[GH-ISSUE #57] [Bug] Train a personalized DeepFloyd model by DreamBooth Lora. Execute stage 1, but there is no personalized effect #57

New issue

Open

opened 2026-10-26 21:03:41 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-26 21:03:41 +00:00

Owner

Originally created by @xdobetter on GitHub (Apr 30, 2024).
Original GitHub issue: https://github.com/deepseek-ai/DreamCraft3D/issues/57

Great job.
You mentioned Train a personalized DeepFloyd model by DreamBooth Lora. I followed the relevant instructions to train a lora weight, and then executed stage 1, but there is no personalized effect, that is to say, it did not generate content related to my topic. Why is this?

The command to train lora is as follows

export MODEL_NAME="DeepFloyd/IF-I-XL-v1.0"
export INSTANCE_DIR=".cache/temp"
export OUTPUT_DIR=".cache/if_dreambooth_mushroom"

accelerate launch threestudio/scripts/train_dreambooth_lora.py \
  --pretrained_model_name_or_path=$MODEL_NAME  \
  --instance_data_dir=$INSTANCE_DIR \
  --output_dir=$OUTPUT_DIR \
  --instance_prompt="a sks mushroom" \
  --resolution=64 \
  --train_batch_size=4 \
  --gradient_accumulation_steps=1 \
  --learning_rate=5e-6 \
  --scale_lr \
  --max_train_steps=1200 \
  --checkpointing_steps=600 \
  --pre_compute_text_embeddings \
  --tokenizer_max_length=77 \
  --text_encoder_use_attention_mask

The command to train stage 1 is as follows

# --------- Stage 1 (NeRF & NeuS) --------- # 
python launch.py --config configs/dreamcraft3d-coarse-nerf.yaml --train system.prompt_processor.prompt="$prompt" data.image_path="$image_path" system.guidance.lora_weights_path=".cache/if_dreambooth_mushroom"

Originally created by @xdobetter on GitHub (Apr 30, 2024). Original GitHub issue: https://github.com/deepseek-ai/DreamCraft3D/issues/57 Great job. You mentioned Train a personalized DeepFloyd model by DreamBooth Lora. I followed the relevant instructions to train a lora weight, and then executed stage 1, but there is no personalized effect, that is to say, it did not generate content related to my topic. Why is this? The command to train lora is as follows ```bash export MODEL_NAME="DeepFloyd/IF-I-XL-v1.0" export INSTANCE_DIR=".cache/temp" export OUTPUT_DIR=".cache/if_dreambooth_mushroom" accelerate launch threestudio/scripts/train_dreambooth_lora.py \ --pretrained_model_name_or_path=$MODEL_NAME \ --instance_data_dir=$INSTANCE_DIR \ --output_dir=$OUTPUT_DIR \ --instance_prompt="a sks mushroom" \ --resolution=64 \ --train_batch_size=4 \ --gradient_accumulation_steps=1 \ --learning_rate=5e-6 \ --scale_lr \ --max_train_steps=1200 \ --checkpointing_steps=600 \ --pre_compute_text_embeddings \ --tokenizer_max_length=77 \ --text_encoder_use_attention_mask ``` The command to train stage 1 is as follows ```bash # --------- Stage 1 (NeRF & NeuS) --------- # python launch.py --config configs/dreamcraft3d-coarse-nerf.yaml --train system.prompt_processor.prompt="$prompt" data.image_path="$image_path" system.guidance.lora_weights_path=".cache/if_dreambooth_mushroom" ```

No labels

pull-request

No milestone

No project

No assignees

1 participant

Notifications

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference

deepseek-ai/DreamCraft3D#57

No description provided.

Rows
Columns