adapter: lora
base_model: EleutherAI/pythia-410m-deduped
bf16: auto
dataset_prepared_path: null
datasets:
- data_files:
  - adcc810a6ceb3267_train_data.json
  ds_type: json
  format: custom
  path: adcc810a6ceb3267_train_data.json
  type:
    field: null
    field_input: domain
    field_instruction: sentence
    field_output: triples
    field_system: null
    format: null
    no_input_format: null
    system_format: '{system}'
    system_prompt: ''
early_stopping_patience: null
evals_per_epoch: 2
gradient_accumulation_steps: 1
group_by_length: false
hub_model_id: taopanda-1/2bc524d3-39c3-413f-86ea-7a7851e7528b
learning_rate: 1.0e-05
load_in_8bit: true
local_rank: null
logging_steps: 1
lora_alpha: 32
lora_dropout: 0.05
lora_fan_in_fan_out: true
lora_model_dir: null
lora_r: 16
lora_target_linear: null
lora_target_modules:
- query_key_value
micro_batch_size: 4
num_epochs: 1
output_dir: ./outputs/lora-alpaca-pythia/taopanda-1_6b76e44c-62d5-4a35-a480-c808b829ee0b
resume_from_checkpoint: null
seed: 96247
sequence_len: 512
special_tokens:
  pad_token: <|endoftext|>
tf32: true
train_on_inputs: false
trust_remote_code: true
val_set_size: 0.05
wandb_entity: fatcat87-taopanda
wandb_log_model: null
wandb_mode: online
wandb_name: taopanda-1_6b76e44c-62d5-4a35-a480-c808b829ee0b
wandb_project: subnet56
wandb_runid: taopanda-1_6b76e44c-62d5-4a35-a480-c808b829ee0b
wandb_watch: null
weight_decay: 0.1

Visualize in Weights & Biases

2bc524d3-39c3-413f-86ea-7a7851e7528b

This model is a fine-tuned version of EleutherAI/pythia-410m-deduped on the None dataset. It achieves the following results on the evaluation set:

Loss: 1.4366

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: 1e-05
train_batch_size: 4
eval_batch_size: 4
seed: 96247
distributed_type: multi-GPU
num_devices: 4
total_train_batch_size: 16
total_eval_batch_size: 16
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: cosine
lr_scheduler_warmup_steps: 23
num_epochs: 1

Training results

Training Loss	Epoch	Step	Validation Loss
3.7662	0.0013	1	3.5014
1.521	0.5006	399	1.4366

Framework versions

PEFT 0.11.1
Transformers 4.42.3
Pytorch 2.3.0+cu121
Datasets 2.19.1
Tokenizers 0.19.1