Kosmas Deligkaris / FrogAI-FirstTest Public
Target: Sony Spresense (Cortex-M4F 156MHz) Clone this project
Experiments EON Tuner
The EON Tuner helps you quickly run hyper-parameter sweeps that explore different pre-processing + model architectures optimized for your defined objectives. Clone this project to use the EON Tuner.

Target

Run #3

Sony Spresense (Cortex-M4F 156MHz)

100 ms

1536 kB

8192 kB

Filters

Status

DSP type

Model type

View

Data set

Variant

Sort

General

F1-score

Precision

Recall

Run #5 (Jan 27 2026, 03:26:42, finished) Run #4 (Jul 10 2025, 01:46:25, finished) Run #3 (Jul 09 2025, 06:25:04, finished) Run #2 (Jul 09 2025, 06:20:02, finished) Run #1 (Jul 01 2025, 08:15:36, finished)
83%
spectr-conv2d-f45
PERFORMANCE
LATENCY
148 ms of 100 ms
Exceeds target by 48 ms
RAM
73 kB of 1536 kB
ROM
63 kB of 8192 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTROGRAM

0.05 | 0.025 | -52

ACCURACY (OVERALL)
CLASSIFICATION

0.005 | 100 | 83%

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

0 7/9/2025, 6:56:51 AM

82%
spectr-conv2d-df9
PERFORMANCE
LATENCY
163 ms of 100 ms
Exceeds target by 63 ms
RAM
73 kB of 1536 kB
ROM
63 kB of 8192 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
250 ms |
Enabled

SPECTROGRAM

0.05 | 0.025 | -52

ACCURACY (OVERALL)
CLASSIFICATION

0.005 | 100 | 82%

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

0 7/9/2025, 7:31:02 AM

70%
spectr-conv1d-7ea
PERFORMANCE
LATENCY
34 ms of 100 ms
RAM
29 kB of 1536 kB
ROM
44 kB of 8192 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTROGRAM

0.05 | 0.025 | -72

ACCURACY (OVERALL)
CLASSIFICATION

0.005 | 100 | 70%

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

0 7/9/2025, 6:38:38 AM

65%
spectr-conv2d-4d3
PERFORMANCE
LATENCY
122 ms of 100 ms
Exceeds target by 22 ms
RAM
73 kB of 1536 kB
ROM
63 kB of 8192 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
250 ms |
Enabled

SPECTROGRAM

0.05 | 0.025 | -32

ACCURACY (OVERALL)
CLASSIFICATION

0.005 | 100 | 65%

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

0 7/9/2025, 8:08:40 AM

49%
spectr-conv2d-366
PERFORMANCE
LATENCY
93 ms of 100 ms
RAM
47 kB of 1536 kB
ROM
41 kB of 8192 kB
DSP NN Unused
TIME-SERIES INPUT

1000 ms |
1000 ms |
Enabled

SPECTROGRAM

0.05 | 0.025 | -52

ACCURACY (OVERALL)
CLASSIFICATION

0.005 | 100 | 49%

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

0 7/9/2025, 6:57:56 AM

Help & documentation Get help on the forum Send feedback Contact sales
Join us
Terms of Service Privacy policy