OPEN SOURCE

→ View original post on X — @saboo_shubham_

Hugging Face releases AI sheets for dataset building without code

By

@saboo_shubham_

–

16 August 2025 17h00

Hugging Face just dropped AI sheets to build and enrich datasets without writing a single line of code.

Works with Qwen, Kimi, Llama 3 and other opensource LLMs.

100% Free, local and Opensource.pic.twitter.com/onvrgg5hqR
— Shubham Saboo (@Saboo_Shubham_) 16 août 2025

Hugging Face just dropped AI sheets to build and enrich datasets without writing a single line of code. Works with Qwen, Kimi, Llama 3 and other opensource LLMs. 100% Free, local and Opensource.

Axolotl: YAML-based LLM fine-tuning and LoRA optimization framework

By

–

16 August 2025 16h13

4. Axolotl • Yaml-based setup for fine-tuning, LoRA/QLoRA, DPO, GRPO, and multimodal workflows
• Includes kernel optimizations for memory-efficient training GitHub repo:

DeepSpeed: Distributed Fine-Tuning Framework for Large Language Models

By

–

16 August 2025 16h13

3. DeepSpeed • Built for large-scale distributed fine-tuning with ZeRO and FSDP
• Optimized for multi-GPU and multi-node training with advanced memory management
• Trusted in production environments for scalable LLM training GitHub repo:

LLaMA Factory: Fine-tune 100+ Models with Simple CLI

By

–

16 August 2025 16h13

2. LLaMA Factory • Fine-tune over 100 models (LLaMA, Mistral, Gemma, etc.) using a simple CLI or WebUI
• Supports LoRA, QLoRA, full or frozen fine-tuning across 2–8‑bit precision GitHub repo:

Unsloth AI: Fast LLM Fine-tuning with 70% Less VRAM

By

–

16 August 2025 16h13

1. Unsloth AI • Fine-tune models like Qwen3, Llama 4, and Gemma 3 up to 2× faster with 70% less VRAM
• Supports low-resource setups and runs on consumer GPUs or even Colab/Kaggle with ~3 GB VRAM GitHub repo:

Four Open Source Libraries Accelerate LLM Fine-tuning

By

–

16 August 2025 16h13

Fine-tuning massive LLMs used to be painfully slow, but not anymore! Here are 4 libraries that accelerates fine-tuning of Large Language Models 100% Open Source

→ View original post on X — @demishassabis

Gemma 3 270M: Efficient Open Model for Edge Devices

By

@demishassabis

–

16 August 2025 1h45

New hyper-efficient addition to our amazing Gemma open models: Gemma 3 270M packs a real punch for its tiny size! It’s super compact and power efficient, so you can easily run your own task-specific fine-tuned systems on edge devices. Enjoy building with it!