Ayyuce Demirbas
@ayyuce.bsky.social
Machine Learning Engineer & Published Researcher, BSc in Computer Engineering | Aspiring PhD, GSoC @TensorFlow, Google Dev Expert in ML
I’ve implemented a multi-modal model that integrates histopathological images and gene expression profiles to classify breast cancer subtypes. The model leverages a Graph Mamba encoder for omics data and a Mamba-based image encoder built on a pretrained ConvNeXt model. github.com/ayyucedemirb...
November 12, 2025 at 5:08 PM
I’ve implemented a multi-modal model that integrates histopathological images and gene expression profiles to classify breast cancer subtypes. The model leverages a Graph Mamba encoder for omics data and a Mamba-based image encoder built on a pretrained ConvNeXt model. github.com/ayyucedemirb...
Compiling the Linux kernel, pretraining my image feature extractor using contrastive learning (on a remote GPU via AnyDesk), and working on my paper.
September 20, 2025 at 10:01 PM
Compiling the Linux kernel, pretraining my image feature extractor using contrastive learning (on a remote GPU via AnyDesk), and working on my paper.
Happy Star Trek day! Live long and prosper. 🖖🏻 #StarTrek #StarTrekDay
September 8, 2025 at 5:40 PM
Happy Star Trek day! Live long and prosper. 🖖🏻 #StarTrek #StarTrekDay
Training a multimodal model for 200 epochs. This should be done in about a week. Wish me luck. 🤞🏻
September 1, 2025 at 9:13 PM
Training a multimodal model for 200 epochs. This should be done in about a week. Wish me luck. 🤞🏻
Instead of treating molecular features as nodes, I made patients the nodes. Now similar patients share information through graph connections, learning from each other's outcomes. Check out the implementation: github.com/ayyucedemirb...
August 28, 2025 at 6:44 PM
Instead of treating molecular features as nodes, I made patients the nodes. Now similar patients share information through graph connections, learning from each other's outcomes. Check out the implementation: github.com/ayyucedemirb...
I’ve implemented suffix arrays and suffix trees from scratch to find repeats, perform fast pattern matching, and generate the Burrows-Wheeler Transform. It also creates visualizations of the data structures. Check out the code on GitHub: github.com/ayyucedemirb...
August 27, 2025 at 11:03 PM
I’ve implemented suffix arrays and suffix trees from scratch to find repeats, perform fast pattern matching, and generate the Burrows-Wheeler Transform. It also creates visualizations of the data structures. Check out the code on GitHub: github.com/ayyucedemirb...
August 26, 2025 at 1:46 PM
August 26, 2025 at 1:42 PM
This is how you can create a longest common subsequence graph using dynamic programming.
github.com/ayyucedemirb...
github.com/ayyucedemirb...
August 24, 2025 at 10:03 PM
This is how you can create a longest common subsequence graph using dynamic programming.
github.com/ayyucedemirb...
github.com/ayyucedemirb...
A GNN project to classify BRCA patient survival using multi-omics data (RNA-seq, CNV, mutations, proteins).
github.com/ayyucedemirb...
github.com/ayyucedemirb...
August 8, 2025 at 3:17 PM
A GNN project to classify BRCA patient survival using multi-omics data (RNA-seq, CNV, mutations, proteins).
github.com/ayyucedemirb...
github.com/ayyucedemirb...
Now you can link your ORCID account to your GitHub account.
June 7, 2025 at 4:40 PM
Now you can link your ORCID account to your GitHub account.
Predicting cancer subtypes using multi-omics data with Graph Convolutional Networks. I know the results are not good enough but I’m working on it. 🧬
GitHub repo: github.com/ayyucedemirb...
GitHub repo: github.com/ayyucedemirb...
June 1, 2025 at 8:29 PM
Predicting cancer subtypes using multi-omics data with Graph Convolutional Networks. I know the results are not good enough but I’m working on it. 🧬
GitHub repo: github.com/ayyucedemirb...
GitHub repo: github.com/ayyucedemirb...
May 23, 2025 at 12:03 PM
You can use the method below to compute BLEU‑1 through BLEU‑4, METEOR, ROUGE‑1/2/L, and BERTScore metrics on your validation set at the end of each epoch.
May 15, 2025 at 5:47 PM
You can use the method below to compute BLEU‑1 through BLEU‑4, METEOR, ROUGE‑1/2/L, and BERTScore metrics on your validation set at the end of each epoch.
If you are getting a warning that says, "A decoder-only architecture is being used, but right-padding was detected! For correct generation results, please set padding_side='left' when initializing the tokenizer." you should add this to your code: processor.tokenizer.padding_side = 'left'.
May 15, 2025 at 5:47 PM
If you are getting a warning that says, "A decoder-only architecture is being used, but right-padding was detected! For correct generation results, please set padding_side='left' when initializing the tokenizer." you should add this to your code: processor.tokenizer.padding_side = 'left'.
A long time ago in a galaxy far, far away…
May the 4th be with you. ✨⚔️
May the 4th be with you. ✨⚔️
May 3, 2025 at 10:07 PM
A long time ago in a galaxy far, far away…
May the 4th be with you. ✨⚔️
May the 4th be with you. ✨⚔️
Wow, that’s some amazing progress. An epoch takes an hour and the validation loss reduces by only 0.00022.
May 1, 2025 at 3:41 PM
Wow, that’s some amazing progress. An epoch takes an hour and the validation loss reduces by only 0.00022.
I’m trying to fine-tune the blip-vqa-base model that consists of 385M parameters on my visual instruction tuning dataset named vqa-rad-instructions. Since the dataset contains only 1793 samples for training, I’m using k-fold to prevent overfitting.
April 12, 2025 at 12:59 PM
I’m trying to fine-tune the blip-vqa-base model that consists of 385M parameters on my visual instruction tuning dataset named vqa-rad-instructions. Since the dataset contains only 1793 samples for training, I’m using k-fold to prevent overfitting.
April 11, 2025 at 11:00 PM
Now you can visualize folder structures with:
- JSON/TXT exports (--output) for automation
- Custom truncation (--truncate N) to control file listings
Try it:
pip install --upgrade PeekDir
PeekDir your_folder --output structure.json --format json --truncate 8
- JSON/TXT exports (--output) for automation
- Custom truncation (--truncate N) to control file listings
Try it:
pip install --upgrade PeekDir
PeekDir your_folder --output structure.json --format json --truncate 8
April 7, 2025 at 2:29 PM
Now you can visualize folder structures with:
- JSON/TXT exports (--output) for automation
- Custom truncation (--truncate N) to control file listings
Try it:
pip install --upgrade PeekDir
PeekDir your_folder --output structure.json --format json --truncate 8
- JSON/TXT exports (--output) for automation
- Custom truncation (--truncate N) to control file listings
Try it:
pip install --upgrade PeekDir
PeekDir your_folder --output structure.json --format json --truncate 8
Just released PeekDir, a Python tool to visualize directory structures. It's useful for understanding the folder structures of complex datasets.
Installation: pip install PeekDir
Usage: PeekDir [directory]
Example usage and output below! 👇
Installation: pip install PeekDir
Usage: PeekDir [directory]
Example usage and output below! 👇
April 5, 2025 at 6:52 PM
Just released PeekDir, a Python tool to visualize directory structures. It's useful for understanding the folder structures of complex datasets.
Installation: pip install PeekDir
Usage: PeekDir [directory]
Example usage and output below! 👇
Installation: pip install PeekDir
Usage: PeekDir [directory]
Example usage and output below! 👇
NeoProtein-GPT is a specialized text-generation model for designing novel protein sequences.
GitHub: github.com/ayyucedemirb...
Model: huggingface.co/ayyuce/NeoPr...
Hugging Face Space: huggingface.co/spaces/ayyuc...
GitHub: github.com/ayyucedemirb...
Model: huggingface.co/ayyuce/NeoPr...
Hugging Face Space: huggingface.co/spaces/ayyuc...
March 22, 2025 at 3:48 PM
NeoProtein-GPT is a specialized text-generation model for designing novel protein sequences.
GitHub: github.com/ayyucedemirb...
Model: huggingface.co/ayyuce/NeoPr...
Hugging Face Space: huggingface.co/spaces/ayyuc...
GitHub: github.com/ayyucedemirb...
Model: huggingface.co/ayyuce/NeoPr...
Hugging Face Space: huggingface.co/spaces/ayyuc...
March 14, 2025 at 4:38 PM
