-
## Issue
Now ONE runtime has multiple backends that run on cpu (`cpu`, `ruy`, and `xnnpack`). Each backend has different characteristics, and some backend may work better than others on specific la…
-
If you have a fixed tensor X and fix some indices, then contracting this tensor with another tensor along the given indices is a linear map on tensors (of each fixed size). It would be nice if this pa…
-
Hello,
I have used tensor-cell2cell but currently using a dataset that doesn't have a clear context/condition to set, hence I have reverted to the previous cell2cell. I wanted to clarify whether ce…
-
### Issue Description
Many normalizing flows require the use of transpositions during the flow, allowing different transformations to operate on different dimensions. Since transpositions don't chang…
-
### Description of the bug:
Hi,
I am trying to covert an PyTorch to TFLite which uses `torch.argmax(..).indicies` and `torch.gather(..`) - hence creating `LongTensor`s (Int64). As my targeted r…
-
This issue occurs in the llama2 fp16 and int4 weights models, as well as a trimmed model that returns after the first GQA node.
-
Machine learning libraries have found a nice API for tensor contractions, which is based on the Einstein notation. Examples are [PyTorch](https://pytorch.org/docs/stable/torch.html?highlight=einsum#to…
-
I run gptq using auto-gptq, and it calls exllamaV2 kernel.
I find it seems to change the layout of q_weight, but I checked the map and map_iv, the are just mapped to themselves in GPTQ.
And aft…
-
Hi,
I have been trying to install this for the past couple of days but I am always running into errors. It would be of great help if you can provide some guidance on this. The following is the conf…
-
Hi,
Could you release the **_code (implented loss funcions, specific optimizers, other parameters, etc.)_** that you folks used **_for training the models_**?
Thanks!