See axolotl config

axolotl version: 0.4.1

adapter: lora
auto_find_batch_size: true
base_model: Korabbit/llama-2-ko-7b
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - e4baa04ca4e4fdd2_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/e4baa04ca4e4fdd2_train_data.json
  type:
    field_instruction: import_statement
    field_output: next_line
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: 3
eval_max_new_tokens: 128
eval_steps: 50
eval_table_size: null
flash_attention: false
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 8
gradient_checkpointing: false
group_by_length: true
hub_model_id: tuantmdev/7a0bb651-9014-440a-993f-305619a36545
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 1e-4
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 40
lora_alpha: 64
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 32
lora_target_linear: true
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 400
micro_batch_size: 2
mlflow_experiment_name: /tmp/e4baa04ca4e4fdd2_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 1
optimizer: adamw_bnb_8bit
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_steps: 50
save_strategy: steps
sequence_len: 512
special_tokens:
  pad_token: </s>
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: null
wandb_mode: online
wandb_name: e3d59494-2ee7-4cc1-b4d8-1ec645f12911
wandb_project: Gradients-On-Demand
wandb_run: unknown
wandb_runid: e3d59494-2ee7-4cc1-b4d8-1ec645f12911
warmup_steps: 80
weight_decay: 0.0
xformers_attention: null

7a0bb651-9014-440a-993f-305619a36545

This model is a fine-tuned version of Korabbit/llama-2-ko-7b on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.4848

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: 0.0001
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 16
• optimizer: Use OptimizerNames.ADAMW_BNB 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: 80
• training_steps: 400

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0007	1	4.7432
2.9064	0.0362	50	1.7389
1.6391	0.0724	100	1.6316
1.5544	0.1086	150	1.5727
1.4779	0.1448	200	1.5435
1.4556	0.1810	250	1.5171
1.4762	0.2172	300	1.5001
1.5009	0.2534	350	1.4819
1.4524	0.2896	400	1.4848

Framework versions

• PEFT 0.13.2
• Transformers 4.46.0
• Pytorch 2.5.0+cu124
• Datasets 3.0.1
• Tokenizers 0.20.1