xtuner_config.py

SYSTEM = 'xtuner.utils.SYSTEM_TEMPLATE.alpaca'
accumulative_counts = 16
alpaca_en = dict(
    dataset=dict(
        data_files=
        '/petrobr/parceirosbr/home/rafael.rodrigues/.cache/huggingface/hub/datasets--recogna-nlp--ultra-alpaca-ptbr/snapshots/e69900d074177d370a911096fc30bdf407eff666/train.json',
        path='json',
        type='datasets.load_dataset'),
    dataset_map_fn='xtuner.dataset.map_fns.ultracabrita_map_fn',
    max_length=2048,
    pack_to_max_length=True,
    remove_unused_columns=True,
    shuffle_before_pack=True,
    template_map_fn=dict(
        template='xtuner.utils.PROMPT_TEMPLATE.gemma',
        type='xtuner.dataset.map_fns.template_map_fn_factory'),
    tokenizer=dict(
        padding_side='right',
        pretrained_model_name_or_path=
        '/petrobr/parceirosbr/home/rafael.rodrigues/.cache/huggingface/hub/models--google--gemma-2b/snapshots/dd006781ade50f5a5216ef690c2e30e7eedf1676/',
        trust_remote_code=True,
        type='transformers.AutoTokenizer.from_pretrained'),
    type='xtuner.dataset.process_hf_dataset',
    use_varlen_attn=False)
alpaca_en_path = '/petrobr/parceirosbr/home/rafael.rodrigues/.cache/huggingface/hub/datasets--recogna-nlp--ultra-alpaca-ptbr/snapshots/e69900d074177d370a911096fc30bdf407eff666/train.json'
batch_size = 1
betas = (
    0.9,
    0.999,
)
custom_hooks = [
    dict(
        tokenizer=dict(
            padding_side='right',
            pretrained_model_name_or_path=
            '/petrobr/parceirosbr/home/rafael.rodrigues/.cache/huggingface/hub/models--google--gemma-2b/snapshots/dd006781ade50f5a5216ef690c2e30e7eedf1676/',
            trust_remote_code=True,
            type='transformers.AutoTokenizer.from_pretrained'),
        type='xtuner.engine.hooks.DatasetInfoHook'),
    dict(
        evaluation_inputs=[
            'O que é um bode?',
            'Qual a capital da França?',
            'Você pode me explicar o Teorema de Pitágoras com um exemplo, por favor?',
            'Olá, tudo bem? Estou procurando livros de ficção científica para ler, você teria sugestões para mim?',
            'Resolva a equação de segundo grau x² - x - 30 = 0',
            'Escreva um código em python para calcular x^y usando divisão e conquista.',
        ],
        every_n_iters=500,
        prompt_template='xtuner.utils.PROMPT_TEMPLATE.gemma',
        system='xtuner.utils.SYSTEM_TEMPLATE.alpaca',
        tokenizer=dict(
            padding_side='right',
            pretrained_model_name_or_path=
            '/petrobr/parceirosbr/home/rafael.rodrigues/.cache/huggingface/hub/models--google--gemma-2b/snapshots/dd006781ade50f5a5216ef690c2e30e7eedf1676/',
            trust_remote_code=True,
            type='transformers.AutoTokenizer.from_pretrained'),
        type='xtuner.engine.hooks.EvaluateChatHook'),
]
dataloader_num_workers = 0
default_hooks = dict(
    checkpoint=dict(
        by_epoch=False,
        interval=500,
        max_keep_ckpts=2,
        type='mmengine.hooks.CheckpointHook'),
    logger=dict(
        interval=10,
        log_metric_by_epoch=False,
        type='mmengine.hooks.LoggerHook'),
    param_scheduler=dict(type='mmengine.hooks.ParamSchedulerHook'),
    sampler_seed=dict(type='mmengine.hooks.DistSamplerSeedHook'),
    timer=dict(type='mmengine.hooks.IterTimerHook'))
env_cfg = dict(
    cudnn_benchmark=False,
    dist_cfg=dict(backend='nccl'),
    mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0))
evaluation_freq = 500
evaluation_inputs = [
    'O que é um bode?',
    'Qual a capital da França?',
    'Você pode me explicar o Teorema de Pitágoras com um exemplo, por favor?',
    'Olá, tudo bem? Estou procurando livros de ficção científica para ler, você teria sugestões para mim?',
    'Resolva a equação de segundo grau x² - x - 30 = 0',
    'Escreva um código em python para calcular x^y usando divisão e conquista.',
]
launcher = 'pytorch'
load_from = None
log_level = 'INFO'
log_processor = dict(by_epoch=False)
lr = 2e-05
max_epochs = 1
max_length = 2048
max_norm = 1
model = dict(
    llm=dict(
        pretrained_model_name_or_path=
        '/petrobr/parceirosbr/home/rafael.rodrigues/.cache/huggingface/hub/models--google--gemma-2b/snapshots/dd006781ade50f5a5216ef690c2e30e7eedf1676/',
        trust_remote_code=True,
        type='transformers.AutoModelForCausalLM.from_pretrained'),
    type='xtuner.model.SupervisedFinetune',
    use_varlen_attn=False)
optim_type = 'torch.optim.AdamW'
optim_wrapper = dict(
    optimizer=dict(
        betas=(
            0.9,
            0.999,
        ),
        lr=2e-05,
        type='torch.optim.AdamW',
        weight_decay=0),
    type='DeepSpeedOptimWrapper')
pack_to_max_length = True
param_scheduler = [
    dict(
        begin=0,
        by_epoch=True,
        convert_to_iter_based=True,
        end=0.03,
        start_factor=1e-05,
        type='mmengine.optim.LinearLR'),
    dict(
        begin=0.03,
        by_epoch=True,
        convert_to_iter_based=True,
        end=1,
        eta_min=0.0,
        type='mmengine.optim.CosineAnnealingLR'),
]
pretrained_model_name_or_path = '/petrobr/parceirosbr/home/rafael.rodrigues/.cache/huggingface/hub/models--google--gemma-2b/snapshots/dd006781ade50f5a5216ef690c2e30e7eedf1676/'
prompt_template = 'xtuner.utils.PROMPT_TEMPLATE.gemma'
randomness = dict(deterministic=False, seed=None)
resume = False
runner_type = 'FlexibleRunner'
save_steps = 500
save_total_limit = 2
strategy = dict(
    config=dict(
        bf16=dict(enabled=False),
        fp16=dict(enabled=True, initial_scale_power=16),
        gradient_accumulation_steps='auto',
        gradient_clipping='auto',
        train_micro_batch_size_per_gpu='auto',
        zero_allow_untested_optimizer=True,
        zero_force_ds_cpu_optimizer=False,
        zero_optimization=dict(overlap_comm=True, stage=2)),
    exclude_frozen_parameters=True,
    gradient_accumulation_steps=16,
    gradient_clipping=1,
    sequence_parallel_size=1,
    train_micro_batch_size_per_gpu=1,
    type='xtuner.engine.DeepSpeedStrategy')
tokenizer = dict(
    padding_side='right',
    pretrained_model_name_or_path=
    '/petrobr/parceirosbr/home/rafael.rodrigues/.cache/huggingface/hub/models--google--gemma-2b/snapshots/dd006781ade50f5a5216ef690c2e30e7eedf1676/',
    trust_remote_code=True,
    type='transformers.AutoTokenizer.from_pretrained')
train_cfg = dict(max_epochs=1, type='xtuner.engine.runner.TrainLoop')
train_dataloader = dict(
    batch_size=1,
    collate_fn=dict(
        type='xtuner.dataset.collate_fns.default_collate_fn',
        use_varlen_attn=False),
    dataset=dict(
        dataset=dict(
            data_files=
            '/petrobr/parceirosbr/home/rafael.rodrigues/.cache/huggingface/hub/datasets--recogna-nlp--ultra-alpaca-ptbr/snapshots/e69900d074177d370a911096fc30bdf407eff666/train.json',
            path='json',
            type='datasets.load_dataset'),
        dataset_map_fn='xtuner.dataset.map_fns.ultracabrita_map_fn',
        max_length=2048,
        pack_to_max_length=True,
        remove_unused_columns=True,
        shuffle_before_pack=True,
        template_map_fn=dict(
            template='xtuner.utils.PROMPT_TEMPLATE.gemma',
            type='xtuner.dataset.map_fns.template_map_fn_factory'),
        tokenizer=dict(
            padding_side='right',
            pretrained_model_name_or_path=
            '/petrobr/parceirosbr/home/rafael.rodrigues/.cache/huggingface/hub/models--google--gemma-2b/snapshots/dd006781ade50f5a5216ef690c2e30e7eedf1676/',
            trust_remote_code=True,
            type='transformers.AutoTokenizer.from_pretrained'),
        type='xtuner.dataset.process_hf_dataset',
        use_varlen_attn=False),
    num_workers=0,
    sampler=dict(shuffle=True, type='mmengine.dataset.DefaultSampler'))
use_varlen_attn = False
visualizer = None
warmup_ratio = 0.03
weight_decay = 0
work_dir = './work_dirs/gemma_2b_full_ultraalpaca'