Aller directement au contenu principal

openhands.core.config.condenser_config

NoOpCondenserConfig Objects​

class NoOpCondenserConfig(BaseModel)

Configuration for NoOpCondenser.

ObservationMaskingCondenserConfig Objects​

class ObservationMaskingCondenserConfig(BaseModel)

Configuration for ObservationMaskingCondenser.

BrowserOutputCondenserConfig Objects​

class BrowserOutputCondenserConfig(BaseModel)

Configuration for the BrowserOutputCondenser.

RecentEventsCondenserConfig Objects​

class RecentEventsCondenserConfig(BaseModel)

Configuration for RecentEventsCondenser.

LLMSummarizingCondenserConfig Objects​

class LLMSummarizingCondenserConfig(BaseModel)

Configuration for LLMCondenser.

AmortizedForgettingCondenserConfig Objects​

class AmortizedForgettingCondenserConfig(BaseModel)

Configuration for AmortizedForgettingCondenser.

LLMAttentionCondenserConfig Objects​

class LLMAttentionCondenserConfig(BaseModel)

Configuration for LLMAttentionCondenser.

condenser_config_from_toml_section​

def condenser_config_from_toml_section(
        data: dict,
        llm_configs: dict | None = None) -> dict[str, CondenserConfig]

Create a CondenserConfig instance from a toml dictionary representing the [condenser] section.

For CondenserConfig, the handling is different since it's a union type. The type of condenser is determined by the 'type' field in the section.

Example:

Parse condenser config like: [condenser] type = "noop"

For condensers that require an LLM config, you can specify the name of an LLM config: [condenser] type = "llm" llm_config = "my_llm" # References [llm.my_llm] section

Arguments:

  • data - The TOML dictionary representing the [condenser] section.
  • llm_configs - Optional dictionary of LLMConfig objects keyed by name.

Returns:

dict[str, CondenserConfig]: A mapping where the key "condenser" corresponds to the configuration.

create_condenser_config​

def create_condenser_config(condenser_type: str,
                            data: dict) -> CondenserConfig

Create a CondenserConfig instance based on the specified type.

Arguments:

  • condenser_type - The type of condenser to create.
  • data - The configuration data.

Returns:

A CondenserConfig instance.

Raises:

  • ValueError - If the condenser type is unknown.
  • ValidationError - If the provided data fails validation for the condenser type.