Functions — github.com/microsoft/fara

CodeHub MCPcopy Create free account

hub / github.com/microsoft/fara / functions

Functions650 in github.com/microsoft/fara

⨍ Functions 650◇ Types & classes 122↳ Endpoints 17

( self, config: MMRubricAgentConfig | dict[str, Any] | None = None, **kwargs: Any )

webeval/src/webeval/rubric_agent/mm_rubric_agent.py:917

(self, data_dir: Path, eval_method = 'AgentTrek_eval', model_c

webeval/src/webeval/benchmarks/om2w/om2w.py:69

Init an OpenAI GPT/Codex engine Args: api_key (_type_, optional): Auth key from OpenAI. Defaults to None. stop (list,

webeval/src/webeval/benchmarks/om2w/impl/src/utils.py:65

( self, data_dir: str, model_client: Optional[ChatCompletionClient] = None, o4

webeval/src/webeval/benchmarks/webtailbench/webtailbench.py:68

( self, data_dir: str, eval_method: str = "exact_match", model_client: ChatCom

webeval/src/webeval/benchmarks/webvoyager/webvoyager.py:58

(self, period_s = 60, thread_safe = True, retries = 2, timeout = 5)

webeval/scripts/mlflow_rate_limiter.py:14

webeval/scripts/webtailbench.py:28

(self, option_strings, dest, type_map=None, **kwargs)

webeval/scripts/arg_parsing.py:9

(self, model_url, model_port, device_id, max_n_images, model_name=None, dtype='auto', enforce_eager=False, use

webeval/scripts/eval_exp.py:82

(self, callbacks = None)

webeval/scripts/eval_exp.py:128

(self, ws = None, user = None, seed = None, max_n_images = 5, save_task_csv = False)

webeval/scripts/eval_exp.py:160

webeval/scripts/om2w.py:16

webeval/scripts/webvoyager.py:17

(self, az_folder, local_path = None, tag = None)

webeval/scripts/analyze_eval_results/artifacts.py:68

webeval/tests/test_webtailbench_dataset.py:166

(self, exc: Exception, endpoint: str = "https://fake")

webeval/tests/test_oai_clients.py:145

(self, cfg=None)

src/fara/_prompts.py:120

( self, browser_manager: Any, client_config: dict, downloads_folder: str | Non

src/fara/fara_agent.py:67

(self, content: str, source: str = "system")

src/fara/fara_types.py:16

( self, content: str | List[Dict[str, Any]], source: str = "user", is_original

src/fara/fara_types.py:23

(self, content: str, source: str = "assistant")

src/fara/fara_types.py:36

A controller for Playwright to interact with web pages. animate_actions: If True, actions will be animated. downloads_folder:

src/fara/browser/playwright_controller.py:250

( self, viewport_height: int, viewport_width: int, headless: bool, pag

src/fara/browser/browser_bb.py:26

(self, name: str, arguments: str)

src/fara/qwen_helpers/schema.py:61

( self, text: Optional[str] = None, image: Optional[str] = None, file: Optiona

src/fara/qwen_helpers/schema.py:75

(self, cfg: Optional[dict] = None)

src/fara/qwen_helpers/base_tool.py:66

Initialize NousFnCallPrompt with a specific template. Args: template_name: Name of the template to use. Options:

src/fara/qwen_helpers/fncall_prompt.py:13

( self, model_url, port, device_id, max_n_images, dtype="auto"

src/fara/vllm/az_vllm.py:139

( self, model_path, max_n_images, device_id="0", host="0.0.0.0",

src/fara/vllm/vllm_facade.py:35

(cls, value: str, description: str = "")

webeval/src/webeval/rubric_agent/_cp_schema.py:55

Method__post_init__

webeval/src/webeval/trajectory.py:26

webeval/src/webeval/trajectory.py:189

webeval/scripts/analyze_eval_results/artifacts.py:54

src/fara/qwen_helpers/schema.py:64

src/fara/qwen_helpers/schema.py:105

src/fara/qwen_helpers/schema.py:157

Method__setitem__

(self, key, value)

src/fara/qwen_helpers/schema.py:30

src/fara/qwen_helpers/schema.py:53

Method_append_late

( d: Dict[str, List[Observation]], key: str, obs: Observation )

webeval/src/webeval/rubric_agent/data_point.py:308

Method_capture_env_state

( self, page, controller, base: str

webeval/src/webeval/systems/websurfer.py:65

Method_check_one

( sa: Dict[str, Any], x: int, y: int )

webeval/src/webeval/rubric_agent/verifier_agent.py:855

Method_download_handler

(download: Download)

src/fara/fara_agent.py:108

src/fara/vllm/vllm_facade.py:73

Function_filter_excluded_subcategories

Remove ``- **N.M …**`` bullet lines whose code is in *exclude_codes*. Each bullet may span multiple lines (the definition wraps). We detect t

webeval/src/webeval/rubric_agent/error_taxonomy_loader.py:143

Function_filter_excluded_summary_rows

Remove summary-table rows whose error code is in *exclude_codes*.

webeval/src/webeval/rubric_agent/error_taxonomy_loader.py:171

Method_generate_reply

(self, input_dict: Dict[str, Any])

webeval/tests/test_verify_trajectories.py:143

Method_get_config_class

webeval/src/webeval/rubric_agent/verifier_agent.py:163

Method_get_config_class

webeval/src/webeval/rubric_agent/critical_point_classifier.py:309

Method_get_config_class

webeval/src/webeval/rubric_agent/task_classification.py:151

Method_get_config_class

webeval/src/webeval/rubric_agent/mm_rubric_agent.py:966

Method_gpt5_client

webeval/src/webeval/rubric_agent/verifier_agent.py:171

Method_gpt5_client

webeval/src/webeval/rubric_agent/mm_rubric_agent.py:974

Method_o4mini_client

webeval/src/webeval/rubric_agent/verifier_agent.py:167

Method_o4mini_client

webeval/src/webeval/rubric_agent/mm_rubric_agent.py:970

Get the current page from browser manager.

src/fara/fara_agent.py:138

Method_playwright_controller

Get the playwright controller from browser manager.

src/fara/fara_agent.py:155

webeval/src/webeval/systems/websurfer.py:104

Method_screenshot_index

(filename: str)

webeval/src/webeval/rubric_agent/mm_rubric_agent.py:1368

Function_screenshot_sort_key

Trailing numeric index from a screenshot filename. Handles ``screenshot_1.png``, ``screenshot0.png``, ``screenshot_post_3.png`` etc.

webeval/src/webeval/benchmarks/webtailbench/shared_data_adapter.py:37

Method_set_action_definitions_from_tools

webeval/src/webeval/rubric_agent/mm_rubric_agent.py:841

Method_set_created_at

(cls, data: Any)

webeval/src/webeval/rubric_agent/data_point.py:611

Method_set_defaults

(cls, data: Any)

webeval/src/webeval/rubric_agent/data_point.py:102

Method_set_defaults

(cls, data: Any)

webeval/src/webeval/rubric_agent/data_point.py:171

Method_wrap_result

webeval/tests/test_verify_trajectories.py:185

Methodactual_usage

webeval/src/webeval/oai_clients/wrapper.py:199

Methodadd_action

(self, action: Action)

webeval/src/webeval/rubric_agent/data_point.py:374

Methodadd_observation

(self, observation: "ObservationEvent")

webeval/src/webeval/rubric_agent/data_point.py:371

Functionadd_om2w_args

webeval/scripts/om2w.py:25

Methodadd_token_usage

(self, key: str, token_usage: RequestUsage | Dict[str, int])

webeval/src/webeval/trajectory.py:41

Functionadd_webtailbench_args

webeval/scripts/webtailbench.py:39

Functionadd_webvoyager_args

webeval/scripts/webvoyager.py:26

Methodanalyze_multi_criteria_screenshot

( screenshot_idx: int, criterion_indices: List[int] )

webeval/src/webeval/rubric_agent/mm_rubric_agent.py:1851

Methodanalyze_single_pair

(criterion_idx: int, screenshot_idx: int)

webeval/src/webeval/rubric_agent/mm_rubric_agent.py:1635

Methodanalyze_single_pair_for_batch

( criterion_idx: int, screenshot_idx: int )

webeval/src/webeval/rubric_agent/mm_rubric_agent.py:1761

Functionare_urls_equal

Check if two URLs are equal after normalization. Args: url1 (str): The first URL. url2 (str): The second URL. Returns:

webeval/src/webeval/evaluators.py:42

Methodargs_format

src/fara/qwen_helpers/base_tool.py:143

Functionattempt_parse_json

(json_str: str)

webeval/src/webeval/utils.py:46

Functionbuild_evaluation_prompt

Constructs the user-facing prompt for the evaluator. This prompt instructs the LLM to verify whether the candidate_answer matches the gol

webeval/src/webeval/evaluators.py:114

(self, params: Union[str, dict], **kwargs)

src/fara/_prompts.py:129

The interface for calling tools. Each tool needs to implement this function, which is the workflow of the tool. Args: pa

src/fara/qwen_helpers/base_tool.py:85

Methodcapabilities

webeval/src/webeval/oai_clients/wrapper.py:206

Methodcapabilities

webeval/src/webeval/oai_clients/wrapper.py:475

Methodcheck_exclusivity

src/fara/qwen_helpers/schema.py:86

Classify a task without a DataPoint / RunContext.

webeval/src/webeval/rubric_agent/critical_point_classifier.py:328

Classify a task without a DataPoint / RunContext.

webeval/src/webeval/rubric_agent/task_classification.py:169

Methodclear_scores_recursive

webeval/src/webeval/rubric_agent/mm_rubric_agent.py:1335

webeval/src/webeval/rubric_agent/base.py:51

webeval/src/webeval/oai_clients/graceful_client.py:176

webeval/src/webeval/oai_clients/wrapper.py:152

webeval/src/webeval/oai_clients/wrapper.py:322

webeval/src/webeval/oai_clients/wrapper.py:387

Per-trajectory teardown. Closes self-built LLM clients (set via ``o4mini_client_config`` / ``gpt5_client_config``); caller-supplied c

webeval/src/webeval/rubric_agent/mm_rubric_agent.py:993

webeval/tests/test_oai_clients.py:161

Functioncompute_aggregate_metrics

Compute aggregate metrics (e.g. average score). Must produce a dictionary. Computes: - mean_score: average over all - max_score:

webeval/src/webeval/evaluators.py:457

Methodcompute_aggregate_metrics

(self, results: List[Any])

webeval/src/webeval/benchmarks/om2w/om2w.py:162

Methodcompute_aggregate_metrics

(self, scores: List[Any])

webeval/src/webeval/benchmarks/webtailbench/webtailbench.py:456

Methodcompute_aggregate_metrics

Compute aggregate metrics for WebVoyager using the shared utility function, and also compute accuracy by web_name.

webeval/src/webeval/benchmarks/webvoyager/webvoyager.py:252

Get the browser context from browser manager.

src/fara/fara_agent.py:150

Get the browser context.

src/fara/browser/browser_bb.py:311

Methodconvert_client_type

(self, new_client_type: Type[ChatCompletionClient])

webeval/src/webeval/oai_clients/graceful_client.py:72

← previous next →401–500 of 650, ranked by callers