File size: 14,405 Bytes
aa7fb02 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 |
import torch
import torch.nn as nn
import torch.nn.functional as F
import numpy as np
from Model.CLIP.cn_clip.clip import load_from_name
import cn_clip.clip as clip
from PIL import Image
from torch.autograd import Variable
from fur_rl.models.StageOne import Stage1
from fur_rl.models.transformer import ModelAttn
import copy
import PIL
# v3 带softmax的模型,actor
class Retriever_v3(nn.Module):
def __init__(self, sentence_clip_model2, sentence_clip_preprocess2, device):
super(Retriever_v3, self).__init__()
self.sentence_clip_model2 = sentence_clip_model2
self.sentence_clip_preprocess2 = sentence_clip_preprocess2
self.tokenize = clip.tokenize
self.Stage1 = Stage1(hid_dim=512)
# self.conv1 = nn.Conv1d(in_channels=3, out_channels=1, kernel_size=3, stride=1, padding=2, dilation=1, groups=1, bias=True, padding_mode='zeros')
# self.conv2 = nn.Conv1d(in_channels=64, out_channels=32, kernel_size=3, stride=1, padding=2, dilation=1, groups=1, bias=True, padding_mode='zeros')
# self.conv3 = nn.Conv1d(in_channels=32, out_channels=16, kernel_size=3, stride=1, padding=2, dilation=1, groups=1, bias=True, padding_mode='zeros')
# self.conv4 = nn.Conv1d(in_channels=16, out_channels=1, kernel_size=1, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros')
self.fc01 = nn.Linear(in_features=512 * 2, out_features=512, bias=True)
self.fc02 = nn.Linear(in_features=512, out_features=1, bias=True)
self.fc03 = nn.Linear(in_features=256, out_features=64, bias=True)
self.fc04 = nn.Linear(in_features=64, out_features=1, bias=True)
# self.fc1 = nn.Linear(in_features=516, out_features=1, bias=True)
self.fc2 = nn.Linear(in_features=512, out_features=512, bias=True)
self.logistic = nn.Sigmoid()
self.relu = nn.ReLU()
self.softmax = nn.Softmax(dim=1)
self.bn = nn.BatchNorm1d(num_features=1)
self.Attn0 = ModelAttn()
self.Attn1 = ModelAttn()
self.norm = nn.LayerNorm(normalized_shape=512)
self.norm50 = nn.LayerNorm(normalized_shape=50)
self.device = device
def img_embed(self, img):
img_embed = self.sentence_clip_model2.encode_image(img)
img_embed = img_embed / img_embed.norm(dim=-1, keepdim=True)
return img_embed
def txt_embed(self, txt_token):
txt_embed = self.sentence_clip_model2.encode_text(txt_token)
txt_embed = txt_embed / txt_embed.norm(dim=-1, keepdim=True)
return txt_embed
def get_env_feature(self, img, txt, reward):
hx = self.Stage1.hx
x = self.Stage1(img, txt, reward)
x = x / x.norm(dim=-1, keepdim=True)
return x, hx
def init_hid(self, batch_size=1):
self.Stage1.init_hid(batch_size=batch_size)
return
def detach_hid(self):
self.Stage1.detach_hid()
return
def get_Q(self, img, txt, action, reward):
# env_feature, hx = self.get_env_feature(img, txt, reward)
img = img.unsqueeze(dim=1)
txt = txt.unsqueeze(dim=1)
env_feature = self.Attn0(img, txt, txt)
# transformers:
x = self.Attn1(env_feature, action, env_feature)
# linear:
x = self.fc02(x)
x = self.logistic(x)
x = torch.clamp(x, min=1e-5, max=1 - 1e-5)
x = x.squeeze(dim=1)
return x
def forward(self, img, txt, action, hx, r):
self.Stage1.hx = hx
P = torch.zeros((action.shape[0], action.shape[1], 1), device=self.device)
for i in range(action.shape[1]):
P[:, i] = self.get_Q(img, txt, action[:, i], r)
P = P.squeeze(dim=2)
# P = self.softmax(P).squeeze(dim=2)
# P = torch.clamp(P, min=1e-6, max=1-1e-6)
return P
# v5 critic模型,只用1个candidate
class Retriever_v5(nn.Module):
def __init__(self, sentence_clip_model2, sentence_clip_preprocess2, device):
super(Retriever_v5, self).__init__()
self.sentence_clip_model2 = sentence_clip_model2
self.sentence_clip_preprocess2 = sentence_clip_preprocess2
self.tokenize = clip.tokenize
self.Stage1 = Stage1(hid_dim=512)
# self.conv1 = nn.Conv1d(in_channels=3, out_channels=1, kernel_size=3, stride=1, padding=2, dilation=1, groups=1, bias=True, padding_mode='zeros')
# self.conv2 = nn.Conv1d(in_channels=64, out_channels=32, kernel_size=3, stride=1, padding=2, dilation=1, groups=1, bias=True, padding_mode='zeros')
# self.conv3 = nn.Conv1d(in_channels=32, out_channels=16, kernel_size=3, stride=1, padding=2, dilation=1, groups=1, bias=True, padding_mode='zeros')
# self.conv4 = nn.Conv1d(in_channels=16, out_channels=1, kernel_size=1, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros')
self.fc01 = nn.Linear(in_features=512 * 2, out_features=512, bias=True)
self.fc02 = nn.Linear(in_features=512, out_features=1, bias=True)
self.fc03 = nn.Linear(in_features=256, out_features=64, bias=True)
self.fc04 = nn.Linear(in_features=64, out_features=1, bias=True)
# self.fc1 = nn.Linear(in_features=516, out_features=1, bias=True)
self.fc2 = nn.Linear(in_features=512, out_features=512, bias=True)
self.logistic = nn.Sigmoid()
self.relu = nn.ReLU()
self.softmax = nn.Softmax(dim=1)
self.bn = nn.BatchNorm1d(num_features=1)
self.Attn0 = ModelAttn()
self.Attn1 = ModelAttn()
self.norm = nn.LayerNorm(normalized_shape=512)
self.norm50 = nn.LayerNorm(normalized_shape=50)
self.device = device
def img_embed(self, img):
img_embed = self.sentence_clip_model2.encode_image(img)
img_embed = img_embed / img_embed.norm(dim=-1, keepdim=True)
return img_embed
def txt_embed(self, txt_token):
txt_embed = self.sentence_clip_model2.encode_text(txt_token)
txt_embed = txt_embed / txt_embed.norm(dim=-1, keepdim=True)
return txt_embed
def get_env_feature(self, img, txt, reward):
hx = self.Stage1.hx
x = self.Stage1(img, txt, reward)
x = x / x.norm(dim=-1, keepdim=True)
return x, hx
def init_hid(self, batch_size=1):
self.Stage1.init_hid(batch_size=batch_size)
return
def detach_hid(self):
self.Stage1.detach_hid()
return
def get_Q(self, img, txt, action, reward):
# env_feature, hx = self.get_env_feature(img, txt, reward)
img = img.unsqueeze(dim=1)
txt = txt.unsqueeze(dim=1)
env_feature = self.Attn0(img, txt, txt)
# transformers:
x = self.Attn1(env_feature, action, env_feature)
# linear:
x = self.fc02(x)
# x = self.logistic(x)
# x = torch.clamp(x, min=1e-5, max=1-1e-5)
x = x.squeeze(dim=1)
return x
def forward(self, img, txt, action, hx, r):
self.Stage1.hx = hx
Q = self.get_Q(img, txt, action, r)
Q = self.logistic(Q)
return Q
class Supervised(nn.Module):
def __init__(self, sentence_clip_model2, sentence_clip_preprocess2, device):
super(Supervised, self).__init__()
self.sentence_clip_model2 = sentence_clip_model2
self.sentence_clip_preprocess2 = sentence_clip_preprocess2
self.tokenize = clip.tokenize
self.Stage1 = Stage1(hid_dim=512)
# self.conv1 = nn.Conv1d(in_channels=3, out_channels=1, kernel_size=3, stride=1, padding=2, dilation=1, groups=1, bias=True, padding_mode='zeros')
# self.conv2 = nn.Conv1d(in_channels=64, out_channels=32, kernel_size=3, stride=1, padding=2, dilation=1, groups=1, bias=True, padding_mode='zeros')
# self.conv3 = nn.Conv1d(in_channels=32, out_channels=16, kernel_size=3, stride=1, padding=2, dilation=1, groups=1, bias=True, padding_mode='zeros')
# self.conv4 = nn.Conv1d(in_channels=16, out_channels=1, kernel_size=1, stride=1, padding=0, dilation=1, groups=1, bias=True, padding_mode='zeros')
self.fc01 = nn.Linear(in_features=512 * 2, out_features=512, bias=True)
self.fc02 = nn.Linear(in_features=512, out_features=1, bias=True)
self.fc03 = nn.Linear(in_features=256, out_features=64, bias=True)
self.fc04 = nn.Linear(in_features=64, out_features=1, bias=True)
# self.fc1 = nn.Linear(in_features=516, out_features=1, bias=True)
self.fc2 = nn.Linear(in_features=512, out_features=512, bias=True)
self.logistic = nn.Sigmoid()
self.relu = nn.ReLU()
self.softmax = nn.Softmax(dim=1)
self.bn = nn.BatchNorm1d(num_features=1)
self.Attn0 = ModelAttn()
self.Attn1 = ModelAttn()
self.norm = nn.LayerNorm(normalized_shape=512)
self.norm50 = nn.LayerNorm(normalized_shape=50)
self.device = device
def img_embed(self, img):
img_embed = self.sentence_clip_model2.encode_image(img)
img_embed = img_embed / img_embed.norm(dim=-1, keepdim=True)
return img_embed
def txt_embed(self, txt_token):
txt_embed = self.sentence_clip_model2.encode_text(txt_token)
txt_embed = txt_embed / txt_embed.norm(dim=-1, keepdim=True)
return txt_embed
def get_env_feature(self, img, txt, reward):
hx = self.Stage1.hx
x = self.Stage1(img, txt, reward)
x = x / x.norm(dim=-1, keepdim=True)
return x, hx
def init_hid(self, batch_size=1):
self.Stage1.init_hid(batch_size=batch_size)
return
def detach_hid(self):
self.Stage1.detach_hid()
return
def get_Q(self, img, txt, action, reward):
# env_feature, hx = self.get_env_feature(img, txt, reward)
img = img.unsqueeze(dim=1)
txt = txt.unsqueeze(dim=1)
env_feature = self.Attn0(img, txt, txt)
env_feature = env_feature / env_feature.norm(dim=-1, keepdim=True)
return env_feature
def forward(self, img, txt, action, hx, r):
self.Stage1.hx = hx
Q = self.get_Q(img, txt, action, r)
return Q
def action(self, img, txt, actions, hx, r):
Q = self.get_Q(img, txt, actions, r)
P = torch.zeros(actions.shape[0], actions.shape[1], 1)
for i in range(actions.shape[1]):
for j in range(actions.shape[0]):
P[j, i] = Q[j] @ (actions[j, i].unsqueeze(0)).t()
return P.squeeze(dim=2)
MEMORY_CAPACITY = 100
N_STATES = 4
GAMMA = 0.95
class DQN_v3():
def __init__(self, sentence_clip_model2, sentence_clip_preprocess2, device, MULTI_GPU=False, device_ids=None):
self.eval_net = Retriever_v5(sentence_clip_model2, sentence_clip_preprocess2, device).to(device)
self.target_net = Retriever_v5(sentence_clip_model2, sentence_clip_preprocess2, device).to(device)
self.actor_net = Retriever_v3(sentence_clip_model2, sentence_clip_preprocess2, device).to(device)
self.supervised_net = Supervised(sentence_clip_model2, sentence_clip_preprocess2, device).to(device)
self.num_train = 0
self.device = device
self.MULTI_GPU = MULTI_GPU
self.device_ids = device_ids
self.memory = [] # initialize memory
self.memory_counter = 0
self.actor_optimizer = torch.optim.Adam(self.actor_net.parameters(), lr=0.0001)
self.critic_optimizer = torch.optim.Adam(self.eval_net.parameters(), lr=0.00001)
self.supervised_optimizer = torch.optim.Adam(self.supervised_net.parameters(), lr=0.001)
self.loss_func1 = nn.MSELoss()
self.loss_func2 = nn.MSELoss()
self.batch_mem = {}
def store_transition(self, g, f, c, hx, a, r, g_, f_, c_, hx_, t, batch_size, net_mem, success_turn, turn):
if turn == 0:
for i in range(batch_size):
self.batch_mem[i] = []
for i in range(batch_size):
if r[i] <= -2000:
continue
else:
g_tmp = copy.deepcopy(g[i])
f_tmp = copy.deepcopy(f[i])
c_tmp = copy.deepcopy(c[i])
hx_tmp = copy.deepcopy(hx[i])
a_tmp = copy.deepcopy(a[i])
reward_temp = copy.deepcopy(r[i])
g_tmp_ = copy.deepcopy(g_[i])
f_tmp_ = copy.deepcopy(f_[i])
c_tmp_ = copy.deepcopy(c_[i])
hx_tmp_ = copy.deepcopy(hx_[i])
t_tmp = copy.deepcopy(t[i])
self.batch_mem[i].append((g_tmp, f_tmp, c_tmp, hx_tmp, a_tmp, reward_temp,
g_tmp_, f_tmp_, c_tmp_, hx_tmp_, t_tmp))
if success_turn[i] >= 0:
# print("len", len(self.memory), len(self.batch_mem[i]))
self.memory.extend(self.batch_mem[i])
# print("len", len(self.memory))
while len(self.memory) > net_mem:
self.memory.pop(0)
self.memory_counter = len(self.memory)
# print("success", success_turn[i], reward_temp, len(self.memory))
def learn(self, batch_size, device=None):
# target parameter update
if self.memory_counter < batch_size:
return 0, 0
else:
sample_index = np.random.choice(len(self.memory), batch_size, replace=False)
sample_index = torch.LongTensor(sample_index).to(device)
g, f, c, hx, a, r, g_, f_, c_, hx_, t = zip(*[self.memory[i] for i in sample_index])
b_g = torch.stack(g)
b_f = torch.stack(f)
b_c = torch.stack(c)
b_hx = torch.stack(hx)
# b_a = torch.stack(a).unsqueeze(1)
b_r = torch.stack(r)
b_g_ = torch.stack(g_)
b_f_ = torch.stack(f_)
b_c_ = torch.stack(c_)
b_hx_ = torch.stack(hx_)
b_t = torch.stack(t)
# print("b_g", b_g.shape, b_f.shape, b_c.shape, b_hx.shape, b_a.shape, b_r.shape, b_g_.shape, b_f_.shape, b_c_.shape, b_hx_.shape)
out_s = self.supervised_net(b_g, b_f, b_c, b_hx, b_r)
loss_s = torch.zeros(out_s.shape[0], 1).to(device)
for i in range(out_s.shape[0]):
loss_s[i] = - out_s[i] @ b_t[i].unsqueeze(0).t()
loss_s = torch.mean(loss_s)
self.supervised_optimizer.zero_grad()
loss_s.backward()
self.supervised_optimizer.step()
return loss_s, 0
|