lfx-nightly 0.2.0.dev25__py3-none-any.whl

lfx/components/ibm/watsonx.py

@@ -0,0 +1,207 @@
+import json
+from typing import Any
+
+import requests
+from langchain_ibm import ChatWatsonx
+from pydantic.v1 import SecretStr
+
+from lfx.base.models.model import LCModelComponent
+from lfx.field_typing import LanguageModel
+from lfx.field_typing.range_spec import RangeSpec
+from lfx.inputs.inputs import BoolInput, DropdownInput, IntInput, SecretStrInput, SliderInput, StrInput
+from lfx.log.logger import logger
+from lfx.schema.dotdict import dotdict
+
+
+class WatsonxAIComponent(LCModelComponent):
+    display_name = "IBM watsonx.ai"
+    description = "Generate text using IBM watsonx.ai foundation models."
+    icon = "WatsonxAI"
+    name = "IBMwatsonxModel"
+    beta = False
+
+    _default_models = ["ibm/granite-3-2b-instruct", "ibm/granite-3-8b-instruct", "ibm/granite-13b-instruct-v2"]
+    _urls = [
+        "https://us-south.ml.cloud.ibm.com",
+        "https://eu-de.ml.cloud.ibm.com",
+        "https://eu-gb.ml.cloud.ibm.com",
+        "https://au-syd.ml.cloud.ibm.com",
+        "https://jp-tok.ml.cloud.ibm.com",
+        "https://ca-tor.ml.cloud.ibm.com",
+    ]
+    inputs = [
+        *LCModelComponent.get_base_inputs(),
+        DropdownInput(
+            name="base_url",
+            display_name="watsonx API Endpoint",
+            info="The base URL of the API.",
+            value=[],
+            options=_urls,
+            real_time_refresh=True,
+            required=True,
+        ),
+        StrInput(
+            name="project_id",
+            display_name="watsonx Project ID",
+            required=True,
+            info="The project ID or deployment space ID that is associated with the foundation model.",
+        ),
+        SecretStrInput(
+            name="api_key",
+            display_name="Watsonx API Key",
+            info="The API Key to use for the model.",
+            required=True,
+        ),
+        DropdownInput(
+            name="model_name",
+            display_name="Model Name",
+            options=[],
+            value=None,
+            real_time_refresh=True,
+            required=True,
+            refresh_button=True,
+        ),
+        IntInput(
+            name="max_tokens",
+            display_name="Max Tokens",
+            advanced=True,
+            info="The maximum number of tokens to generate.",
+            range_spec=RangeSpec(min=1, max=4096),
+            value=1000,
+        ),
+        StrInput(
+            name="stop_sequence",
+            display_name="Stop Sequence",
+            advanced=True,
+            info="Sequence where generation should stop.",
+            field_type="str",
+        ),
+        SliderInput(
+            name="temperature",
+            display_name="Temperature",
+            info="Controls randomness, higher values increase diversity.",
+            value=0.1,
+            range_spec=RangeSpec(min=0, max=2, step=0.01),
+            advanced=True,
+        ),
+        SliderInput(
+            name="top_p",
+            display_name="Top P",
+            info="The cumulative probability cutoff for token selection. "
+            "Lower values mean sampling from a smaller, more top-weighted nucleus.",
+            value=0.9,
+            range_spec=RangeSpec(min=0, max=1, step=0.01),
+            advanced=True,
+        ),
+        SliderInput(
+            name="frequency_penalty",
+            display_name="Frequency Penalty",
+            info="Penalty for frequency of token usage.",
+            value=0.5,
+            range_spec=RangeSpec(min=-2.0, max=2.0, step=0.01),
+            advanced=True,
+        ),
+        SliderInput(
+            name="presence_penalty",
+            display_name="Presence Penalty",
+            info="Penalty for token presence in prior text.",
+            value=0.3,
+            range_spec=RangeSpec(min=-2.0, max=2.0, step=0.01),
+            advanced=True,
+        ),
+        IntInput(
+            name="seed",
+            display_name="Random Seed",
+            advanced=True,
+            info="The random seed for the model.",
+            value=8,
+        ),
+        BoolInput(
+            name="logprobs",
+            display_name="Log Probabilities",
+            advanced=True,
+            info="Whether to return log probabilities of the output tokens.",
+            value=True,
+        ),
+        IntInput(
+            name="top_logprobs",
+            display_name="Top Log Probabilities",
+            advanced=True,
+            info="Number of most likely tokens to return at each position.",
+            value=3,
+            range_spec=RangeSpec(min=1, max=20),
+        ),
+        StrInput(
+            name="logit_bias",
+            display_name="Logit Bias",
+            advanced=True,
+            info='JSON string of token IDs to bias or suppress (e.g., {"1003": -100, "1004": 100}).',
+            field_type="str",
+        ),
+    ]
+
+    @staticmethod
+    def fetch_models(base_url: str) -> list[str]:
+        """Fetch available models from the watsonx.ai API."""
+        try:
+            endpoint = f"{base_url}/ml/v1/foundation_model_specs"
+            params = {"version": "2024-09-16", "filters": "function_text_chat,!lifecycle_withdrawn"}
+            response = requests.get(endpoint, params=params, timeout=10)
+            response.raise_for_status()
+            data = response.json()
+            models = [model["model_id"] for model in data.get("resources", [])]
+            return sorted(models)
+        except Exception:  # noqa: BLE001
+            logger.exception("Error fetching models. Using default models.")
+            return WatsonxAIComponent._default_models
+
+    def update_build_config(self, build_config: dotdict, field_value: Any, field_name: str | None = None):
+        """Update model options when URL or API key changes."""
+        if field_name == "base_url" and field_value:
+            try:
+                models = self.fetch_models(base_url=field_value)
+                build_config["model_name"]["options"] = models
+                if build_config["model_name"]["value"]:
+                    build_config["model_name"]["value"] = models[0]
+                info_message = f"Updated model options: {len(models)} models found in {field_value}"
+                logger.info(info_message)
+            except Exception:  # noqa: BLE001
+                logger.exception("Error updating model options.")
+        if field_name == "model_name" and field_value and field_value in WatsonxAIComponent._urls:
+            build_config["model_name"]["options"] = self.fetch_models(base_url=field_value)
+            build_config["model_name"]["value"] = ""
+        return build_config
+
+    def build_model(self) -> LanguageModel:
+        # Parse logit_bias from JSON string if provided
+        logit_bias = None
+        if hasattr(self, "logit_bias") and self.logit_bias:
+            try:
+                logit_bias = json.loads(self.logit_bias)
+            except json.JSONDecodeError:
+                logger.warning("Invalid logit_bias JSON format. Using default instead.")
+                logit_bias = {"1003": -100, "1004": -100}
+
+        chat_params = {
+            "max_tokens": getattr(self, "max_tokens", None),
+            "temperature": getattr(self, "temperature", None),
+            "top_p": getattr(self, "top_p", None),
+            "frequency_penalty": getattr(self, "frequency_penalty", None),
+            "presence_penalty": getattr(self, "presence_penalty", None),
+            "seed": getattr(self, "seed", None),
+            "stop": [self.stop_sequence] if self.stop_sequence else [],
+            "n": 1,
+            "logprobs": getattr(self, "logprobs", True),
+            "top_logprobs": getattr(self, "top_logprobs", None),
+            "time_limit": 600000,
+            "logit_bias": logit_bias,
+        }
+
+        return ChatWatsonx(
+            apikey=SecretStr(self.api_key).get_secret_value(),
+            url=self.base_url,
+            project_id=self.project_id,
+            model_id=self.model_name,
+            params=chat_params,
+            streaming=self.stream,
+        )

lfx/components/ibm/watsonx_embeddings.py

@@ -0,0 +1,135 @@
+from typing import Any
+
+import requests
+from ibm_watsonx_ai import APIClient, Credentials
+from ibm_watsonx_ai.metanames import EmbedTextParamsMetaNames
+from langchain_ibm import WatsonxEmbeddings
+from pydantic.v1 import SecretStr
+
+from lfx.base.embeddings.model import LCEmbeddingsModel
+from lfx.field_typing import Embeddings
+from lfx.io import BoolInput, DropdownInput, IntInput, SecretStrInput, StrInput
+from lfx.log.logger import logger
+from lfx.schema.dotdict import dotdict
+
+
+class WatsonxEmbeddingsComponent(LCEmbeddingsModel):
+    display_name = "IBM watsonx.ai Embeddings"
+    description = "Generate embeddings using IBM watsonx.ai models."
+    icon = "WatsonxAI"
+    name = "WatsonxEmbeddingsComponent"
+
+    # models present in all the regions
+    _default_models = [
+        "sentence-transformers/all-minilm-l12-v2",
+        "ibm/slate-125m-english-rtrvr-v2",
+        "ibm/slate-30m-english-rtrvr-v2",
+        "intfloat/multilingual-e5-large",
+    ]
+
+    inputs = [
+        DropdownInput(
+            name="url",
+            display_name="watsonx API Endpoint",
+            info="The base URL of the API.",
+            value=None,
+            options=[
+                "https://us-south.ml.cloud.ibm.com",
+                "https://eu-de.ml.cloud.ibm.com",
+                "https://eu-gb.ml.cloud.ibm.com",
+                "https://au-syd.ml.cloud.ibm.com",
+                "https://jp-tok.ml.cloud.ibm.com",
+                "https://ca-tor.ml.cloud.ibm.com",
+            ],
+            real_time_refresh=True,
+        ),
+        StrInput(
+            name="project_id",
+            display_name="watsonx project id",
+            info="The project ID or deployment space ID that is associated with the foundation model.",
+            required=True,
+        ),
+        SecretStrInput(
+            name="api_key",
+            display_name="Watsonx API Key",
+            info="The API Key to use for the model.",
+            required=True,
+        ),
+        DropdownInput(
+            name="model_name",
+            display_name="Model Name",
+            options=[],
+            value=None,
+            dynamic=True,
+            required=True,
+        ),
+        IntInput(
+            name="truncate_input_tokens",
+            display_name="Truncate Input Tokens",
+            advanced=True,
+            value=200,
+        ),
+        BoolInput(
+            name="input_text",
+            display_name="Include the original text in the output",
+            value=True,
+            advanced=True,
+        ),
+    ]
+
+    @staticmethod
+    def fetch_models(base_url: str) -> list[str]:
+        """Fetch available models from the watsonx.ai API."""
+        try:
+            endpoint = f"{base_url}/ml/v1/foundation_model_specs"
+            params = {
+                "version": "2024-09-16",
+                "filters": "function_embedding,!lifecycle_withdrawn:and",
+            }
+            response = requests.get(endpoint, params=params, timeout=10)
+            response.raise_for_status()
+            data = response.json()
+            models = [model["model_id"] for model in data.get("resources", [])]
+            return sorted(models)
+        except Exception:  # noqa: BLE001
+            logger.exception("Error fetching models")
+            return WatsonxEmbeddingsComponent._default_models
+
+    def update_build_config(self, build_config: dotdict, field_value: Any, field_name: str | None = None):
+        """Update model options when URL or API key changes."""
+        logger.debug(
+            "Updating build config. Field name: %s, Field value: %s",
+            field_name,
+            field_value,
+        )
+
+        if field_name == "url" and field_value:
+            try:
+                models = self.fetch_models(base_url=build_config.url.value)
+                build_config.model_name.options = models
+                if build_config.model_name.value:
+                    build_config.model_name.value = models[0]
+                info_message = f"Updated model options: {len(models)} models found in {build_config.url.value}"
+                logger.info(info_message)
+            except Exception:  # noqa: BLE001
+                logger.exception("Error updating model options.")
+
+    def build_embeddings(self) -> Embeddings:
+        credentials = Credentials(
+            api_key=SecretStr(self.api_key).get_secret_value(),
+            url=self.url,
+        )
+
+        api_client = APIClient(credentials)
+
+        params = {
+            EmbedTextParamsMetaNames.TRUNCATE_INPUT_TOKENS: self.truncate_input_tokens,
+            EmbedTextParamsMetaNames.RETURN_OPTIONS: {"input_text": self.input_text},
+        }
+
+        return WatsonxEmbeddings(
+            model_id=self.model_name,
+            params=params,
+            watsonx_client=api_client,
+            project_id=self.project_id,
+        )

lfx/components/icosacomputing/__init__.py

@@ -0,0 +1,5 @@
+from .combinatorial_reasoner import CombinatorialReasonerComponent
+
+__all__ = [
+    "CombinatorialReasonerComponent",
+]

lfx/components/icosacomputing/combinatorial_reasoner.py

@@ -0,0 +1,84 @@
+import requests
+from requests.auth import HTTPBasicAuth
+
+from lfx.base.models.openai_constants import OPENAI_CHAT_MODEL_NAMES
+from lfx.custom.custom_component.component import Component
+from lfx.inputs.inputs import DropdownInput, SecretStrInput, StrInput
+from lfx.io import MessageTextInput, Output
+from lfx.schema.data import Data
+from lfx.schema.message import Message
+
+
+class CombinatorialReasonerComponent(Component):
+    display_name = "Combinatorial Reasoner"
+    description = "Uses Combinatorial Optimization to construct an optimal prompt with embedded reasons. Sign up here:\nhttps://forms.gle/oWNv2NKjBNaqqvCx6"
+    icon = "Icosa"
+    name = "Combinatorial Reasoner"
+
+    inputs = [
+        MessageTextInput(name="prompt", display_name="Prompt", required=True),
+        SecretStrInput(
+            name="openai_api_key",
+            display_name="OpenAI API Key",
+            info="The OpenAI API Key to use for the OpenAI model.",
+            advanced=False,
+            value="OPENAI_API_KEY",
+            required=True,
+        ),
+        StrInput(
+            name="username",
+            display_name="Username",
+            info="Username to authenticate access to Icosa CR API",
+            advanced=False,
+            required=True,
+        ),
+        SecretStrInput(
+            name="password",
+            display_name="Combinatorial Reasoner Password",
+            info="Password to authenticate access to Icosa CR API.",
+            advanced=False,
+            required=True,
+        ),
+        DropdownInput(
+            name="model_name",
+            display_name="Model Name",
+            advanced=False,
+            options=OPENAI_CHAT_MODEL_NAMES,
+            value=OPENAI_CHAT_MODEL_NAMES[0],
+        ),
+    ]
+
+    outputs = [
+        Output(
+            display_name="Optimized Prompt",
+            name="optimized_prompt",
+            method="build_prompt",
+        ),
+        Output(display_name="Selected Reasons", name="reasons", method="build_reasons"),
+    ]
+
+    def build_prompt(self) -> Message:
+        params = {
+            "prompt": self.prompt,
+            "apiKey": self.openai_api_key,
+            "model": self.model_name,
+        }
+
+        creds = HTTPBasicAuth(self.username, password=self.password)
+        response = requests.post(
+            "https://cr-api.icosacomputing.com/cr/langflow",
+            json=params,
+            auth=creds,
+            timeout=100,
+        )
+        response.raise_for_status()
+
+        prompt = response.json()["prompt"]
+
+        self.reasons = response.json()["finalReasons"]
+        return prompt
+
+    def build_reasons(self) -> Data:
+        # list of selected reasons
+        final_reasons = [reason[0] for reason in self.reasons]
+        return Data(value=final_reasons)

lfx/components/input_output/__init__.py

@@ -0,0 +1,40 @@
+from __future__ import annotations
+
+from typing import TYPE_CHECKING, Any
+
+from lfx.components._importing import import_mod
+
+if TYPE_CHECKING:
+    from lfx.components.input_output.chat import ChatInput
+    from lfx.components.input_output.chat_output import ChatOutput
+    from lfx.components.input_output.text import TextInputComponent
+    from lfx.components.input_output.text_output import TextOutputComponent
+    from lfx.components.input_output.webhook import WebhookComponent
+
+_dynamic_imports = {
+    "ChatInput": "chat",
+    "ChatOutput": "chat_output",
+    "TextInputComponent": "text",
+    "TextOutputComponent": "text_output",
+    "WebhookComponent": "webhook",
+}
+
+__all__ = ["ChatInput", "ChatOutput", "TextInputComponent", "TextOutputComponent", "WebhookComponent"]
+
+
+def __getattr__(attr_name: str) -> Any:
+    """Lazily import input/output components on attribute access."""
+    if attr_name not in _dynamic_imports:
+        msg = f"module '{__name__}' has no attribute '{attr_name}'"
+        raise AttributeError(msg)
+    try:
+        result = import_mod(attr_name, _dynamic_imports[attr_name], __spec__.parent)
+    except (ModuleNotFoundError, ImportError, AttributeError) as e:
+        msg = f"Could not import '{attr_name}' from '{__name__}': {e}"
+        raise AttributeError(msg) from e
+    globals()[attr_name] = result
+    return result
+
+
+def __dir__() -> list[str]:
+    return list(__all__)

lfx/components/input_output/chat.py

@@ -0,0 +1,109 @@
+from lfx.base.data.utils import IMG_FILE_TYPES, TEXT_FILE_TYPES
+from lfx.base.io.chat import ChatComponent
+from lfx.inputs.inputs import BoolInput
+from lfx.io import (
+    DropdownInput,
+    FileInput,
+    MessageTextInput,
+    MultilineInput,
+    Output,
+)
+from lfx.schema.message import Message
+from lfx.utils.constants import (
+    MESSAGE_SENDER_AI,
+    MESSAGE_SENDER_NAME_USER,
+    MESSAGE_SENDER_USER,
+)
+
+
+class ChatInput(ChatComponent):
+    display_name = "Chat Input"
+    description = "Get chat inputs from the Playground."
+    documentation: str = "https://docs.langflow.org/chat-input-and-output"
+    icon = "MessagesSquare"
+    name = "ChatInput"
+    minimized = True
+
+    inputs = [
+        MultilineInput(
+            name="input_value",
+            display_name="Input Text",
+            value="",
+            info="Message to be passed as input.",
+            input_types=[],
+        ),
+        BoolInput(
+            name="should_store_message",
+            display_name="Store Messages",
+            info="Store the message in the history.",
+            value=True,
+            advanced=True,
+        ),
+        DropdownInput(
+            name="sender",
+            display_name="Sender Type",
+            options=[MESSAGE_SENDER_AI, MESSAGE_SENDER_USER],
+            value=MESSAGE_SENDER_USER,
+            info="Type of sender.",
+            advanced=True,
+        ),
+        MessageTextInput(
+            name="sender_name",
+            display_name="Sender Name",
+            info="Name of the sender.",
+            value=MESSAGE_SENDER_NAME_USER,
+            advanced=True,
+        ),
+        MessageTextInput(
+            name="session_id",
+            display_name="Session ID",
+            info="The session ID of the chat. If empty, the current session ID parameter will be used.",
+            advanced=True,
+        ),
+        MessageTextInput(
+            name="context_id",
+            display_name="Context ID",
+            info="The context ID of the chat. Adds an extra layer to the local memory.",
+            value="",
+            advanced=True,
+        ),
+        FileInput(
+            name="files",
+            display_name="Files",
+            file_types=TEXT_FILE_TYPES + IMG_FILE_TYPES,
+            info="Files to be sent with the message.",
+            advanced=True,
+            is_list=True,
+            temp_file=True,
+        ),
+    ]
+    outputs = [
+        Output(display_name="Chat Message", name="message", method="message_response"),
+    ]
+
+    async def message_response(self) -> Message:
+        # Ensure files is a list and filter out empty/None values
+        files = self.files if self.files else []
+        if files and not isinstance(files, list):
+            files = [files]
+        # Filter out None/empty values
+        files = [f for f in files if f is not None and f != ""]
+
+        session_id = self.session_id or self.graph.session_id or ""
+        message = await Message.create(
+            text=self.input_value,
+            sender=self.sender,
+            sender_name=self.sender_name,
+            session_id=session_id,
+            context_id=self.context_id,
+            files=files,
+        )
+        if session_id and isinstance(message, Message) and self.should_store_message:
+            stored_message = await self.send_message(
+                message,
+            )
+            self.message.value = stored_message
+            message = stored_message
+
+        self.status = message
+        return message