Reformulate Q-Transformer
This commit is contained in:
D-X-Y
parent
53e1441c8d
commit
c0481a2357
@ -4,7 +4,7 @@
|
||||
# Refer to:
|
||||
# - https://github.com/microsoft/qlib/blob/main/examples/workflow_by_code.ipynb
|
||||
# - https://github.com/microsoft/qlib/blob/main/examples/workflow_by_code.py
|
||||
# python exps/trading/workflow_tt.py --market all --gpu 1
|
||||
# python exps/trading/workflow_tt.py --gpu 1 --market csi300
|
||||
#####################################################
|
||||
import sys, argparse
|
||||
from pathlib import Path
|
||||
@ -63,7 +63,8 @@ def main(xargs):
|
||||
"class": "QuantTransformer",
|
||||
"module_path": "trade_models",
|
||||
"kwargs": {
|
||||
"loss": "mse",
|
||||
"net_config": None,
|
||||
"opt_config": None,
|
||||
"GPU": "0",
|
||||
"metric": "loss",
|
||||
},
|
||||
@ -107,20 +108,23 @@ def main(xargs):
|
||||
provider_uri = "~/.qlib/qlib_data/cn_data"
|
||||
qlib.init(provider_uri=provider_uri, region=REG_CN)
|
||||
|
||||
save_dir = "{:}-{:}".format(xargs.save_dir, xargs.market)
|
||||
dataset = init_instance_by_config(dataset_config)
|
||||
for irun in range(xargs.times):
|
||||
xmodel_config = model_config.copy()
|
||||
xmodel_config = update_gpu(xmodel_config, xags.gpu)
|
||||
task = dict(model=xmodel_config, dataset=dataset_config, record=record_config)
|
||||
run_exp(task_config, dataset, "Transformer", "recorder-{:02d}-{:02d}".format(irun, xargs.times), xargs.save_dir)
|
||||
xmodel_config = update_gpu(xmodel_config, xargs.gpu)
|
||||
task_config = dict(model=xmodel_config, dataset=dataset_config, record=record_config)
|
||||
|
||||
run_exp(task_config, dataset, xargs.name, "recorder-{:02d}-{:02d}".format(irun, xargs.times), save_dir)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
parser = argparse.ArgumentParser("Vanilla Transformable Transformer")
|
||||
parser.add_argument("--save_dir", type=str, default="./outputs/tt-ml-runs", help="The checkpoint directory.")
|
||||
parser.add_argument("--save_dir", type=str, default="./outputs/vtt-runs", help="The checkpoint directory.")
|
||||
parser.add_argument("--name", type=str, default="Transformer", help="The experiment name.")
|
||||
parser.add_argument("--times", type=int, default=10, help="The repeated run times.")
|
||||
parser.add_argument("--gpu", type=int, default=0, help="The GPU ID used for train / test.")
|
||||
parser.add_argument("--market", type=str, default="csi300", help="The market indicator.")
|
||||
parser.add_argument("--market", type=str, default="all", help="The market indicator.")
|
||||
args = parser.parse_args()
|
||||
|
||||
main(args)
|
||||
|
||||
@ -14,10 +14,10 @@ from typing import Optional
|
||||
import logging
|
||||
|
||||
from qlib.utils import (
|
||||
unpack_archive_with_buffer,
|
||||
save_multiple_parts_file,
|
||||
create_save_path,
|
||||
drop_nan_by_y_index,
|
||||
unpack_archive_with_buffer,
|
||||
save_multiple_parts_file,
|
||||
create_save_path,
|
||||
drop_nan_by_y_index,
|
||||
)
|
||||
from qlib.log import get_module_logger, TimeInspector
|
||||
|
||||
@ -25,6 +25,7 @@ import torch
|
||||
import torch.nn as nn
|
||||
import torch.nn.functional as F
|
||||
import torch.optim as optim
|
||||
import torch.utils.data as th_data
|
||||
|
||||
import layers as xlayers
|
||||
from utils import count_parameters
|
||||
@ -34,409 +35,399 @@ from qlib.data.dataset import DatasetH
|
||||
from qlib.data.dataset.handler import DataHandlerLP
|
||||
|
||||
|
||||
default_net_config = dict(d_feat=6, hidden_size=48, depth=5, pos_drop=0.1)
|
||||
|
||||
default_opt_config = dict(epochs=200, lr=0.001, batch_size=2000, early_stop=20, loss="mse", optimizer="adam")
|
||||
|
||||
|
||||
class QuantTransformer(Model):
|
||||
"""Transformer-based Quant Model
|
||||
"""Transformer-based Quant Model"""
|
||||
|
||||
"""
|
||||
def __init__(self, net_config=None, opt_config=None, metric="", GPU=0, seed=None, **kwargs):
|
||||
# Set logger.
|
||||
self.logger = get_module_logger("QuantTransformer")
|
||||
self.logger.info("QuantTransformer pytorch version...")
|
||||
|
||||
def __init__(
|
||||
self,
|
||||
d_feat=6,
|
||||
hidden_size=48,
|
||||
depth=5,
|
||||
pos_dropout=0.1,
|
||||
n_epochs=200,
|
||||
lr=0.001,
|
||||
metric="",
|
||||
batch_size=2000,
|
||||
early_stop=20,
|
||||
loss="mse",
|
||||
optimizer="adam",
|
||||
GPU=0,
|
||||
seed=None,
|
||||
**kwargs
|
||||
):
|
||||
# Set logger.
|
||||
self.logger = get_module_logger("QuantTransformer")
|
||||
self.logger.info("QuantTransformer pytorch version...")
|
||||
# set hyper-parameters.
|
||||
self.net_config = net_config or default_net_config
|
||||
self.opt_config = opt_config or default_opt_config
|
||||
self.metric = metric
|
||||
self.device = torch.device("cuda:{:}".format(GPU) if torch.cuda.is_available() and GPU >= 0 else "cpu")
|
||||
self.seed = seed
|
||||
|
||||
# set hyper-parameters.
|
||||
self.d_feat = d_feat
|
||||
self.hidden_size = hidden_size
|
||||
self.depth = depth
|
||||
self.pos_dropout = pos_dropout
|
||||
self.n_epochs = n_epochs
|
||||
self.lr = lr
|
||||
self.metric = metric
|
||||
self.batch_size = batch_size
|
||||
self.early_stop = early_stop
|
||||
self.optimizer = optimizer.lower()
|
||||
self.loss = loss
|
||||
self.device = torch.device("cuda:{:}".format(GPU) if torch.cuda.is_available() else "cpu")
|
||||
self.use_gpu = torch.cuda.is_available()
|
||||
self.seed = seed
|
||||
self.logger.info(
|
||||
"Transformer parameters setting:"
|
||||
"\nnet_config : {:}"
|
||||
"\nopt_config : {:}"
|
||||
"\nmetric : {:}"
|
||||
"\ndevice : {:}"
|
||||
"\nseed : {:}".format(
|
||||
self.net_config,
|
||||
self.opt_config,
|
||||
self.metric,
|
||||
self.device,
|
||||
self.seed,
|
||||
)
|
||||
)
|
||||
|
||||
self.logger.info(
|
||||
"Transformer parameters setting:"
|
||||
"\nd_feat : {}"
|
||||
"\nhidden_size : {}"
|
||||
"\ndepth : {}"
|
||||
"\ndropout : {}"
|
||||
"\nn_epochs : {}"
|
||||
"\nlr : {}"
|
||||
"\nmetric : {}"
|
||||
"\nbatch_size : {}"
|
||||
"\nearly_stop : {}"
|
||||
"\noptimizer : {}"
|
||||
"\nloss_type : {}"
|
||||
"\nvisible_GPU : {}"
|
||||
"\nuse_GPU : {}"
|
||||
"\nseed : {}".format(
|
||||
d_feat,
|
||||
hidden_size,
|
||||
depth,
|
||||
pos_dropout,
|
||||
n_epochs,
|
||||
lr,
|
||||
metric,
|
||||
batch_size,
|
||||
early_stop,
|
||||
optimizer.lower(),
|
||||
loss,
|
||||
GPU,
|
||||
self.use_gpu,
|
||||
seed,
|
||||
)
|
||||
)
|
||||
if self.seed is not None:
|
||||
np.random.seed(self.seed)
|
||||
torch.manual_seed(self.seed)
|
||||
|
||||
if self.seed is not None:
|
||||
np.random.seed(self.seed)
|
||||
torch.manual_seed(self.seed)
|
||||
self.model = TransformerModel(
|
||||
d_feat=self.net_config["d_feat"],
|
||||
embed_dim=self.net_config["hidden_size"],
|
||||
depth=self.net_config["depth"],
|
||||
pos_drop=self.net_config["pos_drop"],
|
||||
)
|
||||
self.logger.info("model: {:}".format(self.model))
|
||||
self.logger.info("model size: {:.3f} MB".format(count_parameters(self.model)))
|
||||
|
||||
self.model = TransformerModel(d_feat=self.d_feat,
|
||||
embed_dim=self.hidden_size,
|
||||
depth=self.depth,
|
||||
pos_dropout=pos_dropout)
|
||||
self.logger.info('model: {:}'.format(self.model))
|
||||
self.logger.info('model size: {:.3f} MB'.format(count_parameters(self.model)))
|
||||
|
||||
|
||||
if optimizer.lower() == "adam":
|
||||
self.train_optimizer = optim.Adam(self.model.parameters(), lr=self.lr)
|
||||
elif optimizer.lower() == "gd":
|
||||
self.train_optimizer = optim.SGD(self.model.parameters(), lr=self.lr)
|
||||
else:
|
||||
raise NotImplementedError("optimizer {:} is not supported!".format(optimizer))
|
||||
|
||||
self.fitted = False
|
||||
self.model.to(self.device)
|
||||
|
||||
def loss_fn(self, pred, label):
|
||||
mask = ~torch.isnan(label)
|
||||
if self.loss == "mse":
|
||||
return F.mse_loss(pred[mask], label[mask])
|
||||
else:
|
||||
raise ValueError("unknown loss `{:}`".format(self.loss))
|
||||
|
||||
def metric_fn(self, pred, label):
|
||||
|
||||
mask = torch.isfinite(label)
|
||||
|
||||
if self.metric == "" or self.metric == "loss":
|
||||
return -self.loss_fn(pred[mask], label[mask])
|
||||
else:
|
||||
raise ValueError("unknown metric `{:}`".format(self.metric))
|
||||
|
||||
def train_epoch(self, x_train, y_train):
|
||||
|
||||
x_train_values = x_train.values
|
||||
y_train_values = np.squeeze(y_train.values)
|
||||
|
||||
self.model.train()
|
||||
|
||||
indices = np.arange(len(x_train_values))
|
||||
np.random.shuffle(indices)
|
||||
|
||||
for i in range(len(indices))[:: self.batch_size]:
|
||||
|
||||
if len(indices) - i < self.batch_size:
|
||||
break
|
||||
|
||||
feature = torch.from_numpy(x_train_values[indices[i : i + self.batch_size]]).float().to(self.device)
|
||||
label = torch.from_numpy(y_train_values[indices[i : i + self.batch_size]]).float().to(self.device)
|
||||
|
||||
pred = self.model(feature)
|
||||
loss = self.loss_fn(pred, label)
|
||||
|
||||
self.train_optimizer.zero_grad()
|
||||
loss.backward()
|
||||
torch.nn.utils.clip_grad_value_(self.model.parameters(), 3.0)
|
||||
self.train_optimizer.step()
|
||||
|
||||
def test_epoch(self, data_x, data_y):
|
||||
|
||||
# prepare training data
|
||||
x_values = data_x.values
|
||||
y_values = np.squeeze(data_y.values)
|
||||
|
||||
self.model.eval()
|
||||
|
||||
scores = []
|
||||
losses = []
|
||||
|
||||
indices = np.arange(len(x_values))
|
||||
|
||||
for i in range(len(indices))[:: self.batch_size]:
|
||||
|
||||
if len(indices) - i < self.batch_size:
|
||||
break
|
||||
|
||||
feature = torch.from_numpy(x_values[indices[i : i + self.batch_size]]).float().to(self.device)
|
||||
label = torch.from_numpy(y_values[indices[i : i + self.batch_size]]).float().to(self.device)
|
||||
|
||||
pred = self.model(feature)
|
||||
loss = self.loss_fn(pred, label)
|
||||
losses.append(loss.item())
|
||||
|
||||
score = self.metric_fn(pred, label)
|
||||
scores.append(score.item())
|
||||
|
||||
return np.mean(losses), np.mean(scores)
|
||||
|
||||
def fit(
|
||||
self,
|
||||
dataset: DatasetH,
|
||||
evals_result=dict(),
|
||||
verbose=True,
|
||||
save_path=None,
|
||||
):
|
||||
|
||||
df_train, df_valid, df_test = dataset.prepare(
|
||||
["train", "valid", "test"],
|
||||
col_set=["feature", "label"],
|
||||
data_key=DataHandlerLP.DK_L,
|
||||
)
|
||||
|
||||
x_train, y_train = df_train["feature"], df_train["label"]
|
||||
x_valid, y_valid = df_valid["feature"], df_valid["label"]
|
||||
|
||||
if save_path == None:
|
||||
save_path = create_save_path(save_path)
|
||||
stop_steps = 0
|
||||
train_loss = 0
|
||||
best_score = -np.inf
|
||||
best_epoch = 0
|
||||
evals_result["train"] = []
|
||||
evals_result["valid"] = []
|
||||
|
||||
# train
|
||||
self.logger.info("training...")
|
||||
self.fitted = True
|
||||
|
||||
for step in range(self.n_epochs):
|
||||
self.logger.info("Epoch%d:", step)
|
||||
self.logger.info("training...")
|
||||
self.train_epoch(x_train, y_train)
|
||||
self.logger.info("evaluating...")
|
||||
train_loss, train_score = self.test_epoch(x_train, y_train)
|
||||
val_loss, val_score = self.test_epoch(x_valid, y_valid)
|
||||
self.logger.info("train %.6f, valid %.6f" % (train_score, val_score))
|
||||
evals_result["train"].append(train_score)
|
||||
evals_result["valid"].append(val_score)
|
||||
|
||||
if val_score > best_score:
|
||||
best_score = val_score
|
||||
stop_steps = 0
|
||||
best_epoch = step
|
||||
best_param = copy.deepcopy(self.model.state_dict())
|
||||
else:
|
||||
stop_steps += 1
|
||||
if stop_steps >= self.early_stop:
|
||||
self.logger.info("early stop")
|
||||
break
|
||||
|
||||
self.logger.info("best score: %.6lf @ %d" % (best_score, best_epoch))
|
||||
self.model.load_state_dict(best_param)
|
||||
torch.save(best_param, save_path)
|
||||
|
||||
if self.use_gpu:
|
||||
torch.cuda.empty_cache()
|
||||
|
||||
def predict(self, dataset):
|
||||
|
||||
if not self.fitted:
|
||||
raise ValueError("model is not fitted yet!")
|
||||
|
||||
x_test = dataset.prepare("test", col_set="feature")
|
||||
index = x_test.index
|
||||
self.model.eval()
|
||||
x_values = x_test.values
|
||||
sample_num = x_values.shape[0]
|
||||
preds = []
|
||||
|
||||
for begin in range(sample_num)[:: self.batch_size]:
|
||||
|
||||
if sample_num - begin < self.batch_size:
|
||||
end = sample_num
|
||||
else:
|
||||
end = begin + self.batch_size
|
||||
|
||||
x_batch = torch.from_numpy(x_values[begin:end]).float().to(self.device)
|
||||
|
||||
with torch.no_grad():
|
||||
if self.use_gpu:
|
||||
pred = self.model(x_batch).detach().cpu().numpy()
|
||||
if self.opt_config["optimizer"] == "adam":
|
||||
self.train_optimizer = optim.Adam(self.model.parameters(), lr=self.opt_config["lr"])
|
||||
elif self.opt_config["optimizer"] == "adam":
|
||||
self.train_optimizer = optim.SGD(self.model.parameters(), lr=self.opt_config["lr"])
|
||||
else:
|
||||
pred = self.model(x_batch).detach().numpy()
|
||||
raise NotImplementedError("optimizer {:} is not supported!".format(optimizer))
|
||||
|
||||
preds.append(pred)
|
||||
self.fitted = False
|
||||
self.model.to(self.device)
|
||||
|
||||
return pd.Series(np.concatenate(preds), index=index)
|
||||
@property
|
||||
def use_gpu(self):
|
||||
self.device == torch.device("cpu")
|
||||
|
||||
def loss_fn(self, pred, label):
|
||||
mask = ~torch.isnan(label)
|
||||
if self.opt_config["loss"] == "mse":
|
||||
return F.mse_loss(pred[mask], label[mask])
|
||||
else:
|
||||
raise ValueError("unknown loss `{:}`".format(self.loss))
|
||||
|
||||
def metric_fn(self, pred, label):
|
||||
mask = torch.isfinite(label)
|
||||
if self.metric == "" or self.metric == "loss":
|
||||
return -self.loss_fn(pred[mask], label[mask])
|
||||
else:
|
||||
raise ValueError("unknown metric `{:}`".format(self.metric))
|
||||
|
||||
def train_epoch(self, xloader, model, loss_fn, optimizer):
|
||||
model.train()
|
||||
scores, losses = [], []
|
||||
for ibatch, (feats, labels) in enumerate(xloader):
|
||||
feats = feats.to(self.device, non_blocking=True)
|
||||
labels = labels.to(self.device, non_blocking=True)
|
||||
# forward the network
|
||||
preds = model(feats)
|
||||
loss = loss_fn(preds, labels)
|
||||
with torch.no_grad():
|
||||
score = self.metric_fn(preds, labels)
|
||||
losses.append(loss.item())
|
||||
scores.append(loss.item())
|
||||
# optimize the network
|
||||
optimizer.zero_grad()
|
||||
loss.backward()
|
||||
torch.nn.utils.clip_grad_value_(model.parameters(), 3.0)
|
||||
optimizer.step()
|
||||
return np.mean(losses), np.mean(scores)
|
||||
|
||||
def test_epoch(self, xloader, model, loss_fn, metric_fn):
|
||||
model.eval()
|
||||
scores, losses = [], []
|
||||
with torch.no_grad():
|
||||
for ibatch, (feats, labels) in enumerate(xloader):
|
||||
feats = feats.to(self.device, non_blocking=True)
|
||||
labels = labels.to(self.device, non_blocking=True)
|
||||
# forward the network
|
||||
preds = model(feats)
|
||||
loss = loss_fn(preds, labels)
|
||||
score = self.metric_fn(preds, labels)
|
||||
losses.append(loss.item())
|
||||
scores.append(loss.item())
|
||||
return np.mean(losses), np.mean(scores)
|
||||
|
||||
def fit(
|
||||
self,
|
||||
dataset: DatasetH,
|
||||
evals_result=dict(),
|
||||
verbose=True,
|
||||
save_path=None,
|
||||
):
|
||||
def _prepare_dataset(df_data):
|
||||
return th_data.TensorDataset(
|
||||
torch.from_numpy(df_data["feature"].values).float(),
|
||||
torch.from_numpy(df_data["label"].values).squeeze().float(),
|
||||
)
|
||||
|
||||
df_train, df_valid, df_test = dataset.prepare(
|
||||
["train", "valid", "test"],
|
||||
col_set=["feature", "label"],
|
||||
data_key=DataHandlerLP.DK_L,
|
||||
)
|
||||
train_dataset, valid_dataset, test_dataset = (
|
||||
_prepare_dataset(df_train),
|
||||
_prepare_dataset(df_valid),
|
||||
_prepare_dataset(df_test),
|
||||
)
|
||||
|
||||
train_loader = th_data.DataLoader(
|
||||
train_dataset, batch_size=self.opt_config["batch_size"], shuffle=True, drop_last=False, pin_memory=True
|
||||
)
|
||||
valid_loader = th_data.DataLoader(
|
||||
valid_dataset, batch_size=self.opt_config["batch_size"], shuffle=False, drop_last=False, pin_memory=True
|
||||
)
|
||||
test_loader = th_data.DataLoader(
|
||||
test_dataset, batch_size=self.opt_config["batch_size"], shuffle=False, drop_last=False, pin_memory=True
|
||||
)
|
||||
|
||||
if save_path == None:
|
||||
save_path = create_save_path(save_path)
|
||||
stop_steps, best_score, best_epoch = 0, -np.inf, 0
|
||||
train_loss = 0
|
||||
evals_result["train"] = []
|
||||
evals_result["valid"] = []
|
||||
|
||||
# train
|
||||
self.logger.info("Fit procedure for [{:}] with save path={:}".format(self.__class__.__name__, save_path))
|
||||
|
||||
def _internal_test():
|
||||
train_loss, train_score = self.test_epoch(train_loader, self.model, self.loss_fn, self.metric_fn)
|
||||
valid_loss, valid_score = self.test_epoch(valid_loader, self.model, self.loss_fn, self.metric_fn)
|
||||
test_loss, test_score = self.test_epoch(test_loader, self.model, self.loss_fn, self.metric_fn)
|
||||
xstr = "train-score={:.6f}, valid-score={:.6f}, test-score={:.6f}".format(
|
||||
train_score, valid_score, test_score
|
||||
)
|
||||
return dict(train=train_score, valid=valid_score, test=test_score), xstr
|
||||
|
||||
_, eval_str = _internal_test()
|
||||
self.logger.info("Before Training: {:}".format(eval_str))
|
||||
for iepoch in range(self.opt_config["epochs"]):
|
||||
self.logger.info("Epoch={:03d}/{:03d} ::==>>".format(iepoch, self.opt_config["epochs"]))
|
||||
|
||||
train_loss, train_score = self.train_epoch(train_loader, self.model, self.loss_fn, self.train_optimizer)
|
||||
self.logger.info("Training :: loss={:.6f}, score={:.6f}".format(train_loss, train_score))
|
||||
|
||||
eval_score_dict, eval_str = _internal_test()
|
||||
self.logger.info("Evaluating :: {:}".format(eval_str))
|
||||
evals_result["train"].append(eval_score_dict["train"])
|
||||
evals_result["valid"].append(eval_score_dict["valid"])
|
||||
|
||||
if eval_score_dict["valid"] > best_score:
|
||||
stop_steps, best_epoch, best_score = 0, iepoch, eval_score_dict["valid"]
|
||||
best_param = copy.deepcopy(self.model.state_dict())
|
||||
else:
|
||||
stop_steps += 1
|
||||
if stop_steps >= self.opt_config["early_stop"]:
|
||||
self.logger.info("early stop at {:}-th epoch, where the best is @{:}".format(iepoch, best_epoch))
|
||||
break
|
||||
|
||||
self.logger.info("The best score: {:.6f} @ {:02d}-th epoch".format(best_score, best_epoch))
|
||||
self.model.load_state_dict(best_param)
|
||||
torch.save(best_param, save_path)
|
||||
|
||||
if self.use_gpu:
|
||||
torch.cuda.empty_cache()
|
||||
self.fitted = True
|
||||
|
||||
def predict(self, dataset):
|
||||
|
||||
if not self.fitted:
|
||||
raise ValueError("model is not fitted yet!")
|
||||
|
||||
x_test = dataset.prepare("test", col_set="feature")
|
||||
index = x_test.index
|
||||
self.model.eval()
|
||||
x_values = x_test.values
|
||||
sample_num = x_values.shape[0]
|
||||
preds = []
|
||||
|
||||
for begin in range(sample_num)[:: self.batch_size]:
|
||||
|
||||
if sample_num - begin < self.batch_size:
|
||||
end = sample_num
|
||||
else:
|
||||
end = begin + self.batch_size
|
||||
|
||||
x_batch = torch.from_numpy(x_values[begin:end]).float().to(self.device)
|
||||
|
||||
with torch.no_grad():
|
||||
if self.use_gpu:
|
||||
pred = self.model(x_batch).detach().cpu().numpy()
|
||||
else:
|
||||
pred = self.model(x_batch).detach().numpy()
|
||||
|
||||
preds.append(pred)
|
||||
|
||||
return pd.Series(np.concatenate(preds), index=index)
|
||||
|
||||
|
||||
# Real Model
|
||||
|
||||
|
||||
class Attention(nn.Module):
|
||||
def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None, attn_drop=0.0, proj_drop=0.0):
|
||||
super(Attention, self).__init__()
|
||||
self.num_heads = num_heads
|
||||
head_dim = dim // num_heads
|
||||
# NOTE scale factor was wrong in my original version, can set manually to be compat with prev weights
|
||||
self.scale = qk_scale or math.sqrt(head_dim)
|
||||
|
||||
def __init__(self, dim, num_heads=8, qkv_bias=False, qk_scale=None, attn_drop=0., proj_drop=0.):
|
||||
super(Attention, self).__init__()
|
||||
self.num_heads = num_heads
|
||||
head_dim = dim // num_heads
|
||||
# NOTE scale factor was wrong in my original version, can set manually to be compat with prev weights
|
||||
self.scale = qk_scale or math.sqrt(head_dim)
|
||||
self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
|
||||
self.attn_drop = nn.Dropout(attn_drop)
|
||||
self.proj = nn.Linear(dim, dim)
|
||||
self.proj_drop = nn.Dropout(proj_drop)
|
||||
|
||||
self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
|
||||
self.attn_drop = nn.Dropout(attn_drop)
|
||||
self.proj = nn.Linear(dim, dim)
|
||||
self.proj_drop = nn.Dropout(proj_drop)
|
||||
def forward(self, x):
|
||||
B, N, C = x.shape
|
||||
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
|
||||
q, k, v = qkv[0], qkv[1], qkv[2] # make torchscript happy (cannot use tensor as tuple)
|
||||
|
||||
def forward(self, x):
|
||||
B, N, C = x.shape
|
||||
qkv = self.qkv(x).reshape(B, N, 3, self.num_heads, C // self.num_heads).permute(2, 0, 3, 1, 4)
|
||||
q, k, v = qkv[0], qkv[1], qkv[2] # make torchscript happy (cannot use tensor as tuple)
|
||||
attn = (q @ k.transpose(-2, -1)) * self.scale
|
||||
attn = attn.softmax(dim=-1)
|
||||
attn = self.attn_drop(attn)
|
||||
|
||||
attn = (q @ k.transpose(-2, -1)) * self.scale
|
||||
attn = attn.softmax(dim=-1)
|
||||
attn = self.attn_drop(attn)
|
||||
|
||||
x = (attn @ v).transpose(1, 2).reshape(B, N, C)
|
||||
x = self.proj(x)
|
||||
x = self.proj_drop(x)
|
||||
return x
|
||||
x = (attn @ v).transpose(1, 2).reshape(B, N, C)
|
||||
x = self.proj(x)
|
||||
x = self.proj_drop(x)
|
||||
return x
|
||||
|
||||
|
||||
class Block(nn.Module):
|
||||
def __init__(
|
||||
self,
|
||||
dim,
|
||||
num_heads,
|
||||
mlp_ratio=4.0,
|
||||
qkv_bias=False,
|
||||
qk_scale=None,
|
||||
attn_drop=0.0,
|
||||
mlp_drop=0.0,
|
||||
drop_path=0.0,
|
||||
act_layer=nn.GELU,
|
||||
norm_layer=nn.LayerNorm,
|
||||
):
|
||||
super(Block, self).__init__()
|
||||
self.norm1 = norm_layer(dim)
|
||||
self.attn = Attention(
|
||||
dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=mlp_drop
|
||||
)
|
||||
# NOTE: drop path for stochastic depth, we shall see if this is better than dropout here
|
||||
self.drop_path = xlayers.DropPath(drop_path) if drop_path > 0.0 else nn.Identity()
|
||||
self.norm2 = norm_layer(dim)
|
||||
mlp_hidden_dim = int(dim * mlp_ratio)
|
||||
self.mlp = xlayers.MLP(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=mlp_drop)
|
||||
|
||||
def __init__(self, dim, num_heads, mlp_ratio=4., qkv_bias=False, qk_scale=None,
|
||||
attn_drop=0., mlp_drop=0., drop_path=0.,
|
||||
act_layer=nn.GELU, norm_layer=nn.LayerNorm):
|
||||
super(Block, self).__init__()
|
||||
self.norm1 = norm_layer(dim)
|
||||
self.attn = Attention(
|
||||
dim, num_heads=num_heads, qkv_bias=qkv_bias, qk_scale=qk_scale, attn_drop=attn_drop, proj_drop=mlp_drop)
|
||||
# NOTE: drop path for stochastic depth, we shall see if this is better than dropout here
|
||||
self.drop_path = xlayers.DropPath(drop_path) if drop_path > 0. else nn.Identity()
|
||||
self.norm2 = norm_layer(dim)
|
||||
mlp_hidden_dim = int(dim * mlp_ratio)
|
||||
self.mlp = xlayers.MLP(in_features=dim, hidden_features=mlp_hidden_dim, act_layer=act_layer, drop=mlp_drop)
|
||||
|
||||
def forward(self, x):
|
||||
x = x + self.drop_path(self.attn(self.norm1(x)))
|
||||
x = x + self.drop_path(self.mlp(self.norm2(x)))
|
||||
return x
|
||||
def forward(self, x):
|
||||
x = x + self.drop_path(self.attn(self.norm1(x)))
|
||||
x = x + self.drop_path(self.mlp(self.norm2(x)))
|
||||
return x
|
||||
|
||||
|
||||
class SimpleEmbed(nn.Module):
|
||||
def __init__(self, d_feat, embed_dim):
|
||||
super(SimpleEmbed, self).__init__()
|
||||
self.d_feat = d_feat
|
||||
self.embed_dim = embed_dim
|
||||
self.proj = nn.Linear(d_feat, embed_dim)
|
||||
|
||||
def __init__(self, d_feat, embed_dim):
|
||||
super(SimpleEmbed, self).__init__()
|
||||
self.d_feat = d_feat
|
||||
self.embed_dim = embed_dim
|
||||
self.proj = nn.Linear(d_feat, embed_dim)
|
||||
|
||||
def forward(self, x):
|
||||
x = x.reshape(len(x), self.d_feat, -1) # [N, F*T] -> [N, F, T]
|
||||
x = x.permute(0, 2, 1) # [N, F, T] -> [N, T, F]
|
||||
out = self.proj(x) * math.sqrt(self.embed_dim)
|
||||
return out
|
||||
def forward(self, x):
|
||||
x = x.reshape(len(x), self.d_feat, -1) # [N, F*T] -> [N, F, T]
|
||||
x = x.permute(0, 2, 1) # [N, F, T] -> [N, T, F]
|
||||
out = self.proj(x) * math.sqrt(self.embed_dim)
|
||||
return out
|
||||
|
||||
|
||||
class TransformerModel(nn.Module):
|
||||
def __init__(
|
||||
self,
|
||||
d_feat: int,
|
||||
embed_dim: int = 64,
|
||||
depth: int = 4,
|
||||
num_heads: int = 4,
|
||||
mlp_ratio: float = 4.0,
|
||||
qkv_bias: bool = True,
|
||||
qk_scale: Optional[float] = None,
|
||||
pos_drop=0.0,
|
||||
mlp_drop_rate=0.0,
|
||||
attn_drop_rate=0.0,
|
||||
drop_path_rate=0.0,
|
||||
norm_layer=None,
|
||||
):
|
||||
"""
|
||||
Args:
|
||||
d_feat (int, tuple): input image size
|
||||
embed_dim (int): embedding dimension
|
||||
depth (int): depth of transformer
|
||||
num_heads (int): number of attention heads
|
||||
mlp_ratio (int): ratio of mlp hidden dim to embedding dim
|
||||
qkv_bias (bool): enable bias for qkv if True
|
||||
qk_scale (float): override default qk scale of head_dim ** -0.5 if set
|
||||
pos_drop (float): dropout rate for the positional embedding
|
||||
mlp_drop_rate (float): the dropout rate for MLP layers in a block
|
||||
attn_drop_rate (float): attention dropout rate
|
||||
drop_path_rate (float): stochastic depth rate
|
||||
norm_layer: (nn.Module): normalization layer
|
||||
"""
|
||||
super(TransformerModel, self).__init__()
|
||||
self.embed_dim = embed_dim
|
||||
self.num_features = embed_dim
|
||||
norm_layer = norm_layer or partial(nn.LayerNorm, eps=1e-6)
|
||||
|
||||
def __init__(self,
|
||||
d_feat: int,
|
||||
embed_dim: int = 64,
|
||||
depth: int = 4,
|
||||
num_heads: int = 4,
|
||||
mlp_ratio: float = 4.,
|
||||
qkv_bias: bool = True,
|
||||
qk_scale: Optional[float] = None,
|
||||
pos_dropout=0., mlp_drop_rate=0., attn_drop_rate=0., drop_path_rate=0., norm_layer=None):
|
||||
"""
|
||||
Args:
|
||||
d_feat (int, tuple): input image size
|
||||
embed_dim (int): embedding dimension
|
||||
depth (int): depth of transformer
|
||||
num_heads (int): number of attention heads
|
||||
mlp_ratio (int): ratio of mlp hidden dim to embedding dim
|
||||
qkv_bias (bool): enable bias for qkv if True
|
||||
qk_scale (float): override default qk scale of head_dim ** -0.5 if set
|
||||
pos_dropout (float): dropout rate for the positional embedding
|
||||
mlp_drop_rate (float): the dropout rate for MLP layers in a block
|
||||
attn_drop_rate (float): attention dropout rate
|
||||
drop_path_rate (float): stochastic depth rate
|
||||
norm_layer: (nn.Module): normalization layer
|
||||
"""
|
||||
super(TransformerModel, self).__init__()
|
||||
self.embed_dim = embed_dim
|
||||
self.num_features = embed_dim
|
||||
norm_layer = norm_layer or partial(nn.LayerNorm, eps=1e-6)
|
||||
self.input_embed = SimpleEmbed(d_feat, embed_dim=embed_dim)
|
||||
|
||||
self.input_embed = SimpleEmbed(d_feat, embed_dim=embed_dim)
|
||||
self.cls_token = nn.Parameter(torch.zeros(1, 1, embed_dim))
|
||||
self.pos_embed = xlayers.PositionalEncoder(d_model=embed_dim, max_seq_len=65, dropout=pos_drop)
|
||||
|
||||
self.cls_token = nn.Parameter(torch.zeros(1, 1, embed_dim))
|
||||
self.pos_embed = xlayers.PositionalEncoder(d_model=embed_dim, max_seq_len=65, dropout=pos_dropout)
|
||||
dpr = [x.item() for x in torch.linspace(0, drop_path_rate, depth)] # stochastic depth decay rule
|
||||
self.blocks = nn.ModuleList(
|
||||
[
|
||||
Block(
|
||||
dim=embed_dim,
|
||||
num_heads=num_heads,
|
||||
mlp_ratio=mlp_ratio,
|
||||
qkv_bias=qkv_bias,
|
||||
qk_scale=qk_scale,
|
||||
attn_drop=attn_drop_rate,
|
||||
mlp_drop=mlp_drop_rate,
|
||||
drop_path=dpr[i],
|
||||
norm_layer=norm_layer,
|
||||
)
|
||||
for i in range(depth)
|
||||
]
|
||||
)
|
||||
self.norm = norm_layer(embed_dim)
|
||||
|
||||
dpr = [x.item() for x in torch.linspace(0, drop_path_rate, depth)] # stochastic depth decay rule
|
||||
self.blocks = nn.ModuleList([
|
||||
Block(
|
||||
dim=embed_dim, num_heads=num_heads, mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, qk_scale=qk_scale,
|
||||
attn_drop=attn_drop_rate, mlp_drop=mlp_drop_rate, drop_path=dpr[i], norm_layer=norm_layer)
|
||||
for i in range(depth)])
|
||||
self.norm = norm_layer(embed_dim)
|
||||
# regression head
|
||||
self.head = nn.Linear(self.num_features, 1)
|
||||
|
||||
# regression head
|
||||
self.head = nn.Linear(self.num_features, 1)
|
||||
xlayers.trunc_normal_(self.cls_token, std=0.02)
|
||||
self.apply(self._init_weights)
|
||||
|
||||
xlayers.trunc_normal_(self.cls_token, std=.02)
|
||||
self.apply(self._init_weights)
|
||||
def _init_weights(self, m):
|
||||
if isinstance(m, nn.Linear):
|
||||
xlayers.trunc_normal_(m.weight, std=0.02)
|
||||
if isinstance(m, nn.Linear) and m.bias is not None:
|
||||
nn.init.constant_(m.bias, 0)
|
||||
elif isinstance(m, nn.LayerNorm):
|
||||
nn.init.constant_(m.bias, 0)
|
||||
nn.init.constant_(m.weight, 1.0)
|
||||
|
||||
def _init_weights(self, m):
|
||||
if isinstance(m, nn.Linear):
|
||||
xlayers.trunc_normal_(m.weight, std=.02)
|
||||
if isinstance(m, nn.Linear) and m.bias is not None:
|
||||
nn.init.constant_(m.bias, 0)
|
||||
elif isinstance(m, nn.LayerNorm):
|
||||
nn.init.constant_(m.bias, 0)
|
||||
nn.init.constant_(m.weight, 1.0)
|
||||
def forward_features(self, x):
|
||||
batch, flatten_size = x.shape
|
||||
feats = self.input_embed(x) # batch * 60 * 64
|
||||
|
||||
def forward_features(self, x):
|
||||
batch, flatten_size = x.shape
|
||||
feats = self.input_embed(x) # batch * 60 * 64
|
||||
cls_tokens = self.cls_token.expand(batch, -1, -1) # stole cls_tokens impl from Phil Wang, thanks
|
||||
feats_w_ct = torch.cat((cls_tokens, feats), dim=1)
|
||||
feats_w_tp = self.pos_embed(feats_w_ct)
|
||||
|
||||
cls_tokens = self.cls_token.expand(batch, -1, -1) # stole cls_tokens impl from Phil Wang, thanks
|
||||
feats_w_ct = torch.cat((cls_tokens, feats), dim=1)
|
||||
feats_w_tp = self.pos_embed(feats_w_ct)
|
||||
xfeats = feats_w_tp
|
||||
for block in self.blocks:
|
||||
xfeats = block(xfeats)
|
||||
|
||||
xfeats = feats_w_tp
|
||||
for block in self.blocks:
|
||||
xfeats = block(xfeats)
|
||||
xfeats = self.norm(xfeats)[:, 0]
|
||||
return xfeats
|
||||
|
||||
xfeats = self.norm(xfeats)[:, 0]
|
||||
return xfeats
|
||||
|
||||
def forward(self, x):
|
||||
feats = self.forward_features(x)
|
||||
predicts = self.head(feats).squeeze(-1)
|
||||
return predicts
|
||||
def forward(self, x):
|
||||
feats = self.forward_features(x)
|
||||
predicts = self.head(feats).squeeze(-1)
|
||||
return predicts
|
||||
|
||||
Loading…
Reference in New Issue
Block a user