Dheeraj Peri
@arunikayadav42 I don't remember the exact data structure details of SCAN data as it has been a while. The way I create paraphrases (train_enc.txt and train_dec.txt) is [here ](https://github.com/peri044/STT/blob/master/prepare_mscoco_pairs.py). The...
Yes. The image feature (for the image id) is replicated for each of the 20 combinations of the captions. Probably, the data loader script you linked is not the one...
@fortminors The model here is same as the one in this issue : https://github.com/pytorch/TensorRT/issues/3294 I have tried to save the model by re-exporting it using `torch.export.export(model, (inputs,))` (obtained after TRT...
This PR should fix this issue: https://github.com/pytorch/pytorch/pull/142253
Hello @lebionick In the example you've provided, it looks like you're using a torchscript model but you're using our `dynamo` backend. We currently do not support INT8 in dynamo and...
@korkland @MaltoseFlower The error message indicates ```py WARNING:torch_tensorrt [TensorRT Conversion Context]:Calibrator is not being used. Users must provide dynamic range for all tensors that are not Int32 or Bool. ERROR:torch_tensorrt...
Hello @Qi-Zha0 , You should pass torch tensor inputs to the save API. These inputs should be in the range (min_shape, opt_shape, max_shape). eg: `torch_tensorrt.save(trt_gm, "trt_gm.ep", inputs=[torch.randn(4, 3, 224, 224).cuda()])`....
> 1 AI: take > > https://github.com/pytorch/TensorRT/blob/3422c41f165c3cf0833468b0cb548149ca78e057/py/torch_tensorrt/fx/converters/converter_utils.py#L101 > > and copy it into https://github.com/pytorch/TensorRT/blob/3422c41f165c3cf0833468b0cb548149ca78e057/py/torch_tensorrt/dynamo/conversion/converter_utils.py. Modify to use the naming scheme that you have designed to work with refit. > Replace...
- torch_tensorrt.dynamo.cross_compile(model, inputs, *args, **kwargs) -> save EP to disk by calling torch_tensorrt.save()
Options: 1) setup_engine -> execute_engine (Issue here is type is not supported in torch custom ops. For C++ ops, it is working). So POC with custom C++ ops should unblock...