1
00:00:01,500 --> 00:00:01,890
Hello.

2
00:00:01,890 --> 00:00:02,520
Welcome back.

3
00:00:03,180 --> 00:00:08,840
So let's sum this up underscore execute it.

4
00:00:08,910 --> 00:00:15,530
I don't see implement all the functions that we needed to initialize and run our offerings.

5
00:00:15,540 --> 00:00:17,730
There's this function here known as a I in it.

6
00:00:17,740 --> 00:00:24,060
And this function is to create and initialized in your network and it's given explanations here.

7
00:00:24,120 --> 00:00:28,410
All of these explanations were generated by the Cuban mix.

8
00:00:28,440 --> 00:00:34,820
So you start off by this data structure to provide the inference for the activation slush working memory.

9
00:00:34,860 --> 00:00:37,040
This the code that does that.

10
00:00:37,170 --> 00:00:43,380
And then it creates an instance of the neural network and then it's initialized as the neural network.

11
00:00:43,380 --> 00:00:45,510
This is known as the E I in it.

12
00:00:45,630 --> 00:00:51,870
And this function is called You know what it makes cube in its function which should be somewhere here.

13
00:00:52,020 --> 00:00:54,510
So our max cube a.

14
00:00:54,510 --> 00:01:02,810
In its function is simply create chaos A.I. in its imposes activations here this argument this was a

15
00:01:02,860 --> 00:01:09,450
I run looks like it takes a pointer to the input data and then another pointer to the output buffer.

16
00:01:09,450 --> 00:01:14,110
And then over here it checks for input data.

17
00:01:14,160 --> 00:01:16,870
It says if input data doesn't exist.

18
00:01:16,890 --> 00:01:21,060
If it is false or output data is false or neural network is false.

19
00:01:21,180 --> 00:01:22,820
Just return an error.

20
00:01:22,950 --> 00:01:27,040
Ls initialize input and output buffer handles.

21
00:01:27,060 --> 00:01:31,230
And it starts by initializing the batch size to 1 and then input data.

22
00:01:31,230 --> 00:01:39,430
Here is a pointer to in data and then batch size 1 for outputs and then output data.

23
00:01:39,480 --> 00:01:46,980
So this all of this generated by Cuba makes less to some final touches to the um the generated code

24
00:01:47,040 --> 00:01:49,680
that we have over here over here.

25
00:01:49,680 --> 00:01:56,950
You know what a AI in its function which is called by the cube and makes a init function.

26
00:01:57,030 --> 00:02:00,250
I'm going to let's say Prince the arrow.

27
00:02:00,420 --> 00:02:06,390
In case we get an arrow let's be able to print this 2 2 0 0 0 minutes.

28
00:02:07,080 --> 00:02:08,400
So that we can know what's going on.

29
00:02:09,210 --> 00:02:15,210
I'm going to say for instance if I were to type

30
00:02:18,360 --> 00:02:20,880
it's not equal to zero none.

31
00:02:28,510 --> 00:02:32,220
If that's the case then we can go ahead to print the arrow.

32
00:02:32,220 --> 00:02:33,610
I'm simply going to say print f

33
00:02:38,690 --> 00:02:39,200
over here.

34
00:02:39,200 --> 00:02:53,580
I'll see Arrow or I'll use cup to Arrow shoot arrow and then print the type pass percentage T over here

35
00:02:53,730 --> 00:03:01,150
and then the arrow could pass percentage D over here as well and then correct return New Line.

36
00:03:01,150 --> 00:03:10,090
In no way I simply do a prototype and then I wrote a code like this.

37
00:03:12,300 --> 00:03:12,750
Right.

38
00:03:14,220 --> 00:03:14,780
So.

39
00:03:15,120 --> 00:03:18,050
So we get to see the okay.

40
00:03:18,540 --> 00:03:20,980
So let's take a look at our max cube a.

41
00:03:21,020 --> 00:03:23,190
In its function.

42
00:03:23,460 --> 00:03:24,460
Okay.

43
00:03:24,550 --> 00:03:30,520
It's simply initialize as this and then cause a war in its function right.

44
00:03:33,510 --> 00:03:36,780
Let's do the same for a what a run function.

45
00:03:37,700 --> 00:03:41,400
If I hear we I'm gonna simplify this.

46
00:03:41,400 --> 00:03:47,820
We're not going to be using in batch so I'll just take this off like this such that we simply do a run

47
00:03:48,540 --> 00:03:52,570
and then we can check for a work error.

48
00:03:53,580 --> 00:03:55,290
Print out the arrow as well.

49
00:04:01,750 --> 00:04:02,140
K

50
00:04:05,050 --> 00:04:06,860
I'll just simplify this a bit for you.

51
00:04:08,120 --> 00:04:14,730
Rather than simply return the error could I'm going to send it to our serial monitor like we did earlier.

52
00:04:14,750 --> 00:04:15,620
I'll see a print f

53
00:04:19,380 --> 00:04:19,820
arrow

54
00:04:24,080 --> 00:04:25,710
and then type of the arrow

55
00:04:28,610 --> 00:04:36,460
pass percentage T over here and then the error could percentage T over here.

56
00:04:36,500 --> 00:04:43,820
Now do comma erode the type error code like this.

57
00:04:44,040 --> 00:04:46,100
Also I'm going to take this out here.

58
00:04:46,110 --> 00:04:47,890
This check.

59
00:04:47,970 --> 00:04:50,060
Um we always gonna have input data.

60
00:04:50,070 --> 00:04:51,880
We always gonna have output data.

61
00:04:51,940 --> 00:04:58,800
No one in your network you can it or you can keep it if you want and when to simplify to like this often

62
00:04:58,800 --> 00:05:02,290
when data function is that you are using for the first time.

63
00:05:02,460 --> 00:05:10,410
I preferred them you know looking as essential as possible so that you can be able to grasp the core

64
00:05:10,440 --> 00:05:17,060
concept before other supplementary things are added to the function things like checking whether there

65
00:05:17,060 --> 00:05:21,360
is in your network or there is input data or output buffer.

66
00:05:21,510 --> 00:05:22,620
Those aren't part of the core.

67
00:05:22,620 --> 00:05:26,000
This is the core function to run the neural network.

68
00:05:26,040 --> 00:05:32,940
This over here to print the arrow can be deleted as well or you need to do is this over here right.

69
00:05:32,970 --> 00:05:42,780
So this are what I run the a run is called by the is called by our E I O R E A process function and

70
00:05:42,780 --> 00:05:47,410
I'm going to clean everything into a process all of this bits I'm going to clean it and we simply going

71
00:05:47,410 --> 00:05:53,180
to call the E I run this function all of this snow.

72
00:05:53,690 --> 00:05:57,520
It's not really required.

73
00:05:57,520 --> 00:06:00,350
Um where is it a process function

74
00:06:03,650 --> 00:06:05,360
okay starts from here.

75
00:06:08,350 --> 00:06:14,270
So all we have all we have into a process function is simply a run right.

76
00:06:14,330 --> 00:06:14,910
Like this

77
00:06:18,380 --> 00:06:19,820
and we already

78
00:06:28,740 --> 00:06:35,460
we already changed d d e I run we simplify the I run us.

79
00:06:35,460 --> 00:06:35,790
Well

80
00:06:39,820 --> 00:06:41,710
so um one other thing.

81
00:06:42,760 --> 00:06:43,700
Yeah.

82
00:06:44,170 --> 00:06:45,490
This function prototype.

83
00:06:45,490 --> 00:06:49,440
We've changed it so we've got to update this function prototype.

84
00:06:50,200 --> 00:06:54,970
So our main function the prototype for the AI process is this one over here.

85
00:06:54,970 --> 00:06:59,380
It takes to document the input data and the outputs buffer.

86
00:06:59,410 --> 00:07:02,110
So this is going to be the new prototype.

87
00:07:02,110 --> 00:07:08,770
I'm gonna copy this from here and this will become the prototype of the def and this will become the

88
00:07:08,770 --> 00:07:10,060
function definition.

89
00:07:11,430 --> 00:07:19,570
So I'm gonna paste this over here and avoid here and the data type for this is a float pointer

90
00:07:22,590 --> 00:07:24,810
and this is also a float pointer

91
00:07:30,330 --> 00:07:32,660
and I'm going to clean this like this.

92
00:07:32,660 --> 00:07:37,460
And um put this over here.

93
00:07:37,460 --> 00:07:44,400
I'm going to update the DOT each file as well it's over here.

94
00:07:44,400 --> 00:07:46,830
Gonna Replace this over here with this new one.

95
00:07:46,850 --> 00:07:49,810
And then this and the cool on here.

96
00:07:50,030 --> 00:07:50,620
Right.

97
00:07:50,630 --> 00:07:59,980
So this the look of a wah wah wah up underscore X cube dot dot see.

98
00:08:00,110 --> 00:08:05,780
Very simple or two important functions are the A.I. in it and a process.

99
00:08:05,810 --> 00:08:12,740
And most of their work is encapsulated in these internal functions.

100
00:08:12,740 --> 00:08:13,240
Right.

101
00:08:13,250 --> 00:08:19,700
So we we sort of cleaned up what was generated by Cuba next to shorten the code and make it much more

102
00:08:20,030 --> 00:08:26,690
understandable or digestible by Cuban mix generated include in the steps of Flight Number One Number

103
00:08:26,690 --> 00:08:28,030
Two Number Three over here.

104
00:08:28,100 --> 00:08:32,710
All of this generated by Cuba makes and it even tells us what is happening here.

105
00:08:32,720 --> 00:08:39,710
No one specified a data structure to provide the reference of the activation slush working memory chunk

106
00:08:39,770 --> 00:08:42,450
number to create an instance of the API.

107
00:08:42,650 --> 00:08:49,450
Number three initialize the end end ready to be used in all of this was generated by cubit mix right.

108
00:08:49,700 --> 00:08:52,470
We only included the print error here.

109
00:08:52,490 --> 00:08:57,980
This also was generated by Cuban MCs and a lot more for simulating random stuff.

110
00:08:57,980 --> 00:08:59,780
We just cleaned that and kept this.

111
00:09:00,710 --> 00:09:03,700
So really we've not changed much.

112
00:09:04,130 --> 00:09:06,940
Okay so this looks fine now.

113
00:09:06,980 --> 00:09:12,680
I'm going to go back to the main function and add to some final touches to it as well.

114
00:09:13,100 --> 00:09:15,570
So let's see what we have so far.

115
00:09:16,250 --> 00:09:17,010
Okay.

116
00:09:17,100 --> 00:09:18,320
It looks good so far.

117
00:09:18,350 --> 00:09:24,800
We come over here we said if we press the brute the Blue Button user interface resets and then reset

118
00:09:24,800 --> 00:09:25,700
the game as well.

119
00:09:26,460 --> 00:09:30,640
And then after the whole delay one would come over here.

120
00:09:30,770 --> 00:09:32,750
Okay that's a typo over here.

121
00:09:32,750 --> 00:09:34,700
This one here the label should be index.

122
00:09:34,700 --> 00:09:38,060
I like this one.

123
00:09:38,930 --> 00:09:39,490
Okay.

124
00:09:39,500 --> 00:09:45,480
And we have to put an else statement here over here.

125
00:09:45,740 --> 00:09:52,500
I'm gonna see else if on a score second prediction

126
00:09:56,700 --> 00:10:01,200
that probe is less than an underscore.

127
00:10:03,420 --> 00:10:06,360
Output on a score data index.

128
00:10:06,360 --> 00:10:18,950
I if that's the case then what we're going to do is so can prediction label because I.

129
00:10:20,340 --> 00:10:25,220
And then second prediction probability

130
00:10:27,680 --> 00:10:37,580
proud of I here and probe you cause an underscore output on that score data index.

131
00:10:37,590 --> 00:10:43,200
I like this okay right.

132
00:10:48,480 --> 00:10:51,490
So what are we going to do us.

133
00:10:51,640 --> 00:10:53,470
Why is the prediction is done.

134
00:10:53,490 --> 00:11:00,360
We are going to print out the predicted result on D on the liquid crystal display.

135
00:11:01,730 --> 00:11:02,100
Okay.

136
00:11:02,130 --> 00:11:03,890
Oh we've already done that over here.

137
00:11:03,900 --> 00:11:05,890
We have this line for doing that.

138
00:11:05,970 --> 00:11:10,740
We set a color to red font size 20 and then this the background color.

139
00:11:10,770 --> 00:11:13,550
And then there's the position to print this.

140
00:11:13,680 --> 00:11:19,830
And yeah we're print in first s.t. first drink.

141
00:11:19,830 --> 00:11:21,950
We already place this inside us drink.

142
00:11:21,960 --> 00:11:24,710
Okay we're not using the second string buffer.

143
00:11:25,620 --> 00:11:27,750
So I'm going to comment this one out

144
00:11:30,750 --> 00:11:31,510
okay.

145
00:11:31,560 --> 00:11:32,720
Let's see.

146
00:11:33,060 --> 00:11:35,810
And we sent in this to you as well.

147
00:11:36,000 --> 00:11:38,590
Let's see if the US is enabled.

148
00:11:38,940 --> 00:11:43,080
We may have or I may have disabled it some time ago.

149
00:11:43,080 --> 00:11:47,310
Okay I'm going to enable a use it or you.

150
00:11:47,340 --> 00:11:48,480
Over here.

151
00:11:48,660 --> 00:11:51,030
This Amex underscore.

152
00:11:51,560 --> 00:11:54,710
You set one on a score.

153
00:11:54,720 --> 00:11:58,590
You got a score in it like this.

154
00:12:01,090 --> 00:12:01,820
Okay.

155
00:12:01,930 --> 00:12:07,760
It's looking good so we can certainly build now and see what we have.

156
00:12:07,780 --> 00:12:08,920
We may have some arrows.

157
00:12:10,090 --> 00:12:11,240
Um but let's see.

158
00:12:11,260 --> 00:12:12,130
Click here to build

159
00:12:16,500 --> 00:12:17,560
it's built in.

160
00:12:20,440 --> 00:12:21,520
Okay.

161
00:12:21,520 --> 00:12:23,100
It says We have five arrows.

162
00:12:27,080 --> 00:12:28,500
K K

163
00:12:32,380 --> 00:12:33,800
let's see over here.

164
00:12:33,810 --> 00:12:38,390
It says float n n input data.

165
00:12:38,390 --> 00:12:44,610
Data conflict conflicting declaration.

166
00:12:44,660 --> 00:12:49,970
So the question is conflict and declaration with who.

167
00:12:50,010 --> 00:12:59,310
Because if we come over here to the corresponding dots each file there is no conflict in declaration

168
00:12:59,310 --> 00:13:00,040
at all.

169
00:13:00,120 --> 00:13:02,850
It's the same thing here.

170
00:13:03,180 --> 00:13:04,110
Yeah no problem.

171
00:13:04,610 --> 00:13:07,300
Let's see the second one it says in data on declared.

172
00:13:07,410 --> 00:13:08,510
Let's see.

173
00:13:08,510 --> 00:13:09,060
Okay.

174
00:13:09,310 --> 00:13:09,530
Okay.

175
00:13:09,540 --> 00:13:10,470
This is simple one.

176
00:13:10,470 --> 00:13:13,750
So over here we simply need to pass.

177
00:13:14,040 --> 00:13:18,920
I think I should make this clearer over here.

178
00:13:18,990 --> 00:13:22,450
You know what a process a I run in data.

179
00:13:22,530 --> 00:13:29,100
Data here should be the same arguments that would pass to the function so I'll copy this and then paste

180
00:13:29,340 --> 00:13:35,040
the name that we gave it and an input data and I'll copy this as well paste this over here and then

181
00:13:35,040 --> 00:13:39,030
I'll put data I see where the next error says okay.

182
00:13:39,030 --> 00:13:41,550
This output data we've dealt with that one.

183
00:13:41,940 --> 00:13:48,690
So let's view it again and see where we might end up with a conflict in a row.

184
00:13:49,030 --> 00:13:50,340
Okay let's see

185
00:14:00,720 --> 00:14:02,170
it says over here.

186
00:14:02,210 --> 00:14:06,060
Incompatible implicit declaration of print f.

187
00:14:06,420 --> 00:14:15,110
Yes of course because I have not included s d d i ordered H in this file over here.

188
00:14:15,830 --> 00:14:18,100
Um but it is included in the main.

189
00:14:18,140 --> 00:14:28,520
So what we can do is I'm gonna come over here to Maine and we have include sd d i not h over here.

190
00:14:28,550 --> 00:14:34,770
I'm going to cut this and I'm going to put it in a y up hyphen cube.

191
00:14:34,840 --> 00:14:42,890
X h each cause that this I'm gonna put it in this file here because this is called indeed dot c file

192
00:14:43,010 --> 00:14:50,540
as well as the main file or you can simply have double include if you prefer but I've cut it off cut

193
00:14:50,660 --> 00:14:58,700
sd d i ordered H from Main and I'm simply going to put it in this dot h file here up on the score X

194
00:14:58,700 --> 00:15:09,380
cube underscore a dot h I'm going to include s DDA here because this is called by both the d up the

195
00:15:09,380 --> 00:15:15,500
score X cube e I don't see and I mean that c file so once that is done I'm gonna click here to build

196
00:15:19,960 --> 00:15:20,480
okay

197
00:15:23,120 --> 00:15:25,550
include or provide declaration for print if

198
00:15:29,200 --> 00:15:31,730
I see this rather strange

199
00:15:34,490 --> 00:15:36,030
Yeah it it's rather strange.

200
00:15:36,710 --> 00:15:38,690
Okay let's see something here.

201
00:15:38,700 --> 00:15:38,930
Uh

202
00:15:45,950 --> 00:15:47,150
okay let's see

203
00:15:58,640 --> 00:16:05,960
okay so what is happening is that s a sieve this file is not reachable the up on the score X cube hyphen

204
00:16:06,050 --> 00:16:08,420
H is not reachable.

205
00:16:08,420 --> 00:16:09,080
Right.

206
00:16:09,110 --> 00:16:09,920
So let's solve that.

207
00:16:11,660 --> 00:16:17,990
Let's see it s path it s path is in the M S include forward.

208
00:16:18,770 --> 00:16:25,150
I'll see I'll go and verify that it s exist in my in my path.

209
00:16:25,280 --> 00:16:27,810
Let's take a look at the name again.

210
00:16:28,400 --> 00:16:31,430
It's an embedded amnesty include.

211
00:16:31,430 --> 00:16:39,440
Okay so I'm gonna come over here like we did earlier right click on project go to properties over here

212
00:16:39,440 --> 00:16:47,690
like this and then I'm gonna come over here and then go to paths symbols and let's see if we have this

213
00:16:47,690 --> 00:16:57,780
year we have schemes we have em list include over here we have this over here but it's not it's no find

214
00:16:57,770 --> 00:16:58,140
in it.

215
00:16:58,580 --> 00:17:03,920
So what I'm gonna do is I'm simply going to create a new file of I'm gonna create a copy of this and

216
00:17:03,920 --> 00:17:10,520
called that is that there's a workaround I find for this but if one of you know why this is behaving

217
00:17:10,520 --> 00:17:17,810
like this then certainly you can share it's in the forum but I tend to find problems with some of the

218
00:17:17,810 --> 00:17:26,990
path and file location is using this open source um cube IDC show what I'm gonna do is I'm gonna come

219
00:17:27,050 --> 00:17:37,430
over here I'm going to come to the C O selector Ford I'll just put a here or select my project right

220
00:17:37,460 --> 00:17:46,160
click new and then I'm going to create a new file and then I'll simply call this an end not h mean in

221
00:17:46,160 --> 00:17:49,540
neural network to teach and then finish.

222
00:17:49,580 --> 00:17:55,490
So this an empty file and what I'm going to do for this an endless H is I'm going to take everything

223
00:17:55,490 --> 00:18:04,280
in this up underscore X cube a dot H and paste it into my new empty file so I'll do control a two select

224
00:18:04,280 --> 00:18:10,840
to control C to copy and I'll come to my n n not h control V to paste right.

225
00:18:11,870 --> 00:18:19,460
So now this is going to be my head of file for up underscore up on this call cube e I don't see.

226
00:18:19,460 --> 00:18:22,010
So don't let this confuse you.

227
00:18:22,370 --> 00:18:30,170
So what I'm going to do is now when I come to my up on this score cube hyphen a dot see where I'm calling

228
00:18:30,290 --> 00:18:37,370
up underscore execute hyphen a dot h I'm simply going to change it to my n n dot H so someone might

229
00:18:37,370 --> 00:18:43,370
be confused if the F didn't see me do this in the video if you just help the project we don't understand

230
00:18:43,370 --> 00:18:48,920
it you might find this confusing you could give it another name write it down and end what h you could

231
00:18:48,920 --> 00:18:57,200
as well name it up on the score cube on the score e I to dot H or underscore dot H just to make it slightly

232
00:18:57,200 --> 00:19:03,290
different from the original file so now this is going to be what is included here and I'm going to go

233
00:19:03,290 --> 00:19:11,480
to my main dot c file as well and include the same and end it H this is main so over here we call it

234
00:19:11,510 --> 00:19:13,580
up on the score excuse up on the score yeah.

235
00:19:13,650 --> 00:19:14,850
What each.

236
00:19:14,990 --> 00:19:17,800
I'm gonna include and then dot h.

237
00:19:19,130 --> 00:19:19,460
Okay.

238
00:19:19,520 --> 00:19:27,380
And now what I'm gonna do is I'm going to add the path of an end of H to my project's path and symbols

239
00:19:30,550 --> 00:19:31,540
so this is.

240
00:19:31,570 --> 00:19:39,340
So this my project folder amnesty is in my embedded amnesty and an amnesty.

241
00:19:39,340 --> 00:19:47,110
I'm gonna open SDM 32 cube inside a folder and this is the new file we created and then dot H it's found

242
00:19:47,470 --> 00:19:53,790
this is the path of it this thing and then s a dot h file as you can see your here.

243
00:19:53,800 --> 00:20:00,460
So I'm gonna copy this it shows zero because we've not rebuilt our project so I've copied it location

244
00:20:00,640 --> 00:20:01,800
right.

245
00:20:01,830 --> 00:20:06,050
Gonna do control is to save whereas mine and H.

246
00:20:06,130 --> 00:20:08,670
Gonna do contra s to save.

247
00:20:08,720 --> 00:20:17,650
Okay so now I'm gonna come to my project properties and then over here under parts and symbols I'm going

248
00:20:17,650 --> 00:20:26,360
to add new one here control V to paste okay apply and then apply it in close.

249
00:20:26,380 --> 00:20:26,730
Okay.

250
00:20:26,740 --> 00:20:30,940
Once that is done I'm gonna click over here to build

251
00:20:34,290 --> 00:20:34,940
it's built in

252
00:20:42,700 --> 00:20:43,630
it's still built in

253
00:20:54,210 --> 00:20:54,940
it's built in

254
00:21:01,500 --> 00:21:03,570
and it's built successfully.

255
00:21:03,570 --> 00:21:05,840
We've got no error right.

256
00:21:06,000 --> 00:21:11,700
So let's download onto our board and see how our neural network performs.

257
00:21:11,700 --> 00:21:15,240
We're gonna click over here run then debug

258
00:21:23,090 --> 00:21:24,390
it s download in

259
00:21:36,910 --> 00:21:38,070
it's still a download in

260
00:21:41,220 --> 00:21:42,970
says download in progress

261
00:21:48,670 --> 00:21:49,350
okay.

262
00:21:49,400 --> 00:21:50,240
Very fine

263
00:21:57,060 --> 00:21:57,820
okay.

264
00:21:57,910 --> 00:21:59,480
Let's see.

265
00:21:59,620 --> 00:22:01,860
Download verified successfully.

266
00:22:01,870 --> 00:22:05,620
I'm gonna click here to stop and then reset my board and see

267
00:22:08,580 --> 00:22:10,380
so let's see.

268
00:22:10,450 --> 00:22:11,610
This is my board.

269
00:22:11,610 --> 00:22:12,630
I'm gonna run.

270
00:22:13,200 --> 00:22:21,130
I'm gonna write a number in C or right to push the button to predict its predicted a 1.

271
00:22:21,210 --> 00:22:21,560
Okay.

272
00:22:21,570 --> 00:22:22,420
I was wrong.

273
00:22:22,420 --> 00:22:23,370
I'll clean this.

274
00:22:23,370 --> 00:22:24,570
I'll try another number.

275
00:22:24,610 --> 00:22:30,310
Three price to predict predicted a one again.

276
00:22:30,410 --> 00:22:34,410
Oh you raise a right for price to predict.

277
00:22:34,690 --> 00:22:36,160
You predicted it once again.

278
00:22:36,160 --> 00:22:41,040
Obviously a neural network isn't performing good at all right.

279
00:22:41,070 --> 00:22:46,350
So what we're going to do is we're going to update our neural network.

280
00:22:46,440 --> 00:22:51,700
So this gives us the opportunity to see how we can update the neural network.

281
00:22:51,700 --> 00:22:59,960
We're going to take a bigger neural network compress it and then deploy that onto our microcontroller

282
00:23:00,510 --> 00:23:03,830
and we're going to see that it takes just simple steps.

283
00:23:03,830 --> 00:23:11,340
All we need to do is go to Kingdom X and then quantized those back into C code and then replace the

284
00:23:11,370 --> 00:23:12,690
appropriate files.

285
00:23:12,720 --> 00:23:20,160
So we're going to do that and hopefully with that neural network we would see the um the system work

286
00:23:20,220 --> 00:23:21,310
as expected.

287
00:23:21,390 --> 00:23:23,360
So we'll do it in the next lesson.