Tested Phi-3 and Phi-3.5 models on Optical Character Recognition task:
- Phi-3 colab: https://colab.research.google.com/drive/1HFvSQ8ktCfkECqqN38Gu7C_zrRN1ldAz?usp=sharing
- Phi-3.5 colab: https://colab.research.google.com/drive/1P7KeiK4RxuhzXiMviPCjbsbhRNSlE3cl?usp=sharing
Phi-3 gives better results - it just returns the text (in first 2 images), Phi-3.5 explains what's on the image :/
#ai #colab
@doing_something
- Phi-3 colab: https://colab.research.google.com/drive/1HFvSQ8ktCfkECqqN38Gu7C_zrRN1ldAz?usp=sharing
- Phi-3.5 colab: https://colab.research.google.com/drive/1P7KeiK4RxuhzXiMviPCjbsbhRNSlE3cl?usp=sharing
Phi-3 gives better results - it just returns the text (in first 2 images), Phi-3.5 explains what's on the image :/
#ai #colab
@doing_something
doing something
Tested Phi-3 and Phi-3.5 models on Optical Character Recognition task: - Phi-3 colab: https://colab.research.google.com/drive/1HFvSQ8ktCfkECqqN38Gu7C_zrRN1ldAz?usp=sharing - Phi-3.5 colab: https://colab.research.google.com/drive/1P7KeiK4RxuhzXiMviPCjbsbh…
OK, I forgot it's an instructed LLM so just ask it to do what we need...
Good resources about Rust
- https://github.com/ImplFerris/LearnRust
- https://github.com/ImplFerris/rust-in-production
#rust
- https://github.com/ImplFerris/LearnRust
- https://github.com/ImplFerris/rust-in-production
#rust
🤮1
Learned that AdBlocker in my browser Brave is based on Rust
https://github.com/brave/adblock-rust
#rust
https://github.com/brave/adblock-rust
#rust
💩1
Be better in CV with this:
https://www.linkedin.com/feed/update/urn:li:activity:7232348672758358018/
#cv
https://www.linkedin.com/feed/update/urn:li:activity:7232348672758358018/
#cv
Linkedin
How to use Excel files for Computer Vision profiles | Yehor Smoliakov posted on the topic | LinkedIn
I've finished combing through Excel files with profiles of people interested in Computer Vision.
The idea is to gather in one place links to those working in Computer Vision with statistics on their commits (number of repositories/notebooks, published models/datasets…
The idea is to gather in one place links to those working in Computer Vision with statistics on their commits (number of repositories/notebooks, published models/datasets…
Published my go-supervisord fork with many removals and improvements:
https://github.com/egorsmkv/go-supervisord
An example is here.
#go
https://github.com/egorsmkv/go-supervisord
An example is here.
#go
GitHub
GitHub - egorsmkv/go-supervisord: Fork of https://github.com/ochinchina/supervisord with simplification and improvements
Fork of https://github.com/ochinchina/supervisord with simplification and improvements - egorsmkv/go-supervisord
Aesthetic of the code in notebooks should be:
A forum topic on the issue:
https://forums.fast.ai/t/how-to-do-code-formatting-in-jupyter-notebooks/96170
#ai #python
A forum topic on the issue:
https://forums.fast.ai/t/how-to-do-code-formatting-in-jupyter-notebooks/96170
#ai #python
fast.ai Course Forums
How to do code formatting in jupyter notebooks?
I am used to formatting my code using black and isort with normal python scripts. I have been slowly using jupyter notebooks as a development tool. I am curious to know how can I format using black in jupyter notebooks? On a quick google search, I found…
doing something
Aesthetic of the code in notebooks should be: A forum topic on the issue: https://forums.fast.ai/t/how-to-do-code-formatting-in-jupyter-notebooks/96170 #ai #python
ruff supports code formatting in cells for Jupyter Notebooks.- before: https://colab.research.google.com/drive/1pWgjVlkGpKnBUzfSFBnbOl9YWWk8kui5?usp=sharing
- after: https://drive.google.com/file/d/1IXq1c2zpRBoBXRXGEeW6Z74dvsyQSqP9/view?usp=sharing
#rust #ai
This media is not supported in your browser
VIEW IN TELEGRAM
Really useful tool to debug CV models and possible to make demos
https://github.com/rerun-io/rerun
#cv #rust #ai
https://github.com/rerun-io/rerun
#cv #rust #ai
doing something
Really useful tool to debug CV models and possible to make demos https://github.com/rerun-io/rerun #cv #rust #ai
This media is not supported in your browser
VIEW IN TELEGRAM
My code to reproduce the demo but with supervision package:
https://github.com/egorsmkv/rerun-yolov10
#rust #ai #cv
https://github.com/egorsmkv/rerun-yolov10
#rust #ai #cv
doing something
Convenient project to get a foreground mask from images https://github.com/by321/me2net #cv #ai
IS-NET is better in solving of this task
- Colab to test it: https://colab.research.google.com/drive/1ACv8XtterroNOgdvSWUwRftwnyFJUrQw?usp=sharing
#cv #ai
- Colab to test it: https://colab.research.google.com/drive/1ACv8XtterroNOgdvSWUwRftwnyFJUrQw?usp=sharing
#cv #ai
A high-level package to do segmentation
https://github.com/qubvel-org/segmentation_models.pytorch
#cv #ai
https://github.com/qubvel-org/segmentation_models.pytorch
#cv #ai
GitHub
GitHub - qubvel-org/segmentation_models.pytorch: Semantic segmentation models with 500+ pretrained convolutional and transformer…
Semantic segmentation models with 500+ pretrained convolutional and transformer-based backbones. - qubvel-org/segmentation_models.pytorch
doing something
IS-NET is better in solving of this task - Colab to test it: https://colab.research.google.com/drive/1ACv8XtterroNOgdvSWUwRftwnyFJUrQw?usp=sharing #cv #ai
Recently published BiRefNet models have an ability to extract foreground masks
Colab: https://colab.research.google.com/drive/1X3mkiGMCeoePLlEe2522XTPaL-8iDDgv?usp=sharing
#cv #ai
Colab: https://colab.research.google.com/drive/1X3mkiGMCeoePLlEe2522XTPaL-8iDDgv?usp=sharing
#cv #ai