Merge pull request #116 from Grt1228/1.0.12

1.0.12 tokens计算优化、删除模型接口修改、语音接口更新支持官方最新参数
Grt1228 · Apr 12, 2023 · bb22362 · bb22362
2 parents 517f9db + fd346c7
commit bb22362
Show file tree

Hide file tree

Showing 11 changed files with 236 additions and 77 deletions.
diff --git a/README.md b/README.md
@@ -1,9 +1,8 @@
 it’s an “unofficial" or "community-maintained” library.
-
 这是一个非官方的社区维护的库。
-OpenAi官方文档地址：https://platform.openai.com/docs/api-reference
+> 原创发布转载注明出处！
 ## 此项目不仅仅支持chat对话模型，支持openai官方所有api，包括
-- [x] tokens计算
+- [x] tokens            计算
 - [x] Billing           余额查询
 - [x] Models            模型检索
 - [x] Completions       chatgpt对话
@@ -31,6 +30,7 @@ WebSocket参考：[OpenAIWebSocketEventSourceListener](https://github.com/Grt122
 <img src="https://user-images.githubusercontent.com/27008803/225246389-7b452214-f3fe-4a70-bd3e-832a0ed34288.jpg" width="210" height="300" alt="二维码" />  | <img src="https://g-photo.oss-cn-shanghai.aliyuncs.com/hd15.jpg" width="210" height="210" alt="二维码" /> | <img src="https://user-images.githubusercontent.com/27008803/225246581-15e90f78-5438-4637-8e7d-14c68ca13b59.jpg" width="210" height="300" alt="二维码" />
 ---
 ## 更新日志
+- [x] 1.0.12  tokens计算优化、删除模型接口修改、语音接口更新支持官方最新参数
 - [x] 1.0.11  增加新的余额查询接口参考：[OpenAiClientTest](https://github.com/Grt1228/chatgpt-java/blob/main/src/test/java/com/unfbx/chatgpt/OpenAiClientTest.java) 和[OpenAiStreamClientTest](https://github.com/Grt1228/chatgpt-java/blob/main/src/test/java/com/unfbx/chatgpt/OpenAiStreamClientTest.java) ,修复tokens计算慢的问题，
 - [x] 1.0.10  支持tokens计算：[TikTokensTest](https://github.com/Grt1228/chatgpt-java/blob/main/src/test/java/com/unfbx/chatgpt/TikTokensTest.java) ，更多详细的资料参考文档：[Tokens_README.md](https://github.com/Grt1228/chatgpt-java/blob/main/Tokens_README.md)
 - [x] 1.0.9   支持自定义key使用策略参考：[OpenAiClientTest](https://github.com/Grt1228/chatgpt-java/blob/main/src/test/java/com/unfbx/chatgpt/OpenAiClientTest.java) 和[OpenAiStreamClientTest](https://github.com/Grt1228/chatgpt-java/blob/main/src/test/java/com/unfbx/chatgpt/OpenAiStreamClientTest.java) ，弃用ChatGPTClient，优化Moderation接口
@@ -72,7 +72,7 @@ OpenAi官方文档地址：https://platform.openai.com/docs/api-reference
 <dependency>
     <groupId>com.unfbx</groupId>
     <artifactId>chatgpt-java</artifactId>
-    <version>1.0.11</version>
+    <version>1.0.12</version>
 </dependency>
 ```
 ## 2、流式客户端使用示例：

diff --git a/pom.xml b/pom.xml
@@ -5,7 +5,7 @@
 
     <groupId>com.unfbx</groupId>
     <artifactId>chatgpt-java</artifactId>
-    <version>1.0.11</version>
+    <version>1.0.12</version>
     <name>chatgpt-java</name>
     <description>OpenAI Java SDK, OpenAI Api for Java. ChatGPT Java SDK .</description>
     <url>https://www.unfbx.com</url>
@@ -115,7 +115,7 @@
         <dependency>
             <groupId>com.knuddels</groupId>
             <artifactId>jtokkit</artifactId>
-            <version>0.1.0</version>
+            <version>0.2.0</version>
         </dependency>
     </dependencies>
 

diff --git a/src/main/java/com/unfbx/chatgpt/OpenAiApi.java b/src/main/java/com/unfbx/chatgpt/OpenAiApi.java
@@ -240,7 +240,7 @@ Single<UploadFileResponse> uploadFile(@Part MultipartBody.Part file,
      *
      * @return Single DeleteResponse
      */
-    @GET("v1/models/{model}")
+    @DELETE("v1/models/{model}")
     Single<FineTuneDeleteResponse> deleteFineTuneModel(@Path("model") String model);
 
 
@@ -277,26 +277,26 @@ Single<UploadFileResponse> uploadFile(@Part MultipartBody.Part file,
     /**
      * 语音转文字
      *
-     * @param model 模型
-     * @param file  语音文件
+     * @param file              语音文件
+     * @param requestBodyMap    参数
      * @return 文本
      */
     @Multipart
     @POST("v1/audio/transcriptions")
     Single<WhisperResponse> speechToTextTranscriptions(@Part MultipartBody.Part file,
-                                                       @Part("model") RequestBody model);
+                                                       @PartMap() Map<String, RequestBody> requestBodyMap);
 
     /**
      * 语音翻译：目前仅支持翻译为英文
      *
-     * @param model 模型
-     * @param file  语音文件
+     * @param file              语音文件
+     * @param requestBodyMap    参数
      * @return 文本
      */
     @Multipart
     @POST("v1/audio/translations")
     Single<WhisperResponse> speechToTextTranslations(@Part MultipartBody.Part file,
-                                                     @Part("model") RequestBody model);
+                                                     @PartMap() Map<String, RequestBody> requestBodyMap);
 
     /**
      * 余额查询

diff --git a/src/main/java/com/unfbx/chatgpt/OpenAiClient.java b/src/main/java/com/unfbx/chatgpt/OpenAiClient.java
@@ -30,7 +30,8 @@
 import com.unfbx.chatgpt.entity.models.ModelResponse;
 import com.unfbx.chatgpt.entity.moderations.Moderation;
 import com.unfbx.chatgpt.entity.moderations.ModerationResponse;
-import com.unfbx.chatgpt.entity.whisper.Whisper;
+import com.unfbx.chatgpt.entity.whisper.Transcriptions;
+import com.unfbx.chatgpt.entity.whisper.Translations;
 import com.unfbx.chatgpt.entity.whisper.WhisperResponse;
 import com.unfbx.chatgpt.exception.BaseException;
 import com.unfbx.chatgpt.exception.CommonError;
@@ -650,16 +651,32 @@ public ChatCompletionResponse chatCompletion(List<Message> messages) {
     /**
      * 语音转文字
      *
-     * @param model 模型 Whisper.Model
-     * @param file  语音文件 最大支持25MB mp3, mp4, mpeg, mpga, m4a, wav, webm
+     * @param transcriptions 参数
+     * @param file           语音文件 最大支持25MB mp3, mp4, mpeg, mpga, m4a, wav, webm
      * @return 语音文本
      */
-    public WhisperResponse speechToTextTranscriptions(java.io.File file, Whisper.Model model) {
-        checkSpeechFileSize(file);
+    public WhisperResponse speechToTextTranscriptions(java.io.File file, Transcriptions transcriptions) {
+        //文件
         RequestBody fileBody = RequestBody.create(MediaType.parse("multipart/form-data"), file);
         MultipartBody.Part multipartBody = MultipartBody.Part.createFormData("file", file.getName(), fileBody);
-        RequestBody modelBody = RequestBody.create(MediaType.parse("multipart/form-data"), model.getName());
-        Single<WhisperResponse> whisperResponse = this.openAiApi.speechToTextTranscriptions(multipartBody, modelBody);
+        //自定义参数
+        Map<String,RequestBody> requestBodyMap = new HashMap<>();
+        if (StrUtil.isNotBlank(transcriptions.getLanguage())) {
+            requestBodyMap.put(Transcriptions.Fields.language, RequestBody.create(MediaType.parse("multipart/form-data"), transcriptions.getLanguage()));
+        }
+        if (StrUtil.isNotBlank(transcriptions.getModel())) {
+            requestBodyMap.put(Transcriptions.Fields.model, RequestBody.create(MediaType.parse("multipart/form-data"), transcriptions.getModel()));
+        }
+        if (StrUtil.isNotBlank(transcriptions.getPrompt())) {
+            requestBodyMap.put(Transcriptions.Fields.prompt, RequestBody.create(MediaType.parse("multipart/form-data"), transcriptions.getPrompt()));
+        }
+        if (StrUtil.isNotBlank(transcriptions.getResponseFormat())) {
+            requestBodyMap.put(Transcriptions.Fields.responseFormat, RequestBody.create(MediaType.parse("multipart/form-data"), transcriptions.getResponseFormat()));
+        }
+        if (Objects.nonNull(transcriptions.getTemperature())) {
+            requestBodyMap.put(Transcriptions.Fields.temperature, RequestBody.create(MediaType.parse("multipart/form-data"), String.valueOf(transcriptions.getTemperature())));
+        }
+        Single<WhisperResponse> whisperResponse = this.openAiApi.speechToTextTranscriptions(multipartBody, requestBodyMap);
         return whisperResponse.blockingGet();
     }
 
@@ -670,24 +687,39 @@ public WhisperResponse speechToTextTranscriptions(java.io.File file, Whisper.Mod
      * @return 语音文本
      */
     public WhisperResponse speechToTextTranscriptions(java.io.File file) {
-        return this.speechToTextTranscriptions(file, Whisper.Model.WHISPER_1);
+        Transcriptions transcriptions = Transcriptions.builder().build();
+        return this.speechToTextTranscriptions(file, transcriptions);
 
     }
 
 
     /**
      * 语音翻译：目前仅支持翻译为英文
      *
-     * @param model 模型 Whisper.Model
-     * @param file  语音文件 最大支持25MB mp3, mp4, mpeg, mpga, m4a, wav, webm
+     * @param translations 参数
+     * @param file         语音文件 最大支持25MB mp3, mp4, mpeg, mpga, m4a, wav, webm
      * @return 翻译后文本
      */
-    public WhisperResponse speechToTextTranslations(java.io.File file, Whisper.Model model) {
-        checkSpeechFileSize(file);
+    public WhisperResponse speechToTextTranslations(java.io.File file, Translations translations) {
+        //文件
         RequestBody fileBody = RequestBody.create(MediaType.parse("multipart/form-data"), file);
         MultipartBody.Part multipartBody = MultipartBody.Part.createFormData("file", file.getName(), fileBody);
-        RequestBody modelBody = RequestBody.create(MediaType.parse("multipart/form-data"), model.getName());
-        Single<WhisperResponse> whisperResponse = this.openAiApi.speechToTextTranslations(multipartBody, modelBody);
+        //自定义参数
+        Map<String,RequestBody> requestBodyMap = new HashMap<>();
+
+        if (StrUtil.isNotBlank(translations.getModel())) {
+            requestBodyMap.put(Translations.Fields.model, RequestBody.create(MediaType.parse("multipart/form-data"), translations.getModel()));
+        }
+        if (StrUtil.isNotBlank(translations.getPrompt())) {
+            requestBodyMap.put(Translations.Fields.prompt, RequestBody.create(MediaType.parse("multipart/form-data"), translations.getPrompt()));
+        }
+        if (StrUtil.isNotBlank(translations.getResponseFormat())) {
+            requestBodyMap.put(Translations.Fields.responseFormat, RequestBody.create(MediaType.parse("multipart/form-data"), translations.getResponseFormat()));
+        }
+        if (Objects.nonNull(translations.getTemperature())) {
+            requestBodyMap.put(Translations.Fields.temperature, RequestBody.create(MediaType.parse("multipart/form-data"), String.valueOf(translations.getTemperature())));
+        }
+        Single<WhisperResponse> whisperResponse = this.openAiApi.speechToTextTranslations(multipartBody, requestBodyMap);
         return whisperResponse.blockingGet();
     }
 
@@ -698,7 +730,8 @@ public WhisperResponse speechToTextTranslations(java.io.File file, Whisper.Model
      * @return 翻译后文本
      */
     public WhisperResponse speechToTextTranslations(java.io.File file) {
-        return this.speechToTextTranslations(file, Whisper.Model.WHISPER_1);
+        Translations translations = Translations.builder().build();
+        return this.speechToTextTranslations(file, translations);
     }
 
     /**

diff --git a/src/main/java/com/unfbx/chatgpt/entity/fineTune/FineTuneDeleteResponse.java b/src/main/java/com/unfbx/chatgpt/entity/fineTune/FineTuneDeleteResponse.java
@@ -1,11 +1,8 @@
 package com.unfbx.chatgpt.entity.fineTune;
 
-import com.fasterxml.jackson.annotation.JsonProperty;
-import com.unfbx.chatgpt.entity.models.Permission;
 import lombok.Data;
 
 import java.io.Serializable;
-import java.util.List;
 
 @Data
 public class FineTuneDeleteResponse implements Serializable {
@@ -14,16 +11,6 @@ public class FineTuneDeleteResponse implements Serializable {
 
     private String object;
 
-    private long created;
-
-    @JsonProperty("owned_by")
-    private String ownedBy;
-
-    @JsonProperty("permission")
-    private List<Permission> permission;
-
-    private String root;
-
-    private String parent;
+    private boolean deleted;
 
 }
diff --git a/src/main/java/com/unfbx/chatgpt/entity/whisper/Transcriptions.java b/src/main/java/com/unfbx/chatgpt/entity/whisper/Transcriptions.java
@@ -0,0 +1,43 @@
+package com.unfbx.chatgpt.entity.whisper;
+
+import com.fasterxml.jackson.annotation.JsonInclude;
+import com.fasterxml.jackson.annotation.JsonProperty;
+import lombok.*;
+import lombok.experimental.FieldNameConstants;
+
+@Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
+@FieldNameConstants
+@JsonInclude(JsonInclude.Include.NON_NULL)
+public class Transcriptions extends Whisper {
+    /**
+     * 模型目前只支持这一种：WHISPER_1
+     */
+    @Builder.Default
+    private String model = Whisper.Model.WHISPER_1.getName();
+    /**
+     * 提示语，需要与语音语言匹配
+     */
+    private String prompt;
+    /**
+     * 输出的格式，采用以下选项之一：json、text、srt、verbose_json 或 vtt。
+     * 默认值：json
+     */
+    @JsonProperty("response_format")
+    @Builder.Default
+    private String responseFormat = ResponseFormat.JSON.getName();
+    /**
+     * 温度控制随机效果：0-1，值越大输出更加随机
+     * 默认值：0
+     */
+    @Builder.Default
+    private Double temperature = 0d;
+    /**
+     * 输入音频的语言,以 ISO-639-1 格式提供输入语言将提高准确性和延迟。
+     * 参考：<a href=https://baike.baidu.com/item/ISO%20639-1/8292914>ISO-639-1</a>
+     */
+    private String language;
+
+}
diff --git a/src/main/java/com/unfbx/chatgpt/entity/whisper/Translations.java b/src/main/java/com/unfbx/chatgpt/entity/whisper/Translations.java
@@ -0,0 +1,38 @@
+package com.unfbx.chatgpt.entity.whisper;
+
+import com.fasterxml.jackson.annotation.JsonInclude;
+import com.fasterxml.jackson.annotation.JsonProperty;
+import lombok.*;
+import lombok.experimental.FieldNameConstants;
+
+@Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
+@FieldNameConstants
+@JsonInclude(JsonInclude.Include.NON_NULL)
+public class Translations {
+    /**
+     * 模型目前只支持这一种：WHISPER_1
+     */
+    @Builder.Default
+    private String model = Whisper.Model.WHISPER_1.getName();
+    /**
+     * 提示语，需要与语音语言匹配
+     */
+    private String prompt;
+    /**
+     * 输出的格式，采用以下选项之一：json、text、srt、verbose_json 或 vtt。
+     * 默认值：json
+     */
+    @JsonProperty("response_format")
+    @Builder.Default
+    private String responseFormat = Whisper.ResponseFormat.JSON.getName();
+    /**
+     * 温度控制随机效果：0-1，值越大输出更加随机
+     * 默认值：0
+     */
+    @Builder.Default
+    private double temperature = 0;
+
+}
diff --git a/src/main/java/com/unfbx/chatgpt/entity/whisper/Whisper.java b/src/main/java/com/unfbx/chatgpt/entity/whisper/Whisper.java
@@ -10,17 +10,29 @@
  * 描述：语音转文字
  *
  * @author https:www.unfbx.com
- * @since  2023-03-02
+ * @since 2023-03-02
  */
 @Data
 public class Whisper implements Serializable {
 
 
     @Getter
     @AllArgsConstructor
-    public enum Model{
+    public enum Model {
         WHISPER_1("whisper-1"),
         ;
         private String name;
     }
+
+    @Getter
+    @AllArgsConstructor
+    public enum ResponseFormat {
+        JSON("json"),
+        TEXT("text"),
+        SRT("srt"),
+        VERBOSE_JSON("verbose_json"),
+        VTT("vtt"),
+        ;
+        private String name;
+    }
 }