--- license: llama2 library_name: transformers tags: - code model-index: - name: Pandalyst_13B_v1.0 results: - task: type: text-generation metrics: - name: exec@1 type: exec@1 value: 0.71 verified: false language: - en --- ## Pandalyst: A large language model for mastering data analysis using pandas

🐱 Github Repo

**What is Pandalyst** - Pandalyst is a general large language model specifically trained to process and analyze data using the pandas library. **How is Pandalyst** - Pandalyst has strong generalization capabilities for data tables in different fields and different data analysis needs. **Why is Pandalyst** - Pandalyst is open source and free to use, and its small parameter size (7B/13B) allows us to easily deploy it on local PC. - Pandalyst can handle complex data tables (multiple columns and multiple rows), allowing us to enter enough context to describe our table in detail. - Pandalyst has very competitive performance, significantly outperforming models of the same size and even outperforming some of the strongest closed-source models. ## News - 🔥[2023/09/30] We released **Pandalyst-7B-V1.1** , which was trained on **CodeLlama-7b-Python** and achieves the **76.1 exec@1** in our **PandaTest_V1.0** and surpasses **Pandalyst-13B-V1.0**, **WizardCoder-Python-13B-V1.0** and **ChatGPT-3.5 (2023/06/13)**. - 🔥[2023/09/28] We released **Pandalyst-13B-V1.0** , which was trained on **WizardCoder-Python-13B-V1.0** and achieves the **70.7 exec@1** in our **PandaTest_V1.0** and surpasses **WizardCoder-Python-13B-V1.0** and **ChatGPT-3.5 (2023/06/13)**. | Model | Checkpoint | Base Model | PandaTest_V1.0 | EASY | HARD | License | |--------------------|---------------------------------------------------------------------------------------------|------------|----------------|---------------------|---------------------| ----- | | Pandalyst-13B-V1.0 | 🤗 HF Link | WizardCoder-Python-13B-V1.0 | 70.7 | 75.6 | 65.9 | Llama2 | | Pandalyst-7B-V1.1 | 🤗 HF Link | CodeLlama-7b-Python | 76.1 | 85.2 | 67.0 | Llama2 | ## Usage and Human evaluation Please refer to Github.