b76d333f39 | ||
---|---|---|
.vscode | ||
custom_models | ||
.gitignore | ||
LICENSE | ||
README.md | ||
generate.py | ||
lit_export.py | ||
lit_module.py | ||
lit_patches.py | ||
lit_train.py | ||
requirements.txt | ||
utils.py |
README.md
GPT-Pretrain
Usage
Make it simple
python lit_train.py --model_name gpt2 --use_tril_attention_mask
python lit_export.py --version 0
python generate.py --model_name_or_path exports/version_0 --tokenizer_name_or_path gpt2
📝 Note: Training with a "--use_tril_attention_mask" is recommended. However, huggingface model implementions might not support 2D attention mask. You may write a custom model to support 2D attention mask, just like what I did in custom_models/gpt2.
Train on multiple GPUs
python lit_train.py --model_name gpt2 --use_tril_attention_mask --strategy fsdp # default and recommended
python lit_train.py --model_name gpt2 --use_tril_attention_mask --strategy deepspeed
python lit_train.py --model_name gpt2 --use_tril_attention_mask --strategy ddp
Reduce CUDA memory cost
- half precision
python lit_train.py --model_name gpt2 --use_tril_attention_mask --bf16
python lit_train.py --model_name gpt2 --use_tril_attention_mask --fp16
- smaller batch size & accumulate grad batches
python lit_train.py --model_name gpt2 --use_tril_attention_mask --bf16 \ --train_batch_size 2 --val_batch_size 4 --accumulate_grad_batches 128
- cpu_offload
python lit_train.py --model_name gpt2 --use_tril_attention_mask --bf16 \ --strategy fsdp_cpu_offload
python lit_train.py --model_name gpt2 --use_tril_attention_mask --bf16 \ --strategy deepspeed_stage_3_offload