Unsloth & SDPA Integrated in MonsterAPI for 2x LLM Finetuning

I saw the Unsloth work yesterday While it sounds great, it doesn't support multi-GPUmulti-node fine-tuning I'm using trl library with

I used Llama Factory and Unsloth supports multigpu without issue when running in accelerate

unsloth multi gpu GPU support How are we faster? By manually deriving all compute multiple GPU systems We support NVIDIA GPUs from Tesla T4 to H100

unsloth pro price GPU support How are we faster? By manually deriving all compute multiple GPU systems We support NVIDIA GPUs from Tesla T4 to H100