metadata
dataset_info:
features:
- name: Model_name
dtype: string
- name: Train_size
dtype: int64
- name: Test_size
dtype: int64
- name: arg
struct:
- name: auto_find_batch_size
dtype: bool
- name: gradient_accumulation_steps
dtype: int64
- name: learning_rate
dtype: float64
- name: logging_steps
dtype: int64
- name: lr_scheduler_type
dtype: string
- name: num_train_epochs
dtype: int64
- name: output_dir
dtype: string
- name: report_to
dtype: string
- name: save_strategy
dtype: string
- name: save_total_limit
dtype: int64
- name: seed
dtype: int64
- name: warmup_steps
dtype: int64
- name: weight_decay
dtype: float64
- name: lora
dtype: 'null'
- name: Parameters
dtype: int64
- name: Trainable_parameters
dtype: int64
- name: r
dtype: 'null'
- name: Memory Allocation
dtype: string
- name: Training Time
dtype: string
- name: Performance
struct:
- name: accuracy
dtype: float64
- name: f1_macro
dtype: float64
- name: f1_weighted
dtype: float64
- name: precision
dtype: float64
- name: recall
dtype: float64
splits:
- name: train
num_bytes: 1332
num_examples: 6
download_size: 12029
dataset_size: 1332
configs:
- config_name: default
data_files:
- split: train
path: data/train-*