THUDM/slime

slime is an LLM training framework for RL Scaling.

8.7k
Stars
1.3k
Forks
74.2
Radar Score
Rising
Momentum

Scoring Breakdown

Impact89.0
Velocity65.0
Health63.4

Growth Metrics

7-Day Stars
+12
30-Day Stars
+188
Acceleration
0.96x
Language
Python

Why it ranks

Strong momentum — adoption is accelerating week over week.

LLM Frameworks · fine-tuningTop 51% globallyLast push 1d ago+12 stars / 7d+188 stars / 30d
View on GitHub