james-imi

Is there a way to have a docker that has languagetool library install that already downloaded the JAR file so you when you call the library it doesnt need to...

Instance Segmentation Format of YOLOv5 versus v8

6

### Search before asking - [X] I have searched the YOLOv5 [issues](https://github.com/ultralytics/yolov5/issues) and [discussions](https://github.com/ultralytics/yolov5/discussions) and found no similar questions. ### Question Can you also use the txt format used for...

Comet ML doesn't log on segment/train.py

2

### Search before asking - [X] I have searched the YOLOv5 [issues](https://github.com/ultralytics/yolov5/issues) and found no similar bug report. ### YOLOv5 Component Training ### Bug Installed comet ml and did export...

Using class specific score threshold given during training for inference

2

### 💡 Your Question So when training, YOLO-NAS gives you `Best_score_threshold_`. Is there a way to use this on `predict` function instead of a generic conf? ### Versions _No response_

Save the quantized YoloNAS model

4

### 💡 Your Question After doing quantization ``` q_util = SelectiveQuantizer( default_quant_modules_calibrator_weights="max", default_quant_modules_calibrator_inputs="histogram", default_per_channel_quant_weights=True, default_learn_amax=False, verbose=True, ) q_util.quantize_module(model) ``` How do I save the quantized model (not as ONNX)? ###...

Bad support on some medical images with non RGB format

4

### 🐛 Describe the bug Using this code with the latest version ``` for image in tqdm(images): images_predictions = model.predict(image, iou=0.5, conf=0.4, class_agnostic_nms=True) class_names = images_predictions.class_names labels = images_predictions.prediction.labels confidence...

### 💡 Your Question Is there a capability to log image predictions during training? Moreover, it seems like using this code snippet only logs the results after the training is...

ONNX model produces worse result than Pytorch counterpart

1

So I have the following for the Pytorch prediction for a finetuned model with only bounding boxes. ## Pytorch Prediction ``` bbox = [1055, 412, 1286, 991] bbox = np.array(bbox)...

Any recommended training hyperparameters for finetuning in 1 GPU?

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

1

### 💡 Your Question Using ``` model_name = Models.REGSEG48 ... image_processor = ComposeProcessing( [ SegmentationResize((768, 1024)), StandardizeImage(max_value=255.0), NormalizeImage(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]), ImagePermute(permutation=(2, 0, 1)), ] ) best_model =...