james-imi

Results 17 comments of james-imi

Using class specific score threshold given during training for inference

@BloodAxe I assume that the `best score threshold` during training are the recommended scores during training, isn't it? Would it be better then if we can use that during `predict`?...

Using class specific score threshold given during training for inference

> . It is optimized neither for latency nor for granular control of postprocessing. Do you have any function or example for using the outputs from ONNX model without postprocesisng?

AttributeError: 'RegSeg48' object has no attribute 'set_dataset_processing_params'

Same as STDC ``` TypeError: cross_entropy_loss(): argument 'input' (position 1) must be Tensor, not tuple ```

Unable to Type or Copy/Paste in Cursor AI Chat

I am having an annoying issue where on Jupyter Notebook, when i copy from the output, and pastes it on the chat sidebar, it starts to copy a section in...

Error with Model Loaded from Checkpoint

@BloodAxe this is the weirdest API i've seen to a predict method. It's not even finetuning friendly. When exporting with `preprocessing=False`, it doesnt just remove the preprocessing, it also makes...

Error with Model Loaded from Checkpoint

> It tries to pull this meta-information from the model checkpoint. With this information missing, predict() cannot work. I hope it's all clear to this point. I'm pretty sure that...

Sizes of tensors must match except in dimension 1. Expected size 46 but got size 45 for tensor number 1 in the list

Also receiving this particular error @glenn-jocher