Continued-pretraining of qingy2024/Qwen2.5-4B on 143M tokens from HuggingFaceTB/finemath 4-plus
qingy2024/Qwen2.5-4B
Base model