Sijun He 5ac5803d9c | 9 months ago | |
---|---|---|
.. | ||
README.md | 10 months ago | |
argument.py | 10 months ago | |
data.py | 9 months ago | |
finetune_generation.py | 9 months ago | |
merge_lora_params.py | 9 months ago | |
merge_tp_params.py | 10 months ago | |
predict_generation.py | 9 months ago | |
quant.py | 9 months ago | |
utils.py | 9 months ago |
export MODEL='THUDM/chatglm-6b'
export DATA='data'
python -u -m paddle.distributed.launch --gpus "0,1,2,3" finetune_generation.py \
--model_name_or_path $MODEL \
--dataset_name_or_path $DATA \
--output_dir ./checkpoints \
--per_device_train_batch_size 4 \
--gradient_accumulation_steps 2 \
--per_device_eval_batch_size 8 \
--num_train_epochs 3 \
--learning_rate 3e-5 \
--warmup_steps 30 \
--logging_steps 1 \
--evaluation_strategy epoch \
--save_strategy epoch \
--src_length 1024 \
--tgt_length 1024 \
--fp16 \
--fp16_opt_level O2 \
--do_train \
--do_eval \
--disable_tqdm True \
--load_best_model_at_end True \
--metric_for_best_model accuracy \
--eval_with_do_generation False \
--recompute \
--save_total_limit 1 \
--tensor_parallel_degree 4
export MODEL='THUDM/chatglm-6b'
export DATA='data'
python finetune_generation.py \
--model_name_or_path $MODEL \
--dataset_name_or_path $DATA \
--output_dir ./checkpoints \
--per_device_train_batch_size 4 \
--gradient_accumulation_steps 2 \
--per_device_eval_batch_size 8 \
--num_train_epochs 1 \
--learning_rate 3e-4 \
--warmup_steps 30 \
--logging_steps 1 \
--evaluation_strategy epoch \
--save_strategy epoch \
--src_length 1024 \
--tgt_length 1024 \
--fp16 \
--fp16_opt_level O2 \
--do_train \
--do_eval \
--disable_tqdm True \
--load_best_model_at_end True \
--metric_for_best_model accuracy \
--eval_with_do_generation False \
--recompute \
--save_total_limit 1 \
--lora True
export MODEL='THUDM/chatglm-6b'
export DATA='data'
python finetune_generation.py \
--model_name_or_path $MODEL \
--dataset_name_or_path $DATA \
--output_dir ./checkpoints \
--per_device_train_batch_size 4 \
--gradient_accumulation_steps 2 \
--per_device_eval_batch_size 8 \
--num_train_epochs 1 \
--learning_rate 3e-2 \
--warmup_steps 30 \
--logging_steps 1 \
--evaluation_strategy epoch \
--save_strategy epoch \
--src_length 1024 \
--tgt_length 1024 \
--fp16 \
--fp16_opt_level O2 \
--do_train \
--do_eval \
--disable_tqdm True \
--load_best_model_at_end True \
--metric_for_best_model accuracy \
--eval_with_do_generation False \
--recompute \
--save_total_limit 1 \
--prefix_tuning True
👑 Easy-to-use and powerful NLP library with 🤗 Awesome model zoo, supporting wide-range of NLP tasks from research to industrial applications, including 🗂Text Classification, 🔍 Neural Search, ❓ Question Answering, ℹ️ Information Extraction, 📄 Documen
Python C++ Cuda Shell Markdown other
Dear OpenI User
Thank you for your continuous support to the Openl Qizhi Community AI Collaboration Platform. In order to protect your usage rights and ensure network security, we updated the Openl Qizhi Community AI Collaboration Platform Usage Agreement in January 2024. The updated agreement specifies that users are prohibited from using intranet penetration tools. After you click "Agree and continue", you can continue to use our services. Thank you for your cooperation and understanding.
For more agreement content, please refer to the《Openl Qizhi Community AI Collaboration Platform Usage Agreement》