Adriana Meza Soria

Toward Collecting Evidence of Design Work in Software Maintenance Meetings: Preliminary Results

Janet E. Burge, Adriana Meza Soria, André van der Hoek

A software system spends most of its lifetime in the operations and maintenance phase, where developers repair defects and add features, as requested by the customer or otherwise identified. Making such repairs and extensions often requires additional design work. Precisely where, when, …

us (code pays fourni par la source)

TOUCAN: Synthesizing 1.5M Tool-Agentic Data from Real-World MCP Environments

Zhangchen Xu, Adriana Meza Soria, Shawn Zheng Kai Tan, Ashish Sunil Agrawal et autres

Large Language Model (LLM) agents are rapidly emerging as powerful systems for automating tasks across domains. Yet progress in the open-source community is constrained by the lack of high quality permissively licensed tool-agentic training data. Existing datasets are often limited in diversity, …

Power Scheduler: A Batch Size and Token Number Agnostic Learning Rate Scheduler

Yikang Shen, Matthew Stallone, Mayank Mishra, Gaoyuan Zhang et autres

Finding the optimal learning rate for language model pretraining is a challenging task. This is not only because there is a complicated correlation between learning rate, batch size, number of training tokens, model size, and other hyperparameters but also because it is …

Scaling Granite Code Models to 128K Context

Matt Stallone, Vaibhav Saxena, Leonid Karlinsky, Bridget McGinn et autres

This paper introduces long-context Granite code models that support effective context windows of up to 128K tokens. Our solution for scaling context length of Granite 3B/8B code models from 2K/4K to 128K consists of a light-weight continual pretraining by gradually increasing its …

Granite-Function Calling Model: Introducing Function Calling Abilities via Multi-task Learning of Granular Tasks

Ibrahim Abdelaziz, Kinjal Basu, Mayank Agarwal, Sadhana Kumaravel et autres

Large language models (LLMs) have recently shown tremendous promise in serving as the backbone to agentic systems, as demonstrated by their performance in multi-faceted, challenging benchmarks like SWE-Bench and Agent-Bench. However, to realize the true potential of LLMs as autonomous agents, they …

Granite Code Models: A Family of Open Foundation Models for Code Intelligence

Mayank Mishra, Matt Stallone, Gaoyuan Zhang, Yikang Shen et autres

Large Language Models (LLMs) trained on code are revolutionizing the software development process. Increasingly, code LLMs are being integrated into software development environments to improve the productivity of human programmers, and LLM-based agents are beginning to show promise for handling complex tasks …

API Pack: A Massive Multi-Programming Language Dataset for API Call Generation

Zhen Guo, Adriana Meza Soria, Wei Sun, Yikang Shen et autres

We introduce API Pack, a massive multi-programming language dataset containing over one million instruction-API calls for improving the API call generation capabilities of large language models. Our evaluation highlights three key findings: First, fine-tuning on API Pack enables open-source models to outperform …

Characterizing Software Maintenance Meetings: Information Shared, Discussion Outcomes, and Information Captured

Adriana Meza Soria, Taylor Lopez, Elizabeth Seero, Negin Mashhadi et autres

A type of meeting that has been understudied in the software engineering literature to date is what we term the software maintenance meeting: a regularly scheduled team meeting in which emergent issues are addressed that are usually out of scope of the …

us (code pays fourni par la source)

Granite-Function Calling Model: Introducing Function Calling Abilities via Multi-task Learning of Granular Tasks

Ibrahim Abdelaziz, Kinjal Basu, Mayank Agarwal, Sadhana Kumaravel et autres

Ibrahim Abdelaziz, Kinjal Basu, Mayank Agarwal, Sadhana Kumaravel, Matthew Stallone, Rameswar Panda, Yara Rizk, G P Shrivatsa Bhargav, Maxwell Crouse, Chulaka Gunasekara, Shajith Ikbal, Sachindra Joshi, Hima Karanam, Vineet Kumar, Asim Munawar, Sumit Neelam, Dinesh Raghu, Udit Sharma, Adriana Meza Soria, Dheeraj …

Exploring a Research Agenda for Design Knowledge Capture in Meetings

Liz Seero, Janet E. Burge, Adriana Meza Soria, André van der Hoek

Meetings are a frequent part of life for a software developer. Software design is often performed, discussed, and reviewed in these meetings. This means that meetings may contain important design information that could be captured for later use. Meeting design tools may …

us (code pays fourni par la source)

Recurring distributed software maintenance meetings

Adriana Meza Soria, André van der Hoek, Janet E. Burge

Meetings have always been a significant part of working life and software development is no exception, with meetings of all kinds taking place daily. One type of meeting that is critical to software development that has not been widely studied to date …

us (code pays fourni par la source)

Reading to write code

Brooke Jade Ryan, Adriana Meza Soria, Kaj Dreef, André van der Hoek

A substantial portion of any software engineer's job is reading code. Despite the criticality of this skill in a budding software engineer, reading code---and more specifically, techniques on how to read code when integrating oneself into a large existing software project---is often …

us (code pays fourni par la source)