ads

sxcasf's activity

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

commented 2 papers 3 months ago

Lightning OPD: Efficient Post-Training for Large Reasoning Models with Offline On-Policy Distillation

Paper • 2604.13010 • Published Apr 14 • 20 •

7

Lightning OPD: Efficient Post-Training for Large Reasoning Models with Offline On-Policy Distillation

Paper • 2604.13010 • Published Apr 14 • 20 •

7

New activity in HuggingFaceTB/Countdown-Task-GOLD 8 months ago

Inconsistent numbers

6

#1 opened 8 months ago by

MysticJay

New activity in Qwen/Qwen3-1.7B 8 months ago

When enable_thinking=True, why doesn't the chat_template output end with "<think>？

#16 opened 8 months ago by

sxcasf

New activity in VityaVitalich/Qwen3-1.7B 8 months ago

When enable_thinking=True, why doesn't the chat_template output end with "<think>

#1 opened 8 months ago by

sxcasf

New activity in WizardLMTeam/WizardLM-13B-V1.0 almost 2 years ago

Why is it that more than 99.5% of the parameter values in each layer are less than 1e-4

#12 opened almost 2 years ago by

sxcasf