Deep Learning

1.nn.CrossEntropyLoss() 에 softmax 값을 넣지 않는 이유

post-thumbnail

2.Gradient Clipping

post-thumbnail

3.BLEU Score 대략적 지표

post-thumbnail

4.BERT huggingface 사용법

post-thumbnail

6.Auto-regressive model, Auto-encoding model

post-thumbnail

7.Tokenizer

post-thumbnail

8.Multi-GPU with DP, DDP, FSDP

post-thumbnail

9.Chain-of-thought, Zero-shot Chain-of-thought, instruction tuning

post-thumbnail

10.RLHF(reinforcement learning from human feedback)

post-thumbnail

11.WandB

post-thumbnail

12.PEFT(parameter efficient fine-tuning)

post-thumbnail

13.BPE, Sentencepiece

post-thumbnail

14.Batch size & learning rate

post-thumbnail

15.gradScaler, Autocast

post-thumbnail

16.gradient checkpointing

post-thumbnail