Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Training Stability Analysis on Mixtral 8x1B pre-training

0Num Spikes

AdaGC

-2.0811.962640.04Feb 16, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.02
00
2025.02
521.44