Aaron Yang

contrapuntal's activity

• All
• Models
• Datasets
• Spaces
• Buckets
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

New activity in mlx-community/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-mxfp8 about 2 months ago

mxfp8 weights produce all-NaN logits — re-quant from BF16 may be needed

#1 opened about 2 months ago by

contrapuntal