entelecheia / hyfi-absa

A Python package developed as a plugin for the Hydra Fast Interface (HyFI).

HyFI-ABSA

HyFI-ABSA is a Python package developed as a plugin for the Hydra Fast Interface (HyFI).

• Documentation: https://hyfi-absa.entelecheia.ai
• GitHub: https://github.com/entelecheia/hyfi-absa
• PyPI: https://pypi.org/project/hyfi-absa

The main function of HyFI-ABSA is to facilitate Aspect-Based Sentiment Analysis (ABSA), a specialized form of sentiment analysis that allows you to evaluate specific components or aspects of a given text. It also performs general sentiment analysis tasks. One feature of the HyFI-ABSA package is its capacity to employ generative language models for creating training datasets. This feature not only eliminates the extensive manual labor typically associated with data collection and preparation, but also ensures a more diverse and balanced dataset, which subsequently improves the overall performance and accuracy of sentiment analysis. In addition to these primary features, HyFI-ABSA boasts seamless integration with the existing HyFI architecture, thus allowing you to effortlessly combine it with other HyFI plugins and tools.

Changelog

See the CHANGELOG for more information.

Contributing

Contributions are welcome! Please see the contributing guidelines for more information.

License

This project is released under the MIT License.

hyfi-absa's Issues

Add a function to parse the results of the ABSA task

https://github.com/entelecheia/hyfi-absa/blob/4b1c971989bdfb4702282d86a5ec52da70500cac/src/hyabsa/agents/results.py#L13-L42

Add an option to set backoff limit

hyfi-absa/src/hyabsa/llms/openai.py

Lines 86 to 98 in 62b005c

	@backoff.on_exception(
	backoff.expo,
	(
	RateLimitError,
	APIConnectionError,
	APIError,
	ServiceUnavailableError,
	),
	)
	def request_api(engine, args, delay_in_seconds: float = 1):
	time.sleep(delay_in_seconds)
	return engine.create(**args)

Save the ids of the prompts of the requests in the log file

When the job is not successful, the ids should be able to be retrieved from the log file.

Implement API REQUEST PARALLEL PROCESSOR

Refers to: https://github.com/openai/openai-cookbook/blob/5783656852d507c335955d14875ebc9902f628ef/examples/api_request_parallel_processor.py