implement html output, fix log scaling in charts

Author: DaltheCow

.github/workflows/development.yml

@@ -263,6 +263,11 @@ jobs:
         with:
           fetch-depth: 0
 
+      - name: Set up Node.js 22
+        uses: actions/setup-node@v4
+        with:
+          node-version: '22'
+
       - name: Check if UI-related files changed
         id: check-changes
         run: |

.gitignore

@@ -6,6 +6,7 @@ src/guidellm/version.py
 benchmarks.json
 benchmarks.yaml
 benchmarks.csv
+benchmarks.html
 
 # Byte-compiled / optimized / DLL files
 __pycache__/
@@ -227,4 +228,5 @@ src/ui/next-env.d.ts
 !tsconfig.*.json
 !src/ui/lib
 !src/ui/public/manifest.json
+!src/ui/serve.json
 .eslintcache

DEVELOPING.md

@@ -267,6 +267,16 @@ Reference [https://www.npmjs.com/package/jest-runner-groups](jest-runner-groups)
  */
 ```
 
+### 🧪 UI Development Notes
+
+During development, it can be helpful to view sample data. We include a sample benchmark run wired into the Redux store under:
+
+```
+src/ui/lib/store/[runInfo/workloadDetails/benchmarks]WindowData.ts
+```
+
+In the future this will be replaced by a configurable untracked file for dev use.
+
 ## Additional Resources
 
 - [CONTRIBUTING.md](https://github.com/neuralmagic/guidellm/blob/main/CONTRIBUTING.md): Guidelines for contributing to the project.

README.md

@@ -157,17 +157,19 @@ The `guidellm benchmark` command is used to run benchmarks against a generative
 
 GuideLLM UI is a companion frontend for visualizing the results of a GuideLLM benchmark run.
 
-### 🛠 Running the UI
+### 🛠 Generating an HTML report with a benchmark run
 
-1. Use the Hosted Build (Recommended for Most Users)
-
-After running a benchmark with GuideLLM, a report.html file will be generated (by default at guidellm_report/report.html). This file references the latest stable version of the UI hosted at:
+Set the output to benchmarks.html for your run:
 
+```base
+--output-path=benchmarks.html
 ```
-https://neuralmagic.github.io/guidellm/ui/dev/
-```
 
-Open the file in your browser and you're done—no setup required.
+1. Use the Hosted Build (Recommended for Most Users)
+
+This is preconfigured. The latest stable version of the hosted UI (https://neuralmagic.github.io/guidellm/ui/latest) will be used to build the local html file.
+
+Open benchmarks.html in your browser and you're done—no setup required.
 
 2. Build and Serve the UI Locally (For Development) This option is useful if:
 
@@ -180,20 +182,16 @@ Open the file in your browser and you're done—no setup required.
 ```bash
 npm install
 npm run build
-npx serve out
+npm run serve
 ```
 
-This will start a local server (e.g., at http://localhost:3000). Then, in your GuideLLM config or CLI flags, point to this local server as the asset base for report generation.
-
-### 🧪 Development Notes
+This will start a local server (e.g., at http://localhost:3000). Then set the Environment to LOCAL before running your benchmarks.
 
-During UI development, it can be helpful to view sample data. We include a sample benchmark run wired into the Redux store under:
+```bash
+export GUIDELLM__ENV=local
 
+Alternatively, in config.py update the ENV_REPORT_MAPPING used as the asset base for report generation to the LOCAL option.
 ```
-src/lib/store/[runInfo/workloadDetails/benchmarks]WindowData.ts
-```
-
-In the future this will be replaced by a configurable untracked file for dev use.
 
 ## Resources