Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

为什么我实现DQN的时候,奖励值一直在个位数浮动? #88

Open
zhouhuhq opened this issue May 7, 2020 · 3 comments
Open

Comments

@zhouhuhq
Copy link

zhouhuhq commented May 7, 2020

使用的pytorch版本为1.4
Episode: 203 | Ep_r: 3.3
Episode: 204 | Ep_r: 4.04
Episode: 205 | Ep_r: 2.47
Episode: 206 | Ep_r: 0.77
Episode: 207 | Ep_r: 2.5
Episode: 208 | Ep_r: 2.7
Episode: 209 | Ep_r: 2.76
Episode: 210 | Ep_r: 3.7
Episode: 211 | Ep_r: 3.68
Episode: 212 | Ep_r: 2.98
Episode: 213 | Ep_r: 2.87
Episode: 214 | Ep_r: 1.78
Episode: 215 | Ep_r: 2.91
Episode: 216 | Ep_r: 1.71
Episode: 217 | Ep_r: 2.58
Episode: 218 | Ep_r: 2.69
Episode: 219 | Ep_r: 2.79
Episode: 220 | Ep_r: 2.61
Episode: 221 | Ep_r: 2.75
Episode: 222 | Ep_r: 3.97
Episode: 223 | Ep_r: 1.32
Episode: 224 | Ep_r: 2.91
Episode: 225 | Ep_r: 7.97
Episode: 226 | Ep_r: 3.17
Episode: 227 | Ep_r: 2.01
Episode: 228 | Ep_r: 1.29
Episode: 229 | Ep_r: 1.58
Episode: 230 | Ep_r: 2.85
Episode: 231 | Ep_r: 2.62
Episode: 232 | Ep_r: 2.81
Episode: 233 | Ep_r: 1.18
Episode: 234 | Ep_r: 2.53
Episode: 235 | Ep_r: 3.03
Episode: 236 | Ep_r: 4.22
Episode: 237 | Ep_r: 2.72
Episode: 238 | Ep_r: 1.47
Episode: 239 | Ep_r: 2.82
Episode: 240 | Ep_r: 2.32
Episode: 241 | Ep_r: 3.68
Episode: 242 | Ep_r: 13.17
Episode: 243 | Ep_r: 1.98
Episode: 244 | Ep_r: 1.14
Episode: 245 | Ep_r: 2.83
Episode: 246 | Ep_r: 4.7
Episode: 247 | Ep_r: 4.25
Episode: 248 | Ep_r: 1.54
Episode: 249 | Ep_r: 2.72
Episode: 250 | Ep_r: 1.56
Episode: 251 | Ep_r: -0.09
Episode: 252 | Ep_r: 8.29
Episode: 253 | Ep_r: 2.72
Episode: 254 | Ep_r: 3.11
Episode: 255 | Ep_r: 2.9
Episode: 256 | Ep_r: 1.95
Episode: 257 | Ep_r: 2.62
Episode: 258 | Ep_r: 2.75
Episode: 259 | Ep_r: 1.44
Episode: 260 | Ep_r: 2.52
Episode: 261 | Ep_r: 4.06
Episode: 262 | Ep_r: 2.73
Episode: 263 | Ep_r: 2.54
Episode: 264 | Ep_r: 2.34
Episode: 265 | Ep_r: 1.42
Episode: 266 | Ep_r: 2.2
Episode: 267 | Ep_r: 3.05
Episode: 268 | Ep_r: 1.45
Episode: 269 | Ep_r: 1.8
Episode: 270 | Ep_r: 1.09
Episode: 271 | Ep_r: 6.17
Episode: 272 | Ep_r: 2.36
Episode: 273 | Ep_r: 1.4
Episode: 274 | Ep_r: 1.31
Episode: 275 | Ep_r: 1.66
Episode: 276 | Ep_r: 8.7
Episode: 277 | Ep_r: 16.0
Episode: 278 | Ep_r: 1.43
Episode: 279 | Ep_r: 1.34
Episode: 280 | Ep_r: 5.84
Episode: 281 | Ep_r: 2.04
Episode: 282 | Ep_r: 3.77
Episode: 283 | Ep_r: 3.4
Episode: 284 | Ep_r: 2.19
Episode: 285 | Ep_r: 4.2
Episode: 286 | Ep_r: 2.63
Episode: 287 | Ep_r: 1.75
Episode: 288 | Ep_r: 1.95
Episode: 289 | Ep_r: 11.47
Episode: 290 | Ep_r: 3.57
Episode: 291 | Ep_r: 1.54
Episode: 292 | Ep_r: 2.3
Episode: 293 | Ep_r: 2.43
Episode: 294 | Ep_r: 1.16
Episode: 295 | Ep_r: 1.03
Episode: 296 | Ep_r: 5.55
Episode: 297 | Ep_r: 3.22
Episode: 298 | Ep_r: 2.77
Episode: 299 | Ep_r: 2.54
Episode: 300 | Ep_r: 2.37
Episode: 301 | Ep_r: 2.75
Episode: 302 | Ep_r: 1.15
Episode: 303 | Ep_r: 5.51
Episode: 304 | Ep_r: 7.01
Episode: 305 | Ep_r: 1.91
Episode: 306 | Ep_r: 1.06
Episode: 307 | Ep_r: 4.11
Episode: 308 | Ep_r: 4.02
Episode: 309 | Ep_r: 3.77
Episode: 310 | Ep_r: 1.13
Episode: 311 | Ep_r: 7.76
Episode: 312 | Ep_r: 1.37
Episode: 313 | Ep_r: 2.72
Episode: 314 | Ep_r: 6.43
Episode: 315 | Ep_r: 2.58
Episode: 316 | Ep_r: 2.2
Episode: 317 | Ep_r: 1.28
Episode: 318 | Ep_r: 1.67
Episode: 319 | Ep_r: 4.42
Episode: 320 | Ep_r: 1.19
Episode: 321 | Ep_r: 3.16
Episode: 322 | Ep_r: 7.52
Episode: 323 | Ep_r: 1.26
Episode: 324 | Ep_r: 2.74
Episode: 325 | Ep_r: 3.38
Episode: 326 | Ep_r: 3.63
Episode: 327 | Ep_r: 2.27
Episode: 328 | Ep_r: 8.43
Episode: 329 | Ep_r: 1.87
Episode: 330 | Ep_r: 2.65
Episode: 331 | Ep_r: 2.93
Episode: 332 | Ep_r: 2.56
Episode: 333 | Ep_r: 1.45
Episode: 334 | Ep_r: 2.77
Episode: 335 | Ep_r: 1.25
Episode: 336 | Ep_r: 3.33
Episode: 337 | Ep_r: 1.84
Episode: 338 | Ep_r: 6.95
Episode: 339 | Ep_r: 1.01
Episode: 340 | Ep_r: 1.79
Episode: 341 | Ep_r: 5.57
Episode: 342 | Ep_r: 9.24
Episode: 343 | Ep_r: 7.33
Episode: 344 | Ep_r: 2.18
Episode: 345 | Ep_r: 3.22
Episode: 346 | Ep_r: 1.83
Episode: 347 | Ep_r: 2.05
Episode: 348 | Ep_r: 1.11
Episode: 349 | Ep_r: 2.71
Episode: 350 | Ep_r: 1.75
Episode: 351 | Ep_r: 1.13
Episode: 352 | Ep_r: 2.1
Episode: 353 | Ep_r: 1.47
Episode: 354 | Ep_r: 1.76
Episode: 355 | Ep_r: 1.37
Episode: 356 | Ep_r: 3.98
Episode: 357 | Ep_r: 1.16
Episode: 358 | Ep_r: 2.55
Episode: 359 | Ep_r: 3.69
Episode: 360 | Ep_r: 1.46
Episode: 361 | Ep_r: 1.99
Episode: 362 | Ep_r: 1.48
Episode: 363 | Ep_r: 1.75
Episode: 364 | Ep_r: 1.18
Episode: 365 | Ep_r: 2.63
Episode: 366 | Ep_r: 2.21
Episode: 367 | Ep_r: 0.67
Episode: 368 | Ep_r: 4.65
Episode: 369 | Ep_r: 5.28
Episode: 370 | Ep_r: 2.61
Episode: 371 | Ep_r: 2.56
Episode: 372 | Ep_r: 1.63
Episode: 373 | Ep_r: 1.13
Episode: 374 | Ep_r: 4.35
Episode: 375 | Ep_r: 4.3
Episode: 376 | Ep_r: 1.33
Episode: 377 | Ep_r: 1.57
Episode: 378 | Ep_r: 5.16
Episode: 379 | Ep_r: 2.54
Episode: 380 | Ep_r: 1.19
Episode: 381 | Ep_r: 0.95
Episode: 382 | Ep_r: 2.27
Episode: 383 | Ep_r: 1.5
Episode: 384 | Ep_r: 4.11
Episode: 385 | Ep_r: 1.28
Episode: 386 | Ep_r: 2.85
Episode: 387 | Ep_r: 1.57
Episode: 388 | Ep_r: 2.72
Episode: 389 | Ep_r: 2.54
Episode: 390 | Ep_r: 1.52
Episode: 391 | Ep_r: 3.69
Episode: 392 | Ep_r: 9.81
Episode: 393 | Ep_r: 2.85
Episode: 394 | Ep_r: 2.87
Episode: 395 | Ep_r: 0.46
Episode: 396 | Ep_r: 1.26
Episode: 397 | Ep_r: 1.57
Episode: 398 | Ep_r: 1.88
Episode: 399 | Ep_r: 3.0

@backlry
Copy link

backlry commented Dec 7, 2020

兄弟,我以为有这个问题,怎么解决的还记得吗

@zhouhuhq
Copy link
Author

zhouhuhq commented Dec 9, 2020

兄弟,我以为有这个问题,怎么解决的还记得吗

太久远了,我忘了,隐隐约约记得用莫烦的代码就好了,因为我当时是自己写的,没找到原因。

@backlry
Copy link

backlry commented Dec 9, 2020

我也是自己写的,用原版的就行,就是不知道自己的为啥有这个问题

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants