jrahn/gpt2_350M_edu_hermes
Text Generation
•
0.4B
•
Updated
•
15
•
1
llm.c instruction pretraining explorations
Exploring to find a successful training recipe interleaving pre-training with instruction data