got nothing after finetune the llava-v1.5-13b [Usage]

[459737087]

Describe the issue

Issue: got nothing after finetune the llava-v1.5-13b predict "-------------------------" or blank answer

Command:

PROMPT_VERSION=v1
MODEL_VERSION="llava-v1.5-13b"
deepspeed llava/train/train_mem.py \
    --deepspeed ./scripts/zero2.json \
    --lora_enable True \
    --model_name_or_path /output/$MODEL_VERSION \
    --version $PROMPT_VERSION \
    --data_path /output/data.json \
    --image_folder /output/train_data/input1/ \
    --vision_tower openai/clip-vit-large-patch14 \
    --pretrain_mm_mlp_adapter /output/$MODEL_VERSION/mm_projector.bin \
    --mm_vision_select_layer -2 \
    --mm_use_im_start_end False \
    --mm_use_im_patch_token False \
    --bf16 True \
    --output_dir ./checkpoints/$MODEL_VERSION-finetune_lora_0306 \
    --num_train_epochs 100 \
    --per_device_train_batch_size 16 \
    --per_device_eval_batch_size 4 \
    --gradient_accumulation_steps 1 \
    --evaluation_strategy "no" \
    --save_strategy "steps" \
    --save_steps 5000 \
    --save_total_limit 1 \
    --learning_rate 2e-5 \
    --weight_decay 0. \
    --warmup_ratio 0.03 \
    --lr_scheduler_type "cosine" \
    --logging_steps 1 \
    --tf32 True \
    --model_max_length 2048 \
    --gradient_checkpointing True \
    --lazy_preprocess True \
    --dataloader_num_workers 4 \
    --report_to wandb

Log:

PASTE THE LOGS HERE.

Screenshots:

[ykzqjyhhh]

I have the same problem as well. Have you resolved it?