KoboldAI-Client/modeling/inference_models/basic_hf/class.py

from __future__ import annotations

import os
import shutil
import time
from typing import List, Optional, Union

import torch
import transformers
from transformers import LogitsProcessorList
from transformers.models.auto.modeling_auto import AutoModelForCausalLM

import utils
from logger import logger
from modeling import warpers
from modeling.inference_model import (
    GenerationResult,
    GenerationSettings,
    use_core_manipulations,
)
from modeling.inference_models.hf import HFInferenceModel

model_backend_name = "Basic Huggingface"
model_backend_type = "Huggingface"


class model_backend(HFInferenceModel):
    def __init__(self) -> None:
        super().__init__()
        self.model_name = "Basic Huggingface"

        # TODO: These feel weird to be in HFInferenceModel, maybe we could implement
        # them in subclasses?
        self.hf_torch = True
        self.nobreakmodel = True

    def _load(self, save_model: bool, initial_load: bool) -> None:
        utils.koboldai_vars.allowsp = False

        if self.model_name == "NeoCustom":
            self.model_name = os.path.basename(os.path.normpath(self.path))
        utils.koboldai_vars.model = self.model_name

        # If we specify a model and it's in the root directory, we need to move
        # it to the models directory (legacy folder structure to new)
        if self.get_local_model_path(legacy=True):
            shutil.move(
                self.get_local_model_path(legacy=True, ignore_existance=True),
                self.get_local_model_path(ignore_existance=True),
            )

        self.init_model_config()

        self.model = AutoModelForCausalLM.from_pretrained(
            self.get_local_model_path(), low_cpu_mem_usage=True
        )

        if self.usegpu:
            self.model = self.model.to("cuda")

        self.tokenizer = self._get_tokenizer(self.get_local_model_path())

        # Patch sampler to use KAI samplers
        def _patched_get_logits_processor(*args, **kwargs) -> LogitsProcessorList:
            processors = _patched_get_logits_processor.original(*args, **kwargs)
            return processors

        use_core_manipulations.get_logits_processor = _patched_get_logits_processor
        _patched_get_logits_processor.original = (
            transformers.GenerationMixin._get_logits_processor
        )

        class KoboldLogitsWarperList(LogitsProcessorList):
            def __call__(
                _self,  # Unused
                input_ids: torch.LongTensor,
                scores: torch.FloatTensor,
                *args,
                **kwargs,
            ):
                scores = self._apply_warpers(scores=scores, input_ids=input_ids)

                for processor in self.logits_processors:
                    scores = processor(self, scores=scores, input_ids=input_ids)
                    assert (
                        scores is not None
                    ), f"Scores are None; processor '{processor}' is to blame"
                return scores

        def new_sample(self, *args, **kwargs):
            assert kwargs.pop("logits_warper", None) is not None
            kwargs["logits_warper"] = lambda: KoboldLogitsWarperList()

            if utils.koboldai_vars.newlinemode in ["s", "ns"]:
                kwargs["eos_token_id"] = -1
                kwargs.setdefault("pad_token_id", 2)

            return new_sample.old_sample(self, *args, **kwargs)

        new_sample.old_sample = transformers.GenerationMixin.sample
        use_core_manipulations.sample = new_sample

        self.model.kai_model = self
        utils.koboldai_vars.modeldim = self.model.get_input_embeddings().embedding_dim

    def _apply_warpers(
        self, scores: torch.Tensor, input_ids: torch.Tensor
    ) -> torch.Tensor:
        warpers.update_settings()

        for sid in utils.koboldai_vars.sampler_order:
            warper = warpers.Warper.from_id(sid)

            if not warper.value_is_valid():
                continue

            if warper == warpers.RepetitionPenalty:
                # Rep pen needs more data than other samplers
                scores = warper.torch(scores, input_ids=input_ids)
            else:
                scores = warper.torch(scores)

            assert scores is not None, f"Scores are None; warper '{warper}' is to blame"
        return scores

    def _raw_generate(
        self,
        prompt_tokens: Union[List[int], torch.Tensor],
        max_new: int,
        gen_settings: GenerationSettings,
        single_line: bool = False,
        batch_count: int = 1,
        seed: Optional[int] = None,
        **kwargs,
    ) -> GenerationResult:
        if not isinstance(prompt_tokens, torch.Tensor):
            gen_in = torch.tensor(prompt_tokens, dtype=torch.long)[None]
        else:
            gen_in = prompt_tokens
        if not self.usegpu:
            gen_in = gen_in.to("cpu")
        else:
            device = self.get_auxilary_device()
            gen_in = gen_in.to(device)

        additional_bad_words_ids = [self.tokenizer.encode("\n")] if single_line else []

        if seed is not None:
            torch.manual_seed(seed)

        with torch.no_grad():
            start_time = time.time()
            genout = self.model.generate(
                gen_in,
                do_sample=True,
                max_length=min(
                    len(prompt_tokens) + max_new, utils.koboldai_vars.max_length
                ),
                repetition_penalty=1.0,
                bad_words_ids=self.badwordsids + additional_bad_words_ids,
                use_cache=True,
                num_return_sequences=batch_count,
            )
        logger.debug(
            "torch_raw_generate: run generator {}s".format(time.time() - start_time)
        )

        return GenerationResult(
            self,
            out_batches=genout,
            prompt=prompt_tokens,
            is_whole_generation=False,
            output_includes_prompt=True,
        )