iHore/vLLM
8
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity
Files
b515648641227afe0d0fe92be7efac80471b295e
vLLM/recipes/models/Google/gemma-4-26B-A4B-it.yaml
T
Roy WangandClaude Opus 5.5 b515648641 [Doc] Host model recipes in recipes/ (#61302) 
Signed-off-by: esmeetu <jasonailu87@gmail.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-17 09:14:36 +00:00

603 lines
23 KiB
YAML
Raw Blame History

meta:
title: "Gemma 4 26B-A4B IT"
slug: "gemma-4-26b-a4b-it"
provider: "Google"
description: "Google's Gemma 4 MoE multimodal model (26B total / 4B active) with 128 fine-grained experts, top-8 routing, thinking mode, and tool-use protocol."
date_added: 2026-04-02
date_updated: 2026-09-27
tasks:
- multimodal
- text
performance_headline: "MoE multimodal model — 26B total / 4B active, 128 experts with top-8 routing"
related_recipes:
- google/gemma-4-E2B-it
- google/gemma-4-E4B-it
- google/gemma-4-31B-it
platforms:
- id: modal
blurb: "Serverless deploy via the Gemma 4 Modal script bundled with this recipe."
script: https://github.com/vllm-project/recipes/blob/main/Google/gemma4-modal.py
install: |
curl -O https://raw.githubusercontent.com/vllm-project/recipes/main/Google/gemma4-modal.py
pip install modal
modal setup
modal deploy gemma4-modal.py
modal run gemma4-modal.py
hardware:
h100: verified
h200: verified
mi300x: verified
mi325x: verified
mi355x: verified
trillium: verified
ironwood: verified
xeon6: verified
dgx_spark_gb10: verified
dgx_station_gb300: verified
rtx_pro_6000: verified
rtx_5090: verified
arc_pro_b60: verified
arc_pro_b70: verified
model:
model_id: "google/gemma-4-26B-A4B-it"
min_vllm_version: "0.25.0"
architecture: moe
parameter_count: "26B"
active_parameters: "4B"
context_length: 131072
base_args: []
base_env: {}
dependencies: []
features:
tool_calling:
description: "Enable automatic tool choice with Gemma 4 parser and chat template"
args:
- "--enable-auto-tool-choice"
- "--tool-call-parser"
- "gemma4"
- "--chat-template"
- "examples/tool_chat_template_gemma4.jinja"
hardware_overrides:
dgx_spark_gb10:
args:
- "--enable-auto-tool-choice"
- "--tool-call-parser"
- "gemma4"
reasoning:
description: "Enable structured thinking/reasoning output"
args:
- "--reasoning-parser"
- "gemma4"
text_only:
description: "Skip loading the vision encoder for text-only workloads — frees VRAM for KV cache. Mutually exclusive with encoder_parallel."
args:
- "--language-model-only"
spec_decoding:
description: "MTP speculative decoding for accelerated inference"
default_mode: mtp
modes:
mtp:
label: "MTP"
description: "Built-in Multi-Token Prediction, 4 draft tokens."
args:
- "--speculative-config"
- '{"model":"google/gemma-4-26B-A4B-it-assistant","num_speculative_tokens":4}'
# The single-GPU Blackwell desktops run the draft MoE on the Triton
# backend — the fused CUTLASS path the drafter picks by default is not
# built for GB10/GB300/RTX PRO in the validated images.
hardware_overrides:
dgx_spark_gb10: &gemma4_mtp_triton
args:
- "--speculative-config"
- '{"method":"mtp","model":"google/gemma-4-26B-A4B-it-assistant","num_speculative_tokens":4,"moe_backend":"triton"}'
dgx_station_gb300: *gemma4_mtp_triton
rtx_pro_6000: *gemma4_mtp_triton
eagle3:
label: "Eagle3"
description: "Eagle3 speculative decoding with externally trained draft model; the served checkpoint is unchanged."
variants:
- default
args:
- "--speculative-config"
- '{"model": "RedHatAI/gemma-4-26B-A4B-it-speculator.eagle3", "num_speculative_tokens": 3, "method": "eagle3"}'
opt_in_features:
- text_only
- spec_decoding
variants:
default:
precision: bf16
vram_minimum_gb: 64
description: "Full BF16 — single 80 GB NVIDIA GPU or 1x MI300X/MI325X/MI350X/MI355X or 2x Xeon6/Xeon5 NUMA nodes"
supported_hardware: [h100, h200, mi300x, mi325x, mi355x, trillium, ironwood, xeon6, dgx_spark_gb10, dgx_station_gb300, rtx_pro_6000, rtx_5090]
hardware_overrides:
dgx_spark_gb10:
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
dgx_station_gb300:
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
rtx_pro_6000:
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
extra_args:
- "--gpu-memory-utilization"
- "0.9"
fp8:
model_id: "RedHatAI/gemma-4-26B-A4B-it-FP8-dynamic"
precision: fp8
vram_minimum_gb: 32
description: "FP8 (E4M3) weights with dynamic per-token activation quantization — Hopper or Blackwell"
supported_hardware: [h100, h200, dgx_spark_gb10, dgx_station_gb300, rtx_pro_6000, rtx_5090]
extra_args:
- "--kv-cache-dtype"
- "fp8"
hardware_overrides:
dgx_spark_gb10:
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
dgx_station_gb300:
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
rtx_pro_6000:
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
extra_args:
- "--gpu-memory-utilization"
- "0.9"
rtx_5090:
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
extra_args:
- "--gpu-memory-utilization"
- "0.92"
- "--max-model-len"
- "131072"
- "--max-num-seqs"
- "1"
nvfp4:
model_id: "RedHatAI/gemma-4-26B-A4B-it-NVFP4"
precision: nvfp4
label: "NVFP4 - RedHatAI"
vram_minimum_gb: 16
description: "NVFP4 (4-bit) quantized weights — requires Blackwell (B200/B300)"
supported_hardware: [dgx_spark_gb10, dgx_station_gb300, rtx_pro_6000, rtx_5090]
extra_args:
- "--kv-cache-dtype"
- "fp8"
hardware_overrides:
dgx_spark_gb10:
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
dgx_station_gb300:
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
rtx_pro_6000:
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
extra_args:
- "--gpu-memory-utilization"
- "0.9"
rtx_5090:
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
extra_args:
- "--gpu-memory-utilization"
- "0.9"
- "--max-num-batched-tokens"
- "4096"
- "--max-num-seqs"
- "4"
nvidia_nvfp4:
model_id: "nvidia/Gemma-4-26B-A4B-NVFP4"
precision: nvfp4
label: "NVFP4 - NVIDIA"
vram_minimum_gb: 16
description: "NVIDIA ModelOpt NVFP4 (4-bit) quantized weights — validated on the single-GPU Blackwell desktops (DGX Spark GB10, DGX Station GB300, RTX PRO 6000)"
supported_hardware: [dgx_spark_gb10, dgx_station_gb300, rtx_pro_6000]
hardware_overrides:
dgx_spark_gb10:
# vLLM 0.24 and the generic nightly both fail on GB10 with this
# checkpoint; the dated ARM64 community build is the tested runtime, so
# the wheel tab is hidden here rather than offering a command that
# won't start.
docker_image: "eugr/spark-vllm:nightly-20260704"
# This image's entrypoint is nvidia_entrypoint.sh, not `vllm serve`.
docker_serve_command: "vllm serve"
install:
pip: false
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
extra_args:
# GB10 shares its 128 GB with the CPU, so leave headroom for the host.
- "--gpu-memory-utilization"
- "0.8"
- "--max-model-len"
- "262144"
- "--max-num-seqs"
- "8"
- "--max-num-batched-tokens"
- "8192"
- "--enable-prefix-caching"
- "--load-format"
- "fastsafetensors"
dgx_station_gb300:
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
extra_args:
# 252 GB on one GB300 — full context at high utilization and a 64-way
# batch.
- "--gpu-memory-utilization"
- "0.92"
- "--max-model-len"
- "262144"
- "--max-num-seqs"
- "64"
- "--max-num-batched-tokens"
- "8192"
- "--enable-chunked-prefill"
- "--enable-prefix-caching"
- "--load-format"
- "fastsafetensors"
rtx_pro_6000:
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
extra_args:
- "--gpu-memory-utilization"
- "0.92"
- "--max-model-len"
- "262144"
- "--max-num-seqs"
- "8"
- "--max-num-batched-tokens"
- "8192"
- "--enable-prefix-caching"
- "--load-format"
- "fastsafetensors"
rtx_5090:
extra_env:
VLLM_USE_V2_MODEL_RUNNER: "1"
extra_args:
- "--kv-cache-dtype"
- "fp8"
- "--gpu-memory-utilization"
- "0.9"
- "--max-num-batched-tokens"
- "4096"
- "--enable-prefix-caching"
- "--enable-chunked-prefill"
- "--async-scheduling"
- "--max-num-seqs"
- "4"
compatible_strategies:
- single_node_tp
- single_node_tep
- single_node_dep
- multi_node_tp
- multi_node_tep
- multi_node_dep
hardware_overrides:
xpu:
extra_args:
- "--enforce-eager"
- "--attention-backend"
- "flash_attn"
- "--trust-remote-code"
- "--gpu-memory-utilization"
- "0.9"
- "--no-enable-prefix-caching"
- "--max-num-batched-tokens"
- "8192"
- "--max-model-len"
- "5120"
- "--tensor-parallel-size"
- "4"
strategy_overrides:
single_node_tp:
tp: 1
guide: |
## Overview
[Gemma 4 26B-A4B](https://huggingface.co/google/gemma-4-26B-A4B-it) is the Mixture-of-Experts member of Google's Gemma 4 family — 26B total parameters with only 4B active per token via 128 fine-grained experts and top-8 routing. It supports text + images natively, structured thinking, function calling, and dynamic vision resolution.
Intel Xeon 6 CPU support remains verified. Benchmark-specific TP/DP, CPU binding,
sequence limits, and repro-only timeouts remain deployment choices.
### Key Features
- **MoE**: 128 fine-grained experts with top-8 routing and custom GELU-activated FFN.
- **Multimodal**: Text + images natively (video via custom frame-extraction pipeline). Audio is only supported on the smaller E2B/E4B variants.
- **Dual Attention**: Alternating sliding-window (local) and global attention with different head dimensions.
- **Thinking Mode**: Structured reasoning via `<|channel>thought\n...<channel|>` delimiters.
- **Function Calling**: Custom tool-call protocol with dedicated special tokens.
- **Dynamic Vision Resolution**: Per-request configurable vision token budget (70, 140, 280, 560, 1120 tokens).
TPU support is provided through [vLLM TPU](https://github.com/vllm-project/tpu-inference) with recipes for [Trillium](https://github.com/AI-Hypercomputer/tpu-recipes/tree/main/inference/trillium/vLLM/Gemma4) and [Ironwood](https://github.com/AI-Hypercomputer/tpu-recipes/blob/main/inference/ironwood/vLLM/Gemma4/).
## Prerequisites
- Hardware: Intel Xeon 6/Xeon 5 CPUs
- vLLM >= 0.25.0
### pip (NVIDIA CUDA)
```bash
uv venv
source .venv/bin/activate
uv pip install -U vllm --pre \
--extra-index-url https://wheels.vllm.ai/nightly/cu129 \
--extra-index-url https://download.pytorch.org/whl/cu129 \
--index-strategy unsafe-best-match
```
### pip (AMD ROCm: MI300X, MI325X, MI350X, MI355X)
Requires Python 3.12, ROCm 7.2.1, glibc >= 2.35 (Ubuntu 22.04+).
```bash
uv venv --python 3.12
source .venv/bin/activate
uv pip install vllm --pre \
--extra-index-url https://wheels.vllm.ai/rocm/nightly/rocm721 --upgrade
```
### pip (Intel Xeon 6 CPUs)
For Intel and AMD x86 CPUs, follow the [CPU pre-built wheels](https://docs.vllm.ai/en/latest/getting_started/installation/cpu/#pre-built-wheels) installation instructions.
### Docker
```bash
docker pull vllm/vllm-openai:gemma4-0505-cu129 # NVIDIA Hopper (H100/H200, CUDA 12.9)
docker pull vllm/vllm-openai:gemma4-0505-cu130 # NVIDIA Blackwell (B200/B300, CUDA 13.0)
docker pull vllm/vllm-openai-rocm:latest # AMD
docker pull vllm/vllm-openai-cpu:latest-x86_64 # For Intel Xeon 6
docker pull vllm/vllm-openai-xpu:latest # Intel XPU (B60 / B70)
```
TPU images are published separately by [vllm-project/tpu-inference](https://github.com/vllm-project/tpu-inference); see the Trillium / Ironwood tpu-recipes below for the pinned tag.
## Deployment Configurations
### 26B MoE on 1x A100/H100 (BF16)
```bash
vllm serve google/gemma-4-26B-A4B-it \
--max-model-len 32768 \
--gpu-memory-utilization 0.90
```
### Full-Featured Server Launch
Enables text, image, thinking, and tool calling:
```bash
vllm serve google/gemma-4-26B-A4B-it \
--max-model-len 16384 \
--gpu-memory-utilization 0.90 \
--enable-auto-tool-choice \
--reasoning-parser gemma4 \
--tool-call-parser gemma4 \
--chat-template examples/tool_chat_template_gemma4.jinja \
--limit-mm-per-prompt.image 4 \
--async-scheduling \
--host 0.0.0.0 \
--port 8000
```
### Docker (NVIDIA)
```bash
docker run -itd --name gemma4-moe \
--ipc=host --network host --shm-size 16G --gpus all \
-v ~/.cache/huggingface:/root/.cache/huggingface \
vllm/vllm-openai:gemma4-0505-cu129 \
--model google/gemma-4-26B-A4B-it \
--max-model-len 32768 \
--gpu-memory-utilization 0.90 \
--host 0.0.0.0 --port 8000
```
Swap `vllm/vllm-openai:gemma4-0505-cu129` for `vllm/vllm-openai:gemma4-0505-cu130` on Blackwell (B200/B300).
### Docker (AMD MI300X/MI325X/MI350X/MI355X)
```bash
docker run -itd --name gemma4-rocm \
--ipc=host --network=host --privileged \
--cap-add=CAP_SYS_ADMIN --device=/dev/kfd --device=/dev/dri \
--group-add=video --cap-add=SYS_PTRACE \
--security-opt=seccomp=unconfined --shm-size 16G \
-v ~/.cache/huggingface:/root/.cache/huggingface \
vllm/vllm-openai-rocm:latest \
--model google/gemma-4-26B-A4B-it \
--host 0.0.0.0 --port 8000
```
### Docker (Cloud TPU — Trillium / Ironwood)
TPU uses the separate `vllm/vllm-tpu` image (no pip wheel). Pull the tag specified by the upstream [Trillium](https://github.com/AI-Hypercomputer/tpu-recipes/tree/main/inference/trillium/vLLM/Gemma4) or [Ironwood](https://github.com/AI-Hypercomputer/tpu-recipes/blob/main/inference/ironwood/vLLM/Gemma4/) recipe, then run:
```bash
docker run -itd --name gemma4-tpu \
--privileged --network host --shm-size 16G \
-v /dev/shm:/dev/shm -e HF_TOKEN=$HF_TOKEN \
vllm/vllm-tpu:latest \
--model google/gemma-4-26B-A4B-it \
--tensor-parallel-size 8 \
--max-model-len 16384 \
--disable_chunked_mm_input \
--host 0.0.0.0 --port 8000
```
Trillium requires a 4-chip slice minimum; Ironwood runs on a single chip.
### Intel Xeon 6 Deployment via Docker
Launch the x86 CPU vLLM Docker container for `google/gemma-4-26B-A4B-it`:
```bash
docker run \
--privileged --ipc=host -p 8000:8000 \
-v ~/.cache/huggingface:/root/.cache/huggingface \
vllm/vllm-openai-cpu:latest-x86_64 google/gemma-4-26B-A4B-it \
--tensor-parallel-size 1 \
--enable-auto-tool-choice \
--tool-call-parser gemma4 \
--chat-template examples/tool_chat_template_gemma4.jinja \
--reasoning-parser gemma4
```
For additional Intel Xeon 6 deployment details, see the Intel Software Catalog entries for [Gemma 4 26B-A4B IT](https://aiswcatalog.intel.com/models/google-gemma-4-26b-a4b-it).
The following settings are intentionally not prescribed as portable CPU model
defaults because they depend on the workload, hardware, or runtime environment:
- `--max-num-batched-tokens`: scheduler/throughput tuning.
- `--max-num-seqs`: concurrency and scheduler-capacity tuning.
- `--gpu-memory-utilization`: platform memory-budget tuning.
- `--no-enable-prefix-caching`: workload/benchmark cache-policy tuning.
- `VLLM_ENGINE_ITERATION_TIMEOUT_S`: operational runtime timeout.
These settings may still appear in validated hardware-specific overrides. Tune
them at deployment time based on platform resources, workload shape, and
latency/throughput goals.
### Docker (Intel XPU B60 / B70, TP=4, Full-Featured)
Validated on 4x Intel Arc Pro B60 / B70 with the official vLLM XPU image.
```bash
docker run --device /dev/dri \
-v /dev/dri/by-path:/dev/dri/by-path --shm-size=16g \
--privileged --ipc=host -p 8000:8000 \
-v ~/.cache/huggingface:/root/.cache/huggingface \
vllm/vllm-openai-xpu:latest \
google/gemma-4-26B-A4B-it \
--enforce-eager \
--attention-backend flash_attn \
--host 0.0.0.0 \
--port 8000 \
--trust-remote-code \
--enable-auto-tool-choice \
--reasoning-parser gemma4 \
--tool-call-parser gemma4 \
--chat-template examples/tool_chat_template_gemma4.jinja \
--limit-mm-per-prompt.image 4 \
--async-scheduling \
--gpu-memory-utilization 0.9 \
--no-enable-prefix-caching \
--max-num-batched-tokens 8192 \
--max-model-len 5120 \
--tensor-parallel-size 4
```
## Client Usage
### Text Generation
```python
from openai import OpenAI
client = OpenAI(base_url="http://localhost:8000/v1", api_key="EMPTY")
response = client.chat.completions.create(
model="google/gemma-4-26B-A4B-it",
messages=[{"role": "user", "content": "Write a poem about the ocean."}],
max_tokens=512, temperature=0.7,
)
print(response.choices[0].message.content)
```
### Image Understanding
```python
response = client.chat.completions.create(
model="google/gemma-4-26B-A4B-it",
messages=[{"role": "user", "content": [
{"type": "image_url", "image_url": {"url": "https://upload.wikimedia.org/wikipedia/commons/thumb/3/3a/Cat03.jpg/1200px-Cat03.jpg"}},
{"type": "text", "text": "Describe this image in detail."},
]}],
max_tokens=1024,
)
```
### Thinking Mode
```bash
vllm serve google/gemma-4-26B-A4B-it \
--max-model-len 16384 \
--reasoning-parser gemma4 \
--tool-call-parser gemma4 \
--enable-auto-tool-choice \
--chat-template examples/tool_chat_template_gemma4.jinja
```
Enable per-request via `extra_body={"chat_template_kwargs": {"enable_thinking": True}}`.
### Dynamic Vision Resolution
Supported values: 70, 140, 280 (default), 560, 1120 tokens/image.
```bash
vllm serve google/gemma-4-26B-A4B-it \
--mm-processor-kwargs '{"max_soft_tokens": 560}'
```
## Configuration Tips
- Set `--max-model-len` to match your workload.
- `--gpu-memory-utilization 0.90-0.95` maximizes KV cache.
- Text-only workloads: `--limit-mm-per-prompt '{"image": 0, "audio": 0}'`.
- `--async-scheduling` improves throughput.
- FP8 KV cache (`--kv-cache-dtype fp8`) saves ~50% KV memory.
- For MoE, TEP (tensor-expert parallelism) and DEP (data-expert parallelism) strategies scale better than pure TP at large node counts.
## Quantized Variants
Three pre-quantized checkpoints are available:
- [`RedHatAI/gemma-4-26B-A4B-it-FP8-dynamic`](https://huggingface.co/RedHatAI/gemma-4-26B-A4B-it-FP8-dynamic) — FP8 (E4M3) weights with dynamic per-token activation quantization; runs on Hopper and Blackwell.
- [`RedHatAI/gemma-4-26B-A4B-it-NVFP4`](https://huggingface.co/RedHatAI/gemma-4-26B-A4B-it-NVFP4) — NVFP4 (4-bit) weights; requires Blackwell (B200/B300).
- [`nvidia/Gemma-4-26B-A4B-NVFP4`](https://huggingface.co/nvidia/Gemma-4-26B-A4B-NVFP4) — NVIDIA ModelOpt NVFP4 (4-bit) weights; this is the checkpoint validated on the single-GPU Blackwell desktops below.
Pick them from the **Variant** dropdown above, or pass the repo id directly to `vllm serve`.
### NVFP4 on the Blackwell desktops
[`nvidia/Gemma-4-26B-A4B-NVFP4`](https://huggingface.co/nvidia/Gemma-4-26B-A4B-NVFP4)
has been run end to end on DGX Spark (GB10), DGX Station (GB300) and RTX PRO
6000, in each case on one GPU at TP1 with MTP drafting against
[`google/gemma-4-26B-A4B-it-assistant`](https://huggingface.co/google/gemma-4-26B-A4B-it-assistant).
Select the **NVIDIA NVFP4** variant, enable **Spec Decoding**, and switch the
hardware pill — memory utilization, batch size, prefill scheduling and the
drafter's MoE backend all differ per box and are encoded per hardware, so the
generated command is the tested one. Highlights:
- **DGX Spark (GB10)** — 128 GB shared with the CPU, so `--gpu-memory-utilization 0.8`
and an 8-way batch. Serve from `eugr/spark-vllm:nightly-20260704`: vLLM 0.24
and the generic nightly both fail on this checkpoint, so the pip tab is
hidden on this hardware.
- **DGX Station (GB300)** — 252 GB on one GPU, so `--gpu-memory-utilization 0.92`
with chunked prefill and a 64-way batch.
- **RTX PRO 6000** — 96 GB, `--gpu-memory-utilization 0.92` and an 8-way batch.
All three run the full 262,144-token context, above the 131,072 in
`config.json`, and pass `"moe_backend":"triton"` in `--speculative-config`
because the drafter's default fused MoE path is not built in these images.
The GB300 and RTX PRO 6000 runs used an internal NVIDIA nightly vLLM build.
As noted under Speculative Decoding below, Gemma 4 MTP has not landed in a
stable release, so the default `vllm/vllm-openai:latest` tag will not serve
these commands with Spec Decoding enabled — use a nightly wheel or a nightly
image on those two boxes.
## Throughput vs Latency
| Goal | TP | `--max-num-seqs` | Notes |
|------|----|------------------|-------|
| Max throughput | 1-2 | 256-512 | Best tok/s per GPU |
| Min latency | 4-8 | 8-16 | Best TTFT/TPOT |
| Balanced | 2 | 128 | Mixed workloads |
## Speculative Decoding (MTP)
Enable the **Spec Decoding** feature toggle (above) or add `--speculative-config` manually to use MTP drafting with the [assistant model](https://huggingface.co/google/gemma-4-26B-A4B-it-assistant). Recommended `num_speculative_tokens`: 4 for this model. See the [Gemma 4 usage guide](../../Google/Gemma4) for details and benchmarks.
> **Note:** MTP speculative decoding for Gemma 4 is only available on the vLLM nightly build — it has not yet landed in a stable release. Install via the nightly wheel (`uv pip install -U vllm --pre --extra-index-url https://wheels.vllm.ai/nightly/cu129 …`) or use the `vllm/vllm-openai:gemma4-0505-cu129` / `vllm/vllm-openai:gemma4-0505-cu130` images above; the standard `:latest` stable tag does not include this feature.
## References
- [Model card](https://huggingface.co/google/gemma-4-26B-A4B-it)
- [FP8 variant](https://huggingface.co/RedHatAI/gemma-4-26B-A4B-it-FP8-dynamic)
- [NVFP4 variant](https://huggingface.co/RedHatAI/gemma-4-26B-A4B-it-NVFP4)
- [NVIDIA ModelOpt NVFP4 variant](https://huggingface.co/nvidia/Gemma-4-26B-A4B-NVFP4)
- [Gemma docs](https://ai.google.dev/gemma/docs)
- [vLLM Gemma 4 tool-call template](https://github.com/vllm-project/vllm/blob/main/examples/tool_chat_template_gemma4.jinja)
- [TPU recipes: Trillium](https://github.com/AI-Hypercomputer/tpu-recipes/tree/main/inference/trillium/vLLM/Gemma4)
- [TPU recipes: Ironwood](https://github.com/AI-Hypercomputer/tpu-recipes/blob/main/inference/ironwood/vLLM/Gemma4/)
- [Xeon 6 recipe: Gemma 4 26B-A4B IT](https://aiswcatalog.intel.com/models/google-gemma-4-26b-a4b-it)
Reference in New Issue View Git Blame Copy Permalink