Post
129
Uncensoring Mistral,
give it a try
s3nh/Ministral-3-14B-Instruct-2512-BF16-abliterated
give it a try
s3nh/Ministral-3-14B-Instruct-2512-BF16-abliterated
AI & ML interests
LowRes animated waifus (✿◡‿◡)
Recent Activity
View all activity
Post
279
Existing methods — GPTQ, AWQ, llama.cpp's k-quants — minimize empirical loss heuristically. None of them prove they are optimal in any information-theoretic sense. ICRB-Q builds a quantization scheme that is provably optimal via the Cramér-Rao lower bound (CRB): no unbiased estimator of a weight can have lower variance than [F(θ)]⁻¹, where F is the Fisher information matrix.
- 1 reply
Parveshiiii
posted an update 3 months ago
Post
625
🚀 Sonic: A lightweight Python audio processing library with tempo matching, BPM detection, time-stretching, resampling & track blending — now with GPU (CUDA) acceleration for 10x speed!
Perfect for quick remixes, batch edits or syncing tracks.
👉 https://github.com/Parveshiiii/Sonic
#Python #AudioProcessing #OpenSource #PyTorch
Perfect for quick remixes, batch edits or syncing tracks.
👉 https://github.com/Parveshiiii/Sonic
#Python #AudioProcessing #OpenSource #PyTorch
Parveshiiii
posted an update 3 months ago
Post
1642
Excited to announce my latest open-source release on Hugging Face: Parveshiiii/breast-cancer-detector.
This model has been trained and validated on external datasets to support medical research workflows. It is designed to provide reproducible benchmarks and serve as a foundation for further exploration in healthcare AI.
Key highlights:
- Built for medical research and diagnostic study contexts
- Validated against external datasets for reliability
- Openly available to empower the community in building stronger, more effective solutions
This release is part of my ongoing effort to make impactful AI research accessible through **Modotte**. A detailed blog post explaining the methodology, dataset handling, and validation process will be published soon.
You can explore the model here: Parveshiiii/breast-cancer-detector
#AI #MedicalResearch #DeepLearning #Healthcare #OpenSource #HuggingFace
This model has been trained and validated on external datasets to support medical research workflows. It is designed to provide reproducible benchmarks and serve as a foundation for further exploration in healthcare AI.
Key highlights:
- Built for medical research and diagnostic study contexts
- Validated against external datasets for reliability
- Openly available to empower the community in building stronger, more effective solutions
This release is part of my ongoing effort to make impactful AI research accessible through **Modotte**. A detailed blog post explaining the methodology, dataset handling, and validation process will be published soon.
You can explore the model here: Parveshiiii/breast-cancer-detector
#AI #MedicalResearch #DeepLearning #Healthcare #OpenSource #HuggingFace
Parveshiiii
posted an update 4 months ago
Post
2964
Just did something I’ve been meaning to try for ages.
In only 3 hours, on 10 billion+ tokens, I trained a custom BPE + tiktoken-style tokenizer using my new library microtok — and it hits the same token efficiency as Qwen3.
Tokenizers have always felt like black magic to me. We drop them into every LLM project, but actually training one from scratch? That always seemed way too complicated.
Turns out it doesn’t have to be.
microtok makes the whole process stupidly simple — literally just 3 lines of code. No heavy setup, no GPU required. I built it on top of the Hugging Face tokenizers library so it stays clean, fast, and actually understandable.
If you’ve ever wanted to look under the hood and build your own optimized vocabulary instead of just copying someone else’s, this is the entry point you’ve been waiting for.
I wrote up the full story, threw in a ready-to-run Colab template, and dropped the trained tokenizer on Hugging Face.
Blog → https://parveshiiii.github.io/blogs/microtok/
Trained tokenizer → https://huggingface.co/Parveshiiii/microtok
GitHub repo → https://github.com/Parveshiiii/microtok
In only 3 hours, on 10 billion+ tokens, I trained a custom BPE + tiktoken-style tokenizer using my new library microtok — and it hits the same token efficiency as Qwen3.
Tokenizers have always felt like black magic to me. We drop them into every LLM project, but actually training one from scratch? That always seemed way too complicated.
Turns out it doesn’t have to be.
microtok makes the whole process stupidly simple — literally just 3 lines of code. No heavy setup, no GPU required. I built it on top of the Hugging Face tokenizers library so it stays clean, fast, and actually understandable.
If you’ve ever wanted to look under the hood and build your own optimized vocabulary instead of just copying someone else’s, this is the entry point you’ve been waiting for.
I wrote up the full story, threw in a ready-to-run Colab template, and dropped the trained tokenizer on Hugging Face.
Blog → https://parveshiiii.github.io/blogs/microtok/
Trained tokenizer → https://huggingface.co/Parveshiiii/microtok
GitHub repo → https://github.com/Parveshiiii/microtok
Post
7633
We should really have a release date range slider on the /models page. Tired of "trending/most downloaded" being the best way to sort and still seeing models from 2023 on the first page just because they're embedded in enterprise pipelines and get downloaded repeatedly. "Recently Created/Recently Updated" don't solve the discovery problem considering the amount of noise to sift through.
Slight caveat: Trending actually does have some recency bias, but it's not strong/precise enough.
Slight caveat: Trending actually does have some recency bias, but it's not strong/precise enough.
- 3 replies
Parveshiiii
posted an update 5 months ago
Post
352
Introducing Seekify — a truly non‑rate‑limiting search library for Python
Tired of hitting rate limits when building search features? I’ve built Seekify, a lightweight Python library that lets you perform searches without the usual throttling headaches.
🔹 Key highlights
- Simple API — plug it in and start searching instantly
- No rate‑limiting restrictions
- Designed for developers who need reliable search in projects, scripts, or apps
📦 Available now on PyPI:
👉 Check out the repo: https:/github.com/Parveshiiii/Seekify
I’d love feedback, contributions, and ideas for real‑world use cases. Let’s make search smoother together!
Tired of hitting rate limits when building search features? I’ve built Seekify, a lightweight Python library that lets you perform searches without the usual throttling headaches.
🔹 Key highlights
- Simple API — plug it in and start searching instantly
- No rate‑limiting restrictions
- Designed for developers who need reliable search in projects, scripts, or apps
📦 Available now on PyPI:
pip install seekify👉 Check out the repo: https:/github.com/Parveshiiii/Seekify
I’d love feedback, contributions, and ideas for real‑world use cases. Let’s make search smoother together!
[bot] Conversion to Parquet
#3 opened 5 months ago
by 
parquet-converter
not-lain
updated a
dataset 5 months ago
IlyasMoutawwakil
posted an update 6 months ago
Post
3317
Transformers v5 just landed! 🚀
It significantly unifies and reduces modeling code across architectures, while opening the door to a whole new class of performance optimizations.
My favorite new feature? 🤔
The new dynamic weight loader + converter. Here’s why 👇
Over the last few months, the core Transformers maintainers built an incredibly fast weight loader, capable of converting tensors on the fly while loading them in parallel threads. This means we’re no longer constrained by how parameters are laid out inside the safetensors weight files.
In practice, this unlocks two big things:
- Much more modular modeling code. You can now clearly see how architectures build on top of each other (DeepSeek v2 → v3, Qwen v2 → v3 → MoE, etc.). This makes shared bottlenecks obvious and lets us optimize the right building blocks once, for all model families.
- Performance optimizations beyond what torch.compile can do alone. torch.compile operates on the computation graph, but it can’t change parameter layouts. With the new loader, we can restructure weights at load time: fusing MoE expert projections, merging attention QKV projections, and enabling more compute-dense kernels that simply weren’t possible before.
Personally, I'm honored to have contributed in this direction, including the work on optimizing MoE implementations and making modeling code more torch-exportable, so these optimizations can be ported cleanly across runtimes.
Overall, Transformers v5 is a strong signal of where the community and industry are converging: Modularity and Performance, without sacrificing Flexibility.
Transformers v5 makes its signature from_pretrained an entrypoint where you can mix and match:
- Parallelism
- Quantization
- Custom kernels
- Flash/Paged attention
- Continuous batching
- ...
Kudos to everyone involved! I highly recommend the:
Release notes: https://github.com/huggingface/transformers/releases/tag/v5.0.0
Blog post: https://huggingface.co/blog/transformers-v5
It significantly unifies and reduces modeling code across architectures, while opening the door to a whole new class of performance optimizations.
My favorite new feature? 🤔
The new dynamic weight loader + converter. Here’s why 👇
Over the last few months, the core Transformers maintainers built an incredibly fast weight loader, capable of converting tensors on the fly while loading them in parallel threads. This means we’re no longer constrained by how parameters are laid out inside the safetensors weight files.
In practice, this unlocks two big things:
- Much more modular modeling code. You can now clearly see how architectures build on top of each other (DeepSeek v2 → v3, Qwen v2 → v3 → MoE, etc.). This makes shared bottlenecks obvious and lets us optimize the right building blocks once, for all model families.
- Performance optimizations beyond what torch.compile can do alone. torch.compile operates on the computation graph, but it can’t change parameter layouts. With the new loader, we can restructure weights at load time: fusing MoE expert projections, merging attention QKV projections, and enabling more compute-dense kernels that simply weren’t possible before.
Personally, I'm honored to have contributed in this direction, including the work on optimizing MoE implementations and making modeling code more torch-exportable, so these optimizations can be ported cleanly across runtimes.
Overall, Transformers v5 is a strong signal of where the community and industry are converging: Modularity and Performance, without sacrificing Flexibility.
Transformers v5 makes its signature from_pretrained an entrypoint where you can mix and match:
- Parallelism
- Quantization
- Custom kernels
- Flash/Paged attention
- Continuous batching
- ...
Kudos to everyone involved! I highly recommend the:
Release notes: https://github.com/huggingface/transformers/releases/tag/v5.0.0
Blog post: https://huggingface.co/blog/transformers-v5
- 3 replies
Parveshiiii
posted an update 6 months ago
Post
1652
🚀 Wanna train your own AI Model or Tokenizer from scratch?
Building models isn’t just for big labs anymore — with the right data, compute, and workflow, you can create **custom AI models** and **tokenizers** tailored to any domain. Whether it’s NLP, domain‑specific datasets, or experimental architectures, training from scratch gives you full control over vocabulary, embeddings, and performance.
✨ Why train your own?
- Full control over vocabulary & tokenization
- Domain‑specific optimization (medical, legal, technical, etc.)
- Better performance on niche datasets
- Freedom to experiment with architectures
⚡ The best part?
- Tokenizer training (TikToken / BPE) can be done in **just 3 lines of code**.
- Model training runs smoothly on **Google Colab notebooks** — no expensive hardware required.
📂 Try out my work:
- 🔗 https://github.com/OE-Void/Tokenizer-from_scratch
- 🔗 https://github.com/OE-Void/GPT
Building models isn’t just for big labs anymore — with the right data, compute, and workflow, you can create **custom AI models** and **tokenizers** tailored to any domain. Whether it’s NLP, domain‑specific datasets, or experimental architectures, training from scratch gives you full control over vocabulary, embeddings, and performance.
✨ Why train your own?
- Full control over vocabulary & tokenization
- Domain‑specific optimization (medical, legal, technical, etc.)
- Better performance on niche datasets
- Freedom to experiment with architectures
⚡ The best part?
- Tokenizer training (TikToken / BPE) can be done in **just 3 lines of code**.
- Model training runs smoothly on **Google Colab notebooks** — no expensive hardware required.
📂 Try out my work:
- 🔗 https://github.com/OE-Void/Tokenizer-from_scratch
- 🔗 https://github.com/OE-Void/GPT
IlyasMoutawwakil
posted an update 6 months ago
Post
2481
After 2 months of refinement, I'm happy to announce that a lot of Transformers' modeling code is now significantly more torch-compile & export-friendly 🔥
Why it had to be done 👇
PyTorch's Dynamo compiler is increasingly becoming the default interoperability layer for ML systems. Anything that relies on torch.export or torch.compile, from model optimization to cross-framework integrations, benefits directly when models can be captured as a single dynamo-traced graph !
Transformers models are now easier to:
⚙️ Compile end-to-end with torch.compile backends
📦 Export reliably via torch.export and torch.onnx.export
🚀 Deploy to ONNX / ONNX Runtime, Intel Corporation's OpenVINO, NVIDIA AutoDeploy (TRT-LLM), AMD's Quark, Meta's Executorch and more hardware-specific runtimes.
This work aims at unblocking entire TorchDynamo-based toolchains that rely on exporting Transformers across runtimes and accelerators.
We are doubling down on Transformers commitment to be a first-class citizen of the PyTorch ecosystem, more exportable, more optimizable, and easier to deploy everywhere.
There are definitely some edge-cases that we still haven't addressed so don't hesitate to try compiling / exporting your favorite transformers and to open issues / PRs.
PR in the comments ! More updates coming coming soon !
Why it had to be done 👇
PyTorch's Dynamo compiler is increasingly becoming the default interoperability layer for ML systems. Anything that relies on torch.export or torch.compile, from model optimization to cross-framework integrations, benefits directly when models can be captured as a single dynamo-traced graph !
Transformers models are now easier to:
⚙️ Compile end-to-end with torch.compile backends
📦 Export reliably via torch.export and torch.onnx.export
🚀 Deploy to ONNX / ONNX Runtime, Intel Corporation's OpenVINO, NVIDIA AutoDeploy (TRT-LLM), AMD's Quark, Meta's Executorch and more hardware-specific runtimes.
This work aims at unblocking entire TorchDynamo-based toolchains that rely on exporting Transformers across runtimes and accelerators.
We are doubling down on Transformers commitment to be a first-class citizen of the PyTorch ecosystem, more exportable, more optimizable, and easier to deploy everywhere.
There are definitely some edge-cases that we still haven't addressed so don't hesitate to try compiling / exporting your favorite transformers and to open issues / PRs.
PR in the comments ! More updates coming coming soon !
- 1 reply
Parveshiiii
posted an update 6 months ago
Post
275
📢 The Announcement
Subject: XenArcAI is now Modotte – A New Chapter Begins! 🚀
Hello everyone,
We are thrilled to announce that XenArcAI is officially rebranding to Modotte!
Since our journey began, we’ve been committed to pushing the boundaries of AI through open-source innovation, research, and high-quality datasets. As we continue to evolve, we wanted a name that better represents our vision for a modern, interconnected future in the tech space.
What is changing?
The Name: Moving forward, all our projects, models, and community interactions will happen under the Modotte banner.
The Look: You’ll see our new logo and a fresh color palette appearing across our platforms.
What is staying the same?
The Core Team: It’s still the same people behind the scenes, including our founder, Parvesh Rawal.
Our Mission: We remain dedicated to releasing state-of-the-art open-source models and datasets.
Our Continuity: All existing models, datasets, and projects will remain exactly as they are—just with a new home.
This isn’t just a change in appearance; it’s a commitment to our next chapter of growth and discovery. We are so grateful for your ongoing support as we step into this new era.
Welcome to the future. Welcome to Modotte.
Best regards, The Modotte Team
Subject: XenArcAI is now Modotte – A New Chapter Begins! 🚀
Hello everyone,
We are thrilled to announce that XenArcAI is officially rebranding to Modotte!
Since our journey began, we’ve been committed to pushing the boundaries of AI through open-source innovation, research, and high-quality datasets. As we continue to evolve, we wanted a name that better represents our vision for a modern, interconnected future in the tech space.
What is changing?
The Name: Moving forward, all our projects, models, and community interactions will happen under the Modotte banner.
The Look: You’ll see our new logo and a fresh color palette appearing across our platforms.
What is staying the same?
The Core Team: It’s still the same people behind the scenes, including our founder, Parvesh Rawal.
Our Mission: We remain dedicated to releasing state-of-the-art open-source models and datasets.
Our Continuity: All existing models, datasets, and projects will remain exactly as they are—just with a new home.
This isn’t just a change in appearance; it’s a commitment to our next chapter of growth and discovery. We are so grateful for your ongoing support as we step into this new era.
Welcome to the future. Welcome to Modotte.
Best regards, The Modotte Team
ntaquan0125
authored a
paper 6 months ago
Post
3642
Genuine recommendation: You should really use this AutoHotKey macro. Save the file as
In my experience the past few months,
macros.ahk and run it. Before sending a prompt to your coding agent, press Ctrl + Alt + 1 and paste your prompt to any regular chatbot. Then send the output to the agent. This is the actual, boring, real way to "10x your prompting". Use the other number keys to avoid repeating yourself over and over again. I use this macro prolly 100-200 times per day. AutoHotKey isn't as new or hype as a lot of other workflows, but there's a reason it's still widely used after 17 years. Don't overcomplicate it.; Requires AutoHotkey v1.1+
; All macros are `Ctrl + Alt + <variable>`
^!1::
Send, Please help me more clearly articulate what I mean with this message (write the message in a code block):
return
^!2::
Send, Please make the following changes:
return
^!3::
Send, It seems you got cut off by the maximum response limit. Please continue by picking up where you left off.
returnIn my experience the past few months,
Ctrl + Alt + 1 works best with Instruct models (non-thinking). Reasoning causes some models to ramble and miss the point. I've just been using GPT-5.x for this. Parveshiiii
posted an update 7 months ago
Post
3606
Hey everyone!
We’re excited to introduce our new Telegram group: https://t.me/XenArcAI
This space is built for **model builders, tech enthusiasts, and developers** who want to learn, share, and grow together. Whether you’re just starting out or already deep into AI/ML, you’ll find a supportive community ready to help with knowledge, ideas, and collaboration.
💡 Join us to:
- Connect with fellow developers and AI enthusiasts
- Share your projects, insights, and questions
- Learn from others and contribute to a growing knowledge base
👉 If you’re interested, hop in and be part of the conversation: https://t.me/XenArcAI
We’re excited to introduce our new Telegram group: https://t.me/XenArcAI
This space is built for **model builders, tech enthusiasts, and developers** who want to learn, share, and grow together. Whether you’re just starting out or already deep into AI/ML, you’ll find a supportive community ready to help with knowledge, ideas, and collaboration.
💡 Join us to:
- Connect with fellow developers and AI enthusiasts
- Share your projects, insights, and questions
- Learn from others and contribute to a growing knowledge base
👉 If you’re interested, hop in and be part of the conversation: https://t.me/XenArcAI
- 12 replies
Post
2850
🚨 New tool for the Nymbo/Tools MCP server: The new
How it works: The tool exposes the standard discover/info/resources/validate actions. Skills live in
I've included a
Caveat: On HF Spaces,
Try it out ~ https://www.nymbo.net/nymbot
Agent_Skills tool provides full support for Agent Skills (Claude Skills but open-source).How it works: The tool exposes the standard discover/info/resources/validate actions. Skills live in
/Skills under the same File_System root, and any bundled scripts run through Shell_Command, no new infrastructure required.Agent_Skills(action="discover") # List all available skills
Agent_Skills(action="info", skill_name="music-downloader") # Full SKILL.md
Agent_Skills(action="resources", skill_name="music-downloader") # Scripts, refs, assetsI've included a
music-downloader skill as a working demo, it wraps yt-dlp for YouTube/SoundCloud audio extraction.Caveat: On HF Spaces,
Shell_Command works for most tasks, but some operations (like YouTube downloads) are restricted due to the container environment. For full functionality, run the server locally on your machine.Try it out ~ https://www.nymbo.net/nymbot
potato4code
updated a
collection 8 months ago
Post
5307
🚀 I've just shipped a major update to the Nymbo/Tools MCP server: the
Anthropic found 98.7% context savings using code execution with MCP, Cloudflare published similar findings. This is my open-source implementation of the same idea.
# The Problem
Traditional MCP exposes every tool definition directly to the model. With 12 tools, that's thousands of tokens consumed *before the conversation even starts*. Each tool call also passes intermediate results through the context window — a 10,000-row spreadsheet? That's all going into context just to sum a column.
# The Solution: One Tool to Rule Them All
Instead of the model making individual tool calls, it writes Python code that orchestrates the tools directly:
Don't know what tools are available? The agent can discover them at runtime:
The individual direct tool calls are all still there, but they can be disabled if using the
Agent_Terminal, a single "master tool" that cuts token usage by over 90%!Anthropic found 98.7% context savings using code execution with MCP, Cloudflare published similar findings. This is my open-source implementation of the same idea.
# The Problem
Traditional MCP exposes every tool definition directly to the model. With 12 tools, that's thousands of tokens consumed *before the conversation even starts*. Each tool call also passes intermediate results through the context window — a 10,000-row spreadsheet? That's all going into context just to sum a column.
# The Solution: One Tool to Rule Them All
Agent_Terminal wraps all 12 tools (Web_Search, Web_Fetch, File_System, Generate_Image, Generate_Speech, Generate_Video, Deep_Research, Memory_Manager, Obsidian_Vault, Shell_Command, Code_Interpreter) into a single Python code execution gateway.Instead of the model making individual tool calls, it writes Python code that orchestrates the tools directly:
# Search for Bitcoin price
result = Web_Search("current price of bitcoin", max_results=3)
print(result)Don't know what tools are available? The agent can discover them at runtime:
print(search_tools('image')) # Find tools by keyword
print(usage('Generate_Image')) # Get full docs for a specific toolThe individual direct tool calls are all still there, but they can be disabled if using the
Agent_Terminal. Try it now - https://www.nymbo.net/nymbot 
- 1 reply
Parveshiiii
posted an update 8 months ago
Post
1680
Another banger from XenArcAI! 🔥
We’re thrilled to unveil three powerful new releases that push the boundaries of AI research and development:
🔗 https://huggingface.co/XenArcAI/SparkEmbedding-300m
- A lightning-fast embedding model built for scale.
- Optimized for semantic search, clustering, and representation learning.
🔗 https://huggingface.co/datasets/XenArcAI/CodeX-7M-Non-Thinking
- A massive dataset of 7 million code samples.
- Designed for training models on raw coding patterns without reasoning layers.
🔗 https://huggingface.co/datasets/XenArcAI/CodeX-2M-Thinking
- A curated dataset of 2 million code samples.
- Focused on reasoning-driven coding tasks, enabling smarter AI coding assistants.
Together, these projects represent a leap forward in building smarter, faster, and more capable AI systems.
💡 Innovation meets dedication.
🌍 Knowledge meets responsibility.
We’re thrilled to unveil three powerful new releases that push the boundaries of AI research and development:
🔗 https://huggingface.co/XenArcAI/SparkEmbedding-300m
- A lightning-fast embedding model built for scale.
- Optimized for semantic search, clustering, and representation learning.
🔗 https://huggingface.co/datasets/XenArcAI/CodeX-7M-Non-Thinking
- A massive dataset of 7 million code samples.
- Designed for training models on raw coding patterns without reasoning layers.
🔗 https://huggingface.co/datasets/XenArcAI/CodeX-2M-Thinking
- A curated dataset of 2 million code samples.
- Focused on reasoning-driven coding tasks, enabling smarter AI coding assistants.
Together, these projects represent a leap forward in building smarter, faster, and more capable AI systems.
💡 Innovation meets dedication.
🌍 Knowledge meets responsibility.