vision-agent 0.2.118__py3-none-any.whl → 0.2.120__py3-none-any.whl

vision_agent/tools/tools.py

@@ -1,8 +1,9 @@
-import os
 import io
 import json
 import logging
+import os
 import tempfile
+import urllib.request
 from importlib import resources
 from pathlib import Path
 from typing import Any, Dict, List, Optional, Tuple, Union, cast
@@ -15,7 +16,6 @@ from moviepy.editor import ImageSequenceClip
 from PIL import Image, ImageDraw, ImageFont
 from pillow_heif import register_heif_opener  # type: ignore
 from pytube import YouTube  # type: ignore
-import urllib.request
 
 from vision_agent.clients.landing_public_api import LandingPublicAPI
 from vision_agent.tools.tool_utils import (
@@ -1332,7 +1332,7 @@ def save_video(
         video.write_videofile(f.name, codec="libx264")
         f.close()
         _save_video_to_result(f.name)
-        return f.name
+    return f.name
 
 
 def _save_video_to_result(video_uri: str) -> None:

vision_agent/tools/tools_types.py

@@ -1,8 +1,8 @@
-from uuid import UUID
 from enum import Enum
-from typing import List, Tuple, Optional
+from typing import List, Optional, Tuple
+from uuid import UUID
 
-from pydantic import BaseModel, ConfigDict, Field, field_serializer, SerializationInfo
+from pydantic import BaseModel, ConfigDict, Field, SerializationInfo, field_serializer
 
 
 class BboxInput(BaseModel):

vision_agent/utils/execute.py

@@ -5,7 +5,6 @@ import os
 import platform
 import re
 import sys
-import tempfile
 import traceback
 import warnings
 from enum import Enum
@@ -40,6 +39,7 @@ from vision_agent.utils.exceptions import (
 load_dotenv()
 _LOGGER = logging.getLogger(__name__)
 _SESSION_TIMEOUT = 600  # 10 minutes
+WORKSPACE = Path(os.getenv("WORKSPACE", ""))
 
 
 class MimeType(str, Enum):
@@ -384,8 +384,15 @@ class Execution(BaseModel):
 class CodeInterpreter(abc.ABC):
     """Code interpreter interface."""
 
-    def __init__(self, timeout: int, *args: Any, **kwargs: Any) -> None:
+    def __init__(
+        self,
+        timeout: int,
+        remote_path: Optional[Union[str, Path]] = None,
+        *args: Any,
+        **kwargs: Any,
+    ) -> None:
         self.timeout = timeout
+        self.remote_path = Path(remote_path if remote_path is not None else WORKSPACE)
 
     def __enter__(self) -> Self:
         return self
@@ -406,17 +413,21 @@ class CodeInterpreter(abc.ABC):
         self.restart_kernel()
         return self.exec_cell(code)
 
-    def upload_file(self, file: Union[str, Path]) -> str:
+    def upload_file(self, file: Union[str, Path]) -> Path:
         # Default behavior is a no-op (for local code interpreter)
-        return str(file)
+        return Path(file)
 
-    def download_file(self, file_path: str) -> Path:
+    def download_file(
+        self, remote_file_path: Union[str, Path], local_file_path: Union[str, Path]
+    ) -> Path:
         # Default behavior is a no-op (for local code interpreter)
-        return Path(file_path)
+        return Path(local_file_path)
 
 
 class E2BCodeInterpreter(CodeInterpreter):
-    def __init__(self, *args: Any, **kwargs: Any) -> None:
+    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:
@@ -443,6 +454,9 @@ print(f"Vision Agent version: {va_version}")"""
         _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:
@@ -516,19 +530,22 @@ print(f"Vision Agent version: {va_version}")"""
         before_sleep=tenacity.before_sleep_log(_LOGGER, logging.INFO),
         after=tenacity.after_log(_LOGGER, logging.INFO),
     )
-    def upload_file(self, file: Union[str, Path]) -> str:
+    def upload_file(self, file: Union[str, Path]) -> Path:
         file_name = Path(file).name
-        remote_path = f"/home/user/{file_name}"
         with open(file, "rb") as f:
-            self.interpreter.files.write(path=remote_path, data=f)
-            _LOGGER.info(f"File ({file}) is uploaded to: {remote_path}")
-            return remote_path
-
-    def download_file(self, file_path: str) -> Path:
-        with tempfile.NamedTemporaryFile(mode="w+b", delete=False) as file:
-            file.write(self.interpreter.files.read(path=file_path, format="bytes"))
-            _LOGGER.info(f"File ({file_path}) is downloaded to: {file.name}")
-            return Path(file.name)
+            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
@@ -540,7 +557,11 @@ print(f"Vision Agent version: {va_version}")"""
 
 
 class LocalCodeInterpreter(CodeInterpreter):
-    def __init__(self, timeout: int = _SESSION_TIMEOUT) -> None:
+    def __init__(
+        self,
+        timeout: int = _SESSION_TIMEOUT,
+        remote_path: Optional[Union[str, Path]] = None,
+    ) -> None:
         super().__init__(timeout=timeout)
         self.nb = nbformat.v4.new_notebook()
         self.nb_client = NotebookClient(self.nb, timeout=self.timeout)
@@ -554,6 +575,7 @@ Timeout: {self.timeout}"""
         )
         sleep(1)
         self._new_kernel()
+        self.remote_path = Path(remote_path if remote_path is not None else WORKSPACE)
 
     def _new_kernel(self) -> None:
         if self.nb_client.kc is None or not run_sync(self.nb_client.kc.is_alive)():  # type: ignore
@@ -607,6 +629,25 @@ Timeout: {self.timeout}"""
             traceback_raw = traceback.format_exc().splitlines()
             return Execution.from_exception(e, traceback_raw)
 
+    def upload_file(self, file_path: Union[str, Path]) -> Path:
+        with open(file_path, "rb") as f:
+            contents = f.read()
+        with open(self.remote_path / Path(file_path).name, "wb") as f:
+            f.write(contents)
+        _LOGGER.info(f"File ({file_path}) is uploaded to: {str(self.remote_path)}")
+
+        return Path(self.remote_path / file_path)
+
+    def download_file(
+        self, remote_file_path: Union[str, Path], local_file_path: Union[str, Path]
+    ) -> Path:
+        with open(self.remote_path / remote_file_path, "rb") as f:
+            contents = f.read()
+        with open(local_file_path, "wb") as f:
+            f.write(contents)
+        _LOGGER.info(f"File ({remote_file_path}) is downloaded to: {local_file_path}")
+        return Path(local_file_path)
+
 
 class CodeInterpreterFactory:
     """Factory class for creating code interpreters.
@@ -630,13 +671,19 @@ class CodeInterpreterFactory:
         return instance
 
     @staticmethod
-    def new_instance(code_sandbox_runtime: Optional[str] = None) -> CodeInterpreter:
+    def new_instance(
+        code_sandbox_runtime: Optional[str] = None, remote_path: Optional[str] = None
+    ) -> CodeInterpreter:
         if not code_sandbox_runtime:
             code_sandbox_runtime = os.getenv("CODE_SANDBOX_RUNTIME", "local")
         if code_sandbox_runtime == "e2b":
-            instance: CodeInterpreter = E2BCodeInterpreter(timeout=_SESSION_TIMEOUT)
+            instance: CodeInterpreter = E2BCodeInterpreter(
+                timeout=_SESSION_TIMEOUT, remote_path=remote_path
+            )
         elif code_sandbox_runtime == "local":
-            instance = LocalCodeInterpreter(timeout=_SESSION_TIMEOUT)
+            instance = LocalCodeInterpreter(
+                timeout=_SESSION_TIMEOUT, remote_path=remote_path
+            )
         else:
             raise ValueError(
                 f"Unsupported code sandbox runtime: {code_sandbox_runtime}. Supported runtimes: e2b, local"

vision_agent/utils/image_utils.py

@@ -70,7 +70,7 @@ def rle_decode_array(rle: Dict[str, List[int]]) -> np.ndarray:
     r"""Decode a run-length encoded mask. Returns numpy array, 1 - mask, 0 - background.
 
     Parameters:
-        mask: The mask in run-length encoded as an array.
+        rle: The run-length encoded mask.
     """
     size = rle["size"]
     counts = rle["counts"]
@@ -100,7 +100,7 @@ def frames_to_bytes(
     """
     with tempfile.NamedTemporaryFile(delete=True) as temp_file:
         clip = ImageSequenceClip(frames, fps=fps)
-        clip.write_videofile(temp_file.name + f".{file_ext}", fps=fps)
+        clip.write_videofile(temp_file.name + f".{file_ext}", fps=fps, codec="libx264")
         with open(temp_file.name + f".{file_ext}", "rb") as f:
             buffer_bytes = f.read()
     return buffer_bytes

{vision_agent-0.2.118.dist-info → vision_agent-0.2.120.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vision-agent
-Version: 0.2.118
+Version: 0.2.120
 Summary: Toolset for Vision Agent
 Author: Landing AI
 Author-email: dev@landing.ai
@@ -81,15 +81,15 @@ export OPENAI_API_KEY="your-api-key"
 ```
 
 ### Vision Agent
-There are two agents that you can use. Vision Agent is a conversational agent that has
+There are two agents that you can use. `VisionAgent` is a conversational agent that has
 access to tools that allow it to write an navigate python code and file systems. It can
-converse with the user in natural language. VisionAgentCoder is an agent that can write
-code for vision tasks, such as counting people in an image. However, it cannot converse
-and can only respond with code. VisionAgent can call VisionAgentCoder to write vision
-code.
+converse with the user in natural language. `VisionAgentCoder` is an agent specifically
+for writing code for vision tasks, such as counting people in an image. However, it
+cannot chat with you and can only respond with code. `VisionAgent` can call
+`VisionAgentCoder` to write vision code.
 
 #### Basic Usage
-To run the streamlit app locally to chat with Vision Agent, you can run the following
+To run the streamlit app locally to chat with `VisionAgent`, you can run the following
 command:
 
 ```bash
@@ -186,7 +186,7 @@ the code and having it update. You just need to add the code as a response from
 assistant:
 
 ```python
-agent = va.agent.VisionAgent(verbosity=2)
+agent = va.agent.VisionAgentCoder(verbosity=2)
 conv = [
     {
         "role": "user",
@@ -252,6 +252,10 @@ function. Make sure the documentation is in the same format above with descripti
 `Parameters:`, `Returns:`, and `Example\n-------`. You can find an example use case
 [here](examples/custom_tools/) as this is what the agent uses to pick and use the tool.
 
+Can't find the tool you need and want add it to `VisionAgent`? Check out our
+[vision-agent-tools](https://github.com/landing-ai/vision-agent-tools) repository where
+we add the source code for all the tools used in `VisionAgent`.
+
 ## Additional Backends
 ### Ollama
 We also provide a `VisionAgentCoder` that uses Ollama. To get started you must download

{vision_agent-0.2.118.dist-info → vision_agent-0.2.120.dist-info}/RECORD

@@ -1,33 +1,33 @@
 vision_agent/__init__.py,sha256=EAb4-f9iyuEYkBrX4ag1syM8Syx8118_t0R6_C34M9w,57
 vision_agent/agent/__init__.py,sha256=FRwiux1FGvGccetyUCtY46KP01fQteqorm-JtFepovI,176
-vision_agent/agent/agent.py,sha256=Bt8yhjCFXuRdZaHxKEesG40V09nWRt45sZluri1R3AA,575
+vision_agent/agent/agent.py,sha256=2cjIOxEuSJrqbfPXYoV0qER5ihXsPFCoEFJa4jpqan0,597
 vision_agent/agent/agent_utils.py,sha256=22LiPhkJlS5mVeo2dIi259pc2NgA7PGHRpcbnrtKo78,1930
-vision_agent/agent/vision_agent.py,sha256=5rgO-pScVOS3t4sWnLBnGYYkGftGgF4U0FpZzFVrDAY,8447
-vision_agent/agent/vision_agent_coder.py,sha256=tE-15ttnDxUsEdB0XJP4AVNyOU89KS8ZvXZDPcNKA-8,34380
+vision_agent/agent/vision_agent.py,sha256=IEyXT_JPCuWmBHdEnM1Wrsj7hmCe5pKLf0gnZFJTddI,11046
+vision_agent/agent/vision_agent_coder.py,sha256=DOTmDdGPxcI06Jp6yx4ekRMP0vhiVaK9B9Dl8UyJHeo,34396
 vision_agent/agent/vision_agent_coder_prompts.py,sha256=xIya1txRZM8qoQHAWTEkEFCL8L3iZD7QD09t3ZtdxSE,11305
-vision_agent/agent/vision_agent_prompts.py,sha256=ydUU_Wvw-jqdL_vObSUr-VCQvjSwA5Fd74TbbhUzyxk,6112
+vision_agent/agent/vision_agent_prompts.py,sha256=0GliXFtBf32aPu2ClU63FI5ii5CTxWYsvrsmnnDp-gs,7134
 vision_agent/clients/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 vision_agent/clients/http.py,sha256=k883i6M_4nl7zwwHSI-yP5sAgQZIDPM1nrKD6YFJ3Xs,2009
-vision_agent/clients/landing_public_api.py,sha256=6L15zh5lP5JHCpGnYpHMREgrrKiJin_OYdf2vT9HHZQ,1507
+vision_agent/clients/landing_public_api.py,sha256=rGtACkr8o5egDuMHQ5MBO4NuvsgPTp9Ew3rbq4R-vs0,1507
 vision_agent/fonts/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 vision_agent/fonts/default_font_ch_en.ttf,sha256=1YM0Z3XqLDjSNbF7ihQFSAIUdjF9m1rtHiNC_6QosTE,1594400
 vision_agent/lmm/__init__.py,sha256=YuUZRsMHdn8cMOv6iBU8yUqlIOLrbZQqZl9KPnofsHQ,103
-vision_agent/lmm/lmm.py,sha256=xkAxunToISzo5rCcjekqQBvm5SRW-98htieLuztKNbk,20802
-vision_agent/lmm/types.py,sha256=8TSRoTbXyCKVJiH-wHXI2OiGOMSkYv1vLGYeAXtNpOQ,153
-vision_agent/tools/__init__.py,sha256=lUUc2HV13eSxg5KPZop1D-mB4ecmiQ5fYlBTQLNSbYg,2190
-vision_agent/tools/meta_tools.py,sha256=q6h7hZarZrsWRloVE6PbTZwW8J2N1uUM9Ac-XxsT6hk,13365
+vision_agent/lmm/lmm.py,sha256=AYrZNdhghG293wd3aKZ1jK1lUm2NLWwALktbM4wNais,20862
+vision_agent/lmm/types.py,sha256=ZEXR_ptBL0ZwDMTDYkgxUCmSZFmBYPQd2jreNzr_8UY,221
+vision_agent/tools/__init__.py,sha256=i7JOLxRaLdcY7-vCNOGAeOFMBfiAUIwWhnT32FO97VE,2201
+vision_agent/tools/meta_tools.py,sha256=Vu9WnKicGhafx9dPzDbQjQdcIzRCYYFPF68o79hDP-8,14616
 vision_agent/tools/prompts.py,sha256=V1z4YJLXZuUl_iZ5rY0M5hHc_2tmMEUKr0WocXKGt4E,1430
 vision_agent/tools/tool_utils.py,sha256=qMsb9d8QtpXGgF9rpPO2dA390BewKdYO68oWKDu-TGg,6504
-vision_agent/tools/tools.py,sha256=gAW6G9k1vzy8jwRACNnw2Vihsajm_oSlVJqd6E4JSRA,59957
-vision_agent/tools/tools_types.py,sha256=z6_XtUhWgh201yM7Z0CYtiLBEGdHPc_QUydMDHZ84EA,2216
+vision_agent/tools/tools.py,sha256=kbbMToAaHxl42dDEvyz9Mvtpqts0l0hGoC5YQQyozr8,59953
+vision_agent/tools/tools_types.py,sha256=iLWSirheC87fKQolIhx_O4Jk8Lv7DRiLuE8PJqLGiVQ,2216
 vision_agent/utils/__init__.py,sha256=pWk0ktvR4aUEhuEIzSLM9kSgW4WDVqptdvOTeGLkJ6M,230
 vision_agent/utils/exceptions.py,sha256=booSPSuoULF7OXRr_YbC4dtKt6gM_HyiFQHBuaW86C4,2052
-vision_agent/utils/execute.py,sha256=1_pyu16WRlFD81W8Uy_Sv7_jD_qkrkxHdUNeFstBzaA,25082
-vision_agent/utils/image_utils.py,sha256=c1LrmaHD331za8DbA1myJpgUmWoDzePaOK6-dsdpZQo,9847
+vision_agent/utils/execute.py,sha256=Ap8Yx80spQq5f2QtKGx1MK03BR45mJKhlp1kfh-rIao,26751
+vision_agent/utils/image_utils.py,sha256=eNghu_2L8624jEXy8ZZS9OX46Mv0DT9bcvLForujwTs,9848
 vision_agent/utils/sim.py,sha256=ebE9Cs00pVEDI1HMjAzUBk88tQQmc2U-yAzIDinnekU,5572
 vision_agent/utils/type_defs.py,sha256=BE12s3JNQy36QvauXHjwyeffVh5enfcvd4vTzSwvEZI,1384
 vision_agent/utils/video.py,sha256=rNmU9KEIkZB5-EztZNlUiKYN0mm_55A_2VGUM0QpqLA,8779
-vision_agent-0.2.118.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-vision_agent-0.2.118.dist-info/METADATA,sha256=4ilO7j9MOLCtaNekUUVlhMNdDKMk02ecx7ipnXT9RC8,11997
-vision_agent-0.2.118.dist-info/WHEEL,sha256=7Z8_27uaHI_UZAc4Uox4PpBhQ9Y5_modZXWMxtUi4NU,88
-vision_agent-0.2.118.dist-info/RECORD,,
+vision_agent-0.2.120.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+vision_agent-0.2.120.dist-info/METADATA,sha256=-FuNdlrzt5cTK6Ou_HTTROGVvsIwP3trsB5Edt2St3o,12255
+vision_agent-0.2.120.dist-info/WHEEL,sha256=7Z8_27uaHI_UZAc4Uox4PpBhQ9Y5_modZXWMxtUi4NU,88
+vision_agent-0.2.120.dist-info/RECORD,,