Model (Correction for flaw): Gemma-27b-qat

[DIGist]

https://www.reddit.com/r/LocalLLaMA/comments/1jxlqil/psa_google_have_fixed_the_qat_27_model/

A fixed Gemma-3-27b-it-qat:
https://huggingface.co/stduhpf/google-gemma-3-27b-it-qat-q4_0-gguf-small

[DIGist]

Original thread: https://www.reddit.com/r/LocalLLaMA/comments/1jvi860/psa_gemma_3_qat_gguf_models_have_some_wrongly/

Huggingface comment where it's mentioned to the Gemma3 team (14 days ago):
https://huggingface.co/google/gemma-3-12b-it-qat-q4_0-gguf/discussions/3#67f6a2e0207b4bceea793151

Then the Gemma3 team updated the Gemm3 QAT gguf 7 days ago:

As mentioned in the first thread I shared: "Google implemented the same fixes as we did, plus ours have the [general.name] metadata which is still lacking in the ggufs uploaded by Google! So you do not need to redownload the models."

The general.name difference still seems to exist in the upstream gemma3 27b qat but not the one I linked.