adapter: lora
base_model: elyza/Llama-3-ELYZA-JP-8B
bf16: auto
chat_template: llama3
dataset_prepared_path: /workspace/axolotl/data/prepared
datasets:
- ds_type: json
  format: custom
  path: Paladiso/dataset_bdc5d8b7-c291-4772-a9f7-28e2904ac8b5
  type:
    field_instruction: question
    field_output: answer
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: null
eval_max_new_tokens: 128
eval_table_size: null
evals_per_epoch: 4
flash_attention: false
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: false
group_by_length: false
hub_model_id: Paladiso/f88ae1b3-ee70-47d5-93e0-fab71678d2de
hub_private_repo: true
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: 1
lora_alpha: 16
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 8
lora_target_linear: true
lr_scheduler: cosine
max_steps: 10
micro_batch_size: 2
model_type: AutoModelForCausalLM
num_epochs: 1
optimizer: adamw_bnb_8bit
output_dir: /workspace/axolotl/outputs
pad_to_sequence_len: true
push_to_hub: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
save_safetensors: true
saves_per_epoch: 4
sequence_len: 512
special_tokens:
  pad_token: <|eot_id|>
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
use_accelerate: true
val_set_size: 0.05
wandb_entity: null
wandb_mode: online
wandb_name: bdc5d8b7-c291-4772-a9f7-28e2904ac8b5
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: bdc5d8b7-c291-4772-a9f7-28e2904ac8b5
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

f88ae1b3-ee70-47d5-93e0-fab71678d2de

This model is a fine-tuned version of elyza/Llama-3-ELYZA-JP-8B on the Paladiso/dataset_bdc5d8b7-c291-4772-a9f7-28e2904ac8b5 dataset. It achieves the following results on the evaluation set:

Loss: 2.2106

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: 2
eval_batch_size: 2
seed: 42
gradient_accumulation_steps: 4
total_train_batch_size: 8
optimizer: Use adamw_bnb_8bit 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: 10

Training results

Training Loss	Epoch	Step	Validation Loss
2.3538	0.0051	3	2.4673
2.511	0.0101	6	2.4055
2.7125	0.0152	9	2.2106

Framework versions

PEFT 0.14.0
Transformers 4.47.1
Pytorch 2.5.0a0+e000cf0ad9.nv24.10
Datasets 3.1.0
Tokenizers 0.21.0