All Questions
77 questions
0
votes
0
answers
64
views
Memory increasing after hugging face generate method
I wanted to make an inference with codegemma model from huggingface, but when I use model.generate(**inputs) method GPU memory cost increases from 39 GB to 49 GB in peak usage with torch profiler no ...
user17751265
-3
votes
1
answer
34
views
in testing dataset using dataloader , should we set shuffle=true or it doesn't matter?
I have a custom dataset (images of pizza,sushi and steak).
I'm using torch DataLoader for it , now when writing the test dataloader custom should we set shuffle=true or it just doesn't matter??
I ...
0
votes
0
answers
46
views
Pixel mapping to network inputs
I'm working with object detection models in embedded devices and need to run tests comparing performance on the embedded platform vs on a PC. For the integrity of the tests, I need to make sure the ...
0
votes
0
answers
82
views
Can't load my model using load_state_dict: RuntimeError: Error(s) in loading state_dict for UNetGenerator: Unexpected key(s) in state_dict
I am using Kaggle to train my my Pix2pix GAN model
however I had to save every 12 hr. the checkpoints and continue the train again
The problem is when i want to load my model
generator.load_state_dict(...
- 1
0
votes
0
answers
40
views
Reshape RuntimeError: shape '[1, 2]' is invalid for input of size 100
I'm building a simple football scores prediction model in Pytorch.
My steps:
clean and process the data
use team names as inputs and scores as outputs
onehot encode team names
pass two team names to ...
0
votes
1
answer
317
views
Understanding usage of HiFi-GAN by Vits
I'm (trying to) learn AI/ML for speech synthesis and trying to undestand how HiFi-GAN is used by Vits.
From my understanding, Vits will convert text input into mel spectograms which is then converted ...
- 1,246
0
votes
1
answer
339
views
YoloV8 results have no 'box', 'max' properties in it
I've trained a YOLOV8 model to identify objects in an intersection (ie cars, roads etc).
It is working OK and I can get the output as an image with the objects of interested segmented.
However, what I ...
1
vote
1
answer
103
views
Running out of RAM when finetuning model
I am currently trying to finetune Wav2Vec2 model from: https://huggingface.co/dima806/bird_sounds_classification. But my RAM utilisation is running over the free tier on Google Colab.
The following is ...
- 73
0
votes
0
answers
101
views
Correct implementation of F-Beta score as a loss in Pytorch
I'm trying to implement the F-Beta score as a loss. As much as I know, I can't use >= or argmax() functions on the probability as it is not differentiable. Please correct me if I'm wrong.
Closest ...
- 4,232
-2
votes
1
answer
85
views
How to create a tensor based on another one - Studying PyTorch in practice?
I'm studying IA using PyTorch and implementing some toy examples.
First, I created a one-dimensional tensor (X) and a second tensor (y), derived from the first one:
X = torch.arange(0, 100, 1.0)....
0
votes
1
answer
40
views
Problem in defining a ML model for my Npy dataset
I need help in defining a torch model for my data. I have tried various methods but nothing seems to be working out. Error after error related to input size and shaping. How can I resolve these issues?...
2
votes
0
answers
873
views
RuntimeError: Default process group has not been initialized, please make sure to call init_process_group
Issue
I am using CoDi which is multimodal Latent diffusion model. I am trying to remove the modules on images and video from CoDi and fine tune it with text-music pair data.
The training script(train....
- 21
1
vote
0
answers
63
views
SpaCy fine-tuning GPU
I train a text classifier with spaCy and classy classification. But the model does not use gpu during training, fine-tuning is very long
GPU info
$ nvidia-smi
Mon Apr 22 09:41:13 2024
+--------...
- 11
0
votes
1
answer
36
views
What am I doing wrong with my CNN that it is gaining accuracy so slowly
I am using this CNN to detect information in eeg scans. It is gaining accuracy really slowly and I am wondering if I am mmissing anything in any of the layers or am doing anything wrong
class Net(...
0
votes
1
answer
149
views
MLP a2c policy complaining that 0 isn't greater than 0, or infinity isn't greater than 0?
Getting the following error as I'm training some torch models:
ValueError('Expected parameter scale (Tensor of shape (1, 4)) of distribution Normal(loc: torch.Size([1, 4]), scale: torch.Size([1, 4])) ...
- 103