Functions — github.com/discus-labs/discus

hub / github.com/discus-labs/discus / functions

Functions43 in github.com/discus-labs/discus

⨍ Functions 43 ◇ Types & classes 11

↓ 1 callersMethod_get_data_from_model_response

(self, response)

src/discus/models/huggingface.py:63

↓ 1 callersMethod_hf_generate

(self, prompt, context = None)

src/discus/models/huggingface.py:14

↓ 1 callersMethod_load_data

Load data from the CSV file.

src/discus/generator.py:65

↓ 1 callersMethod_openai_generate

(self, prompt, context = None)

src/discus/models/openai.py:16

↓ 1 callersMethod_parse_config

Parse the JSON config file.

src/discus/json/config.py:36

↓ 1 callersMethod_set_attributes

Set instance variables using keys from config data. Raise an error if a key is missing.

src/discus/json/config.py:12

↓ 1 callersMethod_validate_config

Validates the config against supported models and task types.

src/discus/generator.py:20

↓ 1 callersMethodextract_data

TODO: add docstring

src/discus/knowledge.py:34

↓ 1 callersMethodgenerate_prompt

(self, seed_dataset: pd.DataFrame = None, context = None)

src/discus/operations/instruction.py:9

↓ 1 callersMethodgenerate_prompt

(self, seed_dataset: pd.DataFrame = None, context = None)

src/discus/operations/instance.py:9

↓ 1 callersFunctionretrieve_supported_embedding_models

Generate given model providers. This function retrieves the list of embedding model providers from SupportedEmbeddingModels enum.

src/discus/json/architecture.py:11

↓ 1 callersFunctionretrieve_supported_models

Generate given model providers. This function retrieves the list of model providers from SupportedModels enum.

src/discus/json/architecture.py:4

↓ 1 callersFunctionretrieve_task_types

Generate given task types. This function retrieves the list of acceptable task_types from TaskType enum.

src/discus/json/architecture.py:18

↓ 1 callersMethodtransform_dataframe

Transforms a DataFrame into a list of dictionaries. Parameters: df (pandas.DataFrame): The DataFrame to be transformed.

src/discus/operations/instruction.py:59

↓ 1 callersMethodtransform_dataframe

Transforms a DataFrame into a list of dictionaries. Parameters: df (pandas.DataFrame): The DataFrame to be transformed.

src/discus/operations/instance.py:59

Method__init__

Initializes Knowledge class with config

src/discus/knowledge.py:18

Method__init__

Initialize the generator and parse the config.

src/discus/generator.py:15

Method__init__

Initialize the Dataset with a given CSV file path.

src/discus/datasets/dataset.py:8

Method__init__

(self, config: DiscusConfig)

src/discus/models/openai.py:9

Method__init__

(self, config: DiscusConfig)

src/discus/models/huggingface.py:8

Method__init__

(self, config_path: str)

src/discus/json/config.py:8

Method__init__

(self, config: DiscusConfig)

src/discus/operations/instruction.py:6

Method__init__

(self, config: DiscusConfig)

src/discus/operations/instance.py:6

Method__len__

Return the number of examples in the dataset.

src/discus/datasets/dataset.py:168

Method__repr__

Return a string representation of the dataset.

src/discus/datasets/dataset.py:172

Method_get_data_from_model_response

(self, response)

src/discus/models/openai.py:57

Method_load_data

Load data from the CSV file.

src/discus/datasets/dataset.py:12

Method_validate_config

Validates the config against supported models and task types.

src/discus/knowledge.py:28

Methodchunk_data

chunk_proporition is proportion of context window overlap is proportion of chunk which overlaps with the next path - a path t

src/discus/knowledge.py:70

Methodcleaning

Include shnu's cleaning function for datasets.

src/discus/datasets/dataset.py:160

Functioncomplete_instances

Generate completions for a list of input instances using contextual information and OpenAI's chat model. This function takes a list of input

src/discus/knowledge_old.py:123

Functioncreate_vector_store

Create a vector store from text documents using OpenAI embeddings. This function reads text documents from a specified path, processes them

src/discus/knowledge_old.py:13

Methodelements_below_similarity_threshold

Filters elements based on a similarity threshold using TF-IDF cosine similarity. This function takes either a list of text inputs or

src/discus/datasets/dataset.py:97

Methodgenerate_embeddings

(self, collection_name, persist_path)

src/discus/knowledge.py:117

Functiongenerate_prompts

Generate a comprehensive list of prompts based on text documents and OpenAI's chat model. This function reads text documents from a specifie

src/discus/knowledge_old.py:195

Methodget_data

Return the loaded dataset.

src/discus/datasets/dataset.py:164

Methodload_db

(self, collection_name, persist_path)

src/discus/knowledge.py:146

Functionload_vector_store

Load a Chroma vector store from a specified path. This function loads a previously created Chroma vector store from the specified path and r

src/discus/knowledge_old.py:50

Methodn_most_unique_elements

Extracts the n most unique elements from a list of text inputs or a DataFrame. This function takes either a list of text inputs or a

src/discus/datasets/dataset.py:20

Methodnaive_retriever

(self, k=3)

src/discus/knowledge.py:169

Functionquery_vector_store

Query a Chroma vector store for relevant documents and generate a response using OpenAI's chat model. This function queries a Chroma vector

src/discus/knowledge_old.py:71

Methodretriever

(self, query)

src/discus/knowledge.py:176

Methodrun

Load seed dataset and generate the synthetic dataset based on the parsed config.

src/discus/generator.py:29