Edge Impulse Experts / fallBT Public
The EON Tuner helps you find the most optimal architecture for your embedded machine-learning application. Clone this project to use the EON Tuner.

Target

No name set

Arduino Nano 33 BLE Sense (Cortex-M4F 64MHz)

100 ms

256 kB

1024 kB

Filters

Status

DSP type

Model type

View

Data set

Variant

Sort

General

F1-score

Precision

Recall

94%
raw-dense-060
PERFORMANCE
LATENCY
25 ms of 100 ms
RAM
10 kB of 256 kB
ROM
54 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

RAW

1

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 94%

Type Filters Kernel Rate
dense 40 - -
dropout - - 0.25
dense 20 - -
dropout - - 0.25

8/28/2022, 1:55:01 PM

94%
spectr-dense-ad8
PERFORMANCE
LATENCY
37 ms of 100 ms
RAM
12 kB of 256 kB
ROM
43 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTRAL-ANALYSIS

128

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 94%

Type Filters Kernel Rate
dense 40 - -
dropout - - 0.5
dense 20 - -
dropout - - 0.5

8/28/2022, 1:55:06 PM

94%
spectr-dense-e53
PERFORMANCE
LATENCY
31 ms of 100 ms
RAM
12 kB of 256 kB
ROM
30 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTRAL-ANALYSIS

128

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 94%

Type Filters Kernel Rate
dense 20 - -
dropout - - 0.25
dense 10 - -
dropout - - 0.25
dense 5 - -
dropout - - 0.25

8/28/2022, 1:59:37 PM

94%
spectr-dense-0d4
PERFORMANCE
LATENCY
175 ms of 100 ms
Exceeds target by 75 ms
RAM
36 kB of 256 kB
ROM
109 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTRAL-ANALYSIS

1024

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 94%

Type Filters Kernel Rate
dense 20 - -
dropout - - 0.25
dense 10 - -
dropout - - 0.25

8/28/2022, 2:03:56 PM

94%
spectr-dense-d24
PERFORMANCE
LATENCY
175 ms of 100 ms
Exceeds target by 75 ms
RAM
36 kB of 256 kB
ROM
109 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTRAL-ANALYSIS

1024

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 94%

Type Filters Kernel Rate
dense 20 - -
dropout - - 0.25
dense 10 - -
dropout - - 0.25

8/28/2022, 2:05:14 PM

94%
raw-dense-3b9
PERFORMANCE
LATENCY
25 ms of 100 ms
RAM
10 kB of 256 kB
ROM
54 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

RAW

1

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 94%

Type Filters Kernel Rate
dense 40 - -
dropout - - 0.25
dense 20 - -
dropout - - 0.25

8/28/2022, 2:06:33 PM

88%
spectr-dense-25e
PERFORMANCE
LATENCY
36 ms of 100 ms
RAM
12 kB of 256 kB
ROM
30 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTRAL-ANALYSIS

128

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 88%

Type Filters Kernel Rate
dense 20 - -
dropout - - 0.25
dense 10 - -
dropout - - 0.25

8/28/2022, 1:54:09 PM

88%
spectr-dense-9c5
PERFORMANCE
LATENCY
170 ms of 100 ms
Exceeds target by 70 ms
RAM
36 kB of 256 kB
ROM
109 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTRAL-ANALYSIS

1024

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 88%

Type Filters Kernel Rate
dense 20 - -
dropout - - 0.25
dense 10 - -
dropout - - 0.25
dense 5 - -
dropout - - 0.25

8/28/2022, 1:56:04 PM

88%
spectr-dense-bee
PERFORMANCE
LATENCY
109 ms of 100 ms
Exceeds target by 9 ms
RAM
36 kB of 256 kB
ROM
201 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTRAL-ANALYSIS

1024

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 88%

Type Filters Kernel Rate
dense 40 - -
dropout - - 0.5
dense 20 - -
dropout - - 0.5
dense 10 - -
dropout - - 0.5

8/28/2022, 1:57:27 PM

88%
spectr-dense-c65
PERFORMANCE
LATENCY
142 ms of 100 ms
Exceeds target by 42 ms
RAM
36 kB of 256 kB
ROM
200 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTRAL-ANALYSIS

1024

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 88%

Type Filters Kernel Rate
dense 40 - -
dropout - - 0.5
dense 20 - -
dropout - - 0.5

8/28/2022, 2:01:07 PM

88%
spectr-dense-24c
PERFORMANCE
LATENCY
142 ms of 100 ms
Exceeds target by 42 ms
RAM
36 kB of 256 kB
ROM
200 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTRAL-ANALYSIS

1024

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 88%

Type Filters Kernel Rate
dense 40 - -
dropout - - 0.5
dense 20 - -
dropout - - 0.5

8/28/2022, 2:01:54 PM

88%
spectr-dense-459
PERFORMANCE
LATENCY
109 ms of 100 ms
Exceeds target by 9 ms
RAM
36 kB of 256 kB
ROM
201 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTRAL-ANALYSIS

1024

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 88%

Type Filters Kernel Rate
dense 40 - -
dropout - - 0.5
dense 20 - -
dropout - - 0.5
dense 10 - -
dropout - - 0.5

8/28/2022, 2:03:41 PM

88%
spectr-dense-1d8
PERFORMANCE
LATENCY
36 ms of 100 ms
RAM
12 kB of 256 kB
ROM
30 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTRAL-ANALYSIS

128

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 88%

Type Filters Kernel Rate
dense 20 - -
dropout - - 0.25
dense 10 - -
dropout - - 0.25

8/28/2022, 2:04:04 PM

88%
spectr-dense-d30
PERFORMANCE
LATENCY
173 ms of 100 ms
Exceeds target by 73 ms
RAM
36 kB of 256 kB
ROM
200 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTRAL-ANALYSIS

1024

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 88%

Type Filters Kernel Rate
dense 40 - -
dropout - - 0.25
dense 20 - -
dropout - - 0.25

8/28/2022, 2:06:36 PM

82%
raw-dense-c57
PERFORMANCE
LATENCY
27 ms of 100 ms
RAM
10 kB of 256 kB
ROM
54 kB of 1024 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

RAW

1

ACCURACY (KERAS)
CLASSIFICATION (KERAS)

0.0005 | 30 | 82%

Type Filters Kernel Rate
dense 40 - -
dropout - - 0.5
dense 20 - -
dropout - - 0.5

8/28/2022, 1:53:09 PM