From 142bc3e02ffeea1b61c63e805161e5a34c2d7bbd Mon Sep 17 00:00:00 2001 From: waqasm86 Date: Sun, 19 Jul 2026 12:13:31 +0500 Subject: [PATCH] docs: present multi-model GGUF platform --- .github/workflows/profile-validate.yml | 2 +- CONTRIBUTING.md | 2 +- PULL_REQUEST_TEMPLATE.md | 2 +- SUPPORT.md | 2 +- assets/README.md | 21 +++++++++++++++++++++ docs/ORGANIZATION-OPERATIONS.md | 2 +- profile/README.md | 19 ++++++++++--------- 7 files changed, 36 insertions(+), 14 deletions(-) create mode 100644 assets/README.md diff --git a/.github/workflows/profile-validate.yml b/.github/workflows/profile-validate.yml index fb11116..d564605 100644 --- a/.github/workflows/profile-validate.yml +++ b/.github/workflows/profile-validate.yml @@ -34,7 +34,7 @@ jobs: https://github.com/Edge-Computing-LLM/edge-cli \ https://github.com/Edge-Computing-LLM/k3s-nvidia-edge \ https://github.com/Edge-Computing-LLM/llm-observability-stack \ - https://github.com/Edge-Computing-LLM/qwen-gguf-observability \ + https://github.com/Edge-Computing-LLM/gguf-observability \ https://github.com/Edge-Computing-LLM/edge-llm-tests; do curl --fail --location --silent --show-error --output /dev/null "$url" done diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md index 3b6ac82..3c8fb55 100644 --- a/CONTRIBUTING.md +++ b/CONTRIBUTING.md @@ -8,7 +8,7 @@ safe. Start by choosing the repository that owns the change. | Unified install, status, validation, logs, or orchestration | [`edge-cli`](https://github.com/Edge-Computing-LLM/edge-cli) | | Ubuntu, k3s, NVIDIA runtime, GPU Operator, device plugin, or DCGM substrate | [`k3s-nvidia-edge`](https://github.com/Edge-Computing-LLM/k3s-nvidia-edge) | | Ollama, GGUF, WebUI, OpenTelemetry, Prometheus, Grafana, Helm profiles, or benchmarks | [`llm-observability-stack`](https://github.com/Edge-Computing-LLM/llm-observability-stack) | -| Read-only Qwen runtime contract or evidence schema | [`qwen-gguf-observability`](https://github.com/Edge-Computing-LLM/qwen-gguf-observability) | +| Read-only GGUF model runtime contract or evidence schema | [`gguf-observability`](https://github.com/Edge-Computing-LLM/gguf-observability) | | Cross-repository validation and sanitized test reports | [`edge-llm-tests`](https://github.com/Edge-Computing-LLM/edge-llm-tests) | | Organization profile or shared community metadata | [`.github`](https://github.com/Edge-Computing-LLM/.github) | diff --git a/PULL_REQUEST_TEMPLATE.md b/PULL_REQUEST_TEMPLATE.md index 8e6c1a8..9649287 100644 --- a/PULL_REQUEST_TEMPLATE.md +++ b/PULL_REQUEST_TEMPLATE.md @@ -4,7 +4,7 @@ ## Owning layer - + ## Validation diff --git a/SUPPORT.md b/SUPPORT.md index d1b4171..8b5741a 100644 --- a/SUPPORT.md +++ b/SUPPORT.md @@ -5,7 +5,7 @@ Use the repository that owns the affected layer: - control-plane commands and ordering: [`edge-cli`](https://github.com/Edge-Computing-LLM/edge-cli/issues); - Ubuntu, k3s, or NVIDIA infrastructure: [`k3s-nvidia-edge`](https://github.com/Edge-Computing-LLM/k3s-nvidia-edge/issues); - Ollama, WebUI, Helm, or telemetry: [`llm-observability-stack`](https://github.com/Edge-Computing-LLM/llm-observability-stack/issues); -- Qwen runtime evidence: [`qwen-gguf-observability`](https://github.com/Edge-Computing-LLM/qwen-gguf-observability/issues); +- GGUF model runtime evidence: [`gguf-observability`](https://github.com/Edge-Computing-LLM/gguf-observability/issues); - test harness or report format: [`edge-llm-tests`](https://github.com/Edge-Computing-LLM/edge-llm-tests/issues). Before opening an issue, search existing issues and read the repository's diff --git a/assets/README.md b/assets/README.md new file mode 100644 index 0000000..76cfe12 --- /dev/null +++ b/assets/README.md @@ -0,0 +1,21 @@ +# Edge Computing LLM media library + +This directory is the canonical home for first-party organization images and +other publishable multimedia used by Edge Computing LLM. + +| Asset | Purpose | Dimensions | SHA-256 | +|---|---|---:|---| +| `edge-llm-hero.png` | Organization profile hero | 1983ร—793 | `e0441b8aab1e2187ee01e14850abe216cb5068bffee60261ac7e610855602777` | +| `edge-llm-avatar-v2.png` | Current GitHub organization avatar | 1254ร—1254 | `9e1e83b04733bf1369065ac65bafd59e02da23b4ed053d5acd629dd7258e7e62` | +| `edge-llm-avatar.png` | Detailed alternate avatar concept | 1254ร—1254 | `361b741df53cbdaf8416c65c8d95d0c5ffab6c87e9feb5d8f3f19842703add53` | + +There are currently no first-party videos or audio files. Add future profile +screenshots, diagrams, demonstrations, or recordings here with descriptive, +lowercase filenames. Document their source, purpose, license, creation date, +and checksum in this catalog. + +Do not copy media from vendored Helm charts, dashboard-template candidates, +dependency environments, backups, or unrelated event directories into this +library. Those files belong to their upstream projects and may carry different +licenses. Do not store screenshots containing host addresses, tokens, private +prompts, model responses, usernames, or other unsanitized runtime data. diff --git a/docs/ORGANIZATION-OPERATIONS.md b/docs/ORGANIZATION-OPERATIONS.md index 62b0076..c931d64 100644 --- a/docs/ORGANIZATION-OPERATIONS.md +++ b/docs/ORGANIZATION-OPERATIONS.md @@ -28,7 +28,7 @@ table whenever the organization positioning changes. | Field | Current value | Rationale | |---|---|---| | Display name | Edge Computing LLM | Human-readable form of the organization name. | -| Description | Go-first private LLM infrastructure and observability for Ubuntu, k3s, NVIDIA GPUs, and constrained edge systems | Matches the repository portfolio and deployed target platform. | +| Description | Go-first private multi-model LLM infrastructure and observability for Ubuntu, k3s, NVIDIA GPUs, and constrained edge systems | Matches the repository portfolio and deployed target platform. | | URL | `https://github.com/Edge-Computing-LLM/.github#readme` | Sends visitors to the maintained organization overview until a dedicated organization website exists. | | Location | Lahore, Pakistan | Matches the organization owner's published GitHub location. | | X / Twitter | `https://twitter.com/waqasm86` | Matches the organization owner's published GitHub social handle. | diff --git a/profile/README.md b/profile/README.md index 508910e..319e7f5 100644 --- a/profile/README.md +++ b/profile/README.md @@ -30,7 +30,7 @@ edge-cli CI k3s-nvidia-edge CI llm-observability-stack CI - qwen-gguf-observability CI + gguf-observability CI edge-llm-tests CI

@@ -49,7 +49,7 @@ flowchart LR CLI --> L2[llm-observability-stack
Layer 2] L1 --> K[(Ubuntu + k3s
NVIDIA runtime)] L2 --> A[(Ollama + WebUI
telemetry + dashboards)] - K --> Q[qwen-gguf-observability
runtime evidence] + K --> Q[gguf-observability
multi-model runtime evidence] A --> Q CLI --> T[edge-llm-tests
validation evidence] L1 --> T @@ -62,7 +62,7 @@ flowchart LR | ๐Ÿงญ [`edge-cli`](https://github.com/Edge-Computing-LLM/edge-cli) | Unified install, validation, status, logs, and safe uninstall workflows | Platform operators | | โš™๏ธ [`k3s-nvidia-edge`](https://github.com/Edge-Computing-LLM/k3s-nvidia-edge) | Ubuntu, k3s, containerd NVIDIA runtime, GPU Operator, device plugin, and DCGM substrate | Infrastructure contributors | | ๐Ÿ“ˆ [`llm-observability-stack`](https://github.com/Edge-Computing-LLM/llm-observability-stack) | Ollama/GGUF, Open WebUI, OpenTelemetry, Prometheus, Grafana, alerts, benchmarks, and Helm profiles | LLMOps and observability contributors | -| ๐Ÿ”Ž [`qwen-gguf-observability`](https://github.com/Edge-Computing-LLM/qwen-gguf-observability) | Dependency-free, read-only Qwen runtime contracts and privacy-safe evidence | Runtime verification | +| ๐Ÿ”Ž [`gguf-observability`](https://github.com/Edge-Computing-LLM/gguf-observability) | Dependency-free, read-only contracts and privacy-safe evidence for Qwen, Gemma, Llama, and future GGUF models | Runtime verification | | ๐Ÿงช [`edge-llm-tests`](https://github.com/Edge-Computing-LLM/edge-llm-tests) | Cross-repository Go/Git/Helm tests plus sanitized Ubuntu, k3s, NVIDIA, and inference results | Quality and reproducibility | ## Explore by goal @@ -72,7 +72,7 @@ flowchart LR | Install or operate the complete platform | [`edge-cli`](https://github.com/Edge-Computing-LLM/edge-cli#readme) | One control plane orders infrastructure and observability workflows. | | Prepare Ubuntu, k3s, and an NVIDIA GPU | [`k3s-nvidia-edge`](https://github.com/Edge-Computing-LLM/k3s-nvidia-edge#readme) | Layer 1 owns host and Kubernetes accelerator readiness. | | Deploy local inference and telemetry | [`llm-observability-stack`](https://github.com/Edge-Computing-LLM/llm-observability-stack#readme) | Layer 2 owns Ollama, WebUI, OpenTelemetry, metrics, dashboards, and profiles. | -| Verify the deployed Qwen contract | [`qwen-gguf-observability`](https://github.com/Edge-Computing-LLM/qwen-gguf-observability#readme) | Read-only checks capture privacy-safe runtime evidence. | +| Verify a deployed GGUF model contract | [`gguf-observability`](https://github.com/Edge-Computing-LLM/gguf-observability#readme) | Model-selectable, read-only checks capture privacy-safe runtime evidence. | | Reproduce cross-project quality checks | [`edge-llm-tests`](https://github.com/Edge-Computing-LLM/edge-llm-tests#readme) | One Go harness records source, chart, cluster, GPU, and smoke-test results. | ## What makes this stack different @@ -92,15 +92,16 @@ flowchart LR ## Reference edge path The project family is exercised on Ubuntu 24.04 with k3s and an NVIDIA GeForce -940M, including a Qwen 1.8B Chat Q4_K_M GGUF profile. This proves that useful -private inference and observability can fit into a 1 GiB VRAM envelope; it is a -constrained-edge reference, not a claim of fleet-scale production capacity. +940M using low-memory Qwen 1.8B, Gemma 3 1B, and Llama 3.2 1B profiles. Models +run sequentially with partial CUDA layer offload, system-RAM fallback, and a +900 MiB observed VRAM ceiling. This is a constrained-edge reference, not a +claim of fleet-scale production capacity. ```text Ubuntu host โ””โ”€โ”€ k3s + containerd โ”œโ”€โ”€ NVIDIA runtime + GPU resource - โ”œโ”€โ”€ Ollama + Qwen GGUF + โ”œโ”€โ”€ Ollama + selectable Qwen, Gemma, or Llama GGUF model โ”œโ”€โ”€ Open WebUI โ””โ”€โ”€ OpenTelemetry โ†’ Prometheus โ†’ Grafana ``` @@ -108,7 +109,7 @@ Ubuntu host Point-in-time claims belong in versioned evidence, not marketing copy. Browse [`edge-llm-tests`](https://github.com/Edge-Computing-LLM/edge-llm-tests) for the cross-project matrix and -[`qwen-gguf-observability`](https://github.com/Edge-Computing-LLM/qwen-gguf-observability) +[`gguf-observability`](https://github.com/Edge-Computing-LLM/gguf-observability) for the model-runtime contract. ## Get started