mlapiconfig.ini

[general]
# This is an optional file
# If specified, you can specify tokens with secret values in that file
# and onlt refer to the tokens in your main config file

secrets=/mlapi/secrets.ini
#secrets=./secrets.mine

# port that mlapi will listen on. Default 5000
port=5000

# Maximum # of processes that will be forked
# to handle requests. Note that each process will
# have its own copy of the model, so memory can 
# build up very quickly
# This number also dictates how many requests will be executed in parallel
# The rest will be queued

# For now, keep this to 1 if you are on a GPU
processes=1

# the secret key that will be used to sign
# JWT tokens. Make sure you change the value
# in your secrets.ini
mlapi_secret_key=!MLAPI_SECRET_KEY

# folder where images will be uploaded
# default ./images
images_path=/config/images

# folder where the user DB will be stored
db_path=/mlapi/db

# If specified, will limit detected object size to this amount of
# the total image size passed. Can help avoiding weird detections
# You can specify as % or px. Default is px
# Remember the image is resized to 416x416 internally. better
# to keep in %
max_detection_size=100%

# You can now limit the # of detection process
# per target processor. If not specified, default is 1
# Other detection processes will wait to acquire lock

cpu_max_processes=3
tpu_max_processes=1
gpu_max_processes=1

# NEW: Time to wait in seconds per processor to be free, before
# erroring out. Default is 120 (2 mins)
cpu_max_lock_wait=120
tpu_max_lock_wait=120
gpu_max_lock_wait=120


[object]

# for Yolov3
object_framework=opencv
object_processor=cpu
object_config=/config/models/yolov3/yolov3.cfg
object_weights=/config/models/yolov3/yolov3.weights
object_labels=/config/models/yolov3/coco.names

# for Tiny Yolov3
#object_framework=opencv
#object_processor=cpu
#object_config=./models/tinyyolov3/yolov3-tiny.cfg
#object_weights=./models/tinyyolov3/yolov3-tiny.weights
#object_labels=./models/tinyyolov3/coco.names

# for Yolov4
#object_framework=opencv
#object_processor=cpu
#object_config=./models/yolov4/yolov4.cfg
#object_weights=./models/yolov4/yolov4.weights
#object_labels=./models/yolov4/coco.names

# for Tiny Yolov4
#object_framework=opencv
#object_processor=cpu
#object_config=./models/tinyyolov4/yolov4-tiny.cfg
#object_weights=./models/tinyyolov4/yolov4-tiny.weights
#object_labels=./models/tinyyolov4/coco.names

# for Google Coral Edge TPU
#object_framework=coral_edgetpu
#object_processor=tpu
#object_weights=/config/models/coral_edgetpu/ssd_mobilenet_v2_coco_quant_postprocess_edgetpu.tflite
#object_labels=/config/models/coral_edgetpu/coco_indexed.names


[face]
face_detection_framework=dlib
face_recognition_framework=dlib
face_num_jitters=0
face_upsample_times=1
face_model=cnn
face_train_model=hog
face_recog_dist_threshold=0.6
face_recog_knn_algo=ball_tree

known_images_path=/config/known_faces
unknown_images_path=/config/unknown_faces

unknown_face_name=unknown face
save_unknown_faces=yes
save_unknown_faces_leeway_pixels=50

[alpr]

alpr_use_after_detection_only=yes
alpr_api_type=cloud

# -----| If you are using plate recognizer | ------
alpr_service=plate_recognizer
alpr_key=!PLATEREC_ALPR_KEY
platerec_stats=yes
#platerec_regions=['us','cn','kr']
platerec_min_dscore=0.1
platerec_min_score=0.2

# ----| If you are using openALPR |-----
#alpr_service=open_alpr
#alpr_key=!OPENALPR_ALPR_KEY
#openalpr_recognize_vehicle=1
#openalpr_country=us
#openalpr_state=ca
# openalpr returns percents, but we convert to between 0 and 1
#openalpr_min_confidence=0.3

# ----| If you are using openALPR command line |-----
openalpr_cmdline_binary=alpr
openalpr_cmdline_params=-j -d
openalpr_cmdline_min_confidence=0.3