Besimple AI fine-tuned the same Inkling speech model on 1, 25, and 100 hours of proprietary data containing alphanumeric entities.
Public source
Publisher name
Public post
We published how Thinking Machines Lab 's Inkling did on Voice Code Bench few weeks ago, and then we asked, what if we fine-tune Inkling? Inkling's main value prop is it…
Company
Besimple AI
We make AI data simple. https://besimple.com/
- Industry
- Technology, Information and Internet
- Company size
- 11–50 employees
About Besimple AI
Good data is the key to fine-tuning and evaluation of language models, and most models hallucinate without human-in-the-loop. That's why we built Besimple AI for AI teams of all sizes and budgets. We build data flywheel for AI companies so you can continuously improve and monitor your models with human-in-the-loop. We got you covered across text, image, audio and more. Get a quote and start today at: https://besimple.ai/
See moreLatest activity
Latest activity from Besimple AI
4 signals
Research & Knowledge
Besimple AI tested a new ASR model developed by AI at Meta using VoiceCodeBench.
Research & Knowledge
Besimple AI's 100-hour model recovered values the base model missed, including revert-last, tests/auth/login.spec.ts, ALLOWLIST_CIDR, and SN-7KX-9042.
Research & Knowledge
Besimple AI published how Thinking Machines Lab's Inkling speech model was fine-tuned on 1, 25, and 100 hours of proprietary data containing alphanumeric entities.
Discover more
Similar signals
Similar public activity from other companies.
Research & Knowledge
Fluently: AI English tutor
Fluently: AI English tutor trained its pronunciation-sensitive model on real learner calls with a 20.97% error rate.
Research & Knowledge
Silencio Voice AI
Silencio Voice AI validates every clip in delivery by a native-speaker human reviewer.
Research & Knowledge
Speechmatics
Speechmatics benchmarked Melia 1 for switch-point accuracy across Arabic, Mandarin, and Tamil paired with English against Soniox, Amazon, and AssemblyAI.
Research & Knowledge
AudioShake
AudioShake transcribing separated tracks produces 4.1× fewer transcription errors against the strongest open-source speech separator.
Research & Knowledge
Coval