Support ERNIE-Image (#1389)

* ernie-image pipeline

* ernie-image inference and training

* style fix

* ernie docs

* lowvram

* final style fix

* pr-review

* pr-fix round2

* set uniform training weight

* fix

* update lowvram docs

examples/ernie_image/model_training/full/Ernie-Image-T2I.sh

@@ -0,0 +1,17 @@
+# Dataset: data/diffsynth_example_dataset/ernie_image/Ernie-Image-T2I/
+
+accelerate launch --config_file examples/ernie_image/model_training/full/accelerate_config_zero3.yaml \
+  examples/ernie_image/model_training/train.py \
+  --dataset_base_path data/diffsynth_example_dataset/ernie_image/Ernie-Image-T2I \
+  --dataset_metadata_path data/diffsynth_example_dataset/ernie_image/Ernie-Image-T2I/metadata.csv \
+  --max_pixels 1048576 \
+  --dataset_repeat 50 \
+  --model_id_with_origin_paths "baidu/ERNIE-Image:transformer/diffusion_pytorch_model*.safetensors,baidu/ERNIE-Image:text_encoder/model.safetensors,baidu/ERNIE-Image:vae/diffusion_pytorch_model.safetensors" \
+  --learning_rate 1e-5 \
+  --num_epochs 2 \
+  --remove_prefix_in_ckpt "pipe.dit." \
+  --output_path "./models/train/Ernie-Image-T2I_full" \
+  --trainable_models "dit" \
+  --use_gradient_checkpointing \
+  --dataset_num_workers 8 \
+  --find_unused_parameters

examples/ernie_image/model_training/full/accelerate_config_zero3.yaml

@@ -0,0 +1,23 @@
+compute_environment: LOCAL_MACHINE
+debug: false
+deepspeed_config:
+  gradient_accumulation_steps: 1
+  offload_optimizer_device: none
+  offload_param_device: none
+  zero3_init_flag: true
+  zero3_save_16bit_model: true
+  zero_stage: 3
+distributed_type: DEEPSPEED
+downcast_bf16: 'no'
+enable_cpu_affinity: false
+machine_rank: 0
+main_training_function: main
+mixed_precision: bf16
+num_machines: 1
+num_processes: 8
+rdzv_backend: static
+same_network: true
+tpu_env: []
+tpu_use_cluster: false
+tpu_use_sudo: false
+use_cpu: false