Optimized Runtimes — GGUF & ONNX Builds Collection Every quantized and exported artifact in one place. GGUF builds for llama.cpp, Ollama and LM Studio; ONNX exports for ONNX Runtime and browser. • 7 items • Updated about 17 hours ago
Beyond Decoders — Translation & Classification Collection Encoder-decoder and encoder-only models for tasks where a decoder-only LLM isn’t the right tool. • 5 items • Updated about 17 hours ago
Bangla LLM Suite — Fine-tuned, Quantized & Deployed Collection End-to-end Bangla language models: Generation and summarization checkpoints, each shipped with quantized GGUF or ONNX builds and a live demo Sp • 9 items • Updated about 17 hours ago
Task-Specific Fine-Tunes — FFT, LoRA & QLoRA Collection Parameter-efficient fine-tunes across three base families (Qwen2.5, Llama 3.2, Gemma 3) at 1B–3B scale. Text-to-SQL, reflective reasoning, paraphrasin • 6 items • Updated about 17 hours ago