-
### Feature request
The support is [already present in huggingface/transformers](https://github.com/huggingface/transformers/pull/27662).
But when I try to export LLaVA model to neuron format, i…
lifo9 updated
1 month ago
-
```
String response;
//var results = await pool.query("select Ielts.word, Ielts.means from Ielts where id=' "+ j +" ' ");
var results = await pool.query("select Ielts.word, Ielts.means …
-
Could you please update some key codes or configs for helping reproducing the results of TinyImageNet in this paper?
Thanks for your help.
-
Make tests for models that will use appropriate validation data set (e.g., SQUAD, hellaswag) to run model and compare output to expected outputs from data set.
This tests should be organized in a simi…
-
**Different trained TrOCR models require custom arg changes in the script**
Model I am using TrOCR:
The problem arises when using:
* [x ] the official example scripts: (give details below)
Whe…
-
Thank you so much for your great work!
I am so honered to use your method for my custom dataset.
My dataset has only one class and then I should train it. I have followed the instructions as you men…
-
Congrats on your great work!
I am verifying your method on vision tasks and have a small concern on the influence of the "conv_kernel_size" of the 2D group-convolution in your task and I find that …
-
# Issues Discription #
I tried to use the following command to run all model level tests:
```
python run.py -j 16 --report --cachedir cached -v --testsfile models.txt \
--torchmlirbuild /torch…
-
To achieve comparable result as the original paper. LayerNorm is applied to the feature before NF. This is never mentioned in the paper and the usage is very tricky (but this is the only way works for…
-