* Remove all __future__ imports from RLlib.
* Remove (object) again from tf_run_builder.py::TFRunBuilder.
* Fix 2xLINT warnings.
* Fix broken appo_policy import (must be appo_tf_policy)
* Remove future imports from all other ray files (not just RLlib).
* Remove future imports from all other ray files (not just RLlib).
* Remove future import blocks that contain `unicode_literals` as well.
Revert appo_tf_policy.py to appo_policy.py (belongs to another PR).
* Add two empty lines before Schedule class.
* Put back __future__ imports into determine_tests_to_run.py. Fails otherwise on a py2/print related error.
* [Serve] Added deadline awareness
Added deadline awareness while enqueuing a query
Using Blist sorted-list implementation (ascending order) to get queries according to their specified deadlines. [buffer_queues]
Exposed slo_ms via handle/http request
Added slo example
The queries in example will be executed in almost the opposite order of which they are fired
Added slo pytest
Added check for slo_ms to not be negative
Included the changes suggested
* Linting Corrections
* Adding the code changes suggested by format.sh
* Added the suggested changes
Added justification for blist
Added blist in travis/ci/install-dependencies.sh
* Fixed linting issues
* Added blist to ray/doc/requirements-doc.txt
* Cleaner, tabulated progress output.
* Minor HTML changes, trial ID instead of name
* Revert basic variant changes
* Cleanup, address richard's comments, add progress_reporter.py
* Add tabulate dependency
* Added more info to table, auto-hide columns with no data.
* lint
* Address comments
* Replace experiment tag w/ trial ID
* Fixed tests.
* Fixed test
* Added requirement
* Fix formatting
* Implement metric interface
* Address comment: made actor_handles a dict
* Fix iteration
* Lint
* Mark lightweight actors as num_cpus=0 to prevent resource starvation
* Be more explicit about the readiness condition
* Make task_runner non-blocking
* Lint
* Implement flask_request and named python request
* Forgot to include missing files
* Address comment
* Add flask to requirements for doc (lint failed)
* Update doc requirement so lint will build
* Install flask in CI
* Fix typo in .travis.yml
* Commit and format files
* address stylistic concerns
* Replcae "Usage" by "Example" in doc
* Rename srv to serve
* Add serve to CI process; Fix 3.5 compat
* Improve determine_tests_to_run.py
* Quick cosmetic for determien_tests
* Address comments
* Address comments
* Address comment
* Fix typos and grammar
Co-Authored-By: Edward Oakes <ed.nmi.oakes@gmail.com>
* Update python/ray/experimental/serve/global_state.py
Co-Authored-By: Edward Oakes <ed.nmi.oakes@gmail.com>
* Use __init__ for Query and WorkIntent class
* Remove dataclasses dependency
* Rename oid to object_id for clarity
* Rename produce->enqueue_request, consume->dequeue_request
* Address last round of comment
There are several reasons for this:
* We no longer support python2
* There should be only 1 way of installing Modin
* Issue management on these wheels
* I have never heard of anyone using this feature
* It is rarely kept up to date
* Modin depends on specific versions of Ray because of past API changes