takuseno / d3rlpy

An offline deep reinforcement learning library

https://takuseno.github.io/d3rlpy

takuseno/d3rlpy Issues

[QUESTION] Offline Learning via custom MDPDataset
Updated a month ago
[QUESTION] len(observation_shape) == 1
Closed a month ago3
[BUG] saving and loading model with custom network gives KeyError: 'custom'
Closed a month ago3
ValueError: too many values to unpack (expected 4) when using hopper-medium-v0 environment
Closed 2 months ago6
[QUESTION] multidimensional states and actions
Closed 2 months ago8
[BUG] How to continue training from a save checkpoint
Closed 2 months ago2
d3rlpy install d4rl
Closed 2 months ago4
[Question] Is it possible to only collect image observation?
Closed 2 months ago3
[QUESTION] Continously increasing loss and TD error
Closed 2 months ago7
[Question]Just want to make sure that the "environment" metric collected by the logger of offline RL algorithms is the result used in papers
Closed 2 months ago2
[BUG] DiscreteDecisionTransformer Inference Problem, AttributeError: 'numpy.ndarray' object has no attribute 'length'
Closed 2 months ago4
[REQUEST] save decision transformer models as TorchScript/ONNX
Closed 2 months ago2
[QUESTION] Adding a new algorithm to d3rlpy
Closed 2 months ago3
[REQUEST] NotImplementedError: "save_policy method does not support tuple observation yet."
Closed 2 months ago4
[QUESTION] difference between evaluators and OPE ?
Updated 2 months ago2
[BUG] error when running dt on atari
Closed 3 months ago2
Issue: Creating MDPDataset from CSV file in d3rlpy
Closed 3 months ago2
[REQUEST] Allow dynamic neural architectures to be used
Updated 4 months ago2
[BUG] Issue with CUDA Device-Side Assertion Failure During Training
Closed 4 months ago2
FQE support .d3 file model?
Updated 4 months ago4
[QUESTION] Logging metrics at every step of training
Closed 4 months ago5
[BUG] Enabling "use_batch_norm" in VectorEncoderFactory(..., use_batch_norm=True, ...) leads to error
Closed 4 months ago4
[REQUEST / QUESTION] Changing the network architecture in DQN
Closed 4 months ago2
[Question] Reproducing FQE
Closed 4 months ago6
[REQUEST] Support WandB logging
Closed 5 months ago3
n*n text data input issue
Closed 5 months ago8
[REQUEST] More control over logging options
Closed 5 months ago1
BCQ vs Discrete BCQ
Closed 5 months ago2
[BUG] Actor and critic loss are nan as well as predicted action.
Closed 5 months ago
[REQUEST] Possibility of customizing the last activation function in VectorEncoderFactory
Closed 6 months ago2
Question about .fit() in 2.3.0
Closed 6 months ago5
[QUESTION] Plotting average reward
Closed 6 months ago4
[BUG] Error in DDP training of BCQ model
Closed 6 months ago3
[Question] How can I convert a saved reaply_buffer to MDPDataset
Closed 6 months ago3
OPE documentation example not working
Closed 6 months ago3
[BUG] examples/distributed_offline_training.py fails when runing with GPUs
Closed 6 months ago10
Log model that has both built-in scalers and stochastic action output (mu, log_std)
Closed 6 months ago2
[QUESTION] How to add an action mask to bc algorithm?
Closed 7 months ago2
AttributeError: 'KitchenMicrowaveKettleLightSliderV0' object has no attribute 'wrapped_env'
Closed 7 months ago5
[QUESTION] What's the meaning of action_size?
Closed 7 months ago1
[BUG] Potential bug in the function " detect_action_space(actions: NDArray) -> ActionSpace" in d3rlpy/dataset/utils.py
Closed 7 months ago2
[QUESTION] How to add an action mask to DiscreteCQL algorithm?
Updated 7 months ago2
[QUESTION] How to modify the value of a discrete action into an array?
Closed 7 months ago1
[QUESTION] Breaking changes in Replay Buffer and TransitionMinibatch
Closed 7 months ago6
[REQUEST] Error in first page of documentation
Closed 8 months ago2
[BUG] Regarding the division of dataset for five-fold cross-validation
Closed 8 months ago2
[QUESTION] How to customize neural network?
Closed 8 months ago3
Any reason about depcrecating cython
Closed 8 months ago2
[BUG] Cannot use SAC with temperature 0
Closed 8 months ago2
[BUG] The gym version of D4RL and d3rlpy doesn't seem match.
Closed 8 months ago2