Skip to content

enable stable-xl textual inversion #6421

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

Merged
merged 19 commits into from
Jan 9, 2024
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions examples/textual_inversion/README.md
Original file line number Diff line number Diff line change
Expand Up @@ -60,6 +60,8 @@ Now we can launch the training using:

**___Note: Change the `resolution` to 768 if you are using the [stable-diffusion-2](https://huggingface.co/stabilityai/stable-diffusion-2) 768x768 model.___**

**___Note: Please follow the [README_sdxl.md](./README_sdxl.md) if you are using the [stable-diffusion-xl](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0).___**

```bash
export MODEL_NAME="runwayml/stable-diffusion-v1-5"
export DATA_DIR="./cat"
Expand Down
26 changes: 26 additions & 0 deletions examples/textual_inversion/README_sdxl.md
Original file line number Diff line number Diff line change
@@ -0,0 +1,26 @@
## Textual Inversion fine-tuning example for SDXL

```
export MODEL_NAME="stabilityai/stable-diffusion-xl-base-1.0"
export DATA_DIR="./cat"

accelerate launch textual_inversion_sdxl.py \
--pretrained_model_name_or_path=$MODEL_NAME \
--train_data_dir=$DATA_DIR \
--learnable_property="object" \
--placeholder_token="<cat-toy>" \
--initializer_token="toy" \
--mixed_precision="bf16" \
--resolution=768 \
--train_batch_size=1 \
--gradient_accumulation_steps=4 \
--max_train_steps=500 \
--learning_rate=5.0e-04 \
--scale_lr \
--lr_scheduler="constant" \
--lr_warmup_steps=0 \
--save_as_full_pipeline \
--output_dir="./textual_inversion_cat_sdxl"
```

For now, only training of the first text encoder is supported.
152 changes: 152 additions & 0 deletions examples/textual_inversion/test_textual_inversion_sdxl.py
Original file line number Diff line number Diff line change
@@ -0,0 +1,152 @@
# coding=utf-8
# Copyright 2023 HuggingFace Inc.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

import logging
import os
import sys
import tempfile


sys.path.append("..")
from test_examples_utils import ExamplesTestsAccelerate, run_command # noqa: E402


logging.basicConfig(level=logging.DEBUG)

logger = logging.getLogger()
stream_handler = logging.StreamHandler(sys.stdout)
logger.addHandler(stream_handler)


class TextualInversionSdxl(ExamplesTestsAccelerate):
def test_textual_inversion_sdxl(self):
with tempfile.TemporaryDirectory() as tmpdir:
test_args = f"""
examples/textual_inversion/textual_inversion_sdxl.py
--pretrained_model_name_or_path hf-internal-testing/tiny-sdxl-pipe
--train_data_dir docs/source/en/imgs
--learnable_property object
--placeholder_token <cat-toy>
--initializer_token a
--save_steps 1
--num_vectors 2
--resolution 64
--train_batch_size 1
--gradient_accumulation_steps 1
--max_train_steps 2
--learning_rate 5.0e-04
--scale_lr
--lr_scheduler constant
--lr_warmup_steps 0
--output_dir {tmpdir}
""".split()

run_command(self._launch_args + test_args)
# save_pretrained smoke test
self.assertTrue(os.path.isfile(os.path.join(tmpdir, "learned_embeds.safetensors")))

def test_textual_inversion_sdxl_checkpointing(self):
with tempfile.TemporaryDirectory() as tmpdir:
test_args = f"""
examples/textual_inversion/textual_inversion_sdxl.py
--pretrained_model_name_or_path hf-internal-testing/tiny-sdxl-pipe
--train_data_dir docs/source/en/imgs
--learnable_property object
--placeholder_token <cat-toy>
--initializer_token a
--save_steps 1
--num_vectors 2
--resolution 64
--train_batch_size 1
--gradient_accumulation_steps 1
--max_train_steps 3
--learning_rate 5.0e-04
--scale_lr
--lr_scheduler constant
--lr_warmup_steps 0
--output_dir {tmpdir}
--checkpointing_steps=1
--checkpoints_total_limit=2
""".split()

run_command(self._launch_args + test_args)

# check checkpoint directories exist
self.assertEqual(
{x for x in os.listdir(tmpdir) if "checkpoint" in x},
{"checkpoint-2", "checkpoint-3"},
)

def test_textual_inversion_sdxl_checkpointing_checkpoints_total_limit_removes_multiple_checkpoints(self):
with tempfile.TemporaryDirectory() as tmpdir:
test_args = f"""
examples/textual_inversion/textual_inversion_sdxl.py
--pretrained_model_name_or_path hf-internal-testing/tiny-sdxl-pipe
--train_data_dir docs/source/en/imgs
--learnable_property object
--placeholder_token <cat-toy>
--initializer_token a
--save_steps 1
--num_vectors 2
--resolution 64
--train_batch_size 1
--gradient_accumulation_steps 1
--max_train_steps 2
--learning_rate 5.0e-04
--scale_lr
--lr_scheduler constant
--lr_warmup_steps 0
--output_dir {tmpdir}
--checkpointing_steps=1
""".split()

run_command(self._launch_args + test_args)

# check checkpoint directories exist
self.assertEqual(
{x for x in os.listdir(tmpdir) if "checkpoint" in x},
{"checkpoint-1", "checkpoint-2"},
)

resume_run_args = f"""
examples/textual_inversion/textual_inversion_sdxl.py
--pretrained_model_name_or_path hf-internal-testing/tiny-sdxl-pipe
--train_data_dir docs/source/en/imgs
--learnable_property object
--placeholder_token <cat-toy>
--initializer_token a
--save_steps 1
--num_vectors 2
--resolution 64
--train_batch_size 1
--gradient_accumulation_steps 1
--max_train_steps 2
--learning_rate 5.0e-04
--scale_lr
--lr_scheduler constant
--lr_warmup_steps 0
--output_dir {tmpdir}
--checkpointing_steps=1
--resume_from_checkpoint=checkpoint-2
--checkpoints_total_limit=2
""".split()

run_command(self._launch_args + resume_run_args)

# check checkpoint directories exist
self.assertEqual(
{x for x in os.listdir(tmpdir) if "checkpoint" in x},
{"checkpoint-2", "checkpoint-3"},
)
Loading