Loading…

The morning briefing, in your inbox.

Top Cyprus + world stories, summarised. Every day.

Developed by Muflon

Tencent releases Hy3 open-source model matching larger models in performance · Cyprus Report

Tech · World

Tencent releases Hy3 open-source model matching larger models in performance

1 sourceFirst reported 15 days agoUpdated 15 days ago

AI summary: catch up in 20 seconds

Tencent has released an open-source language model called Hy3, featuring 295 billion parameters.
The Hy3 model operates using a mixture-of-experts architecture, activating only 21 billion parameters at any given time.
Tencent claims that Hy3 performs comparably to larger models that have two to five times the number of active parameters.
The company asserts that Hy3 also reduces instances of hallucination compared to other models.

Generated by AI from the 1 sources below · always verify with the originals

AI summary: catch up in 20 seconds

Summarising the coverage…

Generated by AI from the 1 sources below · always verify with the originals

All coverage (1)

Tencent releases Hy3 open-source model that allegedly matches models up to five times its active sizeThe Decoder · 15 days ago

No comments yet. Be the first to join the discussion.

AI summary: catch up in 20 seconds

Tencent has released an open-source language model called Hy3, featuring 295 billion parameters.
The Hy3 model operates using a mixture-of-experts architecture, activating only 21 billion parameters at any given time.
Tencent claims that Hy3 performs comparably to larger models that have two to five times the number of active parameters.
The company asserts that Hy3 also reduces instances of hallucination compared to other models.

Generated by AI from the 1 sources below · always verify with the originals