See axolotl config

axolotl version: 0.4.1

adapter: lora
auto_find_batch_size: true
base_model: unsloth/gemma-2-9b
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 555398d2f4ed4556_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/555398d2f4ed4556_train_data.json
  type:
    field_instruction: input
    field_output: chosen
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
do_eval: true
early_stopping_patience: 3
eval_max_new_tokens: 128
eval_steps: 50
evals_per_epoch: null
flash_attention: true
fp16: false
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 2
gradient_checkpointing: false
group_by_length: true
hub_model_id: lesso08/838b1333-951e-4d23-9a70-09c908d157f2
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.000208
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 10
lora_alpha: 32
lora_dropout: 0.05
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 16
lora_target_linear: true
lr_scheduler: cosine
max_grad_norm: 1.0
max_steps: 500
micro_batch_size: 4
mlflow_experiment_name: /tmp/555398d2f4ed4556_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
saves_per_epoch: null
seed: 80
sequence_len: 512
strict: false
tf32: true
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: null
wandb_mode: online
wandb_name: 3f81c694-97f7-45a8-adf0-cfa9c8ef6d9d
wandb_project: 08a
wandb_run: your_name
wandb_runid: 3f81c694-97f7-45a8-adf0-cfa9c8ef6d9d
warmup_steps: 50
weight_decay: 0.0
xformers_attention: null

838b1333-951e-4d23-9a70-09c908d157f2

This model is a fine-tuned version of unsloth/gemma-2-9b on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.8807

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: 0.000208
• train_batch_size: 4
• eval_batch_size: 4
• seed: 80
• gradient_accumulation_steps: 2
• total_train_batch_size: 8
• 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: 50
• training_steps: 500

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0003	1	1.1669
0.809	0.0168	50	0.9790
1.0684	0.0335	100	1.0705
0.8132	0.0503	150	0.9366
1.0571	0.0670	200	1.0072
0.9024	0.0838	250	0.9066
0.8454	0.1006	300	0.9311
0.6724	0.1173	350	0.8849
1.0545	0.1341	400	0.8880
0.8303	0.1509	450	0.8712
1.1619	0.1676	500	0.8807

Framework versions

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