constants/models.py

MODEL_MAP = {
    "mixtral-8x7b": "mistralai/Mixtral-8x7B-Instruct-v0.1",  # [Recommended]
    "nous-mixtral-8x7b": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO",
    "mistral-7b": "mistralai/Mistral-7B-Instruct-v0.3",
    "yi-1.5-34b": "01-ai/Yi-1.5-34B-Chat",
    "gemma-7b": "google/gemma-1.1-7b-it",
    "openchat-3.5": "openchat/openchat-3.5-0106",
    "command-r-plus": "CohereForAI/c4ai-command-r-plus-4bit",
    "llama3-70b": "meta-llama/Meta-Llama-3-70B-Instruct",
    "zephyr-141b": "HuggingFaceH4/zephyr-orpo-141b-A35b-v0.1",
    "default": "NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO",
    "Qwen2-72B": "Qwen/Qwen2-72B-Instruct",
    "Qwen2-7B-Instruct": "Qwen/Qwen2-7B-Instruct",
    "Qwen2-1.5B-Instruct" : "Qwen/Qwen2-1.5B-Instruct",
    "llama3-8b" : "meta-llama/Meta-Llama-3-8B-Instruct"
}

AVAILABLE_MODELS = list(MODEL_MAP.keys())

# PRO_MODELS = ["command-r-plus", "llama3-70b", "zephyr-141b", "Qwen2-72B"]
PRO_MODELS = []


STOP_SEQUENCES_MAP = {
    # https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1/blob/main/tokenizer_config.json#L33
    "mixtral-8x7b": "</s>",
    # https://huggingface.co/NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO/blob/main/tokenizer_config.json#L50
    "nous-mixtral-8x7b": "<|im_end|>",
    # https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2/blob/main/tokenizer_config.json#L33
    "mistral-7b": "</s>",
    # https://huggingface.co/01-ai/Yi-1.5-34B-Chat/blob/main/tokenizer_config.json#L42
    "yi-1.5-34b": "<|im_end|>",
    # https://huggingface.co/google/gemma-1.1-7b-it/blob/main/tokenizer_config.json#L1509
    "gemma-7b": "<eos>",
    "openchat-3.5": "<|end_of_turn|>",
    # https://huggingface.co/CohereForAI/c4ai-command-r-plus-4bit/blob/main/tokenizer_config.json#L305
    "command-r-plus": "<|END_OF_TURN_TOKEN|>",
    # https://huggingface.co/Qwen/Qwen2-72B-Instruct/blob/main/tokenizer_config.json#L30
    "Qwen2-72B": "<|im_end|>",
    "Qwen2-7B-Instruct": "<|im_end|>",
    "Qwen2-1.5B-Instruct": "<|im_end|>", 
    
    "llama3-8b" : "<|eot_id|>",
    "llama3-70b" : "<|eot_id|>"
}

TOKEN_LIMIT_MAP = {
    "mixtral-8x7b": 32768,
    "nous-mixtral-8x7b": 32768,
    "mistral-7b": 32768,
    "yi-1.5-34b": 4096,
    "gemma-7b": 8192,
    "openchat-3.5": 8192,
    "command-r-plus": 32768,
    "llama3-70b": 8192,
    "zephyr-141b": 2048,
    "gpt-3.5-turbo": 8192,
    "Qwen2-72B": 32768,
    "Qwen2-7B-Instruct": 32768,
    "Qwen2-1.5B-Instruct": 32768,
    "llama3-8b": 8192,
}

TOKEN_RESERVED = 20


# https://platform.openai.com/docs/api-reference/models/list
AVAILABLE_MODELS_DICTS = [
    {
        "id": "mixtral-8x7b",
        "description": "[mistralai/Mixtral-8x7B-Instruct-v0.1]: https://huggingface.co/mistralai/Mixtral-8x7B-Instruct-v0.1",
        "object": "model",
        "created": 1700000000,
        "owned_by": "mistralai",
    },
    {
        "id": "nous-mixtral-8x7b",
        "description": "[NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO]: https://huggingface.co/NousResearch/Nous-Hermes-2-Mixtral-8x7B-DPO",
        "object": "model",
        "created": 1700000000,
        "owned_by": "NousResearch",
    },
    {
        "id": "mistral-7b",
        "description": "[mistralai/Mistral-7B-Instruct-v0.2]: https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.2",
        "object": "model",
        "created": 1700000000,
        "owned_by": "mistralai",
    },
    {
        "id": "yi-1.5-34b",
        "description": "[01-ai/Yi-1.5-34B-Chat]: https://huggingface.co/01-ai/Yi-1.5-34B-Chat",
        "object": "model",
        "created": 1700000000,
        "owned_by": "01-ai",
    },
    {
        "id": "gemma-7b",
        "description": "[google/gemma-1.1-7b-it]: https://huggingface.co/google/gemma-1.1-7b-it",
        "object": "model",
        "created": 1700000000,
        "owned_by": "Google",
    },
    {
        "id": "openchat-3.5",
        "description": "[openchat/openchat-3.5-0106]: https://huggingface.co/openchat/openchat-3.5-0106",
        "object": "model",
        "created": 1700000000,
        "owned_by": "openchat"
    },
    {
        "id": "command-r-plus",
        "description": "[CohereForAI/c4ai-command-r-plus]: https://huggingface.co/CohereForAI/c4ai-command-r-plus",
        "object": "model",
        "created": 1700000000,
        "owned_by": "CohereForAI"
    },
    {
        "id": "llama3-70b",
        "description": "[meta-llama/Meta-Llama-3-70B]: https://huggingface.co/meta-llama/Meta-Llama-3-70B",
        "object": "model",
        "created": 1700000000,
        "owned_by": "Meta"
    },
    {
        "id": "zephyr-141b",
        "description": "[HuggingFaceH4/zephyr-orpo-141b-A35b-v0.1]: https://huggingface.co/HuggingFaceH4/zephyr-orpo-141b-A35b-v0.1",
        "object": "model",
        "created": 1700000000,
        "owned_by": "Huggingface"
    },
    {
        "id": "gpt-3.5-turbo",
        "description": "[openai/gpt-3.5-turbo]: https://platform.openai.com/docs/models/gpt-3-5-turbo",
        "object": "model",
        "created": 1700000000,
        "owned_by": "OpenAI"
    },
    {
        "id": "Qwen2-72B",
        "description": "[Qwen/Qwen2-72B]: https://huggingface.co/Qwen/Qwen2-72B",
        "object": "model",
        "created": 1700000000,
        "owned_by": "Qwen"
    }
]