See axolotl config

axolotl version: 0.4.1

adapter: lora
base_model: unsloth/Llama-3.2-1B-Instruct
bf16: true
chat_template: llama3
dataset_prepared_path: null
datasets:
- data_files:
  - d640525a2dc2a05c_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/d640525a2dc2a05c_train_data.json
  type:
    field_input: ''
    field_instruction: post
    field_output: summary
    format: '{instruction}'
    no_input_format: '{instruction}'
    system_format: '{system}'
    system_prompt: ''
debug: null
deepspeed: null
device_map: auto
do_eval: true
early_stopping_patience: 2
eval_batch_size: 4
eval_max_new_tokens: 128
eval_steps: 150
eval_table_size: null
evals_per_epoch: null
flash_attention: true
fp16: false
fsdp: null
fsdp_config: null
gradient_accumulation_steps: 4
gradient_checkpointing: true
group_by_length: true
hub_model_id: 0x1202/ecae1d40-b1bb-43db-95bd-1c6d3f4a7ec3
hub_repo: null
hub_strategy: checkpoint
hub_token: null
learning_rate: 0.0001
load_in_4bit: false
load_in_8bit: false
local_rank: null
logging_steps: 10
lora_alpha: 128
lora_dropout: 0.3
lora_fan_in_fan_out: null
lora_model_dir: null
lora_r: 64
lora_target_linear: true
lr_scheduler: cosine
max_grad_norm: 1.0
max_memory:
  0: 75GB
max_steps: 4500
micro_batch_size: 8
mlflow_experiment_name: /tmp/d640525a2dc2a05c_train_data.json
model_type: AutoModelForCausalLM
num_epochs: 1
optim_args:
  adam_beta1: 0.9
  adam_beta2: 0.95
  adam_epsilon: 1e-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: 150
saves_per_epoch: null
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: 2a12b116-57ac-4702-b880-01a716a1e301
wandb_project: Gradients-On-Demand
wandb_run: your_name
wandb_runid: 2a12b116-57ac-4702-b880-01a716a1e301
warmup_steps: 50
weight_decay: 0.0
xformers_attention: null

ecae1d40-b1bb-43db-95bd-1c6d3f4a7ec3

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: 2.1301

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: 4
• 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=adam_beta1=0.9,adam_beta2=0.95,adam_epsilon=1e-5
• lr_scheduler_type: cosine
• lr_scheduler_warmup_steps: 50
• training_steps: 3846

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.0003	1	3.1843
2.2768	0.0390	150	2.2907
2.3128	0.0780	300	2.2574
2.1967	0.1170	450	2.2451
2.2483	0.1560	600	2.2303
2.2865	0.1950	750	2.2190
2.273	0.2341	900	2.2104
2.247	0.2731	1050	2.2039
2.2681	0.3121	1200	2.1944
2.1739	0.3511	1350	2.1862
2.1636	0.3901	1500	2.1807
2.2038	0.4291	1650	2.1739
2.1904	0.4681	1800	2.1677
2.1977	0.5071	1950	2.1621
2.1476	0.5461	2100	2.1584
2.1481	0.5851	2250	2.1553
2.1617	0.6241	2400	2.1505
2.202	0.6632	2550	2.1457
2.2041	0.7022	2700	2.1407
2.1973	0.7412	2850	2.1380
2.1571	0.7802	3000	2.1349
2.2192	0.8192	3150	2.1341
2.1933	0.8582	3300	2.1317
2.1233	0.8972	3450	2.1314
2.1085	0.9362	3600	2.1311
2.1266	0.9752	3750	2.1301

Framework versions

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