Been using Humanizer, currently #2 trending in text generation on Hugging Face. It's great! @jialinyyzz Wanted it faster, so I trained a DSpark draft with DraftFit. 7.4s → 3.9s on my Mac. Exact same output in this example. Sharing the weights: https://t.co/krr4ZMu5Wa https:/…

You can deploy an open model using @huggingface Inference Endpoints https://t.co/NNbfoHxRms For anyone interested, here are the settings(screenshot below) I used Model deployed: Qwen3.8-27B-UD-Q4_K_M Then configure Pi @pidotdev agent's models.json: { "providers": { "h…

Hugging Face has a million models and the search is vibes. So I built hugcheck — a CLI that audits models before you pull: pickle RCE checks, typosquat detection, license/card checks. No fake trust scores. npx hugcheck audit <model> https://t.co/4dxaiiy2DQ [https://t.co/DL4Hp…

fine-tuned Llama 3.1-8B with 4-bit quantization and LoRA! Pushed it to Hugging Face hub. #AI #MachineLearning #HuggingFace

I built ImageLoom: a free, open-source AI image studio that runs fully on your own Mac or PC. Chat-style workflow, brush-mask editing, 4K upscaling, SVG / WebP / transparent PNG export. Paste a Hugging Face link and a new model just works. No cloud, no account. MIT. https://t.…

Built my own Indian Legal AI Chatbot 🚀 General LLMs hallucinate on sensitive legal queries, so I fine-tuned Qwen2.5-1.5B with QLoRA on FREE Colab T4 GPU and deployed it on Hugging Face. Live Demo👇 https://t.co/DBNxZWwFhn Model👇 https://t.co/OqUCrh8Whp #AI #LLM #HuggingFace ht…

I pertained and fine-tuned my own LLM chatbot from scratch using Single GPU!! The model is built with PyTorch, using LatentMoE architecture. Following is the Hugging Face Repo. 150M, pure Korean model. 🤗:https://t.co/PFfbtI9KQl @huggingface #AI #LargeLanguageModel

🚀 Validation Loss: 0.1789 in 1 Epoch? I fine-tuned a pre-trained LLM with custom movie reviews using @HuggingFace & got better results than expected! How I Did It: Took a base model Added movie review data Fine-tuned with Hugging Face tools Zero PhD required! Full Guide Inside:…

Simplified Technical English (STE100) is exactly what I have been using to train my latest LLMs. Some previews open-source on Hugging Face, a small dataset too https://t.co/l1qXkTOVKN https://t.co/tEdzS51ENr

Day 17/30 ✅ | GFG Summer SkillUp Challenge Today I explored Hugging Face Transformers and learned how modern AI models are built, fine-tuned, and deployed for real-world applications #SummerSkillUp #GFG #HuggingFace #Transformers #AgenticAI #AI #skillupwithgfg https://t.co/p1I…

I built my first AI chatbot today. Thank you @huggingface. Serverless inference APIs are crazy, helping me embed data to @AstraDB for FREE. Are there any alternatives to hugging face? Should I use AI models by downloading it locally?

I built a lightweight coding assistant by fine-tuned Llama 3.2 1B Instruct using: • Unsloth • Converted to GGUF • ran locally on Ollama • Deployed on Hugging face what should I do next? #LLM #AI #Ollama #OpenSource https://t.co/0ubDCUWlyj

Just fine tuned BLIP on the Flickr8k dataset using LoRA! 🚀 The model is live on Hugging Face, and the training code is on GitHub. Feedback and suggestions are always welcome! #AI #PyTorch #HuggingFace #BuildinPublic https://t.co/HcsQZwKm7y

🚀 Excited to share that I've fine-tuned the LLaMA model using unsloth on a custom dataset for keyword & variable detection! 🎉 Check it out on Hugging Face: https://t.co/rhDWv93rXM. @AIatMeta @llama @llama_index @UnslothAI @huggingface #LLaMA3 #metaai