Recap on ML

Per-paper summaries of 50+ agent evaluation research papers from 2023 to 2025, assessing each work's evaluation approach, focus area, relevance to the Agents-eval project, and a concrete integration example.

A reference list of AI-assisted coding tools covering bug search, IDE and terminal integrations, infrastructure provisioning, full-stack builders, and UI development assistants as of early 2025.

A categorized reference list of AI agent frameworks, research agents, benchmarking tools, execution tracers, and AI-enhanced workflow platforms available as of early 2025.

Benchmark results for fine-tuning and half-precision SegFormer (mit-b0) on scene_parse_150, comparing original, fine-tuned, and float16 model variants across IoU and accuracy metrics on GPU T4 and P100.

A curated reference list of ML tooling spanning end-to-end AutoML frameworks, EDA libraries, feature engineering tools, hyperparameter tuners, experiment trackers, and exploratory runtimes.

Documents runtime errors encountered while loading and training quantized SegFormer models, including device_map incompatibilities, half-precision input mismatches, and bitsandbytes 8-bit training failures.