sdg-hub 0.3.1__py3-none-any.whl → 0.4.1__py3-none-any.whl

sdg_hub/core/flow/base.py

@@ -35,7 +35,7 @@ from ..utils.logger_config import setup_logger
 from ..utils.path_resolution import resolve_path
 from ..utils.yaml_utils import save_flow_yaml
 from .checkpointer import FlowCheckpointer
-from .metadata import DatasetRequirements, FlowMetadata, FlowParameter
+from .metadata import DatasetRequirements, FlowMetadata
 from .migration import FlowMigration
 from .validation import FlowValidator
 
@@ -55,8 +55,6 @@ class Flow(BaseModel):
         Ordered list of blocks to execute in the flow.
     metadata : FlowMetadata
         Flow metadata including name, version, author, etc.
-    parameters : Dict[str, FlowParameter]
-        Runtime parameters that can be overridden during execution.
     """
 
     blocks: list[BaseBlock] = Field(
@@ -66,10 +64,6 @@ class Flow(BaseModel):
     metadata: FlowMetadata = Field(
         description="Flow metadata including name, version, author, etc."
     )
-    parameters: dict[str, FlowParameter] = Field(
-        default_factory=dict,
-        description="Runtime parameters that can be overridden during execution",
-    )
 
     model_config = ConfigDict(extra="forbid", arbitrary_types_allowed=True)
 
@@ -96,32 +90,6 @@ class Flow(BaseModel):
 
         return v
 
-    @field_validator("parameters")
-    @classmethod
-    def validate_parameters(
-        cls, v: dict[str, FlowParameter]
-    ) -> dict[str, FlowParameter]:
-        """Validate parameter names and ensure they are FlowParameter instances."""
-        if not v:
-            return v
-
-        validated = {}
-        for param_name, param_value in v.items():
-            if not isinstance(param_name, str) or not param_name.strip():
-                raise ValueError(
-                    f"Parameter name must be a non-empty string: {param_name}"
-                )
-
-            if not isinstance(param_value, FlowParameter):
-                raise ValueError(
-                    f"Parameter '{param_name}' must be a FlowParameter instance, "
-                    f"got: {type(param_value)}"
-                )
-
-            validated[param_name.strip()] = param_value
-
-        return validated
-
     @model_validator(mode="after")
     def validate_block_names_unique(self) -> "Flow":
         """Ensure all block names are unique within the flow."""
@@ -215,17 +183,6 @@ class Flow(BaseModel):
         except Exception as exc:
             raise FlowValidationError(f"Invalid metadata configuration: {exc}") from exc
 
-        # Extract and validate parameters
-        parameters = {}
-        params_dict = flow_config.get("parameters", {})
-        for param_name, param_config in params_dict.items():
-            try:
-                parameters[param_name] = FlowParameter(**param_config)
-            except Exception as exc:
-                raise FlowValidationError(
-                    f"Invalid parameter '{param_name}': {exc}"
-                ) from exc
-
         # Create blocks with validation
         blocks = []
         block_configs = flow_config.get("blocks", [])
@@ -254,7 +211,7 @@ class Flow(BaseModel):
 
         # Create and validate the flow
         try:
-            flow = cls(blocks=blocks, metadata=metadata, parameters=parameters)
+            flow = cls(blocks=blocks, metadata=metadata)
             # Persist generated id back to the YAML file (only on initial load)
             # If the file had no metadata.id originally, update and rewrite
             if not flow_config.get("metadata", {}).get("id"):
@@ -877,16 +834,19 @@ class Flow(BaseModel):
                             f"Block '{block.block_name}': {param_name} "
                             f"'{old_value}' -> '{param_value}'"
                         )
+                    ## check if allow extra
+                    elif block.model_config["extra"] == "allow":
+                        setattr(block, param_name, param_value)
+                        logger.debug(
+                            f"Block '{block.block_name}': {param_name} "
+                            f"'{old_value}' -> '{param_value}'"
+                        )
                     else:
                         logger.warning(
                             f"Block '{block.block_name}' ({block.__class__.__name__}) "
                             f"does not have attribute '{param_name}' - skipping"
                         )
 
-                # Reinitialize client manager for LLM blocks after updating config
-                if hasattr(block, "_reinitialize_client_manager"):
-                    block._reinitialize_client_manager()
-
                 modified_count += 1
 
         if modified_count > 0:
@@ -1222,17 +1182,12 @@ class Flow(BaseModel):
         # Create new flow with added block
         new_blocks = self.blocks + [block]
 
-        return Flow(
-            blocks=new_blocks, metadata=self.metadata, parameters=self.parameters
-        )
+        return Flow(blocks=new_blocks, metadata=self.metadata)
 
     def get_info(self) -> dict[str, Any]:
         """Get information about the flow."""
         return {
             "metadata": self.metadata.model_dump(),
-            "parameters": {
-                name: param.model_dump() for name, param in self.parameters.items()
-            },
             "blocks": [
                 {
                     "block_type": block.__class__.__name__,
@@ -1336,8 +1291,7 @@ class Flow(BaseModel):
 
         The summary contains:
         1. Flow metadata (name, version, author, description)
-        2. Defined runtime parameters with type hints and defaults
-        3. A table of all blocks with their input and output columns
+        2. A table of all blocks with their input and output columns
 
         Notes
         -----
@@ -1371,17 +1325,6 @@ class Flow(BaseModel):
                 f"Description: [white]{self.metadata.description}[/white]"
             )
 
-        # Parameters section
-        if self.parameters:
-            params_branch = flow_tree.add(
-                "[bold bright_yellow]Parameters[/bold bright_yellow]"
-            )
-            for name, param in self.parameters.items():
-                param_info = f"[bright_cyan]{name}[/bright_cyan]: [white]{param.type_hint}[/white]"
-                if param.default is not None:
-                    param_info += f" = [bright_white]{param.default}[/bright_white]"
-                params_branch.add(param_info)
-
         # Blocks overview
         flow_tree.add(
             f"[bold bright_magenta]Blocks[/bold bright_magenta] ({len(self.blocks)} total)"
@@ -1443,11 +1386,6 @@ class Flow(BaseModel):
             ],
         }
 
-        if self.parameters:
-            config["parameters"] = {
-                name: param.model_dump() for name, param in self.parameters.items()
-            }
-
         save_flow_yaml(output_path, config)
 
     def __len__(self) -> int:

sdg_hub/core/flow/metadata.py

@@ -2,9 +2,8 @@
 """Flow metadata and parameter definitions."""
 
 # Standard
-from datetime import datetime
 from enum import Enum
-from typing import Any, Optional
+from typing import Optional
 
 # Third Party
 from pydantic import BaseModel, Field, field_validator, model_validator
@@ -118,39 +117,6 @@ class RecommendedModels(BaseModel):
         return None
 
 
-class FlowParameter(BaseModel):
-    """Represents a runtime parameter for a flow.
-
-    Attributes
-    ----------
-    default : Any
-        Default value for the parameter.
-    description : str
-        Human-readable description of the parameter.
-    type_hint : str
-        Type hint as string (e.g., "float", "str").
-    required : bool
-        Whether this parameter is required at runtime.
-    constraints : Dict[str, Any]
-        Additional constraints for the parameter.
-    """
-
-    default: Any = Field(..., description="Default value for the parameter")
-    description: str = Field(default="", description="Human-readable description")
-    type_hint: str = Field(default="Any", description="Type hint as string")
-    required: bool = Field(default=False, description="Whether parameter is required")
-    constraints: dict[str, Any] = Field(
-        default_factory=dict, description="Additional constraints for the parameter"
-    )
-
-    @model_validator(mode="after")
-    def validate_required_default(self) -> "FlowParameter":
-        """Validate that required parameters have appropriate defaults."""
-        if self.required and self.default is None:
-            raise ValueError("Required parameters cannot have None as default")
-        return self
-
-
 class DatasetRequirements(BaseModel):
     """Dataset requirements for flow execution.
 
@@ -255,20 +221,10 @@ class FlowMetadata(BaseModel):
         Simplified recommended models structure with default, compatible, and experimental lists.
     tags : List[str]
         Tags for categorization and search.
-    created_at : str
-        Creation timestamp.
-    updated_at : str
-        Last update timestamp.
     license : str
         License identifier.
-    min_sdg_hub_version : str
-        Minimum required SDG Hub version.
     dataset_requirements : Optional[DatasetRequirements]
         Requirements for input datasets.
-    estimated_cost : str
-        Estimated cost tier for running the flow.
-    estimated_duration : str
-        Estimated duration for flow execution.
     """
 
     name: str = Field(..., min_length=1, description="Human-readable name")
@@ -288,29 +244,10 @@ class FlowMetadata(BaseModel):
     tags: list[str] = Field(
         default_factory=list, description="Tags for categorization and search"
     )
-    created_at: str = Field(
-        default_factory=lambda: datetime.now().isoformat(),
-        description="Creation timestamp",
-    )
-    updated_at: str = Field(
-        default_factory=lambda: datetime.now().isoformat(),
-        description="Last update timestamp",
-    )
     license: str = Field(default="Apache-2.0", description="License identifier")
-    min_sdg_hub_version: str = Field(
-        default="", description="Minimum required SDG Hub version"
-    )
     dataset_requirements: Optional[DatasetRequirements] = Field(
         default=None, description="Requirements for input datasets"
     )
-    estimated_cost: str = Field(
-        default="medium",
-        pattern="^(low|medium|high)$",
-        description="Estimated cost tier for running the flow",
-    )
-    estimated_duration: str = Field(
-        default="", description="Estimated duration for flow execution"
-    )
 
     @field_validator("id")
     @classmethod
@@ -352,10 +289,6 @@ class FlowMetadata(BaseModel):
         # Validation is handled within RecommendedModels class
         return v
 
-    def update_timestamp(self) -> None:
-        """Update the updated_at timestamp."""
-        self.updated_at = datetime.now().isoformat()
-
     @model_validator(mode="after")
     def ensure_id(self) -> "FlowMetadata":
         """Ensure id is set.

sdg_hub/core/flow/registry.py

@@ -360,7 +360,6 @@ class FlowRegistry:
                     "tags": ", ".join(metadata.tags) if metadata.tags else "-",
                     "description": metadata.description or "No description",
                     "version": metadata.version,
-                    "cost": metadata.estimated_cost,
                 }
             )
 

sdg_hub/flows/qa_generation/document_grounded_qa/enhanced_multi_summary_qa/detailed_summary/flow.yaml

@@ -17,7 +17,6 @@ metadata:
   - qa-pairs
   - detailed-summaries
   license: Apache-2.0
-  min_sdg_hub_version: 0.2.0
   dataset_requirements:
     required_columns:
     - document
@@ -61,10 +60,16 @@ blocks:
     temperature: 0.7
     n: 50
     async_mode: true
+- block_type: LLMParserBlock
+  block_config:
+    block_name: extract_detailed_summary
+    input_cols: raw_summary
+    extract_content: true
+    expand_lists: true
 - block_type: TextParserBlock
   block_config:
     block_name: parse_detailed_summary
-    input_cols: raw_summary
+    input_cols: extract_detailed_summary_content
     output_cols: summary
     start_tags:
     - ''
@@ -99,10 +104,16 @@ blocks:
     temperature: 0.7
     n: 1
     async_mode: true
+- block_type: LLMParserBlock
+  block_config:
+    block_name: extract_questions
+    input_cols: question_list
+    extract_content: true
+    expand_lists: true
 - block_type: TextParserBlock
   block_config:
     block_name: parse_question_list
-    input_cols: question_list
+    input_cols: extract_questions_content
     output_cols: question
     start_tags:
     - '[QUESTION]'
@@ -127,33 +138,61 @@ blocks:
     temperature: 0.7
     n: 1
     async_mode: true
+- block_type: LLMParserBlock
+  block_config:
+    block_name: extract_answers
+    input_cols: response_dict
+    extract_content: true
+    expand_lists: true
 - block_type: TextParserBlock
   block_config:
     block_name: parse_response_dict
-    input_cols: response_dict
+    input_cols: extract_answers_content
     output_cols: response
     start_tags:
     - ''
     end_tags:
     - ''
     save_reasoning_content: true
-- block_type: EvaluateFaithfulnessBlock
+- block_type: PromptBuilderBlock
   block_config:
-    block_name: eval_faithfulness
+    block_name: eval_faithful_prompt
     input_cols:
     - document
     - response
-    output_cols:
-    - faithfulness_explanation
-    - faithfulness_judgment
+    output_cols: eval_faithful_prompt
     prompt_config_path: ../../multi_summary_qa/instructlab/evaluate_faithfulness.yaml
-    filter_value: 'YES'
-    operation: eq
-    async_mode: true
     format_as_messages: true
+- block_type: LLMChatBlock
+  block_config:
+    block_name: eval_faithful_llm_chat
+    input_cols: eval_faithful_prompt
+    output_cols: eval_faithful_response_dict
+    n: 1
+    async_mode: true
+- block_type: LLMParserBlock
+  block_config:
+    block_name: extract_eval_faithful
+    input_cols: eval_faithful_response_dict
+    extract_content: true
+    
+- block_type: TextParserBlock
+  block_config:
+    block_name: parse_eval_faithful
+    input_cols: extract_eval_faithful_content
+    output_cols: 
+    - faithfulness_explanation
+    - faithfulness_judgment
     start_tags:
     - '[Start of Explanation]'
     - '[Start of Answer]'
     end_tags:
     - '[End of Explanation]'
     - '[End of Answer]'
+- block_type: ColumnValueFilterBlock
+  block_config:
+    block_name: eval_faithful_filter
+    input_cols:
+      - faithfulness_judgment
+    filter_value: 'YES'
+    operation: eq

sdg_hub/flows/qa_generation/document_grounded_qa/enhanced_multi_summary_qa/doc_direct_qa/flow.yaml

@@ -0,0 +1,158 @@
+metadata:
+  name: Document Based Knowledge Tuning Dataset Generation Flow
+  description: Directly generates QA pairs from the raw document.
+  version: 2.0.0
+  author: SDG Hub Contributors
+  recommended_models:
+    default: openai/gpt-oss-120b
+    compatible:
+    - meta-llama/Llama-3.3-70B-Instruct
+    - microsoft/phi-4
+    - mistralai/Mixtral-8x7B-Instruct-v0.1
+    experimental: []
+  tags:
+  - knowledge-tuning
+  - document-internalization
+  - question-generation
+  - qa-pairs
+  - detailed-summaries
+  license: Apache-2.0
+  dataset_requirements:
+    required_columns:
+    - document
+    - document_outline
+    - domain
+    - icl_document
+    - icl_query_1
+    - icl_query_2
+    - icl_query_3
+    description: 'Input dataset should contain documents with text content and domain classification. Each document should be substantial enough for meaningful question generation (minimum 100 words recommended). The flow generates three types
+      of summaries: detailed (n=20), extractive (n=10), and key facts (n=50), each producing corresponding QA pairs designed to help LLMs internalize document knowledge for knowledge tuning.'
+  output_columns:
+  - question
+  - response
+  - raw_document
+  - faithfulness_explanation
+  - faithfulness_judgment
+  id: stellar-peak-605
+blocks:
+- block_type: DuplicateColumnsBlock
+  block_config:
+    block_name: duplicate_document_col
+    input_cols:
+      document: base_document
+- block_type: PromptBuilderBlock
+  block_config:
+    block_name: question_generation_prompt
+    input_cols:
+    - domain
+    - document
+    - document_outline
+    - icl_document
+    - icl_query_1
+    - icl_query_2
+    - icl_query_3
+    output_cols: question_generation_prompt
+    prompt_config_path: ../generate_question_list.yaml
+    format_as_messages: true
+- block_type: LLMChatBlock
+  block_config:
+    block_name: question_generation
+    input_cols: question_generation_prompt
+    output_cols: question_list
+    max_tokens: 256
+    temperature: 1.0
+    n: 1
+    async_mode: true
+- block_type: LLMParserBlock
+  block_config:
+    block_name: extract_questions
+    input_cols: question_list
+    extract_content: true
+    expand_lists: true
+- block_type: TextParserBlock
+  block_config:
+    block_name: parse_question_list
+    input_cols: extract_questions_content
+    output_cols: question
+    start_tags:
+    - '[QUESTION]'
+    end_tags:
+    - '[END]'
+- block_type: PromptBuilderBlock
+  block_config:
+    block_name: answer_generation_prompt
+    input_cols:
+    - question
+    - document
+    - document_outline
+    output_cols: answer_generation_prompt
+    prompt_config_path: ../generate_answers.yaml
+    format_as_messages: true
+- block_type: LLMChatBlock
+  block_config:
+    block_name: answer_generation
+    input_cols: answer_generation_prompt
+    output_cols: response_dict
+    max_tokens: 4096
+    temperature: 1.0
+    n: 1
+    async_mode: true
+- block_type: LLMParserBlock
+  block_config:
+    block_name: extract_answer
+    input_cols: response_dict
+    extract_content: true
+    expand_lists: true
+- block_type: TextParserBlock
+  block_config:
+    block_name: parse_response_dict
+    input_cols: extract_answer_content
+    output_cols: response
+    start_tags:
+    - ''
+    end_tags:
+    - ''
+    save_reasoning_content: true
+- block_type: PromptBuilderBlock
+  block_config:
+    block_name: eval_faithful_prompt
+    input_cols:
+    - document
+    - response
+    output_cols: eval_faithful_prompt
+    prompt_config_path: ../../multi_summary_qa/instructlab/evaluate_faithfulness.yaml
+    format_as_messages: true
+- block_type: LLMChatBlock
+  block_config:
+    block_name: eval_faithful_llm_chat
+    input_cols: eval_faithful_prompt
+    output_cols: eval_faithful_response_dict
+    n: 1
+    async_mode: true
+- block_type: LLMParserBlock
+  block_config:
+    block_name: extract_eval_faithful
+    input_cols: eval_faithful_response_dict
+    extract_content: true
+    
+- block_type: TextParserBlock
+  block_config:
+    block_name: parse_eval_faithful
+    input_cols: extract_eval_faithful_content
+    output_cols: 
+    - faithfulness_explanation
+    - faithfulness_judgment
+    start_tags:
+    - '[Start of Explanation]'
+    - '[Start of Answer]'
+    end_tags:
+    - '[End of Explanation]'
+    - '[End of Answer]'
+- block_type: ColumnValueFilterBlock
+  block_config:
+    block_name: eval_faithful_filter
+    input_cols:
+      - faithfulness_judgment
+    filter_value: 'YES'
+    operation: eq

sdg_hub/flows/qa_generation/document_grounded_qa/enhanced_multi_summary_qa/extractive_summary/flow.yaml

@@ -19,7 +19,6 @@ metadata:
   - qa-pairs
   - extractive-summaries
   license: Apache-2.0
-  min_sdg_hub_version: 0.2.0
   dataset_requirements:
     required_columns:
     - document
@@ -63,10 +62,16 @@ blocks:
     temperature: 0.7
     n: 50
     async_mode: true
+- block_type: LLMParserBlock
+  block_config:
+    block_name: extract_extractive_summary
+    input_cols: raw_summary
+    extract_content: true
+    expand_lists: true
 - block_type: TextParserBlock
   block_config:
     block_name: parse_extractive_summary
-    input_cols: raw_summary
+    input_cols: extract_extractive_summary_content
     output_cols: summary
     start_tags:
     - ''
@@ -101,10 +106,16 @@ blocks:
     temperature: 0.7
     n: 1
     async_mode: true
+- block_type: LLMParserBlock
+  block_config:
+    block_name: extract_questions
+    input_cols: question_list
+    extract_content: true
+    expand_lists: true
 - block_type: TextParserBlock
   block_config:
     block_name: parse_question_list
-    input_cols: question_list
+    input_cols: extract_questions_content
     output_cols: question
     start_tags:
     - '[QUESTION]'
@@ -129,33 +140,61 @@ blocks:
     temperature: 0.7
     n: 1
     async_mode: true
+- block_type: LLMParserBlock
+  block_config:
+    block_name: extract_answers
+    input_cols: response_dict
+    extract_content: true
+    expand_lists: true
 - block_type: TextParserBlock
   block_config:
     block_name: parse_response_dict
-    input_cols: response_dict
+    input_cols: extract_answers_content
     output_cols: response
     start_tags:
     - ''
     end_tags:
     - ''
     save_reasoning_content: true
-- block_type: EvaluateFaithfulnessBlock
+- block_type: PromptBuilderBlock
   block_config:
-    block_name: eval_faithfulness
+    block_name: eval_faithful_prompt
     input_cols:
     - document
     - response
-    output_cols:
-    - faithfulness_explanation
-    - faithfulness_judgment
+    output_cols: eval_faithful_prompt
     prompt_config_path: ../../multi_summary_qa/instructlab/evaluate_faithfulness.yaml
-    filter_value: 'YES'
-    operation: eq
-    async_mode: true
     format_as_messages: true
+- block_type: LLMChatBlock
+  block_config:
+    block_name: eval_faithful_llm_chat
+    input_cols: eval_faithful_prompt
+    output_cols: eval_faithful_response_dict
+    n: 1
+    async_mode: true
+- block_type: LLMParserBlock
+  block_config:
+    block_name: extract_eval_faithful
+    input_cols: eval_faithful_response_dict
+    extract_content: true
+    
+- block_type: TextParserBlock
+  block_config:
+    block_name: parse_eval_faithful
+    input_cols: extract_eval_faithful_content
+    output_cols: 
+    - faithfulness_explanation
+    - faithfulness_judgement
     start_tags:
     - '[Start of Explanation]'
     - '[Start of Answer]'
     end_tags:
     - '[End of Explanation]'
     - '[End of Answer]'
+- block_type: ColumnValueFilterBlock
+  block_config:
+    block_name: eval_faithful_filter
+    input_cols:
+      - faithfulness_judgement
+    filter_value: 'YES'
+    operation: eq