TFException: Shape [-1,24] has negative dimensions

[sterlingcrispin]

I was running a training session in the PPO notebook and stopped it , then exported the graph

INFO:tensorflow:Restoring parameters from ./models/ppo/model-400000.cptk
Converted 4 variables to const ops.
20 ops in the final graph.

but when I use this .bytes file in Unity I get

TFException: Shape [-1,24] has negative dimensions [[Node: epsilon = Placeholder[dtype=DT_FLOAT, shape=[?,24], _device="/job:localhost/replica:0/task:0/cpu:0"]()]] TensorFlow.TFStatus.CheckMaybeRaise (TensorFlow.TFStatus incomingStatus, System.Boolean last) (at <6ed6db22f8874deba74ffe3e566039be>:0) TensorFlow.TFSession.Run (TensorFlow.TFOutput[] inputs, TensorFlow.TFTensor[] inputValues, TensorFlow.TFOutput[] outputs, TensorFlow.TFOperation[] targetOpers, TensorFlow.TFBuffer runMetadata, TensorFlow.TFBuffer runOptions, TensorFlow.TFStatus status) (at <6ed6db22f8874deba74ffe3e566039be>:0) TensorFlow.TFSession+Runner.Run (TensorFlow.TFStatus status) (at <6ed6db22f8874deba74ffe3e566039be>:0) CoreBrainInternal.DecideAction () (at Assets/ML-Agents/Scripts/CoreBrainInternal.cs:244) Brain.DecideAction () (at Assets/ML-Agents/Scripts/Brain.cs:308) Academy.DecideAction () (at Assets/ML-Agents/Scripts/Academy.cs:250) Academy.RunMdp () (a

not sure why this has a -1 value, I've stopped training prematurely on the Ball3D demo without any problems, anyone have any ideas?

[vincentpierre]

Hi, Thank you for you question. It seems to me that you forgot to add a custom placeholder in your Internal Brain. In Embedding the trained model itself it says you need to add a float placeholder called epsilon with values ranging from 0 to 0.
The inspector image here might help.

[awjuliani]

Please let us know if you are able to get the epsilon parameter to work for you to run the agent natively.

[sterlingcrispin]

Ah yep sure enough that's the problem. I remember setting that on the Ball3Ddemo but overlooked it on my own scene, it would be awesome if there was a warning about that built in

[codmw44]

I have similar issues with "Shape [-1,24] has negative dimensions", and after set "graph placeholders" like in Ball3D I get from unity console this error: TFException: Attempting to use uninitialized value dense/kernel
[[Node: dense/kernel/read = IdentityT=DT_FLOAT, _class=["loc:@dense/kernel"], _device="/job:localhost/replica:0/task:0/cpu:0"]]
TensorFlow.TFStatus.CheckMaybeRaise (TensorFlow.TFStatus incomingStatus, System.Boolean last) (at <6ed6db22f8874deba74ffe3e566039be>:0)
TensorFlow.TFSession.Run (TensorFlow.TFOutput[] inputs, TensorFlow.TFTensor[] inputValues, TensorFlow.TFOutput[] outputs, TensorFlow.TFOperation[] targetOpers, TensorFlow.TFBuffer runMetadata, TensorFlow.TFBuffer runOptions, TensorFlow.TFStatus status) (at <6ed6db22f8874deba74ffe3e566039be>:0)
TensorFlow.TFSession+Runner.Run (TensorFlow.TFStatus status) (at <6ed6db22f8874deba74ffe3e566039be>:0)
CoreBrainInternal.DecideAction () (at Assets/ML-Agents/Scripts/CoreBrainInternal.cs:244)
Brain.DecideAction () (at Assets/ML-Agents/Scripts/Brain.cs:308)
Academy.DecideAction () (at Assets/ML-Agents/Scripts/Academy.cs:250)
Academy.RunMdp () (at Assets/ML-Agents/Scripts/Academy.cs:337)
Academy.FixedUpdate () (at Assets/ML-Agents/Scripts/Academy.cs:260)

Maybe I make mistakes in get Graph Model after learning?

[lock]

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.