cpu_oom_config.yaml

gistfile1.txt

checkpoints:
  checkpoint_interval: 10000
  checkpoints_path: /fsx/elie_bakouch/nanotron/debug-ckpt-cpuoom
  checkpoints_path_is_shared_file_system: false
  resume_checkpoint_path: null
  load_lr_scheduler: false
  load_optimizer: false
  save_final_state: true
  save_initial_state: true
data_stages:
- data:
    dataset:
      dataset_folder:
      - /fsx/elie_bakouch/data/fw-edu-dedup
    num_loading_workers: 0
    seed: 8
  name: stable phase
  start_training_step: 1
general:
  benchmark_csv_path: null
  consumed_train_samples: null
  ignore_sanity_checks: true
  project: llama3-3B-finetune
  run: fwedu-60B-resume
  seed: 6
  step: null
logging:
  iteration_step_info_interval: 1
  log_level: info
  log_level_replica: info
model:
  ddp_bucket_cap_mb: 25
  dtype: bfloat16
  init_method:
    std: 0.041666666666666664
  make_vocab_size_divisible_by: 1
  model_config:
    bos_token_id: 128000
    eos_token_id: 128001
    hidden_act: silu
    hidden_size: 3072
    initializer_range: 0.02
    intermediate_size: 8192
    is_llama_config: true
    max_position_embeddings: 4096
    num_attention_heads: 24
    num_hidden_layers: 28
    num_key_value_heads: 8
    pad_token_id: null
    pretraining_tp: 2
    rms_norm_eps: 1.0e-05
    rope_interleaved: false
    rope_scaling:
      factor: 32.0
      high_freq_factor: 4.0
      low_freq_factor: 1.0
      original_max_position_embeddings: 8192
      rope_type: llama3
    rope_theta: 500000.0
    tie_word_embeddings: true
    use_cache: true
    vocab_size: 128256
optimizer:
  accumulate_grad_in_fp32: true
  clip_grad: 1.0
  learning_rate_scheduler:
    learning_rate: 0.00005
    lr_decay_starting_step: 50000
    lr_decay_steps: 10000
    lr_decay_style: linear
    lr_warmup_steps: 1000
    lr_warmup_style: linear
    min_decay_lr: 0
  optimizer_factory:
    adam_beta1: 0.9
    adam_beta2: 0.95
    adam_eps: 1.0e-08
    name: adamW
    torch_adam_is_fused: true
  weight_decay: 0.01
  zero_stage: 1
parallelism:
  dp: 32
  expert_parallel_size: 1
  pp: 1
  pp_engine: 1f1b
  recompute_layer: false
  tp: 2
  tp_linear_async_communication: true
  tp_mode: REDUCE_SCATTER
  tp_recompute_allgather: true
profiler: null
tokenizer:
  tokenizer_max_length: null
  tokenizer_name_or_path: meta-llama/Llama-3.2-3B
  tokenizer_revision: null
tokens:
  batch_accumulation_per_replica: 2
  limit_test_batches: 0
  limit_val_batches: 0
  micro_batch_size: 4
  sequence_length: 4096
  train_steps: 120000
  val_check_interval: -1