Sai Yamanoor / pi_day_audio Public

EON Tuner

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

Keyword spotting

Raspberry Pi RP2040 (Cortex-M0+ 133MHz)

100 ms

264 kB

2048 kB

Filters

Status

DSP type

Network type

View

Data set

Precision

Sort

General

F1-score

Precision

Recall

100%
mfe-conv1d-c4d
PERFORMANCE
LATENCY
554 ms of 100 ms
Exceeds target by 454 ms
RAM
23 kB of 264 kB
ROM
31 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFE

0.05 | 0.05 | 32

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
conv1d 8 3 -
conv1d 16 3 -
conv1d 32 3 -
dropout - - 0.5

2/23/2023, 6:14:20 AM

100%
mfe-conv1d-655
PERFORMANCE
LATENCY
650 ms of 100 ms
Exceeds target by 550 ms
RAM
24 kB of 264 kB
ROM
39 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFE

0.05 | 0.05 | 32

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
conv1d 16 3 -
conv1d 32 3 -
conv1d 64 3 -
dropout - - 0.25

2/23/2023, 6:14:26 AM

100%
mfcc-conv1d-5f7
PERFORMANCE
LATENCY
860 ms of 100 ms
Exceeds target by 760 ms
RAM
25 kB of 264 kB
ROM
38 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFCC

0.05 | 0.05 | 32

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
Data augmentation
conv1d 8 3 -
conv1d 16 3 -
conv1d 32 3 -
conv1d 64 3 -
dropout - - 0.25

2/23/2023, 6:14:33 AM

100%
mfcc-conv1d-bd8
PERFORMANCE
LATENCY
1112 ms of 100 ms
Exceeds target by 1012 ms
RAM
24 kB of 264 kB
ROM
30 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFCC

0.05 | 0.05 | 40

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
Data augmentation
conv1d 16 3 -
conv1d 32 3 -
dropout - - 0.25

2/23/2023, 6:14:37 AM

100%
mfcc-conv1d-0a8
PERFORMANCE
LATENCY
1788 ms of 100 ms
Exceeds target by 1688 ms
RAM
36 kB of 264 kB
ROM
66 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFCC

0.05 | 0.025 | 32

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
Data augmentation
conv1d 16 3 -
conv1d 32 3 -
conv1d 64 3 -
conv1d 128 3 -
dropout - - 0.25

2/23/2023, 6:15:34 AM

100%
mfcc-conv1d-d31
PERFORMANCE
LATENCY
1119 ms of 100 ms
Exceeds target by 1019 ms
RAM
27 kB of 264 kB
ROM
38 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFCC

0.05 | 0.05 | 40

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
Data augmentation
conv1d 8 3 -
conv1d 16 3 -
conv1d 32 3 -
conv1d 64 3 -
dropout - - 0.25

2/23/2023, 6:14:43 AM

100%
mfcc-conv1d-de7
PERFORMANCE
LATENCY
1752 ms of 100 ms
Exceeds target by 1652 ms
RAM
34 kB of 264 kB
ROM
66 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFCC

0.032 | 0.032 | 40

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
conv1d 16 3 -
conv1d 32 3 -
conv1d 64 3 -
conv1d 128 3 -
dropout - - 0.25

2/23/2023, 6:15:41 AM

100%
mfe-conv1d-5ef
PERFORMANCE
LATENCY
566 ms of 100 ms
Exceeds target by 466 ms
RAM
25 kB of 264 kB
ROM
39 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFE

0.05 | 0.05 | 32

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
conv1d 8 3 -
conv1d 16 3 -
conv1d 32 3 -
conv1d 64 3 -
dropout - - 0.5

2/23/2023, 6:15:50 AM

100%
mfe-conv1d-e7c
PERFORMANCE
LATENCY
548 ms of 100 ms
Exceeds target by 448 ms
RAM
21 kB of 264 kB
ROM
28 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFE

0.05 | 0.05 | 32

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
conv1d 8 3 -
conv1d 16 3 -
dropout - - 0.25

2/23/2023, 6:17:31 AM

100%
mfcc-conv1d-b62
PERFORMANCE
LATENCY
1118 ms of 100 ms
Exceeds target by 1018 ms
RAM
25 kB of 264 kB
ROM
30 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFCC

0.05 | 0.05 | 40

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
conv1d 8 3 -
conv1d 16 3 -
conv1d 32 3 -
dropout - - 0.5

2/23/2023, 6:18:17 AM

100%
mfe-conv1d-ca3
PERFORMANCE
LATENCY
634 ms of 100 ms
Exceeds target by 534 ms
RAM
24 kB of 264 kB
ROM
39 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFE

0.05 | 0.05 | 32

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
conv1d 16 3 -
conv1d 32 3 -
conv1d 64 3 -
dropout - - 0.25

2/23/2023, 6:17:02 AM

100%
mfe-conv1d-407
PERFORMANCE
LATENCY
569 ms of 100 ms
Exceeds target by 469 ms
RAM
22 kB of 264 kB
ROM
31 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFE

0.05 | 0.05 | 32

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
conv1d 16 3 -
conv1d 32 3 -
dropout - - 0.5

2/23/2023, 6:15:52 AM

100%
mfcc-conv2d-f93
PERFORMANCE
LATENCY
1742 ms of 100 ms
Exceeds target by 1642 ms
RAM
34 kB of 264 kB
ROM
31 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFCC

0.032 | 0.032 | 40

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
conv2d 8 3 -
conv2d 16 3 -
dropout - - 0.5

2/23/2023, 6:15:47 AM

100%
mfe-conv2d-a4c
PERFORMANCE
LATENCY
919 ms of 100 ms
Exceeds target by 819 ms
RAM
34 kB of 264 kB
ROM
54 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFE

0.05 | 0.05 | 32

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
conv2d 8 3 -
conv2d 16 3 -
conv2d 32 3 -
conv2d 64 3 -
dropout - - 0.25

2/23/2023, 6:20:30 AM

100%
mfe-conv1d-883
PERFORMANCE
LATENCY
1081 ms of 100 ms
Exceeds target by 981 ms
RAM
30 kB of 264 kB
ROM
31 kB of 2048 kB
DSP NN Unused
INPUT

2000 ms | 2000 ms

MFE

0.05 | 0.025 | 32

ACCURACY
CLASSIFICATION

0.005 | 100

Type Filters Kernel Rate
Data augmentation
conv1d 8 3 -
conv1d 16 3 -
conv1d 32 3 -
dropout - - 0.25

2/23/2023, 6:17:11 AM