YOLO installation is failing

Discussion
1

I’m trying to install YOLO package and It’s failing.

I’m tried three different commands:

pip install ultralytics

pip install -U ultralytics

pip install git+https://github.com/ultralytics/ultralytics.git@main

All three fail at “cp39-manylinux_2_28_x86_64.”

Do you have any ideas on what might be causing this or how to fix it?

Thanks,

2

What’s the full log?

3

[ec2-user@ip-172-31-64-237 ~] sudo pip install ultralytics
Collecting ultralytics
Using cached ultralytics-8.3.230-py3-none-any.whl (1.1 MB)
Collecting torch>=1.8.0
Downloading torch-2.8.0-cp39-cp39-manylinux_2_28_x86_64.whl (888.0 MB)
|████████████▌ | 346.4 MB 48 bytes/s eta 130 days, 7:42:58 Killed
[ec2-user@ip-172-31-64-237 ~]

4

You shouldn’t use sudo to install Python packages

How much RAM do you have?

You can try pip install ultralytics --no-cache-dir

5

looks like the download for PyTorch is taking a very long time and might be timing out. Anyway you can connect to a faster network or try downloading when there’s less use of the network by others?

6

Hi, thank for the answer. I added the --no-cache-dir param and set $TMDIR env variable for a bigger filesystem. But I started to receiving the “No space left” message. Do yo know how much space is required?

7

I’m using aws, so network shouldn’t be a problem. May be is a cache space issue as mentioned by Toxite.

8

It depends on whether you are using GPU or not. GPU would require PyTorch with CUDA which is larger than CPU only PyTorch.

Can you provide the output after running this command in terminal: `yolo checks`?

9

Thanks for getting back to me. I don’t have YOLO installed yet because the installation keeps failing. I’ll try freeing up more space and see if that fixes it.

10

Hi;
I added 80 Gb space and I run again pip command:

sudo pip install --prefix /disk1/ultralytics ultralytics --no-cache-dir --cache-dir /disk/tmp &

and I got another interesting error message:

ERROR: THESE PACKAGES DO NOT MATCH THE HASHES FROM THE REQUIREMENTS FILE. If you have updated the package versions, please updat e the hashes. Otherwise, examine the package contents carefully; someone may have tampered with them.
torch>=1.8.0 from https://files.pythonhosted.org/packages/fd/dd/1630cb51b10d3d2e97db95e5a84c32def81fc26b005bce6fc880b0e6db81 /torch-2.8.0-cp39-cp39-manylinux_2_28_x86_64.whl#sha256=06fcee8000e5c62a9f3e52a688b9c5abb7c6228d0e56e3452983416025c41381 (from u ltralytics):
Expected sha256 06fcee8000e5c62a9f3e52a688b9c5abb7c6228d0e56e3452983416025c41381
Got 5596fd98ed1426b0258fa81712aaf6ac0268f5f861911b3f5e9c6b403f5e9e97

I

Please find the complete log below:

[ec2-user@ip-172-31-68-175 disk] sudo pip install --prefix /disk1/ultralytics ultralytics --no-cache-dir --cache-dir /disk/tmp & [1] 172368 [ec2-user@ip-172-31-68-175 disk] Collecting ultralytics
Downloading ultralytics-8.3.232-py3-none-any.whl (1.1 MB)
|████████████████████████████████| 1.1 MB 16.0 MB/s
Collecting pillow>=7.1.2
Downloading pillow-11.3.0-cp39-cp39-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (6.6 MB)
|████████████████████████████████| 6.6 MB 70.1 MB/s
Collecting psutil>=5.8.0
Downloading psutil-7.1.3-cp36-abi3-manylinux2010_x86_64.manylinux_2_12_x86_64.manylinux_2_28_x86_64.whl (263 kB)
|████████████████████████████████| 263 kB 73.6 MB/s
Collecting scipy>=1.4.1
Downloading scipy-1.13.1-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (38.6 MB)
|████████████████████████████████| 38.6 MB 118.0 MB/s
Collecting ultralytics-thop>=2.0.18
Downloading ultralytics_thop-2.0.18-py3-none-any.whl (28 kB)
Collecting opencv-python>=4.6.0
Downloading opencv_python-4.12.0.88-cp37-abi3-manylinux2014_x86_64.manylinux_2_17_x86_64.whl (67.0 MB)
|████████████████████████████████| 67.0 MB 30.6 MB/s
Collecting matplotlib>=3.3.0
Downloading matplotlib-3.9.4-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (8.3 MB)
|████████████████████████████████| 8.3 MB 53.1 MB/s
Requirement already satisfied: requests>=2.23.0 in /usr/lib/python3.9/site-packages (from ultralytics) (2.25.1)
Collecting numpy>=1.23.0
Downloading numpy-2.0.2-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (19.5 MB)
|████████████████████████████████| 19.5 MB 56.5 MB/s
Collecting torchvision>=0.9.0
Downloading torchvision-0.23.0-cp39-cp39-manylinux_2_28_x86_64.whl (8.6 MB)
|████████████████████████████████| 8.6 MB 56.0 MB/s
Collecting polars>=0.20.0
Downloading polars-1.35.2-py3-none-any.whl (783 kB)
|████████████████████████████████| 783 kB 73.1 MB/s
Requirement already satisfied: pyyaml>=5.3.1 in /usr/lib64/python3.9/site-packages (from ultralytics) (5.4.1)
Collecting torch>=1.8.0
Downloading torch-2.8.0-cp39-cp39-manylinux_2_28_x86_64.whl (888.0 MB)
|████████▍ | 231.5 MB 856 bytes/s eta 8 days, 20:56:19
ERROR: THESE PACKAGES DO NOT MATCH THE HASHES FROM THE REQUIREMENTS FILE. If you have updated the package versions, please updat e the hashes. Otherwise, examine the package contents carefully; someone may have tampered with them.
torch>=1.8.0 from https://files.pythonhosted.org/packages/fd/dd/1630cb51b10d3d2e97db95e5a84c32def81fc26b005bce6fc880b0e6db81 /torch-2.8.0-cp39-cp39-manylinux_2_28_x86_64.whl#sha256=06fcee8000e5c62a9f3e52a688b9c5abb7c6228d0e56e3452983416025c41381 (from u ltralytics):
Expected sha256 06fcee8000e5c62a9f3e52a688b9c5abb7c6228d0e56e3452983416025c41381
Got 5596fd98ed1426b0258fa81712aaf6ac0268f5f861911b3f5e9c6b403f5e9e97

Any ideas on how can I fix it?

11

You should probably create an environment with newer Python, like Python 3.12. Choose a Deep Learning AMI with PyTorch for the EC2 instance: