🚀 Gemma 4 DevOps Agents

Welcome to the Gemma-4 DevOps Agents workspace. This repository contains nine specialized, self-hosted AI-driven DevOps/SRE agents powered by Google's Gemma 4 model. These agents are packaged as Model Context Protocol (MCP) servers to analyze, monitor, and troubleshoot infrastructure components.

📂 Project Structure

This workspace is organized into 31 distinct sub-agents, each tailored to a specific environment, model configuration, and serving stack:

Sub-Agent Directory	Purpose / Role	Model Configuration	Serving Engine	Target Infrastructure
g2-4-12B-qat-L4-devops-agent	Serverless cloud SRE leveraging the 12B parameter QAT model on L4 GPU via Cloud Run Gen2.	`google/gemma-4-12B-it-qat-w4a16-ct`	vLLM	Google Cloud Run (L4 GPU)
gpu-12B-6000-devops-agent	Serverless cloud SRE running the 12B configuration on an NVIDIA RTX 6000 GPU.	`gemma-4-12B-it`	vLLM	Google Cloud Run (RTX 6000 GPU)
gpu-12B-L4-devops-agent	Serverless cloud SRE running the 12B configuration on an NVIDIA L4 GPU.	`gemma-4-12B-it`	vLLM	Google Cloud Run (L4 GPU)
gpu-12B-qat-L4-devops-agent	Serverless cloud SRE running the 12B QAT configuration on an NVIDIA L4 GPU.	`google/gemma-4-12B-it-qat-w4a16-ct`	vLLM	Google Cloud Run (L4 GPU)
gpu-12B-qat-mtp-6000-devops-agent	Serverless cloud SRE running the 12B QAT configuration with Multi-Token Prediction (MTP) on an RTX 6000 GPU.	`google/gemma-4-12B-it-qat-w4a16-ct`	vLLM	Google Cloud Run (RTX 6000 GPU)
gpu-12B-qat-mtp-L4-devops-agent	Serverless cloud SRE running the 12B QAT configuration with Multi-Token Prediction (MTP) on an NVIDIA L4 GPU.	`google/gemma-4-12B-it-qat-w4a16-ct`	vLLM	Google Cloud Run (L4 GPU)
gpu-26B-6000-devops-agent	Serverless cloud SRE running the 26B configuration on an NVIDIA RTX 6000 GPU.	`google/gemma-4-26B-it`	vLLM	Google Cloud Run (RTX 6000 GPU)
gpu-26B-L4-devops-agent	Serverless cloud SRE running the 26B configuration on an NVIDIA L4 GPU.	`nvidia/gemma-4-26B-A4B-NVFP4`	vLLM	Google Cloud Run (L4 GPU)
gpu-26B-qat-L4-devops-agent	Serverless cloud SRE leveraging the 26B QAT configuration on an NVIDIA L4 GPU.	`google/gemma-4-26B-A4B-it-qat-w4a16-ct`	vLLM	Google Cloud Run (L4 GPU)
gpu-31B-6000-devops-agent	Serverless cloud SRE running the 31B configuration on an NVIDIA RTX 6000 GPU.	`google/gemma-4-31B-it`	vLLM	Google Cloud Run (RTX 6000 GPU)
gpu-31B-L4-devops-agent	Serverless cloud SRE running the 31B NVFP4 quantized configuration on an NVIDIA L4 GPU.	`nvidia/Gemma-4-31B-IT-NVFP4`	vLLM	Google Cloud Run (L4 GPU)
gpu-31B-qat-L4-devops-agent	Serverless cloud SRE leveraging the 31B QAT configuration on an NVIDIA L4 GPU.	`google/gemma-4-31B-it-qat-w4a16-ct`	vLLM	Google Cloud Run (L4 GPU)
gpu-4B-6000-devops-agent	Serverless cloud SRE running the 4B configuration on an NVIDIA RTX 6000 GPU.	`google/gemma-4-E4B-it`	vLLM	Google Cloud Run (RTX 6000 GPU)
gpu-4B-L4-devops-agent	Serverless cloud SRE running the 4B configuration on an NVIDIA L4 GPU.	`google/gemma-4-E4B-it`	vLLM	Google Cloud Run (L4 GPU)
gpu-4B-qat-L4-devops-agent	Serverless cloud SRE leveraging the 4B QAT configuration on an NVIDIA L4 GPU.	`google/gemma-4-E4B-it-qat-w4a16-ct`	vLLM	Google Cloud Run (L4 GPU)
gpu-6000-devops-agent	Cloud-based SRE managing GPU-accelerated serverless endpoints (RTX 6000 GPU configuration).	`google/gemma-4-E4B-it`	vLLM	Google Cloud Run (RTX 6000 GPU)
gpu-vllm-devops-agent	Cloud-based SRE managing GPU-accelerated serverless endpoints (L4 GPU configuration).	`google/gemma-4-E4B-it`	vLLM	Google Cloud Run (L4 GPU)
local-2B-devops-agent	Local CPU/GPU DevOps/SRE Agent optimized for 2B parameter model execution.	`gemma4:e2b`	Ollama / vLLM	Local Workstation (Docker CPU/GPU)
local-devops-agent	Specialized SRE for local CPU/GPU containerized workloads.	`google/gemma-4-31B-it`	Ollama / vLLM	Local Workstation (Docker CPU/GPU)
mac-2B-devops-agent	Local macOS/Apple Silicon SRE Agent optimized for 2B parameter model execution on local hardware.	`gemma4:e2b`	Ollama / vLLM	Local macOS (Apple Silicon)
mac-4B-devops-agent	Local macOS/Apple Silicon SRE Agent optimized for 4B parameter model execution on local hardware.	`gemma4:e4b`	Ollama / vLLM	Local macOS (Apple Silicon)
tpu-12B-mtp-v6e1-devops-agent	High-performance TPU SRE/DevOps running the 12B configuration with Multi-Token Prediction (MTP) on a v6e-1 TPU VM.	`google/gemma-4-12B-it`	vLLM	Google Cloud TPUs (v6e Trillium v6e-1)
tpu-12B-quant-v6e1-devops-agent	High-performance TPU SRE/DevOps running the quantized 12B configuration (FP8) on a v6e-1 TPU VM.	`vrfai/gemma-4-12B-it-fp8`	vLLM	Google Cloud TPUs (v6e Trillium v6e-1)
tpu-12B-v6e1-devops-agent	High-performance TPU SRE/DevOps managing TPU workloads with the 12B configuration on a v6e-1 TPU VM.	`google/gemma-4-12B-it`	vLLM	Google Cloud TPUs (v6e Trillium v6e-1)
tpu-12B-v6e4-devops-agent	High-performance TPU SRE/DevOps managing TPU workloads with the 12B configuration on a v6e-4 TPU VM.	`google/gemma-4-12B-it`	Ollama / vLLM	Google Cloud TPUs (v6e Trillium v6e-4)
tpu-26B-devops-agent	High-performance TPU SRE/DevOps running the 26B configuration (running 31B model) on a v6e-1 TPU VM.	`google/gemma-4-31B-it`	vLLM	Google Cloud TPUs (v6e Trillium v6e-1)
tpu-2B-v6e1-devops-agent	High-performance TPU SRE/DevOps managing TPU workloads with the 2B configuration on a v6e-1 TPU VM.	`google/gemma-4-E2B-it`	Ollama / vLLM	Google Cloud TPUs (v6e Trillium v6e-1)
tpu-2B-v6e4-devops-agent	High-performance TPU SRE/DevOps managing TPU workloads with the 2B configuration on a v6e-4 TPU VM.	`google/gemma-4-E2B-it`	Ollama / vLLM	Google Cloud TPUs (v6e Trillium v6e-4)
tpu-31B-devops-agent	High-performance TPU SRE/DevOps running the 31B configuration on a v6e-1 TPU VM.	`google/gemma-4-31B-it`	vLLM	Google Cloud TPUs (v6e Trillium v6e-1)
tpu-4B-v6e1-devops-agent	High-performance TPU SRE/DevOps managing TPU workloads with the 4B configuration on a v6e-1 TPU VM.	`google/gemma-4-E4B-it`	Ollama / vLLM	Google Cloud TPUs (v6e Trillium v6e-1)
tpu-4B-v6e4-devops-agent	High-performance TPU SRE/DevOps managing TPU workloads with the 4B configuration on a v6e-4 TPU VM.	`google/gemma-4-E4B-it`	Ollama / vLLM	Google Cloud TPUs (v6e Trillium v6e-4)

🛠 Features & Capabilities

Automated SRE Diagnostics: Fetches and reviews system, container, and Cloud Logging entries using Gemma 4 to identify root causes and generate 3-step remediation plans.
Serving Stack Control: Built-in tools to provision, start, stop, restart, and scale your vLLM and Ollama containers or Cloud TPU Queued Resources.
Observability Dashboards: Real-time dashboards monitoring HBM usage, Tensor Core pressure, Prometheus metrics, and service latencies.
Model Benchmarking: Tools to run load tests and vLLM's internal benchmark suites, returning performance metrics (TTFT, throughput, P95 latency).
Gemini CLI Integration: Custom setup instructions using a LiteLLM Proxy to route standard Gemini CLI commands directly to your private, self-hosted Gemma 4 instance.

🏗 Global Makefile Usage

A root Makefile is provided to manage the sub-agents collectively:

Help / Display commands:
```
make all
```
Install dependencies in all subdirectories:
```
make install
```
Run tests across all agents:
```
make test
```
Lint all Python directories:
```
make lint
```
Clean build/cache folders:
```
make clean
```

🚀 Sub-Agent Overviews

1. GPU Agent (12B QAT L4 / Gen2)

Role: Serverless cloud SRE leveraging the 12B parameter QAT model on L4 GPU via Cloud Run Gen2.
Inference Stack: Runs google/gemma-4-12B-it-qat-w4a16-ct via vLLM.
Documentation: See g2-4-12B-qat-L4-devops-agent/README.md and g2-4-12B-qat-L4-devops-agent/GEMINI.md.

2. GPU Agent (12B 6000)

Role: Serverless cloud SRE running the 12B configuration on an NVIDIA RTX 6000 GPU.
Inference Stack: Runs gemma-4-12B-it via vLLM.
Documentation: See gpu-12B-6000-devops-agent/README.md and gpu-12B-6000-devops-agent/GEMINI.md.

3. GPU Agent (12B L4)

Role: Serverless cloud SRE running the 12B configuration on an NVIDIA L4 GPU.
Inference Stack: Runs gemma-4-12B-it via vLLM.
Documentation: See gpu-12B-L4-devops-agent/README.md and gpu-12B-L4-devops-agent/GEMINI.md.

4. GPU Agent (12B QAT L4)

Role: Serverless cloud SRE running the 12B QAT configuration on an NVIDIA L4 GPU.
Inference Stack: Runs google/gemma-4-12B-it-qat-w4a16-ct via vLLM.
Documentation: See gpu-12B-qat-L4-devops-agent/README.md and gpu-12B-qat-L4-devops-agent/GEMINI.md.

5. GPU Agent (12B QAT MTP 6000)

Role: Serverless cloud SRE running the 12B QAT configuration with Multi-Token Prediction (MTP) on an RTX 6000 GPU.
Inference Stack: Runs google/gemma-4-12B-it-qat-w4a16-ct via vLLM.
Documentation: See gpu-12B-qat-mtp-6000-devops-agent/README.md and gpu-12B-qat-mtp-6000-devops-agent/GEMINI.md.

6. GPU Agent (12B QAT MTP L4)

Role: Serverless cloud SRE running the 12B QAT configuration with Multi-Token Prediction (MTP) on an NVIDIA L4 GPU.
Inference Stack: Runs google/gemma-4-12B-it-qat-w4a16-ct via vLLM.
Documentation: See gpu-12B-qat-mtp-L4-devops-agent/README.md and gpu-12B-qat-mtp-L4-devops-agent/GEMINI.md.

7. GPU Agent (26B 6000)

Role: Serverless cloud SRE running the 26B configuration on an NVIDIA RTX 6000 GPU.
Inference Stack: Runs google/gemma-4-26B-it via vLLM.
Documentation: See gpu-26B-6000-devops-agent/README.md and gpu-26B-6000-devops-agent/GEMINI.md.

8. GPU Agent (26B L4)

Role: Serverless cloud SRE running the 26B configuration on an NVIDIA L4 GPU.
Inference Stack: Runs nvidia/gemma-4-26B-A4B-NVFP4 via vLLM.
Documentation: See gpu-26B-L4-devops-agent/README.md and gpu-26B-L4-devops-agent/GEMINI.md.

9. GPU Agent (26B QAT L4)

Role: Serverless cloud SRE leveraging the 26B QAT configuration on an NVIDIA L4 GPU.
Inference Stack: Runs google/gemma-4-26B-A4B-it-qat-w4a16-ct via vLLM.
Documentation: See gpu-26B-qat-L4-devops-agent/README.md and gpu-26B-qat-L4-devops-agent/GEMINI.md.

10. GPU Agent (31B 6000)

Role: Serverless cloud SRE running the 31B configuration on an NVIDIA RTX 6000 GPU.
Inference Stack: Runs google/gemma-4-31B-it via vLLM.
Documentation: See gpu-31B-6000-devops-agent/README.md and gpu-31B-6000-devops-agent/GEMINI.md.

11. GPU Agent (31B L4)

Role: Serverless cloud SRE running the 31B NVFP4 quantized configuration on an NVIDIA L4 GPU.
Inference Stack: Runs nvidia/Gemma-4-31B-IT-NVFP4 via vLLM.
Documentation: See gpu-31B-L4-devops-agent/README.md and gpu-31B-L4-devops-agent/GEMINI.md.

12. GPU Agent (31B QAT L4)

Role: Serverless cloud SRE leveraging the 31B QAT configuration on an NVIDIA L4 GPU.
Inference Stack: Runs google/gemma-4-31B-it-qat-w4a16-ct via vLLM.
Documentation: See gpu-31B-qat-L4-devops-agent/README.md and gpu-31B-qat-L4-devops-agent/GEMINI.md.

13. GPU Agent (4B 6000)

Role: Serverless cloud SRE running the 4B configuration on an NVIDIA RTX 6000 GPU.
Inference Stack: Runs google/gemma-4-E4B-it via vLLM.
Documentation: See gpu-4B-6000-devops-agent/README.md and gpu-4B-6000-devops-agent/GEMINI.md.

14. GPU Agent (4B L4)

Role: Serverless cloud SRE running the 4B configuration on an NVIDIA L4 GPU.
Inference Stack: Runs google/gemma-4-E4B-it via vLLM.
Documentation: See gpu-4B-L4-devops-agent/README.md and gpu-4B-L4-devops-agent/GEMINI.md.

15. GPU Agent (4B QAT L4)

Role: Serverless cloud SRE leveraging the 4B QAT configuration on an NVIDIA L4 GPU.
Inference Stack: Runs google/gemma-4-E4B-it-qat-w4a16-ct via vLLM.
Documentation: See gpu-4B-qat-L4-devops-agent/README.md and gpu-4B-qat-L4-devops-agent/GEMINI.md.

16. GPU Agent (6000)

Role: Cloud-based SRE managing GPU-accelerated serverless endpoints (RTX 6000 GPU configuration).
Inference Stack: Runs google/gemma-4-E4B-it via vLLM.
Documentation: See gpu-6000-devops-agent/README.md and gpu-6000-devops-agent/GEMINI.md.

17. GPU Agent (vLLM L4)

Role: Cloud-based SRE managing GPU-accelerated serverless endpoints (L4 GPU configuration).
Inference Stack: Runs google/gemma-4-E4B-it via vLLM.
Documentation: See gpu-vllm-devops-agent/README.md and gpu-vllm-devops-agent/GEMINI.md.

18. Local Agent (2B)

Role: Local CPU/GPU DevOps/SRE Agent optimized for 2B parameter model execution.
Inference Stack: Runs gemma4:e2b via Ollama / vLLM.
Documentation: See local-2B-devops-agent/README.md and local-2B-devops-agent/GEMINI.md.

19. Local Agent (31B default)

Role: Specialized SRE for local CPU/GPU containerized workloads.
Inference Stack: Runs google/gemma-4-31B-it via Ollama / vLLM.
Documentation: See local-devops-agent/README.md and local-devops-agent/GEMINI.md.

20. macOS Agent (2B)

Role: Local macOS/Apple Silicon SRE Agent optimized for 2B parameter model execution on local hardware.
Inference Stack: Runs gemma4:e2b via Ollama / vLLM.
Documentation: See mac-2B-devops-agent/README.md and mac-2B-devops-agent/GEMINI.md.

21. macOS Agent (4B)

Role: Local macOS/Apple Silicon SRE Agent optimized for 4B parameter model execution on local hardware.
Inference Stack: Runs gemma4:e4b via Ollama / vLLM.
Documentation: See mac-4B-devops-agent/README.md and mac-4B-devops-agent/GEMINI.md.

22. TPU Agent (12B MTP v6e-1)

Role: High-performance TPU SRE/DevOps running the 12B configuration with Multi-Token Prediction (MTP) on a v6e-1 TPU VM.
Inference Stack: Runs google/gemma-4-12B-it via vLLM.
Documentation: See tpu-12B-mtp-v6e1-devops-agent/README.md and tpu-12B-mtp-v6e1-devops-agent/GEMINI.md.

23. TPU Agent (12B Quant v6e-1)

Role: High-performance TPU SRE/DevOps running the quantized 12B configuration (FP8) on a v6e-1 TPU VM.
Inference Stack: Runs vrfai/gemma-4-12B-it-fp8 via vLLM.
Documentation: See tpu-12B-quant-v6e1-devops-agent/README.md and tpu-12B-quant-v6e1-devops-agent/GEMINI.md.

24. TPU Agent (12B v6e-1)

Role: High-performance TPU SRE/DevOps managing TPU workloads with the 12B configuration on a v6e-1 TPU VM.
Inference Stack: Runs google/gemma-4-12B-it via vLLM.
Documentation: See tpu-12B-v6e1-devops-agent/README.md and tpu-12B-v6e1-devops-agent/GEMINI.md.

25. TPU Agent (12B v6e-4)

Role: High-performance TPU SRE/DevOps managing TPU workloads with the 12B configuration on a v6e-4 TPU VM.
Inference Stack: Runs google/gemma-4-12B-it via Ollama / vLLM.
Documentation: See tpu-12B-v6e4-devops-agent/README.md and tpu-12B-v6e4-devops-agent/GEMINI.md.

26. TPU Agent (26B v6e-1)

Role: High-performance TPU SRE/DevOps running the 26B configuration (running 31B model) on a v6e-1 TPU VM.
Inference Stack: Runs google/gemma-4-31B-it via vLLM.
Documentation: See tpu-26B-devops-agent/README.md and tpu-26B-devops-agent/GEMINI.md.

27. TPU Agent (2B v6e-1)

Role: High-performance TPU SRE/DevOps managing TPU workloads with the 2B configuration on a v6e-1 TPU VM.
Inference Stack: Runs google/gemma-4-E2B-it via Ollama / vLLM.
Documentation: See tpu-2B-v6e1-devops-agent/README.md and tpu-2B-v6e1-devops-agent/GEMINI.md.

28. TPU Agent (2B v6e-4)

Role: High-performance TPU SRE/DevOps managing TPU workloads with the 2B configuration on a v6e-4 TPU VM.
Inference Stack: Runs google/gemma-4-E2B-it via Ollama / vLLM.
Documentation: See tpu-2B-v6e4-devops-agent/README.md and tpu-2B-v6e4-devops-agent/GEMINI.md.

29. TPU Agent (31B v6e-1)

Role: High-performance TPU SRE/DevOps running the 31B configuration on a v6e-1 TPU VM.
Inference Stack: Runs google/gemma-4-31B-it via vLLM.
Documentation: See tpu-31B-devops-agent/README.md and tpu-31B-devops-agent/GEMINI.md.

30. TPU Agent (4B v6e-1)

Role: High-performance TPU SRE/DevOps managing TPU workloads with the 4B configuration on a v6e-1 TPU VM.
Inference Stack: Runs google/gemma-4-E4B-it via Ollama / vLLM.
Documentation: See tpu-4B-v6e1-devops-agent/README.md and tpu-4B-v6e1-devops-agent/GEMINI.md.

31. TPU Agent (4B v6e-4)

Role: High-performance TPU SRE/DevOps managing TPU workloads with the 4B configuration on a v6e-4 TPU VM.
Inference Stack: Runs google/gemma-4-E4B-it via Ollama / vLLM.
Documentation: See tpu-4B-v6e4-devops-agent/README.md and tpu-4B-v6e4-devops-agent/GEMINI.md.

🔒 Security & Credentials

When deploying to Google Cloud or Hugging Face, secure credentials using:

Hugging Face Access Token: Saved locally or to Google Secret Manager.
Application Default Credentials (ADC): Set up using GCP credentials helper scripts.

Credits

Google Cloud credits are provided for this project.

#AgenticArchitect #GoogleAntigravity

Name		Name	Last commit message	Last commit date
Latest commit History 70 Commits
g2-4-12B-qat-L4-devops-agent		g2-4-12B-qat-L4-devops-agent
g2-4-26B-qat-L4-devops-agent		g2-4-26B-qat-L4-devops-agent
g2-4-2B-qat-L4-devops-agent		g2-4-2B-qat-L4-devops-agent
g2-4-31B-qat-L4-devops-agent		g2-4-31B-qat-L4-devops-agent
g2-4-4B-qat-L4-devops-agent		g2-4-4B-qat-L4-devops-agent
gpu-12B-6000-devops-agent		gpu-12B-6000-devops-agent
gpu-12B-L4-devops-agent		gpu-12B-L4-devops-agent
gpu-12B-qat-L4-devops-agent		gpu-12B-qat-L4-devops-agent
gpu-12B-qat-mtp-6000-devops-agent		gpu-12B-qat-mtp-6000-devops-agent
gpu-12B-qat-mtp-L4-devops-agent		gpu-12B-qat-mtp-L4-devops-agent
gpu-26B-6000-devops-agent		gpu-26B-6000-devops-agent
gpu-26B-L4-devops-agent		gpu-26B-L4-devops-agent
gpu-26B-qat-L4-devops-agent		gpu-26B-qat-L4-devops-agent
gpu-31B-6000-devops-agent		gpu-31B-6000-devops-agent
gpu-31B-L4-devops-agent		gpu-31B-L4-devops-agent
gpu-31B-qat-L4-devops-agent		gpu-31B-qat-L4-devops-agent
gpu-4B-6000-devops-agent		gpu-4B-6000-devops-agent
gpu-4B-L4-devops-agent		gpu-4B-L4-devops-agent
gpu-4B-qat-L4-devops-agent		gpu-4B-qat-L4-devops-agent
gpu-6000-devops-agent		gpu-6000-devops-agent
gpu-vllm-devops-agent		gpu-vllm-devops-agent
local-2B-devops-agent		local-2B-devops-agent
local-devops-agent		local-devops-agent
mac-2B-devops-agent		mac-2B-devops-agent
mac-4B-devops-agent		mac-4B-devops-agent
tpu-12B-mtp-v6e1-devops-agent		tpu-12B-mtp-v6e1-devops-agent
tpu-12B-quant-v6e1-devops-agent		tpu-12B-quant-v6e1-devops-agent
tpu-12B-v6e1-devops-agent		tpu-12B-v6e1-devops-agent
tpu-12B-v6e4-devops-agent		tpu-12B-v6e4-devops-agent
tpu-26B-v6e4-devops-agent		tpu-26B-v6e4-devops-agent
tpu-2B-v6e1-devops-agent		tpu-2B-v6e1-devops-agent
tpu-2B-v6e4-devops-agent		tpu-2B-v6e4-devops-agent
tpu-31B-devops-agent		tpu-31B-devops-agent
tpu-31B-v6e4-devops-agent		tpu-31B-v6e4-devops-agent
tpu-31B-v6e8-devops-agent		tpu-31B-v6e8-devops-agent
tpu-4B-v6e1-devops-agent		tpu-4B-v6e1-devops-agent
tpu-4B-v6e4-devops-agent		tpu-4B-v6e4-devops-agent
.gitignore		.gitignore
GEMINI.md		GEMINI.md
Makefile		Makefile
README.md		README.md
set_adc.sh		set_adc.sh

Folders and files

Latest commit

History

Repository files navigation

🚀 Gemma 4 DevOps Agents

📂 Project Structure

🛠 Features & Capabilities

🏗 Global Makefile Usage

🚀 Sub-Agent Overviews

1. GPU Agent (12B QAT L4 / Gen2)

2. GPU Agent (12B 6000)

3. GPU Agent (12B L4)

4. GPU Agent (12B QAT L4)

5. GPU Agent (12B QAT MTP 6000)

6. GPU Agent (12B QAT MTP L4)

7. GPU Agent (26B 6000)

8. GPU Agent (26B L4)

9. GPU Agent (26B QAT L4)

10. GPU Agent (31B 6000)

11. GPU Agent (31B L4)

12. GPU Agent (31B QAT L4)

13. GPU Agent (4B 6000)

14. GPU Agent (4B L4)

15. GPU Agent (4B QAT L4)

16. GPU Agent (6000)

17. GPU Agent (vLLM L4)

18. Local Agent (2B)

19. Local Agent (31B default)

20. macOS Agent (2B)

21. macOS Agent (4B)

22. TPU Agent (12B MTP v6e-1)

23. TPU Agent (12B Quant v6e-1)

24. TPU Agent (12B v6e-1)

25. TPU Agent (12B v6e-4)

26. TPU Agent (26B v6e-1)

27. TPU Agent (2B v6e-1)

28. TPU Agent (2B v6e-4)

29. TPU Agent (31B v6e-1)

30. TPU Agent (4B v6e-1)

31. TPU Agent (4B v6e-4)

🔒 Security & Credentials

Credits

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages