model

superduper.ext.llamacpp.model

Source code

download_uri

download_uri(uri,
     save_path)

Parameter	Description
uri	URI to download
save_path	place to save

Download file.

LlamaCpp

LlamaCpp(self,
     identifier: str,
     db: dataclasses.InitVar[typing.Optional[ForwardRef('Datalayer')]] = None,
     uuid: str = None,
     *,
     artifacts: 'dc.InitVar[t.Optional[t.Dict]]' = None,
     signature: str = 'singleton',
     datatype: 'EncoderArg' = None,
     output_schema: 't.Optional[Schema]' = None,
     flatten: 'bool' = False,
     model_update_kwargs: 't.Dict' = None,
     predict_kwargs: 't.Dict' = None,
     compute_kwargs: 't.Dict' = None,
     validation: 't.Optional[Validation]' = None,
     metric_values: 't.Dict' = None,
     prompt: str = '{input}',
     prompt_func: Optional[Callable] = None,
     max_batch_size: Optional[int] = 4,
     model_name_or_path: str = 'facebook/opt-125m',
     model_kwargs: Dict = None,
     download_dir: str = '.llama_cpp') -> None

Parameter	Description
identifier	Identifier of the leaf.
db	Datalayer instance.
uuid	UUID of the leaf.
artifacts	A dictionary of artifacts paths and DataType objects
signature	Model signature.
datatype	DataType instance.
output_schema	Output schema (mapping of encoders).
flatten	Flatten the model outputs.
model_update_kwargs	The kwargs to use for model update.
predict_kwargs	Additional arguments to use at prediction time.
compute_kwargs	Kwargs used for compute backend job submit. Example (Ray backend): compute_kwargs = dict(resources=...).
validation	The validation Dataset instances to use.
metric_values	The metrics to evaluate on.
prompt	The template to use for the prompt.
prompt_func	The function to use for the prompt.
max_batch_size	The maximum batch size to use for batch generation.
model_name_or_path	path or name of model
model_kwargs	dictionary of init-kwargs
download_dir	local caching directory

Llama.cpp connector.

LlamaCppEmbedding

LlamaCppEmbedding(self,
     identifier: str,
     db: dataclasses.InitVar[typing.Optional[ForwardRef('Datalayer')]] = None,
     uuid: str = None,
     *,
     artifacts: 'dc.InitVar[t.Optional[t.Dict]]' = None,
     signature: str = 'singleton',
     datatype: 'EncoderArg' = None,
     output_schema: 't.Optional[Schema]' = None,
     flatten: 'bool' = False,
     model_update_kwargs: 't.Dict' = None,
     predict_kwargs: 't.Dict' = None,
     compute_kwargs: 't.Dict' = None,
     validation: 't.Optional[Validation]' = None,
     metric_values: 't.Dict' = None,
     prompt: str = '{input}',
     prompt_func: Optional[Callable] = None,
     max_batch_size: Optional[int] = 4,
     model_name_or_path: str = 'facebook/opt-125m',
     model_kwargs: Dict = None,
     download_dir: str = '.llama_cpp') -> None

Parameter	Description
identifier	Identifier of the leaf.
db	Datalayer instance.
uuid	UUID of the leaf.
artifacts	A dictionary of artifacts paths and DataType objects
signature	Model signature.
datatype	DataType instance.
output_schema	Output schema (mapping of encoders).
flatten	Flatten the model outputs.
model_update_kwargs	The kwargs to use for model update.
predict_kwargs	Additional arguments to use at prediction time.
compute_kwargs	Kwargs used for compute backend job submit. Example (Ray backend): compute_kwargs = dict(resources=...).
validation	The validation Dataset instances to use.
metric_values	The metrics to evaluate on.
prompt	The template to use for the prompt.
prompt_func	The function to use for the prompt.
max_batch_size	The maximum batch size to use for batch generation.
model_name_or_path	path or name of model
model_kwargs	dictionary of init-kwargs
download_dir	local caching directory

Llama.cpp connector for embeddings.