llm-kb
← Back to research
Benchmark

Developer proposes STT torture test benchmark on GitHub

A developer drafted a speech-to-text benchmark that goes beyond clean audio and single WER scores, proposing seven challenging scenarios including phone calls from moving cars, speaker interruptions, code-switching, and timestamp drift. The test set is intended as a public GitHub repo to stress-test STT systems with ugly real-world clips.

2 engagement·1 source·Sat, Jul 11, 2026, 07:44 AM
The developer outlined seven scenarios for a torture test: 1) phone call from a moving car with speakerphone and road noise, 2) customer correcting a number mid-speech, 3) silence with background noise (fan, TV, keyboard, hold music), 4) two speakers interrupting each other with partials and diarization stress, 5) addresses and local names with hard entities and weird pauses, 6) long meeting with timestamp and speaker drift, and 7) code-switching. The test set is meant to be a public GitHub repo with ugly clips and clear metrics, avoiding clean audio and single WER scores.

Entities

STT torture test(benchmark)GitHub(tool)

Related

PaperTue, Jul 7, 2026, 04:22 PM

Paper identifies word-level control bottleneck in LLM-based TTS

A new paper highlights the inability of current LLM-based TTS systems to explicitly manipulate word-level acoustic attributes, limiting precise stylistic control and temporal alignment for applications like audiobook narration and video dubbing. The authors attribute this to scarcity of fine-grained annotated datasets and architectural challenges.

0 engagement·1 source·arxiv
arXiv
BenchmarkSat, Jul 11, 2026, 06:55 AM

Meta releases SWE-Together benchmark measuring coding agent steering difficulty

Meta introduced SWE-Together, a new benchmark that evaluates coding agents on interactive, multi-turn tasks rather than single-shot problem solving. The benchmark measures how much human steering an agent requires, which correlates strongly with its capability. This addresses a key limitation of SWE-bench, which tests agents on frozen tickets alone.

0 engagement·1 source·rss
RSS
BenchmarkWed, Jul 8, 2026, 01:00 PM

OpenAI analysis reveals flaws in SWE-Bench Pro coding benchmark

OpenAI published an analysis uncovering reliability issues in SWE-Bench Pro, a popular benchmark for evaluating AI coding models. The findings raise concerns about the accuracy of benchmark scores, potentially affecting how developers and researchers trust model evaluations.

0 engagement·1 source·rss
RSS
BenchmarkThu, Jul 9, 2026, 05:59 PM

UniClawBench benchmark proposed for evaluating proactive AI agents in real-world tasks

Researchers introduced UniClawBench, a universal benchmark for evaluating proactive agents that operate everyday tools in real-world environments. Unlike existing benchmarks that rely on sandboxed settings and single-turn evaluations, UniClawBench aims to isolate specific model capabilities to identify root causes of agent failures.

0 engagement·1 source·arxiv
arXiv
CommunitySun, Jul 12, 2026, 12:12 PM

Developer seeks feedback on fine-tuning LoRA for conversation state extraction in long LLM chats

A developer is working on a side project to improve AI conversation continuity by training a small model to extract structured conversation state from chat chunks, rather than relying on summarization. They are seeking feedback on their approach involving fine-tuning a LoRA, dataset design, and long-context systems.

19 engagement·2 sources·reddit