vision-agent 1.0.2__tar.gz → 1.0.3__tar.gz

{vision_agent-1.0.2 → vision_agent-1.0.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vision-agent
-Version: 1.0.2
+Version: 1.0.3
 Summary: Toolset for Vision Agent
 Author: Landing AI
 Author-email: dev@landing.ai
@@ -12,8 +12,6 @@ Classifier: Programming Language :: Python :: 3.11
 Requires-Dist: anthropic (>=0.31.0,<0.32.0)
 Requires-Dist: av (>=11.0.0,<12.0.0)
 Requires-Dist: dotenv (>=0.9.9,<0.10.0)
-Requires-Dist: e2b (>=0.17.2a50,<0.18.0)
-Requires-Dist: e2b-code-interpreter (==0.0.11a37)
 Requires-Dist: flake8 (>=7.0.0,<8.0.0)
 Requires-Dist: ipykernel (>=6.29.4,<7.0.0)
 Requires-Dist: libcst (>=1.5.0,<2.0.0)
@@ -27,7 +25,7 @@ Requires-Dist: opentelemetry-api (>=1.29.0,<2.0.0)
 Requires-Dist: pandas (>=2.0.0,<3.0.0)
 Requires-Dist: pillow (>=10.0.0,<11.0.0)
 Requires-Dist: pillow-heif (>=0.16.0,<0.17.0)
-Requires-Dist: pydantic (==2.7.4)
+Requires-Dist: pydantic (>=2.0.0,<3.0.0)
 Requires-Dist: pytube (==15.0.0)
 Requires-Dist: requests (>=2.0.0,<3.0.0)
 Requires-Dist: rich (>=13.7.1,<14.0.0)

{vision_agent-1.0.2 → vision_agent-1.0.3}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "poetry.core.masonry.api"
 
 [tool.poetry]
 name = "vision-agent"
-version = "1.0.2"
+version = "1.0.3"
 description = "Toolset for Vision Agent"
 authors = ["Landing AI <dev@landing.ai>"]
 readme = "README.md"
@@ -34,13 +34,11 @@ nbclient = "^0.10.0"
 nbformat = "^5.10.4"
 rich = "^13.7.1"
 ipykernel = "^6.29.4"
-e2b = "^0.17.2a50"
-e2b-code-interpreter = "0.0.11a37"
 tenacity = "^8.3.0"
 pillow-heif = "^0.16.0"
 pytube = "15.0.0"
 anthropic = "^0.31.0"
-pydantic = "2.7.4"
+pydantic = "^2.0.0"
 av = "^11.0.0"
 libcst = "^1.5.0"
 matplotlib = "^3.9.2"
@@ -106,6 +104,4 @@ module = [
     "cv2.*",
     "openai.*",
     "sentence_transformers.*",
-    "e2b_code_interpreter.*",
-    "e2b.*"
 ]

{vision_agent-1.0.2 → vision_agent-1.0.3}/vision_agent/agent/vision_agent.py

@@ -293,7 +293,7 @@ class VisionAgent(Agent):
             callback_message (Optional[Callable[[Dict[str, Any]], None]]): Callback
                 function to send intermediate update messages.
             code_sandbox_runtime (Optional[str]): For string values it can be one of:
-                None, "local" or "e2b". If None, it will read from the environment
+                None or "local". If None, it will read from the environment
                 variable "CODE_SANDBOX_RUNTIME".
         """
 
@@ -560,7 +560,7 @@ class OpenAIVisionAgent(VisionAgent):
             callback_message (Optional[Callable[[Dict[str, Any]], None]]): Callback
                 function to send intermediate update messages.
             code_interpreter (Optional[Union[str, CodeInterpreter]]): For string values
-                it can be one of: None, "local" or "e2b". If None, it will read from
+                it can be one of: None or "local". If None, it will read from
                 the environment variable "CODE_SANDBOX_RUNTIME". If a CodeInterpreter
                 object is provided it will use that.
         """
@@ -591,7 +591,7 @@ class AnthropicVisionAgent(VisionAgent):
             callback_message (Optional[Callable[[Dict[str, Any]], None]]): Callback
                 function to send intermediate update messages.
             code_interpreter (Optional[Union[str, CodeInterpreter]]): For string values
-                it can be one of: None, "local" or "e2b". If None, it will read from
+                it can be one of: None or "local". If None, it will read from
                 the environment variable "CODE_SANDBOX_RUNTIME". If a CodeInterpreter
                 object is provided it will use that.
         """

{vision_agent-1.0.2 → vision_agent-1.0.3}/vision_agent/agent/vision_agent_coder.py

@@ -349,7 +349,7 @@ class VisionAgentCoder(Agent):
                 running in parallel. This callback ensures that the progress are not
                 mixed up.
             code_interpreter (Optional[Union[str, CodeInterpreter]]): For string values
-                it can be one of: None, "local" or "e2b". If None, it will read from
+                it can be one of: None or "local". If None, it will read from
                 the environment variable "CODE_SANDBOX_RUNTIME". If a CodeInterpreter
                 object is provided it will use that.
         """

{vision_agent-1.0.2 → vision_agent-1.0.3}/vision_agent/agent/vision_agent_coder_v2.py

@@ -317,7 +317,7 @@ class VisionAgentCoderV2(AgentCoder):
             hil (bool): Whether to use human-in-the-loop mode.
             verbose (bool): Whether to print out debug information.
             code_sandbox_runtime (Optional[str]): The code sandbox runtime to use, can
-                be one of: None, "local" or "e2b". If None, it will read from the
+                be one of: None or "local". If None, it will read from the
                 environment variable CODE_SANDBOX_RUNTIME.
             update_callback (Callable[[Dict[str, Any]], None]): The callback function
                 that will send back intermediate conversation messages.

{vision_agent-1.0.2 → vision_agent-1.0.3}/vision_agent/agent/vision_agent_planner_v2.py

@@ -442,7 +442,7 @@ class VisionAgentPlannerV2(AgentPlanner):
             hil (bool): Whether to use human-in-the-loop mode.
             verbose (bool): Whether to print out debug information.
             code_sandbox_runtime (Optional[str]): The code sandbox runtime to use, can
-                be one of: None, "local" or "e2b". If None, it will read from the
+                be one of: None or "local". If None, it will read from the
                 environment variable CODE_SANDBOX_RUNTIME.
             update_callback (Callable[[Dict[str, Any]], None]): The callback function
                 that will send back intermediate conversation messages.

{vision_agent-1.0.2 → vision_agent-1.0.3}/vision_agent/agent/vision_agent_v2.py

@@ -180,7 +180,7 @@ class VisionAgentV2(Agent):
             hil (bool): Whether to use human-in-the-loop mode.
             verbose (bool): Whether to print out debug information.
             code_sandbox_runtime (Optional[str]): The code sandbox runtime to use, can
-                be one of: None, "local" or "e2b". If None, it will read from the
+                be one of: None or "local". If None, it will read from the
                 environment variable CODE_SANDBOX_RUNTIME.
             update_callback (Callable[[Dict[str, Any]], None]): The callback function
                 that will send back intermediate conversation messages.

{vision_agent-1.0.2 → vision_agent-1.0.3}/vision_agent/utils/execute.py

@@ -13,16 +13,7 @@ from time import sleep
 from typing import Any, Dict, Iterable, List, Optional, Union
 
 import nbformat
-import tenacity
 from dotenv import load_dotenv
-from e2b.exceptions import SandboxException
-from e2b_code_interpreter import CodeInterpreter as E2BCodeInterpreterImpl
-from e2b_code_interpreter import Execution as E2BExecution
-from e2b_code_interpreter import Result as E2BResult
-from h11._util import LocalProtocolError
-from httpx import ConnectError
-from httpx import RemoteProtocolError as HttpcoreRemoteProtocolError
-from httpx import RemoteProtocolError as HttpxRemoteProtocolError
 from nbclient import NotebookClient
 from nbclient import __version__ as nbclient_version
 from nbclient.exceptions import CellTimeoutError, DeadKernelError
@@ -33,11 +24,6 @@ from opentelemetry.trace import SpanKind, Status, StatusCode, get_tracer
 from pydantic import BaseModel, field_serializer
 from typing_extensions import Self
 
-from vision_agent.utils.exceptions import (
-    RemoteSandboxCreationError,
-    RemoteSandboxExecutionError,
-)
-
 load_dotenv()
 _LOGGER = logging.getLogger(__name__)
 _SESSION_TIMEOUT = 600  # 10 minutes
@@ -215,23 +201,6 @@ class Result:
             formats.extend(iter(self.extra))
         return formats
 
-    @staticmethod
-    def from_e2b_result(result: E2BResult) -> "Result":
-        """
-        Creates a Result object from an E2BResult object.
-        """
-        data = {
-            MimeType.TEXT_PLAIN.value: result.text,
-            MimeType.IMAGE_PNG.value: result.png,
-            MimeType.APPLICATION_JSON.value: result.json,
-        }
-        for k, v in result.extra.items():
-            data[k] = v
-        return Result(
-            is_main_result=result.is_main_result,
-            data=data,
-        )
-
 
 class Logs(BaseModel):
     """Data printed to stdout and stderr during execution, usually by print statements,
@@ -372,26 +341,6 @@ class Execution(BaseModel):
             )
         )
 
-    @staticmethod
-    def from_e2b_execution(exec: E2BExecution) -> "Execution":
-        """Creates an Execution object from an E2BResult object."""
-        return Execution(
-            results=[Result.from_e2b_result(res) for res in exec.results],
-            logs=Logs(stdout=exec.logs.stdout, stderr=exec.logs.stderr),
-            error=(
-                Error(
-                    name=exec.error.name,
-                    value=_remove_escape_and_color_codes(exec.error.value),
-                    traceback_raw=[
-                        _remove_escape_and_color_codes(line)
-                        for line in exec.error.traceback.split("\n")
-                    ],
-                )
-                if exec.error
-                else None
-            ),
-        )
-
 
 class CodeInterpreter(abc.ABC):
     """Code interpreter interface."""
@@ -439,138 +388,6 @@ class CodeInterpreter(abc.ABC):
         return Path(local_file_path)
 
 
-class E2BCodeInterpreter(CodeInterpreter):
-    def __init__(
-        self, remote_path: Optional[Union[str, Path]] = None, *args: Any, **kwargs: Any
-    ) -> None:
-        super().__init__(*args, **kwargs)
-        assert os.getenv("E2B_API_KEY"), "E2B_API_KEY environment variable must be set"
-        try:
-            self.interpreter = E2BCodeInterpreter._new_e2b_interpreter_impl(
-                *args, **kwargs
-            )
-        except Exception as e:
-            raise RemoteSandboxCreationError(
-                f"Failed to create a remote sandbox due to {e}"
-            ) from e
-
-        result = self.exec_cell(
-            """
-import platform
-import sys
-import importlib.metadata
-
-print(f"Python version: {sys.version}")
-print(f"OS version: {platform.system()} {platform.release()} ({platform.architecture()})")
-va_version = importlib.metadata.version("vision-agent")
-print(f"Vision Agent version: {va_version}")"""
-        )
-        sys_versions = "\n".join(result.logs.stdout)
-        _LOGGER.info(
-            f"E2BCodeInterpreter (sandbox id: {self.interpreter.sandbox_id}) initialized:\n{sys_versions}"
-        )
-        self.remote_path = Path(
-            remote_path if remote_path is not None else "/home/user"
-        )
-
-    def close(self, *args: Any, **kwargs: Any) -> None:
-        try:
-            self.interpreter.kill(request_timeout=2)
-            _LOGGER.info(
-                f"The sandbox {self.interpreter.sandbox_id} is closed successfully."
-            )
-        except Exception as e:
-            _LOGGER.warn(
-                f"Failed to close the remote sandbox ({self.interpreter.sandbox_id}) due to {e}. This is not an issue. It's likely that the sandbox is already closed due to timeout."
-            )
-
-    def restart_kernel(self) -> None:
-        self.interpreter.notebook.restart_kernel()
-
-    @tenacity.retry(
-        wait=tenacity.wait_exponential_jitter(),
-        stop=tenacity.stop_after_attempt(3),
-        retry=tenacity.retry_if_exception_type(
-            (
-                LocalProtocolError,
-                HttpxRemoteProtocolError,
-                HttpcoreRemoteProtocolError,
-                ConnectError,
-                SandboxException,
-            )
-        ),
-        before_sleep=tenacity.before_sleep_log(_LOGGER, logging.INFO),
-        after=tenacity.after_log(_LOGGER, logging.INFO),
-    )
-    def exec_cell(self, code: str) -> Execution:
-        self.interpreter.set_timeout(_SESSION_TIMEOUT)  # Extend the life of the sandbox
-        try:
-            _LOGGER.info(
-                f"Start code execution in remote sandbox {self.interpreter.sandbox_id}. Timeout: {_SESSION_TIMEOUT}. Code hash: {hash(code)}"
-            )
-            execution = self.interpreter.notebook.exec_cell(
-                code=code,
-                on_stdout=lambda msg: _LOGGER.info(msg),
-                on_stderr=lambda msg: _LOGGER.info(msg),
-            )
-            _LOGGER.info(
-                f"Finished code execution in remote sandbox {self.interpreter.sandbox_id}. Code hash: {hash(code)}"
-            )
-            return Execution.from_e2b_execution(execution)
-        except (
-            LocalProtocolError,
-            HttpxRemoteProtocolError,
-            HttpcoreRemoteProtocolError,
-            ConnectError,
-            SandboxException,
-        ) as e:
-            raise e
-        except Exception as e:
-            raise RemoteSandboxExecutionError(
-                f"Failed executing code in remote sandbox ({self.interpreter.sandbox_id}) due to error '{type(e).__name__} {str(e)}', code: {code}"
-            ) from e
-
-    @tenacity.retry(
-        wait=tenacity.wait_exponential_jitter(),
-        stop=tenacity.stop_after_attempt(3),
-        retry=tenacity.retry_if_exception_type(
-            (
-                LocalProtocolError,
-                HttpxRemoteProtocolError,
-                HttpcoreRemoteProtocolError,
-                ConnectError,
-                SandboxException,
-            )
-        ),
-        before_sleep=tenacity.before_sleep_log(_LOGGER, logging.INFO),
-        after=tenacity.after_log(_LOGGER, logging.INFO),
-    )
-    def upload_file(self, file: Union[str, Path]) -> Path:
-        file_name = Path(file).name
-        with open(file, "rb") as f:
-            self.interpreter.files.write(path=str(self.remote_path / file_name), data=f)
-        _LOGGER.info(f"File ({file}) is uploaded to: {str(self.remote_path)}")
-        return self.remote_path / file_name
-
-    def download_file(
-        self, remote_file_path: Union[str, Path], local_file_path: Union[str, Path]
-    ) -> Path:
-        with open(local_file_path, "w+b") as f:
-            f.write(
-                self.interpreter.files.read(path=str(remote_file_path), format="bytes")
-            )
-        _LOGGER.info(f"File ({remote_file_path}) is downloaded to: {local_file_path}")
-        return Path(local_file_path)
-
-    @staticmethod
-    def _new_e2b_interpreter_impl(*args, **kwargs) -> E2BCodeInterpreterImpl:  # type: ignore
-        template_name = os.environ.get("E2B_TEMPLATE_NAME", "va-sandbox")
-        _LOGGER.info(
-            f"Creating a new E2BCodeInterpreter using template: {template_name}"
-        )
-        return E2BCodeInterpreterImpl(template=template_name, *args, **kwargs)
-
-
 class LocalCodeInterpreter(CodeInterpreter):
     def __init__(
         self,
@@ -723,12 +540,7 @@ class CodeInterpreterFactory:
     ) -> CodeInterpreter:
         if not code_sandbox_runtime:
             code_sandbox_runtime = os.getenv("CODE_SANDBOX_RUNTIME", "local")
-        if code_sandbox_runtime == "e2b":
-            envs = _get_e2b_env()
-            instance: CodeInterpreter = E2BCodeInterpreter(
-                timeout=_SESSION_TIMEOUT, remote_path=remote_path, envs=envs
-            )
-        elif code_sandbox_runtime == "local":
+        if code_sandbox_runtime == "local":
             instance = LocalCodeInterpreter(
                 timeout=_SESSION_TIMEOUT,
                 remote_path=remote_path,
@@ -736,25 +548,11 @@ class CodeInterpreterFactory:
             )
         else:
             raise ValueError(
-                f"Unsupported code sandbox runtime: {code_sandbox_runtime}. Supported runtimes: e2b, local"
+                f"Unsupported code sandbox runtime: {code_sandbox_runtime}. Supported runtimes: local"
             )
         return instance
 
 
-def _get_e2b_env() -> Union[Dict[str, str], None]:
-    openai_api_key = os.getenv("OPENAI_API_KEY", "")
-    anthropic_api_key = os.getenv("ANTHROPIC_API_KEY", "")
-    if openai_api_key or anthropic_api_key:
-        envs = {"WORKSPACE": os.getenv("WORKSPACE", "/home/user")}
-        if openai_api_key:
-            envs["OPENAI_API_KEY"] = openai_api_key
-        if anthropic_api_key:
-            envs["ANTHROPIC_API_KEY"] = anthropic_api_key
-    else:
-        envs = None
-    return envs
-
-
 def _parse_local_code_interpreter_outputs(outputs: List[Dict[str, Any]]) -> Execution:
     """Parse notebook cell outputs to Execution object. Output types:
     https://nbformat.readthedocs.io/en/latest/format_description.html#code-cell-outputs