See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: unsloth/Llama-3.2-1B-Instruct
bf16: auto
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - MATH-Hard_train_data.json
  ds_type: json
  path: /workspace/input_data/MATH-Hard_train_data.json
  type:
    field_input: problem
    field_instruction: type
    field_output: solution
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
early_stopping_patience: 3
eval_max_new_tokens: 128
eval_steps: 10
eval_table_size: null
flash_attention: true
fp16: null
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
group_by_length: false
hub_model_id: besimray/miner_id_3_ad9b0fa2-323a-4d04-be5e-1304b49c48da_1729735004
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0002
load_in_4bit: false
load_in_8bit: true
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: 32
lora_target_linear: true
lr_scheduler: cosine
max_steps: 1000
micro_batch_size: 5
mlflow_experiment_name: /tmp/MATH-Hard_train_data.json
model_type: LlamaForCausalLM
num_epochs: 5
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: 10
save_strategy: steps
sequence_len: 4096
strict: false
tf32: false
tokenizer_type: AutoTokenizer
train_on_inputs: false
val_set_size: 0.05
wandb_entity: besimray24-rayon
wandb_mode: online
wandb_project: Public_TuningSN
wandb_run: miner_id_24
wandb_runid: ad9b0fa2-323a-4d04-be5e-1304b49c48da
warmup_steps: 10
weight_decay: 0.0
xformers_attention: null

miner_id_3_ad9b0fa2-323a-4d04-be5e-1304b49c48da_1729735004

This model is a fine-tuned version of unsloth/Llama-3.2-1B-Instruct on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.7745

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: 5
• eval_batch_size: 5
• seed: 42
• gradient_accumulation_steps: 4
• total_train_batch_size: 20
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 10
• training_steps: 838

Training results

Training Loss	Epoch	Step	Validation Loss
0.9487	0.0060	1	0.9856
0.8734	0.0596	10	0.8821
0.8004	0.1192	20	0.8355
0.8897	0.1788	30	0.8209
0.8886	0.2385	40	0.8113
0.8671	0.2981	50	0.8041
0.6372	0.3577	60	0.7990
0.7248	0.4173	70	0.7949
0.7752	0.4769	80	0.7907
0.7171	0.5365	90	0.7890
0.7231	0.5961	100	0.7857
0.8939	0.6557	110	0.7840
0.7672	0.7154	120	0.7806
0.7517	0.7750	130	0.7780
0.7914	0.8346	140	0.7770
0.8214	0.8942	150	0.7758
0.7121	0.9538	160	0.7735
0.7407	1.0134	170	0.7732
0.7109	1.0730	180	0.7753
0.6754	1.1326	190	0.7735
0.6476	1.1923	200	0.7745

Framework versions

• PEFT 0.13.2
• Transformers 4.45.2
• Pytorch 2.4.1+cu124
• Datasets 3.0.1
• Tokenizers 0.20.1