File size: 5,030 Bytes
5e0ae28 9d5bde0 5e0ae28 9d5bde0 5e0ae28 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 |
"""
Application constants and enums
"""
from enum import Enum
class ReasoningMode(Enum):
"""
🧠 RESEARCH-ALIGNED REASONING METHODOLOGIES
Based on academic papers and proven techniques
"""
TREE_OF_THOUGHTS = "Tree of Thoughts (ToT)"
CHAIN_OF_THOUGHT = "Chain of Thought (CoT)"
SELF_CONSISTENCY = "Self-Consistency Sampling"
REFLEXION = "Reflexion + Self-Correction"
DEBATE = "Multi-Agent Debate"
ANALOGICAL = "Analogical Reasoning"
SIMPLE = "Simple (Direct Response)"
def __str__(self) -> str:
return self.value
@classmethod
def get_description(cls, mode: 'ReasoningMode') -> str:
"""Get detailed description of reasoning mode"""
descriptions = {
cls.TREE_OF_THOUGHTS: "Explores multiple reasoning paths systematically before converging on the best solution.",
cls.CHAIN_OF_THOUGHT: "Breaks down complex problems into clear, logical steps with explicit reasoning.",
cls.SELF_CONSISTENCY: "Generates multiple independent solutions and identifies the most consistent answer.",
cls.REFLEXION: "Solves problems, critiques reasoning, and refines solutions iteratively.",
cls.DEBATE: "Presents multiple perspectives and synthesizes the strongest arguments.",
cls.ANALOGICAL: "Finds similar problems and applies their solutions to the current problem.",
cls.SIMPLE: "Direct response without structured reasoning - fastest and most natural."
}
return descriptions.get(mode, "Advanced reasoning mode")
class ModelConfig(Enum):
"""
🤖 AVAILABLE GROQ MODELS
"""
# === Meta / Llama Models ===
LLAMA_70B = ("llama-3.3-70b-versatile", 70, 8000, "Best overall performance", "Meta")
LLAMA_70B_V31 = ("llama-3.1-70b-versatile", 70, 8000, "Stable production model", "Meta")
LLAMA_3_1_8B_INSTANT = ("llama-3.1-8b-instant", 8, 131072, "Ultra-fast responses", "Meta")
LLAMA_4_MAVERICK_17B = ("meta-llama/llama-4-maverick-17b-128k", 17, 131072, "Llama 4 experimental", "Meta")
LLAMA_4_SCOUT_17B = ("meta-llama/llama-4-scout-17b-16e-instruct", 17, 16384, "Llama 4 scout model", "Meta")
LLAMA_GUARD_4_12B = ("meta-llama/llama-guard-4-12b", 12, 8192, "Content moderation", "Meta")
LLAMA_PROMPT_GUARD_2_22M = ("meta-llama/llama-prompt-guard-2-22m", 0, 8192, "Prompt injection detection (22M)", "Meta")
LLAMA_PROMPT_GUARD_2_86M = ("meta-llama/llama-prompt-guard-2-86m", 0, 8192, "Prompt injection detection (86M)", "Meta")
# === DeepSeek Models ===
DEEPSEEK_70B = ("deepseek-r1-distill-llama-70b", 70, 8000, "Optimized reasoning", "DeepSeek")
# === Mixtral Models ===
MIXTRAL_8X7B = ("mixtral-8x7b-32768", 47, 32768, "Long context specialist", "Mixtral")
# === Google Gemma Models ===
GEMMA_9B = ("gemma2-9b-it", 9, 8192, "Fast and efficient", "Google")
# === Moonshot AI Models ===
KIMI_K2_INSTRUCT_DEPRECATED = ("moonshotai/kimi-k2-instruct", 0, 200000, "Ultra-long context (Deprecated)", "Moonshot")
KIMI_K2_INSTRUCT_0905 = ("moonshotai/kimi-k2-instruct-0905", 0, 200000, "Ultra-long context specialist", "Moonshot")
# === OpenAI Models ===
GPT_OSS_120B = ("openai/gpt-oss-120b", 120, 8192, "Large open source model", "OpenAI")
GPT_OSS_20B = ("openai/gpt-oss-20b", 20, 8192, "Medium open source model", "OpenAI")
# === Qwen Models ===
QWEN3_32B = ("qwen/qwen3-32b", 32, 32768, "Qwen 3 multilingual", "Qwen")
# === Groq Compound Models ===
GROQ_COMPOUND = ("groq/compound", 0, 8192, "Groq optimized compound", "Groq")
GROQ_COMPOUND_MINI = ("groq/compound-mini", 0, 8192, "Groq mini compound", "Groq")
def __init__(self, model_id: str, params_b: int, max_context: int, description: str, provider: str):
self.model_id = model_id
self.params_b = params_b
self.max_context = max_context
self.description = description
self.provider = provider
def __str__(self) -> str:
return f"{self.provider} - {self.model_id}"
@classmethod
def get_by_id(cls, model_id: str) -> 'ModelConfig':
"""Get model config by ID"""
for model in cls:
if model.model_id == model_id:
return model
raise ValueError(f"Model {model_id} not found")
@classmethod
def get_by_provider(cls, provider: str) -> list:
"""Get all models by provider"""
return [model for model in cls if model.provider == provider]
@classmethod
def get_recommended(cls) -> 'ModelConfig':
"""Get recommended default model"""
return cls.LLAMA_70B
@property
def is_long_context(self) -> bool:
"""Check if model supports long context (>16k)"""
return self.max_context > 16384
@property
def is_fast(self) -> bool:
"""Check if model is optimized for speed"""
return "instant" in self.model_id.lower() or self.params_b < 10
|