See axolotl config

axolotl version: 0.4.1

adapter: lora
auto_find_batch_size: true
base_model: Maykeye/TinyLLama-v0
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 0b711362b5e882a7_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/0b711362b5e882a7_train_data.json
  type:
    field_input: documents
    field_instruction: question
    field_output: answer
    format: '{instruction} {input}'
    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: lesso06/bf6a45f0-c90d-4d36-adeb-9e26868e9576
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.000206
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/0b711362b5e882a7_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: 60
sequence_len: 512
special_tokens:
  pad_token: </s>
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: 0f65094d-5d32-4d26-af34-69fddfd32cf4
wandb_project: 06a
wandb_run: your_name
wandb_runid: 0f65094d-5d32-4d26-af34-69fddfd32cf4
warmup_steps: 50
weight_decay: 0.0
xformers_attention: null

bf6a45f0-c90d-4d36-adeb-9e26868e9576

This model is a fine-tuned version of Maykeye/TinyLLama-v0 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 7.3675

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: 0.000206
• train_batch_size: 4
• eval_batch_size: 4
• seed: 60
• 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	0.0002	1	8.5308
7.756	0.0105	50	7.9158
7.4785	0.0209	100	7.6264
7.3123	0.0314	150	7.5076
7.3021	0.0419	200	7.4542
7.2916	0.0523	250	7.4184
7.1929	0.0628	300	7.4012
7.0804	0.0732	350	7.3856
7.1351	0.0837	400	7.3719
7.1269	0.0942	450	7.3695
7.2618	0.1046	500	7.3675

Framework versions

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