retab 0.0.42__py3-none-any.whl → 0.0.44__py3-none-any.whl

retab/_utils/_model_cards/openai.yaml

@@ -1,301 +0,0 @@
-
-# Reasoning models
-# o1 family 
-- model: "o1"
-  pricing:
-    text:
-      prompt: 15.00
-      cached_discount: 0.5
-      completion: 60.00
-    audio: null
-  capabilities:
-    modalities: ["text", "image"]
-    endpoints: ["chat_completions", "responses", "assistants", "batch"]
-    features: ["streaming", "function_calling", "structured_outputs"]
-  temperature_support: false
-  reasoning_effort_support: true
-
-- model: "o1-2024-12-17"
-  inherits: "o1"
-
-# o3 family
-- model: "o3"
-  pricing:
-    text:
-      prompt: 2.0
-      cached_discount: 0.25
-      completion: 8.0
-    audio: null
-    ft_price_hike: 1.5
-  capabilities:
-    modalities: ["text", "image"]
-    endpoints: ["chat_completions", "responses", "assistants", "batch"]
-    features: ["streaming", "function_calling", "structured_outputs", "schema_generation"]
-  temperature_support: false
-  reasoning_effort_support: true
-  permissions:
-    show_in_free_picker: false
-    show_in_paid_picker: true
-
-- model: "o3-2025-04-16"
-  inherits: "o3"
-  permissions:
-    show_in_free_picker: false
-    show_in_paid_picker: false
-
-# o4-mini family
-- model: "o4-mini"
-  pricing:
-    text:
-      prompt: 1.10
-      cached_discount: 0.25
-      completion: 4.40
-    audio: null
-    ft_price_hike: 1.5
-  capabilities:
-    modalities: ["text", "image"]
-    endpoints: ["chat_completions", "responses", "assistants", "batch"]
-    features: ["streaming", "function_calling", "structured_outputs", "schema_generation"]
-  temperature_support: false
-  reasoning_effort_support: true
-  permissions:
-    show_in_free_picker: false
-    show_in_paid_picker: true
-
-- model: "o4-mini-2025-04-16"
-  inherits: "o4-mini"
-  permissions:
-    show_in_free_picker: false
-    show_in_paid_picker: false
-
-# Chat models
-# gpt-4.1 family
-- model: "gpt-4.1"
-  pricing:
-    text:
-      prompt: 2.00
-      cached_discount: 0.25
-      completion: 8.00
-    audio: null
-    ft_price_hike: 1.5
-  capabilities:
-    modalities: ["text", "image"]
-    endpoints: ["chat_completions", "responses", "assistants", "batch"]
-    features: ["streaming", "function_calling", "structured_outputs", "schema_generation"]
-  temperature_support: true
-  reasoning_effort_support: false
-  permissions:
-    show_in_free_picker: true
-    show_in_paid_picker: true
-
-- model: "gpt-4.1-2025-04-14"
-  inherits: "gpt-4.1"
-  permissions:
-    show_in_free_picker: false
-    show_in_paid_picker: false
-
-- model: "gpt-4.1-mini"
-  pricing:
-    text:
-      prompt: 0.40
-      cached_discount: 0.25
-      completion: 1.60
-    audio: null
-    ft_price_hike: 1.5
-  capabilities:
-    modalities: ["text", "image"]
-    endpoints: ["chat_completions", "responses", "assistants", "batch"]
-    features: ["streaming", "function_calling", "structured_outputs", "schema_generation"]
-  temperature_support: true
-  reasoning_effort_support: false
-  permissions:
-    show_in_free_picker: true
-    show_in_paid_picker: true
-
-- model: "gpt-4.1-mini-2025-04-14"
-  inherits: "gpt-4.1-mini"
-  permissions:
-    show_in_free_picker: false
-    show_in_paid_picker: false
-
-- model: "gpt-4.1-nano"
-  pricing:
-    text:
-      prompt: 0.10
-      cached_discount: 0.25
-      completion: 0.40
-    audio: null
-    ft_price_hike: 1.5
-  capabilities:
-    modalities: ["text", "image"]
-    endpoints: ["chat_completions", "responses", "assistants", "batch"]
-    features: ["streaming", "function_calling", "structured_outputs", "schema_generation"]
-  temperature_support: true
-  reasoning_effort_support: false
-  permissions:
-    show_in_free_picker: true
-    show_in_paid_picker: true
-
-- model: "gpt-4.1-nano-2025-04-14"
-  inherits: "gpt-4.1-nano"
-  permissions:
-    show_in_free_picker: false
-    show_in_paid_picker: false
-
-# gpt-4o family
-- model: "chatgpt-4o-latest"
-  pricing:
-    text:
-      prompt: 2.50
-      cached_discount: 0.5
-      completion: 10.00
-    audio: null
-    ft_price_hike: 1.5
-  capabilities:
-    modalities: ["text", "image"]
-    endpoints: ["chat_completions", "responses", "assistants", "batch", "fine_tuning"]
-    features: ["streaming", "function_calling", "structured_outputs", "fine_tuning", "distillation", "predicted_outputs"]
-  temperature_support: true
-
-- model: "gpt-4o"
-  inherits: "chatgpt-4o-latest"
-
-- model: "gpt-4o-2024-08-06"
-  inherits: "chatgpt-4o-latest"
-
-- model: "gpt-4o-2024-11-20"
-  inherits: "chatgpt-4o-latest"
-
-- model: "gpt-4o-2024-05-13"
-  pricing:
-    text:
-      prompt: 5.00
-      cached_discount: 0.5
-      completion: 15.00
-    audio: null
-    ft_price_hike: 1.5
-  capabilities:
-    modalities: ["text", "image"]
-    endpoints: ["chat_completions", "responses", "assistants", "batch", "fine_tuning"]
-    features: ["streaming", "function_calling", "structured_outputs", "fine_tuning", "distillation", "predicted_outputs"]
-  temperature_support: true
-
-# gpt-4o-audio family
-- model: "gpt-4o-audio-preview-2024-12-17"
-  pricing:
-    text:
-      prompt: 2.50
-      cached_discount: 0.5
-      completion: 10.00
-    audio:
-      prompt: 40.00
-      cached_discount: 0.2
-      completion: 80.00
-  capabilities:
-    modalities: ["text", "audio"]
-    endpoints: ["chat_completions"]
-    features: ["streaming", "function_calling"]
-  temperature_support: true
-
-- model: "gpt-4o-audio-preview-2024-10-01"
-  pricing:
-    text:
-      prompt: 2.50
-      cached_discount: 0.5
-      completion: 10.00
-    audio:
-      prompt: 100.00
-      cached_discount: 0.2
-      completion: 200.00
-  capabilities:
-    modalities: ["text", "audio"]
-    endpoints: ["chat_completions"]
-    features: ["streaming", "function_calling"]
-  temperature_support: true
-
-- model: "gpt-4o-realtime-preview-2024-12-17"
-  pricing:
-    text:
-      prompt: 5.00
-      cached_discount: 0.5
-      completion: 20.00
-    audio:
-      prompt: 40.00
-      cached_discount: 0.2
-      completion: 80.00
-  capabilities:
-    modalities: ["text", "audio"]
-    endpoints: ["chat_completions"]
-    features: ["streaming", "function_calling"]
-  temperature_support: true
-
-- model: "gpt-4o-realtime-preview-2024-10-01"
-  pricing:
-    text:
-      prompt: 5.00
-      cached_discount: 0.5
-      completion: 20.00
-    audio:
-      prompt: 100.00
-      cached_discount: 0.2
-      completion: 200.00
-  capabilities:
-    modalities: ["text", "audio"]
-    endpoints: ["chat_completions"]
-    features: ["streaming", "function_calling"]
-  temperature_support: true
-
-# gpt-4o-mini family
-- model: "gpt-4o-mini"
-  pricing:
-    text:
-      prompt: 0.15
-      cached_discount: 0.5
-      completion: 0.60
-    audio: null
-    ft_price_hike: 2.0
-  capabilities:
-    modalities: ["text", "image"]
-    endpoints: ["chat_completions", "responses", "assistants", "batch", "fine_tuning"]
-    features: ["streaming", "function_calling", "structured_outputs", "fine_tuning"]
-  temperature_support: true
-
-- model: "gpt-4o-mini-2024-07-18"
-  inherits: "gpt-4o-mini"
-
-# gpt-4o-mini-audio family
-- model: "gpt-4o-mini-audio-preview-2024-12-17"
-  pricing:
-    text:
-      prompt: 0.15
-      cached_discount: 0.5
-      completion: 0.60
-    audio:
-      prompt: 10.00
-      cached_discount: 0.2
-      completion: 20.00
-    ft_price_hike: 2.0
-  capabilities:
-    modalities: ["text", "audio"]
-    endpoints: ["chat_completions"]
-    features: ["streaming", "function_calling"]
-  temperature_support: true
-
-- model: "gpt-4o-mini-realtime-preview-2024-12-17"
-  pricing:
-    text:
-      prompt: 0.60
-      cached_discount: 0.5
-      completion: 2.40
-    audio:
-      prompt: 10.00
-      cached_discount: 0.2
-      completion: 20.00
-    ft_price_hike: 2.0
-  capabilities:
-    modalities: ["text", "audio"]
-    endpoints: ["chat_completions"]
-    features: ["streaming", "function_calling"]
-  temperature_support: true
-
-

retab/_utils/_model_cards/xai.yaml

@@ -1,28 +0,0 @@
-# grok3-family
-- model: "grok-3"
-  pricing:
-    text:
-      prompt: 3
-      completion: 15
-    audio: null
-  capabilities:
-    modalities: ["text"]
-    endpoints: ["chat_completions"]
-    features: ["streaming", "structured_outputs"]
-  permissions:
-    show_in_free_picker: true
-    show_in_paid_picker: true
-
-- model: "grok-3-mini"
-  pricing:
-    text:
-      prompt: 0.3
-      completion: 0.5
-    audio: null
-  capabilities:
-    modalities: ["text"]
-    endpoints: ["chat_completions"]
-    features: ["streaming", "structured_outputs"]
-  permissions:
-    show_in_free_picker: true
-    show_in_paid_picker: true

retab/_utils/ai_models.py

@@ -1,138 +0,0 @@
-import os
-import yaml
-from typing import get_args
-
-from ..types.ai_models import AIProvider, GeminiModel, OpenAIModel, xAI_Model, RetabModel, PureLLMModel, ModelCard
-
-MODEL_CARDS_DIR = os.path.join(os.path.dirname(__file__), "_model_cards")
-
-def merge_model_cards(base: dict, override: dict) -> dict:
-    result = base.copy()
-    for key, value in override.items():
-        if key == "inherits":
-            continue
-        if isinstance(value, dict) and key in result:
-            result[key] = merge_model_cards(result[key], value)
-        else:
-            result[key] = value
-    return result
-
-def load_model_cards(yaml_file: str) -> list[ModelCard]:
-    raw_cards = yaml.safe_load(open(yaml_file))
-    name_to_card = {c["model"]: c for c in raw_cards if "inherits" not in c}
-
-    final_cards = []
-    for card in raw_cards:
-        if "inherits" in card:
-            parent = name_to_card[card["inherits"]]
-            merged = merge_model_cards(parent, card)
-            final_cards.append(ModelCard(**merged))
-        else:
-            final_cards.append(ModelCard(**card))
-    return final_cards
-
-# Load all model cards
-model_cards = sum([
-    load_model_cards(os.path.join(MODEL_CARDS_DIR, "openai.yaml")),
-    load_model_cards(os.path.join(MODEL_CARDS_DIR, "anthropic.yaml")),
-    load_model_cards(os.path.join(MODEL_CARDS_DIR, "xai.yaml")),
-    load_model_cards(os.path.join(MODEL_CARDS_DIR, "gemini.yaml")),
-    load_model_cards(os.path.join(MODEL_CARDS_DIR, "auto.yaml")),
-], [])
-model_cards_dict = {card.model: card for card in model_cards}
-
-
-# Validate that model cards
-all_model_names = set(model_cards_dict.keys())
-if all_model_names.symmetric_difference(set(get_args(PureLLMModel))):
-    raise ValueError(f"Mismatch between model cards and PureLLMModel type: {all_model_names.symmetric_difference(set(get_args(PureLLMModel)))}")
-
-
-def get_model_from_model_id(model_id: str) -> str:
-    """
-    Get the model name from the model id.
-    """
-    if model_id.startswith("ft:"):
-        parts = model_id.split(":")
-        return parts[1]
-    else:
-        return model_id
-
-
-def get_model_card(model: str) -> ModelCard:
-    """
-    Get the model card for a specific model.
-
-    Args:
-        model: The model name to look up
-
-    Returns:
-        The ModelCard for the specified model
-
-    Raises:
-        ValueError: If no model card is found for the specified model
-    """
-    model_name = get_model_from_model_id(model)
-    if model_name in model_cards_dict:
-        model_card = ModelCard(**model_cards_dict[model_name].model_dump())
-        if model_name != model:
-            # Fine-tuned model -> Change the name
-            model_card.model = model
-            # Remove the fine-tuning feature (if exists)
-            try:
-                model_card.capabilities.features.remove("fine_tuning")
-            except ValueError:
-                pass
-        return model_card
-
-    raise ValueError(f"No model card found for model: {model_name}")
-
-
-def get_provider_for_model(model_id: str) -> AIProvider:
-    """
-    Determine the AI provider associated with the given model identifier.
-    Returns one of: "Anthropic", "xAI", "OpenAI", "Gemini", "Retab" or None if unknown.
-    """
-    model_name = get_model_from_model_id(model_id)
-    # if model_name in get_args(AnthropicModel):
-    #     return "Anthropic"
-    # if model_name in get_args(xAI_Model):
-    #     return "xAI"
-    if model_name in get_args(OpenAIModel):
-        return "OpenAI"
-    if model_name in get_args(GeminiModel):
-        return "Gemini"
-    if model_name in get_args(RetabModel):
-        return "Retab"
-    raise ValueError(f"Unknown model: {model_name}")
-
-
-def assert_valid_model_extraction(model: str) -> None:
-    try: 
-        get_provider_for_model(model)
-    except ValueError:
-        raise ValueError(
-            f"Invalid model for extraction: {model}.\nValid OpenAI models: {get_args(OpenAIModel)}\n"
-            f"Valid xAI models: {get_args(xAI_Model)}\n"
-            f"Valid Gemini models: {get_args(GeminiModel)}"
-        ) from None
-
-
-def assert_valid_model_schema_generation(model: str) -> None:
-    """Assert that the model is either a standard OpenAI model or a valid fine-tuned model.
-
-    Valid formats:
-    - Standard model: Must be in OpenAIModel
-    - Fine-tuned model: Must be {base_model}:{id} where base_model is in OpenAIModel
-
-    Raises:
-        ValueError: If the model format is invalid
-    """
-    if get_model_from_model_id(model) in get_args(OpenAIModel):
-        return
-    else:
-        raise ValueError(
-                f"Invalid model format: {model}. Must be either:\n"
-                f"1. A standard model: {get_args(OpenAIModel)}\n"
-                f"2. A fine-tuned model in format 'base_model:id' where base_model is one of the standard openai models"
-            ) from None