base_model: ./NeverSleep_Noromaid-13b-v0.1.1/
model_type: LlamaForCausalLM
tokenizer_type: LlamaTokenizer
is_llama_derived_model: true

load_in_8bit: true
load_in_4bit: false
strict: false

datasets:
  - path: lol/LRP.jsonl
    type: completion
dataset_prepared_path:
val_set_size: 0.05
output_dir: ./lora-out5

sequence_len: 4096
sample_packing: false
pad_to_sequence_len: false

adapter: lora
lora_model_dir:
lora_r: 256
lora_alpha: 512
lora_dropout: 0.05
lora_target_linear: true
lora_fan_in_fan_out:
lora_target_modules:
  - gate_proj
  - down_proj
  - up_proj
  - q_proj
  - v_proj
  - k_proj
  - o_proj

wandb_project: leyleytest4-noro
wandb_entity:
wandb_watch:
wandb_run_id:
wandb_log_model:

gradient_accumulation_steps: 2
micro_batch_size: 2
num_epochs: 50
optimizer: adamw_bnb_8bit
lr_scheduler: constant
learning_rate: 0.00000025

train_on_inputs: true
group_by_length: false
bf16: true
fp16: false
tf32: true

gradient_checkpointing: true
early_stopping_patience:
resume_from_checkpoint:
local_rank:
logging_steps: 1
xformers_attention:
flash_attention: true

warmup_steps: 10
eval_steps: 0.05
eval_table_size:
eval_table_max_new_tokens: 128
eval_sample_packing: false
save_steps: 65
debug:
deepspeed:
weight_decay: 0.0
fsdp:
fsdp_config:
special_tokens:
  bos_token: "<s>"
  eos_token: "</s>"
  unk_token: "<unk>"