1
00:00:00,000 --> 00:00:25,600
我們現在看到的就是Kodex的界面。然後我們可以看到這邊,這邊模型這裡我們可以看到這邊是有Deepseek,我已經把它接入了。然後下面有一個Olama本地的Jama4的模型,我已經把它接入了。下面是XAI的。比如說我用這個OlamaJama4這個是本地AI模型。說你好,你是什麼大模型。

2
00:00:25,600 --> 00:00:29,460
你好,我是JamaFall,大语言模型

3
00:00:29,460 --> 00:00:30,780
它是一个本地模型

4
00:00:30,780 --> 00:00:34,380
然后我们再选到Deepseq

5
00:00:34,380 --> 00:00:36,600
然后发给你好

6
00:00:36,600 --> 00:00:40,320
我们可以看到这是Deepseq的回答

7
00:00:40,320 --> 00:00:44,520
然后我们可以看到这个表情是非常经典的Deepseq的回答模式

8
00:00:44,520 --> 00:00:47,280
然后我们再来到OpenCodex

9
00:00:47,280 --> 00:00:49,620
这是它的控制面板的界面

10
00:00:49,620 --> 00:00:50,800
然后我们可以稍微看一下

11
00:00:50,800 --> 00:00:54,300
然后这边是我接入了Deepseq XAI

12
00:00:54,300 --> 00:00:54,940
欧拉玛

13
00:00:54,940 --> 00:00:56,180
欧拉玛是本地的

14
00:00:56,180 --> 00:00:58,460
我们看到这边是本地的一个URL

15
00:00:58,460 --> 00:01:01,180
然后下面是所有可用的模型

16
00:01:01,180 --> 00:01:02,300
也是显示在这的

17
00:01:02,300 --> 00:01:04,000
然后我们再看这边的提供方这里

18
00:01:04,000 --> 00:01:04,640
这里

19
00:01:04,640 --> 00:01:05,840
然后我们可以看到

20
00:01:05,840 --> 00:01:08,920
我们是可以接入很多的API平台的

21
00:01:08,920 --> 00:01:10,100
API平台

22
00:01:10,100 --> 00:01:11,300
然后这边可以点击这里

23
00:01:11,300 --> 00:01:13,020
然后我们看到这边有免费的

24
00:01:13,020 --> 00:01:14,280
然后我们可以看到

25
00:01:14,280 --> 00:01:15,780
这边是可以有很多选择的

26
00:01:15,780 --> 00:01:16,700
然后这边付费的

27
00:01:16,700 --> 00:01:19,380
也有很多接口跟我们选择

28
00:01:19,380 --> 00:01:21,920
然后它是不但可以接入Codex

29
00:01:21,920 --> 00:01:23,960
它还可以接入Clode

30
00:01:23,960 --> 00:01:25,760
我们可以看到这边是有可乐的选择的

31
00:01:26,260 --> 00:01:27,040
然后这边是kodex

32
00:01:27,040 --> 00:01:27,800
这边是kode

33
00:01:28,320 --> 00:01:30,360
然后我接入的是kodex

34
00:01:30,360 --> 00:01:30,880
然后点这里

35
00:01:30,880 --> 00:01:33,680
我们可以看到这是我之前调用的一些信息

36
00:01:33,940 --> 00:01:34,960
大家好我是超哥

37
00:01:34,960 --> 00:01:37,520
今天给大家分享openkodex的配置教程

38
00:01:37,780 --> 00:01:38,800
它可以让kodex

39
00:01:39,060 --> 00:01:42,400
接入任意第三方AI API和本地AI模型

40
00:01:42,900 --> 00:01:44,700
支持几乎所有主流AI

41
00:01:44,960 --> 00:01:46,480
确定摆脱官方token限制

42
00:01:46,740 --> 00:01:48,020
下面我们开始教程

43
00:01:48,280 --> 00:01:50,580
我会把所有用的材料放到我的这个博客的网址里

44
00:01:50,580 --> 00:01:54,240
我们首先是第一个是配置安装OpenCodex

45
00:01:54,240 --> 00:01:56,280
然后这是它的HeadApp项目地址

46
00:01:56,280 --> 00:01:57,360
我们也可以点开看一下

47
00:01:57,360 --> 00:02:00,560
我们往下拉

48
00:02:00,560 --> 00:02:02,640
它这边有个简体中文介绍

49
00:02:02,640 --> 00:02:05,320
介绍是非常详细的

50
00:02:05,320 --> 00:02:06,800
有兴趣的可以看一下

51
00:02:06,800 --> 00:02:09,280
然后我们下面第一个是

52
00:02:09,280 --> 00:02:11,180
安装Note GS环境

53
00:02:11,180 --> 00:02:12,440
然后这是它的下载地址

54
00:02:12,440 --> 00:02:13,320
我们可以点开

55
00:02:13,320 --> 00:02:15,920
点之后我们直接点这边的

56
00:02:15,920 --> 00:02:16,920
有一个下载按钮

57
00:02:16,920 --> 00:02:17,840
我们点击它

58
00:02:17,840 --> 00:02:19,980
然后已经下载好了

59
00:02:19,980 --> 00:02:21,480
我们直接点击安装

60
00:02:21,480 --> 00:02:24,180
这边点下一步

61
00:02:24,180 --> 00:02:25,900
然后这边我们给它勾上

62
00:02:25,900 --> 00:02:27,980
下面我们直接都默认就行了

63
00:02:27,980 --> 00:02:31,820
然后我们点这个finish

64
00:02:31,820 --> 00:02:32,980
然后就安装好了

65
00:02:32,980 --> 00:02:35,320
那我们可以在这边

66
00:02:35,320 --> 00:02:36,940
点这个windows

67
00:02:36,940 --> 00:02:38,880
然后我们搜索cmd

68
00:02:38,880 --> 00:02:42,100
因为这边是需要

69
00:02:42,100 --> 00:02:43,520
打开电脑的cmd

70
00:02:43,520 --> 00:02:45,500
然后之前安装成功检测命令

71
00:02:45,500 --> 00:02:46,580
我们看有没有安装成功

72
00:02:46,580 --> 00:02:48,020
然后我们点这边的copy

73
00:02:49,980 --> 00:02:55,040
让我们在cmd这里输入右击鼠标的右键

74
00:02:55,040 --> 00:02:56,780
让我们仍然粘贴

75
00:02:56,780 --> 00:02:58,380
可以看到这边有个版本号

76
00:02:58,380 --> 00:02:59,620
然后这边还有条命令

77
00:02:59,620 --> 00:03:01,620
让我们再按一次回车

78
00:03:01,620 --> 00:03:03,240
我们看到是两个都有版本号

79
00:03:03,240 --> 00:03:05,340
说明已经安装成功了

80
00:03:05,340 --> 00:03:06,840
然后下面是安装OpenCodex

81
00:03:06,840 --> 00:03:10,240
在安装之前我们先打开我们的TL GT

82
00:03:10,240 --> 00:03:12,240
就是Codex

83
00:03:12,240 --> 00:03:14,820
手机打开它

84
00:03:14,820 --> 00:03:17,020
现在就是这个Codex

85
00:03:17,020 --> 00:03:17,880
这之前是Codex

86
00:03:17,880 --> 00:03:20,320
让我们可以看这边的这个模型

87
00:03:20,320 --> 00:03:23,380
它是只有这一个模型选择的

88
00:03:23,380 --> 00:03:25,280
然后它是选不了别的模型的

89
00:03:25,280 --> 00:03:27,440
让我们先给它关掉

90
00:03:27,440 --> 00:03:29,140
让我们下面是

91
00:03:29,140 --> 00:03:30,480
安装OpenCodex

92
00:03:30,480 --> 00:03:32,120
然后在电脑的CMD里面

93
00:03:32,120 --> 00:03:33,480
执行下面的命令进行安装

94
00:03:33,480 --> 00:03:35,740
执行这条命令

95
00:03:35,740 --> 00:03:36,280
这条

96
00:03:36,280 --> 00:03:38,520
然后copy这条命令

97
00:03:38,520 --> 00:03:41,540
然后给它粘贴到这个CMD里面来

98
00:03:41,540 --> 00:03:44,620
然后按回撤进行安装

99
00:03:44,620 --> 00:03:46,280
然后下面是已经安装好了

100
00:03:46,280 --> 00:03:49,040
然后我们可以看到下面有一个提示

101
00:03:49,040 --> 00:03:50,680
当你碰到这个错误的时候

102
00:03:50,680 --> 00:03:51,680
就准备安装成功

103
00:03:51,680 --> 00:03:53,320
或者是拦截了安装脚本

104
00:03:53,320 --> 00:03:55,240
你可以执行这条命令进行安装

105
00:03:55,240 --> 00:03:56,200
再安装一次就行了

106
00:03:56,200 --> 00:03:57,840
因为我倒是没碰到

107
00:03:57,840 --> 00:03:59,600
所以就已经安装成功了

108
00:03:59,600 --> 00:04:02,400
然后我们下面是需要启动OpenCodex

109
00:04:02,400 --> 00:04:04,440
然后在CMD里面执行下面的命令

110
00:04:04,440 --> 00:04:06,640
这个是启动命令OpenCodex

111
00:04:06,640 --> 00:04:09,000
然后我们copy它

112
00:04:09,000 --> 00:04:10,280
然后打开CMD

113
00:04:10,280 --> 00:04:11,800
然后给它粘贴进来

114
00:04:11,800 --> 00:04:14,160
然后按回车

115
00:04:14,160 --> 00:04:15,380
我们看到这边有个网址

116
00:04:15,380 --> 00:04:18,020
然后下面就是已经启动成功了

117
00:04:18,020 --> 00:04:19,860
然后我们可以看博客这边

118
00:04:19,860 --> 00:04:23,360
启动完成后浏览器里面打开这个URL就行了

119
00:04:23,360 --> 00:04:24,600
就是这个URL

120
00:04:24,600 --> 00:04:25,820
你在这里打开是一样的

121
00:04:25,820 --> 00:04:28,480
然后直接在网页这打开

122
00:04:28,480 --> 00:04:29,500
然后这是它的首页

123
00:04:29,500 --> 00:04:30,200
我们可以看一下

124
00:04:30,200 --> 00:04:32,220
它是一些基础信息

125
00:04:32,220 --> 00:04:33,820
然后这边有个提供方

126
00:04:33,820 --> 00:04:37,000
这里就是添加模型或者API的地方

127
00:04:37,000 --> 00:04:39,400
然后这下面是我们正在使用的模型

128
00:04:39,400 --> 00:04:42,820
现在因为是只有OpenAI的

129
00:04:42,820 --> 00:04:44,180
它就只有OpenAI的

130
00:04:44,180 --> 00:04:45,860
然后下面有一个日志

131
00:04:45,860 --> 00:04:47,620
调试

132
00:04:47,620 --> 00:04:48,640
然后我们可以点开

133
00:04:48,640 --> 00:04:49,660
到时候我们调用的时候

134
00:04:49,660 --> 00:04:51,940
这边是有显示一些调入信息的

135
00:04:51,940 --> 00:04:53,880
然后我们可以来到这边

136
00:04:53,880 --> 00:04:57,040
如果是你想卸载OpenCodex的话

137
00:04:57,040 --> 00:04:57,980
这个是卸载命令

138
00:04:57,980 --> 00:05:00,260
也是粘贴到这个CMD里面

139
00:05:00,260 --> 00:05:00,940
然后执行一下

140
00:05:00,940 --> 00:05:02,080
它就卸载掉了

141
00:05:02,080 --> 00:05:03,760
然后我们可以看到第二个

142
00:05:03,760 --> 00:05:05,220
第二个就是介入第三方API

143
00:05:05,220 --> 00:05:07,940
然后我们是以Deepseq为例

144
00:05:07,940 --> 00:05:09,620
然后我们可以打开Deepseq的官网

145
00:05:09,620 --> 00:05:12,760
打开之后点这边的API开放平台

146
00:05:12,760 --> 00:05:16,680
然后可以点这边的API key

147
00:05:16,680 --> 00:05:18,960
然后点这边的创建API key

148
00:05:18,960 --> 00:05:20,820
然后这边命名一下

149
00:05:20,820 --> 00:05:22,500
比如opencodex

150
00:05:22,500 --> 00:05:25,080
然后给它创建一下

151
00:05:25,080 --> 00:05:28,020
这个就是API key

152
00:05:28,020 --> 00:05:29,060
可以点击复制

153
00:05:29,060 --> 00:05:32,300
然后我们再来到这个opencodex

154
00:05:32,300 --> 00:05:33,900
然后我们点这边的提供方

155
00:05:33,900 --> 00:05:36,200
就是添加API key的

156
00:05:36,200 --> 00:05:39,080
然后这边我们点这边的添加提供方

157
00:05:39,080 --> 00:05:41,760
那这边我们选这个付费

158
00:05:41,760 --> 00:05:43,720
因为DevSec它是付费的

159
00:05:43,720 --> 00:05:46,200
DevSec搜一下

160
00:05:46,200 --> 00:05:47,860
然后直接点这个

161
00:05:47,860 --> 00:05:48,660
这个就是

162
00:05:48,660 --> 00:05:50,740
我们可以看到这边是提供方名字

163
00:05:50,740 --> 00:05:51,600
然后这个是设备

164
00:05:51,600 --> 00:05:52,940
这也都已经自己填好了

165
00:05:52,940 --> 00:05:55,680
我们只需要修改这个APM1这里就行了

166
00:05:55,680 --> 00:05:58,220
然后把这个API可以给它粘贴进来

167
00:05:58,220 --> 00:06:00,540
然后我们直接点这边的添加提供方

168
00:06:00,540 --> 00:06:04,180
现在就已经添加成功了

169
00:06:04,180 --> 00:06:05,200
我们可以在这边看到

170
00:06:05,200 --> 00:06:06,320
这有DevSec的

171
00:06:06,320 --> 00:06:08,340
然后这个是有一个小绿灯

172
00:06:08,340 --> 00:06:11,620
然后这边也显示DevSec一起用

173
00:06:11,620 --> 00:06:13,800
然后我们可以来到

174
00:06:13,800 --> 00:06:15,560
Codex这里

175
00:06:15,560 --> 00:06:17,000
这个我们不要关

176
00:06:17,000 --> 00:06:20,700
这个关掉那个OpenCodex就执行不了了

177
00:06:20,700 --> 00:06:22,280
它这个终端不能关的

178
00:06:22,280 --> 00:06:24,960
然后我们来到Codex

179
00:06:24,960 --> 00:06:27,500
Codex我们可以看到

180
00:06:27,500 --> 00:06:29,820
现在这个模型还是只有OpenAI的

181
00:06:29,820 --> 00:06:31,360
然后我们给它退出一下

182
00:06:31,360 --> 00:06:33,380
点这边的退出

183
00:06:33,380 --> 00:06:34,380
然后我们重新进下去

184
00:06:34,380 --> 00:06:35,780
然后双击打开

185
00:06:35,780 --> 00:06:36,740
重新进一下

186
00:06:36,740 --> 00:06:39,100
重新进来之后

187
00:06:39,100 --> 00:06:40,140
然后我们点击这里

188
00:06:40,140 --> 00:06:42,060
我们可以看到这边就变了

189
00:06:42,060 --> 00:06:43,680
这个模型这里

190
00:06:43,680 --> 00:06:46,040
我们可以看到这边是有DeepSeq的

191
00:06:46,040 --> 00:06:46,920
我们可以选一个

192
00:06:46,920 --> 00:06:49,340
比如说你好

193
00:06:49,340 --> 00:06:52,080
发送

194
00:06:52,080 --> 00:06:56,100
我们可以看到这就是它已经回复

195
00:06:56,100 --> 00:06:59,140
然后我们再来到Omin Codex这边

196
00:06:59,140 --> 00:07:00,580
这边有一个日志这里

197
00:07:00,580 --> 00:07:04,320
我们可以看到它这边之前调用的是GDP的

198
00:07:04,320 --> 00:07:06,620
然后现在是调用的是DeepSeq的

199
00:07:06,620 --> 00:07:08,500
这个时间是16点01

200
00:07:08,500 --> 00:07:11,860
然后这个是时间对得上的

201
00:07:11,860 --> 00:07:16,340
让我们再发一个你是什么大模型

202
00:07:16,340 --> 00:07:20,540
他说Codex是OpenAI的

203
00:07:20,540 --> 00:07:23,500
其实它是它这个可能是底层逻辑问题

204
00:07:23,500 --> 00:07:26,100
所以它显示的还是Codex OpenAI

205
00:07:26,100 --> 00:07:28,580
它其实调用的是Deepseq的

206
00:07:28,580 --> 00:07:29,720
我们可以看到又增加了一条

207
00:07:29,720 --> 00:07:30,560
是Deepseq的

208
00:07:30,560 --> 00:07:33,840
让我们可以来到这里

209
00:07:33,840 --> 00:07:36,360
然后这个先给它关掉

210
00:07:36,360 --> 00:07:37,760
Deepseq的

211
00:07:37,760 --> 00:07:40,900
然后我们来到这边的提供方这里

212
00:07:40,900 --> 00:07:44,200
然后我们可以看到这边它是有很多账号可以登录的

213
00:07:44,200 --> 00:07:46,620
然后XAI就是X的

214
00:07:46,620 --> 00:07:47,880
然后这是Cloud的

215
00:07:47,880 --> 00:07:48,620
然后KMI

216
00:07:48,620 --> 00:07:49,940
然后这边都可以直接登录

217
00:07:49,940 --> 00:07:53,900
登录之后这边就会显示这个TAMIN的模型

218
00:07:53,900 --> 00:07:55,800
如果是你有充钱的话

219
00:07:55,800 --> 00:07:59,360
比如说我登录这个XAI的

220
00:07:59,360 --> 00:08:00,380
然后点这边的登录

221
00:08:12,136 --> 00:08:15,816
然后我在这里粘贴这个收钱码

222
00:08:15,816 --> 00:08:17,076
然后给它提交一下

223
00:08:17,076 --> 00:08:17,616
在这里

224
00:08:17,616 --> 00:08:20,496
再给它粘贴一下提交

225
00:08:20,496 --> 00:08:24,856
现在是已经登录到XAI了

226
00:08:24,856 --> 00:08:25,416
我们可以看到

227
00:08:25,416 --> 00:08:26,976
它这边是个账号显示的

228
00:08:26,976 --> 00:08:27,876
然后这边也是绿灯

229
00:08:27,876 --> 00:08:30,056
然后这边是已经成功了

230
00:08:30,056 --> 00:08:30,656
我给它关掉

231
00:08:30,656 --> 00:08:32,416
关掉

232
00:08:32,416 --> 00:08:35,016
然后已经登录到XAI

233
00:08:35,016 --> 00:08:37,056
然后我们再来到这个Kodex

234
00:08:37,056 --> 00:08:38,836
然后我们可以看到

235
00:08:38,836 --> 00:08:40,396
这边它是没变的

236
00:08:40,396 --> 00:08:43,396
然后我们还是需要重新给它退出一下这个Kodex

237
00:08:44,696 --> 00:08:47,196
让我们再双击这边的ChattTT

238
00:08:47,196 --> 00:08:48,796
就Kodex的图标

239
00:08:50,196 --> 00:08:51,136
让我们看一下

240
00:08:51,136 --> 00:08:52,196
好像是没有的

241
00:08:52,196 --> 00:08:57,396
可能是因为我这个XAI是没有开通这个会员的

242
00:08:57,396 --> 00:08:58,896
所以它是没显示的

243
00:08:58,896 --> 00:08:59,996
然后方法就是这样

244
00:08:59,996 --> 00:09:01,256
就是你在这边登录之后

245
00:09:01,256 --> 00:09:03,996
这边会显示它这个模型

246
00:09:03,996 --> 00:09:05,196
这边会显示它的模型的

247
00:09:05,196 --> 00:09:06,496
因为我这个是没开通会员

248
00:09:07,596 --> 00:09:09,496
然后我们下面看一下

249
00:09:09,496 --> 00:09:12,156
第三个是接入本地AI模型

250
00:09:12,156 --> 00:09:14,796
然后我们需要安装Olama第一步

251
00:09:14,796 --> 00:09:17,596
然后这是Windows下载地址我们点开

252
00:09:17,596 --> 00:09:19,696
然后我们直接点这个下载Windows版就行了

253
00:09:21,496 --> 00:09:22,496
这边允许一下

254
00:09:24,396 --> 00:09:25,296
然后已经下好了

255
00:09:25,296 --> 00:09:27,996
然后点击进行安装

256
00:09:27,996 --> 00:09:29,996
然后点这边的安装按钮

257
00:09:31,856 --> 00:09:33,896
我们看到Olama是已经装好了

258
00:09:33,896 --> 00:09:35,296
然后我们给它关掉

259
00:09:35,296 --> 00:09:36,596
这边给它关掉

260
00:09:36,596 --> 00:09:39,596
然后来到下面是下载本地AI模型

261
00:09:39,596 --> 00:09:41,196
这个是OLAMA模型列表

262
00:09:41,196 --> 00:09:43,436
上面有一些免费的模型让我们下载

263
00:09:43,436 --> 00:09:44,596
下载到本地

264
00:09:44,596 --> 00:09:46,276
然后比如说这边有一个

265
00:09:46,276 --> 00:09:47,996
JFOL点击

266
00:09:47,996 --> 00:09:50,356
然后它就可以下载到本地

267
00:09:50,356 --> 00:09:54,876
然后它的下载OLAMA的模型到本地的命令是这个

268
00:09:54,876 --> 00:09:55,956
OLAMA PURL

269
00:09:55,956 --> 00:09:58,596
然后模型名称

270
00:09:58,596 --> 00:10:01,076
然后点这边的COPY

271
00:10:01,076 --> 00:10:02,836
然后到这边

272
00:10:02,836 --> 00:10:05,276
新建一个吧

273
00:10:05,276 --> 00:10:06,556
新建一个

274
00:10:06,556 --> 00:10:07,636
你在这边搜一样的

275
00:10:07,636 --> 00:10:08,316
在这里

276
00:10:08,316 --> 00:10:08,916
Windows这里

277
00:10:08,916 --> 00:10:10,396
然后我们搜CMD

278
00:10:12,516 --> 00:10:14,276
CMD然后打开

279
00:10:14,276 --> 00:10:15,436
然后粘贴

280
00:10:15,436 --> 00:10:16,436
然后这个模型名称

281
00:10:16,436 --> 00:10:17,916
我们给它改一下

282
00:10:17,916 --> 00:10:19,476
改成这边的模型名称

283
00:10:19,476 --> 00:10:21,956
往下拉它这边是有模型名称的

284
00:10:21,956 --> 00:10:23,916
然后我们点这边后面的复制就行了

285
00:10:23,916 --> 00:10:24,556
这个复制

286
00:10:24,556 --> 00:10:26,636
比如说我这个14B的

287
00:10:26,636 --> 00:10:28,316
然后点这边的复制图标

288
00:10:28,316 --> 00:10:29,876
点击它

289
00:10:29,876 --> 00:10:32,516
然后粘贴到这里

290
00:10:32,516 --> 00:10:33,116
粘贴到这里

291
00:10:33,116 --> 00:10:33,836
Purl

292
00:10:33,836 --> 00:10:34,916
然后后面是空格

293
00:10:34,916 --> 00:10:36,436
然后模型名称给它粘贴进来

294
00:10:36,436 --> 00:10:37,456
然后按回车

295
00:10:37,456 --> 00:10:38,736
它就开始下载了

296
00:10:38,736 --> 00:10:40,336
我们等它下载完成比较大

297
00:10:40,336 --> 00:10:41,676
我们可以看到

298
00:10:41,676 --> 00:10:43,176
模型已经下载到本地了

299
00:10:43,176 --> 00:10:44,816
然后我们给它关掉

300
00:10:44,816 --> 00:10:47,776
然后这个也给它关掉

301
00:10:47,776 --> 00:10:51,776
然后我们再来到这个OpenCodex这里

302
00:10:51,776 --> 00:10:53,916
然后这边还是到提供方这里

303
00:10:53,916 --> 00:10:56,256
然后我们在这边添加提供方

304
00:10:56,256 --> 00:10:58,036
这边有个免费

305
00:10:58,036 --> 00:10:58,956
然后我们往下拉

306
00:10:58,956 --> 00:11:01,516
我们可以看到下面有一个偶拉码

307
00:11:01,516 --> 00:11:02,056
是本地

308
00:11:02,056 --> 00:11:03,236
我们点击它

309
00:11:03,236 --> 00:11:04,856
点击之后

310
00:11:04,856 --> 00:11:06,536
这边我们都不用动

311
00:11:06,536 --> 00:11:07,216
都是填好了

312
00:11:07,216 --> 00:11:09,276
我们直接点这边的添加提供方

313
00:11:09,276 --> 00:11:12,456
我们可以往下拉

314
00:11:12,456 --> 00:11:14,316
现在我们可以看到下面有一个偶拉码

315
00:11:14,316 --> 00:11:16,816
然后这个是之前登录了XAI

316
00:11:16,816 --> 00:11:17,536
这个Deepseek

317
00:11:17,536 --> 00:11:18,596
这三个是都在的

318
00:11:18,596 --> 00:11:20,036
这个偶拉码是本地的

319
00:11:20,036 --> 00:11:23,676
然后我们可以来到我们的Kodex

320
00:11:23,676 --> 00:11:25,856
Kodex这边还是没有的

321
00:11:25,856 --> 00:11:27,276
然后我们还是要给它退出

322
00:11:27,276 --> 00:11:29,436
退出Kodex

323
00:11:29,436 --> 00:11:32,236
然后我们再重新进入一下

324
00:11:32,236 --> 00:11:35,696
然后我们在这边

325
00:11:35,696 --> 00:11:37,036
我们再找一下这个模型

326
00:11:37,036 --> 00:11:38,536
我们可以看到

327
00:11:38,536 --> 00:11:40,896
现在这个是XAI的它模型出来

328
00:11:40,896 --> 00:11:41,676
但是不能用的

329
00:11:41,676 --> 00:11:44,376
因为我的是没有开通会员的

330
00:11:44,376 --> 00:11:46,776
然后现在我们看一下这个

331
00:11:46,776 --> 00:11:48,096
Olama这个

332
00:11:48,096 --> 00:11:50,996
就是刚刚我们下载的本地模型就在这了

333
00:11:50,996 --> 00:11:52,336
然后我们点击它

334
00:11:52,336 --> 00:11:54,876
点着我们在这边输入

335
00:11:54,876 --> 00:11:55,596
比如说你好

336
00:11:55,596 --> 00:11:58,816
你是什么大模型

337
00:11:58,816 --> 00:12:01,296
我们可以看到

338
00:12:01,296 --> 00:12:02,196
他的回答是

339
00:12:02,196 --> 00:12:03,736
我是一个大语言模型

340
00:12:03,736 --> 00:12:05,576
是由谷歌开发的开源模型

341
00:12:05,576 --> 00:12:07,236
这就是本地的

342
00:12:07,236 --> 00:12:08,836
Gemafall的模型

343
00:12:08,836 --> 00:12:11,136
然后我们再来到这里

344
00:12:11,136 --> 00:12:12,916
OpenCodex

345
00:12:12,916 --> 00:12:14,436
然后这边日识这里

346
00:12:14,436 --> 00:12:15,276
我们可以看到

347
00:12:15,276 --> 00:12:17,056
这边的调用的也是Gemafall

348
00:12:17,056 --> 00:12:18,656
是调用的本地的模型的

349
00:12:18,656 --> 00:12:21,316
然后我们再来到提供方这里

350
00:12:21,316 --> 00:12:23,216
我们还是点这边的

351
00:12:23,216 --> 00:12:24,716
添加提供方

352
00:12:24,716 --> 00:12:25,276
我们可以看到

353
00:12:25,276 --> 00:12:26,516
它的免费是有很多

354
00:12:26,516 --> 00:12:28,036
选项是可以选的

355
00:12:28,036 --> 00:12:29,916
也不只是说Olama

356
00:12:29,916 --> 00:12:33,076
像这个LMStudio也可以

357
00:12:33,076 --> 00:12:34,336
然后付费的

358
00:12:34,336 --> 00:12:37,876
几乎所有的主的AI模型都是可以添加的

359
00:12:37,876 --> 00:12:40,056
然后这边还可以直接添加这个账户

360
00:12:40,056 --> 00:12:40,856
然后我们往下拉

361
00:12:40,856 --> 00:12:42,876
这边是可以禁用的

362
00:12:42,876 --> 00:12:44,276
也可以禁用也可以删除

363
00:12:44,276 --> 00:12:44,916
你禁用之后

364
00:12:44,916 --> 00:12:45,956
它这个模型就用不了了

365
00:12:45,956 --> 00:12:46,976
比如说我点这个

366
00:12:46,976 --> 00:12:48,616
Olama的这个禁用

367
00:12:48,616 --> 00:12:51,036
就是Jamafo的这个本地的模型给它禁用

368
00:12:51,036 --> 00:12:53,236
现在已经禁用了

369
00:12:53,236 --> 00:12:55,296
然后再去执行一下

370
00:12:55,296 --> 00:12:56,616
你好

371
00:12:56,616 --> 00:12:58,716
看到能不能回复

372
00:12:58,716 --> 00:13:00,716
我们可以看到现在是已经连不上了

373
00:13:00,716 --> 00:13:04,056
有的朋友他可能会想这些

374
00:13:04,056 --> 00:13:06,436
比如说他想恢复这些模型

375
00:13:06,436 --> 00:13:08,056
就是这些多余的模型都给他去掉

376
00:13:08,056 --> 00:13:09,416
恢复到切记批

377
00:13:09,416 --> 00:13:10,756
直显示切记批的模型

378
00:13:10,756 --> 00:13:12,476
我们是直接把这个

379
00:13:12,476 --> 00:13:14,796
OpenCodex给他删掉

380
00:13:14,796 --> 00:13:15,456
删掉就行了

381
00:13:15,456 --> 00:13:17,036
就是还是之前的这条命令

382
00:13:17,036 --> 00:13:18,736
删掉OpenCodex的命令

383
00:13:18,736 --> 00:13:19,516
给他执行一下

384
00:13:19,516 --> 00:13:21,156
他这边就会恢复

385
00:13:21,156 --> 00:13:25,136
恢复成我们之前只有OpenAdd的这个模型

386
00:13:25,136 --> 00:13:26,336
好吧

387
00:13:26,336 --> 00:13:27,316
今天教程就到这吧

388
00:13:27,316 --> 00:13:28,456
如果大家有什么问题的话

389
00:13:28,456 --> 00:13:30,556
可以在视频下方的评论区留言交流

390
00:13:30,556 --> 00:13:31,696
谢谢大家的观看