76 lines
1.6 KiB
YAML
76 lines
1.6 KiB
YAML
base_model: google/gemma-3n-E2B-it
|
|
processor_type: AutoProcessor
|
|
|
|
# Automatically upload checkpoint and final model to HF
|
|
# hub_model_id: username/custom_model_name
|
|
|
|
plugins:
|
|
- axolotl.integrations.cut_cross_entropy.CutCrossEntropyPlugin
|
|
cut_cross_entropy: true
|
|
|
|
# for use with fft to only train on language model layers
|
|
# unfrozen_parameters:
|
|
# - model.language_model.*
|
|
# - lm_head
|
|
# - embed_tokens
|
|
|
|
load_in_4bit: true
|
|
|
|
# these 3 lines are needed for now to handle vision chat templates w images
|
|
skip_prepare_dataset: true
|
|
remove_unused_columns: false
|
|
sample_packing: false
|
|
|
|
# gemma3 doesn't seem to play nice with ddp
|
|
ddp_find_unused_parameters: true
|
|
|
|
chat_template: gemma3n
|
|
eot_tokens:
|
|
- <end_of_turn>
|
|
datasets:
|
|
- path: HuggingFaceH4/llava-instruct-mix-vsft
|
|
type: chat_template
|
|
split: train[:1%]
|
|
dataset_prepared_path:
|
|
val_set_size: 0.01
|
|
output_dir: ./outputs/out
|
|
|
|
adapter: qlora
|
|
lora_model_dir:
|
|
|
|
sequence_len: 2048
|
|
pad_to_sequence_len: false
|
|
|
|
lora_r: 32
|
|
lora_alpha: 16
|
|
lora_dropout: 0.05
|
|
lora_target_modules: 'model.language_model.layers.[\d]+.(mlp|self_attn).(up|down|gate|q|k|v|o)_proj'
|
|
|
|
wandb_project:
|
|
wandb_entity:
|
|
wandb_watch:
|
|
wandb_name:
|
|
wandb_log_model:
|
|
|
|
gradient_accumulation_steps: 4
|
|
micro_batch_size: 2
|
|
num_epochs: 1
|
|
optimizer: adamw_bnb_8bit
|
|
lr_scheduler: cosine
|
|
learning_rate: 0.0002
|
|
|
|
bf16: true
|
|
fp16:
|
|
tf32: true
|
|
|
|
gradient_checkpointing: true
|
|
gradient_checkpointing_kwargs:
|
|
use_reentrant: false
|
|
logging_steps: 1
|
|
# flash_attention: true # Any attention impl does not work with gemma3n now
|
|
|
|
warmup_ratio: 0.1
|
|
evals_per_epoch: 1
|
|
saves_per_epoch: 1
|
|
weight_decay: 0.0
|