Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8 · Download Model Really slow

try to download model multiple time but usually stuck on 99% for every .safetensors files, need to cancel and try again multiple time
did someone experience same issue?
also try using Git to clone but with no luck

previously downloading DeepSeekR1 full model only took 6 hours without facing any issue

root@tegpu01worker02:/models# HF_HUB_ENABLE_HF_TRANSFER=1 huggingface-cli download Qwen/Qwen3-Coder-480B-A35B-Instruct-FP8 --local-dir models--Qwen--Qwen3-Coder-480B-A35B-Instruct-FP8
Removing incomplete file 'models--Qwen--Qwen3-Coder-480B-A35B-Instruct-FP8/.cache/huggingface/download/iuz8aHLO6Bg3tPS2nIHTyqxKN8g=.8688e8d26a7ca9e3dbef62a7dae247d85014978d558e91e7bba8d7a611548bc7.incomplete' (hf_transfer=True)
Downloading 'model-00006-of-00049.safetensors' to 'models--Qwen--Qwen3-Coder-480B-A35B-Instruct-FP8/.cache/huggingface/download/iuz8aHLO6Bg3tPS2nIHTyqxKN8g=.8688e8d26a7ca9e3dbef62a7dae247d85014978d558e91e7bba8d7a611548bc7.incomplete'
model-00006-of-00049.safetensors: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 10.0G/10.0G [01:25<00:00, 116MB/s]
Download complete. Moving file to models--Qwen--Qwen3-Coder-480B-A35B-Instruct-FP8/model-00006-of-00049.safetensors
Downloading 'model-00007-of-00049.safetensors' to 'models--Qwen--Qwen3-Coder-480B-A35B-Instruct-FP8/.cache/huggingface/download/BS4COEvJKSZFsz-UCd7hwQJLFzI=.5bf602baa71a1e711362e06057dc5b01ce08e11c03bd70596f2fa57b783cf9ba.incomplete'
model-00007-of-00049.safetensors:  99%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████▏ | 9.86G/10.0G [06:23<00:01, 90.2MB/s]