def post(self):
template_params = handle_user()
if not 'admin' in template_params:
self.redirect('/')
return
tournament_url = self.request.get("url")
# Check for input
if tournament_url == "":
self.redirect("/addTournaments")
return
dup_check = TournamentModel.query(
TournamentModel.url == convert_url(tournament_url))
if dup_check.get() is not None:
render_template(self, 'error_dup_league.html', template_params)
return
api.set_credentials(CHALLONGE_USERNAME, CHALLONGE_API)
#Handle tournament not found error
tournament = tournaments.show(tournament_url)
tournament_object = TournamentModel()
tournament_object.name = tournament['name']
tournament_object.url = convert_url(tournament_url)
timestamp = parser.parse(tournament['created-at'][:-6])
tournament_object.timestamp = timestamp
tournament_object.put()
# Extract the matches
# Move participant seach to preindex'd list rather than 3 challonge requests
match_list = matches.index(tournament['id'])
participant_list = participants.index(tournament['id'])
self.response.out.write(participant_list)
for match in match_list:
match_object = MatchModel(parent=tournament_object.key)
# Find names via challonge
#match_object.player1 = participants.show(tournament['id'], match['player1-id'])['name']
#match_object.player2 = participants.show(tournament['id'], match['player2-id'])['name']
# Find names via list
for p in participant_list:
if p['id'] == match['player1-id']:
match_object.player1 = p['name']
if p['id'] == match['player2-id']:
match_object.player2 = p['name']
if match['scores-csv'] != "":
parts = match['scores-csv'].split('-')
match_object.player1Score = int(parts[0])
match_object.player2Score = int(parts[1])
winner = participants.show(tournament['id'], match['winner-id'])
match_object.winner = winner['name']
match_object.label = match['identifier']
timestamp = parser.parse(match['started-at'][:-6])
match_object.timestamp = timestamp
match_object.put()
self.redirect("/listTournaments")
# from models.model_io import ModelInput, ModelOptions, ModelOutput
from utils.flag_parser import parse_arguments
import torch.jit as jit
# embed basemodel.py
import torch
import torch.nn as nn
import torch.nn.functional as F
from utils.net_util import norm_col_init, weights_init
from models import MatchModel
if __name__ == '__main__':
args = parse_arguments()
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
# device = 'cpu'
model = MatchModel(args).to(device)
# model_input = ModelInput()
state = torch.zeros(1, 512, 7, 7).to(device) # [1,512,7,7]
hidden = (
torch.zeros(1, args.hidden_state_sz).to(device), # [1,512]
torch.zeros(1, args.hidden_state_sz).to(device), # [1,512]
)
target_class_embedding = torch.zeros(300).to(device) # [300]
action_probs = torch.zeros(1, args.action_space).to(
device) # [1, #(ACTION_SPACE)]
# model_opts = ModelOptions()
# tw.draw_model(model,([1,512,7,7],
# [1,args.hidden_state_sz],
# [1,args.hidden_state_sz],
