Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Same output after finetuning #29

Open
lichenyigit opened this issue Nov 24, 2023 · 1 comment
Open

Same output after finetuning #29

lichenyigit opened this issue Nov 24, 2023 · 1 comment
Labels
fine-tuning Fine-tuning DISC-LawLLM needs triage The issue needs to be triaged by some maintainer

Comments

@lichenyigit
Copy link

进行了10000次训练力度,相同的问题最终输出结果一样。请问是哪个步骤出了问题?
数据集:
新闻Q&A.json
训练脚本:

torchrun --nproc_per_node 1 src/train_bash.py \
    --stage sft \
    --model_name_or_path ShengbinYue/DISC-LawLLM \
    --do_train \
    --dataset yiqi5-fun \
    --template baichuan2 \
    --finetuning_type lora \
    --lora_rank 8 \
    --lora_target W_pack \
    --output_dir /home/DISC-output-checkpoint \
    --overwrite_output_dir \
    --overwrite_cache \
    --per_device_train_batch_size 4 \
    --per_device_eval_batch_size 4 \
    --gradient_accumulation_steps 8 \
    --preprocessing_num_workers 16 \
    --lr_scheduler_type cosine \
    --logging_steps 10 \
    --save_steps 100 \
    --eval_steps 100 \
    --learning_rate 1e-5 \
    --max_grad_norm 0.5 \
    --num_train_epochs 10000.0 \
    --evaluation_strategy steps \
    --load_best_model_at_end \
    --plot_loss \
    --fp16 \
    --val_size 0.01

@Charlie-XIAO

@Charlie-XIAO Charlie-XIAO added needs triage The issue needs to be triaged by some maintainer fine-tuning Fine-tuning DISC-LawLLM labels Dec 15, 2023
@Charlie-XIAO
Copy link
Collaborator

@yueshengbin who might be more familiar with these.

@Charlie-XIAO Charlie-XIAO changed the title 微调失败 Same output after finetuning Dec 15, 2023
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
fine-tuning Fine-tuning DISC-LawLLM needs triage The issue needs to be triaged by some maintainer
Projects
None yet
Development

No branches or pull requests

2 participants