def upload_image_file():
print("root testing", request.files)
print(">>>>>")
print(user_name)
if request.method == "POST":
# check wether the request value is
print("upload file", request.files)
if "file" in request.files:
# get the multiDict
file = request.files.getlist("file")[0]
print("upload file", request.files)
# secure the filename which will only give file name excluding other parameters
filename = secure_filename(file.filename)
#print(filename , " " , dir(file.stream))
# get the file path
path = os.path.join(UPLOAD_FOLDER, filename)
file.save(path)
infer = Inference(user_name)
result = infer(path)
return render_template("result.html",
name=user_name,
result=result,
image_path=os.path.join(
SUB_UPLOAD_FOLDER, filename))
else:
return render_template("result.html")
else:
return render_template("result.html")
import sys, os, shutil
sys.path.append(os.getcwd())
from src.inference import Inference
from flask import Flask, request, flash, redirect, render_template, url_for, Markup, send_file, send_from_directory
from werkzeug.utils import secure_filename
infer = Inference()
app = Flask(__name__)
app.secret_key = "!@#$%^&*()a-=afs;'';312$%^&*k-[;.sda,./][p;/'=-0989#$%^&0976678v$%^&*(fdsd21234266OJ^&UOKN4odsbd#$%^&*(sadg7(*&^%32b342gd']"
# the upload path for all the files
UPLOAD_FOLDER = "static/uploadFolder"
# a list to track all the files loaded in memory
app.config['UPLOAD_FOLDER'] = UPLOAD_FOLDER
list_of_uploaded_file = []
def clean_upload_folder():
try:
shutil.rmtree(UPLOAD_FOLDER)
except FileNotFoundError as e:
pass
def make_directory():
os.makedirs(UPLOAD_FOLDER, exist_ok=True)
@app.route('/')
def root():
# clean the upload directory every time user use the website and create a new empty directory
def infer():
# load infer data, need to fix
#TODO
loadFile = True
ifLoad, data = False, None
loaddict, dicts = load_file(cfg.dict_path, 'dict', 'pickle')
if not loaddict:
raise (ValueError, 'dict load failed')
if loadFile:
ifLoad, data = load_file(cfg.processed_path, 'processed data',
'pickle')
if not ifLoad or not loadFile:
train_data_obj = Dataset(cfg.train_data_path,
'train',
dicts=dicts,
language_type='es',
unlabeled_file_path=cfg.unlabeled_data_path,
emb_file_path=cfg.emb_es_path)
dev_data_obj = Dataset(cfg.dev_data_path, 'dev', dicts=dicts)
test_data_obj = Dataset(cfg.test_data_path, 'test', dicts=dicts)
save_file(
{
'train_data_obj': train_data_obj,
'dev_data_obj': dev_data_obj,
'test_data_obj': test_data_obj
}, cfg.processed_path)
# train_data_obj.save_dict(cfg.dict_path)
else:
train_data_obj = data['train_data_obj']
dev_data_obj = data['dev_data_obj']
test_data_obj = data['test_data_obj']
infer_data_obj = Dataset(cfg.infer_data_path, 'infer', dicts=dicts)
# load model
emb_mat_token = train_data_obj.emb_mat_token # need to restore model
with tf.variable_scope(network_type) as scope:
if network_type in model_type_set:
model = Model(emb_mat_token, len(train_data_obj.dicts['es']), 100,
scope.name)
graphHandler = GraphHandler(model)
#evaluator = Evaluator(model)
inference = Inference(model)
if cfg.gpu_mem < 1:
gpu_options = tf.GPUOptions(
per_process_gpu_memory_fraction=cfg.gpu_mem, allow_growth=True)
else:
gpu_options = tf.GPUOptions()
graph_config = tf.ConfigProto(gpu_options=gpu_options,
allow_soft_placement=True)
sess = tf.Session(config=graph_config)
graphHandler.initialize(sess)
saver = tf.train.Saver()
step = cfg.load_step
model_path = os.path.join(cfg.ckpt_dir,
'top_result_saver_step_%d.ckpt' % step)
saver.restore(sess, model_path)
logits_array, prob_array = inference.get_inference(sess, infer_data_obj)
inference.save_inference(prob_array, cfg.infer_result_path)
random.seed(seed)
np.random.seed(seed)
torch.manual_seed(seed)
if config['CUDA']:
torch.backends.cudnn.deterministic = True
torch.backends.cudnn.benchmark = False
def handle_model(engine, model_path="/data/egg/lego.egg"):
model = engine.load_model(path=model_path)
meshes = parse_model_geometry(model)
mesh = meshes[0]
renderable_obj = RenderableMesh(mesh)
engine.clear_renderables()
engine.add_renderable(renderable_obj)
engine.start_rendering_loop()
config = load_config()
set_seed(config)
if config['RENDERING']['RENDERING_ENGINE_ON']:
engine = RenderEngine(rendering_config=config['RENDERING'])
else:
engine = None
inferer = Inference(config=config, engine=engine)
inferer.infer()
import sys, os, shutil
sys.path.append(os.getcwd())
from src.inference import Inference
from flask import Flask, request, flash, redirect, render_template, url_for, Markup, send_file, send_from_directory
from werkzeug.utils import secure_filename
infer = None
args = list(sys.argv)
if len(args) > 1:
infer = Inference(args[1])
else:
infer = Inference('densenet')
app = Flask(__name__)
app.secret_key = "!@#$%^&*()a-=afs;'';312$%^&*k-[;.sda,./][p;/'=-0989#$%^&0976678v$%^&*(fdsd21234266OJ^&UOKN4odsbd#$%^&*(sadg7(*&^%32b342gd']"
# the upload path for all the files
SUB_UPLOAD_FOLDER = "static/uploadFolder"
UPLOAD_FOLDER = "flask_api/" + SUB_UPLOAD_FOLDER
# a list to track all the files loaded in memory
app.config['UPLOAD_FOLDER'] = UPLOAD_FOLDER
list_of_uploaded_file = []
def clean_upload_folder():
try:
shutil.rmtree(UPLOAD_FOLDER + "/")
except FileNotFoundError as e:
pass