Hugging Face's logo
Thawatchai's picture

Thawatchai

Grace-03
19
PhysiQuanty's profile picture
·
https://github.com/grace639

AI & ML interests

-

Recent Activity

reacted to harshitkgupta's post with 🚀 2 days ago
Fine-tuned Qwen 2.5 (0.5B → 3B) on real coding-agent traces, 10 controlled runs, one 16GB Mac. Compared PyTorch MPS vs. Apple MLX for local LoRA SFT — and the honest answer is "it depends on what you're optimizing for": • PyTorch MPS: 2.2x–5.7x faster raw throughput, but hits a hard memory wall — can't load a 3B model in FP16 on 16GB. • Apple MLX: 4-bit QLoRA fits 3B+ models with almost flat memory scaling as context grows (+109 MB going from 1k→4k tokens). • 4-bit quantization doesn't cost you convergence — eval loss tracks closely across backends. • The bigger surprise: most of MLX's slowdown isn't the 4-bit dequant tax. Two of the 10 runs went unquantized to isolate it — dequant only explains 1.07x–1.4x of the gap. A ~4.1–4.6x framework-level gap remains either way. All 10 LoRA adapters + Trackio logs are public so the numbers are checkable, not just claimed. Full writeup: https://huggingface.co/blog/harshitkgupta/fine-tuning-coding-agents-on-mac-pytorch-mps-mlx
liked a model 13 days ago
m-a-p/YuE2-3B
liked a model 20 days ago
deepseek-ai/DeepSeek-V4.1-Flash
View all activity

Organizations

None yet