Jiminiya/tiny_step1

See axolotl config

axolotl version: 0.6.0

#base_model: meta-llama/Llama-3.1-8B
base_model: TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T
model_type: LlamaForCausalLM
tokenizer_type: AutoTokenizer
special_tokens:
  pad_token: "</s>"

load_in_8bit: false
load_in_4bit: false
strict: false
#
# max_steps:학습할 step. 논문 상에서는 400이라고 표기
# 단, 학습 환경의 테스트를 위해 50으로 변경
max_steps: 180
pretraining_dataset:
  - path: Jiminiya/INU
    type: pretrain
val_set_size: 0.0
output_dir: ./outputs_continue_1
#dataset_prepared_path:

unfrozen_parameters:
#   - ^lm_head.*
    - ^model.embed_tokens.weight
#     ^model.layer.*

sequence_len: 2048
sample_packing: false
pad_to_sequence_len: false

wandb_project:
wandb_entity:
wandb_watch:
wandb_name:
wandb_log_model:

# gradient_accumulation_steps: 4
# micro_batch_size: 8
gradient_accumulation_steps: 4
micro_batch_size: 2
optimizer: adamw_torch
lr_scheduler: cosine
learning_rate: 4e-5

train_on_inputs: false
group_by_length: false
bf16: auto
fp16:
tf32:

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

warmup_steps: 10
save_steps: 200
debug:
deepspeed:
weight_decay: 0.0
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: true
  fsdp_cpu_ram_efficient_loading: false
  fsdp_auto_wrap_policy: TRANSFORMER_BASED_WRAP
  fsdp_transformer_layer_cls_to_wrap: LlamaDecoderLayer
  fsdp_state_dict_type: FULL_STATE_DICT
  fsdp_sharding_strategy: FULL_SHARD
  fsdp_backward_prefetch: BACKWARD_POST

outputs_continue_1

This model is a fine-tuned version of TinyLlama/TinyLlama-1.1B-intermediate-step-1431k-3T on an unknown dataset.

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 4e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• total_eval_batch_size: 8
• optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 10
• training_steps: 180

Training results

Framework versions

• Transformers 4.48.1
• Pytorch 2.5.1+cu124
• Datasets 3.2.0
• Tokenizers 0.21.0