-
why my task.target_dictionary.indices have a space !!!!
{'': 0, '': 1, '': 2, '': 3, '| ': 4, 'E ': 5, 'A ': 6, 'T ': 7, 'R ': 8, 'O ': 9, 'S ': 10, 'I ': 11, 'N ': 12, 'H ': 13, 'L ': 14, 'D ': 15, …
-
Sorry for asking problem irrelevant to the dataset itself.
But when I'm evaluating Adj-Relation of my results in this dataset, the evaluating code from ICDAR-2019 faced serious performance problem. I…
-
```py
In [1]: from sympy import __version__
In [2]: __version__
Out[2]: '1.8.dev'
In [3]: t = sqrt(3) + 1
In [4]: r = (1 / t).simplify()
In [5]: s = t - 1 / r
In [6]: s
Out[6]:
…
-
I'm evaluating Binary Ninja 4.1 Free on some older DirectShow code. You already have GUIDs and type definitions for `ICreateDevEnum`, `IGraphBuilder`, `IMediaControl`, `IBaseFilter`, `ICaptureGraphBui…
-
**Issue by [W4ngatang](https://github.com/W4ngatang)**
_Wednesday Jul 18, 2018 at 03:15 GMT_
_Originally opened as https://github.com/nyu-mll/jiant/issues/185_
----
Before validating or evaluation,…
-
**Issue by [sleepinyourhat](https://github.com/sleepinyourhat)**
_Monday Jul 23, 2018 at 19:39 GMT_
_Originally opened as https://github.com/nyu-mll/jiant/issues/224_
----
eval_tasks = mnli-diagnos…
-
-
-
Evaluating `(ns ...)` from a buffer (i.e. using ctrl+,,s) and directly in the REPL yields different results. Take following example:
1. Current namespace is `user`, REPL prompt is `user >`
2. Evaluate…
roti updated
11 years ago
-
As discussed last class, can anyone elaborate on the methods we can use to evaluate a clustering algorithm to see if its is better than mere "chance" (i.e. null distribution)?