🤝 Open to Collab
Andrei Ovcharenko
Kreout
AI & ML interests
Java backend development, Spring Boot, Kafka, infrastructure and automation. Interested in local AI tools and practical developer workflows. Open-source developer tools: https://github.com/krotname.
Recent Activity
repliedto jiteshkprasad's post about 6 hours ago
Okay so, we at https://huggingface.co/udgamlabs just released Lekhani, a Hindi transliterator that is 0.87 MB in total. You type Hindi in English letters, like "kaise ho", and it writes कैसे हो.
The whole thing is a 0.81M-parameter model at 4 bits, a 200,000-word list and a 5 KB JavaScript runner with no dependencies. It runs in the browser or in Node, so nothing you type leaves your device.
On the Aksharantar Hindi test it gets 65.9% of words exactly right, 7 points more than IndicXlit's model on its own (58.9%), which has 14x more parameters. IndicXlit's full package is still ahead at 68.7%, but its word list alone takes 358 MB of memory, about 400 times the size of all of Lekhani. I think for a web page or an app, that difference matters a lot.
It reads one word at a time, so "main" and "mein" can come out the same, and a few common words still come out wrong. The card has the full table, the other open models we measured and where it falls short.
It's Apache-2.0 for the model and runner, and CC0 for the word list. If you type Hindi in English letters, try it and tell me which words it gets wrong.
Model: https://huggingface.co/udgamlabs/Lekhani
Try it: https://udgamlabs.com/models/lekhani
Code: https://github.com/udgamlabs/lekhani repliedto Parveshiiii's post 1 day ago
Most AI image detectors work fine in the lab and fail on social media.
The reason is simple. Real photos get compressed, resized and re-uploaded again and again, and detectors trained on clean images read that damage as a sign of fakery. A real photo that has passed through a few platforms often gets flagged as AI-generated.
AIRealNet-2 is built to fix that. It is the upgrade to our original https://huggingface.co/Modotte/AIRealNet, trained on images collected from platforms like Instagram and X, so it has seen the kinds of quality loss and compression real images go through online.
The training set has 98k images, split 50/50 between AI-generated and real, so the model has no class bias to lean on. Personal photos were excluded by design, including AI-edited ones.
The architecture is unchanged: a SwinV2 Tiny backbone with a binary classification head, about 0.2B parameters, working at 256x256 resolution. It stays fast and easy to deploy for research or production.
Model: https://huggingface.co/Modotte/AIRealNet-2 repliedto alphabell-research's post 3 days ago
Ling 3.1 Flash 1019: the public release diff is sharper than the 41 to 44 headline.
At pinned official revisions, eight setup, tokenizer and loader files are byte-identical to the earlier chat model. All 54 weight shards keep the same paths and sizes, but every LFS SHA-256 hash changes.
That shows changed shard bytes, not a training ablation. For agent benchmarks, use the earlier chat checkpoint as the control. The new base checkpoint has a different chat template and EOS ID.
Model: https://huggingface.co/inclusionAI/Ling-3.1-flash-1019
Pinned sources and a read-only reproducer: https://huggingface.co/blog/alphabell/ling-3-1-1019-public-interface-diffOrganizations
