vision-agent 0.2.4__py3-none-any.whl → 0.2.5__py3-none-any.whl

vision_agent/agent/vision_agent.py

@@ -366,6 +366,20 @@ def _handle_viz_tools(
     return image_to_data
 
 
+def sample_n_evenly_spaced(lst: Sequence, n: int) -> Sequence:
+    if n <= 0:
+        return []
+    elif len(lst) == 0:
+        return []
+    elif n == 1:
+        return [lst[0]]
+    elif n >= len(lst):
+        return lst
+
+    spacing = (len(lst) - 1) / (n - 1)
+    return [lst[round(spacing * i)] for i in range(n)]
+
+
 def visualize_result(all_tool_results: List[Dict]) -> Sequence[Union[str, Path]]:
     image_to_data: Dict[str, Dict] = {}
     for tool_result in all_tool_results:
@@ -584,7 +598,7 @@ class VisionAgent(Agent):
             visualized_output = visualize_result(all_tool_results)
             all_tool_results.append({"visualized_output": visualized_output})
             if len(visualized_output) > 0:
-                reflection_images = visualized_output
+                reflection_images = sample_n_evenly_spaced(visualized_output, 3)
             elif image is not None:
                 reflection_images = [image]
             else:

vision_agent/lmm/lmm.py

@@ -9,10 +9,7 @@ from typing import Any, Callable, Dict, List, Optional, Union, cast
 import requests
 from openai import AzureOpenAI, OpenAI
 
-from vision_agent.tools import (
-    CHOOSE_PARAMS,
-    SYSTEM_PROMPT,
-)
+from vision_agent.tools import CHOOSE_PARAMS, SYSTEM_PROMPT
 
 _LOGGER = logging.getLogger(__name__)
 

vision_agent/tools/__init__.py

@@ -12,12 +12,12 @@ from .tools import (  # Counter,
     GroundingDINO,
     GroundingSAM,
     ImageCaption,
-    ZeroShotCounting,
-    VisualPromptCounting,
-    VisualQuestionAnswering,
     ImageQuestionAnswering,
     SegArea,
     SegIoU,
     Tool,
+    VisualPromptCounting,
+    VisualQuestionAnswering,
+    ZeroShotCounting,
     register_tool,
 )

vision_agent/tools/tools.py

@@ -17,9 +17,9 @@ from vision_agent.image_utils import (
     normalize_bbox,
     rle_decode,
 )
+from vision_agent.lmm import OpenAILMM
 from vision_agent.tools.video import extract_frames_from_video
 from vision_agent.type_defs import LandingaiAPIKey
-from vision_agent.lmm import OpenAILMM
 
 _LOGGER = logging.getLogger(__name__)
 _LND_API_KEY = LandingaiAPIKey().api_key

{vision_agent-0.2.4.dist-info → vision_agent-0.2.5.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vision-agent
-Version: 0.2.4
+Version: 0.2.5
 Summary: Toolset for Vision Agent
 Author: Landing AI
 Author-email: dev@landing.ai

{vision_agent-0.2.4.dist-info → vision_agent-0.2.5.dist-info}/RECORD

@@ -5,7 +5,7 @@ vision_agent/agent/easytool.py,sha256=oMHnBg7YBtIPgqQUNcZgq7uMgpPThs99_UnO7ERkMV
 vision_agent/agent/easytool_prompts.py,sha256=zdQQw6WpXOmvwOMtlBlNKY5a3WNlr65dbUvMIGiqdeo,4526
 vision_agent/agent/reflexion.py,sha256=4gz30BuFMeGxSsTzoDV4p91yE0R8LISXp28IaOI6wdM,10506
 vision_agent/agent/reflexion_prompts.py,sha256=G7UAeNz_g2qCb2yN6OaIC7bQVUkda4m3z42EG8wAyfE,9342
-vision_agent/agent/vision_agent.py,sha256=Ehb97lyPs7lYM9ipx07yxm6c2kUqz2OnjGQsv-nMwKA,24849
+vision_agent/agent/vision_agent.py,sha256=SFdw6OBqWj0cr-YthFMM_x-Urg86CggazYQG4wy0n-U,25195
 vision_agent/agent/vision_agent_prompts.py,sha256=W3Z72FpUt71UIJSkjAcgtQqxeMqkYuATqHAN5fYY26c,7342
 vision_agent/fonts/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 vision_agent/fonts/default_font_ch_en.ttf,sha256=1YM0Z3XqLDjSNbF7ihQFSAIUdjF9m1rtHiNC_6QosTE,1594400
@@ -13,13 +13,13 @@ vision_agent/image_utils.py,sha256=YvP5KE9NrWdgJKuHW2NR1glzfObkxtcXBknpmj3Gsbs,7
 vision_agent/llm/__init__.py,sha256=BoUm_zSAKnLlE8s-gKTSQugXDqVZKPqYlWwlTLdhcz4,48
 vision_agent/llm/llm.py,sha256=1BkrSVBWEClyqLc0Rmyw4heLhi_ZVm6JO7-i1wd1ziw,5383
 vision_agent/lmm/__init__.py,sha256=nnNeKD1k7q_4vLb1x51O_EUTYaBgGfeiCx5F433gr3M,67
-vision_agent/lmm/lmm.py,sha256=sECjGMaGrv1QHq7OiFr-9LoBM5uRLjAqd0Ypp-zyFlw,10552
-vision_agent/tools/__init__.py,sha256=X6yJhWa8iKkQm4Mgf1KcV0_o39-Nrg3E56QAB5gWCO0,413
+vision_agent/lmm/lmm.py,sha256=gK90vMxh0OcGSuIZQikBkDXm4pfkdFk1R2y7rtWDl84,10539
+vision_agent/tools/__init__.py,sha256=HfUr0JQUwk0Kyieen93df9lMbbdpVf9Q6CcVFmKv_q4,413
 vision_agent/tools/prompts.py,sha256=V1z4YJLXZuUl_iZ5rY0M5hHc_2tmMEUKr0WocXKGt4E,1430
-vision_agent/tools/tools.py,sha256=hYgRTHMCBwjT0kkT2SY5MN0FK89vuuecu-x1VqRlGbU,42779
+vision_agent/tools/tools.py,sha256=GvRDLeMVS9C7z56hlSpThGoV0r_x5pKSFw-g4JW_qnw,42779
 vision_agent/tools/video.py,sha256=xTElFSFp1Jw4ulOMnk81Vxsh-9dTxcWUO6P9fzEi3AM,7653
 vision_agent/type_defs.py,sha256=4LTnTL4HNsfYqCrDn9Ppjg9bSG2ZGcoKSSd9YeQf4Bw,1792
-vision_agent-0.2.4.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
-vision_agent-0.2.4.dist-info/METADATA,sha256=2T1YLGMh2-n8F0gGf1P2BDhgzxmtmAiylpfW3E3Q4_c,7697
-vision_agent-0.2.4.dist-info/WHEEL,sha256=7Z8_27uaHI_UZAc4Uox4PpBhQ9Y5_modZXWMxtUi4NU,88
-vision_agent-0.2.4.dist-info/RECORD,,
+vision_agent-0.2.5.dist-info/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+vision_agent-0.2.5.dist-info/METADATA,sha256=zSTYpM893hERFpO2j7-YdRmRPKeGI6-qU_wkq5MitFY,7697
+vision_agent-0.2.5.dist-info/WHEEL,sha256=7Z8_27uaHI_UZAc4Uox4PpBhQ9Y5_modZXWMxtUi4NU,88
+vision_agent-0.2.5.dist-info/RECORD,,