See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: unsloth/Llama-3.1-Storm-8B
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - e8ca0ac66aa11e96_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/e8ca0ac66aa11e96_train_data.json
  type:
    field_instruction: Hausa
    field_output: English
    format: '{instruction}'
    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: true
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: false
group_by_length: false
hub_model_id: dimasik87/81ee6aa8-1abc-4905-8446-40b88b66ce39
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: 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_memory:
  0: 70GiB
max_steps: 50
micro_batch_size: 1
mlflow_experiment_name: /tmp/e8ca0ac66aa11e96_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 4
optimizer: adamw_torch
output_dir: miner_id_24
pad_to_sequence_len: true
resume_from_checkpoint: null
s2_attention: null
sample_packing: false
saves_per_epoch: 4
sequence_len: 2028
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: 81ee6aa8-1abc-4905-8446-40b88b66ce39
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 81ee6aa8-1abc-4905-8446-40b88b66ce39
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

81ee6aa8-1abc-4905-8446-40b88b66ce39

This model is a fine-tuned version of unsloth/Llama-3.1-Storm-8B on the None dataset. It achieves the following results on the evaluation set:

• Loss: 3.2479

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: 1
• eval_batch_size: 1
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 4
• optimizer: Use OptimizerNames.ADAMW_TORCH 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: 50

Training results

Training Loss	Epoch	Step	Validation Loss
10.969	0.0003	1	11.5726
9.5194	0.0012	4	11.4298
9.8745	0.0024	8	9.6370
5.6577	0.0036	12	5.3106
2.7349	0.0048	16	4.8557
3.5522	0.0060	20	4.0696
3.0378	0.0072	24	3.7071
3.2366	0.0084	28	3.5185
3.5875	0.0096	32	3.4018
4.0769	0.0108	36	3.3534
2.8831	0.0120	40	3.2889
3.5078	0.0132	44	3.2595
2.8259	0.0144	48	3.2479

Framework versions

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