google/gemma-3-27b-it · Discussions

#90 opened about 17 hours ago by

mondaylord

Support QLoRA Training on AWQ Quantized Models

#89 opened 2 days ago by

s3171103

Is there a minimum CUDA and/or BitsAndBytes version requirement?

#88 opened 3 days ago by

deathknight0

Update config file

#87 opened 16 days ago by

rishirath

therapeuticsgemma3

#86 opened 21 days ago by

mdordevic

Gemma 3 hyperparameters

#84 opened 30 days ago by

Augustas17

How Instructions, Tools and MCP terms are related

#83 opened about 1 month ago by

jvoid

Model Parameters number vs Quantization level tradeoff

#82 opened about 2 months ago by

jvoid

Suffix meaning IT vs PT

5

#81 opened about 2 months ago by

jvoid

No output / Repeated outputs when using Gemma 3 12B/27B on vLLM

#79 opened about 2 months ago by

sanchitahuja205

Genn

#78 opened about 2 months ago by

ezelhan959

qwz

#77 opened about 2 months ago by

Axdar

Specification

#76 opened 2 months ago by

zarifhaikal01

Too much VRAM in vLLM

#75 opened 2 months ago by

cbrug

Gemma-3-27B-it not loading on multi-GPU setup

#74 opened 3 months ago by

glawyer

Passing pictures as arrays to the model (Pytorch or Numpy)

#73 opened 3 months ago by

artyomboyko

Update config.json

#71 opened 3 months ago by

olegshulyakov

vllm and "gemma-3-27b-it" dont work

➕ 8

9

#70 opened 3 months ago by

nastyafairypro

How to eval the video/image sequences?

3

#69 opened 3 months ago by

lkllkl

Wrong context length limit for HuggingChat?

#68 opened 3 months ago by

vesjolovam

Undesired output texts

#67 opened 3 months ago by

Waisullah

Add reasoning capabilities for gemma 3

#66 opened 4 months ago by

devopsML

Getting 401 unaauthorized when trying to use google/gemma-3-27b-it

#65 opened 4 months ago by

Milinddurugkar

Codificação de imagens no prompt

5

#64 opened 4 months ago by

areumtecnologia

pad output?

#63 opened 4 months ago by

Juniworld

Wrong answer by gemma3-27b-it : what is the difference betweem each: google/gemma-3-27b-it, google/gemma-3-27b-pt

#62 opened 4 months ago by

Waisullah

Example of a wrong answer from an Image analysis

#61 opened 4 months ago by

vince62s

Test using M1 Max (64G) and Word

#60 opened 4 months ago by

gptlocalhost

what does "it" in the model name represent?

#59 opened 4 months ago by

jdsouza

Update README.md

#58 opened 4 months ago by

denizzhansahin

Gemma3ForCausalLM has no prepare_4d_causal_attention_mask_with_cache_position

#57 opened 4 months ago by

buckeye17-bah

Inference speed slow?

👍 2

#56 opened 4 months ago by

banank1989

Internal Server Error for url using vllm

#55 opened 4 months ago by

panyupj

Could not find the transformer layer class SiglipMultiheadAttentionPoolingHead in the model.

👍 2

rdtfddgrffdgfdghfghdfujgdhgsf

#54 opened 4 months ago by

neilwu

Keeps repeating itself.

3

#53 opened 4 months ago by

Why MLP so tiny but vision part of the model works quite well

#52 opened 4 months ago by

CCRss

Using google/gemma-3-4b-it as asssitant model for speculative decoding does not work

➕ 1

#51 opened 5 months ago by

sayambhu

Update README.md

#50 opened 5 months ago by

beita6969

SigLIP or SigLIP2 encoder?

12

#48 opened 5 months ago by

orrzohar

Updated model can't stop generation

#47 opened 5 months ago by

JackBAI

I'm releasing the speech version of Gemma-3!

🚀 1

#46 opened 5 months ago by

junnei

Regarding the issue of errors in inference when converting to float16.

#45 opened 5 months ago by

LBJ6666

Please release the weights of Gemini 1.5 Flash

#43 opened 5 months ago by

ZeroWw

Template error: syntax error: Conversation roles must alternate user/assistant/user/assistant/

#42 opened 5 months ago by

stcat

Blank String in System Prompt causes Error - chat_prompt needs some fixes!

#41 opened 5 months ago by

calycekr

Specify maximum context length in config

➕ 1

#40 opened 5 months ago by

saattrupdan

Tokens generated per second

👍 1

3

#39 opened 5 months ago by

rameshch

Image tokenization

#38 opened 5 months ago by

Marbuel

TPU inference

#37 opened 5 months ago by

Arpit-Bansal

How to use the tokenizer of gemma-3