[Dorijan10]

Description

Adds TurinTech's Artemis GGUF builds of Gemma 4 E2B and E4B to the model library, one entry per weight file.

• turintech/gemma-4-E2B-it-Artemis-GGUF: 4 builds. gemma-4-e2b-it-artemis-q3-k-l declares gemma-4-e2b-it-artemis-q3-k-m, gemma-4-e2b-it-artemis-iq3-xxs and gemma-4-e2b-it-artemis-iq2-m as variants.
• turintech/gemma-4-E4B-it-Artemis-GGUF: 3 builds. gemma-4-e4b-it-artemis-q4-0 declares gemma-4-e4b-it-artemis-iq3-m and gemma-4-e4b-it-artemis-q3-k-m as variants.

In each repo the declaring build is the file the model card marks as its default. The other builds declare no variants of their own.

Every entry also installs the repo's F16 vision projector, mmproj-F16.gguf, for image input. All builds of one repo use the same projector path, llama-cpp/mmproj/<repo>/mmproj-F16.gguf, so they share one download. The entries use virtual.yaml with the chat template embedded in the GGUF (use_jinja:true, use_tokenizer_template: true), an 8192-token context and known_usecases chat and vision. The rest of overrides copies the existing gemma-4-e4b-it entry, including function.automatic_tool_parsing_fallback: true. The first URL of each entry is the base model the card names, Google's QAT checkpoint.

Notes for Reviewers

• Same shape as the qwen3.5-0.8b-artemis-* entries from the same publisher, added in feat(gallery): add Qwen3.5 Artemis 0.8B variants #12707 and landed on master in 754800b, plus the F16 vision projector and full quant names.
• Companion to feat(gallery): add Qwen3.5 Artemis 2B, 4B and 9B variants #12928 (Qwen3.5 Artemis 2B, 4B and 9B). This PR inserts directly before the existing qwen3.5-0.8b-artemis-q5 and qwen3.5-0.8b-artemis-q4 entries and feat(gallery): add Qwen3.5 Artemis 2B, 4B and 9B variants #12928 directly after them, so the two apply independently.
• Pinned revisions:
  • turintech/gemma-4-E2B-it-Artemis-GGUF at 13a1c266e43df6e184779f8bc446ca6c66dca152
  • turintech/gemma-4-E4B-it-Artemis-GGUF at 31d9aa277c62a7036cde4e9aaa3c2d330e7775cf
• All sha256 values (seven weights and two F16 projectors) match the LFS metadata in the Hugging Face tree API, the x-linked-etag header at the pinned revision and the checksum table on each model card.
• GGUF architecture is gemma4 for both repos.
• gallery/index.yaml was regenerated with go run ./scripts/models/cmd/modelcatalog --library gallery/models/index.yaml --out gallery/index.yaml.
• These commands passed:
  • go test ./scripts/models/...
  • go test ./core/gallery -run 'TestModelLibrary|TestRealModelLibraryHasTheLibraryShape|TestGeneratedGalleryLoadsWithTheSameEntries|TestCheckedInGalleryIsTheGeneratedCopy|TestDecodeModelIndex' -count=1
  • go test ./core/gallery -run TestGallery -ginkgo.focus='gallery/index.yaml|gallery variant lint helpers' -count=1
  • git diff --check
• Smoke test on commit 5e08aa49 with the LocalAI v4.12.1 release binary on CPU (cpu-llama-cpp backend), serving a local copy of this branch's gallery/index.yaml as a file gallery. Each declaring entry was installed with POST /models/apply, then asked "What is the capital of France? Answer in one word." and, with the TurinTech logo as a data URL, "What text appears in this image?" (max_tokens 3000):
  • gemma-4-e2b-it-artemis-q3-k-l installed its own build, gemma-4-E2B-it-Artemis-Q3_K_L.gguf. Text: Paris. Image: The text that appears in the image is: **TURINTECH** (line break removed).
  • gemma-4-e4b-it-artemis-q4-0 installed its own build, gemma-4-E4B-it-Artemis-Q4_0.gguf. Text: Paris. Image: The text in the image is **TURINTECH**.
  • Three of the four replies carried a reasoning field: both image replies and the E2B text reply. The E4B text reply had none and used 2 completion tokens.
  • The sha256 of all four downloaded files matched the entries.
• The repos ship Google's MTP drafter (mtp-gemma-4-E2B-it-Artemis.gguf and mtp-gemma-4-E4B-it-Artemis.gguf, also under MTP/). These entries leave it unconfigured, with no draft_model, no spec_type option and no mtp tag, so they rank as plain llama.cpp builds.

Sources:

• https://huggingface.co/turintech/gemma-4-E2B-it-Artemis-GGUF
• https://huggingface.co/turintech/gemma-4-E4B-it-Artemis-GGUF
• https://huggingface.co/google/gemma-4-E2B-it-qat-q4_0-unquantized
• https://huggingface.co/google/gemma-4-E4B-it-qat-q4_0-unquantized

Signed commits

• Yes, I signed my commits.
• Documentation updated (docs/content/) for user-facing changes, or not applicable