Dual-Edge-TPU-Adapter
Dual-Edge-TPU-Adapter copied to clipboard
Bring up log
Linux x86_64: nothing with "lspci | grep 1ac1". VM might be an issue. Have to get a new drive for Linux or one of SBCs mentioned here google-coral/edgetpu/issues/256
Raspberry Pi:
pi@raspberrypi:~ $ lspci | grep 1ac1
03:00.0 System peripheral: Device 1ac1:089a
04:00.0 System peripheral: Device 1ac1:089a
AE key
Linux x86_64:
lspci | grep Coral
04:00.0 System peripheral: Global Unichip Corp. Coral Edge TPU
05:00.0 System peripheral: Global Unichip Corp. Coral Edge TPU
Single core
python3 examples/classify_image.py \
> --model test_data/mobilenet_v2_1.0_224_inat_bird_quant_edgetpu.tflite \
> --labels test_data/inat_bird_labels.txt \
> --input test_data/parrot.jpg
----INFERENCE TIME----
Note: The first inference on Edge TPU is slow because it includes loading the model into Edge TPU memory.
11.3ms
2.4ms
2.4ms
2.4ms
2.4ms
-------RESULTS--------
Ara macao (Scarlet Macaw): 0.75781
Pipeline:
python3 examples/model_pipelining_classify_image.py \
> --models \
> test_data/pipeline/inception_v3_299_quant_segment_%d_of_2_edgetpu.tflite \
> --labels test_data/imagenet_labels.txt \
> --input test_data/parrot.jpg
Using devices: ['pci:0', 'pci:1']
Using models: ['test_data/pipeline/inception_v3_299_quant_segment_0_of_2_edgetpu.tflite', 'test_data/pipeline/inception_v3_299_quant_segment_1_of_2_edgetpu.tflite']
WARNING: Logging before InitGoogleLogging() is written to STDERR
I20210628 23:19:34.290771 4441 pipelined_model_runner.cc:139] Thread: 140158135932672 receives empty request
I20210628 23:19:34.290817 4441 pipelined_model_runner.cc:198] Thread: 140158135932672 is shutting down the pipeline...
I20210628 23:19:34.472715 4441 pipelined_model_runner.cc:208] Thread: 140158135932672 Pipeline is off.
-------RESULTS--------
macaw: 0.99609
Average inference time (over 5 iterations): 36.6ms
I20210628 23:19:34.473438 4404 pipelined_model_runner.cc:139] Thread: 140158467786560 receives empty request
I20210628 23:19:34.473474 4404 pipelined_model_runner.cc:193] Thread: 140158467786560 Pipeline was turned off before.
I20210628 23:19:34.473516 4404 pipelined_model_runner.cc:193] Thread: 140158467786560 Pipeline was turned off before.
BM key adapter with m.2 M key NGFF SSD to desktop PCIe adapter
lspci | grep Coral
05:00.0 System peripheral: Global Unichip Corp. Coral Edge TPU
06:00.0 System peripheral: Global Unichip Corp. Coral Edge TPU
Single core
python3 examples/classify_image.py \
> --model test_data/mobilenet_v2_1.0_224_inat_bird_quant_edgetpu.tflite \
> --labels test_data/inat_bird_labels.txt \
> --input test_data/parrot.jpg
----INFERENCE TIME----
Note: The first inference on Edge TPU is slow because it includes loading the model into Edge TPU memory.
11.4ms
2.3ms
2.3ms
2.3ms
2.3ms
-------RESULTS--------
Ara macao (Scarlet Macaw): 0.75781
Pipeline
python3 examples/model_pipelining_classify_image.py \
> --models \
> test_data/pipeline/inception_v3_299_quant_segment_%d_of_2_edgetpu.tflite \
> --labels test_data/imagenet_labels.txt \
> --input test_data/parrot.jpg
Using devices: ['pci:0', 'pci:1']
Using models: ['test_data/pipeline/inception_v3_299_quant_segment_0_of_2_edgetpu.tflite', 'test_data/pipeline/inception_v3_299_quant_segment_1_of_2_edgetpu.tflite']
WARNING: Logging before InitGoogleLogging() is written to STDERR
I20210702 02:12:31.953905 2241 pipelined_model_runner.cc:139] Thread: 140566088103680 receives empty request
I20210702 02:12:31.953933 2241 pipelined_model_runner.cc:198] Thread: 140566088103680 is shutting down the pipeline...
I20210702 02:12:32.135826 2241 pipelined_model_runner.cc:208] Thread: 140566088103680 Pipeline is off.
-------RESULTS--------
macaw: 0.99609
Average inference time (over 5 iterations): 36.7ms
I20210702 02:12:32.136489 2204 pipelined_model_runner.cc:139] Thread: 140566419957568 receives empty request
I20210702 02:12:32.136523 2204 pipelined_model_runner.cc:193] Thread: 140566419957568 Pipeline was turned off before.
I20210702 02:12:32.136554 2204 pipelined_model_runner.cc:193] Thread: 140566419957568 Pipeline was turned off before.