File size: 9,477 Bytes
c02bdcd
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
from typing import List, Optional, Union

from tqdm import tqdm
from transformers import PreTrainedTokenizer, PreTrainedTokenizerFast
from vllm.utils import Counter

from .configs import EngineArgs
from .llm_engine import LLMEngine
from .output import RequestOutput
from .sampling_params import SamplingParams


class LLM:
    """An LLM for generating texts from given prompts and sampling parameters.

    This class includes a tokenizer, a language model (possibly distributed
    across multiple GPUs), and GPU memory space allocated for intermediate
    states (aka KV cache). Given a batch of prompts and sampling parameters,
    this class generates texts from the model, using an intelligent batching
    mechanism and efficient memory management.

    NOTE: This class is intended to be used for offline inference. For online
    serving, use the `AsyncLLMEngine` class instead.
    NOTE: For the comprehensive list of arguments, see `EngineArgs`.

    Args:
        model: The name or path of a HuggingFace Transformers model.
        tokenizer: The name or path of a HuggingFace Transformers tokenizer.
        tokenizer_mode: The tokenizer mode. "auto" will use the fast tokenizer
            if available, and "slow" will always use the slow tokenizer.
        trust_remote_code: Trust remote code (e.g., from HuggingFace) when
            downloading the model and tokenizer.
        tensor_parallel_size: The number of GPUs to use for distributed
            execution with tensor parallelism.
        dtype: The data type for the model weights and activations. Currently,
            we support `float32`, `float16`, and `bfloat16`. If `auto`, we use
            the `torch_dtype` attribute specified in the model config file.
            However, if the `torch_dtype` in the config is `float32`, we will
            use `float16` instead.
        quantization: The method used to quantize the model weights. Currently,
            we support "awq", "gptq" and "squeezellm". If None, we first check
            the `quantization_config` attribute in the model config file. If
            that is None, we assume the model weights are not quantized and use
            `dtype` to determine the data type of the weights.
        revision: The specific model version to use. It can be a branch name,
            a tag name, or a commit id.
        tokenizer_revision: The specific tokenizer version to use. It can be a
            branch name, a tag name, or a commit id.
        seed: The seed to initialize the random number generator for sampling.
        gpu_memory_utilization: The ratio (between 0 and 1) of GPU memory to
            reserve for the model weights, activations, and KV cache. Higher
            values will increase the KV cache size and thus improve the model's
            throughput. However, if the value is too high, it may cause out-of-
            memory (OOM) errors.
        swap_space: The size (GiB) of CPU memory per GPU to use as swap space.
            This can be used for temporarily storing the states of the requests
            when their `best_of` sampling parameters are larger than 1. If all
            requests will have `best_of=1`, you can safely set this to 0.
            Otherwise, too small values may cause out-of-memory (OOM) errors.
        enforce_eager: Whether to enforce eager execution. If True, we will
            disable CUDA graph and always execute the model in eager mode.
            If False, we will use CUDA graph and eager execution in hybrid.
        max_context_len_to_capture: Maximum context len covered by CUDA graphs.
            When a sequence has context length larger than this, we fall back
            to eager mode.
    """

    def __init__(
        self,
        model: str,
        tokenizer: Optional[str] = None,
        tokenizer_mode: str = "auto",
        trust_remote_code: bool = False,
        tensor_parallel_size: int = 1,
        dtype: str = "auto",
        quantization: Optional[str] = None,
        revision: Optional[str] = None,
        tokenizer_revision: Optional[str] = None,
        seed: int = 0,
        gpu_memory_utilization: float = 0.9,
        swap_space: int = 4,
        enforce_eager: bool = False,
        max_context_len_to_capture: int = 8192,
        post_model_path: str = None,
        num_audio_tokens: int = 0,
        num_text_tokens: int = 0,
        **kwargs,
    ) -> None:
        if "disable_log_stats" not in kwargs:
            kwargs["disable_log_stats"] = True
        engine_args = EngineArgs(
            model=model,
            tokenizer=tokenizer,
            tokenizer_mode=tokenizer_mode,
            trust_remote_code=trust_remote_code,
            tensor_parallel_size=tensor_parallel_size,
            dtype=dtype,
            quantization=quantization,
            revision=revision,
            tokenizer_revision=tokenizer_revision,
            seed=seed,
            gpu_memory_utilization=gpu_memory_utilization,
            swap_space=swap_space,
            enforce_eager=enforce_eager,
            max_context_len_to_capture=max_context_len_to_capture,
            num_audio_tokens=num_audio_tokens,
            num_text_tokens=num_text_tokens,
            **kwargs,
        )
        self.llm_engine = LLMEngine.from_engine_args(engine_args, post_model_path)
        self.request_counter = Counter()

    def get_tokenizer(self) -> Union[PreTrainedTokenizer, PreTrainedTokenizerFast]:
        return self.llm_engine.tokenizer

    def set_tokenizer(
        self,
        tokenizer: Union[PreTrainedTokenizer, PreTrainedTokenizerFast],
    ) -> None:
        self.llm_engine.tokenizer = tokenizer

    def generate(
        self,
        prompts: Optional[Union[str, List[str]]] = None,
        sampling_params: Optional[SamplingParams] = None,
        prompt_token_ids: Optional[List[List[int]]] = None,
        use_tqdm: bool = True,
    ) -> List[RequestOutput]:
        """Generates the completions for the input prompts.

        NOTE: This class automatically batches the given prompts, considering
        the memory constraint. For the best performance, put all of your prompts
        into a single list and pass it to this method.

        Args:
            prompts: A list of prompts to generate completions for.
            sampling_params: The sampling parameters for text generation. If
                None, we use the default sampling parameters.
            prompt_token_ids: A list of token IDs for the prompts. If None, we
                use the tokenizer to convert the prompts to token IDs.
            use_tqdm: Whether to use tqdm to display the progress bar.

        Returns:
            A list of `RequestOutput` objects containing the generated
            completions in the same order as the input prompts.
        """
        if prompts is None and prompt_token_ids is None:
            raise ValueError("Either prompts or prompt_token_ids must be " "provided.")
        if isinstance(prompts, str):
            # Convert a single prompt to a list.
            prompts = [prompts]
        if (
            prompts is not None
            and prompt_token_ids is not None
            and len(prompts) != len(prompt_token_ids)
        ):
            raise ValueError(
                "The lengths of prompts and prompt_token_ids " "must be the same."
            )
        if sampling_params is None:
            # Use default sampling params.
            sampling_params = SamplingParams()

        # Add requests to the engine.
        num_requests = len(prompts) if prompts is not None else len(prompt_token_ids)
        for i in range(num_requests):
            prompt = prompts[i] if prompts is not None else None
            token_ids = None if prompt_token_ids is None else prompt_token_ids[i]
            self._add_request(prompt, sampling_params, token_ids)

        rtns = self._run_engine(use_tqdm)
        for i, rtn in enumerate(rtns):
            token_ids = rtn.outputs[0].token_ids
            for j, token_id in enumerate(token_ids):
                if len(token_id) == 1:
                    token_ids[j] = token_id[0]
                else:
                    token_ids[j] = list(token_id)

        return rtns

    def _add_request(
        self,
        prompt: Optional[str],
        sampling_params: SamplingParams,
        prompt_token_ids: Optional[List[int]],
    ) -> None:
        request_id = str(next(self.request_counter))
        self.llm_engine.add_request(
            request_id, prompt, sampling_params, prompt_token_ids
        )

    def _run_engine(self, use_tqdm: bool) -> List[RequestOutput]:
        # Initialize tqdm.
        if use_tqdm:
            num_requests = self.llm_engine.get_num_unfinished_requests()
            pbar = tqdm(total=num_requests, desc="Processed prompts")
        # Run the engine.
        outputs: List[RequestOutput] = []
        while self.llm_engine.has_unfinished_requests():
            step_outputs = self.llm_engine.step()
            for output in step_outputs:
                if output.finished:
                    outputs.append(output)
                    if use_tqdm:
                        pbar.update(1)
        if use_tqdm:
            pbar.close()
        # Sort the outputs by request ID.
        # This is necessary because some requests may be finished earlier than
        # its previous requests.
        outputs = sorted(outputs, key=lambda x: int(x.request_id))
        return outputs