7B_lora_single_device.yaml

# Config for single device full finetuning in full_finetune_single_device.py
# using a Code-Llama2 7B model
#
# This config assumes that you've run the following command before launching
# this run:
#     tune download codellama/CodeLlama-7b-hf --output-dir /tmp/CodeLlama-7b-hf
#
# To launch on a single device, run the following command from root:
#   tune run lora_finetune_single_device --config code_llama2/7B_lora_single_device
#
# You can add specific overrides through the command line. For example
# to override the checkpointer directory while launching training
# you can run:
#   tune run lora_finetune_single_device --config code_llama2/7B_lora_single_device checkpointer.checkpoint_dir=<YOUR_CHECKPOINT_DIR>
#
# This config works only for training on single device.

# Model Arguments
model:
  _component_: torchtune.models.code_llama2.lora_code_llama2_7b
  lora_attn_modules: ['q_proj', 'v_proj']
  apply_lora_to_mlp: False
  apply_lora_to_output: False
  lora_rank: 8
  lora_alpha: 16

# Tokenizer
tokenizer:
  _component_: torchtune.models.llama2.llama2_tokenizer
  path: /tmp/CodeLlama-7b-hf/tokenizer.model

# Dataset
dataset:
  _component_: torchtune.datasets.alpaca_cleaned_dataset
seed: null
shuffle: True


checkpointer:
  _component_: torchtune.utils.FullModelHFCheckpointer
  checkpoint_dir: /tmp/CodeLlama-7b-hf
  checkpoint_files: [
    pytorch_model-00001-of-00003.bin,
    pytorch_model-00002-of-00003.bin,
    pytorch_model-00003-of-00003.bin
  ]
  adapter_checkpoint: null
  recipe_checkpoint: null
  output_dir: /tmp/CodeLlama-7b-hf
  model_type: LLAMA2

# Fine-tuning arguments
batch_size: 2
epochs: 1
max_steps_per_epoch: null
gradient_accumulation_steps: 64
compile: False

optimizer:
  _component_: torch.optim.AdamW
  weight_decay: 0.01
  lr: 3e-4

lr_scheduler:
  _component_: torchtune.modules.get_cosine_schedule_with_warmup
  num_warmup_steps: 100

loss:
  _component_: torch.nn.CrossEntropyLoss


# Training environment
device: cuda
enable_activation_checkpointing: True
dtype: bf16

# Logging
metric_logger:
  _component_: torchtune.utils.metric_logging.DiskLogger
  log_dir: ${output_dir}
output_dir: /tmp/lora_code_llama2_finetune_output
log_every_n_steps: 1
log_peak_memory_stats: False

profiler:
  _component_: torchtune.utils.setup_torch_profiler
  enabled: False

  #Output directory of trace artifacts
  output_dir: ${output_dir}/profiling_outputs

  #`torch.profiler.ProfilerActivity` types to trace
  cpu: True
  cuda: True

  #trace options passed to `torch.profiler.profile`
  profile_memory: False
  with_stack: False
  record_shapes: True
  with_flops: False

  # `torch.profiler.schedule` options:
  # wait_steps -> wait, warmup_steps -> warmup, active_steps -> active, num_cycles -> repeat
  wait_steps: 5
  warmup_steps: 5
  active_steps: 2
  num_cycles: 1