@@ -73,8 +73,8 @@ def create_config(env_id, action_space_size, collector_env_num, evaluator_env_nu
73
73
model_path = None ,
74
74
num_unroll_steps = num_unroll_steps ,
75
75
game_segment_length = 20 ,
76
- # update_per_collect=80,
77
- update_per_collect = 10 , # only for debug
76
+ update_per_collect = 80 ,
77
+ # update_per_collect=10, # only for debug
78
78
replay_ratio = 0.25 ,
79
79
batch_size = batch_size ,
80
80
num_segments = num_segments ,
@@ -96,7 +96,7 @@ def generate_configs(env_id_list, action_space_size, collector_env_num, n_episod
96
96
norm_type , seed , buffer_reanalyze_freq , reanalyze_batch_size , reanalyze_partition ,
97
97
num_segments , total_batch_size ):
98
98
configs = []
99
- exp_name_prefix = f'data_unizero_atari_mt_20250216 /{ len (env_id_list )} games_nlayer8_bs64_brf{ buffer_reanalyze_freq } _seed{ seed } /'
99
+ exp_name_prefix = f'data_unizero_atari_mt_20250217 /{ len (env_id_list )} games_nlayer8_bs64_brf{ buffer_reanalyze_freq } _seed{ seed } _dev-uz-mz /'
100
100
101
101
for task_id , env_id in enumerate (env_id_list ):
102
102
config = create_config (
@@ -164,8 +164,9 @@ def create_env_manager():
164
164
# num_segments = 2
165
165
# n_episode = 2
166
166
# evaluator_env_num = 2
167
- # num_simulations = 2
168
- # batch_size = [4, 4, 4, 4, 4, 4, 4, 4]
167
+ # num_simulations = 5
168
+ # # batch_size = [4, 4, 4, 4, 4, 4, 4, 4]
169
+ # batch_size = [4, 4,4,4]
169
170
170
171
171
172
for seed in [0 ]:
@@ -175,5 +176,5 @@ def create_env_manager():
175
176
num_segments , total_batch_size )
176
177
177
178
with DDPContext ():
178
- # train_unizero_multitask_segment_ddp(configs, seed=seed, max_env_step=max_env_step)
179
- train_unizero_multitask_segment_ddp (configs [:4 ], seed = seed , max_env_step = max_env_step ) # train on the first four tasks, only for debug
179
+ train_unizero_multitask_segment_ddp (configs , seed = seed , max_env_step = max_env_step )
180
+ # train_unizero_multitask_segment_ddp(configs[:4], seed=seed, max_env_step=max_env_step) # train on the first four tasks, only for debug
0 commit comments