Richard Sun
Richard Sun
Hi, I am new to NLP and interested to explore the hype of word2vec. I wanna carry out some intrinsic evaluation such as "man-women=father-mother". In gensim package, we can do...
I would like to make use more than 1 of the servers in my lab. Is it convenient to migrate the fully fine-tuning code to a multi-node environment? Has anybody...
I tried to fully finetune llama-7b with deepspeed based on the code https://github.com/Lightning-AI/lit-llama/blob/main/finetune/full.py. I replaced the FSDPStrategy with DeepSpeedStrategy(offload_optimizer=True, offload_parameters=False, pin_memory=True, offload_optimizer_device='cpu'), but there would always be an OOM error...
I ran into this error when resuming the pre-training on the pre-trained weights: "ImportError: cannot import name '_FaultTolerantMode' from 'pytorch_lightning.utilities.enums' " Has anyone seen this issue before? Thanks! I modified...
how can we add the conversation templates for new LLMs such as Llama3 and Qwen2? Thanks!