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:
  - 4c8e3ef2c206b838_train_data.json
  ds_type: json
  format: custom
  path: /workspace/input_data/4c8e3ef2c206b838_train_data.json
  type:
    field_input: eval_solution
    field_instruction: problem
    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: sn56m3/2bb4fce6-966d-405c-8e32-3a9d757eaf7e
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/4c8e3ef2c206b838_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: sn56-miner
wandb_mode: disabled
wandb_name: 2bb4fce6-966d-405c-8e32-3a9d757eaf7e
wandb_project: god
wandb_run: jr17
wandb_runid: 2bb4fce6-966d-405c-8e32-3a9d757eaf7e
warmup_steps: 10
weight_decay: 0.01
xformers_attention: null

2bb4fce6-966d-405c-8e32-3a9d757eaf7e

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: 10.2736

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
• distributed_type: multi-GPU
• num_devices: 4
• gradient_accumulation_steps: 4
• total_train_batch_size: 128
• total_eval_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	0.0284	1	10.3831
10.3776	0.2553	9	10.3789
10.3679	0.5106	18	10.3654
10.3483	0.7660	27	10.3421
13.012	1.0213	36	10.3126
10.3863	1.2766	45	10.2931
10.4675	1.5319	54	10.2836
10.285	1.7872	63	10.2787
12.8024	2.0426	72	10.2758
10.0497	2.2979	81	10.2743
10.1469	2.5532	90	10.2738
10.3743	2.8085	99	10.2736

Framework versions

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