stanford_alpaca
stanford_alpaca copied to clipboard
Question about training precision
In the provided training command:
torchrun --nproc_per_node=4 --master_port=<your_random_port> train.py \
--model_name_or_path <your_path_to_hf_converted_llama_ckpt_and_tokenizer> \
--data_path ./alpaca_data.json \
--bf16 True \
--output_dir <your_output_dir> \
--num_train_epochs 3 \
--per_device_train_batch_size 4 \
--per_device_eval_batch_size 4 \
--gradient_accumulation_steps 8 \
--evaluation_strategy "no" \
--save_strategy "steps" \
--save_steps 2000 \
--save_total_limit 1 \
--learning_rate 2e-5 \
--weight_decay 0. \
--warmup_ratio 0.03 \
--lr_scheduler_type "cosine" \
--logging_steps 1 \
--fsdp "full_shard auto_wrap" \
--fsdp_transformer_layer_cls_to_wrap 'LLaMADecoderLayer' \
--tf32 True
Why is --bf16
used, if the model checkpoints were originally fp16? Is it simply overridden by the --tf32
flag later?