🚀 Feature

In python x is True and x == True have exactly the same semantics, because True and False are global singletons. Generally, having two different operators that do the same thing in the your IR is not desirable because it means passes have to reason about both cases if they want to optimize the pattern. Additionally, optimizations like Common Subexpression Elimination will no tr

At this moment relu_layer op doesn't allow threshold configuration, and legacy RELU op allows that.
We should add configuration option to relu_layer.

Problem: the approximate method can still be slow for many trees
catboost version: master
Operating System: ubuntu 18.04
CPU: i9
GPU: RTX2080

Would be good to be able to specify how many trees to use for shapley. The model.predict and prediction_type versions allow this. lgbm/xgb allow this.

Our users are often confused by the output from programs such as zip2john sometimes being very large (multi-gigabyte). Maybe we should identify and enhance these programs to output a message to stderr to explain to users that it's normal for the output to be very large - maybe always or maybe only when the output size is above a threshold (e.g., 1 million bytes?)

Hi ,

I have tried out both loss.backward() and model_engine.backward(loss) for my code. There are several subtle differences that I have observed , for one retain_graph = True does not work for model_engine.backward(loss) . This is creating a problem since buffers are not being retained every time I run the code for some reason.

Please look into this if you could.

Today, I can manipulate ListDtype columns and execute operations like segmented sort and unique. Because these operations do not have cross-row (or cross-partition) dependencies, they can be executed in Dask by passing a lambda function to map_partitions.

It would be nice to expose the list accessor on dask-cuDF objects like we do for other accessors. As this is not supported by pandas, per

Current implementation of join can be improved by performing the operation in a single call to the backend kernel instead of multiple calls.

This is a fairly easy kernel and may be a good issue for someone getting to know CUDA/ArrayFire internals. Ping me if you want additional info.

Hi,

We are using pycaret to run a regression model which will also give us feature importance of each of the co-efficients in the model.

When we look at the tuned models feature importance (tuned_model.feature_importances_)
, it is essentially an array of values which we are not sure how it correlates to the independent variables.

![WhatsApp Image 2021-04-30 at 16 54 19](https://user-i

Problem

Cub allows itself to place into a namespace via CUB_NS_PREFIX and CUB_NS_POSTFIX, such that multiple shared libraries can each utilize their own copy of it (and thus different versions can safely coexist). Static variables used for caching could otherwise cause problems (e.g., https://github.com/NVIDIA/cub/blob/main/cub/util_device.cuh#L212).

Thrust however depends on cub and

We would like to forward a particular 'key' column which is part of the features to appear alongside the predictions - this is to be able to identify to which set of features a particular prediction belongs to. Here is an example of predictions output using the tensorflow.contrib.estimator.multi_class_head:

{"classes": ["0", "1", "2", "3", "4", "5", "6", "7", "8", "9"],
 "scores": [0.068196