diff --git a/_modules/ding/worker/collector/base_serial_evaluator.html b/_modules/ding/worker/collector/base_serial_evaluator.html index 3bae17fe1b7d030f59d3be9768f9a93aa0cbb726..e08e3bfdafb10dd42a2a4cab9d4d9f25fc134b75 100644 --- a/_modules/ding/worker/collector/base_serial_evaluator.html +++ b/_modules/ding/worker/collector/base_serial_evaluator.html @@ -316,6 +316,8 @@ Determine whether you need to start the evaluation mode, if the number of training has reached\ the maximum number of times to start the evaluator, return True """ + if train_iter == self._last_eval_iter: + return False if (train_iter - self._last_eval_iter) < self._cfg.eval_freq and train_iter != 0: return False self._last_eval_iter = train_iter diff --git a/_sources/feature/policy_overview_en.rst.txt b/_sources/feature/policy_overview_en.rst.txt index d4680cf9b5012ce09d7f1f00d95ed7140a1ff214..1df1e637faa1fe68acf230678cc2cd614e1b8905 100644 --- a/_sources/feature/policy_overview_en.rst.txt +++ b/_sources/feature/policy_overview_en.rst.txt @@ -58,7 +58,7 @@ Policy Interfaces * ``_forward_learn`` : Forward method for learn mode. - * ``_reset_learn`` : Reset learn mode related varaiables if there are any. Do not need to implement this by force. + * ``_reset_learn`` : Reset learn mode related variables if there are any. Do not need to implement this by force. * ``_monitor_vars_learn`` : Variables that are monitored in learner training process. Those variables will be printed to text and tensorboard logger. @@ -70,7 +70,7 @@ Policy Interfaces * ``_forward_collect`` : Forward method for collect mode. - * ``_reset_collect`` : Reset collect mode related varaiables if there are any. Do not need to implement this by force. + * ``_reset_collect`` : Reset collect mode related variables if there are any. Do not need to implement this by force. * ``_process_transition`` : Process env timestep and policy output into a transition. @@ -84,7 +84,7 @@ Policy Interfaces * ``_forward_eval`` : Forward method for eval mode. - * ``_reset_eval`` : Reset eval mode related varaiables if there are any. Do not need to implement this by force. + * ``_reset_eval`` : Reset eval mode related variables if there are any. Do not need to implement this by force. * ``_state_dict_eval`` : Return model's current state dict. diff --git a/feature/policy_overview_en.html b/feature/policy_overview_en.html index 3dd289778c2cd7d23477c20b207d6c28f6b43e4e..eaa0dc803d4fa3784b245cddddb1afa92fd0abc2 100644 --- a/feature/policy_overview_en.html +++ b/feature/policy_overview_en.html @@ -244,7 +244,7 @@