DouZero
/
Douzero_Resnet
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

调整评估模型加载逻辑

This commit is contained in:
zhiyang7 2021-12-27 14:52:38 +08:00
parent 29866e5dbf
commit 9bdcc85cc7
2 changed files with 15 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
douzero/evaluation/deep_agent.py
View File

@ -23,16 +23,11 @@ def _load_model(position, model_path, model_type, use_legacy, use_lite, use_onnx
else: else:
model = model_dict[position]() model = model_dict[position]()
model_state_dict = model.state_dict() model_state_dict = model.state_dict()
if torch.cuda.is_available():
pretrained = torch.load(model_path, map_location='cuda:0')
else:
pretrained = torch.load(model_path, map_location='cpu') pretrained = torch.load(model_path, map_location='cpu')
pretrained = {k: v for k, v in pretrained.items() if k in model_state_dict} pretrained = {k: v for k, v in pretrained.items() if k in model_state_dict}
model_state_dict.update(pretrained) model_state_dict.update(pretrained)
model.load_state_dict(model_state_dict) model.load_state_dict(model_state_dict)
# torch.save(model.state_dict(), model_path.replace(".ckpt", "_nobn.ckpt")) # torch.save(model.state_dict(), model_path.replace(".ckpt", "_nobn.ckpt"))
if torch.cuda.is_available():
model.cuda()
model.eval() model.eval()
model_path = model_path + '.onnx' model_path = model_path + '.onnx'
if use_onnx and not os.path.exists(model_path): if use_onnx and not os.path.exists(model_path):
@ -58,14 +53,19 @@ class DeepAgent:
self.lite_model = True if "lite" in model_path else False self.lite_model = True if "lite" in model_path else False
self.model_type = "general" if "resnet" in model_path else "old" self.model_type = "general" if "resnet" in model_path else "old"
self.model = _load_model(position, model_path, self.model_type, self.use_legacy, self.lite_model, use_onnx=use_onnx) self.model = _load_model(position, model_path, self.model_type, self.use_legacy, self.lite_model, use_onnx=use_onnx)
if use_onnx: self.onnx_model_path = os.path.abspath(model_path + '.onnx')
self.onnx_model = onnxruntime.InferenceSession(get_example(os.path.abspath(model_path + '.onnx')), providers=['CPUExecutionProvider']) self.use_onnx = use_onnx
else:
self.onnx_model = None self.onnx_model = None
self.EnvCard2RealCard = {3: '3', 4: '4', 5: '5', 6: '6', 7: '7', self.EnvCard2RealCard = {3: '3', 4: '4', 5: '5', 6: '6', 7: '7',
8: '8', 9: '9', 10: 'T', 11: 'J', 12: 'Q', 8: '8', 9: '9', 10: 'T', 11: 'J', 12: 'Q',
13: 'K', 14: 'A', 17: '2', 20: 'X', 30: 'D'} 13: 'K', 14: 'A', 17: '2', 20: 'X', 30: 'D'}
def act(self, infoset, with_confidence = False): def act(self, infoset, with_confidence = False):
if self.use_onnx and self.onnx_model is None:
if torch.cuda.is_available():
self.onnx_model = onnxruntime.InferenceSession(get_example(self.onnx_model_path), providers=['CPUExecutionProvider'])
else:
self.onnx_model = onnxruntime.InferenceSession(get_example(self.onnx_model_path), providers=['CUDAExecutionProvider'])
if not with_confidence and len(infoset.legal_actions) == 1: if not with_confidence and len(infoset.legal_actions) == 1:
return infoset.legal_actions[0] return infoset.legal_actions[0]

8
douzero/evaluation/simulation.py
View File

@ -22,8 +22,7 @@ def load_card_play_models(card_play_model_path_dict):
players[position] = DeepAgent(position, card_play_model_path_dict[position], use_onnx=True) players[position] = DeepAgent(position, card_play_model_path_dict[position], use_onnx=True)
return players return players
def mp_simulate(card_play_data_list, card_play_model_path_dict, q, output, title): def mp_simulate(card_play_data_list, players, q, output, title):
players = load_card_play_models(card_play_model_path_dict)
EnvCard2RealCard = {3: '3', 4: '4', 5: '5', 6: '6', 7: '7', EnvCard2RealCard = {3: '3', 4: '4', 5: '5', 6: '6', 7: '7',
8: '8', 9: '9', 10: 'T', 11: 'J', 12: 'Q', 8: '8', 9: '9', 10: 'T', 11: 'J', 12: 'Q',
13: 'K', 14: 'A', 17: '2', 20: 'X', 30: 'D'} 13: 'K', 14: 'A', 17: '2', 20: 'X', 30: 'D'}
@ -97,11 +96,14 @@ def evaluate(landlord, landlord_up, landlord_front, landlord_down, eval_data, nu
ctx = mp.get_context('spawn') ctx = mp.get_context('spawn')
q = ctx.SimpleQueue() q = ctx.SimpleQueue()
processes = [] processes = []
players = load_card_play_models(card_play_model_path_dict)
for card_paly_data in card_play_data_list_each_worker: for card_paly_data in card_play_data_list_each_worker:
p = ctx.Process( p = ctx.Process(
target=mp_simulate, target=mp_simulate,
args=(card_paly_data, card_play_model_path_dict, q, output, title)) args=(card_paly_data, players, q, output, title))
p.start() p.start()
processes.append(p) processes.append(p)