abstractswarm

swarms/agents/__init__.py

@@ -11,3 +11,4 @@ from swarms.agents.omni_modal_agent import OmniModalAgent
 from swarms.agents.message import Message
 from swarms.agents.stream_response import stream
 from swarms.agents.base import AbstractAgent
+from swarms.agents.registry import Registry

swarms/agents/conversabe_agent.py

@@ -977,7 +977,7 @@ class ConversableAgent(Agent):
                 )
             elif lang in ["python", "Python"]:
                 if code.startswith("# filename: "):
-                    filename = code[11 : code.find("\n")].strip()
+                    filename = code[11: code.find("\n")].strip()
                 else:
                     filename = None
                 exitcode, logs, image = self.run_code(

swarms/agents/meta_prompter.py

@@ -108,7 +108,7 @@ class MetaPrompterAgent:
     def get_new_instructions(self, meta_output):
         """Get New Instructions from the meta_output"""
         delimiter = "Instructions: "
-        new_instructions = meta_output[meta_output.find(delimiter) + len(delimiter) :]
+        new_instructions = meta_output[meta_output.find(delimiter) + len(delimiter):]
         return new_instructions
 
     def run(self, task: str):

swarms/agents/models/groundingdino/datasets/transforms.py

@@ -38,7 +38,7 @@ def crop(image, target, region):
 
     if "masks" in target:
         # FIXME should we update the area here if there are no boxes?
-        target["masks"] = target["masks"][:, i : i + h, j : j + w]
+        target["masks"] = target["masks"][:, i: i + h, j: j + w]
         fields.append("masks")
 
     # remove elements for which the boxes or masks that have zero area

swarms/agents/models/groundingdino/models/GroundingDINO/backbone/backbone.py

@@ -159,7 +159,7 @@ class Backbone(BackboneBase):
         ), "Only resnet50 and resnet101 are available."
         assert return_interm_indices in [[0, 1, 2, 3], [1, 2, 3], [3]]
         num_channels_all = [256, 512, 1024, 2048]
-        num_channels = num_channels_all[4 - len(return_interm_indices) :]
+        num_channels = num_channels_all[4 - len(return_interm_indices):]
         super().__init__(backbone, train_backbone, num_channels, return_interm_indices)
 
 
@@ -224,7 +224,7 @@ def build_backbone(args):
             use_checkpoint=use_checkpoint,
         )
 
-        bb_num_channels = backbone.num_features[4 - len(return_interm_indices) :]
+        bb_num_channels = backbone.num_features[4 - len(return_interm_indices):]
     else:
         raise NotImplementedError("Unknown backbone {}".format(args.backbone))
 

swarms/agents/models/groundingdino/models/GroundingDINO/backbone/swin_transformer.py

@@ -649,7 +649,7 @@ class SwinTransformer(nn.Module):
                 qk_scale=qk_scale,
                 drop=drop_rate,
                 attn_drop=attn_drop_rate,
-                drop_path=dpr[sum(depths[:i_layer]) : sum(depths[: i_layer + 1])],
+                drop_path=dpr[sum(depths[:i_layer]): sum(depths[: i_layer + 1])],
                 norm_layer=norm_layer,
                 # downsample=PatchMerging if (i_layer < self.num_layers - 1) else None,
                 downsample=downsamplelist[i_layer],

swarms/agents/models/groundingdino/models/GroundingDINO/bertwarper.py

@@ -221,9 +221,9 @@ def generate_masks_with_special_tokens(tokenized, special_tokens_list, tokenizer
             position_ids[row, col] = 0
         else:
             attention_mask[
-                row, previous_col + 1 : col + 1, previous_col + 1 : col + 1
+                row, previous_col + 1: col + 1, previous_col + 1: col + 1
             ] = True
-            position_ids[row, previous_col + 1 : col + 1] = torch.arange(
+            position_ids[row, previous_col + 1: col + 1] = torch.arange(
                 0, col - previous_col, device=input_ids.device
             )
 
@@ -273,13 +273,13 @@ def generate_masks_with_special_tokens_and_transfer_map(
             position_ids[row, col] = 0
         else:
             attention_mask[
-                row, previous_col + 1 : col + 1, previous_col + 1 : col + 1
+                row, previous_col + 1: col + 1, previous_col + 1: col + 1
             ] = True
-            position_ids[row, previous_col + 1 : col + 1] = torch.arange(
+            position_ids[row, previous_col + 1: col + 1] = torch.arange(
                 0, col - previous_col, device=input_ids.device
             )
             c2t_maski = torch.zeros((num_token), device=input_ids.device).bool()
-            c2t_maski[previous_col + 1 : col] = True
+            c2t_maski[previous_col + 1: col] = True
             cate_to_token_mask_list[row].append(c2t_maski)
         previous_col = col
 

swarms/agents/models/groundingdino/models/GroundingDINO/utils.py

@@ -76,7 +76,7 @@ def gen_encoder_output_proposals(
     proposals = []
     _cur = 0
     for lvl, (H_, W_) in enumerate(spatial_shapes):
-        mask_flatten_ = memory_padding_mask[:, _cur : (_cur + H_ * W_)].view(
+        mask_flatten_ = memory_padding_mask[:, _cur: (_cur + H_ * W_)].view(
             N_, H_, W_, 1
         )
         valid_H = torch.sum(~mask_flatten_[:, :, 0, 0], 1)

swarms/agents/models/groundingdino/util/utils.py

@@ -619,7 +619,7 @@ def get_phrases_from_posmap(
 ):
     assert isinstance(posmap, torch.Tensor), "posmap must be torch.Tensor"
     if posmap.dim() == 1:
-        posmap[0 : left_idx + 1] = False
+        posmap[0: left_idx + 1] = False
         posmap[right_idx:] = False
         non_zero_idx = posmap.nonzero(as_tuple=True)[0].tolist()
         token_ids = [tokenized["input_ids"][i] for i in non_zero_idx]

swarms/agents/models/groundingdino/util/vl_utils.py

@@ -41,7 +41,7 @@ def create_positive_map_from_span(tokenized, token_span, max_text_len=256):
                 positive_map[j, beg_pos] = 1
                 break
             else:
-                positive_map[j, beg_pos : end_pos + 1].fill_(1)
+                positive_map[j, beg_pos: end_pos + 1].fill_(1)
 
     return positive_map / (positive_map.sum(-1)[:, None] + 1e-6)
 

swarms/agents/models/segment_anything/segment_anything/modeling/mask_decoder.py

@@ -139,7 +139,7 @@ class MaskDecoder(nn.Module):
         # Run the transformer
         hs, src = self.transformer(src, pos_src, tokens)
         iou_token_out = hs[:, 0, :]
-        mask_tokens_out = hs[:, 1 : (1 + self.num_mask_tokens), :]
+        mask_tokens_out = hs[:, 1: (1 + self.num_mask_tokens), :]
 
         # Upscale mask embeddings and predict masks using the mask tokens
         src = src.transpose(1, 2).view(b, c, h, w)

swarms/agents/models/segment_anything/segment_anything/utils/amg.py

@@ -101,7 +101,7 @@ def batch_iterator(batch_size: int, *args) -> Generator[List[Any], None, None]:
     ), "Batched iteration must have inputs of all the same size."
     n_batches = len(args[0]) // batch_size + int(len(args[0]) % batch_size != 0)
     for b in range(n_batches):
-        yield [arg[b * batch_size : (b + 1) * batch_size] for arg in args]
+        yield [arg[b * batch_size: (b + 1) * batch_size] for arg in args]
 
 
 def mask_to_rle_pytorch(tensor: torch.Tensor) -> List[Dict[str, Any]]:
@@ -142,7 +142,7 @@ def rle_to_mask(rle: Dict[str, Any]) -> np.ndarray:
     idx = 0
     parity = False
     for count in rle["counts"]:
-        mask[idx : idx + count] = parity
+        mask[idx: idx + count] = parity
         idx += count
         parity ^= True
     mask = mask.reshape(w, h)

swarms/agents/multi_modal_visual_agent.py

@@ -207,12 +207,12 @@ def blend_gt2pt(old_image, new_image, sigma=0.15, steps=100):
     kernel[steps:-steps, :steps] = left
     kernel[steps:-steps, -steps:] = right
 
-    pt_gt_img = easy_img[pos_h : pos_h + old_size[1], pos_w : pos_w + old_size[0]]
+    pt_gt_img = easy_img[pos_h: pos_h + old_size[1], pos_w: pos_w + old_size[0]]
     gaussian_gt_img = (
         kernel * gt_img_array + (1 - kernel) * pt_gt_img
     )  # gt img with blur img
     gaussian_gt_img = gaussian_gt_img.astype(np.int64)
-    easy_img[pos_h : pos_h + old_size[1], pos_w : pos_w + old_size[0]] = gaussian_gt_img
+    easy_img[pos_h: pos_h + old_size[1], pos_w: pos_w + old_size[0]] = gaussian_gt_img
     gaussian_img = Image.fromarray(easy_img)
     return gaussian_img
 

swarms/agents/multi_modal_workers/omni_agent/omni_chat.py

@@ -317,7 +317,7 @@ def find_json(s):
     s = s.replace("'", '"')
     start = s.find("{")
     end = s.rfind("}")
-    res = s[start : end + 1]
+    res = s[start: end + 1]
     res = res.replace("\n", "")
     return res
 

swarms/agents/registry.py

@@ -0,0 +1,27 @@
+from typing import Dict
+
+from pydantic import BaseModel
+
+
+class Registry(BaseModel):
+    """Registry for storing and building classes."""
+
+    name: str
+    entries: Dict = {}
+
+    def register(self, key: str):
+        def decorator(class_builder):
+            self.entries[key] = class_builder
+            return class_builder
+
+        return decorator
+
+    def build(self, type: str, **kwargs):
+        if type not in self.entries:
+            raise ValueError(
+                f'{type} is not registered. Please register with the .register("{type}") method provided in {self.name} registry'
+            )
+        return self.entries[type](**kwargs)
+
+    def get_all_entries(self):
+        return self.entries

swarms/embeddings/openai.py

@@ -347,7 +347,7 @@ class OpenAIEmbeddings(BaseModel, Embeddings):
                 disallowed_special=self.disallowed_special,
             )
             for j in range(0, len(token), self.embedding_ctx_length):
-                tokens.append(token[j : j + self.embedding_ctx_length])
+                tokens.append(token[j: j + self.embedding_ctx_length])
                 indices.append(i)
 
         batched_embeddings: List[List[float]] = []
@@ -366,7 +366,7 @@ class OpenAIEmbeddings(BaseModel, Embeddings):
         for i in _iter:
             response = embed_with_retry(
                 self,
-                input=tokens[i : i + _chunk_size],
+                input=tokens[i: i + _chunk_size],
                 **self._invocation_params,
             )
             batched_embeddings.extend(r["embedding"] for r in response["data"])
@@ -428,7 +428,7 @@ class OpenAIEmbeddings(BaseModel, Embeddings):
                 disallowed_special=self.disallowed_special,
             )
             for j in range(0, len(token), self.embedding_ctx_length):
-                tokens.append(token[j : j + self.embedding_ctx_length])
+                tokens.append(token[j: j + self.embedding_ctx_length])
                 indices.append(i)
 
         batched_embeddings: List[List[float]] = []
@@ -436,7 +436,7 @@ class OpenAIEmbeddings(BaseModel, Embeddings):
         for i in range(0, len(tokens), _chunk_size):
             response = await async_embed_with_retry(
                 self,
-                input=tokens[i : i + _chunk_size],
+                input=tokens[i: i + _chunk_size],
                 **self._invocation_params,
             )
             batched_embeddings.extend(r["embedding"] for r in response["data"])

swarms/models/chat_openai.py

@@ -458,7 +458,7 @@ class BaseOpenAI(BaseLLM):
                 )
             params["max_tokens"] = self.max_tokens_for_prompt(prompts[0])
         sub_prompts = [
-            prompts[i : i + self.batch_size]
+            prompts[i: i + self.batch_size]
             for i in range(0, len(prompts), self.batch_size)
         ]
         return sub_prompts
@@ -469,7 +469,7 @@ class BaseOpenAI(BaseLLM):
         """Create the LLMResult from the choices and prompts."""
         generations = []
         for i, _ in enumerate(prompts):
-            sub_choices = choices[i * self.n : (i + 1) * self.n]
+            sub_choices = choices[i * self.n: (i + 1) * self.n]
             generations.append(
                 [
                     Generation(

swarms/prompts/python.py

@@ -232,4 +232,3 @@ PY_TEST_GENERATION_COMPLETION_INSTRUCTION = f"""You are an AI coding assistant t
 {PY_TEST_GENERATION_FEW_SHOT}"""
 
 PY_TEST_GENERATION_CHAT_INSTRUCTION = """You are an AI coding assistant that can write unique, diverse, and intuitive unit tests for functions given the signature and docstring."""
-

swarms/structs/nonlinear_workflow.py

@@ -7,6 +7,7 @@ class Task:
     """
     Task is a unit of work that can be executed by an agent
     """
+
     def __init__(
         self, id: str, parents: List["Task"] = None, children: List["Task"] = None
     ):

swarms/swarms/base.py

@@ -1,5 +1,6 @@
 from abc import ABC, abstractmethod
-
+from typing import Optional, List
+from swarms.workers.base import AbstractWorker
 
 class AbstractSwarm(ABC):
     # TODO: Pass in abstract LLM class that can utilize Hf or Anthropic models, Move away from OPENAI
@@ -8,15 +9,52 @@ class AbstractSwarm(ABC):
     # TODO: Add RLHF Data collection, ask user how the swarm is performing
     # TODO: Create an onboarding process if not settings are preconfigured like `from swarms import Swarm, Swarm()` => then initiate onboarding name your swarm + provide purpose + etc
 
-    def __init__(self, agents, vectorstore, tools):
-        self.agents = agents
-        self.vectorstore = vectorstore
-        self.tools = tools
+    @abstractmethod
+    def __init__(self, workers: List["AbstractWorker"]):
+        """Initialize the swarm with workers"""
+        pass
 
     @abstractmethod
     def communicate(self):
+        """Communicate with the swarm through the orchestrator, protocols, and the universal communication layer"""
         pass
 
     @abstractmethod
     def run(self):
+        """Run the swarm"""
+        pass
+
+    @abstractmethod
+    def add_worker(self, worker: "AbstractWorker"):
+        """Add a worker to the swarm"""
+        pass
+
+    @abstractmethod
+    def remove_worker(self, worker: "AbstractWorker"):
+        """Remove a worker from the swarm"""
+        pass
+
+    @abstractmethod
+    def broadcast(self, message: str, sender: Optional["AbstractWorker"] = None):
+        """Broadcast a message to all workers"""
+        pass
+
+    @abstractmethod
+    def reset(self):
+        """Reset the swarm"""
+        pass
+
+    @abstractmethod
+    def plan(self, task: str):
+        """Workers must individually plan using a workflow or pipeline"""
+        pass
+
+    @abstractmethod
+    def direct_message(
+        self,
+        message: str,
+        sender: "AbstractWorker",
+        recipient: "AbstractWorker",
+    ):
+        """Send a direct message to a worker"""
         pass

swarms/tools/autogpt.py

@@ -129,7 +129,7 @@ class WebpageQATool(BaseTool):
         results = []
         # TODO: Handle this with a MapReduceChain
         for i in range(0, len(web_docs), 4):
-            input_docs = web_docs[i : i + 4]
+            input_docs = web_docs[i: i + 4]
             window_result = self.qa_chain(
                 {"input_documents": input_docs, "question": question},
                 return_only_outputs=True,

swarms/tools/developer.py

@@ -306,7 +306,7 @@ class WriteCommand:
     @staticmethod
     def from_str(command: str) -> "WriteCommand":
         filepath = command.split(WriteCommand.separator)[0]
-        return WriteCommand(filepath, command[len(filepath) + 1 :])
+        return WriteCommand(filepath, command[len(filepath) + 1:])
 
 
 class CodeWriter:
@@ -433,7 +433,7 @@ class ReadCommand:
         if self.start == self.end:
             code = code[self.start - 1]
         else:
-            code = "".join(code[self.start - 1 : self.end])
+            code = "".join(code[self.start - 1: self.end])
         return code
 
     @staticmethod
@@ -590,9 +590,9 @@ class PatchCommand:
         lines[self.start.line] = (
             lines[self.start.line][: self.start.col]
             + self.content
-            + lines[self.end.line][self.end.col :]
+            + lines[self.end.line][self.end.col:]
         )
-        lines = lines[: self.start.line + 1] + lines[self.end.line + 1 :]
+        lines = lines[: self.start.line + 1] + lines[self.end.line + 1:]
 
         after = self.write_lines(lines)
 

swarms/utils/main.py

@@ -365,7 +365,7 @@ class FileHandler:
         try:
             if url.startswith(os.environ.get("SERVER", "http://localhost:8000")):
                 local_filepath = url[
-                    len(os.environ.get("SERVER", "http://localhost:8000")) + 1 :
+                    len(os.environ.get("SERVER", "http://localhost:8000")) + 1:
                 ]
                 local_filename = Path("file") / local_filepath.split("/")[-1]
                 src = self.path / local_filepath