See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: elyza/Llama-3-ELYZA-JP-8B
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - a3c5d8b320c8afde_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/a3c5d8b320c8afde_train_data.json
  type:
    field_instruction: question
    field_output: answer
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
do_eval: true
early_stopping_patience: null
eval_max_new_tokens: 128
eval_strategy: steps
eval_table_size: null
evals_per_epoch: 4
flash_attention: false
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
group_by_length: false
hub_model_id: eddysang/e67de51e-8d46-4281-a590-74edc07323ee
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0002
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 5
lora_alpha: 128
lora_dropout: 0.1
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 64
lora_target_linear: true
lora_target_modules:
- q_proj
- k_proj
- v_proj
- o_proj
- gate_proj
- down_proj
- up_proj
lr_scheduler: cosine
max_grad_norm: 1
max_steps: 100
micro_batch_size: 8
mlflow_experiment_name: /tmp/a3c5d8b320c8afde_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 3
optim_args:
  adam_beta1: 0.9
  adam_beta2: 0.95
  adam_epsilon: 2.0e-05
optimizer: adamw_torch
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
saves_per_epoch: 4
sequence_len: 1024
special_tokens:
  pad_token: <|eot_id|>
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.1
wandb_entity: yaudayah0
wandb_mode: online
wandb_name: df6bf9d4-a987-43b4-acb3-402a434df76e
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: df6bf9d4-a987-43b4-acb3-402a434df76e
warmup_steps: 20
weight_decay: 0.02
xformers_attention: false

e67de51e-8d46-4281-a590-74edc07323ee

This model is a fine-tuned version of elyza/Llama-3-ELYZA-JP-8B on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.5872

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: 0.0002
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 32
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=adam_beta1=0.9,adam_beta2=0.95,adam_epsilon=2e-05
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 20
• training_steps: 100

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0002	1	1.3243
1.262	0.0017	9	0.9494
0.8572	0.0033	18	0.7848
0.7654	0.0050	27	0.7108
0.6711	0.0067	36	0.6733
0.6452	0.0083	45	0.6472
0.6194	0.0100	54	0.6276
0.6104	0.0117	63	0.6112
0.6108	0.0133	72	0.6009
0.5846	0.0150	81	0.5924
0.591	0.0166	90	0.5882
0.6007	0.0183	99	0.5872

Framework versions

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