custom_torch_ppg.py 19.6 KB
Newer Older
Dipam Chakraborty's avatar
Dipam Chakraborty committed
1
2
3
4
5
6
7
8
9
10
11
from ray.rllib.policy.torch_policy import TorchPolicy
import numpy as np
from ray.rllib.utils.torch_ops import convert_to_non_torch_type, convert_to_torch_tensor
from ray.rllib.utils import try_import_torch
from ray.rllib.models import ModelCatalog
from ray.rllib.utils.annotations import override
from collections import deque
from .utils import *
import time

torch, nn = try_import_torch()
12
import torch.distributions as td
Dipam Chakraborty's avatar
Dipam Chakraborty committed
13
from torch.cuda.amp import autocast, GradScaler
Dipam Chakraborty's avatar
Dipam Chakraborty committed
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

class CustomTorchPolicy(TorchPolicy):
    """Example of a random policy
    If you are using tensorflow/pytorch to build custom policies,
    you might find `build_tf_policy` and `build_torch_policy` to
    be useful.
    Adopted from examples from https://docs.ray.io/en/master/rllib-concepts.html
    """

    def __init__(self, observation_space, action_space, config):
        self.config = config

        self.device = torch.device("cuda") if torch.cuda.is_available() else torch.device("cpu")
        dist_class, logit_dim = ModelCatalog.get_action_dist(
            action_space, self.config["model"], framework="torch")
        self.model = ModelCatalog.get_model_v2(
                        obs_space=observation_space,
                        action_space=action_space,
                        num_outputs=logit_dim,
                        model_config=self.config["model"],
                        framework="torch",
                        device=self.device,
                     )

        TorchPolicy.__init__(
            self,
            observation_space=observation_space,
            action_space=action_space,
            config=config,
            model=self.model,
            loss=None,
            action_distribution_class=dist_class,
        )
47

Dipam Chakraborty's avatar
Dipam Chakraborty committed
48
        self.framework = "torch"
Dipam Chakraborty's avatar
Dipam Chakraborty committed
49
50
51
52
53
        aux_params = set(self.model.aux_vf.parameters())
        value_params = set(self.model.value_fc.parameters())
        network_params = set(self.model.parameters())
        aux_optim_params = list(network_params - value_params)
        ppo_optim_params = list(network_params - aux_params - value_params)
Dipam Chakraborty's avatar
Dipam Chakraborty committed
54
55
56
57
        if not self.config['single_optimizer']:
            self.optimizer = torch.optim.Adam(ppo_optim_params, lr=self.config['lr'])
        else:
            self.optimizer = torch.optim.Adam(network_params, lr=self.config['lr'])
Dipam Chakraborty's avatar
Dipam Chakraborty committed
58
59
        self.aux_optimizer = torch.optim.Adam(aux_optim_params, lr=self.config['aux_lr'])
        self.value_optimizer = torch.optim.Adam(value_params, lr=self.config['value_lr'])
Dipam Chakraborty's avatar
Dipam Chakraborty committed
60
61
62
63
64
65
66
67
68
69
70
        self.max_reward = self.config['env_config']['return_max']
        self.rewnorm = RewardNormalizer(cliprew=self.max_reward) ## TODO: Might need to go to custom state
        self.reward_deque = deque(maxlen=100)
        self.best_reward = -np.inf
        self.best_weights = None
        self.time_elapsed = 0
        self.batch_end_time = time.time()
        self.timesteps_total = 0
        self.best_rew_tsteps = 0
        
        nw = self.config['num_workers'] if self.config['num_workers'] > 0 else 1
Dipam Chakraborty's avatar
Dipam Chakraborty committed
71
72
73
74
        nenvs = nw * self.config['num_envs_per_worker']
        nsteps = self.config['rollout_fragment_length']
        n_pi = self.config['n_pi']
        self.nbatch = nenvs * nsteps
Dipam Chakraborty's avatar
Dipam Chakraborty committed
75
76
77
78
79
80
81
        self.actual_batch_size = self.nbatch // self.config['updates_per_batch']
        self.accumulate_train_batches = int(np.ceil( self.actual_batch_size / self.config['max_minibatch_size'] ))
        self.mem_limited_batch_size = self.actual_batch_size // self.accumulate_train_batches
        if self.nbatch % self.actual_batch_size != 0 or self.nbatch % self.mem_limited_batch_size != 0:
            print("#################################################")
            print("WARNING: MEMORY LIMITED BATCHING NOT SET PROPERLY")
            print("#################################################")
Dipam Chakraborty's avatar
Dipam Chakraborty committed
82
83
        replay_shape = (n_pi, nsteps, nenvs)
        self.retune_selector = RetuneSelector(nenvs, observation_space, action_space, replay_shape,
Dipam Chakraborty's avatar
Dipam Chakraborty committed
84
85
                                              skips = self.config['skips'], 
                                              n_pi = n_pi,
Dipam Chakraborty's avatar
Dipam Chakraborty committed
86
87
                                              num_retunes = self.config['num_retunes'],
                                              flat_buffer = self.config['flattened_buffer'])
88
        self.save_success = 0
Dipam Chakraborty's avatar
Dipam Chakraborty committed
89
90
        self.target_timesteps = 8_000_000
        self.buffer_time = 20 # TODO: Could try to do a median or mean time step check instead
91
        self.max_time = 100000000
Dipam Chakraborty's avatar
Dipam Chakraborty committed
92
93
94
95
96
97
98
99
        self.maxrewep_lenbuf = deque(maxlen=100)
        self.gamma = self.config['gamma']
        self.adaptive_discount_tuner = AdaptiveDiscountTuner(self.gamma, momentum=0.98, eplenmult=3)
        
        self.lr = config['lr']
        self.ent_coef = config['entropy_coeff']
        
        self.last_dones = np.zeros((nw * self.config['num_envs_per_worker'],))
Dipam Chakraborty's avatar
Dipam Chakraborty committed
100
        self.make_distr = dist_build(action_space)
101
        self.retunes_completed = 0
Dipam Chakraborty's avatar
Dipam Chakraborty committed
102
        self.amp_scaler = GradScaler()
Dipam Chakraborty's avatar
Dipam Chakraborty committed
103
        
Dipam Chakraborty's avatar
Dipam Chakraborty committed
104
105
        self.update_lr()
        
Dipam Chakraborty's avatar
Dipam Chakraborty committed
106
107
    def to_tensor(self, arr):
        return torch.from_numpy(arr).to(self.device)
108
109
110
111
    
    @override(TorchPolicy)
    def extra_action_out(self, input_dict, state_batches, model, action_dist):
        return {'values': model._value.tolist()}
Dipam Chakraborty's avatar
Dipam Chakraborty committed
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
        
    @override(TorchPolicy)
    def learn_on_batch(self, samples):
        """Fused compute gradients and apply gradients call.
        Either this or the combination of compute/apply grads must be
        implemented by subclasses.
        Returns:
            grad_info: dictionary of extra metadata from compute_gradients().
        Examples:
            >>> batch = ev.sample()
            >>> ev.learn_on_batch(samples)
        Reference: https://github.com/ray-project/ray/blob/master/rllib/policy/policy.py#L279-L316
        """
        ## Config data values
        nbatch = self.nbatch
        nbatch_train = self.mem_limited_batch_size 
        gamma, lam = self.gamma, self.config['lambda']
        nsteps = self.config['rollout_fragment_length']
        nenvs = nbatch//nsteps
        ts = (nenvs, nsteps)
        mb_dones = unroll(samples['dones'], ts)
        
        ## Reward Normalization - No reward norm works well for many envs
        if self.config['standardize_rewards']:
            mb_origrewards = unroll(samples['rewards'], ts)
            mb_rewards =  np.zeros_like(mb_origrewards)
Dipam Chakraborty's avatar
Dipam Chakraborty committed
138
            mb_rewards[0] = self.rewnorm.normalize(mb_origrewards[0], self.last_dones, self.config["reset_returns"])
Dipam Chakraborty's avatar
Dipam Chakraborty committed
139
140
141
142
143
            for ii in range(1, nsteps):
                mb_rewards[ii] = self.rewnorm.normalize(mb_origrewards[ii], mb_dones[ii-1])
            self.last_dones = mb_dones[-1]
        else:
            mb_rewards = unroll(samples['rewards'], ts)
Dipam Chakraborty's avatar
Dipam Chakraborty committed
144
       
145
        # Weird hack that helps in many envs (Yes keep it after reward normalization)
Dipam Chakraborty's avatar
Dipam Chakraborty committed
146
147
148
        rew_scale = self.config["scale_reward"]
        if rew_scale != 1.0:
            mb_rewards *= rew_scale
Dipam Chakraborty's avatar
Dipam Chakraborty committed
149
150
151
152
153
154
155
156
157
158
159
        
        should_skip_train_step = self.best_reward_model_select(samples)
        if should_skip_train_step:
            self.update_batch_time()
            return {} # Not doing last optimization step - This is intentional due to noisy gradients
          
        obs = samples['obs']

        ## Value prediction
        next_obs = unroll(samples['new_obs'], ts)[-1]
        last_values, _ = self.model.vf_pi(next_obs, ret_numpy=True, no_grad=True, to_torch=True)
160
        values = samples['values']
Dipam Chakraborty's avatar
Dipam Chakraborty committed
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
        
        ## GAE
        mb_values = unroll(values, ts)
        mb_returns = np.zeros_like(mb_rewards)
        mb_advs = np.zeros_like(mb_rewards)
        lastgaelam = 0
        for t in reversed(range(nsteps)):
            if t == nsteps - 1:
                nextvalues = last_values
            else:
                nextvalues = mb_values[t+1]
            nextnonterminal = 1.0 - mb_dones[t]
            delta = mb_rewards[t] + gamma * nextvalues * nextnonterminal - mb_values[t]
            mb_advs[t] = lastgaelam = delta + gamma * lam * nextnonterminal * lastgaelam
        mb_returns = mb_advs + mb_values
        
        ## Data from config
        cliprange, vfcliprange = self.config['clip_param'], self.config['vf_clip_param']
        max_grad_norm = self.config['grad_clip']
        ent_coef, vf_coef = self.ent_coef, self.config['vf_loss_coeff']
        
182
        logp_actions = samples['action_logp'] ## np.isclose seems to be True always, otherwise compute again if needed
Dipam Chakraborty's avatar
Dipam Chakraborty committed
183
184
185
186
        noptepochs = self.config['num_sgd_iter']
        actions = samples['actions']
        returns = roll(mb_returns)
        
187
188
189
        advs = returns - values
        normalized_advs = (advs - np.mean(advs)) / (np.std(advs) + 1e-8) 
        
Dipam Chakraborty's avatar
Dipam Chakraborty committed
190
191
192
193
194
195
196
197
        ## Train multiple epochs
        optim_count = 0
        inds = np.arange(nbatch)
        for _ in range(noptepochs):
            np.random.shuffle(inds)
            for start in range(0, nbatch, nbatch_train):
                end = start + nbatch_train
                mbinds = inds[start:end]
198
                slices = (self.to_tensor(arr[mbinds]) for arr in (obs, returns, actions, values, logp_actions, normalized_advs))
Dipam Chakraborty's avatar
Dipam Chakraborty committed
199
200
201
                optim_count += 1
                apply_grad = (optim_count % self.accumulate_train_batches) == 0
                self._batch_train(apply_grad, self.accumulate_train_batches,
Dipam Chakraborty's avatar
Dipam Chakraborty committed
202
                                  cliprange, vfcliprange, max_grad_norm, ent_coef, vf_coef, *slices)
Dipam Chakraborty's avatar
Dipam Chakraborty committed
203

Dipam Chakraborty's avatar
Dipam Chakraborty committed
204
        ## Distill with aux head
Dipam Chakraborty's avatar
Dipam Chakraborty committed
205
        should_retune = self.retune_selector.update(unroll(obs, ts), mb_returns)
Dipam Chakraborty's avatar
Dipam Chakraborty committed
206
207
208
        if should_retune:
            self.aux_train()
        
Dipam Chakraborty's avatar
Dipam Chakraborty committed
209
210
211
212
213
214
215
216
217
218
219
220
        self.update_gamma(samples)
        self.update_lr()
        self.update_ent_coef()
            
        self.update_batch_time()
        return {}
    
    def update_batch_time(self):
        self.time_elapsed += time.time() - self.batch_end_time
        self.batch_end_time = time.time()
        
    def _batch_train(self, apply_grad, num_accumulate, 
Dipam Chakraborty's avatar
Dipam Chakraborty committed
221
                     cliprange, vfcliprange, max_grad_norm,
Dipam Chakraborty's avatar
Dipam Chakraborty committed
222
                     ent_coef, vf_coef,
223
                     obs, returns, actions, values, logp_actions_old, advs):
Dipam Chakraborty's avatar
Dipam Chakraborty committed
224
225
        
        vpred, pi_logits = self.model.vf_pi(obs, ret_numpy=False, no_grad=False, to_torch=False)
226
        pd = self.make_distr(pi_logits)
Dipam Chakraborty's avatar
Dipam Chakraborty committed
227
        logp_actions = pd.log_prob(actions[...,None]).squeeze(1)
228
        entropy = torch.mean(pd.entropy())
Dipam Chakraborty's avatar
Dipam Chakraborty committed
229

Dipam Chakraborty's avatar
Dipam Chakraborty committed
230
        vf_loss = .5 * torch.mean(torch.pow((vpred - returns), 2)) * vf_coef
Dipam Chakraborty's avatar
Dipam Chakraborty committed
231

232
        ratio = torch.exp(logp_actions - logp_actions_old)
Dipam Chakraborty's avatar
Dipam Chakraborty committed
233
234
235
236
        pg_losses1 = -advs * ratio
        pg_losses2 = -advs * torch.clamp(ratio, 1-cliprange, 1+cliprange)
        pg_loss = torch.mean(torch.max(pg_losses1, pg_losses2))

Dipam Chakraborty's avatar
Dipam Chakraborty committed
237
        loss = pg_loss - entropy * ent_coef
Dipam Chakraborty's avatar
Dipam Chakraborty committed
238
239
        
        loss = loss / num_accumulate
Dipam Chakraborty's avatar
Dipam Chakraborty committed
240
        vf_loss = vf_loss / num_accumulate
Dipam Chakraborty's avatar
Dipam Chakraborty committed
241
242

        loss.backward()
Dipam Chakraborty's avatar
Dipam Chakraborty committed
243
        vf_loss.backward()
Dipam Chakraborty's avatar
Dipam Chakraborty committed
244
245
246
        if apply_grad:
            self.optimizer.step()
            self.optimizer.zero_grad()
Dipam Chakraborty's avatar
Dipam Chakraborty committed
247
248
249
            if not self.config['single_optimizer']:
                self.value_optimizer.step()
                self.value_optimizer.zero_grad()
Dipam Chakraborty's avatar
Dipam Chakraborty committed
250
251

        
Dipam Chakraborty's avatar
Dipam Chakraborty committed
252
    def aux_train(self):
Dipam Chakraborty's avatar
Dipam Chakraborty committed
253
254
        nbatch_train = self.mem_limited_batch_size 
        retune_epochs = self.config['retune_epochs']
Dipam Chakraborty's avatar
Dipam Chakraborty committed
255
        replay_shape = self.retune_selector.vtarg_replay.shape
Dipam Chakraborty's avatar
Dipam Chakraborty committed
256
        replay_pi = np.empty((*replay_shape, self.retune_selector.ac_space.n), dtype=np.float32)
Dipam Chakraborty's avatar
Dipam Chakraborty committed
257

Dipam Chakraborty's avatar
Dipam Chakraborty committed
258
259
        for nnpi in range(self.retune_selector.n_pi):
            for ne in range(self.retune_selector.nenvs):
Dipam Chakraborty's avatar
Dipam Chakraborty committed
260
                _, replay_pi[nnpi, :, ne] = self.model.vf_pi(self.retune_selector.exp_replay[nnpi, :, ne], 
Dipam Chakraborty's avatar
Dipam Chakraborty committed
261
                                                             ret_numpy=True, no_grad=True, to_torch=True)
Dipam Chakraborty's avatar
Dipam Chakraborty committed
262
        
Dipam Chakraborty's avatar
Dipam Chakraborty committed
263
        # Tune vf and pi heads to older predictions with (augmented?) observations
Dipam Chakraborty's avatar
Dipam Chakraborty committed
264
        for ep in range(retune_epochs):
Dipam Chakraborty's avatar
Dipam Chakraborty committed
265
            for slices in self.retune_selector.make_minibatches(replay_pi):
Dipam Chakraborty's avatar
Dipam Chakraborty committed
266
                self.tune_policy(slices[0], self.to_tensor(slices[1]), self.to_tensor(slices[2]))
267
                
268
        self.retunes_completed += 1
Dipam Chakraborty's avatar
Dipam Chakraborty committed
269
270
        self.retune_selector.retune_done()
 
Dipam Chakraborty's avatar
Dipam Chakraborty committed
271
    def tune_policy(self, obs, target_vf, target_pi):
Dipam Chakraborty's avatar
Dipam Chakraborty committed
272
273
        if self.config['augment_buffer']:
            obs_aug = np.empty(obs.shape, obs.dtype)
Dipam Chakraborty's avatar
Dipam Chakraborty committed
274
            aug_idx = np.random.randint(self.config['augment_randint_num'], size=len(obs))
Dipam Chakraborty's avatar
Dipam Chakraborty committed
275
276
277
278
279
280
            obs_aug[aug_idx == 0] = pad_and_random_crop(obs[aug_idx == 0], 64, 10)
            obs_aug[aug_idx == 1] = random_cutout_color(obs[aug_idx == 1], 10, 30)
            obs_aug[aug_idx >= 2] = obs[aug_idx >= 2]
            obs_in = self.to_tensor(obs_aug)
        else:
            obs_in = self.to_tensor(obs)
281
        
Dipam Chakraborty's avatar
Dipam Chakraborty committed
282
283
284
285
        if not self.config['aux_phase_mixed_precision']:
            loss, vf_loss = self._aux_calc_loss(obs_in, target_vf, target_pi)
            loss.backward()
            vf_loss.backward()
Dipam Chakraborty's avatar
Dipam Chakraborty committed
286
287
288
289
290
            if not self.config['single_optimizer']:
                self.aux_optimizer.step()
                self.value_optimizer.step()
            else:
                self.optimizer.step()
Dipam Chakraborty's avatar
Dipam Chakraborty committed
291
292
293
294
295
296
297
298
            
        else:
            with autocast():
                loss, vf_loss = self._aux_calc_loss(obs_in, target_vf, target_pi)
            
            self.amp_scaler.scale(loss).backward(retain_graph=True)
            self.amp_scaler.scale(vf_loss).backward()

Dipam Chakraborty's avatar
Dipam Chakraborty committed
299
300
301
302
303
            if not self.config['single_optimizer']:
                self.amp_scaler.step(self.aux_optimizer)
                self.amp_scaler.step(self.value_optimizer)
            else:
                self.amp_scaler.step(self.optimizer)
Dipam Chakraborty's avatar
Dipam Chakraborty committed
304
305
306

            self.amp_scaler.update()
            
Dipam Chakraborty's avatar
Dipam Chakraborty committed
307
308
309
310
311
        if not self.config['single_optimizer']:
            self.aux_optimizer.zero_grad()
            self.value_optimizer.zero_grad()
        else:
            self.optimizer.zero_grad()
Dipam Chakraborty's avatar
Dipam Chakraborty committed
312
313
            
    def _aux_calc_loss(self, obs_in, target_vf, target_pi):
Dipam Chakraborty's avatar
Dipam Chakraborty committed
314
        vpred, pi_logits = self.model.vf_pi(obs_in, ret_numpy=False, no_grad=False, to_torch=False)
Dipam Chakraborty's avatar
Dipam Chakraborty committed
315
316
        aux_vpred = self.model.aux_value_function()
        aux_loss = .5 * torch.mean(torch.pow(aux_vpred - target_vf, 2))
Dipam Chakraborty's avatar
Dipam Chakraborty committed
317

318
319
320
        target_pd = self.make_distr(target_pi)
        pd = self.make_distr(pi_logits)
        pi_loss = td.kl_divergence(target_pd, pd).mean()
Dipam Chakraborty's avatar
Dipam Chakraborty committed
321

Dipam Chakraborty's avatar
Dipam Chakraborty committed
322
323
        loss = pi_loss + aux_loss
        vf_loss = .5 * torch.mean(torch.pow(vpred - target_vf, 2))
Dipam Chakraborty's avatar
Dipam Chakraborty committed
324
325
        
        return loss, vf_loss
Dipam Chakraborty's avatar
Dipam Chakraborty committed
326
        
Dipam Chakraborty's avatar
Dipam Chakraborty committed
327
    def best_reward_model_select(self, samples):
328
        self.timesteps_total += len(samples['dones'])
Dipam Chakraborty's avatar
Dipam Chakraborty committed
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
        
        ## Best reward model selection
        eprews = [info['episode']['r'] for info in samples['infos'] if 'episode' in info]
        self.reward_deque.extend(eprews)
        mean_reward = safe_mean(eprews) if len(eprews) >= 100 else safe_mean(self.reward_deque)
        if self.best_reward < mean_reward:
            self.best_reward = mean_reward
            self.best_weights = self.get_weights()["current_weights"]
            self.best_rew_tsteps = self.timesteps_total
           
        if self.timesteps_total > self.target_timesteps or (self.time_elapsed + self.buffer_time) > self.max_time:
            if self.best_weights is not None:
                self.set_model_weights(self.best_weights)
                return True
            
        return False
    
    def update_lr(self):
        if self.config['lr_schedule'] == 'linear':
            self.lr = linear_schedule(initial_val=self.config['lr'],
                                      final_val=self.config['final_lr'],
                                      current_steps=self.timesteps_total,
                                      total_steps=self.target_timesteps)
            
        elif self.config['lr_schedule'] == 'exponential':
            self.lr = 0.997 * self.lr 
Dipam Chakraborty's avatar
Dipam Chakraborty committed
355
356
357
358
359
360
361
362
        
        for g in self.optimizer.param_groups:
            g['lr'] = self.lr
        if self.config['same_lr_everywhere']:
            for g in self.value_optimizer.param_groups:
                g['lr'] = self.lr
            for g in self.aux_optimizer.param_groups:
                g['lr'] = self.lr
Dipam Chakraborty's avatar
Dipam Chakraborty committed
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395

    
    def update_ent_coef(self):
        if self.config['entropy_schedule']:
            self.ent_coef = linear_schedule(initial_val=self.config['entropy_coeff'], 
                                            final_val=self.config['final_entropy_coeff'], 
                                            current_steps=self.timesteps_total, 
                                            total_steps=self.target_timesteps)
    
    def update_gamma(self, samples):
        if self.config['adaptive_gamma']:
            epinfobuf = [info['episode'] for info in samples['infos'] if 'episode' in info]
            self.maxrewep_lenbuf.extend([epinfo['l'] for epinfo in epinfobuf if epinfo['r'] >= self.max_reward])
            sorted_nth = lambda buf, n: np.nan if len(buf) < 100 else sorted(self.maxrewep_lenbuf.copy())[n]
            target_horizon = sorted_nth(self.maxrewep_lenbuf, 80)
            self.gamma = self.adaptive_discount_tuner.update(target_horizon)

        
    def get_custom_state_vars(self):
        return {
            "time_elapsed": self.time_elapsed,
            "timesteps_total": self.timesteps_total,
            "best_weights": self.best_weights,
            "reward_deque": self.reward_deque,
            "batch_end_time": self.batch_end_time,
            "gamma": self.gamma,
            "maxrewep_lenbuf": self.maxrewep_lenbuf,
            "lr": self.lr,
            "ent_coef": self.ent_coef,
            "rewnorm": self.rewnorm,
            "best_rew_tsteps": self.best_rew_tsteps,
            "best_reward": self.best_reward,
            "last_dones": self.last_dones,
396
            "retunes_completed": self.retunes_completed,
Dipam Chakraborty's avatar
Dipam Chakraborty committed
397
398
399
400
401
402
403
404
405
406
        }
    
    def set_custom_state_vars(self, custom_state_vars):
        self.time_elapsed = custom_state_vars["time_elapsed"]
        self.timesteps_total = custom_state_vars["timesteps_total"]
        self.best_weights = custom_state_vars["best_weights"]
        self.reward_deque = custom_state_vars["reward_deque"]
        self.batch_end_time = custom_state_vars["batch_end_time"]
        self.gamma = self.adaptive_discount_tuner.gamma = custom_state_vars["gamma"]
        self.maxrewep_lenbuf = custom_state_vars["maxrewep_lenbuf"]
407
        self.lr = custom_state_vars["lr"]
Dipam Chakraborty's avatar
Dipam Chakraborty committed
408
409
410
411
412
        self.ent_coef = custom_state_vars["ent_coef"]
        self.rewnorm = custom_state_vars["rewnorm"]
        self.best_rew_tsteps = custom_state_vars["best_rew_tsteps"]
        self.best_reward = custom_state_vars["best_reward"]
        self.last_dones = custom_state_vars["last_dones"]
413
        self.retunes_completed = custom_state_vars["retunes_completed"]
Dipam Chakraborty's avatar
Dipam Chakraborty committed
414
415
416
417
418
419
420
421
422
423
424
425
426
427
    
    @override(TorchPolicy)
    def get_weights(self):
        weights = {}
        weights["current_weights"] = {
            k: v.cpu().detach().numpy()
            for k, v in self.model.state_dict().items()
        }
        return weights
    
    @override(TorchPolicy)
    def set_weights(self, weights):
        self.set_model_weights(weights["current_weights"])
        
Dipam Chakraborty's avatar
Dipam Chakraborty committed
428
    def set_optimizer_state(self, optimizer_state, aux_optimizer_state, value_optimizer_state, amp_scaler_state):
Dipam Chakraborty's avatar
Dipam Chakraborty committed
429
430
431
        optimizer_state = convert_to_torch_tensor(optimizer_state, device=self.device)
        self.optimizer.load_state_dict(optimizer_state)
        
Dipam Chakraborty's avatar
Dipam Chakraborty committed
432
433
434
        aux_optimizer_state = convert_to_torch_tensor(aux_optimizer_state, device=self.device)
        self.aux_optimizer.load_state_dict(aux_optimizer_state)
        
Dipam Chakraborty's avatar
Dipam Chakraborty committed
435
436
        value_optimizer_state = convert_to_torch_tensor(value_optimizer_state, device=self.device)
        self.value_optimizer.load_state_dict(value_optimizer_state)
Dipam Chakraborty's avatar
Dipam Chakraborty committed
437
        
Dipam Chakraborty's avatar
Dipam Chakraborty committed
438
439
440
        amp_scaler_state = convert_to_torch_tensor(amp_scaler_state, device=self.device)
        self.amp_scaler.load_state_dict(amp_scaler_state)
        
Dipam Chakraborty's avatar
Dipam Chakraborty committed
441
442
443
    def set_model_weights(self, model_weights):
        model_weights = convert_to_torch_tensor(model_weights, device=self.device)
        self.model.load_state_dict(model_weights)