-
Great work on the implementation!
Very comprehensible and straightforward implementation.
It seems you're performing two forward steps: 1) to choose an action (main.py, line 113), 2) to evaluate t…
-
### What is the problem?
When I try and use a recurrent torch model with A2C I get the following list index out of range error due to a state passed to the model's forward function being empty:…
-
### What is the problem?
Installed ray with the nightly wheel.
I wrote a custom env, model, and action distribution.
I attempt to train it with PPO but there is a key error in one of the int…
ludns updated
4 years ago
-
When I try to run A3C with continuous actions and a `MeanStdFilter` observation filter. I get the following error:
Which is surprising because Im not using the `ConcurrentMeanStdFilter`. Does A3C n…
-
In Azure, we should be able to set a node as spot (preemptible) by setting [properties.priority = 'Spot'](https://docs.microsoft.com/en-us/rest/api/compute/virtualmachines/createorupdate#virtualmachin…
-
When I run with Pong-ram-v4 it works but when I run with PongNoFrameskip-v4 as shown below:
>> python/ray/rllib/train.py --run=IMPALA --config-file=my.yaml
Why is the selection of the game cause …
-
Traceback (most recent call last):
File "train.py", line 49, in
from object_detection.builders import dataset_builder
File "/usr/local/lib/python3.6/dist-packages/object_detection/builders…
-
After running training in the MNISTAnomalyExample the evaluation code triggers a `cudaStreamSynchronize` failure.
Running `1.0.0_changes` branch of examples. Ubuntu 18, snapshots, CUDA 10. Compiled l…
-
Hi,
I'm trying to combine the `A3CLSTMGaussian` and `A3CFFSoftmax` examples to an `A3CLSTMSoftmax` architecture. Is the following the right way to go? Would you change something?
BTW, If I mana…
-
Hello.
When I ran "a3c.py" or "a3c_display.py" , I got error log bellow.
********
I tensorflow/stream_executor/dso_loader.cc:135] successfully opened CUDA library libcublas.so.8.0 locally
I ten…