See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: Eurdem/Defne_llama3_2x8B
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 026982362239f492_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/026982362239f492_train_data.json
  type:
    field_instruction: inputs
    field_output: targets
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
device_map:
  ? ''
  : 0,1,2,3,4,5,6,7
early_stopping_patience: 2
eval_max_new_tokens: 128
eval_steps: 100
eval_table_size: null
flash_attention: true
gradient_accumulation_steps: 8
gradient_checkpointing: true
group_by_length: false
hub_model_id: Alphatao/a838f1d2-844b-4887-a1f8-7528ffbe9df7
hub_repo: null
hub_strategy: null
hub_token: null
learning_rate: 0.0002
load_best_model_at_end: true
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 1
lora_alpha: 32
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 16
lora_target_linear: true
lora_target_modules:
- q_proj
- k_proj
- v_proj
- o_proj
- down_proj
- up_proj
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 1092
micro_batch_size: 4
mlflow_experiment_name: /tmp/026982362239f492_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 2
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: 100
sequence_len: 2048
strict: false
tf32: true
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.04514224320834951
wandb_entity: null
wandb_mode: online
wandb_name: e12ebf95-984c-4ed3-a404-242ec52dd74f
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: e12ebf95-984c-4ed3-a404-242ec52dd74f
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

a838f1d2-844b-4887-a1f8-7528ffbe9df7

This model is a fine-tuned version of Eurdem/Defne_llama3_2x8B on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.8418

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: 4
• eval_batch_size: 4
• seed: 42
• gradient_accumulation_steps: 8
• total_train_batch_size: 32
• 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: 10
• training_steps: 1092

Training results

Training Loss	Epoch	Step	Validation Loss
1.5341	0.0003	1	1.5445
1.024	0.0303	100	1.0019
0.8856	0.0605	200	0.9547
0.9799	0.0908	300	0.9246
0.884	0.1210	400	0.9040
0.8863	0.1513	500	0.8852
0.8652	0.1815	600	0.8703
0.8292	0.2118	700	0.8588
0.8298	0.2420	800	0.8500
0.8653	0.2723	900	0.8445
0.7846	0.3026	1000	0.8418

Framework versions

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