-
### Description of the Issue
Print selected text, but printing without marking required
### Steps to Reproduce the Issue
1. Settings Preferences: ON: Print line numbers, SET: WYSIWIG
2. Use an…
-
Hello, I am reading your code to replicate the experiment. There are some questions about the model in "model.py".
1. in the query() function, "queries" are word lists. So, it can not be input into t…
-
_franzm created the following on May 23:_
The point here is to share `run_ir.py` which will allow to specify the engine (B25, DPR, ColBERT, etc.) and operation (training, indexing, search) by comm…
-
On my first play-through I seem to have hit a dead end.
From the clues I managed to derive a list of names of people I'd like to talk to, as I figure was the intention. But unfortunately it seems the…
-
Hi,
I am wondering Have you ever tried using`bert-base-cased` as the base model? If you tried, does that improve the performance?
By observation, there are a lot of proper nouns in MSMARCO. I am…
-
-
_jbross created the following on May 24:_
1. In https://github.ibm.com/ai-foundation/OneQA/blob/master/oneqa/ir/dense/colbert_top/colbert/indexing/collection_indexer.py#L216, the line
```
sample…
-
Hi again @okhat
Q1: Does Colbert support Persian?
Q2: I tried to train the model using my dataset, but in 5 minutes, the model will be trained. I use a Persian dataset. Do you think this is a reas…
-
_jaydesen created the following on Feb 28:_
_Original issue: https://github.ibm.com/ai-foundation/PrimeQA/issues/49_
-
_jaydesen created the following on Mar 21:_
Confidence Caliberation module needs to be part of PrimeQA release at NAACL.
_Original issue: https://github.ibm.com/ai-foundation/PrimeQA/issues/6…