llm-autotune 0.1.0__tar.gz

llm_autotune-0.1.0/.claude/settings.local.json

@@ -0,0 +1,81 @@
+{
+  "permissions": {
+    "allow": [
+      "Bash(autotune recommend:*)",
+      "WebFetch(domain:ai.google.dev)",
+      "WebFetch(domain:mistral.ai)",
+      "Bash(autotune fetch:*)",
+      "Bash(autotune fetch-many:*)",
+      "Bash(autotune db:*)",
+      "Bash(pip install:*)",
+      "Bash(autotune session:*)",
+      "Bash(python -c \"from autotune.api.server import app; from autotune.api.chat import start_chat; from autotune.api.profiles import PROFILES; print\\(''imports OK''\\); print\\(''profiles:'', list\\(PROFILES.keys\\(\\)\\)\\)\")",
+      "Bash(autotune telemetry:*)",
+      "Bash(find /Users/tanavchinthapatla/Local LLM Optimizer/autotune/session -type f -name *.py)",
+      "Bash(python3 -c ':*)",
+      "Bash(python -m py_compile autotune/cli.py)",
+      "Bash(python -m py_compile autotune/session/advisor.py)",
+      "Bash(python -m py_compile autotune/session/dashboard.py)",
+      "Bash(python -m py_compile autotune/session/controller.py)",
+      "Bash(autotune stress-test:*)",
+      "Bash(python -c \"import sys; sys.path.insert\\(0, '.'\\); from scripts.stress_test import PROMPTS, SUSTAINED_PROMPTS, BIG_CONTEXT_PROMPT; print\\(f'Prompts OK: {len\\(PROMPTS\\)} main, {len\\(SUSTAINED_PROMPTS\\)} sustained'\\); print\\('Import clean'\\)\")",
+      "Bash(curl -s http://localhost:11434/api/tags)",
+      "Bash(python3 -c \"import json,sys; d=json.load\\(sys.stdin\\); print\\([m['name'] for m in d.get\\('models',[]\\)]\\)\")",
+      "Bash(python -c ':*)",
+      "Bash(python scripts/resource_test.py --model phi4-mini:latest --runs 3 --cold 3 --output resource_results.json)",
+      "Read(//private/tmp/**)",
+      "Read(//private/var/folders/**)",
+      "Bash(python3 -c \"from autotune.context import ContextWindow; print\\('ContextWindow imports OK'\\)\")",
+      "Bash(python3 -c \"from rich.console import Console; from rich.table import Table; from rich.progress import Progress; print\\('rich ok'\\)\")",
+      "Bash(timeout 120 python3 scripts/proof.py --model phi4-mini:latest --runs 2 --cold-runs 2 --skip-vram)",
+      "Bash(python3 scripts/proof.py --model phi4-mini:latest --runs 2 --cold-runs 2 --skip-vram)",
+      "Bash(python3 scripts/proof.py --model phi4-mini:latest --runs 3 --cold-runs 3)",
+      "Bash(python3 -c \"import ast; ast.parse\\(open\\('scripts/proof.py'\\).read\\(\\)\\); print\\('syntax ok'\\)\")",
+      "Bash(autotune proof:*)",
+      "Bash(python3 -c \"from autotune.memory.noswap import NoSwapGuard, ModelArch; print\\('noswap import ok'\\)\")",
+      "Bash(python3 scripts/proof.py --model phi4-mini:latest --with-noswap)",
+      "Bash(python -c \"from autotune.cli import cli; ctx = cli.make_context\\('autotune', ['chat', '--help']\\); \")",
+      "Bash(autotune chat:*)",
+      "WebSearch",
+      "Bash(autotune hardware:*)",
+      "Bash(python -c \"from autotune.hardware.profiler import profile_hardware; p = profile_hardware\\(\\); print\\(f'RAM: {p.total_ram_gb}GB, available: {p.available_ram_gb}GB, arch: {p.arch}'\\)\")",
+      "Bash(python -m py_compile autotune/cli.py autotune/api/chat.py autotune/api/kv_manager.py autotune/api/server.py autotune/api/model_selector.py autotune/api/backends/mlx_backend.py autotune/memory/noswap.py autotune/ttft/optimizer.py)",
+      "Bash(python -m py_compile scripts/proof.py scripts/benchmark.py scripts/stress_test.py scripts/resource_test.py)",
+      "Bash(ollama:*)",
+      "Bash(brew upgrade:*)",
+      "Read(//opt/homebrew/bin/**)",
+      "Read(//usr/local/bin/**)",
+      "Bash(python -m py_compile autotune/models/registry.py autotune/models/quality.py autotune/api/ollama_pull.py autotune/api/model_selector.py autotune/session/dashboard.py autotune/api/local_models.py autotune/bench/runner.py)",
+      "Bash(python -m py_compile autotune/api/chat.py autotune/api/conversation.py)",
+      "Bash(curl -s -X DELETE http://localhost:11434/api/delete -d '{\"model\":\"nonexistent\"}')",
+      "Bash(echo \"exit: $?\")",
+      "Bash(python -m autotune delete --help)",
+      "Bash(python -c \"from autotune.api.ollama_pull import delete_model; print\\('delete_model imported OK'\\)\")",
+      "Bash(python3:*)",
+      "Bash(python -m autotune memory --help)",
+      "Bash(autotune memory:*)",
+      "Bash(python -m autotune proof --help)",
+      "Bash(python -m py_compile autotune/ttft/optimizer.py autotune/api/profiles.py autotune/api/backends/chain.py)",
+      "Bash(pytest tests/ -q --tb=short)",
+      "Bash(pip index:*)",
+      "Bash(pip show *)",
+      "Bash(autotune serve *)",
+      "Bash(echo \"Server PID: $!\")",
+      "Bash(curl -s http://localhost:8765/health)",
+      "Bash(curl -s http://localhost:8765/v1/models)",
+      "Bash(pkill -f \"autotune serve\")",
+      "Bash(python -c \"from autotune.api.server import app; print\\('import OK'\\)\")",
+      "Bash(wait)",
+      "Bash(curl -s -X POST http://localhost:8765/v1/chat/completions -H 'Content-Type: application/json' -d '{\"model\":\"qwen3:8b\",\"messages\":[{\"role\":\"user\",\"content\":\"say exactly: Hello world\"}],\"stream\":true,\"max_tokens\":20}' --no-buffer)",
+      "Bash(python -m pytest tests/ -q)",
+      "Bash(python -c \"from autotune.api.server import app, _strip_thinking, _filter_thinking_stream, _is_thinking_model, completions; print\\('import OK'\\)\")",
+      "Bash(python -m pytest tests/)",
+      "Bash(python -m pytest tests/test_server_utils.py -v)",
+      "Bash(python -c \"from autotune.api.server import app, _VERSION; print\\(f'version={_VERSION}, app={app.version}'\\)\")",
+      "Bash(python -m pytest tests/ -v --tb=short)",
+      "Bash(python -m pytest tests/ -v)",
+      "Bash(rm -rf dist/)",
+      "Bash(python -m build)"
+    ]
+  }
+}

llm_autotune-0.1.0/.github/workflows/publish.yml

@@ -0,0 +1,124 @@
+name: Publish to PyPI
+
+on:
+  push:
+    tags:
+      - "v*"  # triggers on v0.1.0, v1.2.3, etc.
+
+permissions:
+  contents: read
+  id-token: write  # required for trusted publishing (OIDC)
+
+jobs:
+  build:
+    name: Build distribution
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+
+      - name: Install build tools
+        run: python -m pip install --upgrade pip build
+
+      - name: Build wheel and sdist
+        run: python -m build
+
+      - name: Upload build artifacts
+        uses: actions/upload-artifact@v4
+        with:
+          name: dist
+          path: dist/
+
+  publish-pypi:
+    name: Publish to PyPI
+    needs: build
+    runs-on: ubuntu-latest
+    environment:
+      name: pypi
+      url: https://pypi.org/project/llm-autotune/
+    steps:
+      - name: Download build artifacts
+        uses: actions/download-artifact@v4
+        with:
+          name: dist
+          path: dist/
+
+      - name: Publish to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1
+        # Uses OIDC trusted publishing — no API token needed.
+        # Set up trusted publisher at: https://pypi.org/manage/account/publishing/
+        # Publisher settings:
+        #   PyPI project name : llm-autotune
+        #   GitHub owner      : tanavc1
+        #   Repository name   : llm-autotune
+        #   Workflow filename  : publish.yml
+        #   Environment name  : pypi
+
+  update-homebrew-tap:
+    name: Update Homebrew tap
+    needs: publish-pypi
+    runs-on: ubuntu-latest
+    steps:
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version: "3.12"
+
+      - name: Install homebrew-pypi-poet
+        run: pip install homebrew-pypi-poet
+
+      - name: Wait for PyPI to propagate and generate formula
+        run: |
+          # Retry until the new version is visible on PyPI (up to ~5 min)
+          VERSION="${{ github.ref_name }}"
+          PKG_VERSION="${VERSION#v}"   # strip leading 'v'
+          for i in $(seq 1 10); do
+            if pip index versions llm-autotune 2>/dev/null | grep -q "$PKG_VERSION"; then
+              echo "Version $PKG_VERSION found on PyPI"
+              break
+            fi
+            echo "Attempt $i: version not yet visible, waiting 30s..."
+            sleep 30
+          done
+          poet -f llm-autotune > llm-autotune.rb
+          echo "--- Generated formula ---"
+          cat llm-autotune.rb
+
+      - name: Push formula to tap repo
+        uses: actions/github-script@v7
+        env:
+          TAP_TOKEN: ${{ secrets.TAP_GITHUB_TOKEN }}
+          VERSION: ${{ github.ref_name }}
+        with:
+          script: |
+            const fs = require('fs');
+            const formula = fs.readFileSync('llm-autotune.rb', 'utf8');
+            const { Octokit } = require('@octokit/rest');
+            const octokit = new Octokit({ auth: process.env.TAP_TOKEN });
+
+            const owner = 'tanavc1';
+            const repo = 'homebrew-autotune';
+            const path = 'Formula/llm-autotune.rb';
+
+            // Get current file SHA if it exists
+            let sha;
+            try {
+              const { data } = await octokit.rest.repos.getContent({ owner, repo, path });
+              sha = data.sha;
+            } catch (e) {
+              // File doesn't exist yet — first publish
+            }
+
+            await octokit.rest.repos.createOrUpdateFileContents({
+              owner,
+              repo,
+              path,
+              message: `chore: update formula for ${process.env.VERSION}`,
+              content: Buffer.from(formula).toString('base64'),
+              sha,
+            });
+            console.log('Tap updated successfully.');

llm_autotune-0.1.0/.github/workflows/test.yml

@@ -0,0 +1,41 @@
+name: Tests
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+    branches: [main]
+
+jobs:
+  test:
+    name: Test (Python ${{ matrix.python-version }})
+    runs-on: ubuntu-latest
+    strategy:
+      fail-fast: false
+      matrix:
+        python-version: ["3.10", "3.11", "3.12"]
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Set up Python ${{ matrix.python-version }}
+        uses: actions/setup-python@v5
+        with:
+          python-version: ${{ matrix.python-version }}
+          cache: pip
+
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          pip install -e ".[dev]"
+
+      - name: Run tests with coverage
+        run: |
+          pytest tests/ --cov=autotune --cov-report=term-missing --cov-report=xml -q
+
+      - name: Upload coverage report
+        uses: codecov/codecov-action@v4
+        if: matrix.python-version == '3.11'
+        with:
+          files: ./coverage.xml
+          fail_ci_if_error: false

llm_autotune-0.1.0/.gitignore

@@ -0,0 +1,33 @@
+__pycache__/
+*.py[cod]
+*.egg-info/
+.eggs/
+dist/
+build/
+*.egg
+.env
+.venv
+venv/
+env/
+
+# Database files (contain machine-specific data)
+*.db
+*.db-shm
+*.db-wal
+
+# Benchmark output (generated, not source)
+benchmark_results.json
+
+# macOS
+.DS_Store
+.AppleDouble
+.LSOverride
+
+# IDE
+.idea/
+.vscode/
+*.swp
+*.swo
+
+# autotune data dir (runtime)
+data/

llm_autotune-0.1.0/Formula/llm-autotune.rb

@@ -0,0 +1,32 @@
+# This file is auto-generated by the publish.yml GitHub Actions workflow
+# using homebrew-pypi-poet (https://github.com/tdsmith/homebrew-pypi-poet).
+#
+# To install:
+#   brew tap tanavc1/autotune
+#   brew install llm-autotune
+#
+# Or one-liner:
+#   brew install tanavc1/autotune/llm-autotune
+
+class LlmAutotune < Formula
+  include Language::Python::Virtualenv
+
+  desc "Automatic local-LLM inference configuration recommender"
+  homepage "https://github.com/tanavc1/local-llm-autotune"
+  url "https://files.pythonhosted.org/packages/source/l/llm-autotune/llm_autotune-0.1.0.tar.gz"
+  sha256 "PLACEHOLDER"  # filled automatically by publish.yml after PyPI upload
+  license "MIT"
+
+  depends_on "python@3.12"
+
+  # Resource SHAs are auto-populated by the publish workflow via `poet -f llm-autotune`.
+  # Do not edit this file by hand — push a new tag to trigger regeneration.
+
+  def install
+    virtualenv_install_with_resources
+  end
+
+  test do
+    assert_match version.to_s, shell_output("#{bin}/autotune --version")
+  end
+end

llm_autotune-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2024 Tanav Chinthapatla
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.