BeefAndPoultry@lemmus.org to

LocalLLaMA@sh.itjust.worksEnglish · 3 days ago

Apparently Gemma 4's chat template can take over 10 seconds just to build the text when there are many tool calls

2

18

Apparently Gemma 4's chat template can take over 10 seconds just to build the text when there are many tool calls

BeefAndPoultry@lemmus.org to

LocalLLaMA@sh.itjust.worksEnglish · 3 days ago

2

google/gemma-4-E4B-it · Make the Gemma 4 chat template render in linear time

The chat template takes quadratic time in the number of messages. A long tool-calling conversation spends seconds in template rendering on every request, before the model does any work.

Not something I ever expected to cost a significant amount of time

You must log in or register to comment.

Chat

Spite_By_Design_LLC@lemmy.dbzer0.com
link
fedilink
English
arrow-up
1·
2 days ago
I thought Google fixed the Gemma template especially for tool calling…or is this ANOTHER regression?
- BeefAndPoultry@lemmus.orgOP
  link
  fedilink
  English
  arrow-up
  1·
  2 days ago
  this is not a regression just a performance inefficiency

LocalLLaMA@sh.itjust.works

localllama@sh.itjust.works

You are not logged in. However you can subscribe from another Fediverse account, for example Lemmy or Mastodon. To do this, paste the following into the search field of your instance: !localllama@sh.itjust.works

Welcome to LocalLLaMA! Here we discuss running and developing machine learning models at home. Lets explore cutting edge open source neural network technology together.

Get support from the community! Ask questions, share prompts, discuss benchmarks, get hyped at the latest and greatest model releases! Enjoy talking about our awesome hobby.

As ambassadors of the self-hosting machine learning community, we strive to support each other and share our enthusiasm in a positive constructive way.

Rules:

Rule 1 - No harassment or personal character attacks of community members. I.E no namecalling, no generalizing entire groups of people that make up our community, no baseless personal insults.

Rule 2 - No comparing artificial intelligence/machine learning models to cryptocurrency. I.E no comparing the usefulness of models to that of NFTs, no comparing the resource usage required to train a model is anything close to maintaining a blockchain/ mining for crypto, no implying its just a fad/bubble that will leave people with nothing of value when it burst.

Rule 3 - No comparing artificial intelligence/machine learning to simple text prediction algorithms. I.E statements such as "llms are basically just simple text predictions like what your phone keyboard autocorrect uses, and they’re still using the same algorithms since <over 10 years ago>.

Rule 4 - No implying that models are devoid of purpose or potential for enriching peoples lives.

Visibility: Public

This community can be federated to other instances and be posted/commented in by their users.

101 users / day
1.38K users / week
1.57K users / month
1.79K users / 6 months
1 local subscriber
5.57K subscribers
151 Posts
987 Comments
Modlog