validating single image predict for fashion_mnist - kamangir/bolt#692

abcli/image_classifier.sh

@@ -68,12 +68,26 @@ function abcli_image_classifier_predict() {
     local data_source=$(abcli_option "$options" "data" object)
     local model_source=$(abcli_option "$options" "model" saved)
 
-    local model_path=$(abcli_huggingface get_model_path image-classifier "$model_name" "$options")
+    if [ "$(abcli_keyword_is $data_object validate)" == true ] ; then
+        if [ "$data_source" == "object" ] ; then
+            abcli_log_error "-imge-classifier: predict: validation object not found."
+            return
+        fi
+
+        if [ "$data_source" == "url" ] ; then
+            local data_object="https://upload.wikimedia.org/wikipedia/commons/4/45/Red_High_Heel_Pumps.jpg"
+        else
+            local data_object="a-validation-filename"
+        fi
+    fi
 
     if [ "$data_source" == "object" ] ; then
         abcli_download object $data_object
     fi
 
+
+    local model_path=$(abcli_huggingface get_model_path image-classifier "$model_name" "$options")
+
     if [ "$model_source" == "object" ] ; then
         local model_object=TBD 
         abcli_download object $model_object
@@ -81,7 +95,7 @@ function abcli_image_classifier_predict() {
 
     abcli_log "image_classifier($model_path).predict($data_object): $options"
 
-    if [ ! -f "$abcli_object_root/$data_object/test_images.pyndarray" ] ; then
+    if [ ! -f "$abcli_object_root/$data_object/test_images.pyndarray" ] && [ "$data_source" == "object" ] ; then
         python3 -m image_classifier \
             preprocess \
             --infer_annotation 0 \
@@ -92,17 +106,34 @@ function abcli_image_classifier_predict() {
             ${@:4}
     fi
 
-    cp -v $abcli_object_root/$data_object/*.pyndarray .
-    cp -v $model_path/image_classifier/model/class_names.json .
+    return
 
-    python3 -m image_classifier \
-        predict \
-        --data_path $abcli_object_root/$data_object \
-        --model_path $model_path \
-        --output_path $abcli_object_path \
-        ${@:4}
+    if [ "$data_source" == "object" ] ; then
+        cp -v $abcli_object_root/$data_object/*.pyndarray .
+        cp -v $model_path/image_classifier/model/class_names.json .
+
+        python3 -m image_classifier \
+            predict \
+            --data_path $abcli_object_root/$data_object \
+            --model_path $model_path \
+            --output_path $abcli_object_path \
+            ${@:4}
 
-    abcli_tag set . image_classifier,predict
+        abcli_tag set . image_classifier,predict
+    else
+        local is_url=0
+        if [ "$data_source" == "url" ] ; then
+            local is_url=1
+        fi
+
+        python3 -m image_classifier \
+            predict_image \
+            --data_path $data_object \
+            --is_url $is_url \
+            --model_path $model_path \
+            --output_path $abcli_object_path \
+            ${@:4}  
+    fi
 }
 
 function abcli_image_classifier_train() {

image_classifier/__init__.py

@@ -1,5 +1,5 @@
 name = "image_classifier"
 
-version = "1.1.144"
+version = "1.1.145"
 
 description = "fashion-mnist + hugging-face + awesome-bash-cli"

image_classifier/__main__.py

@@ -14,7 +14,7 @@ parser.add_argument(
     "task",
     type=str,
     default="",
-    help="describe,eval,ingest,predict,preprocess,train",
+    help="describe,eval,ingest,predict,predict_image,preprocess,train",
 )
 parser.add_argument(
     "--objects",
@@ -60,6 +60,12 @@ parser.add_argument(
     type=str,
     default="",
 )
+parser.add_argument(
+    "--is_url",
+    type=int,
+    default=0,
+    help="0/1",
+)
 parser.add_argument(
     "--model_path",
     type=str,
@@ -111,6 +117,29 @@ elif args.task == "predict":
             test_labels,
             args.output_path,
         )
+elif args.task == "predict_image":
+    success = True
+
+    classifier = Image_Classifier()
+
+    success = classifier.load(args.model_path)
+
+    if success:
+        if args.is_url:
+            image_filename = file.auxiliary("image", file.extension(args.data_path))
+            if not file.download(args.data_path, image_filename):
+                success = False
+        else:
+            image_filename = args.data_path
+
+    if success:
+        success, image = file.load_image(image_filename)
+
+    if success:
+        success = classifier.predict(
+            image / 255.0,
+            output_path=args.output_path,
+        )
 elif args.task == "preprocess":
     success = preprocess(
         args.output_path,