-
Dear authors,
Thank you for your work. I am experiencing the same issue as mentioned in #7 . Could you please let us know when the details of the training data generation will be released? Addition…
-
“What you can do, however, is you can train the net a bit more using your own data. Just prepare one or two starless images from your data and run training for 20 epochs or so. This should significant…
-
```[tasklist]
### Tasks
- [ ] https://github.com/linea-it/pzserver/issues/101
- [ ] https://github.com/linea-it/pzserver/issues/102
```
-
Has anyone trained with 16K data? Is there a significant difference in performance compared to a model trained with 48K data?
-
Eg. in this case for selects:
```2024-09-03 15:53:10,375 [INFO]
Number of options: 16
Number of samples: 500
Languages
es 281
en 207
pt 12
Options
Medidas Provisionales 255
Supervisión de…
-
When I select "Ratio of training to test data: 90%", most of the data on the plot belongs to training data. I think, this expression is misleading. It would be more appropriate to use the "Ratio of tr…
bha38 updated
3 weeks ago
-
Hi,
Thank you for your nice work, and share the code.
I am confuse what the differenc between "boxes" and "hands_boxes" in 1.json? And could you provide the code for data preprocess for Control…
-
Dear Authors,
Thank you for releasing the evaluation code.
Could you provide details on how to setup the training and evaluation datasets/benchmarks? Including the twitter dataset you mention in…
-
Dear Author, I have read HMR 2.0 and I am very excited by the results. I would like to create something based on your article, but I don't know much about dataset preprocessing, so I'd like to ask you…
-
We may want to mask intermediate dialogue turns, e.g. if they arise due to the model making a mistake. I propose adding this by creating a `train` field that is added to datasets in an instance. An in…