(e.g. for links and images), because some of these examples are now being rendered in the docs.

Added by @fchollet in requests for contributions.

It appears that the docs for Logistic Regression differ based on solvers and penalties. The "penalty" parameter states that "The ‘newton-cg’, ‘sag’ and ‘lbfgs’ solvers support only l2 penalties," while the "solver" parameter states that "‘newton-cg’, ‘lbfgs’, ‘sag’ and ‘saga’ handle L2 or no penalty" (attaching some screenshots). This was actually a little unclear to me, as I wasn't sure if the n

Screenshot

Without Maximize

With Maximize

Without Maximize (bar graph)

![max3](https://user-images.githubusercontent.com/80536487/110967599-942b9680-837c-11eb-9774

Hi, is there an easy way to get the memory occupied by some object ref?
E.g. ray.sizeof(ray.put(object))

Travis is not going to automatically offer the free tier for all open source projects; We likely want o migrate away from travis.

Setting up github actions to replace travis would be a welcomed contribution.

In recent versions (can't say from exactly when), there seems to be an off-by-one error in dcc.DatePickerRange. I set max_date_allowed = datetime.today().date(), but in the calendar, yesterday is the maximum date allowed. I see it in my apps, and it is also present in the first example on the DatePickerRange documentation page.

E

NOTE: we'll need to verify that this is indeed what we want the behavior to be before doing any work on this.

Consider the following sequence of actions a user might take

1. Populate a config.toml file by running streamlit config show > ~/.streamlit/config.toml
2. In the file, change the server.port config option to something other than 8501 (the default), let's say it's changed to 8502

Currently, we rely on AllGatherGrad to compute gather for GPUs.

TODO:

• [] Extend this class to support TPU
• [] Add tests

Not a high-priority at all, but it'd be more sensible for such a tutorial/testing utility corpus to be implemented elsewhere - maybe under /test/ or some other data- or doc- related module – rather than in gensim.models.word2vec.

Originally posted by @gojomo in RaRe-Technologies/gensim#2939 (comment)

While setting train_parameters to False very often we also may consider disabling dropout/batchnorm, in other words, to run the pretrained model in eval mode.
We've done a little modification to PretrainedTransformerEmbedder that allows providing whether the token embedder should be forced to eval mode during the training phase.

Do you this feature might be handy? Should I open a PR?

I'm using mxnet to do some work, but there is nothing when I search the mxnet trial and example.

Current pytorch implementation ignores the argument split_f in the function train_batch_ch13 as shown below.

def train_batch_ch13(net, X, y, loss, trainer, devices):
    if isinstance(X, list):
        # Required for BERT Fine-tuning (to be covered later)
        X = [x.to(devices[0]) for x in X]
    else:
        X = X.to(devices[0])
...

Todo: Define the argument `