Precision/behavior failure of TensorRT 10.12 when running mixed FP32/FP16 image restoration on GPU RTX 3080 Laptop #117
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| name: Label New Issues | |
| on: | |
| issues: | |
| types: [opened] | |
| permissions: | |
| issues: write | |
| contents: read | |
| jobs: | |
| label-issue: | |
| runs-on: ubuntu-latest | |
| steps: | |
| - name: Checkout private action repository | |
| uses: actions/checkout@v4 | |
| with: | |
| repository: poweiw/goggles_action | |
| path: ./.github/actions/goggles_action # local path to store the action | |
| token: ${{ secrets.ACTION_REPO_PAT }} # token to access poweiw/goggles_action | |
| ref: v1.2.1 | |
| - name: AI Label Issue | |
| uses: ./.github/actions/goggles_action/actions/llm_label | |
| with: | |
| ACTION_TOKEN: ${{ secrets.GITHUB_TOKEN }} | |
| LLM_MODEL_NAME: ${{ secrets.LLM_MODEL_NAME }} | |
| LLM_TOKEN_SERVER_URL: ${{ secrets.LLM_TOKEN_SERVER_URL }} | |
| LLM_TOKEN_CLIENT_ID: ${{ secrets.LLM_TOKEN_CLIENT_ID }} | |
| LLM_TOKEN_CLIENT_SECRET: ${{ secrets.LLM_TOKEN_CLIENT_SECRET }} | |
| LLM_GENERATE_URL: ${{ secrets.LLM_GENERATE_URL }} | |
| LLM_TOKEN_SCOPE: ${{ secrets.LLM_TOKEN_SCOPE }} | |
| REPO_OWNER: ${{ github.repository_owner }} | |
| REPO_NAME: ${{ github.event.repository.name }} | |
| ISSUE_NUMBER: ${{ github.event.issue.number }} | |
| ISSUE_TITLE: ${{ github.event.issue.title }} | |
| ISSUE_BODY: ${{ github.event.issue.body }} | |
| GITHUB_API_URL: ${{ github.api_url }} | |
| ACTIONS_STEP_VERBOSE: false | |
| EXCLUDED_LABELS: "Investigating,internal-bug-tracked,stale,triaged,wontfix" | |
| LLM_SYSTEM_PROMPT: | | |
| You are an expert GitHub issue labeler. Your task is to analyze the provided issue title, issue body, and a list of available labels with their descriptions. | |
| Based on this information, select the single most appropriate label from the list that best captures the primary issue or request. | |
| Prefer selecting only one label that represents the main topic or problem. Only suggest multiple labels if the issue genuinely spans multiple distinct areas that are equally important. | |
| Respond with ONLY the chosen label name (e.g., 'bug', 'feature-request') or comma-separated names if multiple are truly needed. | |
| If no labels seem appropriate, respond with 'NONE'. | |
| Do not add any other text, explanation, or markdown formatting. |