Flux Lora in quanto error

[Desm0nt]

If I try to train Flux Dev lora with --base_model_precision=fp8-quanto or int8-quanto and --text_encoder_2_precision=fp8-quanto it fails to use any optimiser except adamw_bf16. But on old commit f2f8de305f58163379a2ecf5a95f7fa0cdf51fbe all works fine

Failed with this error

(.venv) root@C.11818744:~/SimpleTuner$ bash train.sh 
VALIDATION_NEGATIVE_PROMPT not set, defaulting to empty.
Your configuration is requesting an incompatible dtype and optimizer combination.
--base_model_default_dtype is set to bf16. You could resolve this by switching it to fp32, at the cost of more VRAM. This would be placed in TRAINER_EXTRA_ARGS.
--adam_bfloat16 could alternatively be provided to resolve the situation. Check your value for OPTIMIZER.
--mixed_precision is bf16, as expected. This value is referred to as MIXED_PRECISION in config.env.
Traceback (most recent call last):
  File "/root/SimpleTuner/train.py", line 2750, in <module>
    main()
  File "/root/SimpleTuner/train.py", line 269, in main
    args = parse_args()
  File "/root/SimpleTuner/helpers/arguments.py", line 1814, in parse_args
    raise ValueError(
ValueError: Your configuration is requesting an incompatible dtype and optimizer combination.
--base_model_default_dtype is set to bf16. You could resolve this by switching it to fp32, at the cost of more VRAM. This would be placed in TRAINER_EXTRA_ARGS.
--adam_bfloat16 could alternatively be provided to resolve the situation. Check your value for OPTIMIZER.
--mixed_precision is bf16, as expected. This value is referred to as MIXED_PRECISION in config.env.

config.env

# Configure these values.

# 'lora' or 'full'
# lora - train a small network for a character or style, or both. quite versatile.
# full - requires lots of vram, trains very slowly, needs a lot of data and concepts.
export MODEL_TYPE='lora'

# Set this to 'true' if you are training a Stable Diffusion 3 checkpoint.
# Use MODEL_NAME="stabilityai/stable-diffusion-3-medium-diffusers"
export STABLE_DIFFUSION_3=false
# Similarly, this is to train PixArt Sigma (1K or 2K) models.
# Use MODEL_NAME="PixArt-alpha/PixArt-Sigma-XL-2-1024-MS"
export PIXART_SIGMA=false
# For old Stable Diffusion 1.x/2.x models, you'll enable this.
# Use MODEL_NAME="stabilityai/stable-diffusion-2-1"
export STABLE_DIFFUSION_LEGACY=false
# For Kwai-Kolors, enable KOLORS.
# Use MODEL_NAME="kwai-kolors/kolors-diffusers"
export KOLORS=false
# For Flux, if you have 8 GPUs and DeepSpeed configured.
# Use MODEL_NAME="black-forest-labs/FLUX.1-dev"
export FLUX=true

# ControlNet model training is only supported when MODEL_TYPE='full'
# See this document for more information: https://github.com/bghira/SimpleTuner/blob/main/documentation/CONTROLNET.md
# DeepFloyd, PixArt, and SD3 do not currently support ControlNet model training.
export CONTROLNET=false

# DoRA enhances the training style of LoRA, but it will run more slowly at the same rank.
# See: https://arxiv.org/abs/2402.09353
# See: https://github.com/huggingface/peft/pull/1474
export USE_DORA=false

# BitFit freeze strategy for the u-net causes everything but the biases to be frozen.
# This may help retain the full model's underlying capabilities. LoRA is currently not tested/known to work.
#if [[ "$MODEL_TYPE" == "full" ]]; then
#    # When training a full model, we will rely on BitFit to keep the u-net intact.
#    export USE_BITFIT=true
#elif [[ "$MODEL_TYPE" == "lora" ]]; then
#    # LoRA can not use BitFit.
#    export USE_BITFIT=false
#elif [[ "$MODEL_TYPE" == "deepfloyd-full" ]]; then
#    export USE_BITFIT=true
#fi

# Restart where we left off. Change this to "checkpoint-1234" to start from a specific checkpoint.
export RESUME_CHECKPOINT="latest"

# How often to checkpoint. Depending on your learning rate, you may wish to change this.
# For the default settings with 10 gradient accumulations, more frequent checkpoints might be preferable at first.
export CHECKPOINTING_STEPS=214
# This is how many checkpoints we will keep. Two is safe, but three is safer.
export CHECKPOINTING_LIMIT=12

# This is decided as a relatively conservative 'constant' learning rate.
# Adjust higher or lower depending on how burnt your model becomes.
export LEARNING_RATE=2.5e-5 #@param {type:"number"}

# Using a Huggingface Hub model:
#export MODEL_NAME="black-forest-labs/FLUX.1-dev"
# Using a local path to a huggingface hub model or saved checkpoint:
export MODEL_NAME="/root/SimpleTuner/FLUX.1-dev"

# Make DEBUG_EXTRA_ARGS empty to disable wandb.
export DEBUG_EXTRA_ARGS="--report_to=wandb"
export TRACKER_PROJECT_NAME="sdxl-training"
export TRACKER_RUN_NAME="simpletuner-flux"

# Max number of steps OR epochs can be used. Not both.
export MAX_NUM_STEPS=0
# Will likely overtrain, but that's fine.
export NUM_EPOCHS=15

# A convenient prefix for all of your training paths.
# These may be absolute or relative paths. Here, we are using relative paths.
# The output will just be in a folder called "output/models" by default.
export DATALOADER_CONFIG="config/multidatabackend.json"
export OUTPUT_DIR="output/models"

# Set this to "true" to push your model to Hugging Face Hub.
export PUSH_TO_HUB="false"
# If PUSH_TO_HUB and PUSH_CHECKPOINTS are both enabled, every saved checkpoint will be pushed to Hugging Face Hub.
export PUSH_CHECKPOINTS="true"
# This will be the model name for your final hub upload, eg. "yourusername/yourmodelname"
# It defaults to the wandb project name, but you can override this here.
export HUB_MODEL_NAME=$TRACKER_PROJECT_NAME

# By default, images will be resized so their SMALLER EDGE is 1024 pixels, maintaining aspect ratio.
# Setting this value to 768px might result in more reasonable training data sizes for SDXL.
#export RESOLUTION=1024
# If you want to have the training data resized by pixel area (Megapixels) rather than edge length,
#  set this value to "area" instead of "pixel", and uncomment the next RESOLUTION declaration.
export RESOLUTION_TYPE="area"
export RESOLUTION=1          # 1.0 Megapixel training sizes
# If RESOLUTION_TYPE="pixel", the minimum resolution specifies the smaller edge length, measured in pixels. Recommended: 1024.
# If RESOLUTION_TYPE="area", the minimum resolution specifies the total image area, measured in megapixels. Recommended: 1.
export MINIMUM_RESOLUTION=$RESOLUTION

# How many decimals to round aspect buckets to.
#export ASPECT_BUCKET_ROUNDING=2

# Use this to append an instance prompt to each caption, used for adding trigger words.
# This has not been tested in SDXL.
#export INSTANCE_PROMPT="lotr style "
# If you also supply a user prompt library or `--use_prompt_library`, this will be added to those lists.
export VALIDATION_PROMPT="art by Raichiyo33, A drawing of woman with long brown hair, wearing elbow glove and a revealing bondage corset showing off her cleavage and shapely figure, poses in front of a yellow background. Her eyes are looking to the left, and she appears to be looking directly at the viewer, holding bag behind her."
export VALIDATION_GUIDANCE=2.7
# You'll want to set this to 0.7 if you are training a terminal SNR model.
export VALIDATION_GUIDANCE_RESCALE=0.0
# How frequently we will save and run a pipeline for validations.
export VALIDATION_STEPS=100
export VALIDATION_NUM_INFERENCE_STEPS=30
export VALIDATION_NEGATIVE_PROMPT=""
export VALIDATION_SEED=42
export VALIDATION_RESOLUTION=1024x1024

# Adjust this for your GPU memory size. This, and resolution, are the biggest VRAM killers.
export TRAIN_BATCH_SIZE=1
# Accumulate your update gradient over many steps, to save VRAM while still having higher effective batch size:
# effective batch size = ($TRAIN_BATCH_SIZE * $GRADIENT_ACCUMULATION_STEPS).
export GRADIENT_ACCUMULATION_STEPS=2

# Use any standard scheduler type. constant, polynomial, constant_with_warmup
export LR_SCHEDULE="constant_with_warmup"
# A warmup period allows the model and the EMA weights more importantly to familiarise itself with the current quanta.
# For the cosine or sine type schedules, the warmup period defines the interval between peaks or valleys.
# Use a sine schedule to simulate a warmup period, or a Cosine period to simulate a polynomial start.
#export LR_WARMUP_STEPS=$((MAX_NUM_STEPS / 10))
export LR_WARMUP_STEPS=100

# Caption dropout probability. Set to 0.1 for 10% of captions dropped out. Set to 0 to disable.
# You may wish to disable dropout if you want to limit your changes strictly to the prompts you show the model.
# You may wish to increase the rate of dropout if you want to more broadly adopt your changes across the model.
export CAPTION_DROPOUT_PROBABILITY=0.1

export METADATA_UPDATE_INTERVAL=65

# How many workers to use for VAE caching.
export MAX_WORKERS=1
# Read and write batch sizes for VAE caching.
export READ_BATCH_SIZE=1
export WRITE_BATCH_SIZE=1
# How many images to encode at once with the VAE. Can increase VRAM use.
export VAE_BATCH_SIZE=1
# How many images to process at once (resize, crop, transform) during VAE caching.
export IMAGE_PROCESSING_BATCH_SIZE=1
# When using large batch sizes, you'll need to increase the pool connection limit.
export AWS_MAX_POOL_CONNECTIONS=128
# For very large systems, setting this can reduce CPU overhead of torch spawning an unnecessarily large number of threads.
export TORCH_NUM_THREADS=8

# If this is set, any images that fail to open will be DELETED to avoid re-checking them every time.
export DELETE_ERRORED_IMAGES=0
# If this is set, any images that are too small for the minimum resolution size will be DELETED.
export DELETE_SMALL_IMAGES=0

# Bytedance recommends these be set to "trailing" so that inference and training behave in a more congruent manner.
# To follow the original SDXL training strategy, use "leading" instead, though results are generally worse.
export TRAINING_SCHEDULER_TIMESTEP_SPACING="trailing"
export INFERENCE_SCHEDULER_TIMESTEP_SPACING="trailing"

# Removing this option or unsetting it uses vanilla training. Setting it reweights the loss by the position of the timestep in the noise schedule.
# A value "5" is recommended by the researchers. A value of "20" is the least impact, and "1" is the most impact.
export MIN_SNR_GAMMA=5

# Set this to an explicit value of "false" to disable Xformers. Probably required for AMD users.
export USE_XFORMERS=true

# There's basically no reason to unset this. However, to disable it, use an explicit value of "false".
# This will save a lot of memory consumption when enabled.
export USE_GRADIENT_CHECKPOINTING=true

##
# Options below here may require a bit more complicated configuration, so they are not simple variables.
##

# TF32 is great on Ampere or Ada, not sure about earlier generations.
export ALLOW_TF32=true
# AdamW 8Bit is a robust and lightweight choice. Adafactor might reduce memory consumption, and Dadaptation is slow and experimental.
# AdamW is the default optimizer, but it uses a lot of memory and is slower than AdamW8Bit or Adafactor.
# When training a quantised base model, you can't use adamw_bf16. Instead, try adafactor or adamw.
# Choices: adamw, adamw8bit, adafactor, dadaptation, adamw_bf16
export OPTIMIZER="adamw8bit"

# EMA is a strong regularisation method that uses a lot of extra VRAM to hold two copies of the weights.
# This is worthwhile on large training runs, but not so much for smaller training runs.
export USE_EMA=false
export EMA_DECAY=0.999

export TRAINER_EXTRA_ARGS="--lora_rank=4 --base_model_precision=int8-quanto --keep_vae_loaded"
## For offset noise training:
# Not recommended for terminal SNR models.
export TRAINER_EXTRA_ARGS="${TRAINER_EXTRA_ARGS} --text_encoder_1_precision=no_change --text_encoder_2_precision=fp8-quanto"

## For terminal SNR training:
#export TRAINER_EXTRA_ARGS="${TRAINER_EXTRA_ARGS} --prediction_type=v_prediction --rescale_betas_zero_snr"
#export TRAINER_EXTRA_ARGS="${TRAINER_EXTRA_ARGS} --training_scheduler_timestep_spacing=trailing --inference_scheduler_timestep_spacing=trailing"
## You may benefit from directing training toward a specific weighted subset of timesteps.
# In this example, we train the final 25% of the timestep schedule with a 3x bias.
#export TRAINER_EXTRA_ARGS="${TRAINER_EXTRA_ARGS} --timestep_bias_strategy=later --timestep_bias_portion=0.25 --timestep_bias_multiplier=3"
# In this example, we train the earliest 25% of the timestep schedule with a 5x bias.
#export TRAINER_EXTRA_ARGS="${TRAINER_EXTRA_ARGS} --timestep_bias_strategy=earlier --timestep_bias_portion=0.25 --timestep_bias_multiplier=5"
# Here, we designate that specifically, timesteps 200 to 500 should be prioritised.
#export TRAINER_EXTRA_ARGS="${TRAINER_EXTRA_ARGS} --timestep_bias_strategy=range --timestep_bias_begin=200 --timestep_bias_end=500 --timestep_bias_multiplier=3"

## For experimental min-SNR weighted loss training (5 is suggested value by the original researchers):
# Not recommended for terminal SNR models.
#export TRAINER_EXTRA_ARGS="${TRAINER_EXTRA_ARGS} --snr_gamma=5.0"

# For Wasabi S3 filesystem backend (experimental)
#export TRAINER_EXTRA_ARGS="${TRAINER_EXTRA_ARGS} --data_backend=aws --aws_bucket_name=test123"
#export TRAINER_EXTRA_ARGS="${TRAINER_EXTRA_ARGS} --aws_endpoint_url=https://s3.wasabisys.com"
#export TRAINER_EXTRA_ARGS="${TRAINER_EXTRA_ARGS} --aws_access_key=1234567890"
#export TRAINER_EXTRA_ARGS="${TRAINER_EXTRA_ARGS} --aws_secret_access_key=0987654321"

# Reproducible training. Set to -1 to disable.
export TRAINING_SEED=42

# Mixed precision is the best. You honestly might need to YOLO it in fp16 mode for Google Colab type setups.
export MIXED_PRECISION="bf16"                # Might not be supported on all GPUs. fp32 will be needed for others.
export PURE_BF16=true

# This has to be changed if you're training with multiple GPUs.
export TRAINING_NUM_PROCESSES=1
export TRAINING_NUM_MACHINES=1
export ACCELERATE_EXTRA_ARGS=""                          # --multi_gpu or other similar flags for huggingface accelerate

# With Pytorch 2.1, you might have pretty good luck here.
# If you're using aspect bucketing however, each resolution change will recompile. Seriously, just don't do it.
# Well, then again... Pytorch 2.2 has support for dynamic shapes. Why not?
export TRAINING_DYNAMO_BACKEND='no'                # or 'no' if you want to disable torch compile in case of performance issues or lack of support (eg. AMD)

export TOKENIZERS_PARALLELISM=false

[bghira]

the error message explains it