Merge pull request #50 from genai-impact/feat/add-google-models

Feat/add google models
genai-impact · Jun 27, 2024 · 808f314 · 808f314
2 parents ad759eb + 701a3cd
commit 808f314
Show file tree

Hide file tree

Showing 14 changed files with 963 additions and 31 deletions.
diff --git a/docs/index.md b/docs/index.md
@@ -36,6 +36,7 @@ EcoLogits currently supports the following providers:
 
 - `anthropic`
 - `cohere`
+- `google-generativeai`
 - `huggingface-hub` (Hugging Face Inference Endpoints)
 - `mistralai`
 - `openai`

diff --git a/docs/providers.md b/docs/providers.md
@@ -6,6 +6,7 @@
 |------------------|------------------------|----------------------------------------------------------------------------------------| 
 | Anthropic        | `anthropic`            | [Guide for Anthropic :octicons-link-16:](tutorial/providers/anthropic.md)              |
 | Cohere           | `cohere`               | [Guide for Cohere :octicons-link-16:](tutorial/providers/cohere.md)                    |
+| Google Gemini    | `google-generativeai`  | [Guide for Google Gemini :octicons-link-16:](tutorial/providers/google.md)             |
 | Hugging Face Hub | `huggingface-hub`      | [Guide for Hugging Face Hub :octicons-link-16:](tutorial/providers/huggingface_hub.md) |
 | Mistral AI       | `mistralai`            | [Guide for Mistral AI :octicons-link-16:](tutorial/providers/mistralai.md)             |
 | OpenAI           | `openai`               | [Guide for OpenAI :octicons-link-16:](tutorial/providers/openai.md)                    |
@@ -17,6 +18,7 @@
 |-----------------|:----------------------------------:|:---------------------------------:|:---------------------------------:|:---------------------------------:|
 | Anthropic       | :material-checkbox-marked-circle:  |  :material-check-circle-outline:  | :material-checkbox-marked-circle: |  :material-check-circle-outline:  |
 | Cohere          | :material-checkbox-marked-circle:  | :material-checkbox-marked-circle: | :material-checkbox-marked-circle: | :material-checkbox-marked-circle: |
+| Google Gemini   | :material-checkbox-marked-circle:  | :material-checkbox-marked-circle: | :material-checkbox-marked-circle: | :material-checkbox-marked-circle: |
 | HuggingFace Hub | :material-checkbox-marked-circle:  | :material-checkbox-marked-circle: | :material-checkbox-marked-circle: | :material-checkbox-marked-circle: |
 | Mistral AI      | :material-checkbox-marked-circle:  | :material-checkbox-marked-circle: | :material-checkbox-marked-circle: | :material-checkbox-marked-circle: |
 | OpenAI          | :material-checkbox-marked-circle:  | :material-checkbox-marked-circle: | :material-checkbox-marked-circle: | :material-checkbox-marked-circle: |

diff --git a/docs/tutorial/providers/google.md b/docs/tutorial/providers/google.md
@@ -0,0 +1,120 @@
+# Google Gemini
+
+This guide is focuses on the integration of :seedling: **EcoLogits** with the [Google Gemini official python client :octicons-link-external-16:](https://github.com/google-gemini/generative-ai-python).
+
+Official links:
+
+* Repository: [:simple-github: google-gemini/generative-ai-python](https://github.com/google-gemini/generative-ai-python)
+* Documentation: [:material-file-document: ai.google.dev](https://ai.google.dev/gemini-api/docs?hl=fr)
+
+
+## Installation
+
+To install EcoLogits along with all necessary dependencies for compatibility with the Google Gemini client, please use the `google-generativeai` extra-dependency option as follows:
+
+```shell
+pip install ecologits[google-generativeai]
+```
+
+This installation command ensures that EcoLogits is set up with the specific libraries required to interface seamlessly with Google Gemini Python client.
+
+
+## Chat Completions
+
+### Example
+
+Integrating EcoLogits with your applications does not alter the standard outputs from the API responses. Instead, it enriches them by adding the `Impacts` object, which contains detailed environmental impact data.
+
+=== "Sync"
+
+    ```python
+    from ecologits import EcoLogits
+    import google.generativeai as genai
+
+    # Initialize EcoLogits
+    EcoLogits.init()
+
+    # Ask something to Google Gemini
+    genai.configure(api_key="<GOOGLE_API_KEY>")
+    model = genai.GenerativeModel("gemini-1.5-flash")
+    response = model.generate_content("Write a story about a magic backpack.")
+
+    # Get estimated environmental impacts of the inference
+    print(response.impacts)
+    ```
+
+=== "Async"
+
+    ```python
+    import asyncio
+    from ecologits import EcoLogits
+    import google.generativeai as genai
+
+    # Initialize EcoLogits
+    EcoLogits.init()
+
+    # Ask something to Google Gemini in async mode
+    async def main() -> None:
+        genai.configure(api_key="<GOOGLE_API_KEY>")
+        model = genai.GenerativeModel("gemini-1.5-flash")
+        response = await model.generate_content_async(
+            "Write a story about a magic backpack."
+        )
+
+        # Get estimated environmental impacts of the inference
+        print(response.impacts)
+
+    asyncio.run(main())
+    ```
+
+### Streaming example
+
+**In streaming mode, the impacts are calculated incrementally**, which means you don't need to sum the impacts from each data chunk. Instead, the impact information in the last chunk reflects the total cumulative environmental impacts for the entire request.
+
+=== "Sync" 
+
+    ```python
+    from ecologits import EcoLogits
+    import google.generativeai as genai
+
+    # Initialize EcoLogits
+    EcoLogits.init()
+
+    # Ask something to Google Gemini in streaming mode
+    genai.configure(api_key="<GOOGLE_API_KEY>")
+    model = genai.GenerativeModel("gemini-1.5-flash")
+    stream = model.generate_content(
+        "Write a story about a magic backpack.", 
+        stream=True
+    )
+
+    # Get cumulative estimated environmental impacts of the inference
+    for chunk in stream:
+        print(chunk.impacts)
+    ```
+
+=== "Async"
+
+    ```python
+    import asyncio
+    from ecologits import EcoLogits
+    import google.generativeai as genai
+
+    # Initialize EcoLogits
+    EcoLogits.init()
+
+    # Ask something to Google Gemini in streaming and async mode
+    async def main() -> None:
+        genai.configure(api_key="<GOOGLE_API_KEY>")
+        model = genai.GenerativeModel("gemini-1.5-flash")
+        stream = await model.generate_content_async(
+            "Write a story about a magic backpack.", 
+            stream=True
+        )
+
+        # Get cumulative estimated environmental impacts of the inference
+        async for chunk in stream:
+            print(chunk.impacts)
+
+    asyncio.run(main())
+    ```
diff --git a/ecologits/data/models.csv b/ecologits/data/models.csv
@@ -108,4 +108,7 @@ huggingface_hub,databricks/dbrx-instruct,132,36,,https://huggingface.co/databric
 huggingface_hub,databricks/dolly-v1-6b,6,6,,https://huggingface.co/databricks/dolly-v1-6b
 huggingface_hub,databricks/dolly-v2-12b,12,12,,https://huggingface.co/databricks/dolly-v2-12b
 huggingface_hub,databricks/dolly-v2-7b,7,7,,https://huggingface.co/databricks/dolly-v2-7b
-huggingface_hub,databricks/dolly-v2-3b,3,3,,https://huggingface.co/databricks/dolly-v2-3b
+huggingface_hub,databricks/dolly-v2-3b,3,3,,https://huggingface.co/databricks/dolly-v2-3b
+google,gemini-1.5-flash,20;70,20;70,model_architecture_not_released,https://deepmind.google/technologies/gemini/flash/
+google,gemini-1.5-pro,440,55;220,model_architecture_not_released,https://deepmind.google/technologies/gemini/pro/
+google,gemini-1.0-pro,20;70,20;70,model_architecture_not_released,https://deepmind.google/technologies/gemini/pro/
diff --git a/ecologits/ecologits.py b/ecologits/ecologits.py
@@ -49,6 +49,7 @@ def init_instruments() -> None:
     init_mistralai_instrumentor()
     init_huggingface_instrumentor()
     init_cohere_instrumentor()
+    init_google_instrumentor()
 
 
 def init_openai_instrumentor() -> None:
@@ -92,3 +93,11 @@ def init_cohere_instrumentor() -> None:
 
         instrumentor = CohereInstrumentor()
         instrumentor.instrument()
+
+
+def init_google_instrumentor() -> None:
+    if importlib.util.find_spec("google") is not None:
+        from ecologits.tracers.google_tracer import GoogleInstrumentor
+
+        instrumentor = GoogleInstrumentor()
+        instrumentor.instrument()
diff --git a/ecologits/model_repository.py b/ecologits/model_repository.py
@@ -11,6 +11,7 @@ class Providers(Enum):
     openai = "openai"
     huggingface_hub = "huggingface_hub"
     cohere = "cohere"
+    google = "google"
 
 
 class Warnings(Enum):
@@ -43,22 +44,28 @@ def find_model(self, provider: str, model_name: str) -> Optional[Model]:
     @classmethod
     def from_csv(cls, filepath: Optional[str] = None) -> "ModelRepository":
         if filepath is None:
-            filepath = os.path.join(os.path.dirname(os.path.realpath(__file__)), "data", "models.csv")
+            filepath = os.path.join(
+                os.path.dirname(os.path.realpath(__file__)), "data", "models.csv"
+            )
         models = []
         with open(filepath) as fd:
             csv = DictReader(fd)
             for row in csv:
                 total_parameters = None
                 total_parameters_range = None
                 if ";" in row["total_parameters"]:
-                    total_parameters_range = [float(p) for p in row["total_parameters"].split(";")]
+                    total_parameters_range = [
+                        float(p) for p in row["total_parameters"].split(";")
+                    ]
                 elif row["total_parameters"] != "":
                     total_parameters = float(row["total_parameters"])
 
                 active_parameters = None
                 active_parameters_range = None
                 if ";" in row["active_parameters"]:
-                    active_parameters_range = [float(p) for p in row["active_parameters"].split(";")]
+                    active_parameters_range = [
+                        float(p) for p in row["active_parameters"].split(";")
+                    ]
                 elif row["active_parameters"] != "":
                     active_parameters = float(row["active_parameters"])
 
@@ -70,16 +77,18 @@ def from_csv(cls, filepath: Optional[str] = None) -> "ModelRepository":
                 if row["sources"] != "":
                     sources = row["sources"].split(";")
 
-                models.append(Model(
-                    provider=Providers(row["provider"]).name,
-                    name=row["name"],
-                    total_parameters=total_parameters,
-                    active_parameters=active_parameters,
-                    total_parameters_range=total_parameters_range,
-                    active_parameters_range=active_parameters_range,
-                    warnings=warnings,
-                    sources=sources
-                ))
+                models.append(
+                    Model(
+                        provider=Providers(row["provider"]).name,
+                        name=row["name"],
+                        total_parameters=total_parameters,
+                        active_parameters=active_parameters,
+                        total_parameters_range=total_parameters_range,
+                        active_parameters_range=active_parameters_range,
+                        warnings=warnings,
+                        sources=sources,
+                    )
+                )
         return cls(models)