-
@natureyoo Hi, thanks for sharing the code, a wonderful work! we have tried replicate the performance by running the coco_adapt.sh, however, the performance with continual setting `R50_ours_continual_…
-
Hugging Face `transformers` is moving to use the `DynamicCache` class as part of the model inputs for the kv cache values. Currently `torch.export` will complain that it is not a tensor. So all models…
-
There are a few commonly used transformer-based optimizations using custom-ops in onnxruntime some of which are listed.
**Custom ops to be converted to functions for ONNX Script** (will add more cu…
-
### System Info
```shell
optimum 1.21.4
optimum-habana 1.14.0.dev0
transformers 4.45.2
+------------------------------------------------------------------…
-
https://huggingface.co/happyme531/Stable-Diffusion-1.5-LCM-ONNX-RKNN2#%E5%B7%B2%E7%9F%A5%E9%97%AE%E9%A2%98
>1. 截至目前，使用最新版本的rknn-toolkit2 2.2.0版本转换的模型仍然存在极其严重的精度损失！即使使用的是fp16数据类型。如图，上方是使用onnx模型推理的结果…
-
https://recap.uni-trier.de/static/377a488cc4cee95714b3ad713aa22fa7/88.pdf
-
## ❓ Questions and Help
How do you run models that are offloaded to the CPU, Trying to work with ```enable_sequential_cpu_offload``` or ```enable_model_cpu_offload```, when running ```torch_xla.sy…
-
### Model Name
Pharmacophore-guided molecular generation
### Model Description
Based on a molecule's pharmacophore, this model generates new molecules de-novo to match the pharmacophore. A graph ne…
-
I was trying to implement a Target based encoder for which I was facing difficulty while implementing a converter
class (CategoricalTransformerOnnx(BaseEstimator, util.TransformerWithTargetMixin…
-
### Describe the issue
Session Creation time takes up to 180 seconds on a Intel(R) Iris(R) Xe Graphics, but it takes only 600 ms on a NVIDIA RTX 4050 on a Windows Runtime solution composed of a "C++ …