-
After working through the Jupyter notebooks, my local git directory listed a large number of files as having been modified:
```git
modified: tutorials/experiment_history/21-05-22_17.05.05/output…
-
The output of get_unusualness_bayesLCA.R is a table with 7 rows per language. For the predict Surprisal, I'd like to take the mean of the Surprisal per language. That okay @blasid ?
-
Hi Jason! I was looking into some results from this package and I found something weird in the rank table output of the third VADIS line. It seems to be ranking the wrong dimensions (or I totally miss…
-
We should shift the unusualness prediction analysis from BRMS to INLA, to simplify and streamline. It's a continuous variable, so it shouldn't be too tricky. The relevant script is this:
[https://g…
-
- Get surprisal metrics for different models and across transcript vs. force-aligned (= no punctuation)
- Look at correlations between models
- Qualitative inspection of examples
- For now, only fo…
rbroc updated
2 years ago
-
There's a draft of implementation for a GPT2 forward language modeling extractor in pliers, which we can use in combo with `MetricExtractor` (as we did for BERT) to get entropy and surprisal estimates…
rbroc updated
2 years ago
-
On my to do-list is to re-write the make precision matrices r script so that it also can make the approrpriate precision matrices for what @blasid wants to do with unusualness and BRMS. Current plan i…
-
微博内容精选
-
## cli command for score (prioritize this)
task: accept sentence(s) and pretty print score-table (shown as below), with scores for each token.
```
score --scorer incremental --model gpt2 --text…
-
Hi,
since fighting browser fingerprinting is only reliably possible in a crowd of users with the same setup, a community effort is needed to make users more homogeneous.
The Arkenfox wiki in it…