-
Everyone wants it to be a thing. LET'S MAKE IT A THING.
-
As mentioned in #1 the design of a benchmarking tool should be considered.
-
-
Can you write a benchmark page, that makes 12.500 classes, plain DOM of 10.000 elements with random classes, makes 1.000.000 addClass operations, 700.000 removeClass operations, 1.500.000 containClass…
-
When harness/run_setup.py fails for a task we get a crash rather than a helpful message.
Also I think it would be better if tasks that fail setup are skipped rather than causing the whole run to f…
-
-
### Environment information
```bash
CLI:
Version: 1.8.3
Color support: true
Platform:
CPU Architecture: aarch64
OS: …
-
Very sorry for the late reply. Regarding evaluation, you can see the latest version of our paper: assets/06203.pdf. We evaluate our model in Our proposed [Scenario Understanding Benchmar…
-
Thanks for your great work! I saw that you used vbench as a benchmark in your paper, but the results do not match human perception. Maybe you can try using `ChronoMagic-Bench`, which `CogVideoX` also …
-
Feature suggestion:
Sell units that haven't got any of the preconfigured traits. That way the trait level can be higher and the bench is less likely to be full.