TrainingCreate.md

TrainingCreate

Properties

Name	Type	Description	Notes
id	str		[readonly]
project_name	str
dataset	str
base_model_name	str
provider	Provider63aEnum		[optional]
text_column	str		[optional]
learning_rate	float		[optional]
batch_size	int		[optional]
num_epochs	int		[optional]
block_size	int		[optional]
warmup_ratio	float		[optional]
lora_r	int		[optional]
lora_alpha	int		[optional]
lora_dropout	float		[optional]
weight_decay	float		[optional]
gradient_accumulation	int		[optional]
use_peft	bool		[optional]
use_fp16	bool		[optional]
use_int4	bool		[optional]
push_to_hub	bool		[optional]
repo_id	str		[optional]
preprocess_dataset	bool		[optional]
prompt_column	str		[optional]
prompt_prefix	str		[optional]
prompt_suffix	str		[optional]
response_prefix	str		[optional]
date_created	datetime		[readonly]
last_modified	datetime		[readonly]

Example

from iblai.models.training_create import TrainingCreate

# TODO update the JSON string below
json = "{}"
# create an instance of TrainingCreate from a JSON string
training_create_instance = TrainingCreate.from_json(json)
# print the JSON string representation of the object
print(TrainingCreate.to_json())

# convert the object into a dict
training_create_dict = training_create_instance.to_dict()
# create an instance of TrainingCreate from a dict
training_create_from_dict = TrainingCreate.from_dict(training_create_dict)

[Back to Model list] [Back to API list] [Back to README]