-
I compared the older map we had been using with the maps from Tanner.
Have a look and get with @tannerzhang to reconcile. This could be bad mapping on my part. It could be unused mappings, though I …
-
When we import new versions of the UMLS Metathesaurus, we set the version numbers for the new submissions to match the Metathesaurus insertion version, e.g. 2023AB, 2024AA, etc. See the MedDRA summary…
-
The dataset types in the gEAR portal are nice and tidy:
```
mysql> select dtype, count(dtype) from dataset group by dtype;
+-----------------------+--------------+
| dtype | coun…
-
Several thesauri (GND, STW, FIV...) include an additional classification to group its concepts. By now Cocoda can only show either the classification or the thesaurus, unless both are combined into on…
-
```2019-12-06 00:18:07] Using single-device training
[2019-12-06 00:18:07] [data] Loading vocabulary from JSON/Yaml file /191206/source_vocab.yml
[2019-12-06 00:18:08] [data] Setting vocabulary size…
sdlmw updated
2 years ago
-
In some cases it is useful to bias a model's output to a large set of tokens (i.e. all tokens that are just ASCII). Currently you'd have to filter all tokens and create a `Vec` containing a potentiall…
-
I'm starting to understand that this repository is about process, how we encourage or steer desired behaviors that result in community success. One thing I noticed yesterday (my time) was that no one …
-
Now we have shown that LLettuce can work for drug concepts, we need to expand to non-drug concepts.
This could be as simple as removing references to "domain = 'Drug'" from queries. However, the OMOP…
-
loop_closure.cpp DBoW3::Vocabulary vocab("./vocab_larger.yml.gz");使用的是文件夹中自带的vocab_larger字典，没有用gen_vocab_large.cpp。结果很不准确，以第十张图为例：
searching for image 9 returns 4 results:
明显跟第一张图没有相似性啊，请问是怎…
-
Hi @danieldeutsch,
Thank you for building this very useful library.
I wanted to point out a small bug that I found while using the QAEval metric. I think the URL for the question generation model …