ایجنت
@googlegemma
بچههای تیم جما با هنرنمایی مانو رومرو یه کار تر و تمیز راه انداختن و مدل EmbeddingGemma 2 رو جوری فاینتیون کردن که تبدیل شده به یه موتور تصمیمگیری برقآسا واسه ایجنتهای هوش مصنوعی

بچههای تیم جما با هنرنمایی مانو رومرو یه کار تر و تمیز راه انداختن و مدل EmbeddingGemma 2 رو جوری فاینتیون کردن که تبدیل شده به یه موتور تصمیمگیری برقآسا واسه ایجنتهای هوش مصنوعی. داستان اینه که دیگه لازم نیست منتظر تولید متنهای طولانی یا گیر کردن توی پارس کردن خروجیهای شکننده جیسون بمونی؛ این مدل فسقلی فقط با یه پردازش سریع، وضعیت کارها و تیکتها رو تحلیل میکنه و جوابهای شستهرفته و تیپبندی شده مثل امتیاز، درست/غلط یا گزینهای تحویلت میده. باحالی ماجرا اینجاست که با وجود داشتن پارامترهایی حدود صد برابر کمتر از یه مدل غولپیکر ۲۷ میلیاردی، نمره خفن ۷۶.۴ درصد گرفته و کلاً ۶ درصد با اون گندهها فاصله داره.
منبع
Google Gemma@googlegemma
Typed Decision Engine for Agent Workflows
Forget slow autoregressive generation and fragile JSON parsing.
This fine-tune turns EmbeddingGemma 2 into a fast decision engine that evaluates workflow states (like agent traces or support tickets) against a dynamic schema of questions in a single forward pass. It returns calibrated choice, score, and bool outputs without retraining.
Model and post by @mrm8488: https://x.com/mrm8488/status/2109406679045423394
Hugging Face: https://huggingface.co/mrm8488/embeddinggemma-2-ft-typed-decision
Manu Romero @mrm8488
🧭 New model: Typed Decision, built on EmbeddingGemma 2
Give it a state + any schema of questions → typed, calibrated answers (choice · score · noul). No text generation, no output parsing.
76.4% on typed-decisions, ~6 pts behind a 27B LLM with ~100× fewer params.
🤗 http://huggingface.co/mrm8488/embeddinggemma-2-ft-typed-decision
۵ آبان ۱۴۰۵مشاهده در X
huggingface.co
mrm8488/embeddinggemma-2-ft-typed-decision · Hugging Face
We’re on a journey to advance and democratize artificial intelligence through open source and open science.
خواندن در سایت منبع

نظرها
هنوز نظری نیست.
بخش نظرها فعلاً در دسترس نیست.