with [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo) and all GPT-3.5 Turbo models newer than `gpt-3.5-turbo-1106`.
 ///
 /// Setting to `{ "type": "json_object" }` enables JSON mode, which guarantees the message the model generates is valid JSON.
 ///

diff --git a/packages/openai_dart/lib/src/generated/schema/schema.freezed.dart b/packages/openai_dart/lib/src/generated/schema/schema.freezed.dart
@@ -3356,6 +3356,10 @@ mixin _$CreateChatCompletionRequest {
   @JsonKey(name: 'frequency_penalty', includeIfNull: false)
   double? get frequencyPenalty => throw _privateConstructorUsedError;
 
+  /// An unique identifier to a custom instance to execute the request. The requesting organization is required to have access to the instance.
+  @JsonKey(name: 'instance_id', includeIfNull: false)
+  String? get instanceId => throw _privateConstructorUsedError;
+
   /// Modify the likelihood of specified tokens appearing in the completion.
   ///
   /// Accepts a JSON object that maps tokens (specified by their token ID in the tokenizer) to an associated bias value from -100 to 100. Mathematically, the bias is added to the logits generated by the model prior to sampling. The exact effect will vary per model, but values between -1 and 1 should decrease or increase likelihood of selection; values like -100 or 100 should result in a ban or exclusive selection of the relevant token.
@@ -3386,7 +3390,7 @@ mixin _$CreateChatCompletionRequest {
   @JsonKey(name: 'presence_penalty', includeIfNull: false)
   double? get presencePenalty => throw _privateConstructorUsedError;
 
-  /// An object specifying the format that the model must output. Compatible with [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo) and `gpt-3.5-turbo-1106`.
+  /// An object specifying the format that the model must output. Compatible with [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo) and all GPT-3.5 Turbo models newer than `gpt-3.5-turbo-1106`.
   ///
   /// Setting to `{ "type": "json_object" }` enables JSON mode, which guarantees the message the model generates is valid JSON.
   ///
@@ -3479,6 +3483,7 @@ abstract class $CreateChatCompletionRequestCopyWith<$Res> {
       List<ChatCompletionMessage> messages,
       @JsonKey(name: 'frequency_penalty', includeIfNull: false)
       double? frequencyPenalty,
+      @JsonKey(name: 'instance_id', includeIfNull: false) String? instanceId,
       @JsonKey(name: 'logit_bias', includeIfNull: false)
       Map<String, int>? logitBias,
       @JsonKey(includeIfNull: false) bool? logprobs,
@@ -3530,6 +3535,7 @@ class _$CreateChatCompletionRequestCopyWithImpl<$Res,
     Object? model = null,
     Object? messages = null,
     Object? frequencyPenalty = freezed,
+    Object? instanceId = freezed,
     Object? logitBias = freezed,
     Object? logprobs = freezed,
     Object? topLogprobs = freezed,
@@ -3561,6 +3567,10 @@ class _$CreateChatCompletionRequestCopyWithImpl<$Res,
           ? _value.frequencyPenalty
           : frequencyPenalty // ignore: cast_nullable_to_non_nullable
               as double?,
+      instanceId: freezed == instanceId
+          ? _value.instanceId
+          : instanceId // ignore: cast_nullable_to_non_nullable
+              as String?,
       logitBias: freezed == logitBias
           ? _value.logitBias
           : logitBias // ignore: cast_nullable_to_non_nullable
@@ -3706,6 +3716,7 @@ abstract class _$$CreateChatCompletionRequestImplCopyWith<$Res>
       List<ChatCompletionMessage> messages,
       @JsonKey(name: 'frequency_penalty', includeIfNull: false)
       double? frequencyPenalty,
+      @JsonKey(name: 'instance_id', includeIfNull: false) String? instanceId,
       @JsonKey(name: 'logit_bias', includeIfNull: false)
       Map<String, int>? logitBias,
       @JsonKey(includeIfNull: false) bool? logprobs,
@@ -3761,6 +3772,7 @@ class __$$CreateChatCompletionRequestImplCopyWithImpl<$Res>
     Object? model = null,
     Object? messages = null,
     Object? frequencyPenalty = freezed,
+    Object? instanceId = freezed,
     Object? logitBias = freezed,
     Object? logprobs = freezed,
     Object? topLogprobs = freezed,
@@ -3792,6 +3804,10 @@ class __$$CreateChatCompletionRequestImplCopyWithImpl<$Res>
           ? _value.frequencyPenalty
           : frequencyPenalty // ignore: cast_nullable_to_non_nullable
               as double?,
+      instanceId: freezed == instanceId
+          ? _value.instanceId
+          : instanceId // ignore: cast_nullable_to_non_nullable
+              as String?,
       logitBias: freezed == logitBias
           ? _value._logitBias
           : logitBias // ignore: cast_nullable_to_non_nullable
@@ -3872,6 +3888,7 @@ class _$CreateChatCompletionRequestImpl extends _CreateChatCompletionRequest {
       required final List<ChatCompletionMessage> messages,
       @JsonKey(name: 'frequency_penalty', includeIfNull: false)
       this.frequencyPenalty = 0.0,
+      @JsonKey(name: 'instance_id', includeIfNull: false) this.instanceId,
       @JsonKey(name: 'logit_bias', includeIfNull: false)
       final Map<String, int>? logitBias,
       @JsonKey(includeIfNull: false) this.logprobs,
@@ -3929,6 +3946,11 @@ class _$CreateChatCompletionRequestImpl extends _CreateChatCompletionRequest {
   @JsonKey(name: 'frequency_penalty', includeIfNull: false)
   final double? frequencyPenalty;
 
+  /// An unique identifier to a custom instance to execute the request. The requesting organization is required to have access to the instance.
+  @override
+  @JsonKey(name: 'instance_id', includeIfNull: false)
+  final String? instanceId;
+
   /// Modify the likelihood of specified tokens appearing in the completion.
   ///
   /// Accepts a JSON object that maps tokens (specified by their token ID in the tokenizer) to an associated bias value from -100 to 100. Mathematically, the bias is added to the logits generated by the model prior to sampling. The exact effect will vary per model, but values between -1 and 1 should decrease or increase likelihood of selection; values like -100 or 100 should result in a ban or exclusive selection of the relevant token.
@@ -3976,7 +3998,7 @@ class _$CreateChatCompletionRequestImpl extends _CreateChatCompletionRequest {
   @JsonKey(name: 'presence_penalty', includeIfNull: false)
   final double? presencePenalty;
 
-  /// An object specifying the format that the model must output. Compatible with [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo) and `gpt-3.5-turbo-1106`.
+  /// An object specifying the format that the model must output. Compatible with [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo) and all GPT-3.5 Turbo models newer than `gpt-3.5-turbo-1106`.
   ///
   /// Setting to `{ "type": "json_object" }` enables JSON mode, which guarantees the message the model generates is valid JSON.
   ///
@@ -4080,7 +4102,7 @@ class _$CreateChatCompletionRequestImpl extends _CreateChatCompletionRequest {
 
   @override
   String toString() {
-    return 'CreateChatCompletionRequest(model: $model, messages: $messages, frequencyPenalty: $frequencyPenalty, logitBias: $logitBias, logprobs: $logprobs, topLogprobs: $topLogprobs, maxTokens: $maxTokens, n: $n, presencePenalty: $presencePenalty, responseFormat: $responseFormat, seed: $seed, stop: $stop, stream: $stream, temperature: $temperature, topP: $topP, tools: $tools, toolChoice: $toolChoice, user: $user, functionCall: $functionCall, functions: $functions)';
+    return 'CreateChatCompletionRequest(model: $model, messages: $messages, frequencyPenalty: $frequencyPenalty, instanceId: $instanceId, logitBias: $logitBias, logprobs: $logprobs, topLogprobs: $topLogprobs, maxTokens: $maxTokens, n: $n, presencePenalty: $presencePenalty, responseFormat: $responseFormat, seed: $seed, stop: $stop, stream: $stream, temperature: $temperature, topP: $topP, tools: $tools, toolChoice: $toolChoice, user: $user, functionCall: $functionCall, functions: $functions)';
   }
 
   @override
@@ -4092,6 +4114,8 @@ class _$CreateChatCompletionRequestImpl extends _CreateChatCompletionRequest {
             const DeepCollectionEquality().equals(other._messages, _messages) &&
             (identical(other.frequencyPenalty, frequencyPenalty) ||
                 other.frequencyPenalty == frequencyPenalty) &&
+            (identical(other.instanceId, instanceId) ||
+                other.instanceId == instanceId) &&
             const DeepCollectionEquality()
                 .equals(other._logitBias, _logitBias) &&
             (identical(other.logprobs, logprobs) ||
@@ -4128,6 +4152,7 @@ class _$CreateChatCompletionRequestImpl extends _CreateChatCompletionRequest {
         model,
         const DeepCollectionEquality().hash(_messages),
         frequencyPenalty,
+        instanceId,
         const DeepCollectionEquality().hash(_logitBias),
         logprobs,
         topLogprobs,
@@ -4170,6 +4195,8 @@ abstract class _CreateChatCompletionRequest
       required final List<ChatCompletionMessage> messages,
       @JsonKey(name: 'frequency_penalty', includeIfNull: false)
       final double? frequencyPenalty,
+      @JsonKey(name: 'instance_id', includeIfNull: false)
+      final String? instanceId,
       @JsonKey(name: 'logit_bias', includeIfNull: false)
       final Map<String, int>? logitBias,
       @JsonKey(includeIfNull: false) final bool? logprobs,
@@ -4222,6 +4249,11 @@ abstract class _CreateChatCompletionRequest
   double? get frequencyPenalty;
   @override
 
+  /// An unique identifier to a custom instance to execute the request. The requesting organization is required to have access to the instance.
+  @JsonKey(name: 'instance_id', includeIfNull: false)
+  String? get instanceId;
+  @override
+
   /// Modify the likelihood of specified tokens appearing in the completion.
   ///
   /// Accepts a JSON object that maps tokens (specified by their token ID in the tokenizer) to an associated bias value from -100 to 100. Mathematically, the bias is added to the logits generated by the model prior to sampling. The exact effect will vary per model, but values between -1 and 1 should decrease or increase likelihood of selection; values like -100 or 100 should result in a ban or exclusive selection of the relevant token.
@@ -4258,7 +4290,7 @@ abstract class _CreateChatCompletionRequest
   double? get presencePenalty;
   @override
 
-  /// An object specifying the format that the model must output. Compatible with [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo) and `gpt-3.5-turbo-1106`.
+  /// An object specifying the format that the model must output. Compatible with [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo) and all GPT-3.5 Turbo models newer than `gpt-3.5-turbo-1106`.
   ///
   /// Setting to `{ "type": "json_object" }` enables JSON mode, which guarantees the message the model generates is valid JSON.
   ///

diff --git a/packages/openai_dart/lib/src/generated/schema/schema.g.dart b/packages/openai_dart/lib/src/generated/schema/schema.g.dart
diff --git a/packages/openai_dart/oas/openapi_curated.yaml b/packages/openai_dart/oas/openapi_curated.yaml
@@ -1481,6 +1481,7 @@ components:
                   "gpt-3.5-turbo-0301",
                   "gpt-3.5-turbo-0613",
                   "gpt-3.5-turbo-1106",
+                  "gpt-3.5-turbo-0125",
                   "gpt-3.5-turbo-16k-0613",
                 ]
         messages:
@@ -1496,6 +1497,11 @@ components:
           maximum: 2
           nullable: true
           description: *completions_frequency_penalty_description
+        instance_id:
+          type: string
+          default: null
+          nullable: true
+          description: An unique identifier to a custom instance to execute the request. The requesting organization is required to have access to the instance.
         logit_bias:
           type: object
           default: null
@@ -1542,7 +1548,7 @@ components:
           title: ChatCompletionResponseFormat
           type: object
           description: |
-            An object specifying the format that the model must output. Compatible with [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo) and `gpt-3.5-turbo-1106`.
+            An object specifying the format that the model must output. Compatible with [GPT-4 Turbo](https://platform.openai.com/docs/models/gpt-4-and-gpt-4-turbo) and all GPT-3.5 Turbo models newer than `gpt-3.5-turbo-1106`.
 
             Setting to `{ "type": "json_object" }` enables JSON mode, which guarantees the message the model generates is valid JSON.
 

diff --git a/packages/openai_dart/oas/openapi_official.yaml b/packages/openai_dart/oas/openapi_official.yaml
@@ -5806,6 +5806,7 @@ components:
                   "gpt-3.5-turbo-0301",
                   "gpt-3.5-turbo-0613",
                   "gpt-3.5-turbo-1106",
+                  "gpt-3.5-turbo-0125",
                   "gpt-3.5-turbo-16k-0613",
                 ]
           x-oaiTypeLabel: string
@@ -5816,6 +5817,11 @@ components:
           maximum: 2
           nullable: true
           description: *completions_frequency_penalty_description
+        instance_id:
+          type: string
+          default: null
+          nullable: true
+          description: An unique identifier to a custom instance to execute the request. The requesting organization is required to have access to the instance.
         logit_bias:
           type: object
           x-oaiTypeLabel: map
@@ -5863,7 +5869,7 @@ components:
         response_format:
           type: object
           description: |
-            An object specifying the format that the model must output. Compatible with [GPT-4 Turbo](/docs/models/gpt-4-and-gpt-4-turbo) and `gpt-3.5-turbo-1106`.
+            An object specifying the format that the model must output. Compatible with [GPT-4 Turbo](/docs/models/gpt-4-and-gpt-4-turbo) and all GPT-3.5 Turbo models newer than `gpt-3.5-turbo-1106`.
 
             Setting to `{ "type": "json_object" }` enables JSON mode, which guarantees the message the model generates is valid JSON.
 
@@ -6839,6 +6845,16 @@ components:
             The sampling temperature, between 0 and 1. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic. If set to 0, the model will use [log probability](https://en.wikipedia.org/wiki/Log_probability) to automatically increase the temperature until certain thresholds are hit.
           type: number
           default: 0
+        timestamp_granularities[]:
+          description: |
+            The timestamp granularities to populate for this transcription. Any of these options: `word`, or `segment`. Note: There is no additional latency for segment timestamps, but generating word timestamps incurs additional latency.
+          type: array
+          items:
+            type: string
+            enum:
+              - word
+              - segment
+          default: [segment]
       required:
         - file
         - model
@@ -8388,7 +8404,6 @@ components:
           description: |
             An array of tool calls the run step was involved in. These can be associated with one of three types of tools: `code_interpreter`, `retrieval`, or `function`.
           items:
-            type: object
             oneOf:
               - $ref: "#/components/schemas/RunStepDetailsToolCallsCodeObject"
               - $ref: "#/components/schemas/RunStepDetailsToolCallsRetrievalObject"