google
/

gemma-4-E4B-it

image-text-to-text

Model card Files Files and versions

Resources

View closed (18)

fix: Remediate 219 glitch tokens in tokenizer.json

#50 opened about 12 hours ago by

gemma4_E4B_it

#49 opened 12 days ago by

[Exploration] Applying Native Dual-Regime & KV-Cache Handoff to this model variant?

#48 opened 16 days ago by

Add ExtractBench evaluation results

#47 opened about 1 month ago by

boyang-runllama

Internal model structure preview

#46 opened about 1 month ago by

fix: chat_template, keep tool-call reasoning across later user turns

#45 opened 2 months ago by

Missing Apache 2.0 LICENSE File

#44 opened 2 months ago by

Update README.md

#43 opened 3 months ago by

Is there any quantization process done to this google/gemma-4-E4B-it model checkpoint?

#41 opened 3 months ago by

Raise per-image vision soft-token budget from 280 to 1120

#40 opened 3 months ago by

lucianommartins

Add IFStruct v1.0 evaluation result

#38 opened 3 months ago by

Chat template may re-inject prior-turn reasoning during multi-turn tool use → repetition loops

#37 opened 4 months ago by

Add MMMU-Pro evaluation result

#34 opened 5 months ago by

Request: DOI

#33 opened 5 months ago by

gemma-4-E4B-it

#32 opened 5 months ago by

GGUF available — Cerebellum v1 & v2 (ablation-guided mixed-precision)

#29 opened 5 months ago by

Image Dimension Test Results

#27 opened 5 months ago by

gemma-4-e4b-it

#25 opened 6 months ago by

Add ParseBench evaluation results

#24 opened 6 months ago by

boyang-runllama

Thanks!

#23 opened 6 months ago by

fix: embed chat_template in tokenizer_config.json

#21 opened 6 months ago by

task-mismatch

#19 opened 6 months ago by

[Bug] google/gemma-4-E4B-it tokenizer fails with AttributeError: 'list' object has no attribute 'keys' — extra_special_tokens format incompatibility with transformers v

#17 opened 6 months ago by

Error in Mistral Common Package

#15 opened 6 months ago by

fix: function calling formatting in chat template

#13 opened 6 months ago by

Unable to quantize model using bitsandbytes

#11 opened 6 months ago by

How Gemma 4 models were trained?

#9 opened 6 months ago by

Gemma4-E4B has emotion vectors — first replication of Anthropic's findings on open-source

#8 opened 6 months ago by

demo

#7 opened 6 months ago by

Gemma 4 MTP

#5 opened 6 months ago by

Add AIME 2026 evaluation result

#4 opened 6 months ago by

Add eval results

#2 opened 6 months ago by