See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: HuggingFaceM4/tiny-random-LlamaForCausalLM
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - 7ea11a9c00b4063e_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/7ea11a9c00b4063e_train_data.json
  type:
    field_input: text
    field_instruction: question_statement
    field_output: answer
    format: '{instruction} {input}'
    no_input_format: '{instruction}'
    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: true
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
group_by_length: false
hub_model_id: ardaspear/75d08b28-e4dc-4393-861b-a803cd5dfd09
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0001
load_in_4bit: false
load_in_8bit: false
local_rank: 0
logging_steps: 3
lora_alpha: 128
lora_dropout: 0.1
lora_fan_in_fan_out: true
lora_model_dir: null
lora_r: 64
lora_target_linear: true
lr_scheduler: cosine
max_memory:
  0: 72GB
max_steps: 100
micro_batch_size: 8
mlflow_experiment_name: /tmp/7ea11a9c00b4063e_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 3
optimizer: adamw_bnb_8bit
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: false
sample_packing: false
saves_per_epoch: 4
sequence_len: 1024
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: leixa-personal
wandb_mode: online
wandb_name: 75d08b28-e4dc-4393-861b-a803cd5dfd09
wandb_project: Gradients-On-Three
wandb_run: your_name
wandb_runid: 75d08b28-e4dc-4393-861b-a803cd5dfd09
warmup_steps: 10
weight_decay: 0.01
xformers_attention: null

75d08b28-e4dc-4393-861b-a803cd5dfd09

This model is a fine-tuned version of HuggingFaceM4/tiny-random-LlamaForCausalLM on the None dataset. It achieves the following results on the evaluation set:

• Loss: 10.2647

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: 0.0001
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• gradient_accumulation_steps: 4
• 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: 100

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0002	1	10.3772
10.3794	0.0020	9	10.3714
10.3554	0.0039	18	10.3539
10.3313	0.0059	27	10.3252
10.3037	0.0079	36	10.2947
10.2756	0.0098	45	10.2780
10.2762	0.0118	54	10.2705
10.2697	0.0138	63	10.2673
10.2635	0.0157	72	10.2658
10.2756	0.0177	81	10.2651
10.2646	0.0197	90	10.2648
10.2826	0.0216	99	10.2647

Framework versions

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