feat: add MiniMax as first-class LLM provider (default: M2.7)

README.md

@@ -137,6 +137,31 @@ weclone-cli make-dataset
 ```
 More Parameter Details: [Data Preprocessing](https://docs.weclone.love/docs/deploy/data_preprocessing.html#related-parameters)
 
+#### Online LLM Providers for Data Cleaning
+
+When using `online_llm_clear`, you can set `llm_provider` to quickly configure a supported provider.  The provider presets automatically populate `base_url` and `model_name` (which can still be overridden explicitly):
+
+| Provider | `llm_provider` | Default Model | API Docs |
+|----------|----------------|---------------|----------|
+| OpenAI | `"openai"` | `gpt-4o-mini` | [platform.openai.com](https://platform.openai.com/docs) |
+| DeepSeek | `"deepseek"` | `deepseek-chat` | [platform.deepseek.com](https://platform.deepseek.com/docs) |
+| MiniMax | `"minimax"` | `MiniMax-M2.7` | [platform.minimax.io](https://platform.minimax.io/docs/api-reference/text-openai-api) |
+| Custom | `"custom"` | *(manual)* | Any OpenAI-compatible API |
+
+Example using MiniMax for data cleaning:
+```jsonc
+"make_dataset_args": {
+    "online_llm_clear": true,
+    "llm_provider": "minimax",
+    "llm_api_key": "your-minimax-api-key",
+    // base_url and model_name are auto-filled:
+    // base_url → https://api.minimax.io/v1
+    // model_name → MiniMax-M2.7
+}
+```
+
+> You can also use the China endpoint by explicitly setting `"base_url": "https://api.minimaxi.com/v1"`.
+
 ## Configure Parameters and Fine-tune Model
 
 - (Optional) Modify `model_name_or_path`, `template`, `lora_target` in `settings.jsonc` to select other locally downloaded models.   

README_zh.md

@@ -134,6 +134,31 @@ weclone-cli make-dataset
 ```
 数据处理更多参数说明：[数据预处理](https://docs.weclone.love/zh/docs/deploy/data_preprocessing.html#%E7%9B%B8%E5%85%B3%E5%8F%82%E6%95%B0)
 
+#### 在线 LLM 供应商（数据清洗）
+
+使用 `online_llm_clear` 时，可通过 `llm_provider` 快速配置供应商，自动填充 `base_url` 和 `model_name`（仍可手动覆盖）：
+
+| 供应商 | `llm_provider` | 默认模型 | API 文档 |
+|--------|----------------|----------|----------|
+| OpenAI | `"openai"` | `gpt-4o-mini` | [platform.openai.com](https://platform.openai.com/docs) |
+| DeepSeek | `"deepseek"` | `deepseek-chat` | [platform.deepseek.com](https://platform.deepseek.com/docs) |
+| MiniMax | `"minimax"` | `MiniMax-M2.7` | [platform.minimax.io](https://platform.minimax.io/docs/api-reference/text-openai-api) |
+| 自定义 | `"custom"` | *(手动填写)* | 任何 OpenAI 兼容 API |
+
+使用 MiniMax 进行数据清洗示例：
+```jsonc
+"make_dataset_args": {
+    "online_llm_clear": true,
+    "llm_provider": "minimax",
+    "llm_api_key": "your-minimax-api-key",
+    // base_url 和 model_name 自动填充：
+    // base_url → https://api.minimax.io/v1
+    // model_name → MiniMax-M2.7
+}
+```
+
+> 国内用户可手动设置 `"base_url": "https://api.minimaxi.com/v1"`（注意多一个 i）。
+
 ## 配置参数并微调模型
 
 - (可选)修改 `settings.jsonc` 的 `model_name_or_path` 、`template`、 `lora_target`选择本地下载好的其他模型。  

examples/mllm.template.jsonc

@@ -41,9 +41,12 @@
             }
         },
         "online_llm_clear": false,
+        // llm_provider: 可选 "openai", "deepseek", "minimax", "custom"
+        // 设置后会自动填充 base_url 和 model_name（若未显式指定）
+        // "llm_provider": "minimax",
         "base_url": "https://xxx/v1",
         "llm_api_key": "xxxxx",
-        "model_name": "xxx", //建议使用参数较大的模型，例如DeepSeek-V3
+        "model_name": "xxx", //建议使用参数较大的模型，例如DeepSeek-V3, MiniMax-M2.7
         "clean_batch_size": 10,
         "vision_api": {
             "enable": false, // 设置为 true 来开启此功能

examples/tg.template.jsonc

@@ -46,9 +46,13 @@
             }
         },
         "online_llm_clear": false,
+        // llm_provider: "openai", "deepseek", "minimax", or "custom"
+        // When set, auto-fills base_url and model_name if not explicitly provided.
+        // e.g. "minimax" → base_url: https://api.minimax.io/v1, model_name: MiniMax-M2.7
+        // "llm_provider": "minimax",
         "base_url": "https://xxx/v1",
         "llm_api_key": "xxxxx",
-        "model_name": "xxx", // Recommend using models with larger parameters, e.g. DeepSeek-V3
+        "model_name": "xxx", // Recommend using models with larger parameters, e.g. DeepSeek-V3, MiniMax-M2.7
         "clean_batch_size": 10,
         "vision_api": {
             "enable": false, // Set to true to enable this feature

settings.template.jsonc

@@ -46,9 +46,13 @@
             }
         },
         "online_llm_clear": false,
+        // llm_provider: 可选 "openai", "deepseek", "minimax", "custom"
+        // 设置后会自动填充 base_url 和 model_name（若未显式指定）
+        // 例如设为 "minimax" 时，base_url 默认 https://api.minimax.io/v1，model_name 默认 MiniMax-M2.7
+        // "llm_provider": "minimax",
         "base_url": "https://xxx/v1",
         "llm_api_key": "xxxxx",
-        "model_name": "xxx", //建议使用参数较大的模型，例如DeepSeek-V3
+        "model_name": "xxx", //建议使用参数较大的模型，例如DeepSeek-V3, MiniMax-M2.7
         "clean_batch_size": 50,
         "vision_api": {
             "enable": false, // 设置为 true 来开启此功能