base_model: $PATH_STORAGE/.cache/huggingface/hub/models--meta-llama--Meta-Llama-3.1-405B/snapshots/222de096204587406c7cadb3e0a101aade116279
tokenizer_type: AutoTokenizer

strict: false

plugins:
  - axolotl.integrations.liger.LigerPlugin
liger_rope: true
liger_rms_norm: true
liger_swiglu: true
liger_fused_linear_cross_entropy: true

chat_template: llama3
datasets:
  - path: mlabonne/FineTome-100k
    type: chat_template
    split: train

dataset_prepared_path: $PATH_STORAGE/axolotl/last_run_prepared
val_set_size: 0.0
output_dir: $PATH_STORAGE/axolotl-artifacts/outputs/llama3_1-405b-finetome
save_safetensors: false  # saving final sharded dict may not work with safetensors

wandb_project: llama-3.1-405b-fft-finetome
wandb_entity:

sequence_len: 4096
sample_packing: true
pad_to_sequence_len: true

gradient_accumulation_steps: 1
micro_batch_size: 1
num_epochs: 2
optimizer: adamw_torch
lr_scheduler: cosine
learning_rate: 1.0e-5

train_on_inputs: false
group_by_length: false
bf16: true
tf32: true

gradient_checkpointing: true
gradient_checkpointing_kwargs:
  use_reentrant: false
logging_steps: 1
flash_attention: true

warmup_steps: 100
saves_per_epoch: 1
weight_decay: 0.1
fsdp_final_state_dict_type: SHARDED_STATE_DICT
fsdp:
  - full_shard
  - auto_wrap
fsdp_config:
  fsdp_limit_all_gathers: true
  fsdp_sync_module_states: true
  fsdp_offload_params: true
  fsdp_use_orig_params: false
  fsdp_cpu_ram_efficient_loading: true
  fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
  fsdp_transformer_layer_cls_to_wrap: LlamaDecoderLayer
  fsdp_state_dict_type: SHARDED_STATE_DICT
  fsdp_sharding_strategy: FULL_SHARD
special_tokens:
  pad_token: <|finetune_right_pad_id|>
  eos_token: <|eot_id|>