def main():
parser = argparse.ArgumentParser()
parser.add_argument("-t", "--time", help="Time elapsed, in seconds, so far in the game",
required=True, type=int)
parser.add_argument("-s", "--spread", help="Spread, relative to visiting team",
required=True, type=float)
parser.add_argument("-f", "--favorite", help="Current point total for the favorite",
required=True, type=int)
parser.add_argument("-u", "--underdog", help="Current point total for the underdog",
required=True, type=int)
args = parser.parse_args()
time_sec = args.time
spread = args.spread
favorite_points = args.favorite
underdog_points = args.underdog
assert time_sec >= 0, "Time must be a positive integer"
assert favorite_points >= 0, "Favorite points must be a positive integer"
assert favorite_points >= 0, "Underdog points must be a positive integer"
model = inference.load_model(MODEL_PATH)
scaler = inference.load_scaler(SCALER_PATH)
pred = inference.run_inference(model, scaler, time_sec, spread,
favorite_points, underdog_points)
print(pred)
def main(image_path, checkpoint_path):
# 加载图像和模型
image = cv2.imread(image_path)
model = load_model(checkpoint_path)
client = net()
flag = 0
while True:
if flag:
word = input("input:\n ")
if not len(word):
continue
word2audio(client, word)
frames = gen_fake(model, 'audio/audio.wav', image)
# 播放语音和视频
pygame.mixer.init()
pygame.mixer.music.load('audio/audio.wav') # 加载音乐
pygame.mixer.music.play() # 播放
for im in frames:
cv2.imshow('frame', im)
cv2.waitKey(1)
time.sleep(1 / fps)
# cv2.waitKey(1000//fps)
pygame.quit()
while True:
frame = image
# 显示结果帧e
cv2.imshow('frame', frame)
if cv2.waitKey(1) == 27: # ord('q') enter:13 esc:27
flag = 1
break
def __init__(self, recOps, people):
self.running = False
self.audio = pyaudio.PyAudio()
self.stream = self.audio.open(format=recOps['Format'],
channels=recOps['Channels'],
rate=recOps['Rate'],
input=True,
frames_per_buffer=recOps['Chunk'])
self.vad = webrtcvad.Vad()
self.vad.set_mode(1)
self.recording = False
self.audioPercent = 100
self.recOps = recOps
self.check = False
self.pastFrames = []
self.lock = threading.Lock()
self.people = people
encoder.load_model("encoder/saved_models/pretrained.pt")
synthesizer = Synthesizer("synthesizer/saved_models/logs-pretrained/".joinpath("taco_pretrained"), low_mem=args.low_mem)
vocoder.load_model("vocoder/saved_models/pretrained/pretrained.pt")
def inferencer(work_queue):
running = True
data = work_queue.get()
if type(data) == bool:
running = data
elif type(data) == ClassifiedImageBundle:
data.set_progress()
while running:
model = load_model(data.get_np_array().shape)
prediction = model.predict(np.array([data.get_np_array() / 255]),
batch_size=1)
print(prediction[0])
data.set_classification(prediction[0])
work_queue.task_done()
data = work_queue.get()
if type(data) == bool:
running = data
elif type(data) == ClassifiedImageBundle:
data.set_progress()
@app.route('/')
def home_endpoint():
return render_template('home.html')
@app.route('/predict', methods=['POST'])
def get_prediction():
image = request.files['file']
if image.filename != '':
fn = os.path.join(app.config['UPLOAD_FOLDER'],
image.filename + str(datetime.now().time()))
image.save(fn)
image = load_image(fn)
res, preprocessed_image = inference(image)
preprocessed_image = Image.fromarray(np.uint8(preprocessed_image *
255)).convert('RGB')
buffer = BytesIO()
preprocessed_image.save(buffer, format="PNG")
myimage = buffer.getvalue()
return jsonify(message=res, image=str(base64.b64encode(myimage))[2:-1])
if __name__ == '__main__':
model = None
load_model('model')
app.run(host="0.0.0.0", port=int(os.environ.get("PORT", 5000)))
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'
import tensorflow as tf
from tensorflow import keras
from tensorflow_core.python.keras.api._v2.keras import layers
from inference import load_model, inference
from PathfindingDll import load_PathfindingDLL
import MCTS
estimate = load_PathfindingDLL()
model = load_model()
model.fit(test_input, test_target)
model.save('path/to/location')
from flask import Flask, flash, request, redirect, url_for, render_template, jsonify
from werkzeug.utils import secure_filename
from inference import load_model, get_result
ALLOWED_EXTENSIONS = set(['png', 'jpg', 'jpeg'])
UPLOAD_FOLDER = 'static/uploads/'
app = Flask(__name__)
app.secret_key = "secret key"
app.config['UPLOAD_FOLDER'] = UPLOAD_FOLDER
filename = ''
netMain, metaMain = load_model()
def allowed_file(filename):
return '.' in filename and filename.rsplit(
'.', 1)[1].lower() in ALLOWED_EXTENSIONS
@app.route('/')
def upload_form():
return render_template('upload.html')
@app.route('/', methods=['POST'])
def upload_image():
if 'file' not in request.files:
# Libs
import torch
import numpy as np
from sklearn.linear_model import LogisticRegression
# Own modules
from infer_sentence import make_ds
from dudu_utils import preprocess, process_block
from inference import load_model, infer_dataset
# Settings
# model_path = r'./models/cnw-2020-04-29_20-02-45.pth.tar'
model_path = r'./models/sswer-2020-04-29_15-47-38.pth.tar'
checkpoint = torch.load(model_path, map_location=lambda storage, loc: storage)
lut = checkpoint['lut']
model = load_model(model_path, len(lut))
window_size = 3
step_size = 3
device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')
BUFFER_DIR = './buffer'
def train_clf(network, dev):
dataset, vocab_size = make_ds()
ftrs, lbls = infer_dataset(network, dataset, dev)
clf = LogisticRegression(max_iter=1000)
clf.fit(ftrs, lbls)
return clf
def pred_sentence():
def load_model_cached(model_file):
model = load_model(model_file)
model.eval()
return model
from MCTS import Node
from PathfindingDll import load_PathfindingDLL
from GameControlerDLL import load_GameControlerDLL
from Game import get_random_start
import numpy as np
from inference import get_move_to_index, load_model
import tensorflow as tf
model = load_model('test')
inputs = np.ones(shape=(1, 9, 9, 4), dtype=np.float32)
policy = model(inputs, training=True)
def save_pt_model(path: str) -> None:
"""
path: path to which .pt file will be exported
"""
model = load_model()
torch.save(model.state_dict(), path)
import cv2
from torchvision import models
import torchvision.transforms as transforms
from PIL import Image
from flask import Flask, jsonify, request
import base64
import numpy as np
import sys
sys.path.append('../ML')
from inference import prepare_image, load_model, inference
class_names = ["KIA Rio", "SKODA OCTAVIA", "Hyundai SOLARIS", "Volkswagen Polo", "Volkswagen Tiguan"]
mpath = '../eval/model.pth'
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
model = load_model(mpath, device)
app = Flask(__name__)
@app.route('/predict', methods=['POST'])
def predict():
if request.method == 'POST':
bm = request.form['content']
bb = bm.encode('ascii')
m = base64.b64decode(bb)
nparr = np.fromstring(m, np.uint8)
image = cv2.imdecode(nparr, cv2.IMREAD_COLOR) / 255.
data = prepare_image(image)['image'].to(device).type(torch.float)
res = inference(model, data)[0]
res_dict = {k: v for k, v in zip(class_names, res)}
return jsonify({'default': res_dict})
app.logger.setLevel(logging.DEBUG)
logHandler = handlers.RotatingFileHandler('logs/app.log',
maxBytes=1000000,
backupCount=5)
formatter = logging.Formatter(
'%(asctime)s - %(name)s - %(levelname)s - %(message)s')
logHandler.setFormatter(formatter)
app.logger.addHandler(logHandler)
MODEL_PATH = 'model/model.h5'
SCALER_PATH = 'model/scaler.pkl'
# load the model the model and scaler globally on startup
global graph # Allow the graph to be accessed within the app context
graph = tf.get_default_graph()
model = inference.load_model(MODEL_PATH)
scaler = inference.load_scaler(SCALER_PATH)
@app.route("/games/<year>/<phase>/<week>")
def get_game_data_by_week(year, phase, week):
app.logger.debug("Starting request for game data for {}:{}{}".format(
year, phase, week))
if phase.upper() not in {"PRE", "REG", "POST"}:
raise BadRequest('Phase must be one of PRE, REG, POST')
try:
game_data = nfl_api.get_game_data(year, phase, week)
except Exception as e:
app.logger.error(e)
raise HTTPException("Error collecting game data")
'home.html')
@app.route('/predict', methods=['POST'])
def get_prediction():
image = request.files['file']
if image.filename != '':
fn = os.path.join(
app.config['UPLOAD_FOLDER'], image.filename +
str(datetime.now().time())
)
image.save(fn)
image = load_image(fn)
os.remove(fn)
res_image, res = run_on_image(image)
res_image = Image.fromarray(np.uint8(res_image)).convert('RGB')
image_height_over_width = res_image.size[1] / res_image.size[0]
#res_image = res_image.resize((260, int(image_height_over_width * 260)))
buffer = BytesIO()
res_image.save(buffer, format="PNG")
return_image = buffer.getvalue()
return jsonify(message=res, image=str(base64.b64encode(return_image))[2:-1])
if __name__ == '__main__':
model = None
load_model('mask_detector/models/model360.pth')
app.run(host="0.0.0.0", port=int(os.environ.get("PORT", 5000)))