Add Gemini LLM and Summarizer (#1176)

* add gemini

* add gemini

* add gemini

* update lockfile and respond to comments

* move types imports out

* updates and fixes

* requires nit

* remove print

* move to ubuntu-latest

.github/workflows/aryn-sdk_release.yml

@@ -11,7 +11,7 @@ permissions:
 jobs:
   release:
     name: Upload to PyPI
-    runs-on: blacksmith-4vcpu-ubuntu-2204
+    runs-on: ubuntu-latest
     environment:
       name: PyPI
       url: https://pypi.org/project/aryn-sdk/
@@ -22,7 +22,7 @@ jobs:
         uses: actions/checkout@v4
       - name: Install poetry
         run: pipx install poetry
-      - uses: useblacksmith/setup-python@v6
+      - uses: actions/setup-python@v4
         with:
           python-version: "3.11"
           cache: "poetry"

.github/workflows/codeql.yml

@@ -21,7 +21,7 @@ on:
 
 jobs:
   analyze:
-    runs-on: 'blacksmith'
+    runs-on: 'ubuntu-latest'
     permissions:
       # required for all workflows
       security-events: write

.github/workflows/dependency-review.yml

@@ -18,7 +18,7 @@ permissions:
 
 jobs:
   dependency-review:
-    runs-on: blacksmith
+    runs-on: ubuntu-latest
     steps:
       - name: 'Checkout repository'
         uses: actions/checkout@v4

.github/workflows/doclint.yml

@@ -9,7 +9,7 @@ on:
 
 jobs:
   doclint:
-    runs-on: blacksmith-4vcpu-ubuntu-2204
+    runs-on: ubuntu-latest
     steps:
       - name: Checkout
         uses: actions/checkout@v4

.github/workflows/draft_release.yml

@@ -10,13 +10,13 @@ permissions:
 
 jobs:
   draft-release:
-    runs-on: blacksmith-4vcpu-ubuntu-2204
+    runs-on: ubuntu-latest
     steps:
       - name: Checkout
         uses: actions/checkout@v4
       - name: Install poetry
         run: pipx install poetry
-      - uses: useblacksmith/setup-python@v6
+      - uses: actions/setup-python@v4
         with:
           python-version: '3.11'
           cache: 'poetry'

.github/workflows/linting.yml

@@ -11,7 +11,7 @@ on:
 # specific.
 jobs:
   mypy:
-    runs-on: blacksmith-4vcpu-ubuntu-2204
+    runs-on: ubuntu-latest
     steps:
       - name: DF-1
         run: df -h
@@ -29,7 +29,7 @@ jobs:
         run: df -h
       - name: Install poetry
         run: pipx install poetry
-      - uses: useblacksmith/setup-python@v6
+      - uses: actions/setup-python@v4
         with:
           python-version: '3.9'
           cache: 'poetry'
@@ -45,13 +45,13 @@ jobs:
           poetry run mypy --install-types --non-interactive .
 
   lint:
-    runs-on: blacksmith-4vcpu-ubuntu-2204
+    runs-on: ubuntu-latest
     steps:
       - name: Checkout
         uses: actions/checkout@v4
       - name: Install poetry
         run: pipx install poetry
-      - uses: useblacksmith/setup-python@v6
+      - uses: actions/setup-python@v4
         with:
           python-version: '3.9'
           cache: 'poetry'

.github/workflows/pypi_release.yml

@@ -5,7 +5,7 @@ on:
 jobs:
   test-pypi-release:
     name: Upload to PyPI
-    runs-on: blacksmith-4vcpu-ubuntu-2204
+    runs-on: ubuntu-latest
     environment:
       name: PyPI
       url: https://pypi.org/project/sycamore-ai/
@@ -16,7 +16,7 @@ jobs:
         uses: actions/checkout@v4
       - name: Install poetry
         run: pipx install poetry
-      - uses: useblacksmith/setup-python@v6
+      - uses: actions/setup-python@v4
         with:
           python-version: "3.11"
           cache: "poetry"

.github/workflows/testing.yml

@@ -12,6 +12,7 @@ on:
 env:
   OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
   ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
+  GEMINI_API_KEY: ${{ secrets.GEMINI_API_KEY }}
   AZURE_OPENAI_API_KEY: ${{ secrets.AZURE_OPENAI_API_KEY }}
   SYCAMORE_S3_TEMP_PATH: s3://aryn-sycamore-integ-temp/
   MODEL_SERVER_KEY: ${{ secrets.MODEL_SERVER_KEY }}
@@ -30,7 +31,7 @@ permissions:
 
 jobs:
   sycamore-unit-tests:
-    runs-on: blacksmith-8vcpu-ubuntu-2204
+    runs-on: ubuntu-latest
     strategy:
       matrix:
         python-version: ["3.9", "3.10", "3.11", "3.12"]
@@ -53,7 +54,7 @@ jobs:
       - name: Install poetry
         run: pipx install poetry
       - name: Set up Python ${{ matrix.python-version }}
-        uses: useblacksmith/setup-python@v6
+        uses: actions/setup-python@v4
         with:
           python-version: ${{ matrix.python-version }}
           cache: "poetry"

lib/sycamore/pyproject.toml

@@ -88,6 +88,7 @@ nltk = { version = "^3.9.0", optional = true }
 
 # LLM Dependencies
 anthropic = { version = "^0.42.0", optional = true }
+google-genai = {version = "^1.0", optional = true}
 tiktoken = "^0.8.0"
 
 [tool.poetry.group.test.dependencies]
@@ -143,6 +144,7 @@ legacy-partitioners = ["unstructured", "python-pptx", "nltk"]
 
 # LLM extras
 anthropic = ["anthropic"]
+google-genai = ["google-genai"]
 
 [tool.ruff]
 line-length = 120

lib/sycamore/sycamore/llms/__init__.py

@@ -4,6 +4,7 @@
 from sycamore.llms.openai import OpenAI, OpenAIClientType, OpenAIModels, OpenAIClientParameters, OpenAIClientWrapper
 from sycamore.llms.bedrock import Bedrock, BedrockModels
 from sycamore.llms.anthropic import Anthropic, AnthropicModels
+from sycamore.llms.gemini import Gemini, GeminiModels
 
 # Register the model constructors.
 MODELS: Dict[str, Callable[..., LLM]] = {}
@@ -16,6 +17,7 @@
 MODELS.update(
     {f"anthropic.{model.value}": lambda **kwargs: Anthropic(model.value, **kwargs) for model in AnthropicModels}
 )
+MODELS.update({f"gemini.{model.value}": lambda **kwargs: Gemini(model.value.name, **kwargs) for model in GeminiModels})
 
 
 def get_llm(model_name: str) -> Callable[..., LLM]:
@@ -38,4 +40,6 @@ def get_llm(model_name: str) -> Callable[..., LLM]:
     "BedrockModels",
     "Anthropic",
     "AnthropicModels",
+    "Gemini",
+    "GeminiModels",
 ]